From 099f53cb50e45ef617a9f1d63ceec799e489418b Mon Sep 17 00:00:00 2001 From: Dan Williams Date: Wed, 8 Apr 2009 14:28:37 -0700 Subject: async_tx: rename zero_sum to val 'zero_sum' does not properly describe the operation of generating parity and checking that it validates against an existing buffer. Change the name of the operation to 'val' (for 'validate'). This is in anticipation of the p+q case where it is a requirement to identify the target parity buffers separately from the source buffers, because the target parity buffers will not have corresponding pq coefficients. Reviewed-by: Andre Noll Acked-by: Maciej Sosnowski Signed-off-by: Dan Williams --- include/linux/async_tx.h | 2 +- include/linux/dmaengine.h | 8 ++++---- 2 files changed, 5 insertions(+), 5 deletions(-) (limited to 'include') diff --git a/include/linux/async_tx.h b/include/linux/async_tx.h index 5fc2ef8d97fa..513150d8c25b 100644 --- a/include/linux/async_tx.h +++ b/include/linux/async_tx.h @@ -117,7 +117,7 @@ async_xor(struct page *dest, struct page **src_list, unsigned int offset, dma_async_tx_callback cb_fn, void *cb_fn_param); struct dma_async_tx_descriptor * -async_xor_zero_sum(struct page *dest, struct page **src_list, +async_xor_val(struct page *dest, struct page **src_list, unsigned int offset, int src_cnt, size_t len, u32 *result, enum async_tx_flags flags, struct dma_async_tx_descriptor *depend_tx, diff --git a/include/linux/dmaengine.h b/include/linux/dmaengine.h index 2e2aa3df170c..6768727d00d7 100644 --- a/include/linux/dmaengine.h +++ b/include/linux/dmaengine.h @@ -55,8 +55,8 @@ enum dma_transaction_type { DMA_PQ_XOR, DMA_DUAL_XOR, DMA_PQ_UPDATE, - DMA_ZERO_SUM, - DMA_PQ_ZERO_SUM, + DMA_XOR_VAL, + DMA_PQ_VAL, DMA_MEMSET, DMA_MEMCPY_CRC32C, DMA_INTERRUPT, @@ -214,7 +214,7 @@ struct dma_async_tx_descriptor { * @device_free_chan_resources: release DMA channel's resources * @device_prep_dma_memcpy: prepares a memcpy operation * @device_prep_dma_xor: prepares a xor operation - * @device_prep_dma_zero_sum: prepares a zero_sum operation + * @device_prep_dma_xor_val: prepares a xor validation operation * @device_prep_dma_memset: prepares a memset operation * @device_prep_dma_interrupt: prepares an end of chain interrupt operation * @device_prep_slave_sg: prepares a slave dma operation @@ -243,7 +243,7 @@ struct dma_device { struct dma_async_tx_descriptor *(*device_prep_dma_xor)( struct dma_chan *chan, dma_addr_t dest, dma_addr_t *src, unsigned int src_cnt, size_t len, unsigned long flags); - struct dma_async_tx_descriptor *(*device_prep_dma_zero_sum)( + struct dma_async_tx_descriptor *(*device_prep_dma_xor_val)( struct dma_chan *chan, dma_addr_t *src, unsigned int src_cnt, size_t len, u32 *result, unsigned long flags); struct dma_async_tx_descriptor *(*device_prep_dma_memset)( -- cgit v1.2.3 From 88ba2aa586c874681c072101287e15d40de7e6e2 Mon Sep 17 00:00:00 2001 From: Dan Williams Date: Thu, 9 Apr 2009 16:16:18 -0700 Subject: async_tx: kill ASYNC_TX_DEP_ACK flag In support of inter-channel chaining async_tx utilizes an ack flag to gate whether a dependent operation can be chained to another. While the flag is not set the chain can be considered open for appending. Setting the ack flag closes the chain and flags the descriptor for garbage collection. The ASYNC_TX_DEP_ACK flag essentially means "close the chain after adding this dependency". Since each operation can only have one child the api now implicitly sets the ack flag at dependency submission time. This removes an unnecessary management burden from clients of the api. [ Impact: clean up and enforce one dependency per operation ] Reviewed-by: Andre Noll Acked-by: Maciej Sosnowski Signed-off-by: Dan Williams --- Documentation/crypto/async-tx-api.txt | 9 ++++----- crypto/async_tx/async_memcpy.c | 2 +- crypto/async_tx/async_memset.c | 2 +- crypto/async_tx/async_tx.c | 4 ++-- crypto/async_tx/async_xor.c | 6 ++---- drivers/md/raid5.c | 25 +++++++++++-------------- include/linux/async_tx.h | 4 +--- 7 files changed, 22 insertions(+), 30 deletions(-) (limited to 'include') diff --git a/Documentation/crypto/async-tx-api.txt b/Documentation/crypto/async-tx-api.txt index 4af12180d191..76feda8541dc 100644 --- a/Documentation/crypto/async-tx-api.txt +++ b/Documentation/crypto/async-tx-api.txt @@ -80,8 +80,8 @@ acknowledged by the application before the offload engine driver is allowed to recycle (or free) the descriptor. A descriptor can be acked by one of the following methods: 1/ setting the ASYNC_TX_ACK flag if no child operations are to be submitted -2/ setting the ASYNC_TX_DEP_ACK flag to acknowledge the parent - descriptor of a new operation. +2/ submitting an unacknowledged descriptor as a dependency to another + async_tx call will implicitly set the acknowledged state. 3/ calling async_tx_ack() on the descriptor. 3.4 When does the operation execute? @@ -136,10 +136,9 @@ int run_xor_copy_xor(struct page **xor_srcs, tx = async_xor(xor_dest, xor_srcs, 0, xor_src_cnt, xor_len, ASYNC_TX_XOR_DROP_DST, NULL, NULL, NULL); - tx = async_memcpy(copy_dest, copy_src, 0, 0, copy_len, - ASYNC_TX_DEP_ACK, tx, NULL, NULL); + tx = async_memcpy(copy_dest, copy_src, 0, 0, copy_len, tx, NULL, NULL); tx = async_xor(xor_dest, xor_srcs, 0, xor_src_cnt, xor_len, - ASYNC_TX_XOR_DROP_DST | ASYNC_TX_DEP_ACK | ASYNC_TX_ACK, + ASYNC_TX_XOR_DROP_DST | ASYNC_TX_ACK, tx, complete_xor_copy_xor, NULL); async_tx_issue_pending_all(); diff --git a/crypto/async_tx/async_memcpy.c b/crypto/async_tx/async_memcpy.c index ddccfb01c416..7117ec6f1b74 100644 --- a/crypto/async_tx/async_memcpy.c +++ b/crypto/async_tx/async_memcpy.c @@ -35,7 +35,7 @@ * @src: src page * @offset: offset in pages to start transaction * @len: length in bytes - * @flags: ASYNC_TX_ACK, ASYNC_TX_DEP_ACK, + * @flags: ASYNC_TX_ACK * @depend_tx: memcpy depends on the result of this transaction * @cb_fn: function to call when the memcpy completes * @cb_param: parameter to pass to the callback routine diff --git a/crypto/async_tx/async_memset.c b/crypto/async_tx/async_memset.c index 5b5eb99bb244..b2f133885b7f 100644 --- a/crypto/async_tx/async_memset.c +++ b/crypto/async_tx/async_memset.c @@ -35,7 +35,7 @@ * @val: fill value * @offset: offset in pages to start transaction * @len: length in bytes - * @flags: ASYNC_TX_ACK, ASYNC_TX_DEP_ACK + * @flags: ASYNC_TX_ACK * @depend_tx: memset depends on the result of this transaction * @cb_fn: function to call when the memcpy completes * @cb_param: parameter to pass to the callback routine diff --git a/crypto/async_tx/async_tx.c b/crypto/async_tx/async_tx.c index 06eb6cc09fef..3766bc3d7d89 100644 --- a/crypto/async_tx/async_tx.c +++ b/crypto/async_tx/async_tx.c @@ -223,7 +223,7 @@ async_tx_submit(struct dma_chan *chan, struct dma_async_tx_descriptor *tx, if (flags & ASYNC_TX_ACK) async_tx_ack(tx); - if (depend_tx && (flags & ASYNC_TX_DEP_ACK)) + if (depend_tx) async_tx_ack(depend_tx); } EXPORT_SYMBOL_GPL(async_tx_submit); @@ -231,7 +231,7 @@ EXPORT_SYMBOL_GPL(async_tx_submit); /** * async_trigger_callback - schedules the callback function to be run after * any dependent operations have been completed. - * @flags: ASYNC_TX_ACK, ASYNC_TX_DEP_ACK + * @flags: ASYNC_TX_ACK * @depend_tx: 'callback' requires the completion of this transaction * @cb_fn: function to call after depend_tx completes * @cb_param: parameter to pass to the callback routine diff --git a/crypto/async_tx/async_xor.c b/crypto/async_tx/async_xor.c index e0580b0ea533..3cc5dc763b54 100644 --- a/crypto/async_tx/async_xor.c +++ b/crypto/async_tx/async_xor.c @@ -105,7 +105,6 @@ do_async_xor(struct dma_chan *chan, struct page *dest, struct page **src_list, _cb_param); depend_tx = tx; - flags |= ASYNC_TX_DEP_ACK; if (src_cnt > xor_src_cnt) { /* drop completed sources */ @@ -168,8 +167,7 @@ do_sync_xor(struct page *dest, struct page **src_list, unsigned int offset, * @offset: offset in pages to start transaction * @src_cnt: number of source pages * @len: length in bytes - * @flags: ASYNC_TX_XOR_ZERO_DST, ASYNC_TX_XOR_DROP_DEST, - * ASYNC_TX_ACK, ASYNC_TX_DEP_ACK + * @flags: ASYNC_TX_XOR_ZERO_DST, ASYNC_TX_XOR_DROP_DEST, ASYNC_TX_ACK * @depend_tx: xor depends on the result of this transaction. * @cb_fn: function to call when the xor completes * @cb_param: parameter to pass to the callback routine @@ -230,7 +228,7 @@ static int page_is_zero(struct page *p, unsigned int offset, size_t len) * @src_cnt: number of source pages * @len: length in bytes * @result: 0 if sum == 0 else non-zero - * @flags: ASYNC_TX_ACK, ASYNC_TX_DEP_ACK + * @flags: ASYNC_TX_ACK * @depend_tx: xor depends on the result of this transaction. * @cb_fn: function to call when the xor completes * @cb_param: parameter to pass to the callback routine diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c index f8d2d35ed298..0ef5362c8d02 100644 --- a/drivers/md/raid5.c +++ b/drivers/md/raid5.c @@ -525,14 +525,12 @@ async_copy_data(int frombio, struct bio *bio, struct page *page, bio_page = bio_iovec_idx(bio, i)->bv_page; if (frombio) tx = async_memcpy(page, bio_page, page_offset, - b_offset, clen, - ASYNC_TX_DEP_ACK, - tx, NULL, NULL); + b_offset, clen, 0, + tx, NULL, NULL); else tx = async_memcpy(bio_page, page, b_offset, - page_offset, clen, - ASYNC_TX_DEP_ACK, - tx, NULL, NULL); + page_offset, clen, 0, + tx, NULL, NULL); } if (clen < len) /* hit end of page */ break; @@ -615,8 +613,7 @@ static void ops_run_biofill(struct stripe_head *sh) } atomic_inc(&sh->count); - async_trigger_callback(ASYNC_TX_DEP_ACK | ASYNC_TX_ACK, tx, - ops_complete_biofill, sh); + async_trigger_callback(ASYNC_TX_ACK, tx, ops_complete_biofill, sh); } static void ops_complete_compute5(void *stripe_head_ref) @@ -701,8 +698,8 @@ ops_run_prexor(struct stripe_head *sh, struct dma_async_tx_descriptor *tx) } tx = async_xor(xor_dest, xor_srcs, 0, count, STRIPE_SIZE, - ASYNC_TX_DEP_ACK | ASYNC_TX_XOR_DROP_DST, tx, - ops_complete_prexor, sh); + ASYNC_TX_XOR_DROP_DST, tx, + ops_complete_prexor, sh); return tx; } @@ -809,7 +806,7 @@ ops_run_postxor(struct stripe_head *sh, struct dma_async_tx_descriptor *tx) * set ASYNC_TX_XOR_DROP_DST and ASYNC_TX_XOR_ZERO_DST * for the synchronous xor case */ - flags = ASYNC_TX_DEP_ACK | ASYNC_TX_ACK | + flags = ASYNC_TX_ACK | (prexor ? ASYNC_TX_XOR_DROP_DST : ASYNC_TX_XOR_ZERO_DST); atomic_inc(&sh->count); @@ -858,7 +855,7 @@ static void ops_run_check(struct stripe_head *sh) &sh->ops.zero_sum_result, 0, NULL, NULL, NULL); atomic_inc(&sh->count); - tx = async_trigger_callback(ASYNC_TX_DEP_ACK | ASYNC_TX_ACK, tx, + tx = async_trigger_callback(ASYNC_TX_ACK, tx, ops_complete_check, sh); } @@ -2687,8 +2684,8 @@ static void handle_stripe_expansion(raid5_conf_t *conf, struct stripe_head *sh, /* place all the copies on one channel */ tx = async_memcpy(sh2->dev[dd_idx].page, - sh->dev[i].page, 0, 0, STRIPE_SIZE, - ASYNC_TX_DEP_ACK, tx, NULL, NULL); + sh->dev[i].page, 0, 0, STRIPE_SIZE, + 0, tx, NULL, NULL); set_bit(R5_Expanded, &sh2->dev[dd_idx].flags); set_bit(R5_UPTODATE, &sh2->dev[dd_idx].flags); diff --git a/include/linux/async_tx.h b/include/linux/async_tx.h index 513150d8c25b..9f14cd540cd2 100644 --- a/include/linux/async_tx.h +++ b/include/linux/async_tx.h @@ -58,13 +58,11 @@ struct dma_chan_ref { * array. * @ASYNC_TX_ACK: immediately ack the descriptor, precludes setting up a * dependency chain - * @ASYNC_TX_DEP_ACK: ack the dependency descriptor. Useful for chaining. */ enum async_tx_flags { ASYNC_TX_XOR_ZERO_DST = (1 << 0), ASYNC_TX_XOR_DROP_DST = (1 << 1), - ASYNC_TX_ACK = (1 << 3), - ASYNC_TX_DEP_ACK = (1 << 4), + ASYNC_TX_ACK = (1 << 2), }; #ifdef CONFIG_DMA_ENGINE -- cgit v1.2.3 From a08abd8ca890a377521d65d493d174bebcaf694b Mon Sep 17 00:00:00 2001 From: Dan Williams Date: Wed, 3 Jun 2009 11:43:59 -0700 Subject: async_tx: structify submission arguments, add scribble Prepare the api for the arrival of a new parameter, 'scribble'. This will allow callers to identify scratchpad memory for dma address or page address conversions. As this adds yet another parameter, take this opportunity to convert the common submission parameters (flags, dependency, callback, and callback argument) into an object that is passed by reference. Also, take this opportunity to fix up the kerneldoc and add notes about the relevant ASYNC_TX_* flags for each routine. [ Impact: moves api pass-by-value parameters to a pass-by-reference struct ] Signed-off-by: Andre Noll Acked-by: Maciej Sosnowski Signed-off-by: Dan Williams --- Documentation/crypto/async-tx-api.txt | 6 +- crypto/async_tx/async_memcpy.c | 26 +++---- crypto/async_tx/async_memset.c | 25 +++---- crypto/async_tx/async_tx.c | 51 +++++++------- crypto/async_tx/async_xor.c | 123 +++++++++++++++++----------------- drivers/md/raid5.c | 59 +++++++++------- include/linux/async_tx.h | 84 ++++++++++++++--------- 7 files changed, 200 insertions(+), 174 deletions(-) (limited to 'include') diff --git a/Documentation/crypto/async-tx-api.txt b/Documentation/crypto/async-tx-api.txt index 76feda8541dc..dfe0475f7919 100644 --- a/Documentation/crypto/async-tx-api.txt +++ b/Documentation/crypto/async-tx-api.txt @@ -54,11 +54,7 @@ features surfaced as a result: 3.1 General format of the API: struct dma_async_tx_descriptor * -async_(, - enum async_tx_flags flags, - struct dma_async_tx_descriptor *dependency, - dma_async_tx_callback callback_routine, - void *callback_parameter); +async_(, struct async_submit ctl *submit) 3.2 Supported operations: memcpy - memory copy between a source and a destination buffer diff --git a/crypto/async_tx/async_memcpy.c b/crypto/async_tx/async_memcpy.c index 7117ec6f1b74..89e05556f3df 100644 --- a/crypto/async_tx/async_memcpy.c +++ b/crypto/async_tx/async_memcpy.c @@ -33,28 +33,28 @@ * async_memcpy - attempt to copy memory with a dma engine. * @dest: destination page * @src: src page - * @offset: offset in pages to start transaction + * @dest_offset: offset into 'dest' to start transaction + * @src_offset: offset into 'src' to start transaction * @len: length in bytes - * @flags: ASYNC_TX_ACK - * @depend_tx: memcpy depends on the result of this transaction - * @cb_fn: function to call when the memcpy completes - * @cb_param: parameter to pass to the callback routine + * @submit: submission / completion modifiers + * + * honored flags: ASYNC_TX_ACK */ struct dma_async_tx_descriptor * async_memcpy(struct page *dest, struct page *src, unsigned int dest_offset, - unsigned int src_offset, size_t len, enum async_tx_flags flags, - struct dma_async_tx_descriptor *depend_tx, - dma_async_tx_callback cb_fn, void *cb_param) + unsigned int src_offset, size_t len, + struct async_submit_ctl *submit) { - struct dma_chan *chan = async_tx_find_channel(depend_tx, DMA_MEMCPY, + struct dma_chan *chan = async_tx_find_channel(submit, DMA_MEMCPY, &dest, 1, &src, 1, len); struct dma_device *device = chan ? chan->device : NULL; struct dma_async_tx_descriptor *tx = NULL; if (device) { dma_addr_t dma_dest, dma_src; - unsigned long dma_prep_flags = cb_fn ? DMA_PREP_INTERRUPT : 0; + unsigned long dma_prep_flags; + dma_prep_flags = submit->cb_fn ? DMA_PREP_INTERRUPT : 0; dma_dest = dma_map_page(device->dev, dest, dest_offset, len, DMA_FROM_DEVICE); @@ -67,13 +67,13 @@ async_memcpy(struct page *dest, struct page *src, unsigned int dest_offset, if (tx) { pr_debug("%s: (async) len: %zu\n", __func__, len); - async_tx_submit(chan, tx, flags, depend_tx, cb_fn, cb_param); + async_tx_submit(chan, tx, submit); } else { void *dest_buf, *src_buf; pr_debug("%s: (sync) len: %zu\n", __func__, len); /* wait for any prerequisite operations */ - async_tx_quiesce(&depend_tx); + async_tx_quiesce(&submit->depend_tx); dest_buf = kmap_atomic(dest, KM_USER0) + dest_offset; src_buf = kmap_atomic(src, KM_USER1) + src_offset; @@ -83,7 +83,7 @@ async_memcpy(struct page *dest, struct page *src, unsigned int dest_offset, kunmap_atomic(dest_buf, KM_USER0); kunmap_atomic(src_buf, KM_USER1); - async_tx_sync_epilog(cb_fn, cb_param); + async_tx_sync_epilog(submit); } return tx; diff --git a/crypto/async_tx/async_memset.c b/crypto/async_tx/async_memset.c index b2f133885b7f..c14437238f4c 100644 --- a/crypto/async_tx/async_memset.c +++ b/crypto/async_tx/async_memset.c @@ -35,26 +35,23 @@ * @val: fill value * @offset: offset in pages to start transaction * @len: length in bytes - * @flags: ASYNC_TX_ACK - * @depend_tx: memset depends on the result of this transaction - * @cb_fn: function to call when the memcpy completes - * @cb_param: parameter to pass to the callback routine + * + * honored flags: ASYNC_TX_ACK */ struct dma_async_tx_descriptor * -async_memset(struct page *dest, int val, unsigned int offset, - size_t len, enum async_tx_flags flags, - struct dma_async_tx_descriptor *depend_tx, - dma_async_tx_callback cb_fn, void *cb_param) +async_memset(struct page *dest, int val, unsigned int offset, size_t len, + struct async_submit_ctl *submit) { - struct dma_chan *chan = async_tx_find_channel(depend_tx, DMA_MEMSET, + struct dma_chan *chan = async_tx_find_channel(submit, DMA_MEMSET, &dest, 1, NULL, 0, len); struct dma_device *device = chan ? chan->device : NULL; struct dma_async_tx_descriptor *tx = NULL; if (device) { dma_addr_t dma_dest; - unsigned long dma_prep_flags = cb_fn ? DMA_PREP_INTERRUPT : 0; + unsigned long dma_prep_flags; + dma_prep_flags = submit->cb_fn ? DMA_PREP_INTERRUPT : 0; dma_dest = dma_map_page(device->dev, dest, offset, len, DMA_FROM_DEVICE); @@ -64,19 +61,19 @@ async_memset(struct page *dest, int val, unsigned int offset, if (tx) { pr_debug("%s: (async) len: %zu\n", __func__, len); - async_tx_submit(chan, tx, flags, depend_tx, cb_fn, cb_param); + async_tx_submit(chan, tx, submit); } else { /* run the memset synchronously */ void *dest_buf; pr_debug("%s: (sync) len: %zu\n", __func__, len); - dest_buf = (void *) (((char *) page_address(dest)) + offset); + dest_buf = page_address(dest) + offset; /* wait for any prerequisite operations */ - async_tx_quiesce(&depend_tx); + async_tx_quiesce(&submit->depend_tx); memset(dest_buf, val, len); - async_tx_sync_epilog(cb_fn, cb_param); + async_tx_sync_epilog(submit); } return tx; diff --git a/crypto/async_tx/async_tx.c b/crypto/async_tx/async_tx.c index 3766bc3d7d89..802a5ce437d9 100644 --- a/crypto/async_tx/async_tx.c +++ b/crypto/async_tx/async_tx.c @@ -45,13 +45,15 @@ static void __exit async_tx_exit(void) /** * __async_tx_find_channel - find a channel to carry out the operation or let * the transaction execute synchronously - * @depend_tx: transaction dependency + * @submit: transaction dependency and submission modifiers * @tx_type: transaction type */ struct dma_chan * -__async_tx_find_channel(struct dma_async_tx_descriptor *depend_tx, - enum dma_transaction_type tx_type) +__async_tx_find_channel(struct async_submit_ctl *submit, + enum dma_transaction_type tx_type) { + struct dma_async_tx_descriptor *depend_tx = submit->depend_tx; + /* see if we can keep the chain on one channel */ if (depend_tx && dma_has_cap(tx_type, depend_tx->chan->device->cap_mask)) @@ -144,13 +146,14 @@ async_tx_channel_switch(struct dma_async_tx_descriptor *depend_tx, /** - * submit_disposition - while holding depend_tx->lock we must avoid submitting - * new operations to prevent a circular locking dependency with - * drivers that already hold a channel lock when calling - * async_tx_run_dependencies. + * submit_disposition - flags for routing an incoming operation * @ASYNC_TX_SUBMITTED: we were able to append the new operation under the lock * @ASYNC_TX_CHANNEL_SWITCH: when the lock is dropped schedule a channel switch * @ASYNC_TX_DIRECT_SUBMIT: when the lock is dropped submit directly + * + * while holding depend_tx->lock we must avoid submitting new operations + * to prevent a circular locking dependency with drivers that already + * hold a channel lock when calling async_tx_run_dependencies. */ enum submit_disposition { ASYNC_TX_SUBMITTED, @@ -160,11 +163,12 @@ enum submit_disposition { void async_tx_submit(struct dma_chan *chan, struct dma_async_tx_descriptor *tx, - enum async_tx_flags flags, struct dma_async_tx_descriptor *depend_tx, - dma_async_tx_callback cb_fn, void *cb_param) + struct async_submit_ctl *submit) { - tx->callback = cb_fn; - tx->callback_param = cb_param; + struct dma_async_tx_descriptor *depend_tx = submit->depend_tx; + + tx->callback = submit->cb_fn; + tx->callback_param = submit->cb_param; if (depend_tx) { enum submit_disposition s; @@ -220,7 +224,7 @@ async_tx_submit(struct dma_chan *chan, struct dma_async_tx_descriptor *tx, tx->tx_submit(tx); } - if (flags & ASYNC_TX_ACK) + if (submit->flags & ASYNC_TX_ACK) async_tx_ack(tx); if (depend_tx) @@ -229,21 +233,20 @@ async_tx_submit(struct dma_chan *chan, struct dma_async_tx_descriptor *tx, EXPORT_SYMBOL_GPL(async_tx_submit); /** - * async_trigger_callback - schedules the callback function to be run after - * any dependent operations have been completed. - * @flags: ASYNC_TX_ACK - * @depend_tx: 'callback' requires the completion of this transaction - * @cb_fn: function to call after depend_tx completes - * @cb_param: parameter to pass to the callback routine + * async_trigger_callback - schedules the callback function to be run + * @submit: submission and completion parameters + * + * honored flags: ASYNC_TX_ACK + * + * The callback is run after any dependent operations have completed. */ struct dma_async_tx_descriptor * -async_trigger_callback(enum async_tx_flags flags, - struct dma_async_tx_descriptor *depend_tx, - dma_async_tx_callback cb_fn, void *cb_param) +async_trigger_callback(struct async_submit_ctl *submit) { struct dma_chan *chan; struct dma_device *device; struct dma_async_tx_descriptor *tx; + struct dma_async_tx_descriptor *depend_tx = submit->depend_tx; if (depend_tx) { chan = depend_tx->chan; @@ -262,14 +265,14 @@ async_trigger_callback(enum async_tx_flags flags, if (tx) { pr_debug("%s: (async)\n", __func__); - async_tx_submit(chan, tx, flags, depend_tx, cb_fn, cb_param); + async_tx_submit(chan, tx, submit); } else { pr_debug("%s: (sync)\n", __func__); /* wait for any prerequisite operations */ - async_tx_quiesce(&depend_tx); + async_tx_quiesce(&submit->depend_tx); - async_tx_sync_epilog(cb_fn, cb_param); + async_tx_sync_epilog(submit); } return tx; diff --git a/crypto/async_tx/async_xor.c b/crypto/async_tx/async_xor.c index 3cc5dc763b54..691fa98a18c4 100644 --- a/crypto/async_tx/async_xor.c +++ b/crypto/async_tx/async_xor.c @@ -34,18 +34,16 @@ static __async_inline struct dma_async_tx_descriptor * do_async_xor(struct dma_chan *chan, struct page *dest, struct page **src_list, unsigned int offset, int src_cnt, size_t len, - enum async_tx_flags flags, - struct dma_async_tx_descriptor *depend_tx, - dma_async_tx_callback cb_fn, void *cb_param) + struct async_submit_ctl *submit) { struct dma_device *dma = chan->device; dma_addr_t *dma_src = (dma_addr_t *) src_list; struct dma_async_tx_descriptor *tx = NULL; int src_off = 0; int i; - dma_async_tx_callback _cb_fn; - void *_cb_param; - enum async_tx_flags async_flags; + dma_async_tx_callback cb_fn_orig = submit->cb_fn; + void *cb_param_orig = submit->cb_param; + enum async_tx_flags flags_orig = submit->flags; enum dma_ctrl_flags dma_flags; int xor_src_cnt; dma_addr_t dma_dest; @@ -63,7 +61,7 @@ do_async_xor(struct dma_chan *chan, struct page *dest, struct page **src_list, } while (src_cnt) { - async_flags = flags; + submit->flags = flags_orig; dma_flags = 0; xor_src_cnt = min(src_cnt, dma->max_xor); /* if we are submitting additional xors, leave the chain open, @@ -71,15 +69,15 @@ do_async_xor(struct dma_chan *chan, struct page *dest, struct page **src_list, * buffer mapped */ if (src_cnt > xor_src_cnt) { - async_flags &= ~ASYNC_TX_ACK; + submit->flags &= ~ASYNC_TX_ACK; dma_flags = DMA_COMPL_SKIP_DEST_UNMAP; - _cb_fn = NULL; - _cb_param = NULL; + submit->cb_fn = NULL; + submit->cb_param = NULL; } else { - _cb_fn = cb_fn; - _cb_param = cb_param; + submit->cb_fn = cb_fn_orig; + submit->cb_param = cb_param_orig; } - if (_cb_fn) + if (submit->cb_fn) dma_flags |= DMA_PREP_INTERRUPT; /* Since we have clobbered the src_list we are committed @@ -90,7 +88,7 @@ do_async_xor(struct dma_chan *chan, struct page *dest, struct page **src_list, xor_src_cnt, len, dma_flags); if (unlikely(!tx)) - async_tx_quiesce(&depend_tx); + async_tx_quiesce(&submit->depend_tx); /* spin wait for the preceeding transactions to complete */ while (unlikely(!tx)) { @@ -101,10 +99,8 @@ do_async_xor(struct dma_chan *chan, struct page *dest, struct page **src_list, dma_flags); } - async_tx_submit(chan, tx, async_flags, depend_tx, _cb_fn, - _cb_param); - - depend_tx = tx; + async_tx_submit(chan, tx, submit); + submit->depend_tx = tx; if (src_cnt > xor_src_cnt) { /* drop completed sources */ @@ -123,8 +119,7 @@ do_async_xor(struct dma_chan *chan, struct page *dest, struct page **src_list, static void do_sync_xor(struct page *dest, struct page **src_list, unsigned int offset, - int src_cnt, size_t len, enum async_tx_flags flags, - dma_async_tx_callback cb_fn, void *cb_param) + int src_cnt, size_t len, struct async_submit_ctl *submit) { int i; int xor_src_cnt; @@ -139,7 +134,7 @@ do_sync_xor(struct page *dest, struct page **src_list, unsigned int offset, /* set destination address */ dest_buf = page_address(dest) + offset; - if (flags & ASYNC_TX_XOR_ZERO_DST) + if (submit->flags & ASYNC_TX_XOR_ZERO_DST) memset(dest_buf, 0, len); while (src_cnt > 0) { @@ -152,33 +147,35 @@ do_sync_xor(struct page *dest, struct page **src_list, unsigned int offset, src_off += xor_src_cnt; } - async_tx_sync_epilog(cb_fn, cb_param); + async_tx_sync_epilog(submit); } /** * async_xor - attempt to xor a set of blocks with a dma engine. - * xor_blocks always uses the dest as a source so the ASYNC_TX_XOR_ZERO_DST - * flag must be set to not include dest data in the calculation. The - * assumption with dma eninges is that they only use the destination - * buffer as a source when it is explicity specified in the source list. * @dest: destination page - * @src_list: array of source pages (if the dest is also a source it must be - * at index zero). The contents of this array may be overwritten. - * @offset: offset in pages to start transaction + * @src_list: array of source pages + * @offset: common src/dst offset to start transaction * @src_cnt: number of source pages * @len: length in bytes - * @flags: ASYNC_TX_XOR_ZERO_DST, ASYNC_TX_XOR_DROP_DEST, ASYNC_TX_ACK - * @depend_tx: xor depends on the result of this transaction. - * @cb_fn: function to call when the xor completes - * @cb_param: parameter to pass to the callback routine + * @submit: submission / completion modifiers + * + * honored flags: ASYNC_TX_ACK, ASYNC_TX_XOR_ZERO_DST, ASYNC_TX_XOR_DROP_DST + * + * xor_blocks always uses the dest as a source so the + * ASYNC_TX_XOR_ZERO_DST flag must be set to not include dest data in + * the calculation. The assumption with dma eninges is that they only + * use the destination buffer as a source when it is explicity specified + * in the source list. + * + * src_list note: if the dest is also a source it must be at index zero. + * The contents of this array will be overwritten if a scribble region + * is not specified. */ struct dma_async_tx_descriptor * async_xor(struct page *dest, struct page **src_list, unsigned int offset, - int src_cnt, size_t len, enum async_tx_flags flags, - struct dma_async_tx_descriptor *depend_tx, - dma_async_tx_callback cb_fn, void *cb_param) + int src_cnt, size_t len, struct async_submit_ctl *submit) { - struct dma_chan *chan = async_tx_find_channel(depend_tx, DMA_XOR, + struct dma_chan *chan = async_tx_find_channel(submit, DMA_XOR, &dest, 1, src_list, src_cnt, len); BUG_ON(src_cnt <= 1); @@ -188,7 +185,7 @@ async_xor(struct page *dest, struct page **src_list, unsigned int offset, pr_debug("%s (async): len: %zu\n", __func__, len); return do_async_xor(chan, dest, src_list, offset, src_cnt, len, - flags, depend_tx, cb_fn, cb_param); + submit); } else { /* run the xor synchronously */ pr_debug("%s (sync): len: %zu\n", __func__, len); @@ -196,16 +193,15 @@ async_xor(struct page *dest, struct page **src_list, unsigned int offset, /* in the sync case the dest is an implied source * (assumes the dest is the first source) */ - if (flags & ASYNC_TX_XOR_DROP_DST) { + if (submit->flags & ASYNC_TX_XOR_DROP_DST) { src_cnt--; src_list++; } /* wait for any prerequisite operations */ - async_tx_quiesce(&depend_tx); + async_tx_quiesce(&submit->depend_tx); - do_sync_xor(dest, src_list, offset, src_cnt, len, - flags, cb_fn, cb_param); + do_sync_xor(dest, src_list, offset, src_cnt, len, submit); return NULL; } @@ -222,25 +218,25 @@ static int page_is_zero(struct page *p, unsigned int offset, size_t len) /** * async_xor_val - attempt a xor parity check with a dma engine. * @dest: destination page used if the xor is performed synchronously - * @src_list: array of source pages. The dest page must be listed as a source - * at index zero. The contents of this array may be overwritten. + * @src_list: array of source pages * @offset: offset in pages to start transaction * @src_cnt: number of source pages * @len: length in bytes * @result: 0 if sum == 0 else non-zero - * @flags: ASYNC_TX_ACK - * @depend_tx: xor depends on the result of this transaction. - * @cb_fn: function to call when the xor completes - * @cb_param: parameter to pass to the callback routine + * @submit: submission / completion modifiers + * + * honored flags: ASYNC_TX_ACK + * + * src_list note: if the dest is also a source it must be at index zero. + * The contents of this array will be overwritten if a scribble region + * is not specified. */ struct dma_async_tx_descriptor * -async_xor_val(struct page *dest, struct page **src_list, - unsigned int offset, int src_cnt, size_t len, - u32 *result, enum async_tx_flags flags, - struct dma_async_tx_descriptor *depend_tx, - dma_async_tx_callback cb_fn, void *cb_param) +async_xor_val(struct page *dest, struct page **src_list, unsigned int offset, + int src_cnt, size_t len, u32 *result, + struct async_submit_ctl *submit) { - struct dma_chan *chan = async_tx_find_channel(depend_tx, DMA_XOR_VAL, + struct dma_chan *chan = async_tx_find_channel(submit, DMA_XOR_VAL, &dest, 1, src_list, src_cnt, len); struct dma_device *device = chan ? chan->device : NULL; @@ -250,11 +246,12 @@ async_xor_val(struct page *dest, struct page **src_list, if (device && src_cnt <= device->max_xor) { dma_addr_t *dma_src = (dma_addr_t *) src_list; - unsigned long dma_prep_flags = cb_fn ? DMA_PREP_INTERRUPT : 0; + unsigned long dma_prep_flags; int i; pr_debug("%s: (async) len: %zu\n", __func__, len); + dma_prep_flags = submit->cb_fn ? DMA_PREP_INTERRUPT : 0; for (i = 0; i < src_cnt; i++) dma_src[i] = dma_map_page(device->dev, src_list[i], offset, len, DMA_TO_DEVICE); @@ -263,7 +260,7 @@ async_xor_val(struct page *dest, struct page **src_list, len, result, dma_prep_flags); if (unlikely(!tx)) { - async_tx_quiesce(&depend_tx); + async_tx_quiesce(&submit->depend_tx); while (!tx) { dma_async_issue_pending(chan); @@ -273,23 +270,23 @@ async_xor_val(struct page *dest, struct page **src_list, } } - async_tx_submit(chan, tx, flags, depend_tx, cb_fn, cb_param); + async_tx_submit(chan, tx, submit); } else { - unsigned long xor_flags = flags; + enum async_tx_flags flags_orig = submit->flags; pr_debug("%s: (sync) len: %zu\n", __func__, len); - xor_flags |= ASYNC_TX_XOR_DROP_DST; - xor_flags &= ~ASYNC_TX_ACK; + submit->flags |= ASYNC_TX_XOR_DROP_DST; + submit->flags &= ~ASYNC_TX_ACK; - tx = async_xor(dest, src_list, offset, src_cnt, len, xor_flags, - depend_tx, NULL, NULL); + tx = async_xor(dest, src_list, offset, src_cnt, len, submit); async_tx_quiesce(&tx); *result = page_is_zero(dest, offset, len) ? 0 : 1; - async_tx_sync_epilog(cb_fn, cb_param); + async_tx_sync_epilog(submit); + submit->flags = flags_orig; } return tx; diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c index 0ef5362c8d02..e1920f23579f 100644 --- a/drivers/md/raid5.c +++ b/drivers/md/raid5.c @@ -499,11 +499,14 @@ async_copy_data(int frombio, struct bio *bio, struct page *page, struct page *bio_page; int i; int page_offset; + struct async_submit_ctl submit; if (bio->bi_sector >= sector) page_offset = (signed)(bio->bi_sector - sector) * 512; else page_offset = (signed)(sector - bio->bi_sector) * -512; + + init_async_submit(&submit, 0, tx, NULL, NULL, NULL); bio_for_each_segment(bvl, bio, i) { int len = bio_iovec_idx(bio, i)->bv_len; int clen; @@ -525,13 +528,14 @@ async_copy_data(int frombio, struct bio *bio, struct page *page, bio_page = bio_iovec_idx(bio, i)->bv_page; if (frombio) tx = async_memcpy(page, bio_page, page_offset, - b_offset, clen, 0, - tx, NULL, NULL); + b_offset, clen, &submit); else tx = async_memcpy(bio_page, page, b_offset, - page_offset, clen, 0, - tx, NULL, NULL); + page_offset, clen, &submit); } + /* chain the operations */ + submit.depend_tx = tx; + if (clen < len) /* hit end of page */ break; page_offset += len; @@ -590,6 +594,7 @@ static void ops_run_biofill(struct stripe_head *sh) { struct dma_async_tx_descriptor *tx = NULL; raid5_conf_t *conf = sh->raid_conf; + struct async_submit_ctl submit; int i; pr_debug("%s: stripe %llu\n", __func__, @@ -613,7 +618,8 @@ static void ops_run_biofill(struct stripe_head *sh) } atomic_inc(&sh->count); - async_trigger_callback(ASYNC_TX_ACK, tx, ops_complete_biofill, sh); + init_async_submit(&submit, ASYNC_TX_ACK, tx, ops_complete_biofill, sh, NULL); + async_trigger_callback(&submit); } static void ops_complete_compute5(void *stripe_head_ref) @@ -645,6 +651,7 @@ static struct dma_async_tx_descriptor *ops_run_compute5(struct stripe_head *sh) struct page *xor_dest = tgt->page; int count = 0; struct dma_async_tx_descriptor *tx; + struct async_submit_ctl submit; int i; pr_debug("%s: stripe %llu block: %d\n", @@ -657,13 +664,12 @@ static struct dma_async_tx_descriptor *ops_run_compute5(struct stripe_head *sh) atomic_inc(&sh->count); + init_async_submit(&submit, ASYNC_TX_XOR_ZERO_DST, NULL, + ops_complete_compute5, sh, NULL); if (unlikely(count == 1)) - tx = async_memcpy(xor_dest, xor_srcs[0], 0, 0, STRIPE_SIZE, - 0, NULL, ops_complete_compute5, sh); + tx = async_memcpy(xor_dest, xor_srcs[0], 0, 0, STRIPE_SIZE, &submit); else - tx = async_xor(xor_dest, xor_srcs, 0, count, STRIPE_SIZE, - ASYNC_TX_XOR_ZERO_DST, NULL, - ops_complete_compute5, sh); + tx = async_xor(xor_dest, xor_srcs, 0, count, STRIPE_SIZE, &submit); return tx; } @@ -683,6 +689,7 @@ ops_run_prexor(struct stripe_head *sh, struct dma_async_tx_descriptor *tx) int disks = sh->disks; struct page *xor_srcs[disks]; int count = 0, pd_idx = sh->pd_idx, i; + struct async_submit_ctl submit; /* existing parity data subtracted */ struct page *xor_dest = xor_srcs[count++] = sh->dev[pd_idx].page; @@ -697,9 +704,9 @@ ops_run_prexor(struct stripe_head *sh, struct dma_async_tx_descriptor *tx) xor_srcs[count++] = dev->page; } - tx = async_xor(xor_dest, xor_srcs, 0, count, STRIPE_SIZE, - ASYNC_TX_XOR_DROP_DST, tx, - ops_complete_prexor, sh); + init_async_submit(&submit, ASYNC_TX_XOR_DROP_DST, tx, + ops_complete_prexor, sh, NULL); + tx = async_xor(xor_dest, xor_srcs, 0, count, STRIPE_SIZE, &submit); return tx; } @@ -772,7 +779,7 @@ ops_run_postxor(struct stripe_head *sh, struct dma_async_tx_descriptor *tx) /* kernel stack size limits the total number of disks */ int disks = sh->disks; struct page *xor_srcs[disks]; - + struct async_submit_ctl submit; int count = 0, pd_idx = sh->pd_idx, i; struct page *xor_dest; int prexor = 0; @@ -811,13 +818,11 @@ ops_run_postxor(struct stripe_head *sh, struct dma_async_tx_descriptor *tx) atomic_inc(&sh->count); - if (unlikely(count == 1)) { - flags &= ~(ASYNC_TX_XOR_DROP_DST | ASYNC_TX_XOR_ZERO_DST); - tx = async_memcpy(xor_dest, xor_srcs[0], 0, 0, STRIPE_SIZE, - flags, tx, ops_complete_postxor, sh); - } else - tx = async_xor(xor_dest, xor_srcs, 0, count, STRIPE_SIZE, - flags, tx, ops_complete_postxor, sh); + init_async_submit(&submit, flags, tx, ops_complete_postxor, sh, NULL); + if (unlikely(count == 1)) + tx = async_memcpy(xor_dest, xor_srcs[0], 0, 0, STRIPE_SIZE, &submit); + else + tx = async_xor(xor_dest, xor_srcs, 0, count, STRIPE_SIZE, &submit); } static void ops_complete_check(void *stripe_head_ref) @@ -838,6 +843,7 @@ static void ops_run_check(struct stripe_head *sh) int disks = sh->disks; struct page *xor_srcs[disks]; struct dma_async_tx_descriptor *tx; + struct async_submit_ctl submit; int count = 0, pd_idx = sh->pd_idx, i; struct page *xor_dest = xor_srcs[count++] = sh->dev[pd_idx].page; @@ -851,12 +857,13 @@ static void ops_run_check(struct stripe_head *sh) xor_srcs[count++] = dev->page; } + init_async_submit(&submit, 0, NULL, NULL, NULL, NULL); tx = async_xor_val(xor_dest, xor_srcs, 0, count, STRIPE_SIZE, - &sh->ops.zero_sum_result, 0, NULL, NULL, NULL); + &sh->ops.zero_sum_result, &submit); atomic_inc(&sh->count); - tx = async_trigger_callback(ASYNC_TX_ACK, tx, - ops_complete_check, sh); + init_async_submit(&submit, ASYNC_TX_ACK, tx, ops_complete_check, sh, NULL); + tx = async_trigger_callback(&submit); } static void raid5_run_ops(struct stripe_head *sh, unsigned long ops_request) @@ -2664,6 +2671,7 @@ static void handle_stripe_expansion(raid5_conf_t *conf, struct stripe_head *sh, if (i != sh->pd_idx && i != sh->qd_idx) { int dd_idx, j; struct stripe_head *sh2; + struct async_submit_ctl submit; sector_t bn = compute_blocknr(sh, i, 1); sector_t s = raid5_compute_sector(conf, bn, 0, @@ -2683,9 +2691,10 @@ static void handle_stripe_expansion(raid5_conf_t *conf, struct stripe_head *sh, } /* place all the copies on one channel */ + init_async_submit(&submit, 0, tx, NULL, NULL, NULL); tx = async_memcpy(sh2->dev[dd_idx].page, sh->dev[i].page, 0, 0, STRIPE_SIZE, - 0, tx, NULL, NULL); + &submit); set_bit(R5_Expanded, &sh2->dev[dd_idx].flags); set_bit(R5_UPTODATE, &sh2->dev[dd_idx].flags); diff --git a/include/linux/async_tx.h b/include/linux/async_tx.h index 9f14cd540cd2..00cfb637ddf2 100644 --- a/include/linux/async_tx.h +++ b/include/linux/async_tx.h @@ -65,6 +65,22 @@ enum async_tx_flags { ASYNC_TX_ACK = (1 << 2), }; +/** + * struct async_submit_ctl - async_tx submission/completion modifiers + * @flags: submission modifiers + * @depend_tx: parent dependency of the current operation being submitted + * @cb_fn: callback routine to run at operation completion + * @cb_param: parameter for the callback routine + * @scribble: caller provided space for dma/page address conversions + */ +struct async_submit_ctl { + enum async_tx_flags flags; + struct dma_async_tx_descriptor *depend_tx; + dma_async_tx_callback cb_fn; + void *cb_param; + void *scribble; +}; + #ifdef CONFIG_DMA_ENGINE #define async_tx_issue_pending_all dma_issue_pending_all #ifdef CONFIG_ARCH_HAS_ASYNC_TX_FIND_CHANNEL @@ -73,8 +89,8 @@ enum async_tx_flags { #define async_tx_find_channel(dep, type, dst, dst_count, src, src_count, len) \ __async_tx_find_channel(dep, type) struct dma_chan * -__async_tx_find_channel(struct dma_async_tx_descriptor *depend_tx, - enum dma_transaction_type tx_type); +__async_tx_find_channel(struct async_submit_ctl *submit, + enum dma_transaction_type tx_type); #endif /* CONFIG_ARCH_HAS_ASYNC_TX_FIND_CHANNEL */ #else static inline void async_tx_issue_pending_all(void) @@ -83,9 +99,10 @@ static inline void async_tx_issue_pending_all(void) } static inline struct dma_chan * -async_tx_find_channel(struct dma_async_tx_descriptor *depend_tx, - enum dma_transaction_type tx_type, struct page **dst, int dst_count, - struct page **src, int src_count, size_t len) +async_tx_find_channel(struct async_submit_ctl *submit, + enum dma_transaction_type tx_type, struct page **dst, + int dst_count, struct page **src, int src_count, + size_t len) { return NULL; } @@ -97,46 +114,53 @@ async_tx_find_channel(struct dma_async_tx_descriptor *depend_tx, * @cb_fn_param: parameter to pass to the callback routine */ static inline void -async_tx_sync_epilog(dma_async_tx_callback cb_fn, void *cb_fn_param) +async_tx_sync_epilog(struct async_submit_ctl *submit) +{ + if (submit->cb_fn) + submit->cb_fn(submit->cb_param); +} + +typedef union { + unsigned long addr; + struct page *page; + dma_addr_t dma; +} addr_conv_t; + +static inline void +init_async_submit(struct async_submit_ctl *args, enum async_tx_flags flags, + struct dma_async_tx_descriptor *tx, + dma_async_tx_callback cb_fn, void *cb_param, + addr_conv_t *scribble) { - if (cb_fn) - cb_fn(cb_fn_param); + args->flags = flags; + args->depend_tx = tx; + args->cb_fn = cb_fn; + args->cb_param = cb_param; + args->scribble = scribble; } -void -async_tx_submit(struct dma_chan *chan, struct dma_async_tx_descriptor *tx, - enum async_tx_flags flags, struct dma_async_tx_descriptor *depend_tx, - dma_async_tx_callback cb_fn, void *cb_fn_param); +void async_tx_submit(struct dma_chan *chan, struct dma_async_tx_descriptor *tx, + struct async_submit_ctl *submit); struct dma_async_tx_descriptor * async_xor(struct page *dest, struct page **src_list, unsigned int offset, - int src_cnt, size_t len, enum async_tx_flags flags, - struct dma_async_tx_descriptor *depend_tx, - dma_async_tx_callback cb_fn, void *cb_fn_param); + int src_cnt, size_t len, struct async_submit_ctl *submit); struct dma_async_tx_descriptor * -async_xor_val(struct page *dest, struct page **src_list, - unsigned int offset, int src_cnt, size_t len, - u32 *result, enum async_tx_flags flags, - struct dma_async_tx_descriptor *depend_tx, - dma_async_tx_callback cb_fn, void *cb_fn_param); +async_xor_val(struct page *dest, struct page **src_list, unsigned int offset, + int src_cnt, size_t len, u32 *result, + struct async_submit_ctl *submit); struct dma_async_tx_descriptor * async_memcpy(struct page *dest, struct page *src, unsigned int dest_offset, - unsigned int src_offset, size_t len, enum async_tx_flags flags, - struct dma_async_tx_descriptor *depend_tx, - dma_async_tx_callback cb_fn, void *cb_fn_param); + unsigned int src_offset, size_t len, + struct async_submit_ctl *submit); struct dma_async_tx_descriptor * async_memset(struct page *dest, int val, unsigned int offset, - size_t len, enum async_tx_flags flags, - struct dma_async_tx_descriptor *depend_tx, - dma_async_tx_callback cb_fn, void *cb_fn_param); + size_t len, struct async_submit_ctl *submit); -struct dma_async_tx_descriptor * -async_trigger_callback(enum async_tx_flags flags, - struct dma_async_tx_descriptor *depend_tx, - dma_async_tx_callback cb_fn, void *cb_fn_param); +struct dma_async_tx_descriptor *async_trigger_callback(struct async_submit_ctl *submit); void async_tx_quiesce(struct dma_async_tx_descriptor **tx); #endif /* _ASYNC_TX_H_ */ -- cgit v1.2.3 From 627ad9fd0733f0a31a266ff98a4a933eee710f0b Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Sat, 20 Jun 2009 23:21:41 -0400 Subject: ext4: Fix type warning on 64-bit platforms in tracing events header Signed-off-by: "Theodore Ts'o" --- include/trace/events/ext4.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/trace/events/ext4.h b/include/trace/events/ext4.h index acf4cc9cd36d..b456fb0a3c57 100644 --- a/include/trace/events/ext4.h +++ b/include/trace/events/ext4.h @@ -34,7 +34,8 @@ TRACE_EVENT(ext4_free_inode, TP_printk("dev %s ino %lu mode %d uid %u gid %u blocks %llu", jbd2_dev_to_name(__entry->dev), __entry->ino, __entry->mode, - __entry->uid, __entry->gid, __entry->blocks) + __entry->uid, __entry->gid, + (unsigned long long) __entry->blocks) ); TRACE_EVENT(ext4_request_inode, -- cgit v1.2.3 From 0dc8335aa3e59f1adbb0fce7c9c0b342146cd036 Mon Sep 17 00:00:00 2001 From: Robert Richter Date: Thu, 7 May 2009 15:09:33 +0200 Subject: oprofile: remove irq_flags in struct op_entry This became obsolete with this commit: 304cc6a ring_buffer: remove unused flags parameter, fix Signed-off-by: Robert Richter --- include/linux/oprofile.h | 1 - 1 file changed, 1 deletion(-) (limited to 'include') diff --git a/include/linux/oprofile.h b/include/linux/oprofile.h index 1d9518bc4c58..dbbe2dbc4418 100644 --- a/include/linux/oprofile.h +++ b/include/linux/oprofile.h @@ -171,7 +171,6 @@ struct op_sample; struct op_entry { struct ring_buffer_event *event; struct op_sample *sample; - unsigned long irq_flags; unsigned long size; unsigned long *data; }; -- cgit v1.2.3 From 51563a0e5650d0d76539625388d72d62b34c726e Mon Sep 17 00:00:00 2001 From: Robert Richter Date: Wed, 3 Jun 2009 20:54:56 +0200 Subject: x86/oprofile: introduce oprofile_add_data64() The IBS implemention writes 64 bit register values to the cpu buffer by writing two 32 values using oprofile_add_data(). This patch introduces oprofile_add_data64() to write a single 64 bit value to the buffer. Signed-off-by: Robert Richter --- arch/x86/oprofile/op_model_amd.c | 27 +++++++++------------------ drivers/oprofile/cpu_buffer.c | 15 +++++++++++++++ include/linux/oprofile.h | 1 + 3 files changed, 25 insertions(+), 18 deletions(-) (limited to 'include') diff --git a/arch/x86/oprofile/op_model_amd.c b/arch/x86/oprofile/op_model_amd.c index 6493ef7ae9ad..cc930467575d 100644 --- a/arch/x86/oprofile/op_model_amd.c +++ b/arch/x86/oprofile/op_model_amd.c @@ -140,13 +140,10 @@ op_amd_handle_ibs(struct pt_regs * const regs, rdmsrl(MSR_AMD64_IBSFETCHLINAD, val); oprofile_write_reserve(&entry, regs, val, IBS_FETCH_CODE, IBS_FETCH_SIZE); - oprofile_add_data(&entry, (u32)val); - oprofile_add_data(&entry, (u32)(val >> 32)); - oprofile_add_data(&entry, (u32)ctl); - oprofile_add_data(&entry, (u32)(ctl >> 32)); + oprofile_add_data64(&entry, val); + oprofile_add_data64(&entry, ctl); rdmsrl(MSR_AMD64_IBSFETCHPHYSAD, val); - oprofile_add_data(&entry, (u32)val); - oprofile_add_data(&entry, (u32)(val >> 32)); + oprofile_add_data64(&entry, val); oprofile_write_commit(&entry); /* reenable the IRQ */ @@ -162,23 +159,17 @@ op_amd_handle_ibs(struct pt_regs * const regs, rdmsrl(MSR_AMD64_IBSOPRIP, val); oprofile_write_reserve(&entry, regs, val, IBS_OP_CODE, IBS_OP_SIZE); - oprofile_add_data(&entry, (u32)val); - oprofile_add_data(&entry, (u32)(val >> 32)); + oprofile_add_data64(&entry, val); rdmsrl(MSR_AMD64_IBSOPDATA, val); - oprofile_add_data(&entry, (u32)val); - oprofile_add_data(&entry, (u32)(val >> 32)); + oprofile_add_data64(&entry, val); rdmsrl(MSR_AMD64_IBSOPDATA2, val); - oprofile_add_data(&entry, (u32)val); - oprofile_add_data(&entry, (u32)(val >> 32)); + oprofile_add_data64(&entry, val); rdmsrl(MSR_AMD64_IBSOPDATA3, val); - oprofile_add_data(&entry, (u32)val); - oprofile_add_data(&entry, (u32)(val >> 32)); + oprofile_add_data64(&entry, val); rdmsrl(MSR_AMD64_IBSDCLINAD, val); - oprofile_add_data(&entry, (u32)val); - oprofile_add_data(&entry, (u32)(val >> 32)); + oprofile_add_data64(&entry, val); rdmsrl(MSR_AMD64_IBSDCPHYSAD, val); - oprofile_add_data(&entry, (u32)val); - oprofile_add_data(&entry, (u32)(val >> 32)); + oprofile_add_data64(&entry, val); oprofile_write_commit(&entry); /* reenable the IRQ */ diff --git a/drivers/oprofile/cpu_buffer.c b/drivers/oprofile/cpu_buffer.c index 50640cc5eef2..a7aae24f2889 100644 --- a/drivers/oprofile/cpu_buffer.c +++ b/drivers/oprofile/cpu_buffer.c @@ -406,6 +406,21 @@ int oprofile_add_data(struct op_entry *entry, unsigned long val) return op_cpu_buffer_add_data(entry, val); } +int oprofile_add_data64(struct op_entry *entry, u64 val) +{ + if (!entry->event) + return 0; + if (op_cpu_buffer_get_size(entry) < 2) + /* + * the function returns 0 to indicate a too small + * buffer, even if there is some space left + */ + return 0; + if (!op_cpu_buffer_add_data(entry, (u32)val)) + return 0; + return op_cpu_buffer_add_data(entry, (u32)(val >> 32)); +} + int oprofile_write_commit(struct op_entry *entry) { if (!entry->event) diff --git a/include/linux/oprofile.h b/include/linux/oprofile.h index dbbe2dbc4418..d68d2ed94f15 100644 --- a/include/linux/oprofile.h +++ b/include/linux/oprofile.h @@ -179,6 +179,7 @@ void oprofile_write_reserve(struct op_entry *entry, struct pt_regs * const regs, unsigned long pc, int code, int size); int oprofile_add_data(struct op_entry *entry, unsigned long val); +int oprofile_add_data64(struct op_entry *entry, u64 val); int oprofile_write_commit(struct op_entry *entry); #endif /* OPROFILE_H */ -- cgit v1.2.3 From a635f9dd83f3382577f4544a96df12356e951a40 Mon Sep 17 00:00:00 2001 From: Jiri Kosina Date: Fri, 12 Jun 2009 15:20:55 +0200 Subject: HID: use debugfs for report dumping descriptor It is a little bit inconvenient for people who have some non-standard HID hardware (usually violating the HID specification) to have to recompile kernel with CONFIG_HID_DEBUG to be able to see kernel's perspective of the HID report descriptor and observe the parsed events. Plus the messages are then mixed up inconveniently with the rest of the dmesg stuff. This patch implements /sys/kernel/debug/hid//rdesc file, which represents the kernel's view of report descriptor (both the raw report descriptor data and parsed contents). With all the device-specific debug data being available through debugfs, there is no need for keeping CONFIG_HID_DEBUG, as the 'debug' parameter to the hid module will now only output only driver-specific debugging options, which has absolutely minimal memory footprint, just a few error messages and one global flag (hid_debug). We use the current set of output formatting functions. The ones that need to be used both for one-shot rdesc seq_file and also for continuous flow of data (individual reports, as being sent by the device) distinguish according to the passed seq_file parameter, and if it is NULL, it still output to kernel ringbuffer, otherwise the corresponding seq_file is used for output. The format of the output is preserved. Signed-off-by: Jiri Kosina --- drivers/hid/Kconfig | 15 --- drivers/hid/Makefile | 5 +- drivers/hid/hid-core.c | 13 ++- drivers/hid/hid-debug.c | 227 +++++++++++++++++++++++++++++------------- drivers/hid/hid-input.c | 13 +-- drivers/hid/usbhid/hid-core.c | 8 +- include/linux/hid-debug.h | 32 +++--- include/linux/hid.h | 4 + 8 files changed, 195 insertions(+), 122 deletions(-) (limited to 'include') diff --git a/drivers/hid/Kconfig b/drivers/hid/Kconfig index 7e67dcb3d4f6..05950a783560 100644 --- a/drivers/hid/Kconfig +++ b/drivers/hid/Kconfig @@ -31,21 +31,6 @@ config HID If unsure, say Y. -config HID_DEBUG - bool "HID debugging support" - default y - depends on HID - ---help--- - This option lets the HID layer output diagnostics about its internal - state, resolve HID usages, dump HID fields, etc. Individual HID drivers - use this debugging facility to output information about individual HID - devices, etc. - - This feature is useful for those who are either debugging the HID parser - or any HID hardware device. - - If unsure, say Y. - config HIDRAW bool "/dev/hidraw raw HID device support" depends on HID diff --git a/drivers/hid/Makefile b/drivers/hid/Makefile index 1f7cb0fd4505..cf3687d1ed63 100644 --- a/drivers/hid/Makefile +++ b/drivers/hid/Makefile @@ -3,9 +3,12 @@ # hid-objs := hid-core.o hid-input.o +ifdef CONFIG_DEBUG_FS + hid-objs += hid-debug.o +endif + obj-$(CONFIG_HID) += hid.o -hid-$(CONFIG_HID_DEBUG) += hid-debug.o hid-$(CONFIG_HIDRAW) += hidraw.o hid-logitech-objs := hid-lg.o diff --git a/drivers/hid/hid-core.c b/drivers/hid/hid-core.c index 8551693d645f..d4317db85b54 100644 --- a/drivers/hid/hid-core.c +++ b/drivers/hid/hid-core.c @@ -44,12 +44,10 @@ #define DRIVER_DESC "HID core driver" #define DRIVER_LICENSE "GPL" -#ifdef CONFIG_HID_DEBUG int hid_debug = 0; module_param_named(debug, hid_debug, int, 0600); MODULE_PARM_DESC(debug, "HID debugging (0=off, 1=probing info, 2=continuous data dumping)"); EXPORT_SYMBOL_GPL(hid_debug); -#endif /* * Register a new report for a device. @@ -987,7 +985,6 @@ int hid_set_field(struct hid_field *field, unsigned offset, __s32 value) if (offset >= field->report_count) { dbg_hid("offset (%d) exceeds report_count (%d)\n", offset, field->report_count); - hid_dump_field(field, 8); return -1; } if (field->logical_minimum < 0) { @@ -1721,6 +1718,8 @@ int hid_add_device(struct hid_device *hdev) if (!ret) hdev->status |= HID_STAT_ADDED; + hid_debug_register(hdev, dev_name(&hdev->dev)); + return ret; } EXPORT_SYMBOL_GPL(hid_add_device); @@ -1768,6 +1767,7 @@ static void hid_remove_device(struct hid_device *hdev) { if (hdev->status & HID_STAT_ADDED) { device_del(&hdev->dev); + hid_debug_unregister(hdev); hdev->status &= ~HID_STAT_ADDED; } } @@ -1843,6 +1843,10 @@ static int __init hid_init(void) { int ret; + if (hid_debug) + printk(KERN_WARNING "HID: hid_debug parameter has been deprecated. " + "Debugging data are now provided via debugfs\n"); + ret = bus_register(&hid_bus_type); if (ret) { printk(KERN_ERR "HID: can't register hid bus\n"); @@ -1853,6 +1857,8 @@ static int __init hid_init(void) if (ret) goto err_bus; + hid_debug_init(); + return 0; err_bus: bus_unregister(&hid_bus_type); @@ -1862,6 +1868,7 @@ err: static void __exit hid_exit(void) { + hid_debug_exit(); hidraw_exit(); bus_unregister(&hid_bus_type); } diff --git a/drivers/hid/hid-debug.c b/drivers/hid/hid-debug.c index 47ac1a7d66e1..067e173aa3e4 100644 --- a/drivers/hid/hid-debug.c +++ b/drivers/hid/hid-debug.c @@ -1,9 +1,9 @@ /* * (c) 1999 Andreas Gal * (c) 2000-2001 Vojtech Pavlik - * (c) 2007 Jiri Kosina + * (c) 2007-2009 Jiri Kosina * - * Some debug stuff for the HID parser. + * HID debugging support */ /* @@ -26,9 +26,13 @@ * Vojtech Pavlik, Simunkova 1594, Prague 8, 182 00 Czech Republic */ +#include +#include #include #include +static struct dentry *hid_debug_root; + struct hid_usage_entry { unsigned page; unsigned usage; @@ -331,72 +335,83 @@ static const struct hid_usage_entry hid_usage_table[] = { { 0, 0, NULL } }; -static void resolv_usage_page(unsigned page) { +static void resolv_usage_page(unsigned page, struct seq_file *f) { const struct hid_usage_entry *p; for (p = hid_usage_table; p->description; p++) if (p->page == page) { - printk("%s", p->description); + if (!f) + printk("%s", p->description); + else + seq_printf(f, "%s", p->description); return; } - printk("%04x", page); + if (!f) + printk("%04x", page); + else + seq_printf(f, "%04x", page); } -void hid_resolv_usage(unsigned usage) { +void hid_resolv_usage(unsigned usage, struct seq_file *f) { const struct hid_usage_entry *p; - if (!hid_debug) - return; - - resolv_usage_page(usage >> 16); - printk("."); + resolv_usage_page(usage >> 16, f); + if (!f) + printk("."); + else + seq_printf(f, "."); for (p = hid_usage_table; p->description; p++) if (p->page == (usage >> 16)) { for(++p; p->description && p->usage != 0; p++) if (p->usage == (usage & 0xffff)) { - printk("%s", p->description); + if (!f) + printk("%s", p->description); + else + seq_printf(f, + "%s", + p->description); return; } break; } - printk("%04x", usage & 0xffff); + if (!f) + printk("%04x", usage & 0xffff); + else + seq_printf(f, "%04x", usage & 0xffff); } EXPORT_SYMBOL_GPL(hid_resolv_usage); -static void tab(int n) { - printk(KERN_DEBUG "%*s", n, ""); +static void tab(int n, struct seq_file *f) { + seq_printf(f, "%*s", n, ""); } -void hid_dump_field(struct hid_field *field, int n) { +void hid_dump_field(struct hid_field *field, int n, struct seq_file *f) { int j; - if (!hid_debug) - return; - if (field->physical) { - tab(n); - printk("Physical("); - hid_resolv_usage(field->physical); printk(")\n"); + tab(n, f); + seq_printf(f, "Physical("); + hid_resolv_usage(field->physical, f); seq_printf(f, ")\n"); } if (field->logical) { - tab(n); - printk("Logical("); - hid_resolv_usage(field->logical); printk(")\n"); + tab(n, f); + seq_printf(f, "Logical("); + hid_resolv_usage(field->logical, f); seq_printf(f, ")\n"); } - tab(n); printk("Usage(%d)\n", field->maxusage); + tab(n, f); seq_printf(f, "Usage(%d)\n", field->maxusage); for (j = 0; j < field->maxusage; j++) { - tab(n+2); hid_resolv_usage(field->usage[j].hid); printk("\n"); + tab(n+2, f); hid_resolv_usage(field->usage[j].hid, f); seq_printf(f, "\n"); } if (field->logical_minimum != field->logical_maximum) { - tab(n); printk("Logical Minimum(%d)\n", field->logical_minimum); - tab(n); printk("Logical Maximum(%d)\n", field->logical_maximum); + tab(n, f); seq_printf(f, "Logical Minimum(%d)\n", field->logical_minimum); + tab(n, f); seq_printf(f, "Logical Maximum(%d)\n", field->logical_maximum); } if (field->physical_minimum != field->physical_maximum) { - tab(n); printk("Physical Minimum(%d)\n", field->physical_minimum); - tab(n); printk("Physical Maximum(%d)\n", field->physical_maximum); + tab(n, f); seq_printf(f, "Physical Minimum(%d)\n", field->physical_minimum); + tab(n, f); seq_printf(f, "Physical Maximum(%d)\n", field->physical_maximum); } if (field->unit_exponent) { - tab(n); printk("Unit Exponent(%d)\n", field->unit_exponent); + tab(n, f); seq_printf(f, "Unit Exponent(%d)\n", field->unit_exponent); } if (field->unit) { static const char *systems[5] = { "None", "SI Linear", "SI Rotation", "English Linear", "English Rotation" }; @@ -417,77 +432,75 @@ void hid_dump_field(struct hid_field *field, int n) { data >>= 4; if(sys > 4) { - tab(n); printk("Unit(Invalid)\n"); + tab(n, f); seq_printf(f, "Unit(Invalid)\n"); } else { int earlier_unit = 0; - tab(n); printk("Unit(%s : ", systems[sys]); + tab(n, f); seq_printf(f, "Unit(%s : ", systems[sys]); for (i=1 ; i>= 4; if (nibble != 0) { if(earlier_unit++ > 0) - printk("*"); - printk("%s", units[sys][i]); + seq_printf(f, "*"); + seq_printf(f, "%s", units[sys][i]); if(nibble != 1) { /* This is a _signed_ nibble(!) */ int val = nibble & 0x7; if(nibble & 0x08) val = -((0x7 & ~val) +1); - printk("^%d", val); + seq_printf(f, "^%d", val); } } } - printk(")\n"); + seq_printf(f, ")\n"); } } - tab(n); printk("Report Size(%u)\n", field->report_size); - tab(n); printk("Report Count(%u)\n", field->report_count); - tab(n); printk("Report Offset(%u)\n", field->report_offset); + tab(n, f); seq_printf(f, "Report Size(%u)\n", field->report_size); + tab(n, f); seq_printf(f, "Report Count(%u)\n", field->report_count); + tab(n, f); seq_printf(f, "Report Offset(%u)\n", field->report_offset); - tab(n); printk("Flags( "); + tab(n, f); seq_printf(f, "Flags( "); j = field->flags; - printk("%s", HID_MAIN_ITEM_CONSTANT & j ? "Constant " : ""); - printk("%s", HID_MAIN_ITEM_VARIABLE & j ? "Variable " : "Array "); - printk("%s", HID_MAIN_ITEM_RELATIVE & j ? "Relative " : "Absolute "); - printk("%s", HID_MAIN_ITEM_WRAP & j ? "Wrap " : ""); - printk("%s", HID_MAIN_ITEM_NONLINEAR & j ? "NonLinear " : ""); - printk("%s", HID_MAIN_ITEM_NO_PREFERRED & j ? "NoPreferredState " : ""); - printk("%s", HID_MAIN_ITEM_NULL_STATE & j ? "NullState " : ""); - printk("%s", HID_MAIN_ITEM_VOLATILE & j ? "Volatile " : ""); - printk("%s", HID_MAIN_ITEM_BUFFERED_BYTE & j ? "BufferedByte " : ""); - printk(")\n"); + seq_printf(f, "%s", HID_MAIN_ITEM_CONSTANT & j ? "Constant " : ""); + seq_printf(f, "%s", HID_MAIN_ITEM_VARIABLE & j ? "Variable " : "Array "); + seq_printf(f, "%s", HID_MAIN_ITEM_RELATIVE & j ? "Relative " : "Absolute "); + seq_printf(f, "%s", HID_MAIN_ITEM_WRAP & j ? "Wrap " : ""); + seq_printf(f, "%s", HID_MAIN_ITEM_NONLINEAR & j ? "NonLinear " : ""); + seq_printf(f, "%s", HID_MAIN_ITEM_NO_PREFERRED & j ? "NoPreferredState " : ""); + seq_printf(f, "%s", HID_MAIN_ITEM_NULL_STATE & j ? "NullState " : ""); + seq_printf(f, "%s", HID_MAIN_ITEM_VOLATILE & j ? "Volatile " : ""); + seq_printf(f, "%s", HID_MAIN_ITEM_BUFFERED_BYTE & j ? "BufferedByte " : ""); + seq_printf(f, ")\n"); } EXPORT_SYMBOL_GPL(hid_dump_field); -void hid_dump_device(struct hid_device *device) { +void hid_dump_device(struct hid_device *device, struct seq_file *f) +{ struct hid_report_enum *report_enum; struct hid_report *report; struct list_head *list; unsigned i,k; static const char *table[] = {"INPUT", "OUTPUT", "FEATURE"}; - if (!hid_debug) - return; - for (i = 0; i < HID_REPORT_TYPES; i++) { report_enum = device->report_enum + i; list = report_enum->report_list.next; while (list != &report_enum->report_list) { report = (struct hid_report *) list; - tab(2); - printk("%s", table[i]); + tab(2, f); + seq_printf(f, "%s", table[i]); if (report->id) - printk("(%d)", report->id); - printk("[%s]", table[report->type]); - printk("\n"); + seq_printf(f, "(%d)", report->id); + seq_printf(f, "[%s]", table[report->type]); + seq_printf(f, "\n"); for (k = 0; k < report->maxfield; k++) { - tab(4); - printk("Field(%d)\n", k); - hid_dump_field(report->field[k], 6); + tab(4, f); + seq_printf(f, "Field(%d)\n", k); + hid_dump_field(report->field[k], 6, f); } list = list->next; } @@ -500,7 +513,7 @@ void hid_dump_input(struct hid_usage *usage, __s32 value) { return; printk(KERN_DEBUG "hid-debug: input "); - hid_resolv_usage(usage->hid); + hid_resolv_usage(usage->hid, NULL); printk(" = %d\n", value); } EXPORT_SYMBOL_GPL(hid_dump_input); @@ -767,12 +780,84 @@ static const char **names[EV_MAX + 1] = { [EV_SND] = sounds, [EV_REP] = repeats, }; -void hid_resolv_event(__u8 type, __u16 code) { - - if (!hid_debug) - return; +void hid_resolv_event(__u8 type, __u16 code, struct seq_file *f) { - printk("%s.%s", events[type] ? events[type] : "?", + seq_printf(f, "%s.%s", events[type] ? events[type] : "?", names[type] ? (names[type][code] ? names[type][code] : "?") : "?"); } -EXPORT_SYMBOL_GPL(hid_resolv_event); + +void hid_dump_input_mapping(struct hid_device *hid, struct seq_file *f) +{ + int i, j, k; + struct hid_report *report; + struct hid_usage *usage; + + for (k = HID_INPUT_REPORT; k <= HID_OUTPUT_REPORT; k++) { + list_for_each_entry(report, &hid->report_enum[k].report_list, list) { + for (i = 0; i < report->maxfield; i++) { + for ( j = 0; j < report->field[i]->maxusage; j++) { + usage = report->field[i]->usage + j; + hid_resolv_usage(usage->hid, f); + seq_printf(f, " ---> "); + hid_resolv_event(usage->type, usage->code, f); + seq_printf(f, "\n"); + } + } + } + } + +} + +static int hid_debug_rdesc_show(struct seq_file *f, void *p) +{ + struct hid_device *hdev = f->private; + int i; + + /* dump HID report descriptor */ + for (i = 0; i < hdev->rsize; i++) + seq_printf(f, "%02x ", hdev->rdesc[i]); + seq_printf(f, "\n\n"); + + /* dump parsed data and input mappings */ + hid_dump_device(hdev, f); + seq_printf(f, "\n"); + hid_dump_input_mapping(hdev, f); + + return 0; +} + +static int hid_debug_rdesc_open(struct inode *inode, struct file *file) +{ + return single_open(file, hid_debug_rdesc_show, inode->i_private); +} + +static const struct file_operations hid_debug_rdesc_fops = { + .open = hid_debug_rdesc_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, +}; + +void hid_debug_register(struct hid_device *hdev, const char *name) +{ + hdev->debug_dir = debugfs_create_dir(name, hid_debug_root); + hdev->debug_rdesc = debugfs_create_file("rdesc", 0400, + hdev->debug_dir, hdev, &hid_debug_rdesc_fops); +} + +void hid_debug_unregister(struct hid_device *hdev) +{ + debugfs_remove(hdev->debug_rdesc); + debugfs_remove(hdev->debug_dir); +} + +void hid_debug_init(void) +{ + hid_debug_root = debugfs_create_dir("hid", NULL); +} + +void hid_debug_exit(void) +{ + debugfs_remove_recursive(hid_debug_root); +} + diff --git a/drivers/hid/hid-input.c b/drivers/hid/hid-input.c index 7f183b7147e1..5862b0f3b55d 100644 --- a/drivers/hid/hid-input.c +++ b/drivers/hid/hid-input.c @@ -159,17 +159,12 @@ static void hidinput_configure_usage(struct hid_input *hidinput, struct hid_fiel field->hidinput = hidinput; - dbg_hid("Mapping: "); - hid_resolv_usage(usage->hid); - dbg_hid_line(" ---> "); - if (field->flags & HID_MAIN_ITEM_CONSTANT) goto ignore; /* only LED usages are supported in output fields */ if (field->report_type == HID_OUTPUT_REPORT && (usage->hid & HID_USAGE_PAGE) != HID_UP_LED) { - dbg_hid_line(" [non-LED output field] "); goto ignore; } @@ -561,15 +556,9 @@ mapped: set_bit(MSC_SCAN, input->mscbit); } - hid_resolv_event(usage->type, usage->code); - - dbg_hid_line("\n"); - - return; - ignore: - dbg_hid_line("IGNORED\n"); return; + } void hidinput_hid_event(struct hid_device *hid, struct hid_field *field, struct hid_usage *usage, __s32 value) diff --git a/drivers/hid/usbhid/hid-core.c b/drivers/hid/usbhid/hid-core.c index ac8049b5f1e9..708aa52d0753 100644 --- a/drivers/hid/usbhid/hid-core.c +++ b/drivers/hid/usbhid/hid-core.c @@ -4,8 +4,8 @@ * Copyright (c) 1999 Andreas Gal * Copyright (c) 2000-2005 Vojtech Pavlik * Copyright (c) 2005 Michael Haboustak for Concept2, Inc - * Copyright (c) 2006-2008 Jiri Kosina * Copyright (c) 2007-2008 Oliver Neukum + * Copyright (c) 2006-2009 Jiri Kosina */ /* @@ -886,11 +886,6 @@ static int usbhid_parse(struct hid_device *hid) goto err; } - dbg_hid("report descriptor (size %u, read %d) = ", rsize, n); - for (n = 0; n < rsize; n++) - dbg_hid_line(" %02x", (unsigned char) rdesc[n]); - dbg_hid_line("\n"); - ret = hid_parse_report(hid, rdesc, rsize); kfree(rdesc); if (ret) { @@ -1005,7 +1000,6 @@ static int usbhid_start(struct hid_device *hid) usbhid->urbctrl->transfer_flags |= (URB_NO_TRANSFER_DMA_MAP | URB_NO_SETUP_DMA_MAP); usbhid_init_reports(hid); - hid_dump_device(hid); set_bit(HID_STARTED, &usbhid->iofl); diff --git a/include/linux/hid-debug.h b/include/linux/hid-debug.h index 50d568ec178a..516e12c33235 100644 --- a/include/linux/hid-debug.h +++ b/include/linux/hid-debug.h @@ -2,7 +2,7 @@ #define __HID_DEBUG_H /* - * Copyright (c) 2007 Jiri Kosina + * Copyright (c) 2007-2009 Jiri Kosina */ /* @@ -22,24 +22,30 @@ * */ -#ifdef CONFIG_HID_DEBUG +#ifdef CONFIG_DEBUG_FS void hid_dump_input(struct hid_usage *, __s32); -void hid_dump_device(struct hid_device *); -void hid_dump_field(struct hid_field *, int); -void hid_resolv_usage(unsigned); -void hid_resolv_event(__u8, __u16); +void hid_dump_device(struct hid_device *, struct seq_file *); +void hid_dump_field(struct hid_field *, int, struct seq_file *); +void hid_resolv_usage(unsigned, struct seq_file *); +void hid_debug_register(struct hid_device *, const char *); +void hid_debug_unregister(struct hid_device *); +void hid_debug_init(void); +void hid_debug_exit(void); #else -#define hid_dump_input(a,b) do { } while (0) -#define hid_dump_device(c) do { } while (0) -#define hid_dump_field(a,b) do { } while (0) -#define hid_resolv_usage(a) do { } while (0) -#define hid_resolv_event(a,b) do { } while (0) - -#endif /* CONFIG_HID_DEBUG */ +#define hid_dump_input(a,b) do { } while (0) +#define hid_dump_device(c) do { } while (0) +#define hid_dump_field(a,b) do { } while (0) +#define hid_resolv_usage(a) do { } while (0) +#define hid_resolv_event(a,b) do { } while (0) +#define hid_debug_register(a, b) do { } while (0) +#define hid_debug_unregister(a) do { } while (0) +#define hid_debug_init() do { } while (0) +#define hid_debug_exit() do { } while (0) +#endif #endif diff --git a/include/linux/hid.h b/include/linux/hid.h index a72876e43589..da09ab140ef1 100644 --- a/include/linux/hid.h +++ b/include/linux/hid.h @@ -451,6 +451,10 @@ struct hid_device { /* device report descriptor */ char phys[64]; /* Device physical location */ char uniq[64]; /* Device unique identifier (serial #) */ + /* debugfs */ + struct dentry *debug_dir; + struct dentry *debug_rdesc; + void *driver_data; /* temporary hid_ff handling (until moved to the drivers) */ -- cgit v1.2.3 From cd667ce24796700e1a0e6e7528efc61c96ff832e Mon Sep 17 00:00:00 2001 From: Jiri Kosina Date: Fri, 12 Jun 2009 15:20:57 +0200 Subject: HID: use debugfs for events/reports dumping This is a followup patch to the one implemeting rdesc representation in debugfs rather than being dependent on compile-time CONFIG_HID_DEBUG setting. The API of the appropriate formatting functions is slightly modified -- if they are passed seq_file pointer, the one-shot output for 'rdesc' file mode is used, and therefore the message is formatted into the corresponding seq_file immediately. Otherwise the called function allocated a new buffer, formats the text into the buffer and returns the pointer to it, so that it can be queued into the ring-buffer of the processess blocked waiting on input on 'events' file in debugfs. 'debug' parameter to the 'hid' module is now used solely for the prupose of inetrnal driver state debugging (parser, transport, etc). Signed-off-by: Jiri Kosina --- drivers/hid/hid-core.c | 42 ++++++-- drivers/hid/hid-debug.c | 239 ++++++++++++++++++++++++++++++++++++++++++---- include/linux/hid-debug.h | 18 +++- include/linux/hid.h | 26 ++--- 4 files changed, 276 insertions(+), 49 deletions(-) (limited to 'include') diff --git a/drivers/hid/hid-core.c b/drivers/hid/hid-core.c index d4317db85b54..449bd747d116 100644 --- a/drivers/hid/hid-core.c +++ b/drivers/hid/hid-core.c @@ -46,7 +46,7 @@ int hid_debug = 0; module_param_named(debug, hid_debug, int, 0600); -MODULE_PARM_DESC(debug, "HID debugging (0=off, 1=probing info, 2=continuous data dumping)"); +MODULE_PARM_DESC(debug, "toggle HID debugging messages"); EXPORT_SYMBOL_GPL(hid_debug); /* @@ -859,7 +859,7 @@ static void hid_process_event(struct hid_device *hid, struct hid_field *field, struct hid_driver *hdrv = hid->driver; int ret; - hid_dump_input(usage, value); + hid_dump_input(hid, usage, value); if (hdrv && hdrv->event && hid_match_usage(hid, usage)) { ret = hdrv->event(hid, field, usage, value); @@ -981,7 +981,7 @@ int hid_set_field(struct hid_field *field, unsigned offset, __s32 value) { unsigned size = field->report_size; - hid_dump_input(field->usage + offset, value); + hid_dump_input(field->report->device, field->usage + offset, value); if (offset >= field->report_count) { dbg_hid("offset (%d) exceeds report_count (%d)\n", offset, field->report_count); @@ -1075,6 +1075,7 @@ int hid_input_report(struct hid_device *hid, int type, u8 *data, int size, int i struct hid_report_enum *report_enum = hid->report_enum + type; struct hid_driver *hdrv = hid->driver; struct hid_report *report; + char *buf; unsigned int i; int ret; @@ -1086,18 +1087,36 @@ int hid_input_report(struct hid_device *hid, int type, u8 *data, int size, int i return -1; } - dbg_hid("report (size %u) (%snumbered)\n", size, report_enum->numbered ? "" : "un"); + buf = kmalloc(sizeof(char) * HID_DEBUG_BUFSIZE, + interrupt ? GFP_ATOMIC : GFP_KERNEL); + + if (!buf) { + report = hid_get_report(report_enum, data); + goto nomem; + } + + snprintf(buf, HID_DEBUG_BUFSIZE - 1, + "\nreport (size %u) (%snumbered)\n", size, report_enum->numbered ? "" : "un"); + hid_debug_event(hid, buf); report = hid_get_report(report_enum, data); if (!report) return -1; /* dump the report */ - dbg_hid("report %d (size %u) = ", report->id, size); - for (i = 0; i < size; i++) - dbg_hid_line(" %02x", data[i]); - dbg_hid_line("\n"); + snprintf(buf, HID_DEBUG_BUFSIZE - 1, + "report %d (size %u) = ", report->id, size); + hid_debug_event(hid, buf); + for (i = 0; i < size; i++) { + snprintf(buf, HID_DEBUG_BUFSIZE - 1, + " %02x", data[i]); + hid_debug_event(hid, buf); + } + hid_debug_event(hid, "\n"); + + kfree(buf); +nomem: if (hdrv && hdrv->raw_event && hid_match_report(hid, report)) { ret = hdrv->raw_event(hid, report, data, size); if (ret != 0) @@ -1756,6 +1775,9 @@ struct hid_device *hid_allocate_device(void) for (i = 0; i < HID_REPORT_TYPES; i++) INIT_LIST_HEAD(&hdev->report_enum[i].report_list); + init_waitqueue_head(&hdev->debug_wait); + INIT_LIST_HEAD(&hdev->debug_list); + return hdev; err: put_device(&hdev->dev); @@ -1844,8 +1866,8 @@ static int __init hid_init(void) int ret; if (hid_debug) - printk(KERN_WARNING "HID: hid_debug parameter has been deprecated. " - "Debugging data are now provided via debugfs\n"); + printk(KERN_WARNING "HID: hid_debug is now used solely for parser and driver debugging.\n" + "HID: debugfs is now used for inspecting the device (report descriptor, reports)\n"); ret = bus_register(&hid_bus_type); if (ret) { diff --git a/drivers/hid/hid-debug.c b/drivers/hid/hid-debug.c index 067e173aa3e4..a331a1821e85 100644 --- a/drivers/hid/hid-debug.c +++ b/drivers/hid/hid-debug.c @@ -28,6 +28,10 @@ #include #include +#include +#include +#include + #include #include @@ -335,49 +339,86 @@ static const struct hid_usage_entry hid_usage_table[] = { { 0, 0, NULL } }; -static void resolv_usage_page(unsigned page, struct seq_file *f) { +/* Either output directly into simple seq_file, or (if f == NULL) + * allocate a separate buffer that will then be passed to the 'events' + * ringbuffer. + * + * This is because these functions can be called both for "one-shot" + * "rdesc" while resolving, or for blocking "events". + * + * This holds both for resolv_usage_page() and hid_resolv_usage(). + */ +static char *resolv_usage_page(unsigned page, struct seq_file *f) { const struct hid_usage_entry *p; + char *buf = NULL; + + if (!f) { + buf = kzalloc(sizeof(char) * HID_DEBUG_BUFSIZE, GFP_ATOMIC); + if (!buf) + return ERR_PTR(-ENOMEM); + } for (p = hid_usage_table; p->description; p++) if (p->page == page) { - if (!f) - printk("%s", p->description); - else + if (!f) { + snprintf(buf, HID_DEBUG_BUFSIZE, "%s", + p->description); + return buf; + } + else { seq_printf(f, "%s", p->description); - return; + return NULL; + } } if (!f) - printk("%04x", page); + snprintf(buf, HID_DEBUG_BUFSIZE, "%04x", page); else seq_printf(f, "%04x", page); + return buf; } -void hid_resolv_usage(unsigned usage, struct seq_file *f) { +char *hid_resolv_usage(unsigned usage, struct seq_file *f) { const struct hid_usage_entry *p; + char *buf = NULL; + int len = 0; - resolv_usage_page(usage >> 16, f); - if (!f) - printk("."); - else + buf = resolv_usage_page(usage >> 16, f); + if (IS_ERR(buf)) { + printk(KERN_ERR "error allocating HID debug buffer\n"); + return NULL; + } + + + if (!f) { + len = strlen(buf); + snprintf(buf+len, max(0, HID_DEBUG_BUFSIZE - len), "."); + len++; + } + else { seq_printf(f, "."); + } for (p = hid_usage_table; p->description; p++) if (p->page == (usage >> 16)) { for(++p; p->description && p->usage != 0; p++) if (p->usage == (usage & 0xffff)) { if (!f) - printk("%s", p->description); + snprintf(buf + len, + max(0,HID_DEBUG_BUFSIZE - len - 1), + "%s", p->description); else seq_printf(f, "%s", p->description); - return; + return buf; } break; } if (!f) - printk("%04x", usage & 0xffff); + snprintf(buf + len, max(0, HID_DEBUG_BUFSIZE - len - 1), + "%04x", usage & 0xffff); else seq_printf(f, "%04x", usage & 0xffff); + return buf; } EXPORT_SYMBOL_GPL(hid_resolv_usage); @@ -508,13 +549,37 @@ void hid_dump_device(struct hid_device *device, struct seq_file *f) } EXPORT_SYMBOL_GPL(hid_dump_device); -void hid_dump_input(struct hid_usage *usage, __s32 value) { - if (hid_debug < 2) +/* enqueue string to 'events' ring buffer */ +void hid_debug_event(struct hid_device *hdev, char *buf) +{ + int i; + struct hid_debug_list *list; + + list_for_each_entry(list, &hdev->debug_list, node) { + for (i = 0; i <= strlen(buf); i++) + list->hid_debug_buf[(list->tail + i) % (HID_DEBUG_BUFSIZE - 1)] = + buf[i]; + list->tail = (list->tail + i) % (HID_DEBUG_BUFSIZE - 1); + } +} +EXPORT_SYMBOL_GPL(hid_debug_event); + +void hid_dump_input(struct hid_device *hdev, struct hid_usage *usage, __s32 value) +{ + char *buf; + int len; + + buf = hid_resolv_usage(usage->hid, NULL); + if (!buf) return; + len = strlen(buf); + snprintf(buf + len, HID_DEBUG_BUFSIZE - len - 1, " = %d\n", value); + + hid_debug_event(hdev, buf); + + kfree(buf); + wake_up_interruptible(&hdev->debug_wait); - printk(KERN_DEBUG "hid-debug: input "); - hid_resolv_usage(usage->hid, NULL); - printk(" = %d\n", value); } EXPORT_SYMBOL_GPL(hid_dump_input); @@ -808,6 +873,7 @@ void hid_dump_input_mapping(struct hid_device *hid, struct seq_file *f) } + static int hid_debug_rdesc_show(struct seq_file *f, void *p) { struct hid_device *hdev = f->private; @@ -831,6 +897,126 @@ static int hid_debug_rdesc_open(struct inode *inode, struct file *file) return single_open(file, hid_debug_rdesc_show, inode->i_private); } +static int hid_debug_events_open(struct inode *inode, struct file *file) +{ + int err = 0; + struct hid_debug_list *list; + + if (!(list = kzalloc(sizeof(struct hid_debug_list), GFP_KERNEL))) { + err = -ENOMEM; + goto out; + } + + if (!(list->hid_debug_buf = kzalloc(sizeof(char) * HID_DEBUG_BUFSIZE, GFP_KERNEL))) { + err = -ENOMEM; + goto out; + } + list->hdev = (struct hid_device *) inode->i_private; + file->private_data = list; + mutex_init(&list->read_mutex); + + list_add_tail(&list->node, &list->hdev->debug_list); + +out: + return err; +} + +static ssize_t hid_debug_events_read(struct file *file, char __user *buffer, + size_t count, loff_t *ppos) +{ + struct hid_debug_list *list = file->private_data; + int ret = 0, len; + DECLARE_WAITQUEUE(wait, current); + + while (ret == 0) { + mutex_lock(&list->read_mutex); + if (list->head == list->tail) { + add_wait_queue(&list->hdev->debug_wait, &wait); + set_current_state(TASK_INTERRUPTIBLE); + + while (list->head == list->tail) { + if (file->f_flags & O_NONBLOCK) { + ret = -EAGAIN; + break; + } + if (signal_pending(current)) { + ret = -ERESTARTSYS; + break; + } + + if (!list->hdev || !list->hdev->debug) { + ret = -EIO; + break; + } + + /* allow O_NONBLOCK from other threads */ + mutex_unlock(&list->read_mutex); + schedule(); + mutex_lock(&list->read_mutex); + set_current_state(TASK_INTERRUPTIBLE); + } + + set_current_state(TASK_RUNNING); + remove_wait_queue(&list->hdev->debug_wait, &wait); + } + + if (ret) + goto out; + + /* pass the ringbuffer contents to userspace */ +copy_rest: + if (list->tail == list->head) + goto out; + if (list->tail > list->head) { + len = list->tail - list->head; + + if (copy_to_user(buffer + ret, &list->hid_debug_buf[list->head], len)) { + ret = -EFAULT; + goto out; + } + ret += len; + list->head += len; + } else { + len = HID_DEBUG_BUFSIZE - list->head; + + if (copy_to_user(buffer, &list->hid_debug_buf[list->head], len)) { + ret = -EFAULT; + goto out; + } + list->head = 0; + ret += len; + goto copy_rest; + } + + } +out: + mutex_unlock(&list->read_mutex); + return ret; +} + +static unsigned int hid_debug_events_poll(struct file *file, poll_table *wait) +{ + struct hid_debug_list *list = file->private_data; + + poll_wait(file, &list->hdev->debug_wait, wait); + if (list->head != list->tail) + return POLLIN | POLLRDNORM; + if (!list->hdev->debug) + return POLLERR | POLLHUP; + return 0; +} + +static int hid_debug_events_release(struct inode *inode, struct file *file) +{ + struct hid_debug_list *list = file->private_data; + + list_del(&list->node); + kfree(list->hid_debug_buf); + kfree(list); + + return 0; +} + static const struct file_operations hid_debug_rdesc_fops = { .open = hid_debug_rdesc_open, .read = seq_read, @@ -838,16 +1024,31 @@ static const struct file_operations hid_debug_rdesc_fops = { .release = single_release, }; +static const struct file_operations hid_debug_events_fops = { + .owner = THIS_MODULE, + .open = hid_debug_events_open, + .read = hid_debug_events_read, + .poll = hid_debug_events_poll, + .release = hid_debug_events_release, +}; + + void hid_debug_register(struct hid_device *hdev, const char *name) { hdev->debug_dir = debugfs_create_dir(name, hid_debug_root); hdev->debug_rdesc = debugfs_create_file("rdesc", 0400, hdev->debug_dir, hdev, &hid_debug_rdesc_fops); + hdev->debug_events = debugfs_create_file("events", 0400, + hdev->debug_dir, hdev, &hid_debug_events_fops); + hdev->debug = 1; } void hid_debug_unregister(struct hid_device *hdev) { + hdev->debug = 0; + wake_up_interruptible(&hdev->debug_wait); debugfs_remove(hdev->debug_rdesc); + debugfs_remove(hdev->debug_events); debugfs_remove(hdev->debug_dir); } diff --git a/include/linux/hid-debug.h b/include/linux/hid-debug.h index 516e12c33235..ec08ac1ad687 100644 --- a/include/linux/hid-debug.h +++ b/include/linux/hid-debug.h @@ -24,14 +24,27 @@ #ifdef CONFIG_DEBUG_FS -void hid_dump_input(struct hid_usage *, __s32); +void hid_dump_input(struct hid_device *, struct hid_usage *, __s32); void hid_dump_device(struct hid_device *, struct seq_file *); void hid_dump_field(struct hid_field *, int, struct seq_file *); -void hid_resolv_usage(unsigned, struct seq_file *); +char *hid_resolv_usage(unsigned, struct seq_file *); void hid_debug_register(struct hid_device *, const char *); void hid_debug_unregister(struct hid_device *); void hid_debug_init(void); void hid_debug_exit(void); +void hid_debug_event(struct hid_device *, char *); + +#define HID_DEBUG_BUFSIZE 512 + +struct hid_debug_list { + char *hid_debug_buf; + int head; + int tail; + struct fasync_struct *fasync; + struct hid_device *hdev; + struct list_head node; + struct mutex read_mutex; +}; #else @@ -44,6 +57,7 @@ void hid_debug_exit(void); #define hid_debug_unregister(a) do { } while (0) #define hid_debug_init() do { } while (0) #define hid_debug_exit() do { } while (0) +#define hid_debug_event(a,b) do { } while (0) #endif diff --git a/include/linux/hid.h b/include/linux/hid.h index da09ab140ef1..60fa52913f89 100644 --- a/include/linux/hid.h +++ b/include/linux/hid.h @@ -451,10 +451,6 @@ struct hid_device { /* device report descriptor */ char phys[64]; /* Device physical location */ char uniq[64]; /* Device unique identifier (serial #) */ - /* debugfs */ - struct dentry *debug_dir; - struct dentry *debug_rdesc; - void *driver_data; /* temporary hid_ff handling (until moved to the drivers) */ @@ -468,6 +464,14 @@ struct hid_device { /* device report descriptor */ /* handler for raw output data, used by hidraw */ int (*hid_output_raw_report) (struct hid_device *, __u8 *, size_t); + + /* debugging support via debugfs */ + unsigned short debug; + struct dentry *debug_dir; + struct dentry *debug_rdesc; + struct dentry *debug_events; + struct list_head debug_list; + wait_queue_head_t debug_wait; }; static inline void *hid_get_drvdata(struct hid_device *hdev) @@ -625,9 +629,7 @@ struct hid_ll_driver { /* HID core API */ -#ifdef CONFIG_HID_DEBUG extern int hid_debug; -#endif extern int hid_add_device(struct hid_device *); extern void hid_destroy_device(struct hid_device *); @@ -783,21 +785,9 @@ int hid_pidff_init(struct hid_device *hid); #define hid_pidff_init NULL #endif -#ifdef CONFIG_HID_DEBUG #define dbg_hid(format, arg...) if (hid_debug) \ printk(KERN_DEBUG "%s: " format ,\ __FILE__ , ## arg) -#define dbg_hid_line(format, arg...) if (hid_debug) \ - printk(format, ## arg) -#else -static inline int __attribute__((format(printf, 1, 2))) -dbg_hid(const char *fmt, ...) -{ - return 0; -} -#define dbg_hid_line dbg_hid -#endif /* HID_DEBUG */ - #define err_hid(format, arg...) printk(KERN_ERR "%s: " format "\n" , \ __FILE__ , ## arg) #endif /* HID_FF */ -- cgit v1.2.3 From 43ce1d23b43330634507a049b55c36e91d27282e Mon Sep 17 00:00:00 2001 From: "Aneesh Kumar K.V" Date: Sun, 14 Jun 2009 17:58:45 -0400 Subject: ext4: Fix mmap/truncate race when blocksize < pagesize && !nodellaoc This patch fixes the mmap/truncate race that was fixed for delayed allocation by merging ext4_{journalled,normal,da}_writepage() into ext4_writepage(). Signed-off-by: Aneesh Kumar K.V Acked-by: Jan Kara Signed-off-by: "Theodore Ts'o" --- fs/ext4/inode.c | 234 +++++++++++--------------------------------- include/trace/events/ext4.h | 45 +-------- 2 files changed, 58 insertions(+), 221 deletions(-) (limited to 'include') diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 1275f34589c7..97c48b5b0578 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -47,6 +47,10 @@ #define MPAGE_DA_EXTENT_TAIL 0x01 +static int __ext4_journalled_writepage(struct page *page, + struct writeback_control *wbc, + unsigned int len); + static inline int ext4_begin_ordered_truncate(struct inode *inode, loff_t new_size) { @@ -2392,7 +2396,7 @@ static int __mpage_da_writepage(struct page *page, * We need to try to allocate * unmapped blocks in the same page. * Otherwise we won't make progress - * with the page in ext4_da_writepage + * with the page in ext4_writepage */ if (ext4_bh_delay_or_unwritten(NULL, bh)) { mpage_add_bh_to_extent(mpd, logical, @@ -2519,13 +2523,47 @@ static int noalloc_get_block_write(struct inode *inode, sector_t iblock, } /* + * Note that we don't need to start a transaction unless we're journaling data + * because we should have holes filled from ext4_page_mkwrite(). We even don't + * need to file the inode to the transaction's list in ordered mode because if + * we are writing back data added by write(), the inode is already there and if + * we are writing back data modified via mmap(), noone guarantees in which + * transaction the data will hit the disk. In case we are journaling data, we + * cannot start transaction directly because transaction start ranks above page + * lock so we have to do some magic. + * * This function can get called via... * - ext4_da_writepages after taking page lock (have journal handle) * - journal_submit_inode_data_buffers (no journal handle) * - shrink_page_list via pdflush (no journal handle) * - grab_page_cache when doing write_begin (have journal handle) + * + * We don't do any block allocation in this function. If we have page with + * multiple blocks we need to write those buffer_heads that are mapped. This + * is important for mmaped based write. So if we do with blocksize 1K + * truncate(f, 1024); + * a = mmap(f, 0, 4096); + * a[0] = 'a'; + * truncate(f, 4096); + * we have in the page first buffer_head mapped via page_mkwrite call back + * but other bufer_heads would be unmapped but dirty(dirty done via the + * do_wp_page). So writepage should write the first block. If we modify + * the mmap area beyond 1024 we will again get a page_fault and the + * page_mkwrite callback will do the block allocation and mark the + * buffer_heads mapped. + * + * We redirty the page if we have any buffer_heads that is either delay or + * unwritten in the page. + * + * We can get recursively called as show below. + * + * ext4_writepage() -> kmalloc() -> __alloc_pages() -> page_launder() -> + * ext4_writepage() + * + * But since we don't do any block allocation we should not deadlock. + * Page also have the dirty flag cleared so we don't get recurive page_lock. */ -static int ext4_da_writepage(struct page *page, +static int ext4_writepage(struct page *page, struct writeback_control *wbc) { int ret = 0; @@ -2534,7 +2572,7 @@ static int ext4_da_writepage(struct page *page, struct buffer_head *page_bufs; struct inode *inode = page->mapping->host; - trace_ext4_da_writepage(inode, page); + trace_ext4_writepage(inode, page); size = i_size_read(inode); if (page->index == size >> PAGE_CACHE_SHIFT) len = size & ~PAGE_CACHE_MASK; @@ -2596,6 +2634,15 @@ static int ext4_da_writepage(struct page *page, block_commit_write(page, 0, len); } + if (PageChecked(page) && ext4_should_journal_data(inode)) { + /* + * It's mmapped pagecache. Add buffers and journal it. There + * doesn't seem much point in redirtying the page here. + */ + ClearPageChecked(page); + return __ext4_journalled_writepage(page, wbc, len); + } + if (test_opt(inode->i_sb, NOBH) && ext4_should_writeback_data(inode)) ret = nobh_writepage(page, noalloc_get_block_write, wbc); else @@ -3135,112 +3182,10 @@ static int bput_one(handle_t *handle, struct buffer_head *bh) return 0; } -/* - * Note that we don't need to start a transaction unless we're journaling data - * because we should have holes filled from ext4_page_mkwrite(). We even don't - * need to file the inode to the transaction's list in ordered mode because if - * we are writing back data added by write(), the inode is already there and if - * we are writing back data modified via mmap(), noone guarantees in which - * transaction the data will hit the disk. In case we are journaling data, we - * cannot start transaction directly because transaction start ranks above page - * lock so we have to do some magic. - * - * In all journaling modes block_write_full_page() will start the I/O. - * - * Problem: - * - * ext4_writepage() -> kmalloc() -> __alloc_pages() -> page_launder() -> - * ext4_writepage() - * - * Similar for: - * - * ext4_file_write() -> generic_file_write() -> __alloc_pages() -> ... - * - * Same applies to ext4_get_block(). We will deadlock on various things like - * lock_journal and i_data_sem - * - * Setting PF_MEMALLOC here doesn't work - too many internal memory - * allocations fail. - * - * 16May01: If we're reentered then journal_current_handle() will be - * non-zero. We simply *return*. - * - * 1 July 2001: @@@ FIXME: - * In journalled data mode, a data buffer may be metadata against the - * current transaction. But the same file is part of a shared mapping - * and someone does a writepage() on it. - * - * We will move the buffer onto the async_data list, but *after* it has - * been dirtied. So there's a small window where we have dirty data on - * BJ_Metadata. - * - * Note that this only applies to the last partial page in the file. The - * bit which block_write_full_page() uses prepare/commit for. (That's - * broken code anyway: it's wrong for msync()). - * - * It's a rare case: affects the final partial page, for journalled data - * where the file is subject to bith write() and writepage() in the same - * transction. To fix it we'll need a custom block_write_full_page(). - * We'll probably need that anyway for journalling writepage() output. - * - * We don't honour synchronous mounts for writepage(). That would be - * disastrous. Any write() or metadata operation will sync the fs for - * us. - * - */ -static int __ext4_normal_writepage(struct page *page, - struct writeback_control *wbc) -{ - struct inode *inode = page->mapping->host; - - if (test_opt(inode->i_sb, NOBH)) - return nobh_writepage(page, noalloc_get_block_write, wbc); - else - return block_write_full_page(page, noalloc_get_block_write, - wbc); -} - -static int ext4_normal_writepage(struct page *page, - struct writeback_control *wbc) -{ - struct inode *inode = page->mapping->host; - loff_t size = i_size_read(inode); - loff_t len; - - trace_ext4_normal_writepage(inode, page); - J_ASSERT(PageLocked(page)); - if (page->index == size >> PAGE_CACHE_SHIFT) - len = size & ~PAGE_CACHE_MASK; - else - len = PAGE_CACHE_SIZE; - - if (page_has_buffers(page)) { - /* if page has buffers it should all be mapped - * and allocated. If there are not buffers attached - * to the page we know the page is dirty but it lost - * buffers. That means that at some moment in time - * after write_begin() / write_end() has been called - * all buffers have been clean and thus they must have been - * written at least once. So they are all mapped and we can - * happily proceed with mapping them and writing the page. - */ - BUG_ON(walk_page_buffers(NULL, page_buffers(page), 0, len, NULL, - ext4_bh_delay_or_unwritten)); - } - - if (!ext4_journal_current_handle()) - return __ext4_normal_writepage(page, wbc); - - redirty_page_for_writepage(wbc, page); - unlock_page(page); - return 0; -} - static int __ext4_journalled_writepage(struct page *page, - struct writeback_control *wbc) + struct writeback_control *wbc, + unsigned int len) { - loff_t size; - unsigned int len; struct address_space *mapping = page->mapping; struct inode *inode = mapping->host; struct buffer_head *page_bufs; @@ -3248,16 +3193,8 @@ static int __ext4_journalled_writepage(struct page *page, int ret = 0; int err; - size = i_size_read(inode); - if (page->index == size >> PAGE_CACHE_SHIFT) - len = size & ~PAGE_CACHE_MASK; - else - len = PAGE_CACHE_SIZE; - ret = block_prepare_write(page, 0, len, noalloc_get_block_write); - if (ret != 0) - goto out_unlock; - page_bufs = page_buffers(page); + BUG_ON(!page_bufs); walk_page_buffers(handle, page_bufs, 0, len, NULL, bget_one); /* As soon as we unlock the page, it can go away, but we have * references to buffers so we are safe */ @@ -3282,67 +3219,10 @@ static int __ext4_journalled_writepage(struct page *page, walk_page_buffers(handle, page_bufs, 0, len, NULL, bput_one); EXT4_I(inode)->i_state |= EXT4_STATE_JDATA; - goto out; - -out_unlock: - unlock_page(page); out: return ret; } -static int ext4_journalled_writepage(struct page *page, - struct writeback_control *wbc) -{ - struct inode *inode = page->mapping->host; - loff_t size = i_size_read(inode); - loff_t len; - - trace_ext4_journalled_writepage(inode, page); - J_ASSERT(PageLocked(page)); - if (page->index == size >> PAGE_CACHE_SHIFT) - len = size & ~PAGE_CACHE_MASK; - else - len = PAGE_CACHE_SIZE; - - if (page_has_buffers(page)) { - /* if page has buffers it should all be mapped - * and allocated. If there are not buffers attached - * to the page we know the page is dirty but it lost - * buffers. That means that at some moment in time - * after write_begin() / write_end() has been called - * all buffers have been clean and thus they must have been - * written at least once. So they are all mapped and we can - * happily proceed with mapping them and writing the page. - */ - BUG_ON(walk_page_buffers(NULL, page_buffers(page), 0, len, NULL, - ext4_bh_delay_or_unwritten)); - } - - if (ext4_journal_current_handle()) - goto no_write; - - if (PageChecked(page)) { - /* - * It's mmapped pagecache. Add buffers and journal it. There - * doesn't seem much point in redirtying the page here. - */ - ClearPageChecked(page); - return __ext4_journalled_writepage(page, wbc); - } else { - /* - * It may be a page full of checkpoint-mode buffers. We don't - * really know unless we go poke around in the buffer_heads. - * But block_write_full_page will do the right thing. - */ - return block_write_full_page(page, noalloc_get_block_write, - wbc); - } -no_write: - redirty_page_for_writepage(wbc, page); - unlock_page(page); - return 0; -} - static int ext4_readpage(struct file *file, struct page *page) { return mpage_readpage(page, ext4_get_block); @@ -3489,7 +3369,7 @@ static int ext4_journalled_set_page_dirty(struct page *page) static const struct address_space_operations ext4_ordered_aops = { .readpage = ext4_readpage, .readpages = ext4_readpages, - .writepage = ext4_normal_writepage, + .writepage = ext4_writepage, .sync_page = block_sync_page, .write_begin = ext4_write_begin, .write_end = ext4_ordered_write_end, @@ -3504,7 +3384,7 @@ static const struct address_space_operations ext4_ordered_aops = { static const struct address_space_operations ext4_writeback_aops = { .readpage = ext4_readpage, .readpages = ext4_readpages, - .writepage = ext4_normal_writepage, + .writepage = ext4_writepage, .sync_page = block_sync_page, .write_begin = ext4_write_begin, .write_end = ext4_writeback_write_end, @@ -3519,7 +3399,7 @@ static const struct address_space_operations ext4_writeback_aops = { static const struct address_space_operations ext4_journalled_aops = { .readpage = ext4_readpage, .readpages = ext4_readpages, - .writepage = ext4_journalled_writepage, + .writepage = ext4_writepage, .sync_page = block_sync_page, .write_begin = ext4_write_begin, .write_end = ext4_journalled_write_end, @@ -3533,7 +3413,7 @@ static const struct address_space_operations ext4_journalled_aops = { static const struct address_space_operations ext4_da_aops = { .readpage = ext4_readpage, .readpages = ext4_readpages, - .writepage = ext4_da_writepage, + .writepage = ext4_writepage, .writepages = ext4_da_writepages, .sync_page = block_sync_page, .write_begin = ext4_da_write_begin, diff --git a/include/trace/events/ext4.h b/include/trace/events/ext4.h index b456fb0a3c57..dfbc9b0edc88 100644 --- a/include/trace/events/ext4.h +++ b/include/trace/events/ext4.h @@ -190,7 +190,7 @@ TRACE_EVENT(ext4_journalled_write_end, __entry->copied) ); -TRACE_EVENT(ext4_da_writepage, +TRACE_EVENT(ext4_writepage, TP_PROTO(struct inode *inode, struct page *page), TP_ARGS(inode, page), @@ -342,49 +342,6 @@ TRACE_EVENT(ext4_da_write_end, __entry->copied) ); -TRACE_EVENT(ext4_normal_writepage, - TP_PROTO(struct inode *inode, struct page *page), - - TP_ARGS(inode, page), - - TP_STRUCT__entry( - __field( dev_t, dev ) - __field( ino_t, ino ) - __field( pgoff_t, index ) - ), - - TP_fast_assign( - __entry->dev = inode->i_sb->s_dev; - __entry->ino = inode->i_ino; - __entry->index = page->index; - ), - - TP_printk("dev %s ino %lu page_index %lu", - jbd2_dev_to_name(__entry->dev), __entry->ino, __entry->index) -); - -TRACE_EVENT(ext4_journalled_writepage, - TP_PROTO(struct inode *inode, struct page *page), - - TP_ARGS(inode, page), - - TP_STRUCT__entry( - __field( dev_t, dev ) - __field( ino_t, ino ) - __field( pgoff_t, index ) - - ), - - TP_fast_assign( - __entry->dev = inode->i_sb->s_dev; - __entry->ino = inode->i_ino; - __entry->index = page->index; - ), - - TP_printk("dev %s ino %lu page_index %lu", - jbd2_dev_to_name(__entry->dev), __entry->ino, __entry->index) -); - TRACE_EVENT(ext4_discard_blocks, TP_PROTO(struct super_block *sb, unsigned long long blk, unsigned long long count), -- cgit v1.2.3 From ca94c442535a44d508c99a77e54f21a59f4fc462 Mon Sep 17 00:00:00 2001 From: Lennart Poettering Date: Mon, 15 Jun 2009 17:17:47 +0200 Subject: sched: Introduce SCHED_RESET_ON_FORK scheduling policy flag This patch introduces a new flag SCHED_RESET_ON_FORK which can be passed to the kernel via sched_setscheduler(), ORed in the policy parameter. If set this will make sure that when the process forks a) the scheduling priority is reset to DEFAULT_PRIO if it was higher and b) the scheduling policy is reset to SCHED_NORMAL if it was either SCHED_FIFO or SCHED_RR. Why have this? Currently, if a process is real-time scheduled this will 'leak' to all its child processes. For security reasons it is often (always?) a good idea to make sure that if a process acquires RT scheduling this is confined to this process and only this process. More specifically this makes the per-process resource limit RLIMIT_RTTIME useful for security purposes, because it makes it impossible to use a fork bomb to circumvent the per-process RLIMIT_RTTIME accounting. This feature is also useful for tools like 'renice' which can then change the nice level of a process without having this spill to all its child processes. Why expose this via sched_setscheduler() and not other syscalls such as prctl() or sched_setparam()? prctl() does not take a pid parameter. Due to that it would be impossible to modify this flag for other processes than the current one. The struct passed to sched_setparam() can unfortunately not be extended without breaking compatibility, since sched_setparam() lacks a size parameter. How to use this from userspace? In your RT program simply replace this: sched_setscheduler(pid, SCHED_FIFO, ¶m); by this: sched_setscheduler(pid, SCHED_FIFO|SCHED_RESET_ON_FORK, ¶m); Signed-off-by: Lennart Poettering Acked-by: Peter Zijlstra LKML-Reference: <20090615152714.GA29092@tango.0pointer.de> Signed-off-by: Ingo Molnar --- include/linux/sched.h | 6 ++++++ kernel/sched.c | 49 ++++++++++++++++++++++++++++++++++++++++--------- 2 files changed, 46 insertions(+), 9 deletions(-) (limited to 'include') diff --git a/include/linux/sched.h b/include/linux/sched.h index 4896fdfec913..d4a2c6662f7d 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -38,6 +38,8 @@ #define SCHED_BATCH 3 /* SCHED_ISO: reserved but not implemented yet */ #define SCHED_IDLE 5 +/* Can be ORed in to make sure the process is reverted back to SCHED_NORMAL on fork */ +#define SCHED_RESET_ON_FORK 0x40000000 #ifdef __KERNEL__ @@ -1209,6 +1211,10 @@ struct task_struct { unsigned did_exec:1; unsigned in_execve:1; /* Tell the LSMs that the process is doing an * execve */ + + /* Revert to default priority/policy when forking */ + unsigned sched_reset_on_fork:1; + pid_t pid; pid_t tgid; diff --git a/kernel/sched.c b/kernel/sched.c index 8ec9d13140be..32e6ede85255 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -2613,12 +2613,28 @@ void sched_fork(struct task_struct *p, int clone_flags) set_task_cpu(p, cpu); /* - * Make sure we do not leak PI boosting priority to the child: + * Revert to default priority/policy on fork if requested. Make sure we + * do not leak PI boosting priority to the child. */ - p->prio = current->normal_prio; + if (current->sched_reset_on_fork && + (p->policy == SCHED_FIFO || p->policy == SCHED_RR)) + p->policy = SCHED_NORMAL; + + if (current->sched_reset_on_fork && + (current->normal_prio < DEFAULT_PRIO)) + p->prio = DEFAULT_PRIO; + else + p->prio = current->normal_prio; + if (!rt_prio(p->prio)) p->sched_class = &fair_sched_class; + /* + * We don't need the reset flag anymore after the fork. It has + * fulfilled its duty: + */ + p->sched_reset_on_fork = 0; + #if defined(CONFIG_SCHEDSTATS) || defined(CONFIG_TASK_DELAY_ACCT) if (likely(sched_info_on())) memset(&p->sched_info, 0, sizeof(p->sched_info)); @@ -6094,17 +6110,25 @@ static int __sched_setscheduler(struct task_struct *p, int policy, unsigned long flags; const struct sched_class *prev_class = p->sched_class; struct rq *rq; + int reset_on_fork; /* may grab non-irq protected spin_locks */ BUG_ON(in_interrupt()); recheck: /* double check policy once rq lock held */ - if (policy < 0) + if (policy < 0) { + reset_on_fork = p->sched_reset_on_fork; policy = oldpolicy = p->policy; - else if (policy != SCHED_FIFO && policy != SCHED_RR && - policy != SCHED_NORMAL && policy != SCHED_BATCH && - policy != SCHED_IDLE) - return -EINVAL; + } else { + reset_on_fork = !!(policy & SCHED_RESET_ON_FORK); + policy &= ~SCHED_RESET_ON_FORK; + + if (policy != SCHED_FIFO && policy != SCHED_RR && + policy != SCHED_NORMAL && policy != SCHED_BATCH && + policy != SCHED_IDLE) + return -EINVAL; + } + /* * Valid priorities for SCHED_FIFO and SCHED_RR are * 1..MAX_USER_RT_PRIO-1, valid priority for SCHED_NORMAL, @@ -6148,6 +6172,10 @@ recheck: /* can't change other user's priorities */ if (!check_same_owner(p)) return -EPERM; + + /* Normal users shall not reset the sched_reset_on_fork flag */ + if (p->sched_reset_on_fork && !reset_on_fork) + return -EPERM; } if (user) { @@ -6191,6 +6219,8 @@ recheck: if (running) p->sched_class->put_prev_task(rq, p); + p->sched_reset_on_fork = reset_on_fork; + oldprio = p->prio; __setscheduler(rq, p, policy, param->sched_priority); @@ -6307,14 +6337,15 @@ SYSCALL_DEFINE1(sched_getscheduler, pid_t, pid) if (p) { retval = security_task_getscheduler(p); if (!retval) - retval = p->policy; + retval = p->policy + | (p->sched_reset_on_fork ? SCHED_RESET_ON_FORK : 0); } read_unlock(&tasklist_lock); return retval; } /** - * sys_sched_getscheduler - get the RT priority of a thread + * sys_sched_getparam - get the RT priority of a thread * @pid: the pid in question. * @param: structure containing the RT priority. */ -- cgit v1.2.3 From 085f30654175a91c28d2b66b9ea6cceab627fed0 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Tue, 16 Jun 2009 13:57:07 +0200 Subject: ALSA: Add new TLV types for dBwith min/max Add new types for TLV dB scale specified with min/max values instead of min/step since the resolution can't match always with the one a device provides. For example, usb audio devices give 1/256 dB resolution while ALSA TLV is based on 1/100 dB resolution. The new min/max types have less problems because the possible rounding error happens only at min/max. Signed-off-by: Takashi Iwai --- include/sound/tlv.h | 14 ++++++++++++++ sound/core/vmaster.c | 8 ++++++-- 2 files changed, 20 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/include/sound/tlv.h b/include/sound/tlv.h index d136ea2181ed..9fd5b19ccf5c 100644 --- a/include/sound/tlv.h +++ b/include/sound/tlv.h @@ -35,6 +35,8 @@ #define SNDRV_CTL_TLVT_DB_SCALE 1 /* dB scale */ #define SNDRV_CTL_TLVT_DB_LINEAR 2 /* linear volume */ #define SNDRV_CTL_TLVT_DB_RANGE 3 /* dB range container */ +#define SNDRV_CTL_TLVT_DB_MINMAX 4 /* dB scale with min/max */ +#define SNDRV_CTL_TLVT_DB_MINMAX_MUTE 5 /* dB scale with min/max with mute */ #define TLV_DB_SCALE_ITEM(min, step, mute) \ SNDRV_CTL_TLVT_DB_SCALE, 2 * sizeof(unsigned int), \ @@ -42,6 +44,18 @@ #define DECLARE_TLV_DB_SCALE(name, min, step, mute) \ unsigned int name[] = { TLV_DB_SCALE_ITEM(min, step, mute) } +/* dB scale specified with min/max values instead of step */ +#define TLV_DB_MINMAX_ITEM(min_dB, max_dB) \ + SNDRV_CTL_TLVT_DB_MINMAX, 2 * sizeof(unsigned int), \ + (min_dB), (max_dB) +#define TLV_DB_MINMAX_MUTE_ITEM(min_dB, max_dB) \ + SNDRV_CTL_TLVT_DB_MINMAX_MUTE, 2 * sizeof(unsigned int), \ + (min_dB), (max_dB) +#define DECLARE_TLV_DB_MINMAX(name, min_dB, max_dB) \ + unsigned int name[] = { TLV_DB_MINMAX_ITEM(min_dB, max_dB) } +#define DECLARE_TLV_DB_MINMAX_MUTE(name, min_dB, max_dB) \ + unsigned int name[] = { TLV_DB_MINMAX_MUTE_ITEM(min_dB, max_dB) } + /* linear volume between min_dB and max_dB (.01dB unit) */ #define TLV_DB_LINEAR_ITEM(min_dB, max_dB) \ SNDRV_CTL_TLVT_DB_LINEAR, 2 * sizeof(unsigned int), \ diff --git a/sound/core/vmaster.c b/sound/core/vmaster.c index 257624bd1997..3b9b550109cb 100644 --- a/sound/core/vmaster.c +++ b/sound/core/vmaster.c @@ -353,7 +353,8 @@ static void master_free(struct snd_kcontrol *kcontrol) * * The optional argument @tlv can be used to specify the TLV information * for dB scale of the master control. It should be a single element - * with #SNDRV_CTL_TLVT_DB_SCALE type, and should be the max 0dB. + * with #SNDRV_CTL_TLVT_DB_SCALE, #SNDRV_CTL_TLV_DB_MINMAX or + * #SNDRV_CTL_TLVT_DB_MINMAX_MUTE type, and should be the max 0dB. */ struct snd_kcontrol *snd_ctl_make_virtual_master(char *name, const unsigned int *tlv) @@ -384,7 +385,10 @@ struct snd_kcontrol *snd_ctl_make_virtual_master(char *name, kctl->private_free = master_free; /* additional (constant) TLV read */ - if (tlv && tlv[0] == SNDRV_CTL_TLVT_DB_SCALE) { + if (tlv && + (tlv[0] == SNDRV_CTL_TLVT_DB_SCALE || + tlv[0] == SNDRV_CTL_TLVT_DB_MINMAX || + tlv[0] == SNDRV_CTL_TLVT_DB_MINMAX_MUTE)) { kctl->vd[0].access |= SNDRV_CTL_ELEM_ACCESS_TLV_READ; memcpy(master->tlv, tlv, sizeof(master->tlv)); kctl->tlv.p = master->tlv; -- cgit v1.2.3 From 5b739ef8a4e8cf5201d21abff897e292c232477b Mon Sep 17 00:00:00 2001 From: Neil Horman Date: Thu, 18 Jun 2009 19:50:21 +0800 Subject: random: Add optional continuous repetition test to entropy store based rngs FIPS-140 requires that all random number generators implement continuous self tests in which each extracted block of data is compared against the last block for repetition. The ansi_cprng implements such a test, but it would be nice if the hw rng's did the same thing. Obviously its not something thats always needed, but it seems like it would be a nice feature to have on occasion. I've written the below patch which allows individual entropy stores to be flagged as desiring a continuous test to be run on them as is extracted. By default this option is off, but is enabled in the event that fips mode is selected during bootup. Signed-off-by: Neil Horman Acked-by: Matt Mackall Signed-off-by: Herbert Xu --- crypto/internal.h | 7 +------ drivers/char/random.c | 14 ++++++++++++++ include/linux/fips.h | 10 ++++++++++ 3 files changed, 25 insertions(+), 6 deletions(-) create mode 100644 include/linux/fips.h (limited to 'include') diff --git a/crypto/internal.h b/crypto/internal.h index 113579a82dff..95baaea21fbc 100644 --- a/crypto/internal.h +++ b/crypto/internal.h @@ -25,12 +25,7 @@ #include #include #include - -#ifdef CONFIG_CRYPTO_FIPS -extern int fips_enabled; -#else -#define fips_enabled 0 -#endif +#include /* Crypto notification events. */ enum { diff --git a/drivers/char/random.c b/drivers/char/random.c index 8c7444857a4b..d8a9255e1a3f 100644 --- a/drivers/char/random.c +++ b/drivers/char/random.c @@ -240,6 +240,7 @@ #include #include #include +#include #ifdef CONFIG_GENERIC_HARDIRQS # include @@ -413,6 +414,7 @@ struct entropy_store { unsigned add_ptr; int entropy_count; int input_rotate; + __u8 *last_data; }; static __u32 input_pool_data[INPUT_POOL_WORDS]; @@ -852,12 +854,21 @@ static ssize_t extract_entropy(struct entropy_store *r, void *buf, { ssize_t ret = 0, i; __u8 tmp[EXTRACT_SIZE]; + unsigned long flags; xfer_secondary_pool(r, nbytes); nbytes = account(r, nbytes, min, reserved); while (nbytes) { extract_buf(r, tmp); + + if (r->last_data) { + spin_lock_irqsave(&r->lock, flags); + if (!memcmp(tmp, r->last_data, EXTRACT_SIZE)) + panic("Hardware RNG duplicated output!\n"); + memcpy(r->last_data, tmp, EXTRACT_SIZE); + spin_unlock_irqrestore(&r->lock, flags); + } i = min_t(int, nbytes, EXTRACT_SIZE); memcpy(buf, tmp, i); nbytes -= i; @@ -940,6 +951,9 @@ static void init_std_data(struct entropy_store *r) now = ktime_get_real(); mix_pool_bytes(r, &now, sizeof(now)); mix_pool_bytes(r, utsname(), sizeof(*(utsname()))); + /* Enable continuous test in fips mode */ + if (fips_enabled) + r->last_data = kmalloc(EXTRACT_SIZE, GFP_KERNEL); } static int rand_initialize(void) diff --git a/include/linux/fips.h b/include/linux/fips.h new file mode 100644 index 000000000000..f8fb07b0b6b8 --- /dev/null +++ b/include/linux/fips.h @@ -0,0 +1,10 @@ +#ifndef _FIPS_H +#define _FIPS_H + +#ifdef CONFIG_CRYPTO_FIPS +extern int fips_enabled; +#else +#define fips_enabled 0 +#endif + +#endif -- cgit v1.2.3 From f29ac756a40d0f1bb07d682ea521e7b666ff06d5 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Fri, 19 Jun 2009 18:27:26 +0200 Subject: perf_counter: Optimize perf_swcounter_event() Similar to tracepoints, use an enable variable to reduce overhead when unused. Only look for a counter of a particular event type when we know there is at least one in the system. Signed-off-by: Peter Zijlstra LKML-Reference: Cc: Mike Galbraith Cc: Paul Mackerras Cc: Arnaldo Carvalho de Melo Signed-off-by: Ingo Molnar --- include/linux/perf_counter.h | 11 ++++++++++- kernel/perf_counter.c | 18 +++++++++++++++--- 2 files changed, 25 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/include/linux/perf_counter.h b/include/linux/perf_counter.h index 89698d8aba5c..e7213e46cf9c 100644 --- a/include/linux/perf_counter.h +++ b/include/linux/perf_counter.h @@ -669,7 +669,16 @@ static inline int is_software_counter(struct perf_counter *counter) (counter->attr.type != PERF_TYPE_HW_CACHE); } -extern void perf_swcounter_event(u32, u64, int, struct pt_regs *, u64); +extern atomic_t perf_swcounter_enabled[PERF_COUNT_SW_MAX]; + +extern void __perf_swcounter_event(u32, u64, int, struct pt_regs *, u64); + +static inline void +perf_swcounter_event(u32 event, u64 nr, int nmi, struct pt_regs *regs, u64 addr) +{ + if (atomic_read(&perf_swcounter_enabled[event])) + __perf_swcounter_event(event, nr, nmi, regs, addr); +} extern void __perf_counter_mmap(struct vm_area_struct *vma); diff --git a/kernel/perf_counter.c b/kernel/perf_counter.c index 1a933a221ea4..7515c7695428 100644 --- a/kernel/perf_counter.c +++ b/kernel/perf_counter.c @@ -3317,8 +3317,8 @@ out: put_cpu_var(perf_cpu_context); } -void -perf_swcounter_event(u32 event, u64 nr, int nmi, struct pt_regs *regs, u64 addr) +void __perf_swcounter_event(u32 event, u64 nr, int nmi, + struct pt_regs *regs, u64 addr) { struct perf_sample_data data = { .regs = regs, @@ -3509,9 +3509,19 @@ static const struct pmu *tp_perf_counter_init(struct perf_counter *counter) } #endif +atomic_t perf_swcounter_enabled[PERF_COUNT_SW_MAX]; + +static void sw_perf_counter_destroy(struct perf_counter *counter) +{ + u64 event = counter->attr.config; + + atomic_dec(&perf_swcounter_enabled[event]); +} + static const struct pmu *sw_perf_counter_init(struct perf_counter *counter) { const struct pmu *pmu = NULL; + u64 event = counter->attr.config; /* * Software counters (currently) can't in general distinguish @@ -3520,7 +3530,7 @@ static const struct pmu *sw_perf_counter_init(struct perf_counter *counter) * to be kernel events, and page faults are never hypervisor * events. */ - switch (counter->attr.config) { + switch (event) { case PERF_COUNT_SW_CPU_CLOCK: pmu = &perf_ops_cpu_clock; @@ -3541,6 +3551,8 @@ static const struct pmu *sw_perf_counter_init(struct perf_counter *counter) case PERF_COUNT_SW_PAGE_FAULTS_MAJ: case PERF_COUNT_SW_CONTEXT_SWITCHES: case PERF_COUNT_SW_CPU_MIGRATIONS: + atomic_inc(&perf_swcounter_enabled[event]); + counter->destroy = sw_perf_counter_destroy; pmu = &perf_ops_generic; break; } -- cgit v1.2.3 From 92722b1bb1ebcba767f9c6ee499992ee33367268 Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Thu, 11 Jun 2009 14:17:48 +0100 Subject: leds: Further document parameters for blink_set() The documentation for the parameters of blink_set() was a bit hard to find so put some where I'd expected to find it. Signed-off-by: Mark Brown Signed-off-by: Richard Purdie --- include/linux/leds.h | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/leds.h b/include/linux/leds.h index 376fe07732ea..c7f0b148df06 100644 --- a/include/linux/leds.h +++ b/include/linux/leds.h @@ -45,7 +45,9 @@ struct led_classdev { /* Get LED brightness level */ enum led_brightness (*brightness_get)(struct led_classdev *led_cdev); - /* Activate hardware accelerated blink */ + /* Activate hardware accelerated blink, delays are in + * miliseconds and if none is provided then a sensible default + * should be chosen. */ int (*blink_set)(struct led_classdev *led_cdev, unsigned long *delay_on, unsigned long *delay_off); -- cgit v1.2.3 From 5054d39e327f76df022163a2ebd02e444c5d65f9 Mon Sep 17 00:00:00 2001 From: Antonio Ospite Date: Fri, 19 Jun 2009 13:55:42 +0200 Subject: leds: LED driver for National Semiconductor LP3944 Funlight Chip LEDs driver for National Semiconductor LP3944 Funlight Chip http://www.national.com/pf/LP/LP3944.html This helper chip can drive up to 8 leds, with two programmable DIM modes; it could even be used as a gpio expander but this driver assumes it is used as a led controller. The DIM modes are used to set _blink_ patterns for leds, the pattern is specified supplying two parameters: - period: from 0s to 1.6s - duty cycle: percentage of the period the led is on, from 0 to 100 LP3944 can be found on Motorola A910 smartphone, where it drives the rgb leds, the camera flash light and the displays backlights. Signed-off-by: Antonio Ospite Signed-off-by: Richard Purdie --- Documentation/leds-lp3944.txt | 50 +++++ drivers/leds/Kconfig | 11 + drivers/leds/Makefile | 1 + drivers/leds/leds-lp3944.c | 466 ++++++++++++++++++++++++++++++++++++++++++ include/linux/leds-lp3944.h | 53 +++++ 5 files changed, 581 insertions(+) create mode 100644 Documentation/leds-lp3944.txt create mode 100644 drivers/leds/leds-lp3944.c create mode 100644 include/linux/leds-lp3944.h (limited to 'include') diff --git a/Documentation/leds-lp3944.txt b/Documentation/leds-lp3944.txt new file mode 100644 index 000000000000..c6eda18b15ef --- /dev/null +++ b/Documentation/leds-lp3944.txt @@ -0,0 +1,50 @@ +Kernel driver lp3944 +==================== + + * National Semiconductor LP3944 Fun-light Chip + Prefix: 'lp3944' + Addresses scanned: None (see the Notes section below) + Datasheet: Publicly available at the National Semiconductor website + http://www.national.com/pf/LP/LP3944.html + +Authors: + Antonio Ospite + + +Description +----------- +The LP3944 is a helper chip that can drive up to 8 leds, with two programmable +DIM modes; it could even be used as a gpio expander but this driver assumes it +is used as a led controller. + +The DIM modes are used to set _blink_ patterns for leds, the pattern is +specified supplying two parameters: + - period: from 0s to 1.6s + - duty cycle: percentage of the period the led is on, from 0 to 100 + +Setting a led in DIM0 or DIM1 mode makes it blink according to the pattern. +See the datasheet for details. + +LP3944 can be found on Motorola A910 smartphone, where it drives the rgb +leds, the camera flash light and the lcds power. + + +Notes +----- +The chip is used mainly in embedded contexts, so this driver expects it is +registered using the i2c_board_info mechanism. + +To register the chip at address 0x60 on adapter 0, set the platform data +according to include/linux/leds-lp3944.h, set the i2c board info: + + static struct i2c_board_info __initdata a910_i2c_board_info[] = { + { + I2C_BOARD_INFO("lp3944", 0x60), + .platform_data = &a910_lp3944_leds, + }, + }; + +and register it in the platform init function + + i2c_register_board_info(0, a910_i2c_board_info, + ARRAY_SIZE(a910_i2c_board_info)); diff --git a/drivers/leds/Kconfig b/drivers/leds/Kconfig index cfcd6bf831c9..7c8e7122aaa9 100644 --- a/drivers/leds/Kconfig +++ b/drivers/leds/Kconfig @@ -146,6 +146,17 @@ config LEDS_GPIO_OF of_platform devices. For instance, LEDs which are listed in a "dts" file. +config LEDS_LP3944 + tristate "LED Support for N.S. LP3944 (Fun Light) I2C chip" + depends on LEDS_CLASS && I2C + help + This option enables support for LEDs connected to the National + Semiconductor LP3944 Lighting Management Unit (LMU) also known as + Fun Light Chip. + + To compile this driver as a module, choose M here: the + module will be called leds-lp3944. + config LEDS_CLEVO_MAIL tristate "Mail LED on Clevo notebook" depends on LEDS_CLASS && X86 && SERIO_I8042 && DMI diff --git a/drivers/leds/Makefile b/drivers/leds/Makefile index 2d41c4dcf92f..e8cdcf77a4c3 100644 --- a/drivers/leds/Makefile +++ b/drivers/leds/Makefile @@ -20,6 +20,7 @@ obj-$(CONFIG_LEDS_COBALT_RAQ) += leds-cobalt-raq.o obj-$(CONFIG_LEDS_SUNFIRE) += leds-sunfire.o obj-$(CONFIG_LEDS_PCA9532) += leds-pca9532.o obj-$(CONFIG_LEDS_GPIO) += leds-gpio.o +obj-$(CONFIG_LEDS_LP3944) += leds-lp3944.o obj-$(CONFIG_LEDS_CLEVO_MAIL) += leds-clevo-mail.o obj-$(CONFIG_LEDS_HP6XX) += leds-hp6xx.o obj-$(CONFIG_LEDS_FSG) += leds-fsg.o diff --git a/drivers/leds/leds-lp3944.c b/drivers/leds/leds-lp3944.c new file mode 100644 index 000000000000..5946208ba26e --- /dev/null +++ b/drivers/leds/leds-lp3944.c @@ -0,0 +1,466 @@ +/* + * leds-lp3944.c - driver for National Semiconductor LP3944 Funlight Chip + * + * Copyright (C) 2009 Antonio Ospite + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + */ + +/* + * I2C driver for National Semiconductor LP3944 Funlight Chip + * http://www.national.com/pf/LP/LP3944.html + * + * This helper chip can drive up to 8 leds, with two programmable DIM modes; + * it could even be used as a gpio expander but this driver assumes it is used + * as a led controller. + * + * The DIM modes are used to set _blink_ patterns for leds, the pattern is + * specified supplying two parameters: + * - period: from 0s to 1.6s + * - duty cycle: percentage of the period the led is on, from 0 to 100 + * + * LP3944 can be found on Motorola A910 smartphone, where it drives the rgb + * leds, the camera flash light and the displays backlights. + */ + +#include +#include +#include +#include +#include +#include + +/* Read Only Registers */ +#define LP3944_REG_INPUT1 0x00 /* LEDs 0-7 InputRegister (Read Only) */ +#define LP3944_REG_REGISTER1 0x01 /* None (Read Only) */ + +#define LP3944_REG_PSC0 0x02 /* Frequency Prescaler 0 (R/W) */ +#define LP3944_REG_PWM0 0x03 /* PWM Register 0 (R/W) */ +#define LP3944_REG_PSC1 0x04 /* Frequency Prescaler 1 (R/W) */ +#define LP3944_REG_PWM1 0x05 /* PWM Register 1 (R/W) */ +#define LP3944_REG_LS0 0x06 /* LEDs 0-3 Selector (R/W) */ +#define LP3944_REG_LS1 0x07 /* LEDs 4-7 Selector (R/W) */ + +/* These registers are not used to control leds in LP3944, they can store + * arbitrary values which the chip will ignore. + */ +#define LP3944_REG_REGISTER8 0x08 +#define LP3944_REG_REGISTER9 0x09 + +#define LP3944_DIM0 0 +#define LP3944_DIM1 1 + +/* period in ms */ +#define LP3944_PERIOD_MIN 0 +#define LP3944_PERIOD_MAX 1600 + +/* duty cycle is a percentage */ +#define LP3944_DUTY_CYCLE_MIN 0 +#define LP3944_DUTY_CYCLE_MAX 100 + +#define ldev_to_led(c) container_of(c, struct lp3944_led_data, ldev) + +/* Saved data */ +struct lp3944_led_data { + u8 id; + enum lp3944_type type; + enum lp3944_status status; + struct led_classdev ldev; + struct i2c_client *client; + struct work_struct work; +}; + +struct lp3944_data { + struct mutex lock; + struct i2c_client *client; + struct lp3944_led_data leds[LP3944_LEDS_MAX]; +}; + +static int lp3944_reg_read(struct i2c_client *client, u8 reg, u8 *value) +{ + int tmp; + + tmp = i2c_smbus_read_byte_data(client, reg); + if (tmp < 0) + return -EINVAL; + + *value = tmp; + + return 0; +} + +static int lp3944_reg_write(struct i2c_client *client, u8 reg, u8 value) +{ + return i2c_smbus_write_byte_data(client, reg, value); +} + +/** + * Set the period for DIM status + * + * @client: the i2c client + * @dim: either LP3944_DIM0 or LP3944_DIM1 + * @period: period of a blink, that is a on/off cycle, expressed in ms. + */ +static int lp3944_dim_set_period(struct i2c_client *client, u8 dim, u16 period) +{ + u8 psc_reg; + u8 psc_value; + int err; + + if (dim == LP3944_DIM0) + psc_reg = LP3944_REG_PSC0; + else if (dim == LP3944_DIM1) + psc_reg = LP3944_REG_PSC1; + else + return -EINVAL; + + /* Convert period to Prescaler value */ + if (period > LP3944_PERIOD_MAX) + return -EINVAL; + + psc_value = (period * 255) / LP3944_PERIOD_MAX; + + err = lp3944_reg_write(client, psc_reg, psc_value); + + return err; +} + +/** + * Set the duty cycle for DIM status + * + * @client: the i2c client + * @dim: either LP3944_DIM0 or LP3944_DIM1 + * @duty_cycle: percentage of a period during which a led is ON + */ +static int lp3944_dim_set_dutycycle(struct i2c_client *client, u8 dim, + u8 duty_cycle) +{ + u8 pwm_reg; + u8 pwm_value; + int err; + + if (dim == LP3944_DIM0) + pwm_reg = LP3944_REG_PWM0; + else if (dim == LP3944_DIM1) + pwm_reg = LP3944_REG_PWM1; + else + return -EINVAL; + + /* Convert duty cycle to PWM value */ + if (duty_cycle > LP3944_DUTY_CYCLE_MAX) + return -EINVAL; + + pwm_value = (duty_cycle * 255) / LP3944_DUTY_CYCLE_MAX; + + err = lp3944_reg_write(client, pwm_reg, pwm_value); + + return err; +} + +/** + * Set the led status + * + * @led: a lp3944_led_data structure + * @status: one of LP3944_LED_STATUS_OFF + * LP3944_LED_STATUS_ON + * LP3944_LED_STATUS_DIM0 + * LP3944_LED_STATUS_DIM1 + */ +static int lp3944_led_set(struct lp3944_led_data *led, u8 status) +{ + struct lp3944_data *data = i2c_get_clientdata(led->client); + u8 id = led->id; + u8 reg; + u8 val = 0; + int err; + + dev_dbg(&led->client->dev, "%s: %s, status before normalization:%d\n", + __func__, led->ldev.name, status); + + switch (id) { + case LP3944_LED0: + case LP3944_LED1: + case LP3944_LED2: + case LP3944_LED3: + reg = LP3944_REG_LS0; + break; + case LP3944_LED4: + case LP3944_LED5: + case LP3944_LED6: + case LP3944_LED7: + id -= LP3944_LED4; + reg = LP3944_REG_LS1; + break; + default: + return -EINVAL; + } + + if (status > LP3944_LED_STATUS_DIM1) + return -EINVAL; + + /* invert only 0 and 1, leave unchanged the other values, + * remember we are abusing status to set blink patterns + */ + if (led->type == LP3944_LED_TYPE_LED_INVERTED && status < 2) + status = 1 - status; + + mutex_lock(&data->lock); + lp3944_reg_read(led->client, reg, &val); + + val &= ~(LP3944_LED_STATUS_MASK << (id << 1)); + val |= (status << (id << 1)); + + dev_dbg(&led->client->dev, "%s: %s, reg:%d id:%d status:%d val:%#x\n", + __func__, led->ldev.name, reg, id, status, val); + + /* set led status */ + err = lp3944_reg_write(led->client, reg, val); + mutex_unlock(&data->lock); + + return err; +} + +static int lp3944_led_set_blink(struct led_classdev *led_cdev, + unsigned long *delay_on, + unsigned long *delay_off) +{ + struct lp3944_led_data *led = ldev_to_led(led_cdev); + u16 period; + u8 duty_cycle; + int err; + + /* units are in ms */ + if (*delay_on + *delay_off > LP3944_PERIOD_MAX) + return -EINVAL; + + if (*delay_on == 0 && *delay_off == 0) { + /* Special case: the leds subsystem requires a default user + * friendly blink pattern for the LED. Let's blink the led + * slowly (1Hz). + */ + *delay_on = 500; + *delay_off = 500; + } + + period = (*delay_on) + (*delay_off); + + /* duty_cycle is the percentage of period during which the led is ON */ + duty_cycle = 100 * (*delay_on) / period; + + /* invert duty cycle for inverted leds, this has the same effect of + * swapping delay_on and delay_off + */ + if (led->type == LP3944_LED_TYPE_LED_INVERTED) + duty_cycle = 100 - duty_cycle; + + /* NOTE: using always the first DIM mode, this means that all leds + * will have the same blinking pattern. + * + * We could find a way later to have two leds blinking in hardware + * with different patterns at the same time, falling back to software + * control for the other ones. + */ + err = lp3944_dim_set_period(led->client, LP3944_DIM0, period); + if (err) + return err; + + err = lp3944_dim_set_dutycycle(led->client, LP3944_DIM0, duty_cycle); + if (err) + return err; + + dev_dbg(&led->client->dev, "%s: OK hardware accelerated blink!\n", + __func__); + + led->status = LP3944_LED_STATUS_DIM0; + schedule_work(&led->work); + + return 0; +} + +static void lp3944_led_set_brightness(struct led_classdev *led_cdev, + enum led_brightness brightness) +{ + struct lp3944_led_data *led = ldev_to_led(led_cdev); + + dev_dbg(&led->client->dev, "%s: %s, %d\n", + __func__, led_cdev->name, brightness); + + led->status = brightness; + schedule_work(&led->work); +} + +static void lp3944_led_work(struct work_struct *work) +{ + struct lp3944_led_data *led; + + led = container_of(work, struct lp3944_led_data, work); + lp3944_led_set(led, led->status); +} + +static int lp3944_configure(struct i2c_client *client, + struct lp3944_data *data, + struct lp3944_platform_data *pdata) +{ + int i, err = 0; + + for (i = 0; i < pdata->leds_size; i++) { + struct lp3944_led *pled = &pdata->leds[i]; + struct lp3944_led_data *led = &data->leds[i]; + led->client = client; + led->id = i; + + switch (pled->type) { + + case LP3944_LED_TYPE_LED: + case LP3944_LED_TYPE_LED_INVERTED: + led->type = pled->type; + led->status = pled->status; + led->ldev.name = pled->name; + led->ldev.max_brightness = 1; + led->ldev.brightness_set = lp3944_led_set_brightness; + led->ldev.blink_set = lp3944_led_set_blink; + led->ldev.flags = LED_CORE_SUSPENDRESUME; + + INIT_WORK(&led->work, lp3944_led_work); + err = led_classdev_register(&client->dev, &led->ldev); + if (err < 0) { + dev_err(&client->dev, + "couldn't register LED %s\n", + led->ldev.name); + goto exit; + } + + /* to expose the default value to userspace */ + led->ldev.brightness = led->status; + + /* Set the default led status */ + err = lp3944_led_set(led, led->status); + if (err < 0) { + dev_err(&client->dev, + "%s couldn't set STATUS %d\n", + led->ldev.name, led->status); + goto exit; + } + break; + + case LP3944_LED_TYPE_NONE: + default: + break; + + } + } + return 0; + +exit: + if (i > 0) + for (i = i - 1; i >= 0; i--) + switch (pdata->leds[i].type) { + + case LP3944_LED_TYPE_LED: + case LP3944_LED_TYPE_LED_INVERTED: + led_classdev_unregister(&data->leds[i].ldev); + cancel_work_sync(&data->leds[i].work); + break; + + case LP3944_LED_TYPE_NONE: + default: + break; + } + + return err; +} + +static int __devinit lp3944_probe(struct i2c_client *client, + const struct i2c_device_id *id) +{ + struct lp3944_platform_data *lp3944_pdata = client->dev.platform_data; + struct lp3944_data *data; + + if (lp3944_pdata == NULL) { + dev_err(&client->dev, "no platform data\n"); + return -EINVAL; + } + + /* Let's see whether this adapter can support what we need. */ + if (!i2c_check_functionality(client->adapter, + I2C_FUNC_SMBUS_BYTE_DATA)) { + dev_err(&client->dev, "insufficient functionality!\n"); + return -ENODEV; + } + + data = kzalloc(sizeof(struct lp3944_data), GFP_KERNEL); + if (!data) + return -ENOMEM; + + data->client = client; + i2c_set_clientdata(client, data); + + mutex_init(&data->lock); + + dev_info(&client->dev, "lp3944 enabled\n"); + + lp3944_configure(client, data, lp3944_pdata); + return 0; +} + +static int __devexit lp3944_remove(struct i2c_client *client) +{ + struct lp3944_platform_data *pdata = client->dev.platform_data; + struct lp3944_data *data = i2c_get_clientdata(client); + int i; + + for (i = 0; i < pdata->leds_size; i++) + switch (data->leds[i].type) { + case LP3944_LED_TYPE_LED: + case LP3944_LED_TYPE_LED_INVERTED: + led_classdev_unregister(&data->leds[i].ldev); + cancel_work_sync(&data->leds[i].work); + break; + + case LP3944_LED_TYPE_NONE: + default: + break; + } + + kfree(data); + i2c_set_clientdata(client, NULL); + + return 0; +} + +/* lp3944 i2c driver struct */ +static const struct i2c_device_id lp3944_id[] = { + {"lp3944", 0}, + {} +}; + +MODULE_DEVICE_TABLE(i2c, lp3944_id); + +static struct i2c_driver lp3944_driver = { + .driver = { + .name = "lp3944", + }, + .probe = lp3944_probe, + .remove = __devexit_p(lp3944_remove), + .id_table = lp3944_id, +}; + +static int __init lp3944_module_init(void) +{ + return i2c_add_driver(&lp3944_driver); +} + +static void __exit lp3944_module_exit(void) +{ + i2c_del_driver(&lp3944_driver); +} + +module_init(lp3944_module_init); +module_exit(lp3944_module_exit); + +MODULE_AUTHOR("Antonio Ospite "); +MODULE_DESCRIPTION("LP3944 Fun Light Chip"); +MODULE_LICENSE("GPL"); diff --git a/include/linux/leds-lp3944.h b/include/linux/leds-lp3944.h new file mode 100644 index 000000000000..afc9f9fd70f5 --- /dev/null +++ b/include/linux/leds-lp3944.h @@ -0,0 +1,53 @@ +/* + * leds-lp3944.h - platform data structure for lp3944 led controller + * + * Copyright (C) 2009 Antonio Ospite + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + */ + +#ifndef __LINUX_LEDS_LP3944_H +#define __LINUX_LEDS_LP3944_H + +#include +#include + +#define LP3944_LED0 0 +#define LP3944_LED1 1 +#define LP3944_LED2 2 +#define LP3944_LED3 3 +#define LP3944_LED4 4 +#define LP3944_LED5 5 +#define LP3944_LED6 6 +#define LP3944_LED7 7 +#define LP3944_LEDS_MAX 8 + +#define LP3944_LED_STATUS_MASK 0x03 +enum lp3944_status { + LP3944_LED_STATUS_OFF = 0x0, + LP3944_LED_STATUS_ON = 0x1, + LP3944_LED_STATUS_DIM0 = 0x2, + LP3944_LED_STATUS_DIM1 = 0x3 +}; + +enum lp3944_type { + LP3944_LED_TYPE_NONE, + LP3944_LED_TYPE_LED, + LP3944_LED_TYPE_LED_INVERTED, +}; + +struct lp3944_led { + char *name; + enum lp3944_type type; + enum lp3944_status status; +}; + +struct lp3944_platform_data { + struct lp3944_led leds[LP3944_LEDS_MAX]; + u8 leds_size; +}; + +#endif /* __LINUX_LEDS_LP3944_H */ -- cgit v1.2.3 From ed88bae6918fa990cbfe47316bd0f790121aaf00 Mon Sep 17 00:00:00 2001 From: Trent Piepho Date: Tue, 12 May 2009 15:33:12 -0700 Subject: leds: Add options to have GPIO LEDs start on or keep their state There already is a "default-on" trigger but there are problems with it. For one, it's a inefficient way to do it and requires led trigger support to be compiled in. But the real reason is that is produces a glitch on the LED. The GPIO is allocate with the LED *off*, then *later* when the trigger runs it is turned back on. If the LED was already on via the GPIO's reset default or action of the firmware, this produces a glitch where the LED goes from on to off to on. While normally this is fast enough that it wouldn't be noticeable to a human observer, there are still serious problems. One is that there may be something else on the GPIO line, like a hardware alarm or watchdog, that is fast enough to notice the glitch. Another is that the kernel may panic before the LED is turned back on, thus hanging with the LED in the wrong state. This is not just speculation, but actually happened to me with an embedded system that has an LED which should turn off when the kernel finishes booting, which was left in the incorrect state due to a bug in the OF LED binding code. We also let GPIO LEDs get their initial value from whatever the current state of the GPIO line is. On some systems the LEDs are put into some state by the firmware or hardware before Linux boots, and it is desired to have them keep this state which is otherwise unknown to Linux. This requires that the underlying GPIO driver support reading the value of output GPIOs. Some drivers support this and some do not. The platform device binding gains a field in the platform data "default_state" that controls this. There are three constants defined to select from on, off, or keeping the current state. The OpenFirmware binding uses a property named "default-state" that can be set to "on", "off", or "keep". The default if the property isn't present is off. Signed-off-by: Trent Piepho Acked-by: Grant Likely Acked-by: Wolfram Sang Acked-by: Sean MacLennan Signed-off-by: Richard Purdie --- Documentation/powerpc/dts-bindings/gpio/led.txt | 17 ++++++++++++++++- drivers/leds/leds-gpio.c | 20 +++++++++++++++++--- include/linux/leds.h | 9 +++++++-- 3 files changed, 40 insertions(+), 6 deletions(-) (limited to 'include') diff --git a/Documentation/powerpc/dts-bindings/gpio/led.txt b/Documentation/powerpc/dts-bindings/gpio/led.txt index 4fe14deedc0a..064db928c3c1 100644 --- a/Documentation/powerpc/dts-bindings/gpio/led.txt +++ b/Documentation/powerpc/dts-bindings/gpio/led.txt @@ -16,10 +16,17 @@ LED sub-node properties: string defining the trigger assigned to the LED. Current triggers are: "backlight" - LED will act as a back-light, controlled by the framebuffer system - "default-on" - LED will turn on + "default-on" - LED will turn on, but see "default-state" below "heartbeat" - LED "double" flashes at a load average based rate "ide-disk" - LED indicates disk activity "timer" - LED flashes at a fixed, configurable rate +- default-state: (optional) The initial state of the LED. Valid + values are "on", "off", and "keep". If the LED is already on or off + and the default-state property is set the to same value, then no + glitch should be produced where the LED momentarily turns off (or + on). The "keep" setting will keep the LED at whatever its current + state is, without producing a glitch. The default is off if this + property is not present. Examples: @@ -30,14 +37,22 @@ leds { gpios = <&mcu_pio 0 1>; /* Active low */ linux,default-trigger = "ide-disk"; }; + + fault { + gpios = <&mcu_pio 1 0>; + /* Keep LED on if BIOS detected hardware fault */ + default-state = "keep"; + }; }; run-control { compatible = "gpio-leds"; red { gpios = <&mpc8572 6 0>; + default-state = "off"; }; green { gpios = <&mpc8572 7 0>; + default-state = "on"; }; } diff --git a/drivers/leds/leds-gpio.c b/drivers/leds/leds-gpio.c index 76895e691042..6b06638eb5b4 100644 --- a/drivers/leds/leds-gpio.c +++ b/drivers/leds/leds-gpio.c @@ -76,7 +76,7 @@ static int __devinit create_gpio_led(const struct gpio_led *template, struct gpio_led_data *led_dat, struct device *parent, int (*blink_set)(unsigned, unsigned long *, unsigned long *)) { - int ret; + int ret, state; /* skip leds that aren't available */ if (!gpio_is_valid(template->gpio)) { @@ -99,11 +99,15 @@ static int __devinit create_gpio_led(const struct gpio_led *template, led_dat->cdev.blink_set = gpio_blink_set; } led_dat->cdev.brightness_set = gpio_led_set; - led_dat->cdev.brightness = LED_OFF; + if (template->default_state == LEDS_GPIO_DEFSTATE_KEEP) + state = !!gpio_get_value(led_dat->gpio) ^ led_dat->active_low; + else + state = (template->default_state == LEDS_GPIO_DEFSTATE_ON); + led_dat->cdev.brightness = state ? LED_FULL : LED_OFF; if (!template->retain_state_suspended) led_dat->cdev.flags |= LED_CORE_SUSPENDRESUME; - ret = gpio_direction_output(led_dat->gpio, led_dat->active_low); + ret = gpio_direction_output(led_dat->gpio, led_dat->active_low ^ state); if (ret < 0) goto err; @@ -223,12 +227,22 @@ static int __devinit of_gpio_leds_probe(struct of_device *ofdev, memset(&led, 0, sizeof(led)); for_each_child_of_node(np, child) { enum of_gpio_flags flags; + const char *state; led.gpio = of_get_gpio_flags(child, 0, &flags); led.active_low = flags & OF_GPIO_ACTIVE_LOW; led.name = of_get_property(child, "label", NULL) ? : child->name; led.default_trigger = of_get_property(child, "linux,default-trigger", NULL); + state = of_get_property(child, "default-state", NULL); + if (state) { + if (!strcmp(state, "keep")) + led.default_state = LEDS_GPIO_DEFSTATE_KEEP; + else if(!strcmp(state, "on")) + led.default_state = LEDS_GPIO_DEFSTATE_ON; + else + led.default_state = LEDS_GPIO_DEFSTATE_OFF; + } ret = create_gpio_led(&led, &pdata->led_data[pdata->num_leds++], &ofdev->dev, NULL); diff --git a/include/linux/leds.h b/include/linux/leds.h index c7f0b148df06..62af62915cf7 100644 --- a/include/linux/leds.h +++ b/include/linux/leds.h @@ -143,9 +143,14 @@ struct gpio_led { const char *name; const char *default_trigger; unsigned gpio; - u8 active_low : 1; - u8 retain_state_suspended : 1; + unsigned active_low : 1; + unsigned retain_state_suspended : 1; + unsigned default_state : 2; + /* default_state should be one of LEDS_GPIO_DEFSTATE_(ON|OFF|KEEP) */ }; +#define LEDS_GPIO_DEFSTATE_OFF 0 +#define LEDS_GPIO_DEFSTATE_ON 1 +#define LEDS_GPIO_DEFSTATE_KEEP 2 struct gpio_led_platform_data { int num_leds; -- cgit v1.2.3 From a1dd8c617217322614f0465ae347895c4b58e1ab Mon Sep 17 00:00:00 2001 From: Richard Purdie Date: Mon, 22 Jun 2009 14:54:13 +0100 Subject: leds: Futher document blink_set Futher document blink_set function pointer Signed-off-by: Richard Purdie --- include/linux/leds.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/leds.h b/include/linux/leds.h index 62af62915cf7..d8bf9665e70c 100644 --- a/include/linux/leds.h +++ b/include/linux/leds.h @@ -47,7 +47,8 @@ struct led_classdev { /* Activate hardware accelerated blink, delays are in * miliseconds and if none is provided then a sensible default - * should be chosen. */ + * should be chosen. The call can adjust the timings if it can't + * match the values specified exactly. */ int (*blink_set)(struct led_classdev *led_cdev, unsigned long *delay_on, unsigned long *delay_off); -- cgit v1.2.3 From e14cbee401cd00779a5267128371506b22c77bc9 Mon Sep 17 00:00:00 2001 From: Michel Dänzer Date: Tue, 23 Jun 2009 12:36:32 +0200 Subject: drm: Fix shifts which were miscalculated when converting from bitfields. MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Looks like I managed to mess up most shifts when converting from bitfields. :( The patch below works on my Thinkpad T500 (as well as on my PowerBook, where the previous change worked as well, maybe out of luck...). I'd appreciate more testing and eyes looking over it though. Signed-off-by: Michel Dänzer Tested-by: Michael Pyne Signed-off-by: Dave Airlie --- drivers/gpu/drm/drm_edid.c | 12 ++++++------ include/drm/drm_edid.h | 38 +++++++++++++++++++------------------- 2 files changed, 25 insertions(+), 25 deletions(-) (limited to 'include') diff --git a/drivers/gpu/drm/drm_edid.c b/drivers/gpu/drm/drm_edid.c index 7d0835226f6e..80cc6d06d61b 100644 --- a/drivers/gpu/drm/drm_edid.c +++ b/drivers/gpu/drm/drm_edid.c @@ -294,10 +294,10 @@ static struct drm_display_mode *drm_mode_detailed(struct drm_device *dev, unsigned vactive = (pt->vactive_vblank_hi & 0xf0) << 4 | pt->vactive_lo; unsigned hblank = (pt->hactive_hblank_hi & 0xf) << 8 | pt->hblank_lo; unsigned vblank = (pt->vactive_vblank_hi & 0xf) << 8 | pt->vblank_lo; - unsigned hsync_offset = (pt->hsync_vsync_offset_pulse_width_hi & 0x3) << 8 | pt->hsync_offset_lo; - unsigned hsync_pulse_width = (pt->hsync_vsync_offset_pulse_width_hi & 0xc) << 6 | pt->hsync_pulse_width_lo; - unsigned vsync_offset = (pt->hsync_vsync_offset_pulse_width_hi & 0x30) | (pt->vsync_offset_pulse_width_lo & 0xf); - unsigned vsync_pulse_width = (pt->hsync_vsync_offset_pulse_width_hi & 0xc0) >> 2 | pt->vsync_offset_pulse_width_lo >> 4; + unsigned hsync_offset = (pt->hsync_vsync_offset_pulse_width_hi & 0xc0) << 2 | pt->hsync_offset_lo; + unsigned hsync_pulse_width = (pt->hsync_vsync_offset_pulse_width_hi & 0x30) << 4 | pt->hsync_pulse_width_lo; + unsigned vsync_offset = (pt->hsync_vsync_offset_pulse_width_hi & 0xc) >> 2 | pt->vsync_offset_pulse_width_lo >> 4; + unsigned vsync_pulse_width = (pt->hsync_vsync_offset_pulse_width_hi & 0x3) << 4 | (pt->vsync_offset_pulse_width_lo & 0xf); /* ignore tiny modes */ if (hactive < 64 || vactive < 64) @@ -347,8 +347,8 @@ static struct drm_display_mode *drm_mode_detailed(struct drm_device *dev, mode->flags |= (pt->misc & DRM_EDID_PT_VSYNC_POSITIVE) ? DRM_MODE_FLAG_PVSYNC : DRM_MODE_FLAG_NVSYNC; - mode->width_mm = pt->width_mm_lo | (pt->width_height_mm_hi & 0xf) << 8; - mode->height_mm = pt->height_mm_lo | (pt->width_height_mm_hi & 0xf0) << 4; + mode->width_mm = pt->width_mm_lo | (pt->width_height_mm_hi & 0xf0) << 4; + mode->height_mm = pt->height_mm_lo | (pt->width_height_mm_hi & 0xf) << 8; if (quirks & EDID_QUIRK_DETAILED_IN_CM) { mode->width_mm *= 10; diff --git a/include/drm/drm_edid.h b/include/drm/drm_edid.h index c263e4d71754..7d6c9a2dfcbb 100644 --- a/include/drm/drm_edid.h +++ b/include/drm/drm_edid.h @@ -35,11 +35,11 @@ struct est_timings { } __attribute__((packed)); /* 00=16:10, 01=4:3, 10=5:4, 11=16:9 */ -#define EDID_TIMING_ASPECT_SHIFT 0 +#define EDID_TIMING_ASPECT_SHIFT 6 #define EDID_TIMING_ASPECT_MASK (0x3 << EDID_TIMING_ASPECT_SHIFT) /* need to add 60 */ -#define EDID_TIMING_VFREQ_SHIFT 2 +#define EDID_TIMING_VFREQ_SHIFT 0 #define EDID_TIMING_VFREQ_MASK (0x3f << EDID_TIMING_VFREQ_SHIFT) struct std_timing { @@ -47,11 +47,11 @@ struct std_timing { u8 vfreq_aspect; } __attribute__((packed)); -#define DRM_EDID_PT_HSYNC_POSITIVE (1 << 6) -#define DRM_EDID_PT_VSYNC_POSITIVE (1 << 5) +#define DRM_EDID_PT_HSYNC_POSITIVE (1 << 1) +#define DRM_EDID_PT_VSYNC_POSITIVE (1 << 2) #define DRM_EDID_PT_SEPARATE_SYNC (3 << 3) -#define DRM_EDID_PT_STEREO (1 << 2) -#define DRM_EDID_PT_INTERLACED (1 << 1) +#define DRM_EDID_PT_STEREO (1 << 5) +#define DRM_EDID_PT_INTERLACED (1 << 7) /* If detailed data is pixel timing */ struct detailed_pixel_timing { @@ -93,7 +93,7 @@ struct detailed_data_monitor_range { } __attribute__((packed)); struct detailed_data_wpindex { - u8 white_xy_lo; /* Upper 2 bits each */ + u8 white_yx_lo; /* Lower 2 bits each */ u8 white_x_hi; u8 white_y_hi; u8 gamma; /* need to divide by 100 then add 1 */ @@ -135,21 +135,21 @@ struct detailed_timing { } data; } __attribute__((packed)); -#define DRM_EDID_INPUT_SERRATION_VSYNC (1 << 7) -#define DRM_EDID_INPUT_SYNC_ON_GREEN (1 << 5) -#define DRM_EDID_INPUT_COMPOSITE_SYNC (1 << 4) +#define DRM_EDID_INPUT_SERRATION_VSYNC (1 << 0) +#define DRM_EDID_INPUT_SYNC_ON_GREEN (1 << 1) +#define DRM_EDID_INPUT_COMPOSITE_SYNC (1 << 2) #define DRM_EDID_INPUT_SEPARATE_SYNCS (1 << 3) -#define DRM_EDID_INPUT_BLANK_TO_BLACK (1 << 2) -#define DRM_EDID_INPUT_VIDEO_LEVEL (3 << 1) -#define DRM_EDID_INPUT_DIGITAL (1 << 0) /* bits above must be zero if set */ +#define DRM_EDID_INPUT_BLANK_TO_BLACK (1 << 4) +#define DRM_EDID_INPUT_VIDEO_LEVEL (3 << 5) +#define DRM_EDID_INPUT_DIGITAL (1 << 7) /* bits below must be zero if set */ -#define DRM_EDID_FEATURE_DEFAULT_GTF (1 << 7) -#define DRM_EDID_FEATURE_PREFERRED_TIMING (1 << 6) -#define DRM_EDID_FEATURE_STANDARD_COLOR (1 << 5) +#define DRM_EDID_FEATURE_DEFAULT_GTF (1 << 0) +#define DRM_EDID_FEATURE_PREFERRED_TIMING (1 << 1) +#define DRM_EDID_FEATURE_STANDARD_COLOR (1 << 2) #define DRM_EDID_FEATURE_DISPLAY_TYPE (3 << 3) /* 00=mono, 01=rgb, 10=non-rgb, 11=unknown */ -#define DRM_EDID_FEATURE_PM_ACTIVE_OFF (1 << 2) -#define DRM_EDID_FEATURE_PM_SUSPEND (1 << 1) -#define DRM_EDID_FEATURE_PM_STANDBY (1 << 0) +#define DRM_EDID_FEATURE_PM_ACTIVE_OFF (1 << 5) +#define DRM_EDID_FEATURE_PM_SUSPEND (1 << 6) +#define DRM_EDID_FEATURE_PM_STANDBY (1 << 7) struct edid { u8 header[8]; -- cgit v1.2.3 From e74e396204bfcb67570ba4517b08f5918e69afea Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Mon, 30 Mar 2009 19:07:44 +0900 Subject: percpu: use dynamic percpu allocator as the default percpu allocator This patch makes most !CONFIG_HAVE_SETUP_PER_CPU_AREA archs use dynamic percpu allocator. The first chunk is allocated using embedding helper and 8k is reserved for modules. This ensures that the new allocator behaves almost identically to the original allocator as long as static percpu variables are concerned, so it shouldn't introduce much breakage. s390 and alpha use custom SHIFT_PERCPU_PTR() to work around addressing range limit the addressing model imposes. Unfortunately, this breaks if the address is specified using a variable, so for now, the two archs aren't converted. The following architectures are affected by this change. * sh * arm * cris * mips * sparc(32) * blackfin * avr32 * parisc (broken, under investigation) * m32r * powerpc(32) As this change makes the dynamic allocator the default one, CONFIG_HAVE_DYNAMIC_PER_CPU_AREA is replaced with its invert - CONFIG_HAVE_LEGACY_PER_CPU_AREA, which is added to yet-to-be converted archs. These archs implement their own setup_per_cpu_areas() and the conversion is not trivial. * powerpc(64) * sparc(64) * ia64 * alpha * s390 Boot and batch alloc/free tests on x86_32 with debug code (x86_32 doesn't use default first chunk initialization). Compile tested on sparc(32), powerpc(32), arm and alpha. Kyle McMartin reported that this change breaks parisc. The problem is still under investigation and he is okay with pushing this patch forward and fixing parisc later. [ Impact: use dynamic allocator for most archs w/o custom percpu setup ] Signed-off-by: Tejun Heo Acked-by: Rusty Russell Acked-by: David S. Miller Acked-by: Benjamin Herrenschmidt Acked-by: Martin Schwidefsky Reviewed-by: Christoph Lameter Cc: Paul Mundt Cc: Russell King Cc: Mikael Starvik Cc: Ralf Baechle Cc: Bryan Wu Cc: Kyle McMartin Cc: Matthew Wilcox Cc: Grant Grundler Cc: Hirokazu Takata Cc: Richard Henderson Cc: Ivan Kokshaysky Cc: Heiko Carstens Cc: Ingo Molnar --- arch/alpha/Kconfig | 3 +++ arch/ia64/Kconfig | 3 +++ arch/powerpc/Kconfig | 3 +++ arch/s390/Kconfig | 3 +++ arch/sparc/Kconfig | 3 +++ arch/x86/Kconfig | 3 --- include/linux/percpu.h | 12 +++++++++--- init/main.c | 24 ------------------------ kernel/module.c | 6 +++--- mm/Makefile | 2 +- mm/allocpercpu.c | 28 ++++++++++++++++++++++++++++ mm/percpu.c | 40 +++++++++++++++++++++++++++++++++++++++- 12 files changed, 95 insertions(+), 35 deletions(-) (limited to 'include') diff --git a/arch/alpha/Kconfig b/arch/alpha/Kconfig index 9fb8aae5c391..05d86407188c 100644 --- a/arch/alpha/Kconfig +++ b/arch/alpha/Kconfig @@ -70,6 +70,9 @@ config AUTO_IRQ_AFFINITY depends on SMP default y +config HAVE_LEGACY_PER_CPU_AREA + def_bool y + source "init/Kconfig" source "kernel/Kconfig.freezer" diff --git a/arch/ia64/Kconfig b/arch/ia64/Kconfig index 170042b420d4..328d2f8b8c3f 100644 --- a/arch/ia64/Kconfig +++ b/arch/ia64/Kconfig @@ -89,6 +89,9 @@ config GENERIC_TIME_VSYSCALL bool default y +config HAVE_LEGACY_PER_CPU_AREA + def_bool y + config HAVE_SETUP_PER_CPU_AREA def_bool y diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig index bf6cedfa05db..a774c2acbe69 100644 --- a/arch/powerpc/Kconfig +++ b/arch/powerpc/Kconfig @@ -46,6 +46,9 @@ config GENERIC_HARDIRQS_NO__DO_IRQ bool default y +config HAVE_LEGACY_PER_CPU_AREA + def_bool PPC64 + config HAVE_SETUP_PER_CPU_AREA def_bool PPC64 diff --git a/arch/s390/Kconfig b/arch/s390/Kconfig index a14dba0e4d67..f4a3cc62d28f 100644 --- a/arch/s390/Kconfig +++ b/arch/s390/Kconfig @@ -75,6 +75,9 @@ config VIRT_CPU_ACCOUNTING config ARCH_SUPPORTS_DEBUG_PAGEALLOC def_bool y +config HAVE_LEGACY_PER_CPU_AREA + def_bool y + mainmenu "Linux Kernel Configuration" config S390 diff --git a/arch/sparc/Kconfig b/arch/sparc/Kconfig index 3f8b6a92eabd..7a8698b913fe 100644 --- a/arch/sparc/Kconfig +++ b/arch/sparc/Kconfig @@ -92,6 +92,9 @@ config AUDIT_ARCH bool default y +config HAVE_LEGACY_PER_CPU_AREA + def_bool y if SPARC64 + config HAVE_SETUP_PER_CPU_AREA def_bool y if SPARC64 diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index d1430ef6b4f9..a48a90076d83 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -149,9 +149,6 @@ config ARCH_HAS_CACHE_LINE_SIZE config HAVE_SETUP_PER_CPU_AREA def_bool y -config HAVE_DYNAMIC_PER_CPU_AREA - def_bool y - config HAVE_CPUMASK_OF_CPU_MAP def_bool X86_64_SMP diff --git a/include/linux/percpu.h b/include/linux/percpu.h index 26fd9d12f050..e5000343dd61 100644 --- a/include/linux/percpu.h +++ b/include/linux/percpu.h @@ -34,7 +34,7 @@ #ifdef CONFIG_SMP -#ifdef CONFIG_HAVE_DYNAMIC_PER_CPU_AREA +#ifndef CONFIG_HAVE_LEGACY_PER_CPU_AREA /* minimum unit size, also is the maximum supported allocation size */ #define PCPU_MIN_UNIT_SIZE PFN_ALIGN(64 << 10) @@ -80,7 +80,7 @@ extern ssize_t __init pcpu_embed_first_chunk( extern void *__alloc_reserved_percpu(size_t size, size_t align); -#else /* CONFIG_HAVE_DYNAMIC_PER_CPU_AREA */ +#else /* CONFIG_HAVE_LEGACY_PER_CPU_AREA */ struct percpu_data { void *ptrs[1]; @@ -99,11 +99,15 @@ struct percpu_data { (__typeof__(ptr))__p->ptrs[(cpu)]; \ }) -#endif /* CONFIG_HAVE_DYNAMIC_PER_CPU_AREA */ +#endif /* CONFIG_HAVE_LEGACY_PER_CPU_AREA */ extern void *__alloc_percpu(size_t size, size_t align); extern void free_percpu(void *__pdata); +#ifndef CONFIG_HAVE_SETUP_PER_CPU_AREA +extern void __init setup_per_cpu_areas(void); +#endif + #else /* CONFIG_SMP */ #define per_cpu_ptr(ptr, cpu) ({ (void)(cpu); (ptr); }) @@ -124,6 +128,8 @@ static inline void free_percpu(void *p) kfree(p); } +static inline void __init setup_per_cpu_areas(void) { } + #endif /* CONFIG_SMP */ #define alloc_percpu(type) (type *)__alloc_percpu(sizeof(type), \ diff --git a/init/main.c b/init/main.c index 09131ec090c1..602d724afa5c 100644 --- a/init/main.c +++ b/init/main.c @@ -357,7 +357,6 @@ static void __init smp_init(void) #define smp_init() do { } while (0) #endif -static inline void setup_per_cpu_areas(void) { } static inline void setup_nr_cpu_ids(void) { } static inline void smp_prepare_cpus(unsigned int maxcpus) { } @@ -378,29 +377,6 @@ static void __init setup_nr_cpu_ids(void) nr_cpu_ids = find_last_bit(cpumask_bits(cpu_possible_mask),NR_CPUS) + 1; } -#ifndef CONFIG_HAVE_SETUP_PER_CPU_AREA -unsigned long __per_cpu_offset[NR_CPUS] __read_mostly; - -EXPORT_SYMBOL(__per_cpu_offset); - -static void __init setup_per_cpu_areas(void) -{ - unsigned long size, i; - char *ptr; - unsigned long nr_possible_cpus = num_possible_cpus(); - - /* Copy section for each CPU (we discard the original) */ - size = ALIGN(PERCPU_ENOUGH_ROOM, PAGE_SIZE); - ptr = alloc_bootmem_pages(size * nr_possible_cpus); - - for_each_possible_cpu(i) { - __per_cpu_offset[i] = ptr - __per_cpu_start; - memcpy(ptr, __per_cpu_start, __per_cpu_end - __per_cpu_start); - ptr += size; - } -} -#endif /* CONFIG_HAVE_SETUP_PER_CPU_AREA */ - /* Called by boot processor to activate the rest. */ static void __init smp_init(void) { diff --git a/kernel/module.c b/kernel/module.c index 38928fcaff2b..f5934954fa99 100644 --- a/kernel/module.c +++ b/kernel/module.c @@ -364,7 +364,7 @@ EXPORT_SYMBOL_GPL(find_module); #ifdef CONFIG_SMP -#ifdef CONFIG_HAVE_DYNAMIC_PER_CPU_AREA +#ifndef CONFIG_HAVE_LEGACY_PER_CPU_AREA static void *percpu_modalloc(unsigned long size, unsigned long align, const char *name) @@ -389,7 +389,7 @@ static void percpu_modfree(void *freeme) free_percpu(freeme); } -#else /* ... !CONFIG_HAVE_DYNAMIC_PER_CPU_AREA */ +#else /* ... CONFIG_HAVE_LEGACY_PER_CPU_AREA */ /* Number of blocks used and allocated. */ static unsigned int pcpu_num_used, pcpu_num_allocated; @@ -535,7 +535,7 @@ static int percpu_modinit(void) } __initcall(percpu_modinit); -#endif /* CONFIG_HAVE_DYNAMIC_PER_CPU_AREA */ +#endif /* CONFIG_HAVE_LEGACY_PER_CPU_AREA */ static unsigned int find_pcpusec(Elf_Ehdr *hdr, Elf_Shdr *sechdrs, diff --git a/mm/Makefile b/mm/Makefile index 5e0bd6426693..c77c6487552f 100644 --- a/mm/Makefile +++ b/mm/Makefile @@ -33,7 +33,7 @@ obj-$(CONFIG_FAILSLAB) += failslab.o obj-$(CONFIG_MEMORY_HOTPLUG) += memory_hotplug.o obj-$(CONFIG_FS_XIP) += filemap_xip.o obj-$(CONFIG_MIGRATION) += migrate.o -ifdef CONFIG_HAVE_DYNAMIC_PER_CPU_AREA +ifndef CONFIG_HAVE_LEGACY_PER_CPU_AREA obj-$(CONFIG_SMP) += percpu.o else obj-$(CONFIG_SMP) += allocpercpu.o diff --git a/mm/allocpercpu.c b/mm/allocpercpu.c index dfdee6a47359..df34ceae0c67 100644 --- a/mm/allocpercpu.c +++ b/mm/allocpercpu.c @@ -5,6 +5,8 @@ */ #include #include +#include +#include #ifndef cache_line_size #define cache_line_size() L1_CACHE_BYTES @@ -147,3 +149,29 @@ void free_percpu(void *__pdata) kfree(__percpu_disguise(__pdata)); } EXPORT_SYMBOL_GPL(free_percpu); + +/* + * Generic percpu area setup. + */ +#ifndef CONFIG_HAVE_SETUP_PER_CPU_AREA +unsigned long __per_cpu_offset[NR_CPUS] __read_mostly; + +EXPORT_SYMBOL(__per_cpu_offset); + +void __init setup_per_cpu_areas(void) +{ + unsigned long size, i; + char *ptr; + unsigned long nr_possible_cpus = num_possible_cpus(); + + /* Copy section for each CPU (we discard the original) */ + size = ALIGN(PERCPU_ENOUGH_ROOM, PAGE_SIZE); + ptr = alloc_bootmem_pages(size * nr_possible_cpus); + + for_each_possible_cpu(i) { + __per_cpu_offset[i] = ptr - __per_cpu_start; + memcpy(ptr, __per_cpu_start, __per_cpu_end - __per_cpu_start); + ptr += size; + } +} +#endif /* CONFIG_HAVE_SETUP_PER_CPU_AREA */ diff --git a/mm/percpu.c b/mm/percpu.c index b70f2acd8853..b14984566f5a 100644 --- a/mm/percpu.c +++ b/mm/percpu.c @@ -43,7 +43,7 @@ * * To use this allocator, arch code should do the followings. * - * - define CONFIG_HAVE_DYNAMIC_PER_CPU_AREA + * - drop CONFIG_HAVE_LEGACY_PER_CPU_AREA * * - define __addr_to_pcpu_ptr() and __pcpu_ptr_to_addr() to translate * regular address to percpu pointer and back if they need to be @@ -1275,3 +1275,41 @@ ssize_t __init pcpu_embed_first_chunk(size_t static_size, size_t reserved_size, reserved_size, dyn_size, pcpue_unit_size, pcpue_ptr, NULL); } + +/* + * Generic percpu area setup. + * + * The embedding helper is used because its behavior closely resembles + * the original non-dynamic generic percpu area setup. This is + * important because many archs have addressing restrictions and might + * fail if the percpu area is located far away from the previous + * location. As an added bonus, in non-NUMA cases, embedding is + * generally a good idea TLB-wise because percpu area can piggy back + * on the physical linear memory mapping which uses large page + * mappings on applicable archs. + */ +#ifndef CONFIG_HAVE_SETUP_PER_CPU_AREA +unsigned long __per_cpu_offset[NR_CPUS] __read_mostly; +EXPORT_SYMBOL(__per_cpu_offset); + +void __init setup_per_cpu_areas(void) +{ + size_t static_size = __per_cpu_end - __per_cpu_start; + ssize_t unit_size; + unsigned long delta; + unsigned int cpu; + + /* + * Always reserve area for module percpu variables. That's + * what the legacy allocator did. + */ + unit_size = pcpu_embed_first_chunk(static_size, PERCPU_MODULE_RESERVE, + PERCPU_DYNAMIC_RESERVE, -1); + if (unit_size < 0) + panic("Failed to initialized percpu areas."); + + delta = (unsigned long)pcpu_base_addr - (unsigned long)__per_cpu_start; + for_each_possible_cpu(cpu) + __per_cpu_offset[cpu] = delta + cpu * unit_size; +} +#endif /* CONFIG_HAVE_SETUP_PER_CPU_AREA */ -- cgit v1.2.3 From 405d967dc70002991f8fc35c20e0d3cbc7614f63 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Wed, 24 Jun 2009 15:13:38 +0900 Subject: linker script: throw away .discard section x86 throws away .discard section but no other archs do. Also, .discard is not thrown away while linking modules. Make every arch and module linking throw it away. This will be used to define dummy variables for percpu declarations and definitions. This patch is based on Ivan Kokshaysky's alpha percpu patch. [ Impact: always throw away everything in .discard ] Signed-off-by: Tejun Heo Cc: Ivan Kokshaysky Cc: Richard Henderson Cc: Russell King Cc: Haavard Skinnemoen Cc: Bryan Wu Cc: Mikael Starvik Cc: Jesper Nilsson Cc: David Howells Cc: Yoshinori Sato Cc: Tony Luck Cc: Hirokazu Takata Cc: Geert Uytterhoeven Cc: Michal Simek Cc: Ralf Baechle Cc: Kyle McMartin Cc: Benjamin Herrenschmidt Cc: Paul Mackerras Cc: Martin Schwidefsky Cc: Heiko Carstens Cc: Paul Mundt Cc: David S. Miller Cc: Jeff Dike Cc: Chris Zankel Cc: Rusty Russell Cc: Ingo Molnar --- Makefile | 2 +- arch/alpha/kernel/vmlinux.lds.S | 1 + arch/arm/kernel/vmlinux.lds.S | 1 + arch/avr32/kernel/vmlinux.lds.S | 1 + arch/blackfin/kernel/vmlinux.lds.S | 1 + arch/cris/kernel/vmlinux.lds.S | 1 + arch/frv/kernel/vmlinux.lds.S | 2 ++ arch/h8300/kernel/vmlinux.lds.S | 1 + arch/ia64/kernel/vmlinux.lds.S | 1 + arch/m32r/kernel/vmlinux.lds.S | 1 + arch/m68k/kernel/vmlinux-std.lds | 1 + arch/m68k/kernel/vmlinux-sun3.lds | 1 + arch/m68knommu/kernel/vmlinux.lds.S | 1 + arch/microblaze/kernel/vmlinux.lds.S | 2 ++ arch/mips/kernel/vmlinux.lds.S | 1 + arch/mn10300/kernel/vmlinux.lds.S | 1 + arch/parisc/kernel/vmlinux.lds.S | 1 + arch/powerpc/kernel/vmlinux.lds.S | 1 + arch/s390/kernel/vmlinux.lds.S | 1 + arch/sh/kernel/vmlinux.lds.S | 1 + arch/sparc/kernel/vmlinux.lds.S | 1 + arch/um/kernel/dyn.lds.S | 2 ++ arch/um/kernel/uml.lds.S | 2 ++ arch/xtensa/kernel/vmlinux.lds.S | 1 + include/asm-generic/vmlinux.lds.h | 8 ++++++++ scripts/module-common.lds | 8 ++++++++ 26 files changed, 44 insertions(+), 1 deletion(-) create mode 100644 scripts/module-common.lds (limited to 'include') diff --git a/Makefile b/Makefile index 46e1c9d03d51..12245be05122 100644 --- a/Makefile +++ b/Makefile @@ -327,7 +327,7 @@ CHECKFLAGS := -D__linux__ -Dlinux -D__STDC__ -Dunix -D__unix__ \ MODFLAGS = -DMODULE CFLAGS_MODULE = $(MODFLAGS) AFLAGS_MODULE = $(MODFLAGS) -LDFLAGS_MODULE = +LDFLAGS_MODULE = -T $(srctree)/scripts/module-common.lds CFLAGS_KERNEL = AFLAGS_KERNEL = CFLAGS_GCOV = -fprofile-arcs -ftest-coverage diff --git a/arch/alpha/kernel/vmlinux.lds.S b/arch/alpha/kernel/vmlinux.lds.S index b9d6568e5f7f..75fe1d6877e9 100644 --- a/arch/alpha/kernel/vmlinux.lds.S +++ b/arch/alpha/kernel/vmlinux.lds.S @@ -139,6 +139,7 @@ SECTIONS EXIT_TEXT EXIT_DATA *(.exitcall.exit) + *(.discard) } .mdebug 0 : { diff --git a/arch/arm/kernel/vmlinux.lds.S b/arch/arm/kernel/vmlinux.lds.S index 6c0779792546..e256c57b8981 100644 --- a/arch/arm/kernel/vmlinux.lds.S +++ b/arch/arm/kernel/vmlinux.lds.S @@ -82,6 +82,7 @@ SECTIONS EXIT_TEXT EXIT_DATA *(.exitcall.exit) + *(.discard) *(.ARM.exidx.exit.text) *(.ARM.extab.exit.text) #ifndef CONFIG_MMU diff --git a/arch/avr32/kernel/vmlinux.lds.S b/arch/avr32/kernel/vmlinux.lds.S index 7910d41eb886..b8324608ec0c 100644 --- a/arch/avr32/kernel/vmlinux.lds.S +++ b/arch/avr32/kernel/vmlinux.lds.S @@ -131,6 +131,7 @@ SECTIONS /DISCARD/ : { EXIT_DATA *(.exitcall.exit) + *(.discard) } DWARF_DEBUG diff --git a/arch/blackfin/kernel/vmlinux.lds.S b/arch/blackfin/kernel/vmlinux.lds.S index 6ac307ca0d80..6e8eabd8f0a6 100644 --- a/arch/blackfin/kernel/vmlinux.lds.S +++ b/arch/blackfin/kernel/vmlinux.lds.S @@ -280,5 +280,6 @@ SECTIONS /DISCARD/ : { *(.exitcall.exit) + *(.discard) } } diff --git a/arch/cris/kernel/vmlinux.lds.S b/arch/cris/kernel/vmlinux.lds.S index 0d2adfc794d4..a3175ebb38cc 100644 --- a/arch/cris/kernel/vmlinux.lds.S +++ b/arch/cris/kernel/vmlinux.lds.S @@ -145,6 +145,7 @@ SECTIONS EXIT_TEXT EXIT_DATA *(.exitcall.exit) + *(.discard) } dram_end = dram_start + (CONFIG_ETRAX_DRAM_SIZE - __CONFIG_ETRAX_VMEM_SIZE)*1024*1024; diff --git a/arch/frv/kernel/vmlinux.lds.S b/arch/frv/kernel/vmlinux.lds.S index 22d9787406ed..64b5a5e4d35e 100644 --- a/arch/frv/kernel/vmlinux.lds.S +++ b/arch/frv/kernel/vmlinux.lds.S @@ -177,6 +177,8 @@ SECTIONS .debug_ranges 0 : { *(.debug_ranges) } .comment 0 : { *(.comment) } + + /DISCARD/ : { *(.discard) } } __kernel_image_size_no_bss = __bss_start - __kernel_image_start; diff --git a/arch/h8300/kernel/vmlinux.lds.S b/arch/h8300/kernel/vmlinux.lds.S index 43a87b9085b6..03d6c0df33db 100644 --- a/arch/h8300/kernel/vmlinux.lds.S +++ b/arch/h8300/kernel/vmlinux.lds.S @@ -154,6 +154,7 @@ SECTIONS } /DISCARD/ : { *(.exitcall.exit) + *(.discard) } .romfs : { diff --git a/arch/ia64/kernel/vmlinux.lds.S b/arch/ia64/kernel/vmlinux.lds.S index 4a95e86b9ac2..13d958975874 100644 --- a/arch/ia64/kernel/vmlinux.lds.S +++ b/arch/ia64/kernel/vmlinux.lds.S @@ -29,6 +29,7 @@ SECTIONS EXIT_TEXT EXIT_DATA *(.exitcall.exit) + *(.discard) *(.IA_64.unwind.exit.text) *(.IA_64.unwind_info.exit.text) } diff --git a/arch/m32r/kernel/vmlinux.lds.S b/arch/m32r/kernel/vmlinux.lds.S index 4179adf6c624..480a49944cfd 100644 --- a/arch/m32r/kernel/vmlinux.lds.S +++ b/arch/m32r/kernel/vmlinux.lds.S @@ -125,6 +125,7 @@ SECTIONS EXIT_TEXT EXIT_DATA *(.exitcall.exit) + *(.discard) } /* Stabs debugging sections. */ diff --git a/arch/m68k/kernel/vmlinux-std.lds b/arch/m68k/kernel/vmlinux-std.lds index 01d212bb05a6..905a797ada93 100644 --- a/arch/m68k/kernel/vmlinux-std.lds +++ b/arch/m68k/kernel/vmlinux-std.lds @@ -87,6 +87,7 @@ SECTIONS EXIT_TEXT EXIT_DATA *(.exitcall.exit) + *(.discard) } /* Stabs debugging sections. */ diff --git a/arch/m68k/kernel/vmlinux-sun3.lds b/arch/m68k/kernel/vmlinux-sun3.lds index c192f773db96..47d04be322aa 100644 --- a/arch/m68k/kernel/vmlinux-sun3.lds +++ b/arch/m68k/kernel/vmlinux-sun3.lds @@ -82,6 +82,7 @@ __init_begin = .; EXIT_TEXT EXIT_DATA *(.exitcall.exit) + *(.discard) } .crap : { diff --git a/arch/m68knommu/kernel/vmlinux.lds.S b/arch/m68knommu/kernel/vmlinux.lds.S index b7fe505e358d..68111a61a77f 100644 --- a/arch/m68knommu/kernel/vmlinux.lds.S +++ b/arch/m68knommu/kernel/vmlinux.lds.S @@ -188,6 +188,7 @@ SECTIONS { EXIT_TEXT EXIT_DATA *(.exitcall.exit) + *(.discard) } .bss : { diff --git a/arch/microblaze/kernel/vmlinux.lds.S b/arch/microblaze/kernel/vmlinux.lds.S index d34d38dcd12c..a207543c5927 100644 --- a/arch/microblaze/kernel/vmlinux.lds.S +++ b/arch/microblaze/kernel/vmlinux.lds.S @@ -162,4 +162,6 @@ SECTIONS { } . = ALIGN(4096); _end = .; + + /DISCARD/ : { *(.discard) } } diff --git a/arch/mips/kernel/vmlinux.lds.S b/arch/mips/kernel/vmlinux.lds.S index 58738c8d754f..45901609b741 100644 --- a/arch/mips/kernel/vmlinux.lds.S +++ b/arch/mips/kernel/vmlinux.lds.S @@ -179,6 +179,7 @@ SECTIONS /* Sections to be discarded */ /DISCARD/ : { *(.exitcall.exit) + *(.discard) /* ABI crap starts here */ *(.MIPS.options) diff --git a/arch/mn10300/kernel/vmlinux.lds.S b/arch/mn10300/kernel/vmlinux.lds.S index 24de6b90f401..5d9f2f96ad92 100644 --- a/arch/mn10300/kernel/vmlinux.lds.S +++ b/arch/mn10300/kernel/vmlinux.lds.S @@ -146,6 +146,7 @@ SECTIONS /* Sections to be discarded */ /DISCARD/ : { *(.exitcall.exit) + *(.discard) } STABS_DEBUG diff --git a/arch/parisc/kernel/vmlinux.lds.S b/arch/parisc/kernel/vmlinux.lds.S index fd2cc4fd2b65..ccf58341845a 100644 --- a/arch/parisc/kernel/vmlinux.lds.S +++ b/arch/parisc/kernel/vmlinux.lds.S @@ -240,6 +240,7 @@ SECTIONS /* Sections to be discarded */ /DISCARD/ : { *(.exitcall.exit) + *(.discard) #ifdef CONFIG_64BIT /* temporary hack until binutils is fixed to not emit these * for static binaries diff --git a/arch/powerpc/kernel/vmlinux.lds.S b/arch/powerpc/kernel/vmlinux.lds.S index 8ef8a14abc95..7fca9355fd3d 100644 --- a/arch/powerpc/kernel/vmlinux.lds.S +++ b/arch/powerpc/kernel/vmlinux.lds.S @@ -40,6 +40,7 @@ SECTIONS /* Sections to be discarded. */ /DISCARD/ : { *(.exitcall.exit) + *(.discard) EXIT_DATA } diff --git a/arch/s390/kernel/vmlinux.lds.S b/arch/s390/kernel/vmlinux.lds.S index a53db23ee092..98867dfea469 100644 --- a/arch/s390/kernel/vmlinux.lds.S +++ b/arch/s390/kernel/vmlinux.lds.S @@ -161,6 +161,7 @@ SECTIONS /DISCARD/ : { EXIT_DATA *(.exitcall.exit) + *(.discard) } /* Debugging sections. */ diff --git a/arch/sh/kernel/vmlinux.lds.S b/arch/sh/kernel/vmlinux.lds.S index f53c76acaede..766976d27b21 100644 --- a/arch/sh/kernel/vmlinux.lds.S +++ b/arch/sh/kernel/vmlinux.lds.S @@ -171,6 +171,7 @@ SECTIONS */ /DISCARD/ : { *(.exitcall.exit) + *(.discard) } STABS_DEBUG diff --git a/arch/sparc/kernel/vmlinux.lds.S b/arch/sparc/kernel/vmlinux.lds.S index fcbbd000ec08..d63cf914667d 100644 --- a/arch/sparc/kernel/vmlinux.lds.S +++ b/arch/sparc/kernel/vmlinux.lds.S @@ -175,6 +175,7 @@ SECTIONS EXIT_TEXT EXIT_DATA *(.exitcall.exit) + *(.discard) } STABS_DEBUG diff --git a/arch/um/kernel/dyn.lds.S b/arch/um/kernel/dyn.lds.S index 9975e1ab44fb..2916d6eadffd 100644 --- a/arch/um/kernel/dyn.lds.S +++ b/arch/um/kernel/dyn.lds.S @@ -156,4 +156,6 @@ SECTIONS STABS_DEBUG DWARF_DEBUG + + /DISCARD/ : { *(.discard) } } diff --git a/arch/um/kernel/uml.lds.S b/arch/um/kernel/uml.lds.S index 11b835248b86..1f8a622cabe1 100644 --- a/arch/um/kernel/uml.lds.S +++ b/arch/um/kernel/uml.lds.S @@ -100,4 +100,6 @@ SECTIONS STABS_DEBUG DWARF_DEBUG + + /DISCARD/ : { *(.discard) } } diff --git a/arch/xtensa/kernel/vmlinux.lds.S b/arch/xtensa/kernel/vmlinux.lds.S index 41c159cd872f..b1e24638acd7 100644 --- a/arch/xtensa/kernel/vmlinux.lds.S +++ b/arch/xtensa/kernel/vmlinux.lds.S @@ -287,6 +287,7 @@ SECTIONS EXIT_TEXT EXIT_DATA *(.exitcall.exit) + *(.discard) } .xt.lit : { *(.xt.lit) } diff --git a/include/asm-generic/vmlinux.lds.h b/include/asm-generic/vmlinux.lds.h index 55413e568f07..a19120c4e109 100644 --- a/include/asm-generic/vmlinux.lds.h +++ b/include/asm-generic/vmlinux.lds.h @@ -628,6 +628,14 @@ #define INITRAMFS #endif +#define DISCARDS \ + /DISCARD/ : { \ + EXIT_TEXT \ + EXIT_DATA \ + *(.exitcall.exit) \ + *(.discard) \ + } + /** * PERCPU_VADDR - define output section for percpu area * @vaddr: explicit base address (optional) diff --git a/scripts/module-common.lds b/scripts/module-common.lds new file mode 100644 index 000000000000..47a1f9ae0ede --- /dev/null +++ b/scripts/module-common.lds @@ -0,0 +1,8 @@ +/* + * Common module linker script, always used when linking a module. + * Archs are free to supply their own linker scripts. ld will + * combine them automatically. + */ +SECTIONS { + /DISCARD/ : { *(.discard) } +} -- cgit v1.2.3 From 7c756e6e19e71f0327760d8955f7077118ebb2b1 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Wed, 24 Jun 2009 15:13:50 +0900 Subject: percpu: implement optional weak percpu definitions Some archs (alpha and s390) need to use weak definitions for percpu variables in modules so that the compiler generates external references for them. This patch implements weak percpu definitions which arch can enable by defining ARCH_NEEDS_WEAK_PER_CPU in arch percpu header file. This weak definition adds the following two restrictions on percpu variable definitions. 1. percpu symbols must be unique whether static or not 2. percpu variables can't be defined inside a function To ensure that these restrictions are observed in generic code, config option DEBUG_FORCE_WEAK_PER_CPU enables weak percpu definitions for all cases. This patch is inspired by Ivan Kokshaysky's alpha percpu patch. [ Impact: stricter rules for percpu variables, one more debug config option ] Signed-off-by: Tejun Heo Cc: Ingo Molnar Cc: David Howells Cc: Ivan Kokshaysky --- include/linux/percpu-defs.h | 65 ++++++++++++++++++++++++++++++++++++++------- lib/Kconfig.debug | 15 +++++++++++ 2 files changed, 71 insertions(+), 9 deletions(-) (limited to 'include') diff --git a/include/linux/percpu-defs.h b/include/linux/percpu-defs.h index 8f921d74f49f..cf32838ad0fa 100644 --- a/include/linux/percpu-defs.h +++ b/include/linux/percpu-defs.h @@ -10,21 +10,68 @@ /* * Base implementations of per-CPU variable declarations and definitions, where * the section in which the variable is to be placed is provided by the - * 'section' argument. This may be used to affect the parameters governing the + * 'sec' argument. This may be used to affect the parameters governing the * variable's storage. * * NOTE! The sections for the DECLARE and for the DEFINE must match, lest * linkage errors occur due the compiler generating the wrong code to access * that section. */ -#define DECLARE_PER_CPU_SECTION(type, name, section) \ - extern \ - __attribute__((__section__(PER_CPU_BASE_SECTION section))) \ - PER_CPU_ATTRIBUTES __typeof__(type) per_cpu__##name - -#define DEFINE_PER_CPU_SECTION(type, name, section) \ - __attribute__((__section__(PER_CPU_BASE_SECTION section))) \ - PER_CPU_ATTRIBUTES __typeof__(type) per_cpu__##name +#define __PCPU_ATTRS(sec) \ + __attribute__((section(PER_CPU_BASE_SECTION sec))) \ + PER_CPU_ATTRIBUTES + +#define __PCPU_DUMMY_ATTRS \ + __attribute__((section(".discard"), unused)) + +/* + * s390 and alpha modules require percpu variables to be defined as + * weak to force the compiler to generate GOT based external + * references for them. This is necessary because percpu sections + * will be located outside of the usually addressable area. + * + * This definition puts the following two extra restrictions when + * defining percpu variables. + * + * 1. The symbol must be globally unique, even the static ones. + * 2. Static percpu variables cannot be defined inside a function. + * + * Archs which need weak percpu definitions should define + * ARCH_NEEDS_WEAK_PER_CPU in asm/percpu.h when necessary. + * + * To ensure that the generic code observes the above two + * restrictions, if CONFIG_DEBUG_FORCE_WEAK_PER_CPU is set weak + * definition is used for all cases. + */ +#if defined(ARCH_NEEDS_WEAK_PER_CPU) || defined(CONFIG_DEBUG_FORCE_WEAK_PER_CPU) +/* + * __pcpu_scope_* dummy variable is used to enforce scope. It + * receives the static modifier when it's used in front of + * DEFINE_PER_CPU() and will trigger build failure if + * DECLARE_PER_CPU() is used for the same variable. + * + * __pcpu_unique_* dummy variable is used to enforce symbol uniqueness + * such that hidden weak symbol collision, which will cause unrelated + * variables to share the same address, can be detected during build. + */ +#define DECLARE_PER_CPU_SECTION(type, name, sec) \ + extern __PCPU_DUMMY_ATTRS char __pcpu_scope_##name; \ + extern __PCPU_ATTRS(sec) __weak __typeof__(type) per_cpu__##name + +#define DEFINE_PER_CPU_SECTION(type, name, sec) \ + __PCPU_DUMMY_ATTRS char __pcpu_scope_##name; \ + __PCPU_DUMMY_ATTRS char __pcpu_unique_##name; \ + __PCPU_ATTRS(sec) __weak __typeof__(type) per_cpu__##name +#else +/* + * Normal declaration and definition macros. + */ +#define DECLARE_PER_CPU_SECTION(type, name, sec) \ + extern __PCPU_ATTRS(sec) __typeof__(type) per_cpu__##name + +#define DEFINE_PER_CPU_SECTION(type, name, sec) \ + __PCPU_ATTRS(sec) __typeof__(type) per_cpu__##name +#endif /* * Variant on the per-CPU variable declaration/definition theme used for diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index 23067ab1a73c..77e0d8b1b7c5 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -777,6 +777,21 @@ config DEBUG_BLOCK_EXT_DEVT Say N if you are unsure. +config DEBUG_FORCE_WEAK_PER_CPU + bool "Force weak per-cpu definitions" + depends on DEBUG_KERNEL + help + s390 and alpha require percpu variables in modules to be + defined weak to work around addressing range issue which + puts the following two restrictions on percpu variable + definitions. + + 1. percpu symbols must be unique whether static or not + 2. percpu variables can't be defined inside a function + + To ensure that generic code follows the above rules, this + option forces all percpu variables to be defined as weak. + config LKDTM tristate "Linux Kernel Dump Test Tool Module" depends on DEBUG_KERNEL -- cgit v1.2.3 From 346c17a6cf60375323adfaa4b8a9d841049f890e Mon Sep 17 00:00:00 2001 From: Bartlomiej Zolnierkiewicz Date: Mon, 22 Jun 2009 07:38:26 +0000 Subject: ide: relax DMA info validity checking There are some broken devices that report multiple DMA xfer modes enabled at once (ATA spec doesn't allow it) but otherwise work fine with DMA so just delete ide_id_dma_bug(). [ As discovered by detective work by Frans and Bart, due to how handling of the ID block was handled before commit c419993 ("ide-iops: only clear DMA words on setting DMA mode") this check was always seeing zeros in the fields or other similar garbage. Therefore this check wasn't actually checking anything. Now that the tests actually check the real bits, all we see are devices that trigger the check yet work perfectly fine, therefore killing this useless check is the best thing to do. -DaveM ] Reported-by: Frans Pop Signed-off-by: Bartlomiej Zolnierkiewicz Signed-off-by: David S. Miller --- drivers/ide/ide-dma.c | 21 --------------------- drivers/ide/ide-iops.c | 3 --- include/linux/ide.h | 2 -- 3 files changed, 26 deletions(-) (limited to 'include') diff --git a/drivers/ide/ide-dma.c b/drivers/ide/ide-dma.c index 219e6fb78dc6..ee58c88dee5a 100644 --- a/drivers/ide/ide-dma.c +++ b/drivers/ide/ide-dma.c @@ -361,9 +361,6 @@ static int ide_tune_dma(ide_drive_t *drive) if (__ide_dma_bad_drive(drive)) return 0; - if (ide_id_dma_bug(drive)) - return 0; - if (hwif->host_flags & IDE_HFLAG_TRUST_BIOS_FOR_DMA) return config_drive_for_dma(drive); @@ -394,24 +391,6 @@ static int ide_dma_check(ide_drive_t *drive) return -1; } -int ide_id_dma_bug(ide_drive_t *drive) -{ - u16 *id = drive->id; - - if (id[ATA_ID_FIELD_VALID] & 4) { - if ((id[ATA_ID_UDMA_MODES] >> 8) && - (id[ATA_ID_MWDMA_MODES] >> 8)) - goto err_out; - } else if ((id[ATA_ID_MWDMA_MODES] >> 8) && - (id[ATA_ID_SWDMA_MODES] >> 8)) - goto err_out; - - return 0; -err_out: - printk(KERN_ERR "%s: bad DMA info in identify block\n", drive->name); - return 1; -} - int ide_set_dma(ide_drive_t *drive) { int rc; diff --git a/drivers/ide/ide-iops.c b/drivers/ide/ide-iops.c index fa047150a1c6..917186ec4966 100644 --- a/drivers/ide/ide-iops.c +++ b/drivers/ide/ide-iops.c @@ -329,9 +329,6 @@ int ide_driveid_update(ide_drive_t *drive) kfree(id); - if ((drive->dev_flags & IDE_DFLAG_USING_DMA) && ide_id_dma_bug(drive)) - ide_dma_off(drive); - return 1; out_err: if (rc == 2) diff --git a/include/linux/ide.h b/include/linux/ide.h index 95c6e00a72e8..cf1f3888067c 100644 --- a/include/linux/ide.h +++ b/include/linux/ide.h @@ -1361,7 +1361,6 @@ int ide_in_drive_list(u16 *, const struct drive_list_entry *); #ifdef CONFIG_BLK_DEV_IDEDMA int ide_dma_good_drive(ide_drive_t *); int __ide_dma_bad_drive(ide_drive_t *); -int ide_id_dma_bug(ide_drive_t *); u8 ide_find_dma_mode(ide_drive_t *, u8); @@ -1402,7 +1401,6 @@ void ide_dma_lost_irq(ide_drive_t *); ide_startstop_t ide_dma_timeout_retry(ide_drive_t *, int); #else -static inline int ide_id_dma_bug(ide_drive_t *drive) { return 0; } static inline u8 ide_find_dma_mode(ide_drive_t *drive, u8 speed) { return 0; } static inline u8 ide_max_dma_mode(ide_drive_t *drive) { return 0; } static inline void ide_dma_off_quietly(ide_drive_t *drive) { ; } -- cgit v1.2.3 From 507e123151149e578c9aae33eb876c49824da5f8 Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Tue, 23 Jun 2009 17:38:15 +0200 Subject: timer stats: Optimize by adding quick check to avoid function calls When the kernel is configured with CONFIG_TIMER_STATS but timer stats are runtime disabled we still get calls to __timer_stats_timer_set_start_info which initializes some fields in the corresponding struct timer_list. So add some quick checks in the the timer stats setup functions to avoid function calls to __timer_stats_timer_set_start_info when timer stats are disabled. In an artificial workload that does nothing but playing ping pong with a single tcp packet via loopback this decreases cpu consumption by 1 - 1.5%. This is part of a modified function trace output on SLES11: perl-2497 [00] 28630647177732388 [+ 125]: sk_reset_timer <-tcp_v4_rcv perl-2497 [00] 28630647177732513 [+ 125]: mod_timer <-sk_reset_timer perl-2497 [00] 28630647177732638 [+ 125]: __timer_stats_timer_set_start_info <-mod_timer perl-2497 [00] 28630647177732763 [+ 125]: __mod_timer <-mod_timer perl-2497 [00] 28630647177732888 [+ 125]: __timer_stats_timer_set_start_info <-__mod_timer perl-2497 [00] 28630647177733013 [+ 93]: lock_timer_base <-__mod_timer Signed-off-by: Heiko Carstens Cc: Andrew Morton Cc: Martin Schwidefsky Cc: Mustafa Mesanovic Cc: Arjan van de Ven LKML-Reference: <20090623153811.GA4641@osiris.boeblingen.de.ibm.com> Signed-off-by: Ingo Molnar --- include/linux/hrtimer.h | 5 +++++ include/linux/timer.h | 4 ++++ kernel/time/timer_stats.c | 16 ++++++++-------- kernel/timer.c | 2 ++ 4 files changed, 19 insertions(+), 8 deletions(-) (limited to 'include') diff --git a/include/linux/hrtimer.h b/include/linux/hrtimer.h index 7400900de94a..54648e625efd 100644 --- a/include/linux/hrtimer.h +++ b/include/linux/hrtimer.h @@ -21,6 +21,7 @@ #include #include #include +#include struct hrtimer_clock_base; @@ -447,6 +448,8 @@ extern void timer_stats_update_stats(void *timer, pid_t pid, void *startf, static inline void timer_stats_account_hrtimer(struct hrtimer *timer) { + if (likely(!timer->start_pid)) + return; timer_stats_update_stats(timer, timer->start_pid, timer->start_site, timer->function, timer->start_comm, 0); } @@ -456,6 +459,8 @@ extern void __timer_stats_hrtimer_set_start_info(struct hrtimer *timer, static inline void timer_stats_hrtimer_set_start_info(struct hrtimer *timer) { + if (likely(!timer_stats_active)) + return; __timer_stats_hrtimer_set_start_info(timer, __builtin_return_address(0)); } diff --git a/include/linux/timer.h b/include/linux/timer.h index ccf882eed8f8..be62ec2ebea5 100644 --- a/include/linux/timer.h +++ b/include/linux/timer.h @@ -190,6 +190,8 @@ extern unsigned long get_next_timer_interrupt(unsigned long now); */ #ifdef CONFIG_TIMER_STATS +extern int timer_stats_active; + #define TIMER_STATS_FLAG_DEFERRABLE 0x1 extern void init_timer_stats(void); @@ -203,6 +205,8 @@ extern void __timer_stats_timer_set_start_info(struct timer_list *timer, static inline void timer_stats_timer_set_start_info(struct timer_list *timer) { + if (likely(!timer_stats_active)) + return; __timer_stats_timer_set_start_info(timer, __builtin_return_address(0)); } diff --git a/kernel/time/timer_stats.c b/kernel/time/timer_stats.c index c994530d166d..4cde8b9c716f 100644 --- a/kernel/time/timer_stats.c +++ b/kernel/time/timer_stats.c @@ -96,7 +96,7 @@ static DEFINE_MUTEX(show_mutex); /* * Collection status, active/inactive: */ -static int __read_mostly active; +int __read_mostly timer_stats_active; /* * Beginning/end timestamps of measurement: @@ -242,7 +242,7 @@ void timer_stats_update_stats(void *timer, pid_t pid, void *startf, struct entry *entry, input; unsigned long flags; - if (likely(!active)) + if (likely(!timer_stats_active)) return; lock = &per_cpu(lookup_lock, raw_smp_processor_id()); @@ -254,7 +254,7 @@ void timer_stats_update_stats(void *timer, pid_t pid, void *startf, input.timer_flag = timer_flag; spin_lock_irqsave(lock, flags); - if (!active) + if (!timer_stats_active) goto out_unlock; entry = tstat_lookup(&input, comm); @@ -290,7 +290,7 @@ static int tstats_show(struct seq_file *m, void *v) /* * If still active then calculate up to now: */ - if (active) + if (timer_stats_active) time_stop = ktime_get(); time = ktime_sub(time_stop, time_start); @@ -368,18 +368,18 @@ static ssize_t tstats_write(struct file *file, const char __user *buf, mutex_lock(&show_mutex); switch (ctl[0]) { case '0': - if (active) { - active = 0; + if (timer_stats_active) { + timer_stats_active = 0; time_stop = ktime_get(); sync_access(); } break; case '1': - if (!active) { + if (!timer_stats_active) { reset_entries(); time_start = ktime_get(); smp_mb(); - active = 1; + timer_stats_active = 1; } break; default: diff --git a/kernel/timer.c b/kernel/timer.c index 54d3912f8cad..0b36b9e5cc8b 100644 --- a/kernel/timer.c +++ b/kernel/timer.c @@ -380,6 +380,8 @@ static void timer_stats_account_timer(struct timer_list *timer) { unsigned int flag = 0; + if (likely(!timer->start_site)) + return; if (unlikely(tbase_get_deferrable(timer->base))) flag |= TIMER_STATS_FLAG_DEFERRABLE; -- cgit v1.2.3 From c17ef45342cc033fdf7bdd5b28615e0090f8d2e7 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Tue, 23 Jun 2009 17:12:47 -0700 Subject: rcu: Remove Classic RCU Remove Classic RCU, given that the combination of Tree RCU and the proposed Bloatwatch RCU do everything that Classic RCU can with fewer bugs. Tree RCU has been default in x86 builds for almost six months, and seems to be quite reliable, so there does not seem to be much justification for keeping the Classic RCU code and config complexity around anymore. Signed-off-by: Paul E. McKenney Cc: akpm@linux-foundation.org Cc: niv@us.ibm.com Cc: dvhltc@us.ibm.com Cc: dipankar@in.ibm.com Cc: dhowells@redhat.com Cc: lethal@linux-sh.org Cc: kernel@wantstofly.org Signed-off-by: Ingo Molnar --- include/linux/rcuclassic.h | 178 ---------- include/linux/rcupdate.h | 4 +- init/Kconfig | 12 +- kernel/Makefile | 1 - kernel/rcuclassic.c | 807 --------------------------------------------- 5 files changed, 3 insertions(+), 999 deletions(-) delete mode 100644 include/linux/rcuclassic.h delete mode 100644 kernel/rcuclassic.c (limited to 'include') diff --git a/include/linux/rcuclassic.h b/include/linux/rcuclassic.h deleted file mode 100644 index bfd92e1e5d2c..000000000000 --- a/include/linux/rcuclassic.h +++ /dev/null @@ -1,178 +0,0 @@ -/* - * Read-Copy Update mechanism for mutual exclusion (classic version) - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. - * - * Copyright IBM Corporation, 2001 - * - * Author: Dipankar Sarma - * - * Based on the original work by Paul McKenney - * and inputs from Rusty Russell, Andrea Arcangeli and Andi Kleen. - * Papers: - * http://www.rdrop.com/users/paulmck/paper/rclockpdcsproof.pdf - * http://lse.sourceforge.net/locking/rclock_OLS.2001.05.01c.sc.pdf (OLS2001) - * - * For detailed explanation of Read-Copy Update mechanism see - - * Documentation/RCU - * - */ - -#ifndef __LINUX_RCUCLASSIC_H -#define __LINUX_RCUCLASSIC_H - -#include -#include -#include -#include -#include - -#ifdef CONFIG_RCU_CPU_STALL_DETECTOR -#define RCU_SECONDS_TILL_STALL_CHECK (10 * HZ) /* for rcp->jiffies_stall */ -#define RCU_SECONDS_TILL_STALL_RECHECK (30 * HZ) /* for rcp->jiffies_stall */ -#endif /* #ifdef CONFIG_RCU_CPU_STALL_DETECTOR */ - -/* Global control variables for rcupdate callback mechanism. */ -struct rcu_ctrlblk { - long cur; /* Current batch number. */ - long completed; /* Number of the last completed batch */ - long pending; /* Number of the last pending batch */ -#ifdef CONFIG_RCU_CPU_STALL_DETECTOR - unsigned long gp_start; /* Time at which GP started in jiffies. */ - unsigned long jiffies_stall; - /* Time at which to check for CPU stalls. */ -#endif /* #ifdef CONFIG_RCU_CPU_STALL_DETECTOR */ - - int signaled; - - spinlock_t lock ____cacheline_internodealigned_in_smp; - DECLARE_BITMAP(cpumask, NR_CPUS); /* CPUs that need to switch for */ - /* current batch to proceed. */ -} ____cacheline_internodealigned_in_smp; - -/* Is batch a before batch b ? */ -static inline int rcu_batch_before(long a, long b) -{ - return (a - b) < 0; -} - -/* Is batch a after batch b ? */ -static inline int rcu_batch_after(long a, long b) -{ - return (a - b) > 0; -} - -/* Per-CPU data for Read-Copy UPdate. */ -struct rcu_data { - /* 1) quiescent state handling : */ - long quiescbatch; /* Batch # for grace period */ - int passed_quiesc; /* User-mode/idle loop etc. */ - int qs_pending; /* core waits for quiesc state */ - - /* 2) batch handling */ - /* - * if nxtlist is not NULL, then: - * batch: - * The batch # for the last entry of nxtlist - * [*nxttail[1], NULL = *nxttail[2]): - * Entries that batch # <= batch - * [*nxttail[0], *nxttail[1]): - * Entries that batch # <= batch - 1 - * [nxtlist, *nxttail[0]): - * Entries that batch # <= batch - 2 - * The grace period for these entries has completed, and - * the other grace-period-completed entries may be moved - * here temporarily in rcu_process_callbacks(). - */ - long batch; - struct rcu_head *nxtlist; - struct rcu_head **nxttail[3]; - long qlen; /* # of queued callbacks */ - struct rcu_head *donelist; - struct rcu_head **donetail; - long blimit; /* Upper limit on a processed batch */ - int cpu; - struct rcu_head barrier; -}; - -/* - * Increment the quiescent state counter. - * The counter is a bit degenerated: We do not need to know - * how many quiescent states passed, just if there was at least - * one since the start of the grace period. Thus just a flag. - */ -extern void rcu_qsctr_inc(int cpu); -extern void rcu_bh_qsctr_inc(int cpu); - -extern int rcu_pending(int cpu); -extern int rcu_needs_cpu(int cpu); - -#ifdef CONFIG_DEBUG_LOCK_ALLOC -extern struct lockdep_map rcu_lock_map; -# define rcu_read_acquire() \ - lock_acquire(&rcu_lock_map, 0, 0, 2, 1, NULL, _THIS_IP_) -# define rcu_read_release() lock_release(&rcu_lock_map, 1, _THIS_IP_) -#else -# define rcu_read_acquire() do { } while (0) -# define rcu_read_release() do { } while (0) -#endif - -#define __rcu_read_lock() \ - do { \ - preempt_disable(); \ - __acquire(RCU); \ - rcu_read_acquire(); \ - } while (0) -#define __rcu_read_unlock() \ - do { \ - rcu_read_release(); \ - __release(RCU); \ - preempt_enable(); \ - } while (0) -#define __rcu_read_lock_bh() \ - do { \ - local_bh_disable(); \ - __acquire(RCU_BH); \ - rcu_read_acquire(); \ - } while (0) -#define __rcu_read_unlock_bh() \ - do { \ - rcu_read_release(); \ - __release(RCU_BH); \ - local_bh_enable(); \ - } while (0) - -#define __synchronize_sched() synchronize_rcu() - -#define call_rcu_sched(head, func) call_rcu(head, func) - -extern void __rcu_init(void); -#define rcu_init_sched() do { } while (0) -extern void rcu_check_callbacks(int cpu, int user); -extern void rcu_restart_cpu(int cpu); - -extern long rcu_batches_completed(void); -extern long rcu_batches_completed_bh(void); - -#define rcu_enter_nohz() do { } while (0) -#define rcu_exit_nohz() do { } while (0) - -/* A context switch is a grace period for rcuclassic. */ -static inline int rcu_blocking_is_gp(void) -{ - return num_online_cpus() == 1; -} - -#endif /* __LINUX_RCUCLASSIC_H */ diff --git a/include/linux/rcupdate.h b/include/linux/rcupdate.h index 15fbb3ca634d..0cdfdb622faa 100644 --- a/include/linux/rcupdate.h +++ b/include/linux/rcupdate.h @@ -54,9 +54,7 @@ struct rcu_head { /* Internal to kernel, but needed by rcupreempt.h. */ extern int rcu_scheduler_active; -#if defined(CONFIG_CLASSIC_RCU) -#include -#elif defined(CONFIG_TREE_RCU) +#if defined(CONFIG_TREE_RCU) #include #elif defined(CONFIG_PREEMPT_RCU) #include diff --git a/init/Kconfig b/init/Kconfig index 1ce05a4cb5f6..d10f31dfa0b2 100644 --- a/init/Kconfig +++ b/init/Kconfig @@ -316,21 +316,13 @@ choice prompt "RCU Implementation" default TREE_RCU -config CLASSIC_RCU - bool "Classic RCU" - help - This option selects the classic RCU implementation that is - designed for best read-side performance on non-realtime - systems. - - Select this option if you are unsure. - config TREE_RCU bool "Tree-based hierarchical RCU" help This option selects the RCU implementation that is designed for very large SMP system with hundreds or - thousands of CPUs. + thousands of CPUs. It also scales down nicely to + smaller systems. config PREEMPT_RCU bool "Preemptible RCU" diff --git a/kernel/Makefile b/kernel/Makefile index 0a32cb21ec97..d6042bcc9e99 100644 --- a/kernel/Makefile +++ b/kernel/Makefile @@ -80,7 +80,6 @@ obj-$(CONFIG_DETECT_HUNG_TASK) += hung_task.o obj-$(CONFIG_GENERIC_HARDIRQS) += irq/ obj-$(CONFIG_SECCOMP) += seccomp.o obj-$(CONFIG_RCU_TORTURE_TEST) += rcutorture.o -obj-$(CONFIG_CLASSIC_RCU) += rcuclassic.o obj-$(CONFIG_TREE_RCU) += rcutree.o obj-$(CONFIG_PREEMPT_RCU) += rcupreempt.o obj-$(CONFIG_TREE_RCU_TRACE) += rcutree_trace.o diff --git a/kernel/rcuclassic.c b/kernel/rcuclassic.c deleted file mode 100644 index 0f2b0b311304..000000000000 --- a/kernel/rcuclassic.c +++ /dev/null @@ -1,807 +0,0 @@ -/* - * Read-Copy Update mechanism for mutual exclusion - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. - * - * Copyright IBM Corporation, 2001 - * - * Authors: Dipankar Sarma - * Manfred Spraul - * - * Based on the original work by Paul McKenney - * and inputs from Rusty Russell, Andrea Arcangeli and Andi Kleen. - * Papers: - * http://www.rdrop.com/users/paulmck/paper/rclockpdcsproof.pdf - * http://lse.sourceforge.net/locking/rclock_OLS.2001.05.01c.sc.pdf (OLS2001) - * - * For detailed explanation of Read-Copy Update mechanism see - - * Documentation/RCU - * - */ -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include - -#ifdef CONFIG_DEBUG_LOCK_ALLOC -static struct lock_class_key rcu_lock_key; -struct lockdep_map rcu_lock_map = - STATIC_LOCKDEP_MAP_INIT("rcu_read_lock", &rcu_lock_key); -EXPORT_SYMBOL_GPL(rcu_lock_map); -#endif - - -/* Definition for rcupdate control block. */ -static struct rcu_ctrlblk rcu_ctrlblk = { - .cur = -300, - .completed = -300, - .pending = -300, - .lock = __SPIN_LOCK_UNLOCKED(&rcu_ctrlblk.lock), - .cpumask = CPU_BITS_NONE, -}; - -static struct rcu_ctrlblk rcu_bh_ctrlblk = { - .cur = -300, - .completed = -300, - .pending = -300, - .lock = __SPIN_LOCK_UNLOCKED(&rcu_bh_ctrlblk.lock), - .cpumask = CPU_BITS_NONE, -}; - -static DEFINE_PER_CPU(struct rcu_data, rcu_data); -static DEFINE_PER_CPU(struct rcu_data, rcu_bh_data); - -/* - * Increment the quiescent state counter. - * The counter is a bit degenerated: We do not need to know - * how many quiescent states passed, just if there was at least - * one since the start of the grace period. Thus just a flag. - */ -void rcu_qsctr_inc(int cpu) -{ - struct rcu_data *rdp = &per_cpu(rcu_data, cpu); - rdp->passed_quiesc = 1; -} - -void rcu_bh_qsctr_inc(int cpu) -{ - struct rcu_data *rdp = &per_cpu(rcu_bh_data, cpu); - rdp->passed_quiesc = 1; -} - -static int blimit = 10; -static int qhimark = 10000; -static int qlowmark = 100; - -#ifdef CONFIG_SMP -static void force_quiescent_state(struct rcu_data *rdp, - struct rcu_ctrlblk *rcp) -{ - int cpu; - unsigned long flags; - - set_need_resched(); - spin_lock_irqsave(&rcp->lock, flags); - if (unlikely(!rcp->signaled)) { - rcp->signaled = 1; - /* - * Don't send IPI to itself. With irqs disabled, - * rdp->cpu is the current cpu. - * - * cpu_online_mask is updated by the _cpu_down() - * using __stop_machine(). Since we're in irqs disabled - * section, __stop_machine() is not exectuting, hence - * the cpu_online_mask is stable. - * - * However, a cpu might have been offlined _just_ before - * we disabled irqs while entering here. - * And rcu subsystem might not yet have handled the CPU_DEAD - * notification, leading to the offlined cpu's bit - * being set in the rcp->cpumask. - * - * Hence cpumask = (rcp->cpumask & cpu_online_mask) to prevent - * sending smp_reschedule() to an offlined CPU. - */ - for_each_cpu_and(cpu, - to_cpumask(rcp->cpumask), cpu_online_mask) { - if (cpu != rdp->cpu) - smp_send_reschedule(cpu); - } - } - spin_unlock_irqrestore(&rcp->lock, flags); -} -#else -static inline void force_quiescent_state(struct rcu_data *rdp, - struct rcu_ctrlblk *rcp) -{ - set_need_resched(); -} -#endif - -static void __call_rcu(struct rcu_head *head, struct rcu_ctrlblk *rcp, - struct rcu_data *rdp) -{ - long batch; - - head->next = NULL; - smp_mb(); /* Read of rcu->cur must happen after any change by caller. */ - - /* - * Determine the batch number of this callback. - * - * Using ACCESS_ONCE to avoid the following error when gcc eliminates - * local variable "batch" and emits codes like this: - * 1) rdp->batch = rcp->cur + 1 # gets old value - * ...... - * 2)rcu_batch_after(rcp->cur + 1, rdp->batch) # gets new value - * then [*nxttail[0], *nxttail[1]) may contain callbacks - * that batch# = rdp->batch, see the comment of struct rcu_data. - */ - batch = ACCESS_ONCE(rcp->cur) + 1; - - if (rdp->nxtlist && rcu_batch_after(batch, rdp->batch)) { - /* process callbacks */ - rdp->nxttail[0] = rdp->nxttail[1]; - rdp->nxttail[1] = rdp->nxttail[2]; - if (rcu_batch_after(batch - 1, rdp->batch)) - rdp->nxttail[0] = rdp->nxttail[2]; - } - - rdp->batch = batch; - *rdp->nxttail[2] = head; - rdp->nxttail[2] = &head->next; - - if (unlikely(++rdp->qlen > qhimark)) { - rdp->blimit = INT_MAX; - force_quiescent_state(rdp, &rcu_ctrlblk); - } -} - -#ifdef CONFIG_RCU_CPU_STALL_DETECTOR - -static void record_gp_stall_check_time(struct rcu_ctrlblk *rcp) -{ - rcp->gp_start = jiffies; - rcp->jiffies_stall = jiffies + RCU_SECONDS_TILL_STALL_CHECK; -} - -static void print_other_cpu_stall(struct rcu_ctrlblk *rcp) -{ - int cpu; - long delta; - unsigned long flags; - - /* Only let one CPU complain about others per time interval. */ - - spin_lock_irqsave(&rcp->lock, flags); - delta = jiffies - rcp->jiffies_stall; - if (delta < 2 || rcp->cur != rcp->completed) { - spin_unlock_irqrestore(&rcp->lock, flags); - return; - } - rcp->jiffies_stall = jiffies + RCU_SECONDS_TILL_STALL_RECHECK; - spin_unlock_irqrestore(&rcp->lock, flags); - - /* OK, time to rat on our buddy... */ - - printk(KERN_ERR "INFO: RCU detected CPU stalls:"); - for_each_possible_cpu(cpu) { - if (cpumask_test_cpu(cpu, to_cpumask(rcp->cpumask))) - printk(" %d", cpu); - } - printk(" (detected by %d, t=%ld jiffies)\n", - smp_processor_id(), (long)(jiffies - rcp->gp_start)); -} - -static void print_cpu_stall(struct rcu_ctrlblk *rcp) -{ - unsigned long flags; - - printk(KERN_ERR "INFO: RCU detected CPU %d stall (t=%lu/%lu jiffies)\n", - smp_processor_id(), jiffies, - jiffies - rcp->gp_start); - dump_stack(); - spin_lock_irqsave(&rcp->lock, flags); - if ((long)(jiffies - rcp->jiffies_stall) >= 0) - rcp->jiffies_stall = - jiffies + RCU_SECONDS_TILL_STALL_RECHECK; - spin_unlock_irqrestore(&rcp->lock, flags); - set_need_resched(); /* kick ourselves to get things going. */ -} - -static void check_cpu_stall(struct rcu_ctrlblk *rcp) -{ - long delta; - - delta = jiffies - rcp->jiffies_stall; - if (cpumask_test_cpu(smp_processor_id(), to_cpumask(rcp->cpumask)) && - delta >= 0) { - - /* We haven't checked in, so go dump stack. */ - print_cpu_stall(rcp); - - } else if (rcp->cur != rcp->completed && delta >= 2) { - - /* They had two seconds to dump stack, so complain. */ - print_other_cpu_stall(rcp); - } -} - -#else /* #ifdef CONFIG_RCU_CPU_STALL_DETECTOR */ - -static void record_gp_stall_check_time(struct rcu_ctrlblk *rcp) -{ -} - -static inline void check_cpu_stall(struct rcu_ctrlblk *rcp) -{ -} - -#endif /* #else #ifdef CONFIG_RCU_CPU_STALL_DETECTOR */ - -/** - * call_rcu - Queue an RCU callback for invocation after a grace period. - * @head: structure to be used for queueing the RCU updates. - * @func: actual update function to be invoked after the grace period - * - * The update function will be invoked some time after a full grace - * period elapses, in other words after all currently executing RCU - * read-side critical sections have completed. RCU read-side critical - * sections are delimited by rcu_read_lock() and rcu_read_unlock(), - * and may be nested. - */ -void call_rcu(struct rcu_head *head, - void (*func)(struct rcu_head *rcu)) -{ - unsigned long flags; - - head->func = func; - local_irq_save(flags); - __call_rcu(head, &rcu_ctrlblk, &__get_cpu_var(rcu_data)); - local_irq_restore(flags); -} -EXPORT_SYMBOL_GPL(call_rcu); - -/** - * call_rcu_bh - Queue an RCU for invocation after a quicker grace period. - * @head: structure to be used for queueing the RCU updates. - * @func: actual update function to be invoked after the grace period - * - * The update function will be invoked some time after a full grace - * period elapses, in other words after all currently executing RCU - * read-side critical sections have completed. call_rcu_bh() assumes - * that the read-side critical sections end on completion of a softirq - * handler. This means that read-side critical sections in process - * context must not be interrupted by softirqs. This interface is to be - * used when most of the read-side critical sections are in softirq context. - * RCU read-side critical sections are delimited by rcu_read_lock() and - * rcu_read_unlock(), * if in interrupt context or rcu_read_lock_bh() - * and rcu_read_unlock_bh(), if in process context. These may be nested. - */ -void call_rcu_bh(struct rcu_head *head, - void (*func)(struct rcu_head *rcu)) -{ - unsigned long flags; - - head->func = func; - local_irq_save(flags); - __call_rcu(head, &rcu_bh_ctrlblk, &__get_cpu_var(rcu_bh_data)); - local_irq_restore(flags); -} -EXPORT_SYMBOL_GPL(call_rcu_bh); - -/* - * Return the number of RCU batches processed thus far. Useful - * for debug and statistics. - */ -long rcu_batches_completed(void) -{ - return rcu_ctrlblk.completed; -} -EXPORT_SYMBOL_GPL(rcu_batches_completed); - -/* - * Return the number of RCU batches processed thus far. Useful - * for debug and statistics. - */ -long rcu_batches_completed_bh(void) -{ - return rcu_bh_ctrlblk.completed; -} -EXPORT_SYMBOL_GPL(rcu_batches_completed_bh); - -/* Raises the softirq for processing rcu_callbacks. */ -static inline void raise_rcu_softirq(void) -{ - raise_softirq(RCU_SOFTIRQ); -} - -/* - * Invoke the completed RCU callbacks. They are expected to be in - * a per-cpu list. - */ -static void rcu_do_batch(struct rcu_data *rdp) -{ - unsigned long flags; - struct rcu_head *next, *list; - int count = 0; - - list = rdp->donelist; - while (list) { - next = list->next; - prefetch(next); - list->func(list); - list = next; - if (++count >= rdp->blimit) - break; - } - rdp->donelist = list; - - local_irq_save(flags); - rdp->qlen -= count; - local_irq_restore(flags); - if (rdp->blimit == INT_MAX && rdp->qlen <= qlowmark) - rdp->blimit = blimit; - - if (!rdp->donelist) - rdp->donetail = &rdp->donelist; - else - raise_rcu_softirq(); -} - -/* - * Grace period handling: - * The grace period handling consists out of two steps: - * - A new grace period is started. - * This is done by rcu_start_batch. The start is not broadcasted to - * all cpus, they must pick this up by comparing rcp->cur with - * rdp->quiescbatch. All cpus are recorded in the - * rcu_ctrlblk.cpumask bitmap. - * - All cpus must go through a quiescent state. - * Since the start of the grace period is not broadcasted, at least two - * calls to rcu_check_quiescent_state are required: - * The first call just notices that a new grace period is running. The - * following calls check if there was a quiescent state since the beginning - * of the grace period. If so, it updates rcu_ctrlblk.cpumask. If - * the bitmap is empty, then the grace period is completed. - * rcu_check_quiescent_state calls rcu_start_batch(0) to start the next grace - * period (if necessary). - */ - -/* - * Register a new batch of callbacks, and start it up if there is currently no - * active batch and the batch to be registered has not already occurred. - * Caller must hold rcu_ctrlblk.lock. - */ -static void rcu_start_batch(struct rcu_ctrlblk *rcp) -{ - if (rcp->cur != rcp->pending && - rcp->completed == rcp->cur) { - rcp->cur++; - record_gp_stall_check_time(rcp); - - /* - * Accessing nohz_cpu_mask before incrementing rcp->cur needs a - * Barrier Otherwise it can cause tickless idle CPUs to be - * included in rcp->cpumask, which will extend graceperiods - * unnecessarily. - */ - smp_mb(); - cpumask_andnot(to_cpumask(rcp->cpumask), - cpu_online_mask, nohz_cpu_mask); - - rcp->signaled = 0; - } -} - -/* - * cpu went through a quiescent state since the beginning of the grace period. - * Clear it from the cpu mask and complete the grace period if it was the last - * cpu. Start another grace period if someone has further entries pending - */ -static void cpu_quiet(int cpu, struct rcu_ctrlblk *rcp) -{ - cpumask_clear_cpu(cpu, to_cpumask(rcp->cpumask)); - if (cpumask_empty(to_cpumask(rcp->cpumask))) { - /* batch completed ! */ - rcp->completed = rcp->cur; - rcu_start_batch(rcp); - } -} - -/* - * Check if the cpu has gone through a quiescent state (say context - * switch). If so and if it already hasn't done so in this RCU - * quiescent cycle, then indicate that it has done so. - */ -static void rcu_check_quiescent_state(struct rcu_ctrlblk *rcp, - struct rcu_data *rdp) -{ - unsigned long flags; - - if (rdp->quiescbatch != rcp->cur) { - /* start new grace period: */ - rdp->qs_pending = 1; - rdp->passed_quiesc = 0; - rdp->quiescbatch = rcp->cur; - return; - } - - /* Grace period already completed for this cpu? - * qs_pending is checked instead of the actual bitmap to avoid - * cacheline trashing. - */ - if (!rdp->qs_pending) - return; - - /* - * Was there a quiescent state since the beginning of the grace - * period? If no, then exit and wait for the next call. - */ - if (!rdp->passed_quiesc) - return; - rdp->qs_pending = 0; - - spin_lock_irqsave(&rcp->lock, flags); - /* - * rdp->quiescbatch/rcp->cur and the cpu bitmap can come out of sync - * during cpu startup. Ignore the quiescent state. - */ - if (likely(rdp->quiescbatch == rcp->cur)) - cpu_quiet(rdp->cpu, rcp); - - spin_unlock_irqrestore(&rcp->lock, flags); -} - - -#ifdef CONFIG_HOTPLUG_CPU - -/* warning! helper for rcu_offline_cpu. do not use elsewhere without reviewing - * locking requirements, the list it's pulling from has to belong to a cpu - * which is dead and hence not processing interrupts. - */ -static void rcu_move_batch(struct rcu_data *this_rdp, struct rcu_head *list, - struct rcu_head **tail, long batch) -{ - unsigned long flags; - - if (list) { - local_irq_save(flags); - this_rdp->batch = batch; - *this_rdp->nxttail[2] = list; - this_rdp->nxttail[2] = tail; - local_irq_restore(flags); - } -} - -static void __rcu_offline_cpu(struct rcu_data *this_rdp, - struct rcu_ctrlblk *rcp, struct rcu_data *rdp) -{ - unsigned long flags; - - /* - * if the cpu going offline owns the grace period - * we can block indefinitely waiting for it, so flush - * it here - */ - spin_lock_irqsave(&rcp->lock, flags); - if (rcp->cur != rcp->completed) - cpu_quiet(rdp->cpu, rcp); - rcu_move_batch(this_rdp, rdp->donelist, rdp->donetail, rcp->cur + 1); - rcu_move_batch(this_rdp, rdp->nxtlist, rdp->nxttail[2], rcp->cur + 1); - spin_unlock(&rcp->lock); - - this_rdp->qlen += rdp->qlen; - local_irq_restore(flags); -} - -static void rcu_offline_cpu(int cpu) -{ - struct rcu_data *this_rdp = &get_cpu_var(rcu_data); - struct rcu_data *this_bh_rdp = &get_cpu_var(rcu_bh_data); - - __rcu_offline_cpu(this_rdp, &rcu_ctrlblk, - &per_cpu(rcu_data, cpu)); - __rcu_offline_cpu(this_bh_rdp, &rcu_bh_ctrlblk, - &per_cpu(rcu_bh_data, cpu)); - put_cpu_var(rcu_data); - put_cpu_var(rcu_bh_data); -} - -#else - -static void rcu_offline_cpu(int cpu) -{ -} - -#endif - -/* - * This does the RCU processing work from softirq context. - */ -static void __rcu_process_callbacks(struct rcu_ctrlblk *rcp, - struct rcu_data *rdp) -{ - unsigned long flags; - long completed_snap; - - if (rdp->nxtlist) { - local_irq_save(flags); - completed_snap = ACCESS_ONCE(rcp->completed); - - /* - * move the other grace-period-completed entries to - * [rdp->nxtlist, *rdp->nxttail[0]) temporarily - */ - if (!rcu_batch_before(completed_snap, rdp->batch)) - rdp->nxttail[0] = rdp->nxttail[1] = rdp->nxttail[2]; - else if (!rcu_batch_before(completed_snap, rdp->batch - 1)) - rdp->nxttail[0] = rdp->nxttail[1]; - - /* - * the grace period for entries in - * [rdp->nxtlist, *rdp->nxttail[0]) has completed and - * move these entries to donelist - */ - if (rdp->nxttail[0] != &rdp->nxtlist) { - *rdp->donetail = rdp->nxtlist; - rdp->donetail = rdp->nxttail[0]; - rdp->nxtlist = *rdp->nxttail[0]; - *rdp->donetail = NULL; - - if (rdp->nxttail[1] == rdp->nxttail[0]) - rdp->nxttail[1] = &rdp->nxtlist; - if (rdp->nxttail[2] == rdp->nxttail[0]) - rdp->nxttail[2] = &rdp->nxtlist; - rdp->nxttail[0] = &rdp->nxtlist; - } - - local_irq_restore(flags); - - if (rcu_batch_after(rdp->batch, rcp->pending)) { - unsigned long flags2; - - /* and start it/schedule start if it's a new batch */ - spin_lock_irqsave(&rcp->lock, flags2); - if (rcu_batch_after(rdp->batch, rcp->pending)) { - rcp->pending = rdp->batch; - rcu_start_batch(rcp); - } - spin_unlock_irqrestore(&rcp->lock, flags2); - } - } - - rcu_check_quiescent_state(rcp, rdp); - if (rdp->donelist) - rcu_do_batch(rdp); -} - -static void rcu_process_callbacks(struct softirq_action *unused) -{ - /* - * Memory references from any prior RCU read-side critical sections - * executed by the interrupted code must be see before any RCU - * grace-period manupulations below. - */ - - smp_mb(); /* See above block comment. */ - - __rcu_process_callbacks(&rcu_ctrlblk, &__get_cpu_var(rcu_data)); - __rcu_process_callbacks(&rcu_bh_ctrlblk, &__get_cpu_var(rcu_bh_data)); - - /* - * Memory references from any later RCU read-side critical sections - * executed by the interrupted code must be see after any RCU - * grace-period manupulations above. - */ - - smp_mb(); /* See above block comment. */ -} - -static int __rcu_pending(struct rcu_ctrlblk *rcp, struct rcu_data *rdp) -{ - /* Check for CPU stalls, if enabled. */ - check_cpu_stall(rcp); - - if (rdp->nxtlist) { - long completed_snap = ACCESS_ONCE(rcp->completed); - - /* - * This cpu has pending rcu entries and the grace period - * for them has completed. - */ - if (!rcu_batch_before(completed_snap, rdp->batch)) - return 1; - if (!rcu_batch_before(completed_snap, rdp->batch - 1) && - rdp->nxttail[0] != rdp->nxttail[1]) - return 1; - if (rdp->nxttail[0] != &rdp->nxtlist) - return 1; - - /* - * This cpu has pending rcu entries and the new batch - * for then hasn't been started nor scheduled start - */ - if (rcu_batch_after(rdp->batch, rcp->pending)) - return 1; - } - - /* This cpu has finished callbacks to invoke */ - if (rdp->donelist) - return 1; - - /* The rcu core waits for a quiescent state from the cpu */ - if (rdp->quiescbatch != rcp->cur || rdp->qs_pending) - return 1; - - /* nothing to do */ - return 0; -} - -/* - * Check to see if there is any immediate RCU-related work to be done - * by the current CPU, returning 1 if so. This function is part of the - * RCU implementation; it is -not- an exported member of the RCU API. - */ -int rcu_pending(int cpu) -{ - return __rcu_pending(&rcu_ctrlblk, &per_cpu(rcu_data, cpu)) || - __rcu_pending(&rcu_bh_ctrlblk, &per_cpu(rcu_bh_data, cpu)); -} - -/* - * Check to see if any future RCU-related work will need to be done - * by the current CPU, even if none need be done immediately, returning - * 1 if so. This function is part of the RCU implementation; it is -not- - * an exported member of the RCU API. - */ -int rcu_needs_cpu(int cpu) -{ - struct rcu_data *rdp = &per_cpu(rcu_data, cpu); - struct rcu_data *rdp_bh = &per_cpu(rcu_bh_data, cpu); - - return !!rdp->nxtlist || !!rdp_bh->nxtlist || rcu_pending(cpu); -} - -/* - * Top-level function driving RCU grace-period detection, normally - * invoked from the scheduler-clock interrupt. This function simply - * increments counters that are read only from softirq by this same - * CPU, so there are no memory barriers required. - */ -void rcu_check_callbacks(int cpu, int user) -{ - if (user || - (idle_cpu(cpu) && rcu_scheduler_active && - !in_softirq() && hardirq_count() <= (1 << HARDIRQ_SHIFT))) { - - /* - * Get here if this CPU took its interrupt from user - * mode or from the idle loop, and if this is not a - * nested interrupt. In this case, the CPU is in - * a quiescent state, so count it. - * - * Also do a memory barrier. This is needed to handle - * the case where writes from a preempt-disable section - * of code get reordered into schedule() by this CPU's - * write buffer. The memory barrier makes sure that - * the rcu_qsctr_inc() and rcu_bh_qsctr_inc() are see - * by other CPUs to happen after any such write. - */ - - smp_mb(); /* See above block comment. */ - rcu_qsctr_inc(cpu); - rcu_bh_qsctr_inc(cpu); - - } else if (!in_softirq()) { - - /* - * Get here if this CPU did not take its interrupt from - * softirq, in other words, if it is not interrupting - * a rcu_bh read-side critical section. This is an _bh - * critical section, so count it. The memory barrier - * is needed for the same reason as is the above one. - */ - - smp_mb(); /* See above block comment. */ - rcu_bh_qsctr_inc(cpu); - } - raise_rcu_softirq(); -} - -static void __cpuinit rcu_init_percpu_data(int cpu, struct rcu_ctrlblk *rcp, - struct rcu_data *rdp) -{ - unsigned long flags; - - spin_lock_irqsave(&rcp->lock, flags); - memset(rdp, 0, sizeof(*rdp)); - rdp->nxttail[0] = rdp->nxttail[1] = rdp->nxttail[2] = &rdp->nxtlist; - rdp->donetail = &rdp->donelist; - rdp->quiescbatch = rcp->completed; - rdp->qs_pending = 0; - rdp->cpu = cpu; - rdp->blimit = blimit; - spin_unlock_irqrestore(&rcp->lock, flags); -} - -static void __cpuinit rcu_online_cpu(int cpu) -{ - struct rcu_data *rdp = &per_cpu(rcu_data, cpu); - struct rcu_data *bh_rdp = &per_cpu(rcu_bh_data, cpu); - - rcu_init_percpu_data(cpu, &rcu_ctrlblk, rdp); - rcu_init_percpu_data(cpu, &rcu_bh_ctrlblk, bh_rdp); - open_softirq(RCU_SOFTIRQ, rcu_process_callbacks); -} - -static int __cpuinit rcu_cpu_notify(struct notifier_block *self, - unsigned long action, void *hcpu) -{ - long cpu = (long)hcpu; - - switch (action) { - case CPU_UP_PREPARE: - case CPU_UP_PREPARE_FROZEN: - rcu_online_cpu(cpu); - break; - case CPU_DEAD: - case CPU_DEAD_FROZEN: - rcu_offline_cpu(cpu); - break; - default: - break; - } - return NOTIFY_OK; -} - -static struct notifier_block __cpuinitdata rcu_nb = { - .notifier_call = rcu_cpu_notify, -}; - -/* - * Initializes rcu mechanism. Assumed to be called early. - * That is before local timer(SMP) or jiffie timer (uniproc) is setup. - * Note that rcu_qsctr and friends are implicitly - * initialized due to the choice of ``0'' for RCU_CTR_INVALID. - */ -void __init __rcu_init(void) -{ -#ifdef CONFIG_RCU_CPU_STALL_DETECTOR - printk(KERN_INFO "RCU-based detection of stalled CPUs is enabled.\n"); -#endif /* #ifdef CONFIG_RCU_CPU_STALL_DETECTOR */ - rcu_cpu_notify(&rcu_nb, CPU_UP_PREPARE, - (void *)(long)smp_processor_id()); - /* Register notifier for non-boot CPUs */ - register_cpu_notifier(&rcu_nb); -} - -module_param(blimit, int, 0); -module_param(qhimark, int, 0); -module_param(qlowmark, int, 0); -- cgit v1.2.3 From 9e48858f7d36a6a3849f1d1b40c3bf5624b4ee7c Mon Sep 17 00:00:00 2001 From: Ingo Molnar Date: Thu, 7 May 2009 19:26:19 +1000 Subject: security: rename ptrace_may_access => ptrace_access_check The ->ptrace_may_access() methods are named confusingly - the real ptrace_may_access() returns a bool, while these security checks have a retval convention. Rename it to ptrace_access_check, to reduce the confusion factor. [ Impact: cleanup, no code changed ] Signed-off-by: Ingo Molnar Signed-off-by: James Morris --- include/linux/security.h | 14 +++++++------- kernel/ptrace.c | 2 +- security/capability.c | 2 +- security/commoncap.c | 4 ++-- security/security.c | 4 ++-- security/selinux/hooks.c | 6 +++--- security/smack/smack_lsm.c | 8 ++++---- 7 files changed, 20 insertions(+), 20 deletions(-) (limited to 'include') diff --git a/include/linux/security.h b/include/linux/security.h index 5eff459b3833..145909165dbf 100644 --- a/include/linux/security.h +++ b/include/linux/security.h @@ -52,7 +52,7 @@ struct audit_krule; extern int cap_capable(struct task_struct *tsk, const struct cred *cred, int cap, int audit); extern int cap_settime(struct timespec *ts, struct timezone *tz); -extern int cap_ptrace_may_access(struct task_struct *child, unsigned int mode); +extern int cap_ptrace_access_check(struct task_struct *child, unsigned int mode); extern int cap_ptrace_traceme(struct task_struct *parent); extern int cap_capget(struct task_struct *target, kernel_cap_t *effective, kernel_cap_t *inheritable, kernel_cap_t *permitted); extern int cap_capset(struct cred *new, const struct cred *old, @@ -1209,7 +1209,7 @@ static inline void security_free_mnt_opts(struct security_mnt_opts *opts) * @alter contains the flag indicating whether changes are to be made. * Return 0 if permission is granted. * - * @ptrace_may_access: + * @ptrace_access_check: * Check permission before allowing the current process to trace the * @child process. * Security modules may also want to perform a process tracing check @@ -1224,7 +1224,7 @@ static inline void security_free_mnt_opts(struct security_mnt_opts *opts) * Check that the @parent process has sufficient permission to trace the * current process before allowing the current process to present itself * to the @parent process for tracing. - * The parent process will still have to undergo the ptrace_may_access + * The parent process will still have to undergo the ptrace_access_check * checks before it is allowed to trace this one. * @parent contains the task_struct structure for debugger process. * Return 0 if permission is granted. @@ -1336,7 +1336,7 @@ static inline void security_free_mnt_opts(struct security_mnt_opts *opts) struct security_operations { char name[SECURITY_NAME_MAX + 1]; - int (*ptrace_may_access) (struct task_struct *child, unsigned int mode); + int (*ptrace_access_check) (struct task_struct *child, unsigned int mode); int (*ptrace_traceme) (struct task_struct *parent); int (*capget) (struct task_struct *target, kernel_cap_t *effective, @@ -1617,7 +1617,7 @@ extern int security_module_enable(struct security_operations *ops); extern int register_security(struct security_operations *ops); /* Security operations */ -int security_ptrace_may_access(struct task_struct *child, unsigned int mode); +int security_ptrace_access_check(struct task_struct *child, unsigned int mode); int security_ptrace_traceme(struct task_struct *parent); int security_capget(struct task_struct *target, kernel_cap_t *effective, @@ -1798,10 +1798,10 @@ static inline int security_init(void) return 0; } -static inline int security_ptrace_may_access(struct task_struct *child, +static inline int security_ptrace_access_check(struct task_struct *child, unsigned int mode) { - return cap_ptrace_may_access(child, mode); + return cap_ptrace_access_check(child, mode); } static inline int security_ptrace_traceme(struct task_struct *parent) diff --git a/kernel/ptrace.c b/kernel/ptrace.c index 61c78b2c07ba..9a4184e04f29 100644 --- a/kernel/ptrace.c +++ b/kernel/ptrace.c @@ -152,7 +152,7 @@ int __ptrace_may_access(struct task_struct *task, unsigned int mode) if (!dumpable && !capable(CAP_SYS_PTRACE)) return -EPERM; - return security_ptrace_may_access(task, mode); + return security_ptrace_access_check(task, mode); } bool ptrace_may_access(struct task_struct *task, unsigned int mode) diff --git a/security/capability.c b/security/capability.c index 21b6cead6a8e..f218dd361647 100644 --- a/security/capability.c +++ b/security/capability.c @@ -863,7 +863,7 @@ struct security_operations default_security_ops = { void security_fixup_ops(struct security_operations *ops) { - set_to_cap_if_null(ops, ptrace_may_access); + set_to_cap_if_null(ops, ptrace_access_check); set_to_cap_if_null(ops, ptrace_traceme); set_to_cap_if_null(ops, capget); set_to_cap_if_null(ops, capset); diff --git a/security/commoncap.c b/security/commoncap.c index 48b7e0228fa3..aa97704564d4 100644 --- a/security/commoncap.c +++ b/security/commoncap.c @@ -101,7 +101,7 @@ int cap_settime(struct timespec *ts, struct timezone *tz) } /** - * cap_ptrace_may_access - Determine whether the current process may access + * cap_ptrace_access_check - Determine whether the current process may access * another * @child: The process to be accessed * @mode: The mode of attachment. @@ -109,7 +109,7 @@ int cap_settime(struct timespec *ts, struct timezone *tz) * Determine whether a process may access another, returning 0 if permission * granted, -ve if denied. */ -int cap_ptrace_may_access(struct task_struct *child, unsigned int mode) +int cap_ptrace_access_check(struct task_struct *child, unsigned int mode) { int ret = 0; diff --git a/security/security.c b/security/security.c index dc7674fbfc7a..4501c5e1f988 100644 --- a/security/security.c +++ b/security/security.c @@ -124,9 +124,9 @@ int register_security(struct security_operations *ops) /* Security operations */ -int security_ptrace_may_access(struct task_struct *child, unsigned int mode) +int security_ptrace_access_check(struct task_struct *child, unsigned int mode) { - return security_ops->ptrace_may_access(child, mode); + return security_ops->ptrace_access_check(child, mode); } int security_ptrace_traceme(struct task_struct *parent) diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c index d6f64783acd1..e3b4f3083dd7 100644 --- a/security/selinux/hooks.c +++ b/security/selinux/hooks.c @@ -1854,12 +1854,12 @@ static inline u32 open_file_to_av(struct file *file) /* Hook functions begin here. */ -static int selinux_ptrace_may_access(struct task_struct *child, +static int selinux_ptrace_access_check(struct task_struct *child, unsigned int mode) { int rc; - rc = cap_ptrace_may_access(child, mode); + rc = cap_ptrace_access_check(child, mode); if (rc) return rc; @@ -5315,7 +5315,7 @@ static int selinux_key_getsecurity(struct key *key, char **_buffer) static struct security_operations selinux_ops = { .name = "selinux", - .ptrace_may_access = selinux_ptrace_may_access, + .ptrace_access_check = selinux_ptrace_access_check, .ptrace_traceme = selinux_ptrace_traceme, .capget = selinux_capget, .capset = selinux_capset, diff --git a/security/smack/smack_lsm.c b/security/smack/smack_lsm.c index 0023182078c7..1c9bdbcbe3d2 100644 --- a/security/smack/smack_lsm.c +++ b/security/smack/smack_lsm.c @@ -91,7 +91,7 @@ struct inode_smack *new_inode_smack(char *smack) */ /** - * smack_ptrace_may_access - Smack approval on PTRACE_ATTACH + * smack_ptrace_access_check - Smack approval on PTRACE_ATTACH * @ctp: child task pointer * @mode: ptrace attachment mode * @@ -99,13 +99,13 @@ struct inode_smack *new_inode_smack(char *smack) * * Do the capability checks, and require read and write. */ -static int smack_ptrace_may_access(struct task_struct *ctp, unsigned int mode) +static int smack_ptrace_access_check(struct task_struct *ctp, unsigned int mode) { int rc; struct smk_audit_info ad; char *sp, *tsp; - rc = cap_ptrace_may_access(ctp, mode); + rc = cap_ptrace_access_check(ctp, mode); if (rc != 0) return rc; @@ -3032,7 +3032,7 @@ static void smack_release_secctx(char *secdata, u32 seclen) struct security_operations smack_ops = { .name = "smack", - .ptrace_may_access = smack_ptrace_may_access, + .ptrace_access_check = smack_ptrace_access_check, .ptrace_traceme = smack_ptrace_traceme, .syslog = smack_syslog, -- cgit v1.2.3 From c7a1a4c80f873d5d6ecd173035bb80eba489f380 Mon Sep 17 00:00:00 2001 From: Rémi Denis-Courmont Date: Wed, 24 Jun 2009 01:07:44 +0000 Subject: Phonet: publicize the Netlink notification function MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: Rémi Denis-Courmont Signed-off-by: David S. Miller --- include/net/phonet/pn_dev.h | 1 + net/phonet/pn_netlink.c | 4 ++-- 2 files changed, 3 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/include/net/phonet/pn_dev.h b/include/net/phonet/pn_dev.h index 5054dc5ea2c2..29d126736611 100644 --- a/include/net/phonet/pn_dev.h +++ b/include/net/phonet/pn_dev.h @@ -45,6 +45,7 @@ int phonet_address_add(struct net_device *dev, u8 addr); int phonet_address_del(struct net_device *dev, u8 addr); u8 phonet_address_get(struct net_device *dev, u8 addr); int phonet_address_lookup(struct net *net, u8 addr); +void phonet_address_notify(int event, struct net_device *dev, u8 addr); #define PN_NO_ADDR 0xff diff --git a/net/phonet/pn_netlink.c b/net/phonet/pn_netlink.c index cec4e5951681..f8b4cee434c2 100644 --- a/net/phonet/pn_netlink.c +++ b/net/phonet/pn_netlink.c @@ -32,7 +32,7 @@ static int fill_addr(struct sk_buff *skb, struct net_device *dev, u8 addr, u32 pid, u32 seq, int event); -static void rtmsg_notify(int event, struct net_device *dev, u8 addr) +void phonet_address_notify(int event, struct net_device *dev, u8 addr) { struct sk_buff *skb; int err = -ENOBUFS; @@ -94,7 +94,7 @@ static int addr_doit(struct sk_buff *skb, struct nlmsghdr *nlh, void *attr) else err = phonet_address_del(dev, pnaddr); if (!err) - rtmsg_notify(nlh->nlmsg_type, dev, pnaddr); + phonet_address_notify(nlh->nlmsg_type, dev, pnaddr); return err; } -- cgit v1.2.3 From dcf52fb71d988ba945054308f661bddf9b2455fb Mon Sep 17 00:00:00 2001 From: Bjorn Helgaas Date: Mon, 22 Jun 2009 20:41:45 +0000 Subject: ACPI: remove unused acpi_device_ops .stop method No drivers use the .stop method, so remove it. Signed-off-by: Bjorn Helgaas Reviewed-by: Alex Chiang Signed-off-by: Len Brown --- drivers/acpi/scan.c | 5 ----- include/acpi/acpi_bus.h | 2 -- 2 files changed, 7 deletions(-) (limited to 'include') diff --git a/drivers/acpi/scan.c b/drivers/acpi/scan.c index 781435d7e369..4a89f081160f 100644 --- a/drivers/acpi/scan.c +++ b/drivers/acpi/scan.c @@ -426,9 +426,6 @@ static int acpi_device_probe(struct device * dev) if (acpi_drv->ops.notify) { ret = acpi_device_install_notify_handler(acpi_dev); if (ret) { - if (acpi_drv->ops.stop) - acpi_drv->ops.stop(acpi_dev, - acpi_dev->removal_type); if (acpi_drv->ops.remove) acpi_drv->ops.remove(acpi_dev, acpi_dev->removal_type); @@ -452,8 +449,6 @@ static int acpi_device_remove(struct device * dev) if (acpi_drv) { if (acpi_drv->ops.notify) acpi_device_remove_notify_handler(acpi_dev); - if (acpi_drv->ops.stop) - acpi_drv->ops.stop(acpi_dev, acpi_dev->removal_type); if (acpi_drv->ops.remove) acpi_drv->ops.remove(acpi_dev, acpi_dev->removal_type); } diff --git a/include/acpi/acpi_bus.h b/include/acpi/acpi_bus.h index c65e4ce6c3af..79a6c5ebe908 100644 --- a/include/acpi/acpi_bus.h +++ b/include/acpi/acpi_bus.h @@ -89,7 +89,6 @@ struct acpi_device; typedef int (*acpi_op_add) (struct acpi_device * device); typedef int (*acpi_op_remove) (struct acpi_device * device, int type); typedef int (*acpi_op_start) (struct acpi_device * device); -typedef int (*acpi_op_stop) (struct acpi_device * device, int type); typedef int (*acpi_op_suspend) (struct acpi_device * device, pm_message_t state); typedef int (*acpi_op_resume) (struct acpi_device * device); @@ -106,7 +105,6 @@ struct acpi_device_ops { acpi_op_add add; acpi_op_remove remove; acpi_op_start start; - acpi_op_stop stop; acpi_op_suspend suspend; acpi_op_resume resume; acpi_op_bind bind; -- cgit v1.2.3 From 6fdc03709433ccc2005f0f593ae9d9dd04f7b485 Mon Sep 17 00:00:00 2001 From: Stefan Richter Date: Sat, 20 Jun 2009 13:23:59 +0200 Subject: firewire: core: do not DMA-map stack addresses The DMA mapping API cannot map on-stack addresses, as explained in Documentation/DMA-mapping.txt. Convert the two cases of on-stack packet payload buffers in firewire-core (payload of lock requests in the bus manager work and in iso resource management) to slab-allocated memory. There are a number on-stack buffers for quadlet write or quadlet read requests in firewire-core and firewire-sbp2. These are harmless; they are copied to/ from card driver internal DMA buffers since quadlet payloads are inlined with packet headers. Signed-off-by: Stefan Richter --- drivers/firewire/core-card.c | 14 +++++++------- drivers/firewire/core-cdev.c | 4 +++- drivers/firewire/core-iso.c | 24 +++++++++++++----------- drivers/firewire/core.h | 3 ++- include/linux/firewire.h | 1 + 5 files changed, 26 insertions(+), 20 deletions(-) (limited to 'include') diff --git a/drivers/firewire/core-card.c b/drivers/firewire/core-card.c index 543fccac81bb..f74edae5cb4c 100644 --- a/drivers/firewire/core-card.c +++ b/drivers/firewire/core-card.c @@ -196,8 +196,8 @@ static void allocate_broadcast_channel(struct fw_card *card, int generation) { int channel, bandwidth = 0; - fw_iso_resource_manage(card, generation, 1ULL << 31, - &channel, &bandwidth, true); + fw_iso_resource_manage(card, generation, 1ULL << 31, &channel, + &bandwidth, true, card->bm_transaction_data); if (channel == 31) { card->broadcast_channel_allocated = true; device_for_each_child(card->device, (void *)(long)generation, @@ -230,7 +230,6 @@ static void fw_card_bm_work(struct work_struct *work) bool do_reset = false; bool root_device_is_running; bool root_device_is_cmc; - __be32 lock_data[2]; spin_lock_irqsave(&card->lock, flags); @@ -273,22 +272,23 @@ static void fw_card_bm_work(struct work_struct *work) goto pick_me; } - lock_data[0] = cpu_to_be32(0x3f); - lock_data[1] = cpu_to_be32(local_id); + card->bm_transaction_data[0] = cpu_to_be32(0x3f); + card->bm_transaction_data[1] = cpu_to_be32(local_id); spin_unlock_irqrestore(&card->lock, flags); rcode = fw_run_transaction(card, TCODE_LOCK_COMPARE_SWAP, irm_id, generation, SCODE_100, CSR_REGISTER_BASE + CSR_BUS_MANAGER_ID, - lock_data, sizeof(lock_data)); + card->bm_transaction_data, + sizeof(card->bm_transaction_data)); if (rcode == RCODE_GENERATION) /* Another bus reset, BM work has been rescheduled. */ goto out; if (rcode == RCODE_COMPLETE && - lock_data[0] != cpu_to_be32(0x3f)) { + card->bm_transaction_data[0] != cpu_to_be32(0x3f)) { /* Somebody else is BM. Only act as IRM. */ if (local_id == irm_id) diff --git a/drivers/firewire/core-cdev.c b/drivers/firewire/core-cdev.c index d1d30c615b0f..ced186d7e9a9 100644 --- a/drivers/firewire/core-cdev.c +++ b/drivers/firewire/core-cdev.c @@ -125,6 +125,7 @@ struct iso_resource { int generation; u64 channels; s32 bandwidth; + __be32 transaction_data[2]; struct iso_resource_event *e_alloc, *e_dealloc; }; @@ -1049,7 +1050,8 @@ static void iso_resource_work(struct work_struct *work) r->channels, &channel, &bandwidth, todo == ISO_RES_ALLOC || todo == ISO_RES_REALLOC || - todo == ISO_RES_ALLOC_ONCE); + todo == ISO_RES_ALLOC_ONCE, + r->transaction_data); /* * Is this generation outdated already? As long as this resource sticks * in the idr, it will be scheduled again for a newer generation or at diff --git a/drivers/firewire/core-iso.c b/drivers/firewire/core-iso.c index 166f19c6d38d..110e731f5574 100644 --- a/drivers/firewire/core-iso.c +++ b/drivers/firewire/core-iso.c @@ -177,9 +177,8 @@ EXPORT_SYMBOL(fw_iso_context_stop); */ static int manage_bandwidth(struct fw_card *card, int irm_id, int generation, - int bandwidth, bool allocate) + int bandwidth, bool allocate, __be32 data[2]) { - __be32 data[2]; int try, new, old = allocate ? BANDWIDTH_AVAILABLE_INITIAL : 0; /* @@ -215,9 +214,9 @@ static int manage_bandwidth(struct fw_card *card, int irm_id, int generation, } static int manage_channel(struct fw_card *card, int irm_id, int generation, - u32 channels_mask, u64 offset, bool allocate) + u32 channels_mask, u64 offset, bool allocate, __be32 data[2]) { - __be32 data[2], c, all, old; + __be32 c, all, old; int i, retry = 5; old = all = allocate ? cpu_to_be32(~0) : 0; @@ -260,7 +259,7 @@ static int manage_channel(struct fw_card *card, int irm_id, int generation, } static void deallocate_channel(struct fw_card *card, int irm_id, - int generation, int channel) + int generation, int channel, __be32 buffer[2]) { u32 mask; u64 offset; @@ -269,7 +268,7 @@ static void deallocate_channel(struct fw_card *card, int irm_id, offset = channel < 32 ? CSR_REGISTER_BASE + CSR_CHANNELS_AVAILABLE_HI : CSR_REGISTER_BASE + CSR_CHANNELS_AVAILABLE_LO; - manage_channel(card, irm_id, generation, mask, offset, false); + manage_channel(card, irm_id, generation, mask, offset, false, buffer); } /** @@ -298,7 +297,7 @@ static void deallocate_channel(struct fw_card *card, int irm_id, */ void fw_iso_resource_manage(struct fw_card *card, int generation, u64 channels_mask, int *channel, int *bandwidth, - bool allocate) + bool allocate, __be32 buffer[2]) { u32 channels_hi = channels_mask; /* channels 31...0 */ u32 channels_lo = channels_mask >> 32; /* channels 63...32 */ @@ -310,10 +309,12 @@ void fw_iso_resource_manage(struct fw_card *card, int generation, if (channels_hi) c = manage_channel(card, irm_id, generation, channels_hi, - CSR_REGISTER_BASE + CSR_CHANNELS_AVAILABLE_HI, allocate); + CSR_REGISTER_BASE + CSR_CHANNELS_AVAILABLE_HI, + allocate, buffer); if (channels_lo && c < 0) { c = manage_channel(card, irm_id, generation, channels_lo, - CSR_REGISTER_BASE + CSR_CHANNELS_AVAILABLE_LO, allocate); + CSR_REGISTER_BASE + CSR_CHANNELS_AVAILABLE_LO, + allocate, buffer); if (c >= 0) c += 32; } @@ -325,12 +326,13 @@ void fw_iso_resource_manage(struct fw_card *card, int generation, if (*bandwidth == 0) return; - ret = manage_bandwidth(card, irm_id, generation, *bandwidth, allocate); + ret = manage_bandwidth(card, irm_id, generation, *bandwidth, + allocate, buffer); if (ret < 0) *bandwidth = 0; if (allocate && ret < 0 && c >= 0) { - deallocate_channel(card, irm_id, generation, c); + deallocate_channel(card, irm_id, generation, c, buffer); *channel = ret; } } diff --git a/drivers/firewire/core.h b/drivers/firewire/core.h index c3cfc647e5e3..6052816be353 100644 --- a/drivers/firewire/core.h +++ b/drivers/firewire/core.h @@ -120,7 +120,8 @@ void fw_node_event(struct fw_card *card, struct fw_node *node, int event); int fw_iso_buffer_map(struct fw_iso_buffer *buffer, struct vm_area_struct *vma); void fw_iso_resource_manage(struct fw_card *card, int generation, - u64 channels_mask, int *channel, int *bandwidth, bool allocate); + u64 channels_mask, int *channel, int *bandwidth, + bool allocate, __be32 buffer[2]); /* -topology */ diff --git a/include/linux/firewire.h b/include/linux/firewire.h index 9823946adbc5..192d1e43c43c 100644 --- a/include/linux/firewire.h +++ b/include/linux/firewire.h @@ -127,6 +127,7 @@ struct fw_card { struct delayed_work work; int bm_retries; int bm_generation; + __be32 bm_transaction_data[2]; bool broadcast_channel_allocated; u32 broadcast_channel; -- cgit v1.2.3 From 9d73777e500929b71dcfed16eec05f6760e345a6 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Thu, 25 Jun 2009 11:58:55 +0200 Subject: clarify get_user_pages() prototype Currently the 4th parameter of get_user_pages() is called len, but its in pages, not bytes. Rename the thing to nr_pages to avoid future confusion. Signed-off-by: Peter Zijlstra Signed-off-by: Linus Torvalds --- include/linux/mm.h | 2 +- mm/memory.c | 26 ++++++++++++-------------- mm/nommu.c | 12 +++++------- 3 files changed, 18 insertions(+), 22 deletions(-) (limited to 'include') diff --git a/include/linux/mm.h b/include/linux/mm.h index d006e93d5c93..ba3a7cb1eaa0 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -826,7 +826,7 @@ extern int make_pages_present(unsigned long addr, unsigned long end); extern int access_process_vm(struct task_struct *tsk, unsigned long addr, void *buf, int len, int write); int get_user_pages(struct task_struct *tsk, struct mm_struct *mm, - unsigned long start, int len, int write, int force, + unsigned long start, int nr_pages, int write, int force, struct page **pages, struct vm_area_struct **vmas); int get_user_pages_fast(unsigned long start, int nr_pages, int write, struct page **pages); diff --git a/mm/memory.c b/mm/memory.c index f46ac18ba231..65216194eb8d 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -1207,8 +1207,8 @@ static inline int use_zero_page(struct vm_area_struct *vma) int __get_user_pages(struct task_struct *tsk, struct mm_struct *mm, - unsigned long start, int len, int flags, - struct page **pages, struct vm_area_struct **vmas) + unsigned long start, int nr_pages, int flags, + struct page **pages, struct vm_area_struct **vmas) { int i; unsigned int vm_flags = 0; @@ -1217,7 +1217,7 @@ int __get_user_pages(struct task_struct *tsk, struct mm_struct *mm, int ignore = !!(flags & GUP_FLAGS_IGNORE_VMA_PERMISSIONS); int ignore_sigkill = !!(flags & GUP_FLAGS_IGNORE_SIGKILL); - if (len <= 0) + if (nr_pages <= 0) return 0; /* * Require read or write permissions. @@ -1269,7 +1269,7 @@ int __get_user_pages(struct task_struct *tsk, struct mm_struct *mm, vmas[i] = gate_vma; i++; start += PAGE_SIZE; - len--; + nr_pages--; continue; } @@ -1280,7 +1280,7 @@ int __get_user_pages(struct task_struct *tsk, struct mm_struct *mm, if (is_vm_hugetlb_page(vma)) { i = follow_hugetlb_page(mm, vma, pages, vmas, - &start, &len, i, write); + &start, &nr_pages, i, write); continue; } @@ -1357,9 +1357,9 @@ int __get_user_pages(struct task_struct *tsk, struct mm_struct *mm, vmas[i] = vma; i++; start += PAGE_SIZE; - len--; - } while (len && start < vma->vm_end); - } while (len); + nr_pages--; + } while (nr_pages && start < vma->vm_end); + } while (nr_pages); return i; } @@ -1368,7 +1368,7 @@ int __get_user_pages(struct task_struct *tsk, struct mm_struct *mm, * @tsk: task_struct of target task * @mm: mm_struct of target mm * @start: starting user address - * @len: number of pages from start to pin + * @nr_pages: number of pages from start to pin * @write: whether pages will be written to by the caller * @force: whether to force write access even if user mapping is * readonly. This will result in the page being COWed even @@ -1380,7 +1380,7 @@ int __get_user_pages(struct task_struct *tsk, struct mm_struct *mm, * Or NULL if the caller does not require them. * * Returns number of pages pinned. This may be fewer than the number - * requested. If len is 0 or negative, returns 0. If no pages + * requested. If nr_pages is 0 or negative, returns 0. If no pages * were pinned, returns -errno. Each page returned must be released * with a put_page() call when it is finished with. vmas will only * remain valid while mmap_sem is held. @@ -1414,7 +1414,7 @@ int __get_user_pages(struct task_struct *tsk, struct mm_struct *mm, * See also get_user_pages_fast, for performance critical applications. */ int get_user_pages(struct task_struct *tsk, struct mm_struct *mm, - unsigned long start, int len, int write, int force, + unsigned long start, int nr_pages, int write, int force, struct page **pages, struct vm_area_struct **vmas) { int flags = 0; @@ -1424,9 +1424,7 @@ int get_user_pages(struct task_struct *tsk, struct mm_struct *mm, if (force) flags |= GUP_FLAGS_FORCE; - return __get_user_pages(tsk, mm, - start, len, flags, - pages, vmas); + return __get_user_pages(tsk, mm, start, nr_pages, flags, pages, vmas); } EXPORT_SYMBOL(get_user_pages); diff --git a/mm/nommu.c b/mm/nommu.c index 2fd2ad5da98e..bf0cc762a7d2 100644 --- a/mm/nommu.c +++ b/mm/nommu.c @@ -173,8 +173,8 @@ unsigned int kobjsize(const void *objp) } int __get_user_pages(struct task_struct *tsk, struct mm_struct *mm, - unsigned long start, int len, int flags, - struct page **pages, struct vm_area_struct **vmas) + unsigned long start, int nr_pages, int flags, + struct page **pages, struct vm_area_struct **vmas) { struct vm_area_struct *vma; unsigned long vm_flags; @@ -189,7 +189,7 @@ int __get_user_pages(struct task_struct *tsk, struct mm_struct *mm, vm_flags = write ? (VM_WRITE | VM_MAYWRITE) : (VM_READ | VM_MAYREAD); vm_flags &= force ? (VM_MAYREAD | VM_MAYWRITE) : (VM_READ | VM_WRITE); - for (i = 0; i < len; i++) { + for (i = 0; i < nr_pages; i++) { vma = find_vma(mm, start); if (!vma) goto finish_or_fault; @@ -224,7 +224,7 @@ finish_or_fault: * - don't permit access to VMAs that don't support it, such as I/O mappings */ int get_user_pages(struct task_struct *tsk, struct mm_struct *mm, - unsigned long start, int len, int write, int force, + unsigned long start, int nr_pages, int write, int force, struct page **pages, struct vm_area_struct **vmas) { int flags = 0; @@ -234,9 +234,7 @@ int get_user_pages(struct task_struct *tsk, struct mm_struct *mm, if (force) flags |= GUP_FLAGS_FORCE; - return __get_user_pages(tsk, mm, - start, len, flags, - pages, vmas); + return __get_user_pages(tsk, mm, start, nr_pages, flags, pages, vmas); } EXPORT_SYMBOL(get_user_pages); -- cgit v1.2.3 From 41f95331b972a039f519ae0c70f051b7121f7346 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Tue, 23 Jun 2009 17:55:18 +0200 Subject: perf_counter: Split the mmap control page in two parts Since there are two distinct sections to the control page, move them apart so that possible extentions don't overlap. Signed-off-by: Peter Zijlstra LKML-Reference: Signed-off-by: Ingo Molnar --- include/linux/perf_counter.h | 6 ++++++ 1 file changed, 6 insertions(+) (limited to 'include') diff --git a/include/linux/perf_counter.h b/include/linux/perf_counter.h index e7213e46cf9c..489d5cbfbcca 100644 --- a/include/linux/perf_counter.h +++ b/include/linux/perf_counter.h @@ -233,6 +233,12 @@ struct perf_counter_mmap_page { __u32 index; /* hardware counter identifier */ __s64 offset; /* add to hardware counter value */ + /* + * Hole for extension of the self monitor capabilities + */ + + __u64 __reserved[125]; /* align to 1k */ + /* * Control data for the mmap() data buffer. * -- cgit v1.2.3 From 7f8b4e4e0988dadfd22330fd147ad2453e19f510 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Mon, 22 Jun 2009 14:34:35 +0200 Subject: perf_counter: Add scale information to the mmap control page Add the needed time scale to the self-profile mmap information. Signed-off-by: Peter Zijlstra LKML-Reference: Signed-off-by: Ingo Molnar --- include/linux/perf_counter.h | 4 +++- kernel/perf_counter.c | 6 ++++++ 2 files changed, 9 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/perf_counter.h b/include/linux/perf_counter.h index 489d5cbfbcca..bcbf1c43ed42 100644 --- a/include/linux/perf_counter.h +++ b/include/linux/perf_counter.h @@ -232,12 +232,14 @@ struct perf_counter_mmap_page { __u32 lock; /* seqlock for synchronization */ __u32 index; /* hardware counter identifier */ __s64 offset; /* add to hardware counter value */ + __u64 time_enabled; /* time counter active */ + __u64 time_running; /* time counter on cpu */ /* * Hole for extension of the self monitor capabilities */ - __u64 __reserved[125]; /* align to 1k */ + __u64 __reserved[123]; /* align to 1k */ /* * Control data for the mmap() data buffer. diff --git a/kernel/perf_counter.c b/kernel/perf_counter.c index c2b19c111718..23614adab475 100644 --- a/kernel/perf_counter.c +++ b/kernel/perf_counter.c @@ -1782,6 +1782,12 @@ void perf_counter_update_userpage(struct perf_counter *counter) if (counter->state == PERF_COUNTER_STATE_ACTIVE) userpg->offset -= atomic64_read(&counter->hw.prev_count); + userpg->time_enabled = counter->total_time_enabled + + atomic64_read(&counter->child_total_time_enabled); + + userpg->time_running = counter->total_time_running + + atomic64_read(&counter->child_total_time_running); + barrier(); ++userpg->lock; preempt_enable(); -- cgit v1.2.3 From 38b200d67636a30cb8dc1508137908e7a649b5c9 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Tue, 23 Jun 2009 20:13:11 +0200 Subject: perf_counter: Add PERF_EVENT_READ Provide a read() like event which can be used to log the counter value at specific sites such as child->parent folding on exit. In order to be useful, we log the counter parent ID, not the actual counter ID, since userspace can only relate parent IDs to perf_counter_attr constructs. Signed-off-by: Peter Zijlstra LKML-Reference: Signed-off-by: Ingo Molnar --- include/linux/perf_counter.h | 12 ++++++++ kernel/perf_counter.c | 72 +++++++++++++++++++++++++++++++++++++++++--- 2 files changed, 80 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/include/linux/perf_counter.h b/include/linux/perf_counter.h index bcbf1c43ed42..6a384f04755a 100644 --- a/include/linux/perf_counter.h +++ b/include/linux/perf_counter.h @@ -334,6 +334,18 @@ enum perf_event_type { */ PERF_EVENT_FORK = 7, + /* + * struct { + * struct perf_event_header header; + * u32 pid, tid; + * u64 value; + * { u64 time_enabled; } && PERF_FORMAT_ENABLED + * { u64 time_running; } && PERF_FORMAT_RUNNING + * { u64 parent_id; } && PERF_FORMAT_ID + * }; + */ + PERF_EVENT_READ = 8, + /* * When header.misc & PERF_EVENT_MISC_OVERFLOW the event_type field * will be PERF_SAMPLE_* diff --git a/kernel/perf_counter.c b/kernel/perf_counter.c index 02994a719e27..a72c20e91953 100644 --- a/kernel/perf_counter.c +++ b/kernel/perf_counter.c @@ -2623,6 +2623,66 @@ static void perf_counter_output(struct perf_counter *counter, int nmi, perf_output_end(&handle); } +/* + * read event + */ + +struct perf_read_event { + struct perf_event_header header; + + u32 pid; + u32 tid; + u64 value; + u64 format[3]; +}; + +static void +perf_counter_read_event(struct perf_counter *counter, + struct task_struct *task) +{ + struct perf_output_handle handle; + struct perf_read_event event = { + .header = { + .type = PERF_EVENT_READ, + .misc = 0, + .size = sizeof(event) - sizeof(event.format), + }, + .pid = perf_counter_pid(counter, task), + .tid = perf_counter_tid(counter, task), + .value = atomic64_read(&counter->count), + }; + int ret, i = 0; + + if (counter->attr.read_format & PERF_FORMAT_TOTAL_TIME_ENABLED) { + event.header.size += sizeof(u64); + event.format[i++] = counter->total_time_enabled; + } + + if (counter->attr.read_format & PERF_FORMAT_TOTAL_TIME_RUNNING) { + event.header.size += sizeof(u64); + event.format[i++] = counter->total_time_running; + } + + if (counter->attr.read_format & PERF_FORMAT_ID) { + u64 id; + + event.header.size += sizeof(u64); + if (counter->parent) + id = counter->parent->id; + else + id = counter->id; + + event.format[i++] = id; + } + + ret = perf_output_begin(&handle, counter, event.header.size, 0, 0); + if (ret) + return; + + perf_output_copy(&handle, &event, event.header.size); + perf_output_end(&handle); +} + /* * fork tracking */ @@ -3985,10 +4045,13 @@ static int inherit_group(struct perf_counter *parent_counter, } static void sync_child_counter(struct perf_counter *child_counter, - struct perf_counter *parent_counter) + struct task_struct *child) { + struct perf_counter *parent_counter = child_counter->parent; u64 child_val; + perf_counter_read_event(child_counter, child); + child_val = atomic64_read(&child_counter->count); /* @@ -4017,7 +4080,8 @@ static void sync_child_counter(struct perf_counter *child_counter, static void __perf_counter_exit_task(struct perf_counter *child_counter, - struct perf_counter_context *child_ctx) + struct perf_counter_context *child_ctx, + struct task_struct *child) { struct perf_counter *parent_counter; @@ -4031,7 +4095,7 @@ __perf_counter_exit_task(struct perf_counter *child_counter, * counters need to be zapped - but otherwise linger. */ if (parent_counter) { - sync_child_counter(child_counter, parent_counter); + sync_child_counter(child_counter, child); free_counter(child_counter); } } @@ -4093,7 +4157,7 @@ void perf_counter_exit_task(struct task_struct *child) again: list_for_each_entry_safe(child_counter, tmp, &child_ctx->counter_list, list_entry) - __perf_counter_exit_task(child_counter, child_ctx); + __perf_counter_exit_task(child_counter, child_ctx, child); /* * If the last counter was a group counter, it will have appended all -- cgit v1.2.3 From bfbd3381e63aa2a14c6706afb50ce4630aa0d9a2 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Wed, 24 Jun 2009 21:11:59 +0200 Subject: perf_counter: Implement more accurate per task statistics With the introduction of PERF_EVENT_READ we have the possibility to provide accurate counter values for individual tasks in a task hierarchy. However, due to the lazy context switching used for similar counter contexts our current per task counts are way off. In order to maintain some of the lazy switch benefits we don't disable it out-right, but simply iterate the active counters and flip the values between the contexts. This only reads the counters but does not need to reprogram the full PMU. Signed-off-by: Peter Zijlstra LKML-Reference: Signed-off-by: Ingo Molnar --- include/linux/perf_counter.h | 4 ++- kernel/perf_counter.c | 83 ++++++++++++++++++++++++++++++++++++++++++-- 2 files changed, 83 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/include/linux/perf_counter.h b/include/linux/perf_counter.h index 6a384f04755a..de70a10b5ec8 100644 --- a/include/linux/perf_counter.h +++ b/include/linux/perf_counter.h @@ -178,8 +178,9 @@ struct perf_counter_attr { mmap : 1, /* include mmap data */ comm : 1, /* include comm data */ freq : 1, /* use freq, not period */ + inherit_stat : 1, /* per task counts */ - __reserved_1 : 53; + __reserved_1 : 52; __u32 wakeup_events; /* wakeup every n events */ __u32 __reserved_2; @@ -602,6 +603,7 @@ struct perf_counter_context { int nr_counters; int nr_active; int is_active; + int nr_stat; atomic_t refcount; struct task_struct *task; diff --git a/kernel/perf_counter.c b/kernel/perf_counter.c index a72c20e91953..385ca51c6e60 100644 --- a/kernel/perf_counter.c +++ b/kernel/perf_counter.c @@ -236,6 +236,8 @@ list_add_counter(struct perf_counter *counter, struct perf_counter_context *ctx) list_add_rcu(&counter->event_entry, &ctx->event_list); ctx->nr_counters++; + if (counter->attr.inherit_stat) + ctx->nr_stat++; } /* @@ -250,6 +252,8 @@ list_del_counter(struct perf_counter *counter, struct perf_counter_context *ctx) if (list_empty(&counter->list_entry)) return; ctx->nr_counters--; + if (counter->attr.inherit_stat) + ctx->nr_stat--; list_del_init(&counter->list_entry); list_del_rcu(&counter->event_entry); @@ -1006,6 +1010,76 @@ static int context_equiv(struct perf_counter_context *ctx1, && !ctx1->pin_count && !ctx2->pin_count; } +static void __perf_counter_read(void *counter); + +static void __perf_counter_sync_stat(struct perf_counter *counter, + struct perf_counter *next_counter) +{ + u64 value; + + if (!counter->attr.inherit_stat) + return; + + /* + * Update the counter value, we cannot use perf_counter_read() + * because we're in the middle of a context switch and have IRQs + * disabled, which upsets smp_call_function_single(), however + * we know the counter must be on the current CPU, therefore we + * don't need to use it. + */ + switch (counter->state) { + case PERF_COUNTER_STATE_ACTIVE: + __perf_counter_read(counter); + break; + + case PERF_COUNTER_STATE_INACTIVE: + update_counter_times(counter); + break; + + default: + break; + } + + /* + * In order to keep per-task stats reliable we need to flip the counter + * values when we flip the contexts. + */ + value = atomic64_read(&next_counter->count); + value = atomic64_xchg(&counter->count, value); + atomic64_set(&next_counter->count, value); + + /* + * XXX also sync time_enabled and time_running ? + */ +} + +#define list_next_entry(pos, member) \ + list_entry(pos->member.next, typeof(*pos), member) + +static void perf_counter_sync_stat(struct perf_counter_context *ctx, + struct perf_counter_context *next_ctx) +{ + struct perf_counter *counter, *next_counter; + + if (!ctx->nr_stat) + return; + + counter = list_first_entry(&ctx->event_list, + struct perf_counter, event_entry); + + next_counter = list_first_entry(&next_ctx->event_list, + struct perf_counter, event_entry); + + while (&counter->event_entry != &ctx->event_list && + &next_counter->event_entry != &next_ctx->event_list) { + + __perf_counter_sync_stat(counter, next_counter); + + counter = list_next_entry(counter, event_entry); + next_counter = list_next_entry(counter, event_entry); + } +} + /* * Called from scheduler to remove the counters of the current task, * with interrupts disabled. @@ -1061,6 +1135,8 @@ void perf_counter_task_sched_out(struct task_struct *task, ctx->task = next; next_ctx->task = task; do_switch = 0; + + perf_counter_sync_stat(ctx, next_ctx); } spin_unlock(&next_ctx->lock); spin_unlock(&ctx->lock); @@ -1350,7 +1426,7 @@ void perf_counter_task_tick(struct task_struct *curr, int cpu) /* * Cross CPU call to read the hardware counter */ -static void __read(void *info) +static void __perf_counter_read(void *info) { struct perf_counter *counter = info; struct perf_counter_context *ctx = counter->ctx; @@ -1372,7 +1448,7 @@ static u64 perf_counter_read(struct perf_counter *counter) */ if (counter->state == PERF_COUNTER_STATE_ACTIVE) { smp_call_function_single(counter->oncpu, - __read, counter, 1); + __perf_counter_read, counter, 1); } else if (counter->state == PERF_COUNTER_STATE_INACTIVE) { update_counter_times(counter); } @@ -4050,7 +4126,8 @@ static void sync_child_counter(struct perf_counter *child_counter, struct perf_counter *parent_counter = child_counter->parent; u64 child_val; - perf_counter_read_event(child_counter, child); + if (child_counter->attr.inherit_stat) + perf_counter_read_event(child_counter, child); child_val = atomic64_read(&child_counter->count); -- cgit v1.2.3 From e6e18ec79b023d5fe84226cef533cf0e3770ce93 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Thu, 25 Jun 2009 11:27:12 +0200 Subject: perf_counter: Rework the sample ABI The PERF_EVENT_READ implementation made me realize we don't actually need the sample_type int the output sample, since we already have that in the perf_counter_attr information. Therefore, remove the PERF_EVENT_MISC_OVERFLOW bit and the event->type overloading, and imply put counter overflow samples in a PERF_EVENT_SAMPLE type. This also fixes the issue that event->type was only 32-bit and sample_type had 64 usable bits. Signed-off-by: Peter Zijlstra LKML-Reference: Signed-off-by: Ingo Molnar --- include/linux/perf_counter.h | 10 +++++----- kernel/perf_counter.c | 36 +++++++++++++++--------------------- tools/perf/builtin-annotate.c | 8 ++++---- tools/perf/builtin-report.c | 32 +++++++++++++++++++------------- tools/perf/builtin-top.c | 11 ++++++----- 5 files changed, 49 insertions(+), 48 deletions(-) (limited to 'include') diff --git a/include/linux/perf_counter.h b/include/linux/perf_counter.h index de70a10b5ec8..3078e23c91eb 100644 --- a/include/linux/perf_counter.h +++ b/include/linux/perf_counter.h @@ -262,7 +262,6 @@ struct perf_counter_mmap_page { #define PERF_EVENT_MISC_KERNEL (1 << 0) #define PERF_EVENT_MISC_USER (2 << 0) #define PERF_EVENT_MISC_HYPERVISOR (3 << 0) -#define PERF_EVENT_MISC_OVERFLOW (1 << 2) struct perf_event_header { __u32 type; @@ -348,9 +347,6 @@ enum perf_event_type { PERF_EVENT_READ = 8, /* - * When header.misc & PERF_EVENT_MISC_OVERFLOW the event_type field - * will be PERF_SAMPLE_* - * * struct { * struct perf_event_header header; * @@ -358,8 +354,9 @@ enum perf_event_type { * { u32 pid, tid; } && PERF_SAMPLE_TID * { u64 time; } && PERF_SAMPLE_TIME * { u64 addr; } && PERF_SAMPLE_ADDR - * { u64 config; } && PERF_SAMPLE_CONFIG + * { u64 id; } && PERF_SAMPLE_ID * { u32 cpu, res; } && PERF_SAMPLE_CPU + * { u64 period; } && PERF_SAMPLE_PERIOD * * { u64 nr; * { u64 id, val; } cnt[nr]; } && PERF_SAMPLE_GROUP @@ -368,6 +365,9 @@ enum perf_event_type { * u64 ips[nr]; } && PERF_SAMPLE_CALLCHAIN * }; */ + PERF_EVENT_SAMPLE = 9, + + PERF_EVENT_MAX, /* non-ABI */ }; enum perf_callchain_context { diff --git a/kernel/perf_counter.c b/kernel/perf_counter.c index 385ca51c6e60..f2f232696587 100644 --- a/kernel/perf_counter.c +++ b/kernel/perf_counter.c @@ -2575,15 +2575,14 @@ static void perf_counter_output(struct perf_counter *counter, int nmi, u32 cpu, reserved; } cpu_entry; - header.type = 0; + header.type = PERF_EVENT_SAMPLE; header.size = sizeof(header); - header.misc = PERF_EVENT_MISC_OVERFLOW; + header.misc = 0; header.misc |= perf_misc_flags(data->regs); if (sample_type & PERF_SAMPLE_IP) { ip = perf_instruction_pointer(data->regs); - header.type |= PERF_SAMPLE_IP; header.size += sizeof(ip); } @@ -2592,7 +2591,6 @@ static void perf_counter_output(struct perf_counter *counter, int nmi, tid_entry.pid = perf_counter_pid(counter, current); tid_entry.tid = perf_counter_tid(counter, current); - header.type |= PERF_SAMPLE_TID; header.size += sizeof(tid_entry); } @@ -2602,34 +2600,25 @@ static void perf_counter_output(struct perf_counter *counter, int nmi, */ time = sched_clock(); - header.type |= PERF_SAMPLE_TIME; header.size += sizeof(u64); } - if (sample_type & PERF_SAMPLE_ADDR) { - header.type |= PERF_SAMPLE_ADDR; + if (sample_type & PERF_SAMPLE_ADDR) header.size += sizeof(u64); - } - if (sample_type & PERF_SAMPLE_ID) { - header.type |= PERF_SAMPLE_ID; + if (sample_type & PERF_SAMPLE_ID) header.size += sizeof(u64); - } if (sample_type & PERF_SAMPLE_CPU) { - header.type |= PERF_SAMPLE_CPU; header.size += sizeof(cpu_entry); cpu_entry.cpu = raw_smp_processor_id(); } - if (sample_type & PERF_SAMPLE_PERIOD) { - header.type |= PERF_SAMPLE_PERIOD; + if (sample_type & PERF_SAMPLE_PERIOD) header.size += sizeof(u64); - } if (sample_type & PERF_SAMPLE_GROUP) { - header.type |= PERF_SAMPLE_GROUP; header.size += sizeof(u64) + counter->nr_siblings * sizeof(group_entry); } @@ -2639,10 +2628,9 @@ static void perf_counter_output(struct perf_counter *counter, int nmi, if (callchain) { callchain_size = (1 + callchain->nr) * sizeof(u64); - - header.type |= PERF_SAMPLE_CALLCHAIN; header.size += callchain_size; - } + } else + header.size += sizeof(u64); } ret = perf_output_begin(&handle, counter, header.size, nmi, 1); @@ -2693,8 +2681,14 @@ static void perf_counter_output(struct perf_counter *counter, int nmi, } } - if (callchain) - perf_output_copy(&handle, callchain, callchain_size); + if (sample_type & PERF_SAMPLE_CALLCHAIN) { + if (callchain) + perf_output_copy(&handle, callchain, callchain_size); + else { + u64 nr = 0; + perf_output_put(&handle, nr); + } + } perf_output_end(&handle); } diff --git a/tools/perf/builtin-annotate.c b/tools/perf/builtin-annotate.c index 7e58e3ad1508..722c0f54e549 100644 --- a/tools/perf/builtin-annotate.c +++ b/tools/perf/builtin-annotate.c @@ -855,7 +855,7 @@ static unsigned long total = 0, total_unknown = 0; static int -process_overflow_event(event_t *event, unsigned long offset, unsigned long head) +process_sample_event(event_t *event, unsigned long offset, unsigned long head) { char level; int show = 0; @@ -1013,10 +1013,10 @@ process_period_event(event_t *event, unsigned long offset, unsigned long head) static int process_event(event_t *event, unsigned long offset, unsigned long head) { - if (event->header.misc & PERF_EVENT_MISC_OVERFLOW) - return process_overflow_event(event, offset, head); - switch (event->header.type) { + case PERF_EVENT_SAMPLE: + return process_sample_event(event, offset, head); + case PERF_EVENT_MMAP: return process_mmap_event(event, offset, head); diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c index e575f3039766..ec5361c67bf5 100644 --- a/tools/perf/builtin-report.c +++ b/tools/perf/builtin-report.c @@ -53,6 +53,8 @@ static regex_t parent_regex; static int exclude_other = 1; +static u64 sample_type; + struct ip_event { struct perf_event_header header; u64 ip; @@ -1135,7 +1137,7 @@ static int validate_chain(struct ip_callchain *chain, event_t *event) } static int -process_overflow_event(event_t *event, unsigned long offset, unsigned long head) +process_sample_event(event_t *event, unsigned long offset, unsigned long head) { char level; int show = 0; @@ -1147,12 +1149,12 @@ process_overflow_event(event_t *event, unsigned long offset, unsigned long head) void *more_data = event->ip.__more_data; struct ip_callchain *chain = NULL; - if (event->header.type & PERF_SAMPLE_PERIOD) { + if (sample_type & PERF_SAMPLE_PERIOD) { period = *(u64 *)more_data; more_data += sizeof(u64); } - dprintf("%p [%p]: PERF_EVENT (IP, %d): %d: %p period: %Ld\n", + dprintf("%p [%p]: PERF_EVENT_SAMPLE (IP, %d): %d: %p period: %Ld\n", (void *)(offset + head), (void *)(long)(event->header.size), event->header.misc, @@ -1160,7 +1162,7 @@ process_overflow_event(event_t *event, unsigned long offset, unsigned long head) (void *)(long)ip, (long long)period); - if (event->header.type & PERF_SAMPLE_CALLCHAIN) { + if (sample_type & PERF_SAMPLE_CALLCHAIN) { int i; chain = (void *)more_data; @@ -1352,10 +1354,10 @@ process_event(event_t *event, unsigned long offset, unsigned long head) { trace_event(event); - if (event->header.misc & PERF_EVENT_MISC_OVERFLOW) - return process_overflow_event(event, offset, head); - switch (event->header.type) { + case PERF_EVENT_SAMPLE: + return process_sample_event(event, offset, head); + case PERF_EVENT_MMAP: return process_mmap_event(event, offset, head); @@ -1388,18 +1390,21 @@ process_event(event_t *event, unsigned long offset, unsigned long head) static struct perf_header *header; -static int perf_header__has_sample(u64 sample_mask) +static u64 perf_header__sample_type(void) { + u64 sample_type = 0; int i; for (i = 0; i < header->attrs; i++) { struct perf_header_attr *attr = header->attr[i]; - if (!(attr->attr.sample_type & sample_mask)) - return 0; + if (!sample_type) + sample_type = attr->attr.sample_type; + else if (sample_type != attr->attr.sample_type) + die("non matching sample_type"); } - return 1; + return sample_type; } static int __cmd_report(void) @@ -1437,8 +1442,9 @@ static int __cmd_report(void) header = perf_header__read(input); head = header->data_offset; - if (sort__has_parent && - !perf_header__has_sample(PERF_SAMPLE_CALLCHAIN)) { + sample_type = perf_header__sample_type(); + + if (sort__has_parent && !(sample_type & PERF_SAMPLE_CALLCHAIN)) { fprintf(stderr, "selected --sort parent, but no callchain data\n"); exit(-1); } diff --git a/tools/perf/builtin-top.c b/tools/perf/builtin-top.c index 5352b5e352ed..cf0d21f1ae10 100644 --- a/tools/perf/builtin-top.c +++ b/tools/perf/builtin-top.c @@ -392,11 +392,11 @@ static void record_ip(u64 ip, int counter) samples--; } -static void process_event(u64 ip, int counter) +static void process_event(u64 ip, int counter, int user) { samples++; - if (ip < min_ip || ip > max_ip) { + if (user) { userspace_samples++; return; } @@ -509,9 +509,10 @@ static void mmap_read_counter(struct mmap_data *md) old += size; - if (event->header.misc & PERF_EVENT_MISC_OVERFLOW) { - if (event->header.type & PERF_SAMPLE_IP) - process_event(event->ip.ip, md->counter); + if (event->header.type == PERF_EVENT_SAMPLE) { + int user = + (event->header.misc & PERF_EVENT_MISC_CPUMODE_MASK) == PERF_EVENT_MISC_USER; + process_event(event->ip.ip, md->counter, user); } } -- cgit v1.2.3 From 5211a242d0cbdded372aee59da18f80552b0a80a Mon Sep 17 00:00:00 2001 From: Kurt Garloff Date: Wed, 24 Jun 2009 14:32:11 -0700 Subject: x86: Add sysctl to allow panic on IOCK NMI error This patch introduces a new sysctl: /proc/sys/kernel/panic_on_io_nmi which defaults to 0 (off). When enabled, the kernel panics when the kernel receives an NMI caused by an IO error. The IO error triggered NMI indicates a serious system condition, which could result in IO data corruption. Rather than contiuing, panicing and dumping might be a better choice, so one can figure out what's causing the IO error. This could be especially important to companies running IO intensive applications where corruption must be avoided, e.g. a bank's databases. [ SuSE has been shipping it for a while, it was done at the request of a large database vendor, for their users. ] Signed-off-by: Kurt Garloff Signed-off-by: Roberto Angelino Signed-off-by: Greg Kroah-Hartman Cc: "Eric W. Biederman" LKML-Reference: <20090624213211.GA11291@kroah.com> Signed-off-by: Ingo Molnar --- arch/x86/kernel/dumpstack.c | 1 + arch/x86/kernel/traps.c | 3 +++ include/linux/kernel.h | 1 + kernel/sysctl.c | 8 ++++++++ 4 files changed, 13 insertions(+) (limited to 'include') diff --git a/arch/x86/kernel/dumpstack.c b/arch/x86/kernel/dumpstack.c index 95ea5fa7d444..c8405718a4c3 100644 --- a/arch/x86/kernel/dumpstack.c +++ b/arch/x86/kernel/dumpstack.c @@ -22,6 +22,7 @@ #include "dumpstack.h" int panic_on_unrecovered_nmi; +int panic_on_io_nmi; unsigned int code_bytes = 64; int kstack_depth_to_print = 3 * STACKSLOTS_PER_LINE; static int die_counter; diff --git a/arch/x86/kernel/traps.c b/arch/x86/kernel/traps.c index a0f48f5671c0..5204332f475d 100644 --- a/arch/x86/kernel/traps.c +++ b/arch/x86/kernel/traps.c @@ -346,6 +346,9 @@ io_check_error(unsigned char reason, struct pt_regs *regs) printk(KERN_EMERG "NMI: IOCK error (debug interrupt?)\n"); show_registers(regs); + if (panic_on_io_nmi) + panic("NMI IOCK error: Not continuing"); + /* Re-enable the IOCK line, wait for a few seconds */ reason = (reason & 0xf) | 8; outb(reason, 0x61); diff --git a/include/linux/kernel.h b/include/linux/kernel.h index fac104e7186a..d6320a3e8def 100644 --- a/include/linux/kernel.h +++ b/include/linux/kernel.h @@ -303,6 +303,7 @@ extern int oops_in_progress; /* If set, an oops, panic(), BUG() or die() is in extern int panic_timeout; extern int panic_on_oops; extern int panic_on_unrecovered_nmi; +extern int panic_on_io_nmi; extern const char *print_tainted(void); extern void add_taint(unsigned flag); extern int test_taint(unsigned flag); diff --git a/kernel/sysctl.c b/kernel/sysctl.c index 62e4ff9968b5..fba42eda8de2 100644 --- a/kernel/sysctl.c +++ b/kernel/sysctl.c @@ -743,6 +743,14 @@ static struct ctl_table kern_table[] = { .mode = 0644, .proc_handler = &proc_dointvec, }, + { + .ctl_name = CTL_UNNUMBERED, + .procname = "panic_on_io_nmi", + .data = &panic_on_io_nmi, + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = &proc_dointvec, + }, { .ctl_name = KERN_BOOTLOADER_TYPE, .procname = "bootloader_type", -- cgit v1.2.3 From 5e955245d6cf49c5ed26c7add7392ff5a6762bf4 Mon Sep 17 00:00:00 2001 From: Bartlomiej Zolnierkiewicz Date: Tue, 23 Jun 2009 11:27:27 +0000 Subject: ide: always kill the whole request on error * Use blk_rq_bytes() instead of obsolete ide_rq_bytes() in ide_kill_rq() and ide_floppy_do_request() for failed requests. [ bugfix part ] * Use blk_rq_bytes() instead of obsolete ide_rq_bytes() in ide_do_devset() and ide_complete_drive_reset(). Then remove ide_rq_bytes(). [ cleanup part ] Signed-off-by: Bartlomiej Zolnierkiewicz Signed-off-by: David S. Miller --- drivers/ide/ide-devsets.c | 2 +- drivers/ide/ide-eh.c | 2 +- drivers/ide/ide-floppy.c | 2 +- drivers/ide/ide-io.c | 14 ++------------ include/linux/ide.h | 1 - 5 files changed, 5 insertions(+), 16 deletions(-) (limited to 'include') diff --git a/drivers/ide/ide-devsets.c b/drivers/ide/ide-devsets.c index 5bf958e5b1d5..1099bf7cf968 100644 --- a/drivers/ide/ide-devsets.c +++ b/drivers/ide/ide-devsets.c @@ -183,6 +183,6 @@ ide_startstop_t ide_do_devset(ide_drive_t *drive, struct request *rq) err = setfunc(drive, *(int *)&rq->cmd[1]); if (err) rq->errors = err; - ide_complete_rq(drive, err, ide_rq_bytes(rq)); + ide_complete_rq(drive, err, blk_rq_bytes(rq)); return ide_stopped; } diff --git a/drivers/ide/ide-eh.c b/drivers/ide/ide-eh.c index 2b9141979613..e9abf2c3c335 100644 --- a/drivers/ide/ide-eh.c +++ b/drivers/ide/ide-eh.c @@ -149,7 +149,7 @@ static inline void ide_complete_drive_reset(ide_drive_t *drive, int err) if (rq && blk_special_request(rq) && rq->cmd[0] == REQ_DRIVE_RESET) { if (err <= 0 && rq->errors == 0) rq->errors = -EIO; - ide_complete_rq(drive, err ? err : 0, ide_rq_bytes(rq)); + ide_complete_rq(drive, err ? err : 0, blk_rq_bytes(rq)); } } diff --git a/drivers/ide/ide-floppy.c b/drivers/ide/ide-floppy.c index 8b3f204f7d73..fefbdfc8db06 100644 --- a/drivers/ide/ide-floppy.c +++ b/drivers/ide/ide-floppy.c @@ -293,7 +293,7 @@ out_end: drive->failed_pc = NULL; if (blk_fs_request(rq) == 0 && rq->errors == 0) rq->errors = -EIO; - ide_complete_rq(drive, -EIO, ide_rq_bytes(rq)); + ide_complete_rq(drive, -EIO, blk_rq_bytes(rq)); return ide_stopped; } diff --git a/drivers/ide/ide-io.c b/drivers/ide/ide-io.c index 95db5f03f6a2..d5f3c77beadd 100644 --- a/drivers/ide/ide-io.c +++ b/drivers/ide/ide-io.c @@ -112,16 +112,6 @@ void ide_complete_cmd(ide_drive_t *drive, struct ide_cmd *cmd, u8 stat, u8 err) } } -/* obsolete, blk_rq_bytes() should be used instead */ -unsigned int ide_rq_bytes(struct request *rq) -{ - if (blk_pc_request(rq)) - return blk_rq_bytes(rq); - else - return blk_rq_cur_sectors(rq) << 9; -} -EXPORT_SYMBOL_GPL(ide_rq_bytes); - int ide_complete_rq(ide_drive_t *drive, int error, unsigned int nr_bytes) { ide_hwif_t *hwif = drive->hwif; @@ -152,14 +142,14 @@ void ide_kill_rq(ide_drive_t *drive, struct request *rq) if ((media == ide_floppy || media == ide_tape) && drv_req) { rq->errors = 0; - ide_complete_rq(drive, -EIO, blk_rq_bytes(rq)); } else { if (media == ide_tape) rq->errors = IDE_DRV_ERROR_GENERAL; else if (blk_fs_request(rq) == 0 && rq->errors == 0) rq->errors = -EIO; - ide_complete_rq(drive, -EIO, ide_rq_bytes(rq)); } + + ide_complete_rq(drive, -EIO, blk_rq_bytes(rq)); } static void ide_tf_set_specify_cmd(ide_drive_t *drive, struct ide_taskfile *tf) diff --git a/include/linux/ide.h b/include/linux/ide.h index cf1f3888067c..c6af7c44d46c 100644 --- a/include/linux/ide.h +++ b/include/linux/ide.h @@ -1062,7 +1062,6 @@ int generic_ide_ioctl(ide_drive_t *, struct block_device *, unsigned, unsigned l extern int ide_vlb_clk; extern int ide_pci_clk; -unsigned int ide_rq_bytes(struct request *); int ide_end_rq(ide_drive_t *, struct request *, int, unsigned int); void ide_kill_rq(ide_drive_t *, struct request *); -- cgit v1.2.3 From 38b7f49a0654cb52cac61c6455807248eee3059d Mon Sep 17 00:00:00 2001 From: Jiri Kosina Date: Fri, 26 Jun 2009 10:48:34 +0200 Subject: HID: fix debugfs build with !CONFIG_DEBUG_FS Fix the debug function prototypes to be correct even in the !CONFIG_DEBUG_FS case. Reported-by: Stephen Rothwell Signed-off-by: Jiri Kosina --- include/linux/hid-debug.h | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) (limited to 'include') diff --git a/include/linux/hid-debug.h b/include/linux/hid-debug.h index ec08ac1ad687..53744fa1c8b7 100644 --- a/include/linux/hid-debug.h +++ b/include/linux/hid-debug.h @@ -22,6 +22,8 @@ * */ +#define HID_DEBUG_BUFSIZE 512 + #ifdef CONFIG_DEBUG_FS void hid_dump_input(struct hid_device *, struct hid_usage *, __s32); @@ -34,7 +36,6 @@ void hid_debug_init(void); void hid_debug_exit(void); void hid_debug_event(struct hid_device *, char *); -#define HID_DEBUG_BUFSIZE 512 struct hid_debug_list { char *hid_debug_buf; @@ -48,11 +49,10 @@ struct hid_debug_list { #else -#define hid_dump_input(a,b) do { } while (0) -#define hid_dump_device(c) do { } while (0) -#define hid_dump_field(a,b) do { } while (0) -#define hid_resolv_usage(a) do { } while (0) -#define hid_resolv_event(a,b) do { } while (0) +#define hid_dump_input(a,b,c) do { } while (0) +#define hid_dump_device(a,b) do { } while (0) +#define hid_dump_field(a,b,c) do { } while (0) +#define hid_resolv_usage(a,b) do { } while (0) #define hid_debug_register(a, b) do { } while (0) #define hid_debug_unregister(a) do { } while (0) #define hid_debug_init() do { } while (0) -- cgit v1.2.3 From 73f1d9391a6aa72efdcea2f302ee7bfcd313c631 Mon Sep 17 00:00:00 2001 From: Paul Mundt Date: Wed, 24 Jun 2009 01:04:36 +0900 Subject: asm-generic/vmlinux.lds.h: Fix up RW_DATA_SECTION definition. RW_DATA_SECTION is defined to take 4 different alignment parameters, while NOSAVE_DATA currently uses a fixed PAGE_SIZE alignment as noted in the comments. There are presently no in-tree users of this at present, and I just stumbled across this while implementing the simplified script on a new architecture port, which subsequently resulted in a syntax error. Signed-off-by: Paul Mundt Signed-off-by: Sam Ravnborg --- include/asm-generic/vmlinux.lds.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/include/asm-generic/vmlinux.lds.h b/include/asm-generic/vmlinux.lds.h index 92b73b6140ff..f92e730695c8 100644 --- a/include/asm-generic/vmlinux.lds.h +++ b/include/asm-generic/vmlinux.lds.h @@ -704,7 +704,7 @@ * matches the requirment of PAGE_ALIGNED_DATA. * * use 0 as page_align if page_aligned data is not used */ -#define RW_DATA_SECTION(cacheline, nosave, pagealigned, inittask) \ +#define RW_DATA_SECTION(cacheline, pagealigned, inittask) \ . = ALIGN(PAGE_SIZE); \ .data : AT(ADDR(.data) - LOAD_OFFSET) { \ INIT_TASK(inittask) \ @@ -712,7 +712,7 @@ READ_MOSTLY_DATA(cacheline) \ DATA_DATA \ CONSTRUCTORS \ - NOSAVE_DATA(nosave) \ + NOSAVE_DATA \ PAGE_ALIGNED_DATA(pagealigned) \ } -- cgit v1.2.3 From d2af12aeadaedf657c9fb9c3df984d2c5ab25f4c Mon Sep 17 00:00:00 2001 From: Tim Abbott Date: Tue, 23 Jun 2009 19:59:35 -0400 Subject: Add new macros for page-aligned data and bss sections. This patch is preparation for replacing most uses of ".bss.page_aligned" and ".data.page_aligned" in the kernel with macros, so that the section name can later be changed without having to touch a lot of the kernel. The long-term goal here is to be able to change the kernel's magic section names to those that are compatible with -ffunction-sections -fdata-sections. This requires renaming all magic sections with names of the form ".data.foo". Signed-off-by: Tim Abbott Acked-by: David Howells Signed-off-by: Sam Ravnborg --- include/linux/linkage.h | 9 +++++++++ 1 file changed, 9 insertions(+) (limited to 'include') diff --git a/include/linux/linkage.h b/include/linux/linkage.h index fee9e59649c1..691f59171c6c 100644 --- a/include/linux/linkage.h +++ b/include/linux/linkage.h @@ -21,6 +21,15 @@ #define __page_aligned_data __section(.data.page_aligned) __aligned(PAGE_SIZE) #define __page_aligned_bss __section(.bss.page_aligned) __aligned(PAGE_SIZE) +/* + * For assembly routines. + * + * Note when using these that you must specify the appropriate + * alignment directives yourself + */ +#define __PAGE_ALIGNED_DATA .section ".data.page_aligned", "aw" +#define __PAGE_ALIGNED_BSS .section ".bss.page_aligned", "aw" + /* * This is used by architectures to keep arguments on the stack * untouched by the compiler by keeping them live until the end. -- cgit v1.2.3 From 39a449d96ac3db9b6d498b6ffbf4c763746d5e8b Mon Sep 17 00:00:00 2001 From: Tim Abbott Date: Tue, 23 Jun 2009 18:53:15 -0400 Subject: asm-generic/vmlinux.lds.h: shuffle INIT_TASK* macro names in vmlinux.lds.h We recently added a INIT_TASK(align) in include/asm-generic/vmlinux.lds.h, but there is already a macro INIT_TASK in include/linux/init_task.h, which is quite confusing. We should switch the macro in the linker script to INIT_TASK_DATA. (Sorry that I missed this in reviewing the patch). Since the macros are new, there is only one user of the INIT_TASK in vmlinux.lds.h, arch/mn10300/kernel/vmlinux.lds.S. However, we are currently using INIT_TASK_DATA for laying down an entire .data.init_task section. So rename that to INIT_TASK_DATA_SECTION. I would be worried about changing the meaning of INIT_TASK_DATA, but the old INIT_TASK_DATA implementation had no users, and in fact if anyone had tried to use it, it would have failed to compile because it didn't pass the alignment to the old INIT_TASK. Signed-off-by: Tim Abbott Cc: David Howells Cc: Jesper Nilsson --- arch/mn10300/kernel/vmlinux.lds.S | 2 +- include/asm-generic/vmlinux.lds.h | 8 ++++---- 2 files changed, 5 insertions(+), 5 deletions(-) (limited to 'include') diff --git a/arch/mn10300/kernel/vmlinux.lds.S b/arch/mn10300/kernel/vmlinux.lds.S index bcebcefb4ad7..c96ba3da95ac 100644 --- a/arch/mn10300/kernel/vmlinux.lds.S +++ b/arch/mn10300/kernel/vmlinux.lds.S @@ -61,7 +61,7 @@ SECTIONS _edata = .; /* End of data section */ } - .data.init_task : { INIT_TASK(THREAD_SIZE); } + .data.init_task : { INIT_TASK_DATA(THREAD_SIZE); } /* might get freed after init */ . = ALIGN(PAGE_SIZE); diff --git a/include/asm-generic/vmlinux.lds.h b/include/asm-generic/vmlinux.lds.h index f92e730695c8..720af4c72206 100644 --- a/include/asm-generic/vmlinux.lds.h +++ b/include/asm-generic/vmlinux.lds.h @@ -191,7 +191,7 @@ . = ALIGN(align); \ *(.data.cacheline_aligned) -#define INIT_TASK(align) \ +#define INIT_TASK_DATA(align) \ . = ALIGN(align); \ *(.data.init_task) @@ -434,10 +434,10 @@ /* * Init task */ -#define INIT_TASK_DATA(align) \ +#define INIT_TASK_DATA_SECTION(align) \ . = ALIGN(align); \ .data.init_task : { \ - INIT_TASK \ + INIT_TASK_DATA(align) \ } #ifdef CONFIG_CONSTRUCTORS @@ -707,7 +707,7 @@ #define RW_DATA_SECTION(cacheline, pagealigned, inittask) \ . = ALIGN(PAGE_SIZE); \ .data : AT(ADDR(.data) - LOAD_OFFSET) { \ - INIT_TASK(inittask) \ + INIT_TASK_DATA(inittask) \ CACHELINE_ALIGNED_DATA(cacheline) \ READ_MOSTLY_DATA(cacheline) \ DATA_DATA \ -- cgit v1.2.3 From 857eceebd2803c9a3459f784acf45e5266921e4d Mon Sep 17 00:00:00 2001 From: Tim Abbott Date: Tue, 23 Jun 2009 19:59:36 -0400 Subject: Add new __init_task_data macro to be used in arch init_task.c files. This patch is preparation for replacing most ".data.init_task" in the kernel with macros, so that the section name can later be changed without having to touch a lot of the kernel. The long-term goal here is to be able to change the kernel's magic section names to those that are compatible with -ffunction-sections -fdata-sections. This requires renaming all magic sections with names of the form ".data.foo". Signed-off-by: Tim Abbott Signed-off-by: Sam Ravnborg --- include/linux/init_task.h | 3 +++ 1 file changed, 3 insertions(+) (limited to 'include') diff --git a/include/linux/init_task.h b/include/linux/init_task.h index 5368fbdc7801..7fc01b13be43 100644 --- a/include/linux/init_task.h +++ b/include/linux/init_task.h @@ -183,5 +183,8 @@ extern struct cred init_cred; LIST_HEAD_INIT(cpu_timers[2]), \ } +/* Attach to the init_task data structure for proper alignment */ +#define __init_task_data __attribute__((__section__(".data.init_task"))) + #endif -- cgit v1.2.3 From 84261923d3dddb766736023bead6fa07b7e218d5 Mon Sep 17 00:00:00 2001 From: Marcelo Tosatti Date: Wed, 17 Jun 2009 10:53:47 -0300 Subject: KVM: protect concurrent make_all_cpus_request make_all_cpus_request contains a race condition which can trigger false request completed status, as follows: CPU0 CPU1 if (test_and_set_bit(req,&vcpu->requests)) .... if (test_and_set_bit(req,&vcpu->requests)) .. return proceed to smp_call_function_many(wait=1) Use a spinlock to serialize concurrent CPUs. Cc: stable@kernel.org Signed-off-by: Andrea Arcangeli Signed-off-by: Marcelo Tosatti Signed-off-by: Avi Kivity --- include/linux/kvm_host.h | 1 + virt/kvm/kvm_main.c | 3 +++ 2 files changed, 4 insertions(+) (limited to 'include') diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h index aacc5449f586..16713dc672e4 100644 --- a/include/linux/kvm_host.h +++ b/include/linux/kvm_host.h @@ -125,6 +125,7 @@ struct kvm_kernel_irq_routing_entry { struct kvm { struct mutex lock; /* protects the vcpus array and APIC accesses */ spinlock_t mmu_lock; + spinlock_t requests_lock; struct rw_semaphore slots_lock; struct mm_struct *mm; /* userspace tied to this vm */ int nmemslots; diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index 013a5b3e9f75..2884baf1d5f9 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -746,6 +746,7 @@ static bool make_all_cpus_request(struct kvm *kvm, unsigned int req) cpumask_clear(cpus); me = get_cpu(); + spin_lock(&kvm->requests_lock); for (i = 0; i < KVM_MAX_VCPUS; ++i) { vcpu = kvm->vcpus[i]; if (!vcpu) @@ -762,6 +763,7 @@ static bool make_all_cpus_request(struct kvm *kvm, unsigned int req) smp_call_function_many(cpus, ack_flush, NULL, 1); else called = false; + spin_unlock(&kvm->requests_lock); put_cpu(); free_cpumask_var(cpus); return called; @@ -982,6 +984,7 @@ static struct kvm *kvm_create_vm(void) kvm->mm = current->mm; atomic_inc(&kvm->mm->mm_count); spin_lock_init(&kvm->mmu_lock); + spin_lock_init(&kvm->requests_lock); kvm_io_bus_init(&kvm->pio_bus); mutex_init(&kvm->lock); kvm_io_bus_init(&kvm->mmio_bus); -- cgit v1.2.3 From 94e5d714f604d4cb4cb13163f01ede278e69258b Mon Sep 17 00:00:00 2001 From: Mimi Zohar Date: Fri, 26 Jun 2009 14:05:27 -0400 Subject: integrity: add ima_counts_put (updated) This patch fixes an imbalance message as reported by J.R. Okajima. The IMA file counters are incremented in ima_path_check. If the actual open fails, such as ETXTBSY, decrement the counters to prevent unnecessary imbalance messages. Reported-by: J.R. Okajima Signed-off-by: Mimi Zohar Signed-off-by: James Morris --- fs/namei.c | 7 +++++++ include/linux/ima.h | 6 ++++++ security/integrity/ima/ima_main.c | 29 ++++++++++++++++++++++++++++- 3 files changed, 41 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/fs/namei.c b/fs/namei.c index 5b961eb71cbf..f3c5b278895a 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -1761,6 +1761,10 @@ do_last: goto exit; } filp = nameidata_to_filp(&nd, open_flag); + if (IS_ERR(filp)) + ima_counts_put(&nd.path, + acc_mode & (MAY_READ | MAY_WRITE | + MAY_EXEC)); mnt_drop_write(nd.path.mnt); if (nd.root.mnt) path_put(&nd.root); @@ -1817,6 +1821,9 @@ ok: goto exit; } filp = nameidata_to_filp(&nd, open_flag); + if (IS_ERR(filp)) + ima_counts_put(&nd.path, + acc_mode & (MAY_READ | MAY_WRITE | MAY_EXEC)); /* * It is now safe to drop the mnt write * because the filp has had a write taken diff --git a/include/linux/ima.h b/include/linux/ima.h index b1b827d091a9..0e3f2a4c25f6 100644 --- a/include/linux/ima.h +++ b/include/linux/ima.h @@ -24,6 +24,7 @@ extern int ima_path_check(struct path *path, int mask, int update_counts); extern void ima_file_free(struct file *file); extern int ima_file_mmap(struct file *file, unsigned long prot); extern void ima_counts_get(struct file *file); +extern void ima_counts_put(struct path *path, int mask); #else static inline int ima_bprm_check(struct linux_binprm *bprm) @@ -60,5 +61,10 @@ static inline void ima_counts_get(struct file *file) { return; } + +static inline void ima_counts_put(struct path *path, int mask) +{ + return; +} #endif /* CONFIG_IMA_H */ #endif /* _LINUX_IMA_H */ diff --git a/security/integrity/ima/ima_main.c b/security/integrity/ima/ima_main.c index 6f611874d10e..101c512564ec 100644 --- a/security/integrity/ima/ima_main.c +++ b/security/integrity/ima/ima_main.c @@ -238,7 +238,34 @@ out: } /* - * ima_opens_get - increment file counts + * ima_counts_put - decrement file counts + * + * File counts are incremented in ima_path_check. On file open + * error, such as ETXTBSY, decrement the counts to prevent + * unnecessary imbalance messages. + */ +void ima_counts_put(struct path *path, int mask) +{ + struct inode *inode = path->dentry->d_inode; + struct ima_iint_cache *iint; + + if (!ima_initialized || !S_ISREG(inode->i_mode)) + return; + iint = ima_iint_find_insert_get(inode); + if (!iint) + return; + + mutex_lock(&iint->mutex); + iint->opencount--; + if ((mask & MAY_WRITE) || (mask == 0)) + iint->writecount--; + else if (mask & (MAY_READ | MAY_EXEC)) + iint->readcount--; + mutex_unlock(&iint->mutex); +} + +/* + * ima_counts_get - increment file counts * * - for IPC shm and shmat file. * - for nfsd exported files. -- cgit v1.2.3 From bab7614d6d1b1fc96ec6c5a7ca34c8641060e659 Mon Sep 17 00:00:00 2001 From: Eric Miao Date: Mon, 29 Jun 2009 00:20:52 -0700 Subject: Input: add support for generic GPIO-based matrix keypad Original patch by Marek Vasut, modified by Eric in: 1. use delayed work to simplify the debouncing 2. combine col_polarity/row_polarity into a single active_low field 3. use a generic bit array based XOR algorithm to detect key press/release, which should make the column assertion time shorter and code a bit cleaner 4. remove the ALT_FN handling, which is no way generic, the ALT_FN key should be treated as no different from other keys, and translation will be done by user space by commands like 'loadkeys'. 5. explicitly disable row IRQs and flush potential pending work, and schedule an immediate scan after resuming as suggested by Uli Luckas 6. incorporate review comments from many others Patch tested on Littleton/PXA310 (though PXA310 has a dedicate keypad controller, I have to configure those pins as generic GPIO to use this driver, works quite well, though), and Sharp Zaurus model SL-C7x0 and SL-C1000. [dtor@mail.ru: fix error unwinding path, support changing keymap from userspace] Signed-off-by: Marek Vasut Reviewed-by: Trilok Soni Reviewed-by: Uli Luckas Reviewed-by: Russell King Reviewed-by: Robert Jarzmik Signed-off-by: Eric Miao Signed-off-by: Dmitry Torokhov --- drivers/input/keyboard/Kconfig | 13 +- drivers/input/keyboard/Makefile | 1 + drivers/input/keyboard/matrix_keypad.c | 453 +++++++++++++++++++++++++++++++++ include/linux/input/matrix_keypad.h | 65 +++++ 4 files changed, 530 insertions(+), 2 deletions(-) create mode 100644 drivers/input/keyboard/matrix_keypad.c create mode 100644 include/linux/input/matrix_keypad.h (limited to 'include') diff --git a/drivers/input/keyboard/Kconfig b/drivers/input/keyboard/Kconfig index d2df1030675a..a6b989a9dc07 100644 --- a/drivers/input/keyboard/Kconfig +++ b/drivers/input/keyboard/Kconfig @@ -158,7 +158,16 @@ config KEYBOARD_GPIO with configuration data saying which GPIOs are used. To compile this driver as a module, choose M here: the - module will be called gpio-keys. + module will be called gpio_keys. + +config KEYBOARD_MATRIX + tristate "GPIO driven matrix keypad support" + depends on GENERIC_GPIO + help + Enable support for GPIO driven matrix keypad. + + To compile this driver as a module, choose M here: the + module will be called matrix_keypad. config KEYBOARD_HIL_OLD tristate "HP HIL keyboard support (simple driver)" @@ -254,7 +263,7 @@ config KEYBOARD_PXA27x tristate "PXA27x/PXA3xx keypad support" depends on PXA27x || PXA3xx help - Enable support for PXA27x/PXA3xx keypad controller + Enable support for PXA27x/PXA3xx keypad controller. To compile this driver as a module, choose M here: the module will be called pxa27x_keypad. diff --git a/drivers/input/keyboard/Makefile b/drivers/input/keyboard/Makefile index 632efbc18c44..b5b5eae9724f 100644 --- a/drivers/input/keyboard/Makefile +++ b/drivers/input/keyboard/Makefile @@ -20,6 +20,7 @@ obj-$(CONFIG_KEYBOARD_LKKBD) += lkkbd.o obj-$(CONFIG_KEYBOARD_LM8323) += lm8323.o obj-$(CONFIG_KEYBOARD_LOCOMO) += locomokbd.o obj-$(CONFIG_KEYBOARD_MAPLE) += maple_keyb.o +obj-$(CONFIG_KEYBOARD_MATRIX) += matrix_keypad.o obj-$(CONFIG_KEYBOARD_NEWTON) += newtonkbd.o obj-$(CONFIG_KEYBOARD_OMAP) += omap-keypad.o obj-$(CONFIG_KEYBOARD_PXA27x) += pxa27x_keypad.o diff --git a/drivers/input/keyboard/matrix_keypad.c b/drivers/input/keyboard/matrix_keypad.c new file mode 100644 index 000000000000..e9b2e7cb05be --- /dev/null +++ b/drivers/input/keyboard/matrix_keypad.c @@ -0,0 +1,453 @@ +/* + * GPIO driven matrix keyboard driver + * + * Copyright (c) 2008 Marek Vasut + * + * Based on corgikbd.c + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +struct matrix_keypad { + const struct matrix_keypad_platform_data *pdata; + struct input_dev *input_dev; + unsigned short *keycodes; + + uint32_t last_key_state[MATRIX_MAX_COLS]; + struct delayed_work work; + bool scan_pending; + bool stopped; + spinlock_t lock; +}; + +/* + * NOTE: normally the GPIO has to be put into HiZ when de-activated to cause + * minmal side effect when scanning other columns, here it is configured to + * be input, and it should work on most platforms. + */ +static void __activate_col(const struct matrix_keypad_platform_data *pdata, + int col, bool on) +{ + bool level_on = !pdata->active_low; + + if (on) { + gpio_direction_output(pdata->col_gpios[col], level_on); + } else { + gpio_set_value_cansleep(pdata->col_gpios[col], !level_on); + gpio_direction_input(pdata->col_gpios[col]); + } +} + +static void activate_col(const struct matrix_keypad_platform_data *pdata, + int col, bool on) +{ + __activate_col(pdata, col, on); + + if (on && pdata->col_scan_delay_us) + udelay(pdata->col_scan_delay_us); +} + +static void activate_all_cols(const struct matrix_keypad_platform_data *pdata, + bool on) +{ + int col; + + for (col = 0; col < pdata->num_col_gpios; col++) + __activate_col(pdata, col, on); +} + +static bool row_asserted(const struct matrix_keypad_platform_data *pdata, + int row) +{ + return gpio_get_value_cansleep(pdata->row_gpios[row]) ? + !pdata->active_low : pdata->active_low; +} + +static void enable_row_irqs(struct matrix_keypad *keypad) +{ + const struct matrix_keypad_platform_data *pdata = keypad->pdata; + int i; + + for (i = 0; i < pdata->num_row_gpios; i++) + enable_irq(gpio_to_irq(pdata->row_gpios[i])); +} + +static void disable_row_irqs(struct matrix_keypad *keypad) +{ + const struct matrix_keypad_platform_data *pdata = keypad->pdata; + int i; + + for (i = 0; i < pdata->num_row_gpios; i++) + disable_irq_nosync(gpio_to_irq(pdata->row_gpios[i])); +} + +/* + * This gets the keys from keyboard and reports it to input subsystem + */ +static void matrix_keypad_scan(struct work_struct *work) +{ + struct matrix_keypad *keypad = + container_of(work, struct matrix_keypad, work.work); + struct input_dev *input_dev = keypad->input_dev; + const struct matrix_keypad_platform_data *pdata = keypad->pdata; + uint32_t new_state[MATRIX_MAX_COLS]; + int row, col, code; + + /* de-activate all columns for scanning */ + activate_all_cols(pdata, false); + + memset(new_state, 0, sizeof(new_state)); + + /* assert each column and read the row status out */ + for (col = 0; col < pdata->num_col_gpios; col++) { + + activate_col(pdata, col, true); + + for (row = 0; row < pdata->num_row_gpios; row++) + new_state[col] |= + row_asserted(pdata, row) ? (1 << row) : 0; + + activate_col(pdata, col, false); + } + + for (col = 0; col < pdata->num_col_gpios; col++) { + uint32_t bits_changed; + + bits_changed = keypad->last_key_state[col] ^ new_state[col]; + if (bits_changed == 0) + continue; + + for (row = 0; row < pdata->num_row_gpios; row++) { + if ((bits_changed & (1 << row)) == 0) + continue; + + code = (row << 4) + col; + input_event(input_dev, EV_MSC, MSC_SCAN, code); + input_report_key(input_dev, + keypad->keycodes[code], + new_state[col] & (1 << row)); + } + } + input_sync(input_dev); + + memcpy(keypad->last_key_state, new_state, sizeof(new_state)); + + activate_all_cols(pdata, true); + + /* Enable IRQs again */ + spin_lock_irq(&keypad->lock); + keypad->scan_pending = false; + enable_row_irqs(keypad); + spin_unlock_irq(&keypad->lock); +} + +static irqreturn_t matrix_keypad_interrupt(int irq, void *id) +{ + struct matrix_keypad *keypad = id; + unsigned long flags; + + spin_lock_irqsave(&keypad->lock, flags); + + /* + * See if another IRQ beaten us to it and scheduled the + * scan already. In that case we should not try to + * disable IRQs again. + */ + if (unlikely(keypad->scan_pending || keypad->stopped)) + goto out; + + disable_row_irqs(keypad); + keypad->scan_pending = true; + schedule_delayed_work(&keypad->work, + msecs_to_jiffies(keypad->pdata->debounce_ms)); + +out: + spin_unlock_irqrestore(&keypad->lock, flags); + return IRQ_HANDLED; +} + +static int matrix_keypad_start(struct input_dev *dev) +{ + struct matrix_keypad *keypad = input_get_drvdata(dev); + + keypad->stopped = false; + mb(); + + /* + * Schedule an immediate key scan to capture current key state; + * columns will be activated and IRQs be enabled after the scan. + */ + schedule_delayed_work(&keypad->work, 0); + + return 0; +} + +static void matrix_keypad_stop(struct input_dev *dev) +{ + struct matrix_keypad *keypad = input_get_drvdata(dev); + + keypad->stopped = true; + mb(); + flush_work(&keypad->work.work); + /* + * matrix_keypad_scan() will leave IRQs enabled; + * we should disable them now. + */ + disable_row_irqs(keypad); +} + +#ifdef CONFIG_PM +static int matrix_keypad_suspend(struct platform_device *pdev, pm_message_t state) +{ + struct matrix_keypad *keypad = platform_get_drvdata(pdev); + const struct matrix_keypad_platform_data *pdata = keypad->pdata; + int i; + + matrix_keypad_stop(keypad->input_dev); + + if (device_may_wakeup(&pdev->dev)) + for (i = 0; i < pdata->num_row_gpios; i++) + enable_irq_wake(gpio_to_irq(pdata->row_gpios[i])); + + return 0; +} + +static int matrix_keypad_resume(struct platform_device *pdev) +{ + struct matrix_keypad *keypad = platform_get_drvdata(pdev); + const struct matrix_keypad_platform_data *pdata = keypad->pdata; + int i; + + if (device_may_wakeup(&pdev->dev)) + for (i = 0; i < pdata->num_row_gpios; i++) + disable_irq_wake(gpio_to_irq(pdata->row_gpios[i])); + + matrix_keypad_start(keypad->input_dev); + + return 0; +} +#else +#define matrix_keypad_suspend NULL +#define matrix_keypad_resume NULL +#endif + +static int __devinit init_matrix_gpio(struct platform_device *pdev, + struct matrix_keypad *keypad) +{ + const struct matrix_keypad_platform_data *pdata = keypad->pdata; + int i, err = -EINVAL; + + /* initialized strobe lines as outputs, activated */ + for (i = 0; i < pdata->num_col_gpios; i++) { + err = gpio_request(pdata->col_gpios[i], "matrix_kbd_col"); + if (err) { + dev_err(&pdev->dev, + "failed to request GPIO%d for COL%d\n", + pdata->col_gpios[i], i); + goto err_free_cols; + } + + gpio_direction_output(pdata->col_gpios[i], !pdata->active_low); + } + + for (i = 0; i < pdata->num_row_gpios; i++) { + err = gpio_request(pdata->row_gpios[i], "matrix_kbd_row"); + if (err) { + dev_err(&pdev->dev, + "failed to request GPIO%d for ROW%d\n", + pdata->row_gpios[i], i); + goto err_free_rows; + } + + gpio_direction_input(pdata->row_gpios[i]); + } + + for (i = 0; i < pdata->num_row_gpios; i++) { + err = request_irq(gpio_to_irq(pdata->row_gpios[i]), + matrix_keypad_interrupt, + IRQF_DISABLED | + IRQF_TRIGGER_RISING | IRQF_TRIGGER_FALLING, + "matrix-keypad", keypad); + if (err) { + dev_err(&pdev->dev, + "Unable to acquire interrupt for GPIO line %i\n", + pdata->row_gpios[i]); + goto err_free_irqs; + } + } + + /* initialized as disabled - enabled by input->open */ + disable_row_irqs(keypad); + return 0; + +err_free_irqs: + while (--i >= 0) + free_irq(gpio_to_irq(pdata->row_gpios[i]), keypad); + i = pdata->num_row_gpios; +err_free_rows: + while (--i >= 0) + gpio_free(pdata->row_gpios[i]); + i = pdata->num_col_gpios; +err_free_cols: + while (--i >= 0) + gpio_free(pdata->col_gpios[i]); + + return err; +} + +static int __devinit matrix_keypad_probe(struct platform_device *pdev) +{ + const struct matrix_keypad_platform_data *pdata; + const struct matrix_keymap_data *keymap_data; + struct matrix_keypad *keypad; + struct input_dev *input_dev; + unsigned short *keycodes; + int i; + int err; + + pdata = pdev->dev.platform_data; + if (!pdata) { + dev_err(&pdev->dev, "no platform data defined\n"); + return -EINVAL; + } + + keymap_data = pdata->keymap_data; + if (!keymap_data) { + dev_err(&pdev->dev, "no keymap data defined\n"); + return -EINVAL; + } + + if (!keymap_data->max_keymap_size) { + dev_err(&pdev->dev, "invalid keymap data supplied\n"); + return -EINVAL; + } + + keypad = kzalloc(sizeof(struct matrix_keypad), GFP_KERNEL); + keycodes = kzalloc(keymap_data->max_keymap_size * + sizeof(keypad->keycodes), + GFP_KERNEL); + input_dev = input_allocate_device(); + if (!keypad || !keycodes || !input_dev) { + err = -ENOMEM; + goto err_free_mem; + } + + keypad->input_dev = input_dev; + keypad->pdata = pdata; + keypad->keycodes = keycodes; + keypad->stopped = true; + INIT_DELAYED_WORK(&keypad->work, matrix_keypad_scan); + spin_lock_init(&keypad->lock); + + input_dev->name = pdev->name; + input_dev->id.bustype = BUS_HOST; + input_dev->dev.parent = &pdev->dev; + input_dev->evbit[0] = BIT_MASK(EV_KEY) | BIT_MASK(EV_REP); + input_dev->open = matrix_keypad_start; + input_dev->close = matrix_keypad_stop; + + input_dev->keycode = keycodes; + input_dev->keycodesize = sizeof(*keycodes); + input_dev->keycodemax = keymap_data->max_keymap_size; + + for (i = 0; i < keymap_data->keymap_size; i++) { + unsigned int key = keymap_data->keymap[i]; + unsigned int row = KEY_ROW(key); + unsigned int col = KEY_COL(key); + unsigned short code = KEY_VAL(key); + + keycodes[(row << 4) + col] = code; + __set_bit(code, input_dev->keybit); + } + __clear_bit(KEY_RESERVED, input_dev->keybit); + + input_set_capability(input_dev, EV_MSC, MSC_SCAN); + input_set_drvdata(input_dev, keypad); + + err = init_matrix_gpio(pdev, keypad); + if (err) + goto err_free_mem; + + err = input_register_device(keypad->input_dev); + if (err) + goto err_free_mem; + + device_init_wakeup(&pdev->dev, pdata->wakeup); + platform_set_drvdata(pdev, keypad); + + return 0; + +err_free_mem: + input_free_device(input_dev); + kfree(keycodes); + kfree(keypad); + return err; +} + +static int __devexit matrix_keypad_remove(struct platform_device *pdev) +{ + struct matrix_keypad *keypad = platform_get_drvdata(pdev); + const struct matrix_keypad_platform_data *pdata = keypad->pdata; + int i; + + device_init_wakeup(&pdev->dev, 0); + + for (i = 0; i < pdata->num_row_gpios; i++) { + free_irq(gpio_to_irq(pdata->row_gpios[i]), keypad); + gpio_free(pdata->row_gpios[i]); + } + + for (i = 0; i < pdata->num_col_gpios; i++) + gpio_free(pdata->col_gpios[i]); + + input_unregister_device(keypad->input_dev); + platform_set_drvdata(pdev, NULL); + kfree(keypad->keycodes); + kfree(keypad); + + return 0; +} + +static struct platform_driver matrix_keypad_driver = { + .probe = matrix_keypad_probe, + .remove = __devexit_p(matrix_keypad_remove), + .suspend = matrix_keypad_suspend, + .resume = matrix_keypad_resume, + .driver = { + .name = "matrix-keypad", + .owner = THIS_MODULE, + }, +}; + +static int __init matrix_keypad_init(void) +{ + return platform_driver_register(&matrix_keypad_driver); +} + +static void __exit matrix_keypad_exit(void) +{ + platform_driver_unregister(&matrix_keypad_driver); +} + +module_init(matrix_keypad_init); +module_exit(matrix_keypad_exit); + +MODULE_AUTHOR("Marek Vasut "); +MODULE_DESCRIPTION("GPIO Driven Matrix Keypad Driver"); +MODULE_LICENSE("GPL v2"); +MODULE_ALIAS("platform:matrix-keypad"); diff --git a/include/linux/input/matrix_keypad.h b/include/linux/input/matrix_keypad.h new file mode 100644 index 000000000000..7964516c6954 --- /dev/null +++ b/include/linux/input/matrix_keypad.h @@ -0,0 +1,65 @@ +#ifndef _MATRIX_KEYPAD_H +#define _MATRIX_KEYPAD_H + +#include +#include + +#define MATRIX_MAX_ROWS 16 +#define MATRIX_MAX_COLS 16 + +#define KEY(row, col, val) ((((row) & (MATRIX_MAX_ROWS - 1)) << 24) |\ + (((col) & (MATRIX_MAX_COLS - 1)) << 16) |\ + (val & 0xffff)) + +#define KEY_ROW(k) (((k) >> 24) & 0xff) +#define KEY_COL(k) (((k) >> 16) & 0xff) +#define KEY_VAL(k) ((k) & 0xffff) + +/** + * struct matrix_keymap_data - keymap for matrix keyboards + * @keymap: pointer to array of uint32 values encoded with KEY() macro + * representing keymap + * @keymap_size: number of entries (initialized) in this keymap + * @max_keymap_size: maximum size of keymap supported by the device + * + * This structure is supposed to be used by platform code to supply + * keymaps to drivers that implement matrix-like keypads/keyboards. + */ +struct matrix_keymap_data { + const uint32_t *keymap; + unsigned int keymap_size; + unsigned int max_keymap_size; +}; + +/** + * struct matrix_keypad_platform_data - platform-dependent keypad data + * @keymap_data: pointer to &matrix_keymap_data + * @row_gpios: array of gpio numbers reporesenting rows + * @col_gpios: array of gpio numbers reporesenting colums + * @num_row_gpios: actual number of row gpios used by device + * @num_col_gpios: actual number of col gpios used by device + * @col_scan_delay_us: delay, measured in microseconds, that is + * needed before we can keypad after activating column gpio + * @debounce_ms: debounce interval in milliseconds + * + * This structure represents platform-specific data that use used by + * matrix_keypad driver to perform proper initialization. + */ +struct matrix_keypad_platform_data { + const struct matrix_keymap_data *keymap_data; + + unsigned int row_gpios[MATRIX_MAX_ROWS]; + unsigned int col_gpios[MATRIX_MAX_COLS]; + unsigned int num_row_gpios; + unsigned int num_col_gpios; + + unsigned int col_scan_delay_us; + + /* key debounce interval in milli-second */ + unsigned int debounce_ms; + + bool active_low; + bool wakeup; +}; + +#endif /* _MATRIX_KEYPAD_H */ -- cgit v1.2.3 From 1a8dd307cc0a2119be4e578c517795464e6dabba Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Mon, 29 Jun 2009 17:45:39 +0900 Subject: percpu: use __weak only in the definition of weak percpu variables __weak is necessary only for definition and might even not work in declaration. Drop it from declaration. This change was suggested by Ivan Kokshaysky. Signed-off-by: Tejun Heo Acked-by: Ivan Kokshaysky --- include/linux/percpu-defs.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/percpu-defs.h b/include/linux/percpu-defs.h index cf32838ad0fa..9b7a53cc16eb 100644 --- a/include/linux/percpu-defs.h +++ b/include/linux/percpu-defs.h @@ -56,7 +56,7 @@ */ #define DECLARE_PER_CPU_SECTION(type, name, sec) \ extern __PCPU_DUMMY_ATTRS char __pcpu_scope_##name; \ - extern __PCPU_ATTRS(sec) __weak __typeof__(type) per_cpu__##name + extern __PCPU_ATTRS(sec) __typeof__(type) per_cpu__##name #define DEFINE_PER_CPU_SECTION(type, name, sec) \ __PCPU_DUMMY_ATTRS char __pcpu_scope_##name; \ -- cgit v1.2.3 From a3a9f79e361e864f0e9d75ebe2a0cb43d17c4272 Mon Sep 17 00:00:00 2001 From: Patrick McHardy Date: Mon, 29 Jun 2009 14:07:56 +0200 Subject: netfilter: tcp conntrack: fix unacknowledged data detection with NAT When NAT helpers change the TCP packet size, the highest seen sequence number needs to be corrected. This is currently only done upwards, when the packet size is reduced the sequence number is unchanged. This causes TCP conntrack to falsely detect unacknowledged data and decrease the timeout. Fix by updating the highest seen sequence number in both directions after packet mangling. Tested-by: Krzysztof Piotr Oledzki Signed-off-by: Patrick McHardy --- include/net/netfilter/nf_conntrack.h | 4 ++-- net/ipv4/netfilter/nf_nat_helper.c | 17 +++++++++++------ net/netfilter/nf_conntrack_proto_tcp.c | 6 +++--- 3 files changed, 16 insertions(+), 11 deletions(-) (limited to 'include') diff --git a/include/net/netfilter/nf_conntrack.h b/include/net/netfilter/nf_conntrack.h index a632689b61b4..cbdd6284996d 100644 --- a/include/net/netfilter/nf_conntrack.h +++ b/include/net/netfilter/nf_conntrack.h @@ -258,8 +258,8 @@ static inline bool nf_ct_kill(struct nf_conn *ct) /* Update TCP window tracking data when NAT mangles the packet */ extern void nf_conntrack_tcp_update(const struct sk_buff *skb, unsigned int dataoff, - struct nf_conn *ct, - int dir); + struct nf_conn *ct, int dir, + s16 offset); /* Fake conntrack entry for untracked connections */ extern struct nf_conn nf_conntrack_untracked; diff --git a/net/ipv4/netfilter/nf_nat_helper.c b/net/ipv4/netfilter/nf_nat_helper.c index 155c008626c8..09172a65d9b6 100644 --- a/net/ipv4/netfilter/nf_nat_helper.c +++ b/net/ipv4/netfilter/nf_nat_helper.c @@ -191,7 +191,8 @@ nf_nat_mangle_tcp_packet(struct sk_buff *skb, ct, ctinfo); /* Tell TCP window tracking about seq change */ nf_conntrack_tcp_update(skb, ip_hdrlen(skb), - ct, CTINFO2DIR(ctinfo)); + ct, CTINFO2DIR(ctinfo), + (int)rep_len - (int)match_len); nf_conntrack_event_cache(IPCT_NATSEQADJ, ct); } @@ -377,6 +378,7 @@ nf_nat_seq_adjust(struct sk_buff *skb, struct tcphdr *tcph; int dir; __be32 newseq, newack; + s16 seqoff, ackoff; struct nf_conn_nat *nat = nfct_nat(ct); struct nf_nat_seq *this_way, *other_way; @@ -390,15 +392,18 @@ nf_nat_seq_adjust(struct sk_buff *skb, tcph = (void *)skb->data + ip_hdrlen(skb); if (after(ntohl(tcph->seq), this_way->correction_pos)) - newseq = htonl(ntohl(tcph->seq) + this_way->offset_after); + seqoff = this_way->offset_after; else - newseq = htonl(ntohl(tcph->seq) + this_way->offset_before); + seqoff = this_way->offset_before; if (after(ntohl(tcph->ack_seq) - other_way->offset_before, other_way->correction_pos)) - newack = htonl(ntohl(tcph->ack_seq) - other_way->offset_after); + ackoff = other_way->offset_after; else - newack = htonl(ntohl(tcph->ack_seq) - other_way->offset_before); + ackoff = other_way->offset_before; + + newseq = htonl(ntohl(tcph->seq) + seqoff); + newack = htonl(ntohl(tcph->ack_seq) - ackoff); inet_proto_csum_replace4(&tcph->check, skb, tcph->seq, newseq, 0); inet_proto_csum_replace4(&tcph->check, skb, tcph->ack_seq, newack, 0); @@ -413,7 +418,7 @@ nf_nat_seq_adjust(struct sk_buff *skb, if (!nf_nat_sack_adjust(skb, tcph, ct, ctinfo)) return 0; - nf_conntrack_tcp_update(skb, ip_hdrlen(skb), ct, dir); + nf_conntrack_tcp_update(skb, ip_hdrlen(skb), ct, dir, seqoff); return 1; } diff --git a/net/netfilter/nf_conntrack_proto_tcp.c b/net/netfilter/nf_conntrack_proto_tcp.c index 33fc0a443f3d..97a82ba75376 100644 --- a/net/netfilter/nf_conntrack_proto_tcp.c +++ b/net/netfilter/nf_conntrack_proto_tcp.c @@ -720,8 +720,8 @@ static bool tcp_in_window(const struct nf_conn *ct, /* Caller must linearize skb at tcp header. */ void nf_conntrack_tcp_update(const struct sk_buff *skb, unsigned int dataoff, - struct nf_conn *ct, - int dir) + struct nf_conn *ct, int dir, + s16 offset) { const struct tcphdr *tcph = (const void *)skb->data + dataoff; const struct ip_ct_tcp_state *sender = &ct->proto.tcp.seen[dir]; @@ -734,7 +734,7 @@ void nf_conntrack_tcp_update(const struct sk_buff *skb, /* * We have to worry for the ack in the reply packet only... */ - if (after(end, ct->proto.tcp.seen[dir].td_end)) + if (ct->proto.tcp.seen[dir].td_end + offset == end) ct->proto.tcp.seen[dir].td_end = end; ct->proto.tcp.last_end = end; spin_unlock_bh(&ct->lock); -- cgit v1.2.3 From 8a3af79361e85db6fec4173ef1916322471c19e3 Mon Sep 17 00:00:00 2001 From: Jaswinder Singh Rajput Date: Mon, 29 Jun 2009 14:28:27 +0200 Subject: netfilter: headers_check fix: linux/netfilter/xt_osf.h fix the following 'make headers_check' warnings: usr/include/linux/netfilter/xt_osf.h:40: found __[us]{8,16,32,64} type without #include Signed-off-by: Jaswinder Singh Rajput Signed-off-by: Patrick McHardy --- include/linux/netfilter/xt_osf.h | 2 ++ 1 file changed, 2 insertions(+) (limited to 'include') diff --git a/include/linux/netfilter/xt_osf.h b/include/linux/netfilter/xt_osf.h index fd2272e0959a..18afa495f973 100644 --- a/include/linux/netfilter/xt_osf.h +++ b/include/linux/netfilter/xt_osf.h @@ -20,6 +20,8 @@ #ifndef _XT_OSF_H #define _XT_OSF_H +#include + #define MAXGENRELEN 32 #define XT_OSF_GENRE (1<<0) -- cgit v1.2.3 From d6d3f08b0fd998b647a05540cedd11a067b72867 Mon Sep 17 00:00:00 2001 From: Jan Engelhardt Date: Mon, 29 Jun 2009 14:31:46 +0200 Subject: netfilter: xtables: conntrack match revision 2 As reported by Philip, the UNTRACKED state bit does not fit within the 8-bit state_mask member. Enlarge state_mask and give status_mask a few more bits too. Reported-by: Philip Craig References: http://markmail.org/thread/b7eg6aovfh4agyz7 Signed-off-by: Jan Engelhardt Signed-off-by: Patrick McHardy --- include/linux/netfilter/xt_conntrack.h | 13 +++++++ net/netfilter/xt_conntrack.c | 66 ++++++++++++++++++++++++++++++---- 2 files changed, 73 insertions(+), 6 deletions(-) (limited to 'include') diff --git a/include/linux/netfilter/xt_conntrack.h b/include/linux/netfilter/xt_conntrack.h index 3430c7751948..7ae05338e94c 100644 --- a/include/linux/netfilter/xt_conntrack.h +++ b/include/linux/netfilter/xt_conntrack.h @@ -81,4 +81,17 @@ struct xt_conntrack_mtinfo1 { __u8 state_mask, status_mask; }; +struct xt_conntrack_mtinfo2 { + union nf_inet_addr origsrc_addr, origsrc_mask; + union nf_inet_addr origdst_addr, origdst_mask; + union nf_inet_addr replsrc_addr, replsrc_mask; + union nf_inet_addr repldst_addr, repldst_mask; + __u32 expires_min, expires_max; + __u16 l4proto; + __be16 origsrc_port, origdst_port; + __be16 replsrc_port, repldst_port; + __u16 match_flags, invert_flags; + __u16 state_mask, status_mask; +}; + #endif /*_XT_CONNTRACK_H*/ diff --git a/net/netfilter/xt_conntrack.c b/net/netfilter/xt_conntrack.c index 0b7139f3dd78..fc581800698e 100644 --- a/net/netfilter/xt_conntrack.c +++ b/net/netfilter/xt_conntrack.c @@ -129,7 +129,7 @@ conntrack_addrcmp(const union nf_inet_addr *kaddr, static inline bool conntrack_mt_origsrc(const struct nf_conn *ct, - const struct xt_conntrack_mtinfo1 *info, + const struct xt_conntrack_mtinfo2 *info, u_int8_t family) { return conntrack_addrcmp(&ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple.src.u3, @@ -138,7 +138,7 @@ conntrack_mt_origsrc(const struct nf_conn *ct, static inline bool conntrack_mt_origdst(const struct nf_conn *ct, - const struct xt_conntrack_mtinfo1 *info, + const struct xt_conntrack_mtinfo2 *info, u_int8_t family) { return conntrack_addrcmp(&ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple.dst.u3, @@ -147,7 +147,7 @@ conntrack_mt_origdst(const struct nf_conn *ct, static inline bool conntrack_mt_replsrc(const struct nf_conn *ct, - const struct xt_conntrack_mtinfo1 *info, + const struct xt_conntrack_mtinfo2 *info, u_int8_t family) { return conntrack_addrcmp(&ct->tuplehash[IP_CT_DIR_REPLY].tuple.src.u3, @@ -156,7 +156,7 @@ conntrack_mt_replsrc(const struct nf_conn *ct, static inline bool conntrack_mt_repldst(const struct nf_conn *ct, - const struct xt_conntrack_mtinfo1 *info, + const struct xt_conntrack_mtinfo2 *info, u_int8_t family) { return conntrack_addrcmp(&ct->tuplehash[IP_CT_DIR_REPLY].tuple.dst.u3, @@ -164,7 +164,7 @@ conntrack_mt_repldst(const struct nf_conn *ct, } static inline bool -ct_proto_port_check(const struct xt_conntrack_mtinfo1 *info, +ct_proto_port_check(const struct xt_conntrack_mtinfo2 *info, const struct nf_conn *ct) { const struct nf_conntrack_tuple *tuple; @@ -204,7 +204,7 @@ ct_proto_port_check(const struct xt_conntrack_mtinfo1 *info, static bool conntrack_mt(const struct sk_buff *skb, const struct xt_match_param *par) { - const struct xt_conntrack_mtinfo1 *info = par->matchinfo; + const struct xt_conntrack_mtinfo2 *info = par->matchinfo; enum ip_conntrack_info ctinfo; const struct nf_conn *ct; unsigned int statebit; @@ -278,6 +278,16 @@ conntrack_mt(const struct sk_buff *skb, const struct xt_match_param *par) return true; } +static bool +conntrack_mt_v1(const struct sk_buff *skb, const struct xt_match_param *par) +{ + const struct xt_conntrack_mtinfo2 *const *info = par->matchinfo; + struct xt_match_param newpar = *par; + + newpar.matchinfo = *info; + return conntrack_mt(skb, &newpar); +} + static bool conntrack_mt_check(const struct xt_mtchk_param *par) { if (nf_ct_l3proto_try_module_get(par->family) < 0) { @@ -288,11 +298,45 @@ static bool conntrack_mt_check(const struct xt_mtchk_param *par) return true; } +static bool conntrack_mt_check_v1(const struct xt_mtchk_param *par) +{ + struct xt_conntrack_mtinfo1 *info = par->matchinfo; + struct xt_conntrack_mtinfo2 *up; + int ret = conntrack_mt_check(par); + + if (ret < 0) + return ret; + + up = kmalloc(sizeof(*up), GFP_KERNEL); + if (up == NULL) { + nf_ct_l3proto_module_put(par->family); + return -ENOMEM; + } + + /* + * The strategy here is to minimize the overhead of v1 matching, + * by prebuilding a v2 struct and putting the pointer into the + * v1 dataspace. + */ + memcpy(up, info, offsetof(typeof(*info), state_mask)); + up->state_mask = info->state_mask; + up->status_mask = info->status_mask; + *(void **)info = up; + return true; +} + static void conntrack_mt_destroy(const struct xt_mtdtor_param *par) { nf_ct_l3proto_module_put(par->family); } +static void conntrack_mt_destroy_v1(const struct xt_mtdtor_param *par) +{ + struct xt_conntrack_mtinfo2 **info = par->matchinfo; + kfree(*info); + conntrack_mt_destroy(par); +} + #ifdef CONFIG_COMPAT struct compat_xt_conntrack_info { @@ -363,6 +407,16 @@ static struct xt_match conntrack_mt_reg[] __read_mostly = { .revision = 1, .family = NFPROTO_UNSPEC, .matchsize = sizeof(struct xt_conntrack_mtinfo1), + .match = conntrack_mt_v1, + .checkentry = conntrack_mt_check_v1, + .destroy = conntrack_mt_destroy_v1, + .me = THIS_MODULE, + }, + { + .name = "conntrack", + .revision = 2, + .family = NFPROTO_UNSPEC, + .matchsize = sizeof(struct xt_conntrack_mtinfo2), .match = conntrack_mt, .checkentry = conntrack_mt_check, .destroy = conntrack_mt_destroy, -- cgit v1.2.3 From e6ce3066010a21bde961d8f8cefe0b69cae78a0f Mon Sep 17 00:00:00 2001 From: Jeremy Kerr Date: Mon, 29 Jun 2009 14:31:58 +0800 Subject: fs: allow d_instantiate to be called with negative parent dentry The new fsnotify infrastructure (starting at 90586523) causes an oops in spufs, where we populate a directory with files before instantiating the directory itself. The new changes seem to have introduced an assumption that a dentry's parent will be positive when instantiating. This change makes it once again possible to d_instantiate a dentry with a negative parent, and brings __fsnotify_d_instantiate() into line with inotify_d_instantiate(), which already has this NULL check. Signed-off-by: Jeremy Kerr Signed-off-by: Eric Paris --- include/linux/fsnotify_backend.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/fsnotify_backend.h b/include/linux/fsnotify_backend.h index 44848aa830dc..6c3de999fb34 100644 --- a/include/linux/fsnotify_backend.h +++ b/include/linux/fsnotify_backend.h @@ -280,7 +280,7 @@ static inline void __fsnotify_update_dcache_flags(struct dentry *dentry) assert_spin_locked(&dentry->d_lock); parent = dentry->d_parent; - if (fsnotify_inode_watches_children(parent->d_inode)) + if (parent->d_inode && fsnotify_inode_watches_children(parent->d_inode)) dentry->d_flags |= DCACHE_FSNOTIFY_PARENT_WATCHED; else dentry->d_flags &= ~DCACHE_FSNOTIFY_PARENT_WATCHED; -- cgit v1.2.3 From 2fc90f6133a87da8177636866557d4cc5f56e661 Mon Sep 17 00:00:00 2001 From: Alexey Zaytsev Date: Wed, 24 Jun 2009 16:22:30 +0400 Subject: PCI: make pci_name() take const argument Since this function should never modify it (saves warnings when called with const args too). Signed-off-by: Alexey Zaytsev Signed-off-by: Jesse Barnes --- include/linux/pci.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/pci.h b/include/linux/pci.h index d304ddf412d0..115fb7ba5089 100644 --- a/include/linux/pci.h +++ b/include/linux/pci.h @@ -1145,7 +1145,7 @@ static inline void pci_set_drvdata(struct pci_dev *pdev, void *data) /* If you want to know what to call your pci_dev, ask this function. * Again, it's a wrapper around the generic device. */ -static inline const char *pci_name(struct pci_dev *pdev) +static inline const char *pci_name(const struct pci_dev *pdev) { return dev_name(&pdev->dev); } -- cgit v1.2.3 From 2bf427b25b79eb7cea27963a66c3d4684cae0e0c Mon Sep 17 00:00:00 2001 From: Bartlomiej Zolnierkiewicz Date: Mon, 29 Jun 2009 19:20:42 -0700 Subject: ide: fix resume for CONFIG_BLK_DEV_IDEACPI=y commit 2f0d0fd2a605666d38e290c5c0d2907484352dc4 ("ide-acpi: cleanup do_drive_get_GTF()") didn't account for the lack of hwif->acpidata check in generic_ide_suspend() [ indirect user of do_drive_get_GTF() through ide_acpi_exec_tfs() ] resulting in broken resume when ACPI support is enabled but ACPI data is unavailable. Fix it by adding ide_port_acpi() helper for checking if port needs ACPI handling and cleaning generic_ide_{suspend,resume}() to use it instead of hiding hwif->acpidata and ide_noacpi checks in IDE ACPI helpers (this should help in preventing similar bugs in the future). While at it: - kill superfluous debugging printks in ide_acpi_{get,push}_timing() Reported-and-tested-by: Etienne Basset Also-reported-and-tested-by: Jeff Chua Signed-off-by: Bartlomiej Zolnierkiewicz Signed-off-by: David S. Miller --- drivers/ide/ide-acpi.c | 37 +++++++------------------------------ drivers/ide/ide-pm.c | 30 ++++++++++++++++++------------ include/linux/ide.h | 2 ++ 3 files changed, 27 insertions(+), 42 deletions(-) (limited to 'include') diff --git a/drivers/ide/ide-acpi.c b/drivers/ide/ide-acpi.c index 77f79d26b264..c509c9916464 100644 --- a/drivers/ide/ide-acpi.c +++ b/drivers/ide/ide-acpi.c @@ -92,6 +92,11 @@ int ide_acpi_init(void) return 0; } +bool ide_port_acpi(ide_hwif_t *hwif) +{ + return ide_noacpi == 0 && hwif->acpidata; +} + /** * ide_get_dev_handle - finds acpi_handle and PCI device.function * @dev: device to locate @@ -352,9 +357,6 @@ int ide_acpi_exec_tfs(ide_drive_t *drive) unsigned long gtf_address; unsigned long obj_loc; - if (ide_noacpi) - return 0; - DEBPRINT("call get_GTF, drive=%s port=%d\n", drive->name, drive->dn); ret = do_drive_get_GTF(drive, >f_length, >f_address, &obj_loc); @@ -389,16 +391,6 @@ void ide_acpi_get_timing(ide_hwif_t *hwif) struct acpi_buffer output; union acpi_object *out_obj; - if (ide_noacpi) - return; - - DEBPRINT("ENTER:\n"); - - if (!hwif->acpidata) { - DEBPRINT("no ACPI data for %s\n", hwif->name); - return; - } - /* Setting up output buffer for _GTM */ output.length = ACPI_ALLOCATE_BUFFER; output.pointer = NULL; /* ACPI-CA sets this; save/free it later */ @@ -479,16 +471,6 @@ void ide_acpi_push_timing(ide_hwif_t *hwif) struct ide_acpi_drive_link *master = &hwif->acpidata->master; struct ide_acpi_drive_link *slave = &hwif->acpidata->slave; - if (ide_noacpi) - return; - - DEBPRINT("ENTER:\n"); - - if (!hwif->acpidata) { - DEBPRINT("no ACPI data for %s\n", hwif->name); - return; - } - /* Give the GTM buffer + drive Identify data to the channel via the * _STM method: */ /* setup input parameters buffer for _STM */ @@ -527,16 +509,11 @@ void ide_acpi_set_state(ide_hwif_t *hwif, int on) ide_drive_t *drive; int i; - if (ide_noacpi || ide_noacpi_psx) + if (ide_noacpi_psx) return; DEBPRINT("ENTER:\n"); - if (!hwif->acpidata) { - DEBPRINT("no ACPI data for %s\n", hwif->name); - return; - } - /* channel first and then drives for power on and verse versa for power off */ if (on) acpi_bus_set_power(hwif->acpidata->obj_handle, ACPI_STATE_D0); @@ -616,7 +593,7 @@ void ide_acpi_port_init_devices(ide_hwif_t *hwif) drive->name, err); } - if (!ide_acpionboot) { + if (ide_noacpi || ide_acpionboot == 0) { DEBPRINT("ACPI methods disabled on boot\n"); return; } diff --git a/drivers/ide/ide-pm.c b/drivers/ide/ide-pm.c index c14ca144cffe..ad7be2669dcb 100644 --- a/drivers/ide/ide-pm.c +++ b/drivers/ide/ide-pm.c @@ -10,9 +10,11 @@ int generic_ide_suspend(struct device *dev, pm_message_t mesg) struct request_pm_state rqpm; int ret; - /* call ACPI _GTM only once */ - if ((drive->dn & 1) == 0 || pair == NULL) - ide_acpi_get_timing(hwif); + if (ide_port_acpi(hwif)) { + /* call ACPI _GTM only once */ + if ((drive->dn & 1) == 0 || pair == NULL) + ide_acpi_get_timing(hwif); + } memset(&rqpm, 0, sizeof(rqpm)); rq = blk_get_request(drive->queue, READ, __GFP_WAIT); @@ -26,9 +28,11 @@ int generic_ide_suspend(struct device *dev, pm_message_t mesg) ret = blk_execute_rq(drive->queue, NULL, rq, 0); blk_put_request(rq); - /* call ACPI _PS3 only after both devices are suspended */ - if (ret == 0 && ((drive->dn & 1) || pair == NULL)) - ide_acpi_set_state(hwif, 0); + if (ret == 0 && ide_port_acpi(hwif)) { + /* call ACPI _PS3 only after both devices are suspended */ + if ((drive->dn & 1) || pair == NULL) + ide_acpi_set_state(hwif, 0); + } return ret; } @@ -42,13 +46,15 @@ int generic_ide_resume(struct device *dev) struct request_pm_state rqpm; int err; - /* call ACPI _PS0 / _STM only once */ - if ((drive->dn & 1) == 0 || pair == NULL) { - ide_acpi_set_state(hwif, 1); - ide_acpi_push_timing(hwif); - } + if (ide_port_acpi(hwif)) { + /* call ACPI _PS0 / _STM only once */ + if ((drive->dn & 1) == 0 || pair == NULL) { + ide_acpi_set_state(hwif, 1); + ide_acpi_push_timing(hwif); + } - ide_acpi_exec_tfs(drive); + ide_acpi_exec_tfs(drive); + } memset(&rqpm, 0, sizeof(rqpm)); rq = blk_get_request(drive->queue, READ, __GFP_WAIT); diff --git a/include/linux/ide.h b/include/linux/ide.h index c6af7c44d46c..edc93a6d931d 100644 --- a/include/linux/ide.h +++ b/include/linux/ide.h @@ -1419,6 +1419,7 @@ static inline void ide_dma_unmap_sg(ide_drive_t *drive, #ifdef CONFIG_BLK_DEV_IDEACPI int ide_acpi_init(void); +bool ide_port_acpi(ide_hwif_t *hwif); extern int ide_acpi_exec_tfs(ide_drive_t *drive); extern void ide_acpi_get_timing(ide_hwif_t *hwif); extern void ide_acpi_push_timing(ide_hwif_t *hwif); @@ -1427,6 +1428,7 @@ void ide_acpi_port_init_devices(ide_hwif_t *); extern void ide_acpi_set_state(ide_hwif_t *hwif, int on); #else static inline int ide_acpi_init(void) { return 0; } +static inline bool ide_port_acpi(ide_hwif_t *hwif) { return 0; } static inline int ide_acpi_exec_tfs(ide_drive_t *drive) { return 0; } static inline void ide_acpi_get_timing(ide_hwif_t *hwif) { ; } static inline void ide_acpi_push_timing(ide_hwif_t *hwif) { ; } -- cgit v1.2.3 From 57e7986ed142417498155ebcd5eaf617ac37136d Mon Sep 17 00:00:00 2001 From: Paul Mackerras Date: Tue, 30 Jun 2009 16:07:19 +1000 Subject: perf_counter: Provide a way to enable counters on exec This provides a way to mark a counter to be enabled on the next exec. This is useful for measuring the total activity of a program without including overhead from the process that launches it. This also changes the perf stat command to use this new facility. Signed-off-by: Paul Mackerras Cc: Peter Zijlstra LKML-Reference: <19017.43927.838745.689203@cargo.ozlabs.ibm.com> Signed-off-by: Ingo Molnar --- include/linux/perf_counter.h | 3 ++- kernel/perf_counter.c | 50 ++++++++++++++++++++++++++++++++++++++++++++ tools/perf/builtin-stat.c | 6 +++--- 3 files changed, 55 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/include/linux/perf_counter.h b/include/linux/perf_counter.h index 3078e23c91eb..5e970c7d3fd5 100644 --- a/include/linux/perf_counter.h +++ b/include/linux/perf_counter.h @@ -179,8 +179,9 @@ struct perf_counter_attr { comm : 1, /* include comm data */ freq : 1, /* use freq, not period */ inherit_stat : 1, /* per task counts */ + enable_on_exec : 1, /* next exec enables */ - __reserved_1 : 52; + __reserved_1 : 51; __u32 wakeup_events; /* wakeup every n events */ __u32 __reserved_2; diff --git a/kernel/perf_counter.c b/kernel/perf_counter.c index 66ab1e9d1294..d55a50da2347 100644 --- a/kernel/perf_counter.c +++ b/kernel/perf_counter.c @@ -1428,6 +1428,53 @@ void perf_counter_task_tick(struct task_struct *curr, int cpu) perf_counter_task_sched_in(curr, cpu); } +/* + * Enable all of a task's counters that have been marked enable-on-exec. + * This expects task == current. + */ +static void perf_counter_enable_on_exec(struct task_struct *task) +{ + struct perf_counter_context *ctx; + struct perf_counter *counter; + unsigned long flags; + int enabled = 0; + + local_irq_save(flags); + ctx = task->perf_counter_ctxp; + if (!ctx || !ctx->nr_counters) + goto out; + + __perf_counter_task_sched_out(ctx); + + spin_lock(&ctx->lock); + + list_for_each_entry(counter, &ctx->counter_list, list_entry) { + if (!counter->attr.enable_on_exec) + continue; + counter->attr.enable_on_exec = 0; + if (counter->state >= PERF_COUNTER_STATE_INACTIVE) + continue; + counter->state = PERF_COUNTER_STATE_INACTIVE; + counter->tstamp_enabled = + ctx->time - counter->total_time_enabled; + enabled = 1; + } + + /* + * Unclone this context if we enabled any counter. + */ + if (enabled && ctx->parent_ctx) { + put_ctx(ctx->parent_ctx); + ctx->parent_ctx = NULL; + } + + spin_unlock(&ctx->lock); + + perf_counter_task_sched_in(task, smp_processor_id()); + out: + local_irq_restore(flags); +} + /* * Cross CPU call to read the hardware counter */ @@ -2949,6 +2996,9 @@ void perf_counter_comm(struct task_struct *task) { struct perf_comm_event comm_event; + if (task->perf_counter_ctxp) + perf_counter_enable_on_exec(task); + if (!atomic_read(&nr_comm_counters)) return; diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c index 201ef2367dcb..2e03524a1de0 100644 --- a/tools/perf/builtin-stat.c +++ b/tools/perf/builtin-stat.c @@ -116,8 +116,9 @@ static void create_perf_stat_counter(int counter, int pid) fd[cpu][counter], strerror(errno)); } } else { - attr->inherit = inherit; - attr->disabled = 1; + attr->inherit = inherit; + attr->disabled = 1; + attr->enable_on_exec = 1; fd[0][counter] = sys_perf_counter_open(attr, pid, -1, -1, 0); if (fd[0][counter] < 0 && verbose) @@ -262,7 +263,6 @@ static int run_perf_stat(int argc, const char **argv) * Enable counters and exec the command: */ t0 = rdclock(); - prctl(PR_TASK_PERF_COUNTERS_ENABLE); close(go_pipe[1]); wait(&status); -- cgit v1.2.3 From e0a43ddcc08c34dbd666d93600fd23914505f4aa Mon Sep 17 00:00:00 2001 From: Miklos Szeredi Date: Tue, 30 Jun 2009 20:12:23 +0200 Subject: fuse: allow umask processing in userspace MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This patch lets filesystems handle masking the file mode on creation. This is needed if filesystem is using ACLs. - The CREATE, MKDIR and MKNOD requests are extended with a "umask" parameter. - A new FUSE_DONT_MASK flag is added to the INIT request/reply. With this the filesystem may request that the create mode is not masked. CC: Jean-Pierre André Signed-off-by: Miklos Szeredi --- fs/fuse/dir.c | 20 +++++++++++++++++--- fs/fuse/fuse_i.h | 3 +++ fs/fuse/inode.c | 9 ++++++++- include/linux/fuse.h | 20 ++++++++++++++++++-- 4 files changed, 46 insertions(+), 6 deletions(-) (limited to 'include') diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c index b3089a083d30..6b700734e519 100644 --- a/fs/fuse/dir.c +++ b/fs/fuse/dir.c @@ -375,7 +375,7 @@ static int fuse_create_open(struct inode *dir, struct dentry *entry, int mode, struct fuse_conn *fc = get_fuse_conn(dir); struct fuse_req *req; struct fuse_req *forget_req; - struct fuse_open_in inarg; + struct fuse_create_in inarg; struct fuse_open_out outopen; struct fuse_entry_out outentry; struct fuse_file *ff; @@ -399,15 +399,20 @@ static int fuse_create_open(struct inode *dir, struct dentry *entry, int mode, if (!ff) goto out_put_request; + if (!fc->dont_mask) + mode &= ~current_umask(); + flags &= ~O_NOCTTY; memset(&inarg, 0, sizeof(inarg)); memset(&outentry, 0, sizeof(outentry)); inarg.flags = flags; inarg.mode = mode; + inarg.umask = current_umask(); req->in.h.opcode = FUSE_CREATE; req->in.h.nodeid = get_node_id(dir); req->in.numargs = 2; - req->in.args[0].size = sizeof(inarg); + req->in.args[0].size = fc->minor < 12 ? sizeof(struct fuse_open_in) : + sizeof(inarg); req->in.args[0].value = &inarg; req->in.args[1].size = entry->d_name.len + 1; req->in.args[1].value = entry->d_name.name; @@ -546,12 +551,17 @@ static int fuse_mknod(struct inode *dir, struct dentry *entry, int mode, if (IS_ERR(req)) return PTR_ERR(req); + if (!fc->dont_mask) + mode &= ~current_umask(); + memset(&inarg, 0, sizeof(inarg)); inarg.mode = mode; inarg.rdev = new_encode_dev(rdev); + inarg.umask = current_umask(); req->in.h.opcode = FUSE_MKNOD; req->in.numargs = 2; - req->in.args[0].size = sizeof(inarg); + req->in.args[0].size = fc->minor < 12 ? FUSE_COMPAT_MKNOD_IN_SIZE : + sizeof(inarg); req->in.args[0].value = &inarg; req->in.args[1].size = entry->d_name.len + 1; req->in.args[1].value = entry->d_name.name; @@ -578,8 +588,12 @@ static int fuse_mkdir(struct inode *dir, struct dentry *entry, int mode) if (IS_ERR(req)) return PTR_ERR(req); + if (!fc->dont_mask) + mode &= ~current_umask(); + memset(&inarg, 0, sizeof(inarg)); inarg.mode = mode; + inarg.umask = current_umask(); req->in.h.opcode = FUSE_MKDIR; req->in.numargs = 2; req->in.args[0].size = sizeof(inarg); diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h index aaf2f9ff970e..ede4f77b2d6c 100644 --- a/fs/fuse/fuse_i.h +++ b/fs/fuse/fuse_i.h @@ -446,6 +446,9 @@ struct fuse_conn { /** Do multi-page cached writes */ unsigned big_writes:1; + /** Don't apply umask to creation modes */ + unsigned dont_mask:1; + /** The number of requests waiting for completion */ atomic_t num_waiting; diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c index d8673ccf90b7..6cc501bd0187 100644 --- a/fs/fuse/inode.c +++ b/fs/fuse/inode.c @@ -725,6 +725,8 @@ static void process_init_reply(struct fuse_conn *fc, struct fuse_req *req) } if (arg->flags & FUSE_BIG_WRITES) fc->big_writes = 1; + if (arg->flags & FUSE_DONT_MASK) + fc->dont_mask = 1; } else { ra_pages = fc->max_read / PAGE_CACHE_SIZE; fc->no_lock = 1; @@ -748,7 +750,7 @@ static void fuse_send_init(struct fuse_conn *fc, struct fuse_req *req) arg->minor = FUSE_KERNEL_MINOR_VERSION; arg->max_readahead = fc->bdi.ra_pages * PAGE_CACHE_SIZE; arg->flags |= FUSE_ASYNC_READ | FUSE_POSIX_LOCKS | FUSE_ATOMIC_O_TRUNC | - FUSE_EXPORT_SUPPORT | FUSE_BIG_WRITES; + FUSE_EXPORT_SUPPORT | FUSE_BIG_WRITES | FUSE_DONT_MASK; req->in.h.opcode = FUSE_INIT; req->in.numargs = 1; req->in.args[0].size = sizeof(*arg); @@ -864,6 +866,11 @@ static int fuse_fill_super(struct super_block *sb, void *data, int silent) if (err) goto err_put_conn; + /* Handle umasking inside the fuse code */ + if (sb->s_flags & MS_POSIXACL) + fc->dont_mask = 1; + sb->s_flags |= MS_POSIXACL; + fc->release = fuse_free_conn; fc->flags = d.flags; fc->user_id = d.user_id; diff --git a/include/linux/fuse.h b/include/linux/fuse.h index d41ed593f79f..e2b816a62488 100644 --- a/include/linux/fuse.h +++ b/include/linux/fuse.h @@ -25,6 +25,9 @@ * - add IOCTL message * - add unsolicited notification support * - add POLL message and NOTIFY_POLL notification + * + * 7.12 + * - add umask flag to input argument of open, mknod and mkdir */ #ifndef _LINUX_FUSE_H @@ -36,7 +39,7 @@ #define FUSE_KERNEL_VERSION 7 /** Minor version number of this interface */ -#define FUSE_KERNEL_MINOR_VERSION 11 +#define FUSE_KERNEL_MINOR_VERSION 12 /** The node ID of the root inode */ #define FUSE_ROOT_ID 1 @@ -112,6 +115,7 @@ struct fuse_file_lock { * INIT request/reply flags * * FUSE_EXPORT_SUPPORT: filesystem handles lookups of "." and ".." + * FUSE_DONT_MASK: don't apply umask to file mode on create operations */ #define FUSE_ASYNC_READ (1 << 0) #define FUSE_POSIX_LOCKS (1 << 1) @@ -119,6 +123,7 @@ struct fuse_file_lock { #define FUSE_ATOMIC_O_TRUNC (1 << 3) #define FUSE_EXPORT_SUPPORT (1 << 4) #define FUSE_BIG_WRITES (1 << 5) +#define FUSE_DONT_MASK (1 << 6) /** * CUSE INIT request/reply flags @@ -262,14 +267,18 @@ struct fuse_attr_out { struct fuse_attr attr; }; +#define FUSE_COMPAT_MKNOD_IN_SIZE 8 + struct fuse_mknod_in { __u32 mode; __u32 rdev; + __u32 umask; + __u32 padding; }; struct fuse_mkdir_in { __u32 mode; - __u32 padding; + __u32 umask; }; struct fuse_rename_in { @@ -300,8 +309,15 @@ struct fuse_setattr_in { }; struct fuse_open_in { + __u32 flags; + __u32 unused; +}; + +struct fuse_create_in { __u32 flags; __u32 mode; + __u32 umask; + __u32 padding; }; struct fuse_open_out { -- cgit v1.2.3 From 3b463ae0c6264f70e5d4c0a9c46af20fed43c96e Mon Sep 17 00:00:00 2001 From: John Muir Date: Sun, 31 May 2009 11:13:57 -0400 Subject: fuse: invalidation reverse calls Add notification messages that allow the filesystem to invalidate VFS caches. Two notifications are added: 1) inode invalidation - invalidate cached attributes - invalidate a range of pages in the page cache (this is optional) 2) dentry invalidation - try to invalidate a subtree in the dentry cache Care must be taken while accessing the 'struct super_block' for the mount, as it can go away while an invalidation is in progress. To prevent this, introduce a rw-semaphore, that is taken for read during the invalidation and taken for write in the ->kill_sb callback. Cc: Csaba Henk Cc: Anand Avati Signed-off-by: Miklos Szeredi --- fs/fuse/dev.c | 81 ++++++++++++++++++++++++++++++++++++++++++++++++++++ fs/fuse/dir.c | 37 ++++++++++++++++++++++++ fs/fuse/fuse_i.h | 24 ++++++++++++++++ fs/fuse/inode.c | 59 ++++++++++++++++++++++++++++++++++++-- include/linux/fuse.h | 16 +++++++++++ 5 files changed, 214 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index 8a11a8c67c42..f58ecbc416c8 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -849,6 +849,81 @@ err: return err; } +static int fuse_notify_inval_inode(struct fuse_conn *fc, unsigned int size, + struct fuse_copy_state *cs) +{ + struct fuse_notify_inval_inode_out outarg; + int err = -EINVAL; + + if (size != sizeof(outarg)) + goto err; + + err = fuse_copy_one(cs, &outarg, sizeof(outarg)); + if (err) + goto err; + fuse_copy_finish(cs); + + down_read(&fc->killsb); + err = -ENOENT; + if (!fc->sb) + goto err_unlock; + + err = fuse_reverse_inval_inode(fc->sb, outarg.ino, + outarg.off, outarg.len); + +err_unlock: + up_read(&fc->killsb); + return err; + +err: + fuse_copy_finish(cs); + return err; +} + +static int fuse_notify_inval_entry(struct fuse_conn *fc, unsigned int size, + struct fuse_copy_state *cs) +{ + struct fuse_notify_inval_entry_out outarg; + int err = -EINVAL; + char buf[FUSE_NAME_MAX+1]; + struct qstr name; + + if (size < sizeof(outarg)) + goto err; + + err = fuse_copy_one(cs, &outarg, sizeof(outarg)); + if (err) + goto err; + + err = -ENAMETOOLONG; + if (outarg.namelen > FUSE_NAME_MAX) + goto err; + + name.name = buf; + name.len = outarg.namelen; + err = fuse_copy_one(cs, buf, outarg.namelen + 1); + if (err) + goto err; + fuse_copy_finish(cs); + buf[outarg.namelen] = 0; + name.hash = full_name_hash(name.name, name.len); + + down_read(&fc->killsb); + err = -ENOENT; + if (!fc->sb) + goto err_unlock; + + err = fuse_reverse_inval_entry(fc->sb, outarg.parent, &name); + +err_unlock: + up_read(&fc->killsb); + return err; + +err: + fuse_copy_finish(cs); + return err; +} + static int fuse_notify(struct fuse_conn *fc, enum fuse_notify_code code, unsigned int size, struct fuse_copy_state *cs) { @@ -856,6 +931,12 @@ static int fuse_notify(struct fuse_conn *fc, enum fuse_notify_code code, case FUSE_NOTIFY_POLL: return fuse_notify_poll(fc, size, cs); + case FUSE_NOTIFY_INVAL_INODE: + return fuse_notify_inval_inode(fc, size, cs); + + case FUSE_NOTIFY_INVAL_ENTRY: + return fuse_notify_inval_entry(fc, size, cs); + default: fuse_copy_finish(cs); return -EINVAL; diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c index 6b700734e519..e703654e7f40 100644 --- a/fs/fuse/dir.c +++ b/fs/fuse/dir.c @@ -859,6 +859,43 @@ int fuse_update_attributes(struct inode *inode, struct kstat *stat, return err; } +int fuse_reverse_inval_entry(struct super_block *sb, u64 parent_nodeid, + struct qstr *name) +{ + int err = -ENOTDIR; + struct inode *parent; + struct dentry *dir; + struct dentry *entry; + + parent = ilookup5(sb, parent_nodeid, fuse_inode_eq, &parent_nodeid); + if (!parent) + return -ENOENT; + + mutex_lock(&parent->i_mutex); + if (!S_ISDIR(parent->i_mode)) + goto unlock; + + err = -ENOENT; + dir = d_find_alias(parent); + if (!dir) + goto unlock; + + entry = d_lookup(dir, name); + dput(dir); + if (!entry) + goto unlock; + + fuse_invalidate_attr(parent); + fuse_invalidate_entry(entry); + dput(entry); + err = 0; + + unlock: + mutex_unlock(&parent->i_mutex); + iput(parent); + return err; +} + /* * Calling into a user-controlled filesystem gives the filesystem * daemon ptrace-like capabilities over the requester process. This diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h index ede4f77b2d6c..52b641fc0faf 100644 --- a/fs/fuse/fuse_i.h +++ b/fs/fuse/fuse_i.h @@ -484,6 +484,12 @@ struct fuse_conn { /** Called on final put */ void (*release)(struct fuse_conn *); + + /** Super block for this connection. */ + struct super_block *sb; + + /** Read/write semaphore to hold when accessing sb. */ + struct rw_semaphore killsb; }; static inline struct fuse_conn *get_fuse_conn_super(struct super_block *sb) @@ -511,6 +517,11 @@ extern const struct file_operations fuse_dev_operations; extern const struct dentry_operations fuse_dentry_operations; +/** + * Inode to nodeid comparison. + */ +int fuse_inode_eq(struct inode *inode, void *_nodeidp); + /** * Get a filled in inode */ @@ -711,6 +722,19 @@ void fuse_release_nowrite(struct inode *inode); u64 fuse_get_attr_version(struct fuse_conn *fc); +/** + * File-system tells the kernel to invalidate cache for the given node id. + */ +int fuse_reverse_inval_inode(struct super_block *sb, u64 nodeid, + loff_t offset, loff_t len); + +/** + * File-system tells the kernel to invalidate parent attributes and + * the dentry matching parent/name. + */ +int fuse_reverse_inval_entry(struct super_block *sb, u64 parent_nodeid, + struct qstr *name); + int fuse_do_open(struct fuse_conn *fc, u64 nodeid, struct file *file, bool isdir); ssize_t fuse_direct_io(struct file *file, const char __user *buf, diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c index 6cc501bd0187..f91ccc4a189d 100644 --- a/fs/fuse/inode.c +++ b/fs/fuse/inode.c @@ -206,7 +206,7 @@ static void fuse_init_inode(struct inode *inode, struct fuse_attr *attr) BUG(); } -static int fuse_inode_eq(struct inode *inode, void *_nodeidp) +int fuse_inode_eq(struct inode *inode, void *_nodeidp) { u64 nodeid = *(u64 *) _nodeidp; if (get_node_id(inode) == nodeid) @@ -257,6 +257,31 @@ struct inode *fuse_iget(struct super_block *sb, u64 nodeid, return inode; } +int fuse_reverse_inval_inode(struct super_block *sb, u64 nodeid, + loff_t offset, loff_t len) +{ + struct inode *inode; + pgoff_t pg_start; + pgoff_t pg_end; + + inode = ilookup5(sb, nodeid, fuse_inode_eq, &nodeid); + if (!inode) + return -ENOENT; + + fuse_invalidate_attr(inode); + if (offset >= 0) { + pg_start = offset >> PAGE_CACHE_SHIFT; + if (len <= 0) + pg_end = -1; + else + pg_end = (offset + len - 1) >> PAGE_CACHE_SHIFT; + invalidate_inode_pages2_range(inode->i_mapping, + pg_start, pg_end); + } + iput(inode); + return 0; +} + static void fuse_umount_begin(struct super_block *sb) { fuse_abort_conn(get_fuse_conn_super(sb)); @@ -480,6 +505,7 @@ void fuse_conn_init(struct fuse_conn *fc) memset(fc, 0, sizeof(*fc)); spin_lock_init(&fc->lock); mutex_init(&fc->inst_mutex); + init_rwsem(&fc->killsb); atomic_set(&fc->count, 1); init_waitqueue_head(&fc->waitq); init_waitqueue_head(&fc->blocked_waitq); @@ -862,6 +888,7 @@ static int fuse_fill_super(struct super_block *sb, void *data, int silent) fuse_conn_init(fc); fc->dev = sb->s_dev; + fc->sb = sb; err = fuse_bdi_init(fc, sb); if (err) goto err_put_conn; @@ -948,12 +975,25 @@ static int fuse_get_sb(struct file_system_type *fs_type, return get_sb_nodev(fs_type, flags, raw_data, fuse_fill_super, mnt); } +static void fuse_kill_sb_anon(struct super_block *sb) +{ + struct fuse_conn *fc = get_fuse_conn_super(sb); + + if (fc) { + down_write(&fc->killsb); + fc->sb = NULL; + up_write(&fc->killsb); + } + + kill_anon_super(sb); +} + static struct file_system_type fuse_fs_type = { .owner = THIS_MODULE, .name = "fuse", .fs_flags = FS_HAS_SUBTYPE, .get_sb = fuse_get_sb, - .kill_sb = kill_anon_super, + .kill_sb = fuse_kill_sb_anon, }; #ifdef CONFIG_BLOCK @@ -965,11 +1005,24 @@ static int fuse_get_sb_blk(struct file_system_type *fs_type, mnt); } +static void fuse_kill_sb_blk(struct super_block *sb) +{ + struct fuse_conn *fc = get_fuse_conn_super(sb); + + if (fc) { + down_write(&fc->killsb); + fc->sb = NULL; + up_write(&fc->killsb); + } + + kill_block_super(sb); +} + static struct file_system_type fuseblk_fs_type = { .owner = THIS_MODULE, .name = "fuseblk", .get_sb = fuse_get_sb_blk, - .kill_sb = kill_block_super, + .kill_sb = fuse_kill_sb_blk, .fs_flags = FS_REQUIRES_DEV | FS_HAS_SUBTYPE, }; diff --git a/include/linux/fuse.h b/include/linux/fuse.h index e2b816a62488..cf593bf9fd32 100644 --- a/include/linux/fuse.h +++ b/include/linux/fuse.h @@ -28,6 +28,8 @@ * * 7.12 * - add umask flag to input argument of open, mknod and mkdir + * - add notification messages for invalidation of inodes and + * directory entries */ #ifndef _LINUX_FUSE_H @@ -229,6 +231,8 @@ enum fuse_opcode { enum fuse_notify_code { FUSE_NOTIFY_POLL = 1, + FUSE_NOTIFY_INVAL_INODE = 2, + FUSE_NOTIFY_INVAL_ENTRY = 3, FUSE_NOTIFY_CODE_MAX, }; @@ -524,4 +528,16 @@ struct fuse_dirent { #define FUSE_DIRENT_SIZE(d) \ FUSE_DIRENT_ALIGN(FUSE_NAME_OFFSET + (d)->namelen) +struct fuse_notify_inval_inode_out { + __u64 ino; + __s64 off; + __s64 len; +}; + +struct fuse_notify_inval_entry_out { + __u64 parent; + __u32 namelen; + __u32 padding; +}; + #endif /* _LINUX_FUSE_H */ -- cgit v1.2.3 From b294a290d24d1196d68399cc3a9b8c50bfb55abd Mon Sep 17 00:00:00 2001 From: Andres Salomon Date: Tue, 30 Jun 2009 02:13:01 -0400 Subject: Revert "power: remove POWER_SUPPLY_PROP_CAPACITY_LEVEL" This reverts commit 8efe444038a205e79b38b7ad03878824901849a8 and 4cbc76eadf56399cd11fb736b33c53aec9caab8c. Richard@laptop.org was apparently using CAPACITY_LEVEL for debugging battery/EC problems, and was upset that it was removed. This readds it. Conflicts: Documentation/power_supply_class.txt Signed-off-by: Andres Salomon Signed-off-by: Anton Vorontsov --- Documentation/power/power_supply_class.txt | 2 ++ drivers/power/olpc_battery.c | 9 +++++++++ drivers/power/power_supply_sysfs.c | 6 ++++++ include/linux/power_supply.h | 10 ++++++++++ 4 files changed, 27 insertions(+) (limited to 'include') diff --git a/Documentation/power/power_supply_class.txt b/Documentation/power/power_supply_class.txt index c6cd4956047c..709d95571d7b 100644 --- a/Documentation/power/power_supply_class.txt +++ b/Documentation/power/power_supply_class.txt @@ -108,6 +108,8 @@ relative, time-based measurements. ENERGY_FULL, ENERGY_EMPTY - same as above but for energy. CAPACITY - capacity in percents. +CAPACITY_LEVEL - capacity level. This corresponds to +POWER_SUPPLY_CAPACITY_LEVEL_*. TEMP - temperature of the power supply. TEMP_AMBIENT - ambient temperature. diff --git a/drivers/power/olpc_battery.c b/drivers/power/olpc_battery.c index 58e419299cd6..3a589df09376 100644 --- a/drivers/power/olpc_battery.c +++ b/drivers/power/olpc_battery.c @@ -276,6 +276,14 @@ static int olpc_bat_get_property(struct power_supply *psy, return ret; val->intval = ec_byte; break; + case POWER_SUPPLY_PROP_CAPACITY_LEVEL: + if (ec_byte & BAT_STAT_FULL) + val->intval = POWER_SUPPLY_CAPACITY_LEVEL_FULL; + else if (ec_byte & BAT_STAT_LOW) + val->intval = POWER_SUPPLY_CAPACITY_LEVEL_LOW; + else + val->intval = POWER_SUPPLY_CAPACITY_LEVEL_NORMAL; + break; case POWER_SUPPLY_PROP_TEMP: ret = olpc_ec_cmd(EC_BAT_TEMP, NULL, 0, (void *)&ec_word, 2); if (ret) @@ -321,6 +329,7 @@ static enum power_supply_property olpc_bat_props[] = { POWER_SUPPLY_PROP_VOLTAGE_AVG, POWER_SUPPLY_PROP_CURRENT_AVG, POWER_SUPPLY_PROP_CAPACITY, + POWER_SUPPLY_PROP_CAPACITY_LEVEL, POWER_SUPPLY_PROP_TEMP, POWER_SUPPLY_PROP_TEMP_AMBIENT, POWER_SUPPLY_PROP_MANUFACTURER, diff --git a/drivers/power/power_supply_sysfs.c b/drivers/power/power_supply_sysfs.c index da73591017f9..9deabbde6fd6 100644 --- a/drivers/power/power_supply_sysfs.c +++ b/drivers/power/power_supply_sysfs.c @@ -51,6 +51,9 @@ static ssize_t power_supply_show_property(struct device *dev, "Unknown", "NiMH", "Li-ion", "Li-poly", "LiFe", "NiCd", "LiMn" }; + static char *capacity_level_text[] = { + "Unknown", "Critical", "Low", "Normal", "High", "Full" + }; ssize_t ret; struct power_supply *psy = dev_get_drvdata(dev); const ptrdiff_t off = attr - power_supply_attrs; @@ -71,6 +74,8 @@ static ssize_t power_supply_show_property(struct device *dev, return sprintf(buf, "%s\n", health_text[value.intval]); else if (off == POWER_SUPPLY_PROP_TECHNOLOGY) return sprintf(buf, "%s\n", technology_text[value.intval]); + else if (off == POWER_SUPPLY_PROP_CAPACITY_LEVEL) + return sprintf(buf, "%s\n", capacity_level_text[value.intval]); else if (off >= POWER_SUPPLY_PROP_MODEL_NAME) return sprintf(buf, "%s\n", value.strval); @@ -109,6 +114,7 @@ static struct device_attribute power_supply_attrs[] = { POWER_SUPPLY_ATTR(energy_now), POWER_SUPPLY_ATTR(energy_avg), POWER_SUPPLY_ATTR(capacity), + POWER_SUPPLY_ATTR(capacity_level), POWER_SUPPLY_ATTR(temp), POWER_SUPPLY_ATTR(temp_ambient), POWER_SUPPLY_ATTR(time_to_empty_now), diff --git a/include/linux/power_supply.h b/include/linux/power_supply.h index 594c494ac3f0..0ab6aa171241 100644 --- a/include/linux/power_supply.h +++ b/include/linux/power_supply.h @@ -58,6 +58,15 @@ enum { POWER_SUPPLY_TECHNOLOGY_LiMn, }; +enum { + POWER_SUPPLY_CAPACITY_LEVEL_UNKNOWN = 0, + POWER_SUPPLY_CAPACITY_LEVEL_CRITICAL, + POWER_SUPPLY_CAPACITY_LEVEL_LOW, + POWER_SUPPLY_CAPACITY_LEVEL_NORMAL, + POWER_SUPPLY_CAPACITY_LEVEL_HIGH, + POWER_SUPPLY_CAPACITY_LEVEL_FULL, +}; + enum power_supply_property { /* Properties of type `int' */ POWER_SUPPLY_PROP_STATUS = 0, @@ -89,6 +98,7 @@ enum power_supply_property { POWER_SUPPLY_PROP_ENERGY_NOW, POWER_SUPPLY_PROP_ENERGY_AVG, POWER_SUPPLY_PROP_CAPACITY, /* in percents! */ + POWER_SUPPLY_PROP_CAPACITY_LEVEL, POWER_SUPPLY_PROP_TEMP, POWER_SUPPLY_PROP_TEMP_AMBIENT, POWER_SUPPLY_PROP_TIME_TO_EMPTY_NOW, -- cgit v1.2.3 From 133890103b9de08904f909995973e4b5c08a780e Mon Sep 17 00:00:00 2001 From: Davide Libenzi Date: Tue, 30 Jun 2009 11:41:11 -0700 Subject: eventfd: revised interface and cleanups Change the eventfd interface to de-couple the eventfd memory context, from the file pointer instance. Without such change, there is no clean way to racely free handle the POLLHUP event sent when the last instance of the file* goes away. Also, now the internal eventfd APIs are using the eventfd context instead of the file*. This patch is required by KVM's IRQfd code, which is still under development. Signed-off-by: Davide Libenzi Cc: Gregory Haskins Cc: Rusty Russell Cc: Benjamin LaHaise Cc: Avi Kivity Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/lguest/lg.h | 2 +- drivers/lguest/lguest_user.c | 4 +- fs/aio.c | 24 +++------ fs/eventfd.c | 122 ++++++++++++++++++++++++++++++++++++++----- include/linux/aio.h | 4 +- include/linux/eventfd.h | 35 ++++++++++--- 6 files changed, 149 insertions(+), 42 deletions(-) (limited to 'include') diff --git a/drivers/lguest/lg.h b/drivers/lguest/lg.h index d4e8979735cb..9c3138265f8e 100644 --- a/drivers/lguest/lg.h +++ b/drivers/lguest/lg.h @@ -82,7 +82,7 @@ struct lg_cpu { struct lg_eventfd { unsigned long addr; - struct file *event; + struct eventfd_ctx *event; }; struct lg_eventfd_map { diff --git a/drivers/lguest/lguest_user.c b/drivers/lguest/lguest_user.c index 32e297121058..9f9a2953b383 100644 --- a/drivers/lguest/lguest_user.c +++ b/drivers/lguest/lguest_user.c @@ -50,7 +50,7 @@ static int add_eventfd(struct lguest *lg, unsigned long addr, int fd) /* Now append new entry. */ new->map[new->num].addr = addr; - new->map[new->num].event = eventfd_fget(fd); + new->map[new->num].event = eventfd_ctx_fdget(fd); if (IS_ERR(new->map[new->num].event)) { kfree(new); return PTR_ERR(new->map[new->num].event); @@ -357,7 +357,7 @@ static int close(struct inode *inode, struct file *file) /* Release any eventfds they registered. */ for (i = 0; i < lg->eventfds->num; i++) - fput(lg->eventfds->map[i].event); + eventfd_ctx_put(lg->eventfds->map[i].event); kfree(lg->eventfds); /* If lg->dead doesn't contain an error code it will be NULL or a diff --git a/fs/aio.c b/fs/aio.c index 76da12537956..d065b2c3273e 100644 --- a/fs/aio.c +++ b/fs/aio.c @@ -485,6 +485,8 @@ static inline void really_put_req(struct kioctx *ctx, struct kiocb *req) { assert_spin_locked(&ctx->ctx_lock); + if (req->ki_eventfd != NULL) + eventfd_ctx_put(req->ki_eventfd); if (req->ki_dtor) req->ki_dtor(req); if (req->ki_iovec != &req->ki_inline_vec) @@ -509,8 +511,6 @@ static void aio_fput_routine(struct work_struct *data) /* Complete the fput(s) */ if (req->ki_filp != NULL) __fput(req->ki_filp); - if (req->ki_eventfd != NULL) - __fput(req->ki_eventfd); /* Link the iocb into the context's free list */ spin_lock_irq(&ctx->ctx_lock); @@ -528,8 +528,6 @@ static void aio_fput_routine(struct work_struct *data) */ static int __aio_put_req(struct kioctx *ctx, struct kiocb *req) { - int schedule_putreq = 0; - dprintk(KERN_DEBUG "aio_put(%p): f_count=%ld\n", req, atomic_long_read(&req->ki_filp->f_count)); @@ -549,24 +547,16 @@ static int __aio_put_req(struct kioctx *ctx, struct kiocb *req) * we would not be holding the last reference to the file*, so * this function will be executed w/out any aio kthread wakeup. */ - if (unlikely(atomic_long_dec_and_test(&req->ki_filp->f_count))) - schedule_putreq++; - else - req->ki_filp = NULL; - if (req->ki_eventfd != NULL) { - if (unlikely(atomic_long_dec_and_test(&req->ki_eventfd->f_count))) - schedule_putreq++; - else - req->ki_eventfd = NULL; - } - if (unlikely(schedule_putreq)) { + if (unlikely(atomic_long_dec_and_test(&req->ki_filp->f_count))) { get_ioctx(ctx); spin_lock(&fput_lock); list_add(&req->ki_list, &fput_head); spin_unlock(&fput_lock); queue_work(aio_wq, &fput_work); - } else + } else { + req->ki_filp = NULL; really_put_req(ctx, req); + } return 1; } @@ -1622,7 +1612,7 @@ static int io_submit_one(struct kioctx *ctx, struct iocb __user *user_iocb, * an eventfd() fd, and will be signaled for each completed * event using the eventfd_signal() function. */ - req->ki_eventfd = eventfd_fget((int) iocb->aio_resfd); + req->ki_eventfd = eventfd_ctx_fdget((int) iocb->aio_resfd); if (IS_ERR(req->ki_eventfd)) { ret = PTR_ERR(req->ki_eventfd); req->ki_eventfd = NULL; diff --git a/fs/eventfd.c b/fs/eventfd.c index 3f0e1974abdc..31d12de83a2a 100644 --- a/fs/eventfd.c +++ b/fs/eventfd.c @@ -14,35 +14,44 @@ #include #include #include -#include #include #include +#include +#include struct eventfd_ctx { + struct kref kref; wait_queue_head_t wqh; /* * Every time that a write(2) is performed on an eventfd, the * value of the __u64 being written is added to "count" and a * wakeup is performed on "wqh". A read(2) will return the "count" * value to userspace, and will reset "count" to zero. The kernel - * size eventfd_signal() also, adds to the "count" counter and + * side eventfd_signal() also, adds to the "count" counter and * issue a wakeup. */ __u64 count; unsigned int flags; }; -/* - * Adds "n" to the eventfd counter "count". Returns "n" in case of - * success, or a value lower then "n" in case of coutner overflow. - * This function is supposed to be called by the kernel in paths - * that do not allow sleeping. In this function we allow the counter - * to reach the ULLONG_MAX value, and we signal this as overflow - * condition by returining a POLLERR to poll(2). +/** + * eventfd_signal - Adds @n to the eventfd counter. + * @ctx: [in] Pointer to the eventfd context. + * @n: [in] Value of the counter to be added to the eventfd internal counter. + * The value cannot be negative. + * + * This function is supposed to be called by the kernel in paths that do not + * allow sleeping. In this function we allow the counter to reach the ULLONG_MAX + * value, and we signal this as overflow condition by returining a POLLERR + * to poll(2). + * + * Returns @n in case of success, a non-negative number lower than @n in case + * of overflow, or the following error codes: + * + * -EINVAL : The value of @n is negative. */ -int eventfd_signal(struct file *file, int n) +int eventfd_signal(struct eventfd_ctx *ctx, int n) { - struct eventfd_ctx *ctx = file->private_data; unsigned long flags; if (n < 0) @@ -59,9 +68,45 @@ int eventfd_signal(struct file *file, int n) } EXPORT_SYMBOL_GPL(eventfd_signal); +static void eventfd_free(struct kref *kref) +{ + struct eventfd_ctx *ctx = container_of(kref, struct eventfd_ctx, kref); + + kfree(ctx); +} + +/** + * eventfd_ctx_get - Acquires a reference to the internal eventfd context. + * @ctx: [in] Pointer to the eventfd context. + * + * Returns: In case of success, returns a pointer to the eventfd context. + */ +struct eventfd_ctx *eventfd_ctx_get(struct eventfd_ctx *ctx) +{ + kref_get(&ctx->kref); + return ctx; +} +EXPORT_SYMBOL_GPL(eventfd_ctx_get); + +/** + * eventfd_ctx_put - Releases a reference to the internal eventfd context. + * @ctx: [in] Pointer to eventfd context. + * + * The eventfd context reference must have been previously acquired either + * with eventfd_ctx_get() or eventfd_ctx_fdget()). + */ +void eventfd_ctx_put(struct eventfd_ctx *ctx) +{ + kref_put(&ctx->kref, eventfd_free); +} +EXPORT_SYMBOL_GPL(eventfd_ctx_put); + static int eventfd_release(struct inode *inode, struct file *file) { - kfree(file->private_data); + struct eventfd_ctx *ctx = file->private_data; + + wake_up_poll(&ctx->wqh, POLLHUP); + eventfd_ctx_put(ctx); return 0; } @@ -185,6 +230,16 @@ static const struct file_operations eventfd_fops = { .write = eventfd_write, }; +/** + * eventfd_fget - Acquire a reference of an eventfd file descriptor. + * @fd: [in] Eventfd file descriptor. + * + * Returns a pointer to the eventfd file structure in case of success, or the + * following error pointer: + * + * -EBADF : Invalid @fd file descriptor. + * -EINVAL : The @fd file descriptor is not an eventfd file. + */ struct file *eventfd_fget(int fd) { struct file *file; @@ -201,6 +256,48 @@ struct file *eventfd_fget(int fd) } EXPORT_SYMBOL_GPL(eventfd_fget); +/** + * eventfd_ctx_fdget - Acquires a reference to the internal eventfd context. + * @fd: [in] Eventfd file descriptor. + * + * Returns a pointer to the internal eventfd context, otherwise the error + * pointers returned by the following functions: + * + * eventfd_fget + */ +struct eventfd_ctx *eventfd_ctx_fdget(int fd) +{ + struct file *file; + struct eventfd_ctx *ctx; + + file = eventfd_fget(fd); + if (IS_ERR(file)) + return (struct eventfd_ctx *) file; + ctx = eventfd_ctx_get(file->private_data); + fput(file); + + return ctx; +} +EXPORT_SYMBOL_GPL(eventfd_ctx_fdget); + +/** + * eventfd_ctx_fileget - Acquires a reference to the internal eventfd context. + * @file: [in] Eventfd file pointer. + * + * Returns a pointer to the internal eventfd context, otherwise the error + * pointer: + * + * -EINVAL : The @fd file descriptor is not an eventfd file. + */ +struct eventfd_ctx *eventfd_ctx_fileget(struct file *file) +{ + if (file->f_op != &eventfd_fops) + return ERR_PTR(-EINVAL); + + return eventfd_ctx_get(file->private_data); +} +EXPORT_SYMBOL_GPL(eventfd_ctx_fileget); + SYSCALL_DEFINE2(eventfd2, unsigned int, count, int, flags) { int fd; @@ -217,6 +314,7 @@ SYSCALL_DEFINE2(eventfd2, unsigned int, count, int, flags) if (!ctx) return -ENOMEM; + kref_init(&ctx->kref); init_waitqueue_head(&ctx->wqh); ctx->count = count; ctx->flags = flags; diff --git a/include/linux/aio.h b/include/linux/aio.h index b16a957030f8..47f7d932a01d 100644 --- a/include/linux/aio.h +++ b/include/linux/aio.h @@ -121,9 +121,9 @@ struct kiocb { /* * If the aio_resfd field of the userspace iocb is not zero, - * this is the underlying file* to deliver event to. + * this is the underlying eventfd context to deliver events to. */ - struct file *ki_eventfd; + struct eventfd_ctx *ki_eventfd; }; #define is_sync_kiocb(iocb) ((iocb)->ki_key == KIOCB_SYNC_KEY) diff --git a/include/linux/eventfd.h b/include/linux/eventfd.h index f45a8ae5f828..3b85ba6479f4 100644 --- a/include/linux/eventfd.h +++ b/include/linux/eventfd.h @@ -8,10 +8,8 @@ #ifndef _LINUX_EVENTFD_H #define _LINUX_EVENTFD_H -#ifdef CONFIG_EVENTFD - -/* For O_CLOEXEC and O_NONBLOCK */ #include +#include /* * CAREFUL: Check include/asm-generic/fcntl.h when defining @@ -27,16 +25,37 @@ #define EFD_SHARED_FCNTL_FLAGS (O_CLOEXEC | O_NONBLOCK) #define EFD_FLAGS_SET (EFD_SHARED_FCNTL_FLAGS | EFD_SEMAPHORE) +#ifdef CONFIG_EVENTFD + +struct eventfd_ctx *eventfd_ctx_get(struct eventfd_ctx *ctx); +void eventfd_ctx_put(struct eventfd_ctx *ctx); struct file *eventfd_fget(int fd); -int eventfd_signal(struct file *file, int n); +struct eventfd_ctx *eventfd_ctx_fdget(int fd); +struct eventfd_ctx *eventfd_ctx_fileget(struct file *file); +int eventfd_signal(struct eventfd_ctx *ctx, int n); #else /* CONFIG_EVENTFD */ -#define eventfd_fget(fd) ERR_PTR(-ENOSYS) -static inline int eventfd_signal(struct file *file, int n) -{ return 0; } +/* + * Ugly ugly ugly error layer to support modules that uses eventfd but + * pretend to work in !CONFIG_EVENTFD configurations. Namely, AIO. + */ +static inline struct eventfd_ctx *eventfd_ctx_fdget(int fd) +{ + return ERR_PTR(-ENOSYS); +} + +static inline int eventfd_signal(struct eventfd_ctx *ctx, int n) +{ + return -ENOSYS; +} + +static inline void eventfd_ctx_put(struct eventfd_ctx *ctx) +{ + +} -#endif /* CONFIG_EVENTFD */ +#endif #endif /* _LINUX_EVENTFD_H */ -- cgit v1.2.3 From 2a2325e6e8a3782795fb520220c36fd805775972 Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Tue, 30 Jun 2009 11:41:13 -0700 Subject: gcov: fix __ctors_start alignment The ctors section for each object file is eight byte aligned (on 64 bit). However the __ctors_start symbol starts at an arbitrary address dependent on the size of the previous sections. Therefore the linker may add some zeroes after __ctors_start to make sure the ctors contents are properly aligned. However the extra zeroes at the beginning aren't expected by the code. When walking the functions pointers contained in there and extra zeroes are added this may result in random jumps. So make sure that the __ctors_start symbol is always aligned as well. Fixes this crash on an allyesconfig on s390: [ 0.582482] Kernel BUG at 0000000000000012 [verbose debug info unavailable] [ 0.582489] illegal operation: 0001 [#1] SMP DEBUG_PAGEALLOC [ 0.582496] Modules linked in: [ 0.582501] CPU: 0 Tainted: G W 2.6.31-rc1-dirty #273 [ 0.582506] Process swapper (pid: 1, task: 000000003f218000, ksp: 000000003f2238e8) [ 0.582510] Krnl PSW : 0704200180000000 0000000000000012 (0x12) [ 0.582518] R:0 T:1 IO:1 EX:1 Key:0 M:1 W:0 P:0 AS:0 CC:2 PM:0 EA:3 [ 0.582524] Krnl GPRS: 0000000000036727 0000000000000010 0000000000000001 0000000000000001 [ 0.582529] 00000000001dfefa 0000000000000000 0000000000000000 0000000000000040 [ 0.582534] 0000000001fff0f0 0000000001790628 0000000002296048 0000000002296048 [ 0.582540] 00000000020c438e 0000000001786000 0000000002014a66 000000003f223e60 [ 0.582553] Krnl Code:>0000000000000012: 0000 unknown [ 0.582559] 0000000000000014: 0000 unknown [ 0.582564] 0000000000000016: 0000 unknown [ 0.582570] 0000000000000018: 0000 unknown [ 0.582575] 000000000000001a: 0000 unknown [ 0.582580] 000000000000001c: 0000 unknown [ 0.582585] 000000000000001e: 0000 unknown [ 0.582591] 0000000000000020: 0000 unknown [ 0.582596] Call Trace: [ 0.582599] ([<0000000002014a46>] kernel_init+0x622/0x7a0) [ 0.582607] [<0000000000113e22>] kernel_thread_starter+0x6/0xc [ 0.582615] [<0000000000113e1c>] kernel_thread_starter+0x0/0xc [ 0.582621] INFO: lockdep is turned off. [ 0.582624] Last Breaking-Event-Address: [ 0.582627] [<0000000002014a64>] kernel_init+0x640/0x7a0 Cc: Peter Oberparleiter Cc: Ingo Molnar Cc: Martin Schwidefsky Signed-off-by: Heiko Carstens Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/asm-generic/vmlinux.lds.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/asm-generic/vmlinux.lds.h b/include/asm-generic/vmlinux.lds.h index 92b73b6140ff..dccdbed05848 100644 --- a/include/asm-generic/vmlinux.lds.h +++ b/include/asm-generic/vmlinux.lds.h @@ -441,7 +441,8 @@ } #ifdef CONFIG_CONSTRUCTORS -#define KERNEL_CTORS() VMLINUX_SYMBOL(__ctors_start) = .; \ +#define KERNEL_CTORS() . = ALIGN(8); \ + VMLINUX_SYMBOL(__ctors_start) = .; \ *(.ctors) \ VMLINUX_SYMBOL(__ctors_end) = .; #else -- cgit v1.2.3 From b01e8dc34379f4ba2f454390e340a025edbaaa7e Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Tue, 30 Jun 2009 11:41:18 -0700 Subject: alpha: fix percpu build breakage alpha percpu access requires custom SHIFT_PERCPU_PTR() definition for modules to work around addressing range limitation. This is done via generating inline assembly using C preprocessing which forces the assembler to generate external reference. This happens behind the compiler's back and makes the compiler think that static percpu variables in modules are unused. This used to be worked around by using __unused attribute for percpu variables which prevent the compiler from omitting the variable; however, recent declare/definition attribute unification change broke this as __used can't be used for declaration. Also, in the process, PER_CPU_ATTRIBUTES definition in alpha percpu.h got broken. This patch adds PER_CPU_DEF_ATTRIBUTES which is only used for definitions and make alpha use it to add __used for percpu variables in modules. This also fixes the PER_CPU_ATTRIBUTES double definition bug. Signed-off-by: Tejun Heo Tested-by: maximilian attems Acked-by: Ivan Kokshaysky Cc: Richard Henderson Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/alpha/include/asm/percpu.h | 6 +++--- include/asm-generic/percpu.h | 4 ++++ include/linux/percpu-defs.h | 3 ++- 3 files changed, 9 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/arch/alpha/include/asm/percpu.h b/arch/alpha/include/asm/percpu.h index 06c5c7a4afd3..b663f1f10b6a 100644 --- a/arch/alpha/include/asm/percpu.h +++ b/arch/alpha/include/asm/percpu.h @@ -30,7 +30,7 @@ extern unsigned long __per_cpu_offset[NR_CPUS]; #ifndef MODULE #define SHIFT_PERCPU_PTR(var, offset) RELOC_HIDE(&per_cpu_var(var), (offset)) -#define PER_CPU_ATTRIBUTES +#define PER_CPU_DEF_ATTRIBUTES #else /* * To calculate addresses of locally defined variables, GCC uses 32-bit @@ -49,7 +49,7 @@ extern unsigned long __per_cpu_offset[NR_CPUS]; : "=&r"(__ptr), "=&r"(tmp_gp)); \ (typeof(&per_cpu_var(var)))(__ptr + (offset)); }) -#define PER_CPU_ATTRIBUTES __used +#define PER_CPU_DEF_ATTRIBUTES __used #endif /* MODULE */ @@ -71,7 +71,7 @@ extern unsigned long __per_cpu_offset[NR_CPUS]; #define __get_cpu_var(var) per_cpu_var(var) #define __raw_get_cpu_var(var) per_cpu_var(var) -#define PER_CPU_ATTRIBUTES +#define PER_CPU_DEF_ATTRIBUTES #endif /* SMP */ diff --git a/include/asm-generic/percpu.h b/include/asm-generic/percpu.h index d7d50d7ee51e..aa00800adacc 100644 --- a/include/asm-generic/percpu.h +++ b/include/asm-generic/percpu.h @@ -97,4 +97,8 @@ extern void setup_per_cpu_areas(void); #define PER_CPU_ATTRIBUTES #endif +#ifndef PER_CPU_DEF_ATTRIBUTES +#define PER_CPU_DEF_ATTRIBUTES +#endif + #endif /* _ASM_GENERIC_PERCPU_H_ */ diff --git a/include/linux/percpu-defs.h b/include/linux/percpu-defs.h index 8f921d74f49f..68438e18fff4 100644 --- a/include/linux/percpu-defs.h +++ b/include/linux/percpu-defs.h @@ -24,7 +24,8 @@ #define DEFINE_PER_CPU_SECTION(type, name, section) \ __attribute__((__section__(PER_CPU_BASE_SECTION section))) \ - PER_CPU_ATTRIBUTES __typeof__(type) per_cpu__##name + PER_CPU_ATTRIBUTES PER_CPU_DEF_ATTRIBUTES \ + __typeof__(type) per_cpu__##name /* * Variant on the per-CPU variable declaration/definition theme used for -- cgit v1.2.3 From c4285b47b0514e2103584ee829246f813e7ae323 Mon Sep 17 00:00:00 2001 From: Michael Buesch Date: Tue, 30 Jun 2009 11:41:21 -0700 Subject: parport/serial: add support for NetMos 9901 Multi-IO card Add support for the PCI-Express NetMos 9901 Multi-IO card. 0001:06:00.0 Serial controller [0700]: NetMos Technology Device [9710:9901] (prog-if 02 [16550]) Subsystem: Device [a000:1000] Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx- Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- SERR- Kernel driver in use: serial Kernel modules: 8250_pci 0001:06:00.1 Serial controller [0700]: NetMos Technology Device [9710:9901] (prog-if 02 [16550]) Subsystem: Device [a000:1000] Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx- Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- SERR- Kernel driver in use: serial Kernel modules: 8250_pci 0001:06:00.2 Parallel controller [0701]: NetMos Technology Device [9710:9901] (prog-if 03 [IEEE1284]) Subsystem: Device [a000:2000] Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx- Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- SERR- Region 2: Memory at 80101000 (32-bit, non-prefetchable) [size=4K] Region 4: Memory at 80100000 (32-bit, non-prefetchable) [size=4K] Capabilities: Kernel driver in use: parport_pc Kernel modules: parport_pc [ 16.760181] PCI parallel port detected: 416c:0100, I/O at 0x812010(0x0), IRQ 65 [ 16.760225] parport0: PC-style at 0x812010, irq 65 [PCSPP,TRISTATE,EPP] [ 16.851842] serial 0001:06:00.0: enabling device (0004 -> 0007) [ 16.883776] 0001:06:00.0: ttyS0 at I/O 0x812030 (irq = 65) is a ST16650V2 [ 16.893832] serial 0001:06:00.1: enabling device (0004 -> 0007) [ 16.926537] 0001:06:00.1: ttyS1 at I/O 0x812020 (irq = 65) is a ST16650V2 Signed-off-by: Michael Buesch Cc: Alan Cox Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/parport/parport_pc.c | 5 ++++- drivers/serial/8250_pci.c | 6 ++++++ include/linux/pci_ids.h | 1 + 3 files changed, 11 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/drivers/parport/parport_pc.c b/drivers/parport/parport_pc.c index 1032d5fdbd42..2597145a066e 100644 --- a/drivers/parport/parport_pc.c +++ b/drivers/parport/parport_pc.c @@ -2907,6 +2907,7 @@ enum parport_pc_pci_cards { netmos_9755, netmos_9805, netmos_9815, + netmos_9901, quatech_sppxp100, }; @@ -2987,7 +2988,7 @@ static struct parport_pc_pci { /* netmos_9755 */ { 2, { { 0, 1 }, { 2, 3 },} }, /* netmos_9805 */ { 1, { { 0, -1 }, } }, /* netmos_9815 */ { 2, { { 0, -1 }, { 2, -1 }, } }, - + /* netmos_9901 */ { 1, { { 0, -1 }, } }, /* quatech_sppxp100 */ { 1, { { 0, 1 }, } }, }; @@ -3089,6 +3090,8 @@ static const struct pci_device_id parport_pc_pci_tbl[] = { PCI_ANY_ID, PCI_ANY_ID, 0, 0, netmos_9805 }, { PCI_VENDOR_ID_NETMOS, PCI_DEVICE_ID_NETMOS_9815, PCI_ANY_ID, PCI_ANY_ID, 0, 0, netmos_9815 }, + { PCI_VENDOR_ID_NETMOS, PCI_DEVICE_ID_NETMOS_9901, + 0xA000, 0x2000, 0, 0, netmos_9901 }, /* Quatech SPPXP-100 Parallel port PCI ExpressCard */ { PCI_VENDOR_ID_QUATECH, PCI_DEVICE_ID_QUATECH_SPPXP_100, PCI_ANY_ID, PCI_ANY_ID, 0, 0, quatech_sppxp100 }, diff --git a/drivers/serial/8250_pci.c b/drivers/serial/8250_pci.c index a07015d646dd..6160e03f410c 100644 --- a/drivers/serial/8250_pci.c +++ b/drivers/serial/8250_pci.c @@ -759,6 +759,8 @@ static int pci_netmos_init(struct pci_dev *dev) /* subdevice 0x00PS means

parallel, serial */ unsigned int num_serial = dev->subsystem_device & 0xf; + if (dev->device == PCI_DEVICE_ID_NETMOS_9901) + return 0; if (dev->subsystem_vendor == PCI_VENDOR_ID_IBM && dev->subsystem_device == 0x0299) return 0; @@ -3557,6 +3559,10 @@ static struct pci_device_id serial_pci_tbl[] = { PCI_VENDOR_ID_IBM, 0x0299, 0, 0, pbn_b0_bt_2_115200 }, + { PCI_VENDOR_ID_NETMOS, PCI_DEVICE_ID_NETMOS_9901, + 0xA000, 0x1000, + 0, 0, pbn_b0_1_115200 }, + /* * These entries match devices with class COMMUNICATION_SERIAL, * COMMUNICATION_MODEM or COMMUNICATION_MULTISERIAL diff --git a/include/linux/pci_ids.h b/include/linux/pci_ids.h index a3b000365795..73b46b6b904f 100644 --- a/include/linux/pci_ids.h +++ b/include/linux/pci_ids.h @@ -2645,6 +2645,7 @@ #define PCI_DEVICE_ID_NETMOS_9835 0x9835 #define PCI_DEVICE_ID_NETMOS_9845 0x9845 #define PCI_DEVICE_ID_NETMOS_9855 0x9855 +#define PCI_DEVICE_ID_NETMOS_9901 0x9901 #define PCI_VENDOR_ID_3COM_2 0xa727 -- cgit v1.2.3 From 341c87bf346f57748230628c5ad6ee69219250e8 Mon Sep 17 00:00:00 2001 From: KAMEZAWA Hiroyuki Date: Tue, 30 Jun 2009 11:41:23 -0700 Subject: elf: limit max map count to safe value With ELF, at generating coredump, some more headers other than used vmas are added. When max_map_count == 65536, a core generated by following kinds of code can be unreadable because the number of ELF's program header is written in 16bit in Ehdr (please see elf.h) and the number overflows. == ... = mmap(); (munmap, mprotect, etc...) if (failed) abort(); == This can happen in mmap/munmap/mprotect/etc...which calls split_vma(). I think 65536 is not safe as _default_ and reduce it to 65530 is good for avoiding unexpected corrupted core. Anyway, max_map_count can be enlarged by sysctl if a user is brave.. Signed-off-by: KAMEZAWA Hiroyuki Cc: Hugh Dickins Cc: Jakub Jelinek Acked-by: Roland McGrath Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/binfmt_elf.c | 5 ++++- include/linux/sched.h | 16 ++++++++++++++-- 2 files changed, 18 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/fs/binfmt_elf.c b/fs/binfmt_elf.c index 9fa212b014a5..f1867900e459 100644 --- a/fs/binfmt_elf.c +++ b/fs/binfmt_elf.c @@ -1929,7 +1929,10 @@ static int elf_core_dump(long signr, struct pt_regs *regs, struct file *file, un elf = kmalloc(sizeof(*elf), GFP_KERNEL); if (!elf) goto out; - + /* + * The number of segs are recored into ELF header as 16bit value. + * Please check DEFAULT_MAX_MAP_COUNT definition when you modify here. + */ segs = current->mm->map_count; #ifdef ELF_CORE_EXTRA_PHDRS segs += ELF_CORE_EXTRA_PHDRS; diff --git a/include/linux/sched.h b/include/linux/sched.h index 4d0754269884..0085d758d645 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -349,8 +349,20 @@ extern int mutex_spin_on_owner(struct mutex *lock, struct thread_info *owner); struct nsproxy; struct user_namespace; -/* Maximum number of active map areas.. This is a random (large) number */ -#define DEFAULT_MAX_MAP_COUNT 65536 +/* + * Default maximum number of active map areas, this limits the number of vmas + * per mm struct. Users can overwrite this number by sysctl but there is a + * problem. + * + * When a program's coredump is generated as ELF format, a section is created + * per a vma. In ELF, the number of sections is represented in unsigned short. + * This means the number of sections should be smaller than 65535 at coredump. + * Because the kernel adds some informative sections to a image of program at + * generating coredump, we need some margin. The number of extra sections is + * 1-3 now and depends on arch. We use "5" as safe margin, here. + */ +#define MAPCOUNT_ELF_CORE_MARGIN (5) +#define DEFAULT_MAX_MAP_COUNT (USHORT_MAX - MAPCOUNT_ELF_CORE_MARGIN) extern int sysctl_max_map_count; -- cgit v1.2.3 From b55f627feeb9d48fdbde3835e18afbc76712e49b Mon Sep 17 00:00:00 2001 From: David Brownell Date: Tue, 30 Jun 2009 11:41:26 -0700 Subject: spi: new spi->mode bits Add two new spi_device.mode bits to accomodate more protocol options, and pass them through to usermode drivers: * SPI_NO_CS ... a second 3-wire variant, where the chipselect line is removed instead of a data line; transfers are still full duplex. This obviously has STRONG protocol implications since the chipselect transitions can't be used to synchronize state transitions with the SPI master. * SPI_READY ... defines open drain signal that's pulled low to pause the clock. This defines a 5-wire variant (normal 4-wire SPI plus READY) and two 4-wire variants (READY plus each of the 3-wire flavors). Such hardware flow control can be a big win. There are ADC converters and flash chips that expose READY signals, but not many host controllers support it today. The spi_bitbang code should be changed to use SPI_NO_CS instead of its current nonportable hack. That's a mode most hardware can easily support (unlike SPI_READY). Signed-off-by: David Brownell Cc: "Paulraj, Sandeep" Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- Documentation/spi/spidev_test.c | 10 +++++++++- drivers/spi/spidev.c | 17 +++++++++++------ include/linux/spi/spi.h | 2 ++ include/linux/spi/spidev.h | 2 ++ 4 files changed, 24 insertions(+), 7 deletions(-) (limited to 'include') diff --git a/Documentation/spi/spidev_test.c b/Documentation/spi/spidev_test.c index cf0e3ce0d526..c1a5aad3c75a 100644 --- a/Documentation/spi/spidev_test.c +++ b/Documentation/spi/spidev_test.c @@ -99,11 +99,13 @@ void parse_opts(int argc, char *argv[]) { "lsb", 0, 0, 'L' }, { "cs-high", 0, 0, 'C' }, { "3wire", 0, 0, '3' }, + { "no-cs", 0, 0, 'N' }, + { "ready", 0, 0, 'R' }, { NULL, 0, 0, 0 }, }; int c; - c = getopt_long(argc, argv, "D:s:d:b:lHOLC3", lopts, NULL); + c = getopt_long(argc, argv, "D:s:d:b:lHOLC3NR", lopts, NULL); if (c == -1) break; @@ -139,6 +141,12 @@ void parse_opts(int argc, char *argv[]) case '3': mode |= SPI_3WIRE; break; + case 'N': + mode |= SPI_NO_CS; + break; + case 'R': + mode |= SPI_READY; + break; default: print_usage(argv[0]); break; diff --git a/drivers/spi/spidev.c b/drivers/spi/spidev.c index 5d869c4d3eb2..606e7a40a8da 100644 --- a/drivers/spi/spidev.c +++ b/drivers/spi/spidev.c @@ -58,15 +58,20 @@ static unsigned long minors[N_SPI_MINORS / BITS_PER_LONG]; /* Bit masks for spi_device.mode management. Note that incorrect - * settings for CS_HIGH and 3WIRE can cause *lots* of trouble for other - * devices on a shared bus: CS_HIGH, because this device will be - * active when it shouldn't be; 3WIRE, because when active it won't - * behave as it should. + * settings for some settings can cause *lots* of trouble for other + * devices on a shared bus: * - * REVISIT should changing those two modes be privileged? + * - CS_HIGH ... this device will be active when it shouldn't be + * - 3WIRE ... when active, it won't behave as it should + * - NO_CS ... there will be no explicit message boundaries; this + * is completely incompatible with the shared bus model + * - READY ... transfers may proceed when they shouldn't. + * + * REVISIT should changing those flags be privileged? */ #define SPI_MODE_MASK (SPI_CPHA | SPI_CPOL | SPI_CS_HIGH \ - | SPI_LSB_FIRST | SPI_3WIRE | SPI_LOOP) + | SPI_LSB_FIRST | SPI_3WIRE | SPI_LOOP \ + | SPI_NO_CS | SPI_READY) struct spidev_data { dev_t devt; diff --git a/include/linux/spi/spi.h b/include/linux/spi/spi.h index 9c4cd27f4685..743c933ac4e7 100644 --- a/include/linux/spi/spi.h +++ b/include/linux/spi/spi.h @@ -80,6 +80,8 @@ struct spi_device { #define SPI_LSB_FIRST 0x08 /* per-word bits-on-wire */ #define SPI_3WIRE 0x10 /* SI/SO signals shared */ #define SPI_LOOP 0x20 /* loopback mode */ +#define SPI_NO_CS 0x40 /* 1 dev/bus, no chipselect */ +#define SPI_READY 0x80 /* slave pulls low to pause */ u8 bits_per_word; int irq; void *controller_state; diff --git a/include/linux/spi/spidev.h b/include/linux/spi/spidev.h index 95251ccd5a07..bf0570a84f7a 100644 --- a/include/linux/spi/spidev.h +++ b/include/linux/spi/spidev.h @@ -40,6 +40,8 @@ #define SPI_LSB_FIRST 0x08 #define SPI_3WIRE 0x10 #define SPI_LOOP 0x20 +#define SPI_NO_CS 0x40 +#define SPI_READY 0x80 /*---------------------------------------------------------------------------*/ -- cgit v1.2.3 From 70d6027ff2bc8bab180273b77e7ab3e8a62cca51 Mon Sep 17 00:00:00 2001 From: David Brownell Date: Tue, 30 Jun 2009 11:41:27 -0700 Subject: spi: add spi_master flag word Add a new spi_master.flags word listing constraints relevant to that controller. Define the first constraint bit: a half duplex restriction. Include that constraint in the OMAP1 MicroWire controller driver. Have the mmc_spi host be the first customer of this flag. Its coding relies heavily on full duplex transfers, so it must fail when the underlying controller driver won't perform them. (The spi_write_then_read routine could use it too: use the temporarily-withdrawn full-duplex speedup unless this flag is set, in which case the existing code applies. Similarly, any spi_master implementing only SPI_3WIRE should set the flag.) Signed-off-by: David Brownell Cc: Marek Szyprowski Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/mmc/host/mmc_spi.c | 6 ++++++ drivers/spi/omap_uwire.c | 2 ++ include/linux/spi/spi.h | 4 ++++ 3 files changed, 12 insertions(+) (limited to 'include') diff --git a/drivers/mmc/host/mmc_spi.c b/drivers/mmc/host/mmc_spi.c index 240608cc7ae9..a461017ce5ce 100644 --- a/drivers/mmc/host/mmc_spi.c +++ b/drivers/mmc/host/mmc_spi.c @@ -1313,6 +1313,12 @@ static int mmc_spi_probe(struct spi_device *spi) struct mmc_spi_host *host; int status; + /* We rely on full duplex transfers, mostly to reduce + * per-transfer overheads (by making fewer transfers). + */ + if (spi->master->flags & SPI_MASTER_HALF_DUPLEX) + return -EINVAL; + /* MMC and SD specs only seem to care that sampling is on the * rising edge ... meaning SPI modes 0 or 3. So either SPI mode * should be legit. We'll use mode 0 since the steady state is 0, diff --git a/drivers/spi/omap_uwire.c b/drivers/spi/omap_uwire.c index aa90ddb37066..8980a5640bd9 100644 --- a/drivers/spi/omap_uwire.c +++ b/drivers/spi/omap_uwire.c @@ -514,6 +514,8 @@ static int __init uwire_probe(struct platform_device *pdev) /* the spi->mode bits understood by this driver: */ master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH; + master->flags = SPI_MASTER_HALF_DUPLEX; + master->bus_num = 2; /* "official" */ master->num_chipselect = 4; master->setup = uwire_setup; diff --git a/include/linux/spi/spi.h b/include/linux/spi/spi.h index 743c933ac4e7..c47c4b4da97e 100644 --- a/include/linux/spi/spi.h +++ b/include/linux/spi/spi.h @@ -250,6 +250,10 @@ struct spi_master { /* spi_device.mode flags understood by this controller driver */ u16 mode_bits; + /* other constraints relevant to this driver */ + u16 flags; +#define SPI_MASTER_HALF_DUPLEX BIT(0) /* can't do full duplex */ + /* Setup mode and clock, etc (spi driver may call many times). * * IMPORTANT: this may be called when transfers to another -- cgit v1.2.3 From 537a1bf059fa312355696fa6db80726e655e7f17 Mon Sep 17 00:00:00 2001 From: Krzysztof Helt Date: Tue, 30 Jun 2009 11:41:29 -0700 Subject: fbdev: add mutex for fb_mmap locking Add a mutex to avoid a circular locking problem between the mm layer semaphore and fbdev ioctl mutex through the fb_mmap() call. Also, add mutex to all places where smem_start and smem_len fields change so the mutex inside the fb_mmap() is actually used. Changing of these fields before calling the framebuffer_register() are not mutexed. This is 2.6.31 material. It removes one lockdep (fb_mmap() and register_framebuffer()) but there is still another one (fb_release() and register_framebuffer()). It also cleans up handling of the smem_start and smem_len fields used by mutexed section of the fb_mmap(). Signed-off-by: Krzysztof Helt Cc: Peter Zijlstra Cc: "Rafael J. Wysocki" Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/video/atafb.c | 7 ++++++- drivers/video/atmel_lcdfb.c | 2 ++ drivers/video/fbmem.c | 13 +++++-------- drivers/video/fsl-diu-fb.c | 14 +++++++++----- drivers/video/i810/i810_main.c | 2 ++ drivers/video/matrox/matroxfb_base.c | 3 +++ drivers/video/matrox/matroxfb_crtc2.c | 5 ++++- drivers/video/mx3fb.c | 17 +++++++++++------ drivers/video/omap/omapfb_main.c | 4 ++++ drivers/video/platinumfb.c | 2 ++ drivers/video/pxafb.c | 2 ++ drivers/video/sh7760fb.c | 19 ++++++------------- drivers/video/sis/sis_main.c | 2 ++ drivers/video/sm501fb.c | 21 +++++++++++++-------- drivers/video/w100fb.c | 2 ++ include/linux/fb.h | 1 + 16 files changed, 74 insertions(+), 42 deletions(-) (limited to 'include') diff --git a/drivers/video/atafb.c b/drivers/video/atafb.c index 018850c116c6..497ff8af03ed 100644 --- a/drivers/video/atafb.c +++ b/drivers/video/atafb.c @@ -2414,7 +2414,10 @@ static int atafb_get_fix(struct fb_fix_screeninfo *fix, struct fb_info *info) if (err) return err; memset(fix, 0, sizeof(struct fb_fix_screeninfo)); - return fbhw->encode_fix(fix, &par); + mutex_lock(&info->mm_lock); + err = fbhw->encode_fix(fix, &par); + mutex_unlock(&info->mm_lock); + return err; } static int atafb_get_var(struct fb_var_screeninfo *var, struct fb_info *info) @@ -2743,7 +2746,9 @@ static int atafb_set_par(struct fb_info *info) /* Decode wanted screen parameters */ fbhw->decode_var(&info->var, par); + mutex_lock(&info->mm_lock); fbhw->encode_fix(&info->fix, par); + mutex_unlock(&info->mm_lock); /* Set new videomode */ ata_set_par(par); diff --git a/drivers/video/atmel_lcdfb.c b/drivers/video/atmel_lcdfb.c index 5afd64482f55..cb88394ba995 100644 --- a/drivers/video/atmel_lcdfb.c +++ b/drivers/video/atmel_lcdfb.c @@ -270,7 +270,9 @@ static int atmel_lcdfb_alloc_video_memory(struct atmel_lcdfb_info *sinfo) smem_len = (var->xres_virtual * var->yres_virtual * ((var->bits_per_pixel + 7) / 8)); + mutex_lock(&info->mm_lock); info->fix.smem_len = max(smem_len, sinfo->smem_len); + mutex_unlock(&info->mm_lock); info->screen_base = dma_alloc_writecombine(info->device, info->fix.smem_len, (dma_addr_t *)&info->fix.smem_start, GFP_KERNEL); diff --git a/drivers/video/fbmem.c b/drivers/video/fbmem.c index f8a09bf8d0cd..53ea05645ff8 100644 --- a/drivers/video/fbmem.c +++ b/drivers/video/fbmem.c @@ -1310,8 +1310,6 @@ static long fb_compat_ioctl(struct file *file, unsigned int cmd, static int fb_mmap(struct file *file, struct vm_area_struct * vma) -__acquires(&info->lock) -__releases(&info->lock) { int fbidx = iminor(file->f_path.dentry->d_inode); struct fb_info *info = registered_fb[fbidx]; @@ -1325,16 +1323,14 @@ __releases(&info->lock) off = vma->vm_pgoff << PAGE_SHIFT; if (!fb) return -ENODEV; + mutex_lock(&info->mm_lock); if (fb->fb_mmap) { int res; - mutex_lock(&info->lock); res = fb->fb_mmap(info, vma); - mutex_unlock(&info->lock); + mutex_unlock(&info->mm_lock); return res; } - mutex_lock(&info->lock); - /* frame buffer memory */ start = info->fix.smem_start; len = PAGE_ALIGN((start & ~PAGE_MASK) + info->fix.smem_len); @@ -1342,13 +1338,13 @@ __releases(&info->lock) /* memory mapped io */ off -= len; if (info->var.accel_flags) { - mutex_unlock(&info->lock); + mutex_unlock(&info->mm_lock); return -EINVAL; } start = info->fix.mmio_start; len = PAGE_ALIGN((start & ~PAGE_MASK) + info->fix.mmio_len); } - mutex_unlock(&info->lock); + mutex_unlock(&info->mm_lock); start &= PAGE_MASK; if ((vma->vm_end - vma->vm_start + off) > len) return -EINVAL; @@ -1518,6 +1514,7 @@ register_framebuffer(struct fb_info *fb_info) break; fb_info->node = i; mutex_init(&fb_info->lock); + mutex_init(&fb_info->mm_lock); fb_info->dev = device_create(fb_class, fb_info->device, MKDEV(FB_MAJOR, i), NULL, "fb%d", i); diff --git a/drivers/video/fsl-diu-fb.c b/drivers/video/fsl-diu-fb.c index f153c581cbd7..0bf2190928d0 100644 --- a/drivers/video/fsl-diu-fb.c +++ b/drivers/video/fsl-diu-fb.c @@ -750,24 +750,26 @@ static void update_lcdc(struct fb_info *info) static int map_video_memory(struct fb_info *info) { phys_addr_t phys; + u32 smem_len = info->fix.line_length * info->var.yres_virtual; pr_debug("info->var.xres_virtual = %d\n", info->var.xres_virtual); pr_debug("info->var.yres_virtual = %d\n", info->var.yres_virtual); pr_debug("info->fix.line_length = %d\n", info->fix.line_length); + pr_debug("MAP_VIDEO_MEMORY: smem_len = %u\n", smem_len); - info->fix.smem_len = info->fix.line_length * info->var.yres_virtual; - pr_debug("MAP_VIDEO_MEMORY: smem_len = %d\n", info->fix.smem_len); - info->screen_base = fsl_diu_alloc(info->fix.smem_len, &phys); + info->screen_base = fsl_diu_alloc(smem_len, &phys); if (info->screen_base == NULL) { printk(KERN_ERR "Unable to allocate fb memory\n"); return -ENOMEM; } + mutex_lock(&info->mm_lock); info->fix.smem_start = (unsigned long) phys; + info->fix.smem_len = smem_len; + mutex_unlock(&info->mm_lock); info->screen_size = info->fix.smem_len; pr_debug("Allocated fb @ paddr=0x%08lx, size=%d.\n", - info->fix.smem_start, - info->fix.smem_len); + info->fix.smem_start, info->fix.smem_len); pr_debug("screen base %p\n", info->screen_base); return 0; @@ -776,9 +778,11 @@ static int map_video_memory(struct fb_info *info) static void unmap_video_memory(struct fb_info *info) { fsl_diu_free(info->screen_base, info->fix.smem_len); + mutex_lock(&info->mm_lock); info->screen_base = NULL; info->fix.smem_start = 0; info->fix.smem_len = 0; + mutex_unlock(&info->mm_lock); } /* diff --git a/drivers/video/i810/i810_main.c b/drivers/video/i810/i810_main.c index 2e940199fc89..71960672d721 100644 --- a/drivers/video/i810/i810_main.c +++ b/drivers/video/i810/i810_main.c @@ -1090,8 +1090,10 @@ static int encode_fix(struct fb_fix_screeninfo *fix, struct fb_info *info) memset(fix, 0, sizeof(struct fb_fix_screeninfo)); strcpy(fix->id, "I810"); + mutex_lock(&info->mm_lock); fix->smem_start = par->fb.physical; fix->smem_len = par->fb.size; + mutex_unlock(&info->mm_lock); fix->type = FB_TYPE_PACKED_PIXELS; fix->type_aux = 0; fix->xpanstep = 8; diff --git a/drivers/video/matrox/matroxfb_base.c b/drivers/video/matrox/matroxfb_base.c index 8e7a275df50c..59c3a2e14913 100644 --- a/drivers/video/matrox/matroxfb_base.c +++ b/drivers/video/matrox/matroxfb_base.c @@ -724,8 +724,10 @@ static void matroxfb_update_fix(WPMINFO2) struct fb_fix_screeninfo *fix = &ACCESS_FBINFO(fbcon).fix; DBG(__func__) + mutex_lock(&ACCESS_FBINFO(fbcon).mm_lock); fix->smem_start = ACCESS_FBINFO(video.base) + ACCESS_FBINFO(curr.ydstorg.bytes); fix->smem_len = ACCESS_FBINFO(video.len_usable) - ACCESS_FBINFO(curr.ydstorg.bytes); + mutex_unlock(&ACCESS_FBINFO(fbcon).mm_lock); } static int matroxfb_check_var(struct fb_var_screeninfo *var, struct fb_info *info) @@ -2081,6 +2083,7 @@ static int matroxfb_probe(struct pci_dev* pdev, const struct pci_device_id* dumm spin_lock_init(&ACCESS_FBINFO(lock.accel)); init_rwsem(&ACCESS_FBINFO(crtc2.lock)); init_rwsem(&ACCESS_FBINFO(altout.lock)); + mutex_init(&ACCESS_FBINFO(fbcon).mm_lock); ACCESS_FBINFO(irq_flags) = 0; init_waitqueue_head(&ACCESS_FBINFO(crtc1.vsync.wait)); init_waitqueue_head(&ACCESS_FBINFO(crtc2.vsync.wait)); diff --git a/drivers/video/matrox/matroxfb_crtc2.c b/drivers/video/matrox/matroxfb_crtc2.c index 7ac4c5f6145d..909e10a11898 100644 --- a/drivers/video/matrox/matroxfb_crtc2.c +++ b/drivers/video/matrox/matroxfb_crtc2.c @@ -289,13 +289,16 @@ static int matroxfb_dh_release(struct fb_info* info, int user) { #undef m2info } -static void matroxfb_dh_init_fix(struct matroxfb_dh_fb_info *m2info) { +static void matroxfb_dh_init_fix(struct matroxfb_dh_fb_info *m2info) +{ struct fb_fix_screeninfo *fix = &m2info->fbcon.fix; strcpy(fix->id, "MATROX DH"); + mutex_lock(&m2info->fbcon.mm_lock); fix->smem_start = m2info->video.base; fix->smem_len = m2info->video.len_usable; + mutex_unlock(&m2info->fbcon.mm_lock); fix->ypanstep = 1; fix->ywrapstep = 0; fix->xpanstep = 8; /* TBD */ diff --git a/drivers/video/mx3fb.c b/drivers/video/mx3fb.c index b7af5256e887..567fb944bd2a 100644 --- a/drivers/video/mx3fb.c +++ b/drivers/video/mx3fb.c @@ -669,7 +669,7 @@ static uint32_t bpp_to_pixfmt(int bpp) } static int mx3fb_blank(int blank, struct fb_info *fbi); -static int mx3fb_map_video_memory(struct fb_info *fbi); +static int mx3fb_map_video_memory(struct fb_info *fbi, unsigned int mem_len); static int mx3fb_unmap_video_memory(struct fb_info *fbi); /** @@ -742,8 +742,7 @@ static int mx3fb_set_par(struct fb_info *fbi) if (fbi->fix.smem_start) mx3fb_unmap_video_memory(fbi); - fbi->fix.smem_len = mem_len; - if (mx3fb_map_video_memory(fbi) < 0) { + if (mx3fb_map_video_memory(fbi, mem_len) < 0) { mutex_unlock(&mx3_fbi->mutex); return -ENOMEM; } @@ -1198,6 +1197,7 @@ static int mx3fb_resume(struct platform_device *pdev) /** * mx3fb_map_video_memory() - allocates the DRAM memory for the frame buffer. * @fbi: framebuffer information pointer + * @mem_len: length of mapped memory * @return: Error code indicating success or failure * * This buffer is remapped into a non-cached, non-buffered, memory region to @@ -1205,23 +1205,26 @@ static int mx3fb_resume(struct platform_device *pdev) * area is remapped, all virtual memory access to the video memory should occur * at the new region. */ -static int mx3fb_map_video_memory(struct fb_info *fbi) +static int mx3fb_map_video_memory(struct fb_info *fbi, unsigned int mem_len) { int retval = 0; dma_addr_t addr; fbi->screen_base = dma_alloc_writecombine(fbi->device, - fbi->fix.smem_len, + mem_len, &addr, GFP_DMA); if (!fbi->screen_base) { dev_err(fbi->device, "Cannot allocate %u bytes framebuffer memory\n", - fbi->fix.smem_len); + mem_len); retval = -EBUSY; goto err0; } + mutex_lock(&fbi->mm_lock); fbi->fix.smem_start = addr; + fbi->fix.smem_len = mem_len; + mutex_unlock(&fbi->mm_lock); dev_dbg(fbi->device, "allocated fb @ p=0x%08x, v=0x%p, size=%d.\n", (uint32_t) fbi->fix.smem_start, fbi->screen_base, fbi->fix.smem_len); @@ -1251,8 +1254,10 @@ static int mx3fb_unmap_video_memory(struct fb_info *fbi) fbi->screen_base, fbi->fix.smem_start); fbi->screen_base = 0; + mutex_lock(&fbi->mm_lock); fbi->fix.smem_start = 0; fbi->fix.smem_len = 0; + mutex_unlock(&fbi->mm_lock); return 0; } diff --git a/drivers/video/omap/omapfb_main.c b/drivers/video/omap/omapfb_main.c index 060d72fe57cb..4ea99bfc37b4 100644 --- a/drivers/video/omap/omapfb_main.c +++ b/drivers/video/omap/omapfb_main.c @@ -393,8 +393,10 @@ static void set_fb_fix(struct fb_info *fbi) rg = &plane->fbdev->mem_desc.region[plane->idx]; fbi->screen_base = rg->vaddr; + mutex_lock(&fbi->mm_lock); fix->smem_start = rg->paddr; fix->smem_len = rg->size; + mutex_unlock(&fbi->mm_lock); fix->type = FB_TYPE_PACKED_PIXELS; bpp = var->bits_per_pixel; @@ -886,8 +888,10 @@ static int omapfb_setup_mem(struct fb_info *fbi, struct omapfb_mem_info *mi) * plane memory is dealloce'd, the other * screen parameters in var / fix are invalid. */ + mutex_lock(&fbi->mm_lock); fbi->fix.smem_start = 0; fbi->fix.smem_len = 0; + mutex_unlock(&fbi->mm_lock); } } } diff --git a/drivers/video/platinumfb.c b/drivers/video/platinumfb.c index 03b3670130a0..bacfabd9ce16 100644 --- a/drivers/video/platinumfb.c +++ b/drivers/video/platinumfb.c @@ -141,7 +141,9 @@ static int platinumfb_set_par (struct fb_info *info) offset = 0x10; info->screen_base = pinfo->frame_buffer + init->fb_offset + offset; + mutex_lock(&info->mm_lock); info->fix.smem_start = (pinfo->frame_buffer_phys) + init->fb_offset + offset; + mutex_unlock(&info->mm_lock); info->fix.visual = (pinfo->cmode == CMODE_8) ? FB_VISUAL_PSEUDOCOLOR : FB_VISUAL_DIRECTCOLOR; info->fix.line_length = vmode_attrs[pinfo->vmode-1].hres * (1<cmode) diff --git a/drivers/video/pxafb.c b/drivers/video/pxafb.c index 0889d50c3288..6506117c134b 100644 --- a/drivers/video/pxafb.c +++ b/drivers/video/pxafb.c @@ -815,8 +815,10 @@ static int overlayfb_map_video_memory(struct pxafb_layer *ofb) ofb->video_mem_phys = virt_to_phys(ofb->video_mem); ofb->video_mem_size = size; + mutex_lock(&ofb->fb.mm_lock); ofb->fb.fix.smem_start = ofb->video_mem_phys; ofb->fb.fix.smem_len = ofb->fb.fix.line_length * var->yres_virtual; + mutex_unlock(&ofb->fb.mm_lock); ofb->fb.screen_base = ofb->video_mem; return 0; } diff --git a/drivers/video/sh7760fb.c b/drivers/video/sh7760fb.c index 653bdfee3057..9f6d6e61f0cc 100644 --- a/drivers/video/sh7760fb.c +++ b/drivers/video/sh7760fb.c @@ -120,18 +120,6 @@ static int sh7760_setcolreg (u_int regno, return 0; } -static void encode_fix(struct fb_fix_screeninfo *fix, struct fb_info *info, - unsigned long stride) -{ - memset(fix, 0, sizeof(struct fb_fix_screeninfo)); - strcpy(fix->id, "sh7760-lcdc"); - - fix->smem_start = (unsigned long)info->screen_base; - fix->smem_len = info->screen_size; - - fix->line_length = stride; -} - static int sh7760fb_get_color_info(struct device *dev, u16 lddfr, int *bpp, int *gray) { @@ -334,7 +322,8 @@ static int sh7760fb_set_par(struct fb_info *info) iowrite32(ldsarl, par->base + LDSARL); /* mem for lower half of DSTN */ - encode_fix(&info->fix, info, stride); + info->fix.line_length = stride; + sh7760fb_check_var(&info->var, info); sh7760fb_blank(FB_BLANK_UNBLANK, info); /* panel on! */ @@ -435,6 +424,8 @@ static int sh7760fb_alloc_mem(struct fb_info *info) info->screen_base = fbmem; info->screen_size = vram; + info->fix.smem_start = (unsigned long)info->screen_base; + info->fix.smem_len = info->screen_size; return 0; } @@ -520,6 +511,8 @@ static int __devinit sh7760fb_probe(struct platform_device *pdev) info->var.transp.length = 0; info->var.transp.msb_right = 0; + strcpy(info->fix.id, "sh7760-lcdc"); + /* set the DON2 bit now, before cmap allocation, as it will randomize * palette memory. */ diff --git a/drivers/video/sis/sis_main.c b/drivers/video/sis/sis_main.c index 7072d19080d5..fd33455389b8 100644 --- a/drivers/video/sis/sis_main.c +++ b/drivers/video/sis/sis_main.c @@ -1847,8 +1847,10 @@ sisfb_get_fix(struct fb_fix_screeninfo *fix, int con, struct fb_info *info) strcpy(fix->id, ivideo->myid); + mutex_lock(&info->mm_lock); fix->smem_start = ivideo->video_base + ivideo->video_offset; fix->smem_len = ivideo->sisfb_mem; + mutex_unlock(&info->mm_lock); fix->type = FB_TYPE_PACKED_PIXELS; fix->type_aux = 0; fix->visual = (ivideo->video_bpp == 8) ? FB_VISUAL_PSEUDOCOLOR : FB_VISUAL_TRUECOLOR; diff --git a/drivers/video/sm501fb.c b/drivers/video/sm501fb.c index eb5d73a06702..98f24f0ec00d 100644 --- a/drivers/video/sm501fb.c +++ b/drivers/video/sm501fb.c @@ -145,7 +145,7 @@ static inline void sm501fb_sync_regs(struct sm501fb_info *info) #define SM501_MEMF_ACCEL (8) static int sm501_alloc_mem(struct sm501fb_info *inf, struct sm501_mem *mem, - unsigned int why, size_t size) + unsigned int why, size_t size, u32 smem_len) { struct sm501fb_par *par; struct fb_info *fbi; @@ -172,7 +172,7 @@ static int sm501_alloc_mem(struct sm501fb_info *inf, struct sm501_mem *mem, if (ptr > 0) ptr &= ~(PAGE_SIZE - 1); - if (fbi && ptr < fbi->fix.smem_len) + if (fbi && ptr < smem_len) return -ENOMEM; break; @@ -197,7 +197,7 @@ static int sm501_alloc_mem(struct sm501fb_info *inf, struct sm501_mem *mem, case SM501_MEMF_ACCEL: fbi = inf->fb[HEAD_CRT]; - ptr = fbi ? fbi->fix.smem_len : 0; + ptr = fbi ? smem_len : 0; fbi = inf->fb[HEAD_PANEL]; if (fbi) { @@ -413,6 +413,7 @@ static int sm501fb_set_par_common(struct fb_info *info, unsigned int mem_type; unsigned int clock_type; unsigned int head_addr; + unsigned int smem_len; dev_dbg(fbi->dev, "%s: %dx%d, bpp = %d, virtual %dx%d\n", __func__, var->xres, var->yres, var->bits_per_pixel, @@ -453,18 +454,20 @@ static int sm501fb_set_par_common(struct fb_info *info, /* allocate fb memory within 501 */ info->fix.line_length = (var->xres_virtual * var->bits_per_pixel)/8; - info->fix.smem_len = info->fix.line_length * var->yres_virtual; + smem_len = info->fix.line_length * var->yres_virtual; dev_dbg(fbi->dev, "%s: line length = %u\n", __func__, info->fix.line_length); - if (sm501_alloc_mem(fbi, &par->screen, mem_type, - info->fix.smem_len)) { + if (sm501_alloc_mem(fbi, &par->screen, mem_type, smem_len, smem_len)) { dev_err(fbi->dev, "no memory available\n"); return -ENOMEM; } + mutex_lock(&info->mm_lock); info->fix.smem_start = fbi->fbmem_res->start + par->screen.sm_addr; + info->fix.smem_len = smem_len; + mutex_unlock(&info->mm_lock); info->screen_base = fbi->fbmem + par->screen.sm_addr; info->screen_size = info->fix.smem_len; @@ -637,7 +640,8 @@ static int sm501fb_set_par_crt(struct fb_info *info) if ((control & SM501_DC_CRT_CONTROL_SEL) == 0) { /* the head is displaying panel data... */ - sm501_alloc_mem(fbi, &par->screen, SM501_MEMF_CRT, 0); + sm501_alloc_mem(fbi, &par->screen, SM501_MEMF_CRT, 0, + info->fix.smem_len); goto out_update; } @@ -1289,7 +1293,8 @@ static int sm501_init_cursor(struct fb_info *fbi, unsigned int reg_base) par->cursor_regs = info->regs + reg_base; - ret = sm501_alloc_mem(info, &par->cursor, SM501_MEMF_CURSOR, 1024); + ret = sm501_alloc_mem(info, &par->cursor, SM501_MEMF_CURSOR, 1024, + fbi->fix.smem_len); if (ret < 0) return ret; diff --git a/drivers/video/w100fb.c b/drivers/video/w100fb.c index d0674f1e3f10..8a141c2c637b 100644 --- a/drivers/video/w100fb.c +++ b/drivers/video/w100fb.c @@ -523,6 +523,7 @@ static int w100fb_set_par(struct fb_info *info) info->fix.ywrapstep = 0; info->fix.line_length = par->xres * BITS_PER_PIXEL / 8; + mutex_lock(&info->mm_lock); if ((par->xres*par->yres*BITS_PER_PIXEL/8) > (MEM_INT_SIZE+1)) { par->extmem_active = 1; info->fix.smem_len = par->mach->mem->size+1; @@ -530,6 +531,7 @@ static int w100fb_set_par(struct fb_info *info) par->extmem_active = 0; info->fix.smem_len = MEM_INT_SIZE+1; } + mutex_unlock(&info->mm_lock); w100fb_activate_var(par); } diff --git a/include/linux/fb.h b/include/linux/fb.h index dd68358996b7..f847df9e99b6 100644 --- a/include/linux/fb.h +++ b/include/linux/fb.h @@ -819,6 +819,7 @@ struct fb_info { int node; int flags; struct mutex lock; /* Lock for open/release/ioctl funcs */ + struct mutex mm_lock; /* Lock for fb_mmap and smem_* fields */ struct fb_var_screeninfo var; /* Current var */ struct fb_fix_screeninfo fix; /* Current fix */ struct fb_monspecs monspecs; /* Current Monitor specs */ -- cgit v1.2.3 From d9d62f3f2c6fa609883714f6fd6cd710a83d307f Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Mon, 29 Jun 2009 16:54:12 +0000 Subject: usbnet: Remove private stats structure Now that nothing uses the private stats structure we can remove it. Signed-off-by: Herbert Xu Signed-off-by: David S. Miller --- include/linux/usb/usbnet.h | 1 - 1 file changed, 1 deletion(-) (limited to 'include') diff --git a/include/linux/usb/usbnet.h b/include/linux/usb/usbnet.h index 5d44059f6d63..310e18a880ff 100644 --- a/include/linux/usb/usbnet.h +++ b/include/linux/usb/usbnet.h @@ -42,7 +42,6 @@ struct usbnet { /* protocol/interface state */ struct net_device *net; - struct net_device_stats stats; int msg_enable; unsigned long data [5]; u32 xid; -- cgit v1.2.3 From 7878cba9f0037f5599004b03a1260b32d9050360 Mon Sep 17 00:00:00 2001 From: "Martin K. Petersen" Date: Fri, 26 Jun 2009 15:37:49 +0200 Subject: block: Create bip slabs with embedded integrity vectors This patch restores stacking ability to the block layer integrity infrastructure by creating a set of dedicated bip slabs. Each bip slab has an embedded bio_vec array at the end. This cuts down on memory allocations and also simplifies the code compared to the original bvec version. Only the largest bip slab is backed by a mempool. The pool is contained in the bio_set so stacking drivers can ensure forward progress. Signed-off-by: Martin K. Petersen Signed-off-by: Jens Axboe --- block/blk-core.c | 2 +- drivers/md/dm.c | 4 +- fs/bio-integrity.c | 170 ++++++++++++++++++++++++++++++++++++++-------------- fs/bio.c | 11 +++- include/linux/bio.h | 22 +++++-- 5 files changed, 152 insertions(+), 57 deletions(-) (limited to 'include') diff --git a/block/blk-core.c b/block/blk-core.c index b06cf5c2a829..345d99da8d41 100644 --- a/block/blk-core.c +++ b/block/blk-core.c @@ -2365,7 +2365,7 @@ int blk_rq_prep_clone(struct request *rq, struct request *rq_src, __bio_clone(bio, bio_src); if (bio_integrity(bio_src) && - bio_integrity_clone(bio, bio_src, gfp_mask)) + bio_integrity_clone(bio, bio_src, gfp_mask, bs)) goto free_and_out; if (bio_ctr && bio_ctr(bio, bio_src, data)) diff --git a/drivers/md/dm.c b/drivers/md/dm.c index 3c6d4ee8921d..9acd54a5cffb 100644 --- a/drivers/md/dm.c +++ b/drivers/md/dm.c @@ -1017,7 +1017,7 @@ static struct bio *split_bvec(struct bio *bio, sector_t sector, clone->bi_flags |= 1 << BIO_CLONED; if (bio_integrity(bio)) { - bio_integrity_clone(clone, bio, GFP_NOIO); + bio_integrity_clone(clone, bio, GFP_NOIO, bs); bio_integrity_trim(clone, bio_sector_offset(bio, idx, offset), len); } @@ -1045,7 +1045,7 @@ static struct bio *clone_bio(struct bio *bio, sector_t sector, clone->bi_flags &= ~(1 << BIO_SEG_VALID); if (bio_integrity(bio)) { - bio_integrity_clone(clone, bio, GFP_NOIO); + bio_integrity_clone(clone, bio, GFP_NOIO, bs); if (idx != bio->bi_idx || clone->bi_size < bio->bi_size) bio_integrity_trim(clone, diff --git a/fs/bio-integrity.c b/fs/bio-integrity.c index 31c46a241bac..49a34e7f7306 100644 --- a/fs/bio-integrity.c +++ b/fs/bio-integrity.c @@ -1,7 +1,7 @@ /* * bio-integrity.c - bio data integrity extensions * - * Copyright (C) 2007, 2008 Oracle Corporation + * Copyright (C) 2007, 2008, 2009 Oracle Corporation * Written by: Martin K. Petersen * * This program is free software; you can redistribute it and/or @@ -25,63 +25,121 @@ #include #include -static struct kmem_cache *bio_integrity_slab __read_mostly; -static mempool_t *bio_integrity_pool; -static struct bio_set *integrity_bio_set; +struct integrity_slab { + struct kmem_cache *slab; + unsigned short nr_vecs; + char name[8]; +}; + +#define IS(x) { .nr_vecs = x, .name = "bip-"__stringify(x) } +struct integrity_slab bip_slab[BIOVEC_NR_POOLS] __read_mostly = { + IS(1), IS(4), IS(16), IS(64), IS(128), IS(BIO_MAX_PAGES), +}; +#undef IS + static struct workqueue_struct *kintegrityd_wq; +static inline unsigned int vecs_to_idx(unsigned int nr) +{ + switch (nr) { + case 1: + return 0; + case 2 ... 4: + return 1; + case 5 ... 16: + return 2; + case 17 ... 64: + return 3; + case 65 ... 128: + return 4; + case 129 ... BIO_MAX_PAGES: + return 5; + default: + BUG(); + } +} + +static inline int use_bip_pool(unsigned int idx) +{ + if (idx == BIOVEC_NR_POOLS) + return 1; + + return 0; +} + /** - * bio_integrity_alloc - Allocate integrity payload and attach it to bio + * bio_integrity_alloc_bioset - Allocate integrity payload and attach it to bio * @bio: bio to attach integrity metadata to * @gfp_mask: Memory allocation mask * @nr_vecs: Number of integrity metadata scatter-gather elements + * @bs: bio_set to allocate from * * Description: This function prepares a bio for attaching integrity * metadata. nr_vecs specifies the maximum number of pages containing * integrity metadata that can be attached. */ -struct bio_integrity_payload *bio_integrity_alloc(struct bio *bio, - gfp_t gfp_mask, - unsigned int nr_vecs) +struct bio_integrity_payload *bio_integrity_alloc_bioset(struct bio *bio, + gfp_t gfp_mask, + unsigned int nr_vecs, + struct bio_set *bs) { struct bio_integrity_payload *bip; - struct bio_vec *iv; - unsigned long idx; + unsigned int idx = vecs_to_idx(nr_vecs); BUG_ON(bio == NULL); + bip = NULL; - bip = mempool_alloc(bio_integrity_pool, gfp_mask); - if (unlikely(bip == NULL)) { - printk(KERN_ERR "%s: could not alloc bip\n", __func__); - return NULL; - } + /* Lower order allocations come straight from slab */ + if (!use_bip_pool(idx)) + bip = kmem_cache_alloc(bip_slab[idx].slab, gfp_mask); - memset(bip, 0, sizeof(*bip)); + /* Use mempool if lower order alloc failed or max vecs were requested */ + if (bip == NULL) { + bip = mempool_alloc(bs->bio_integrity_pool, gfp_mask); - iv = bvec_alloc_bs(gfp_mask, nr_vecs, &idx, integrity_bio_set); - if (unlikely(iv == NULL)) { - printk(KERN_ERR "%s: could not alloc bip_vec\n", __func__); - mempool_free(bip, bio_integrity_pool); - return NULL; + if (unlikely(bip == NULL)) { + printk(KERN_ERR "%s: could not alloc bip\n", __func__); + return NULL; + } } - bip->bip_pool = idx; - bip->bip_vec = iv; + memset(bip, 0, sizeof(*bip)); + + bip->bip_slab = idx; bip->bip_bio = bio; bio->bi_integrity = bip; return bip; } +EXPORT_SYMBOL(bio_integrity_alloc_bioset); + +/** + * bio_integrity_alloc - Allocate integrity payload and attach it to bio + * @bio: bio to attach integrity metadata to + * @gfp_mask: Memory allocation mask + * @nr_vecs: Number of integrity metadata scatter-gather elements + * + * Description: This function prepares a bio for attaching integrity + * metadata. nr_vecs specifies the maximum number of pages containing + * integrity metadata that can be attached. + */ +struct bio_integrity_payload *bio_integrity_alloc(struct bio *bio, + gfp_t gfp_mask, + unsigned int nr_vecs) +{ + return bio_integrity_alloc_bioset(bio, gfp_mask, nr_vecs, fs_bio_set); +} EXPORT_SYMBOL(bio_integrity_alloc); /** * bio_integrity_free - Free bio integrity payload * @bio: bio containing bip to be freed + * @bs: bio_set this bio was allocated from * * Description: Used to free the integrity portion of a bio. Usually * called from bio_free(). */ -void bio_integrity_free(struct bio *bio) +void bio_integrity_free(struct bio *bio, struct bio_set *bs) { struct bio_integrity_payload *bip = bio->bi_integrity; @@ -92,8 +150,10 @@ void bio_integrity_free(struct bio *bio) && bip->bip_buf != NULL) kfree(bip->bip_buf); - bvec_free_bs(integrity_bio_set, bip->bip_vec, bip->bip_pool); - mempool_free(bip, bio_integrity_pool); + if (use_bip_pool(bip->bip_slab)) + mempool_free(bip, bs->bio_integrity_pool); + else + kmem_cache_free(bip_slab[bip->bip_slab].slab, bip); bio->bi_integrity = NULL; } @@ -114,7 +174,7 @@ int bio_integrity_add_page(struct bio *bio, struct page *page, struct bio_integrity_payload *bip = bio->bi_integrity; struct bio_vec *iv; - if (bip->bip_vcnt >= bvec_nr_vecs(bip->bip_pool)) { + if (bip->bip_vcnt >= bvec_nr_vecs(bip->bip_slab)) { printk(KERN_ERR "%s: bip_vec full\n", __func__); return 0; } @@ -647,8 +707,8 @@ void bio_integrity_split(struct bio *bio, struct bio_pair *bp, int sectors) bp->iv1 = bip->bip_vec[0]; bp->iv2 = bip->bip_vec[0]; - bp->bip1.bip_vec = &bp->iv1; - bp->bip2.bip_vec = &bp->iv2; + bp->bip1.bip_vec[0] = bp->iv1; + bp->bip2.bip_vec[0] = bp->iv2; bp->iv1.bv_len = sectors * bi->tuple_size; bp->iv2.bv_offset += sectors * bi->tuple_size; @@ -667,17 +727,19 @@ EXPORT_SYMBOL(bio_integrity_split); * @bio: New bio * @bio_src: Original bio * @gfp_mask: Memory allocation mask + * @bs: bio_set to allocate bip from * * Description: Called to allocate a bip when cloning a bio */ -int bio_integrity_clone(struct bio *bio, struct bio *bio_src, gfp_t gfp_mask) +int bio_integrity_clone(struct bio *bio, struct bio *bio_src, + gfp_t gfp_mask, struct bio_set *bs) { struct bio_integrity_payload *bip_src = bio_src->bi_integrity; struct bio_integrity_payload *bip; BUG_ON(bip_src == NULL); - bip = bio_integrity_alloc(bio, gfp_mask, bip_src->bip_vcnt); + bip = bio_integrity_alloc_bioset(bio, gfp_mask, bip_src->bip_vcnt, bs); if (bip == NULL) return -EIO; @@ -693,25 +755,43 @@ int bio_integrity_clone(struct bio *bio, struct bio *bio_src, gfp_t gfp_mask) } EXPORT_SYMBOL(bio_integrity_clone); -static int __init bio_integrity_init(void) +int bioset_integrity_create(struct bio_set *bs, int pool_size) { - kintegrityd_wq = create_workqueue("kintegrityd"); + unsigned int max_slab = vecs_to_idx(BIO_MAX_PAGES); + + bs->bio_integrity_pool = + mempool_create_slab_pool(pool_size, bip_slab[max_slab].slab); + if (!bs->bio_integrity_pool) + return -1; + + return 0; +} +EXPORT_SYMBOL(bioset_integrity_create); + +void bioset_integrity_free(struct bio_set *bs) +{ + if (bs->bio_integrity_pool) + mempool_destroy(bs->bio_integrity_pool); +} +EXPORT_SYMBOL(bioset_integrity_free); + +void __init bio_integrity_init(void) +{ + unsigned int i; + + kintegrityd_wq = create_workqueue("kintegrityd"); if (!kintegrityd_wq) panic("Failed to create kintegrityd\n"); - bio_integrity_slab = KMEM_CACHE(bio_integrity_payload, - SLAB_HWCACHE_ALIGN|SLAB_PANIC); + for (i = 0 ; i < BIOVEC_NR_POOLS ; i++) { + unsigned int size; - bio_integrity_pool = mempool_create_slab_pool(BIO_POOL_SIZE, - bio_integrity_slab); - if (!bio_integrity_pool) - panic("bio_integrity: can't allocate bip pool\n"); + size = sizeof(struct bio_integrity_payload) + + bip_slab[i].nr_vecs * sizeof(struct bio_vec); - integrity_bio_set = bioset_create(BIO_POOL_SIZE, 0); - if (!integrity_bio_set) - panic("bio_integrity: can't allocate bio_set\n"); - - return 0; + bip_slab[i].slab = + kmem_cache_create(bip_slab[i].name, size, 0, + SLAB_HWCACHE_ALIGN|SLAB_PANIC, NULL); + } } -subsys_initcall(bio_integrity_init); diff --git a/fs/bio.c b/fs/bio.c index 24c914043532..1486b19fc431 100644 --- a/fs/bio.c +++ b/fs/bio.c @@ -238,7 +238,7 @@ void bio_free(struct bio *bio, struct bio_set *bs) bvec_free_bs(bs, bio->bi_io_vec, BIO_POOL_IDX(bio)); if (bio_integrity(bio)) - bio_integrity_free(bio); + bio_integrity_free(bio, bs); /* * If we have front padding, adjust the bio pointer before freeing @@ -341,7 +341,7 @@ struct bio *bio_alloc(gfp_t gfp_mask, int nr_iovecs) static void bio_kmalloc_destructor(struct bio *bio) { if (bio_integrity(bio)) - bio_integrity_free(bio); + bio_integrity_free(bio, fs_bio_set); kfree(bio); } @@ -472,7 +472,7 @@ struct bio *bio_clone(struct bio *bio, gfp_t gfp_mask) if (bio_integrity(bio)) { int ret; - ret = bio_integrity_clone(b, bio, gfp_mask); + ret = bio_integrity_clone(b, bio, gfp_mask, fs_bio_set); if (ret < 0) { bio_put(b); @@ -1539,6 +1539,7 @@ void bioset_free(struct bio_set *bs) if (bs->bio_pool) mempool_destroy(bs->bio_pool); + bioset_integrity_free(bs); biovec_free_pools(bs); bio_put_slab(bs); @@ -1579,6 +1580,9 @@ struct bio_set *bioset_create(unsigned int pool_size, unsigned int front_pad) if (!bs->bio_pool) goto bad; + if (bioset_integrity_create(bs, pool_size)) + goto bad; + if (!biovec_create_pools(bs, pool_size)) return bs; @@ -1616,6 +1620,7 @@ static int __init init_bio(void) if (!bio_slabs) panic("bio: can't allocate bios\n"); + bio_integrity_init(); biovec_init_slabs(); fs_bio_set = bioset_create(BIO_POOL_SIZE, 0); diff --git a/include/linux/bio.h b/include/linux/bio.h index 2a04eb54c0dd..2892b710771c 100644 --- a/include/linux/bio.h +++ b/include/linux/bio.h @@ -319,7 +319,6 @@ static inline int bio_has_allocated_vec(struct bio *bio) */ struct bio_integrity_payload { struct bio *bip_bio; /* parent bio */ - struct bio_vec *bip_vec; /* integrity data vector */ sector_t bip_sector; /* virtual start sector */ @@ -328,11 +327,12 @@ struct bio_integrity_payload { unsigned int bip_size; - unsigned short bip_pool; /* pool the ivec came from */ + unsigned short bip_slab; /* slab the bip came from */ unsigned short bip_vcnt; /* # of integrity bio_vecs */ unsigned short bip_idx; /* current bip_vec index */ struct work_struct bip_work; /* I/O completion */ + struct bio_vec bip_vec[0]; /* embedded bvec array */ }; #endif /* CONFIG_BLK_DEV_INTEGRITY */ @@ -430,6 +430,9 @@ struct bio_set { unsigned int front_pad; mempool_t *bio_pool; +#if defined(CONFIG_BLK_DEV_INTEGRITY) + mempool_t *bio_integrity_pool; +#endif mempool_t *bvec_pool; }; @@ -634,8 +637,9 @@ static inline struct bio *bio_list_get(struct bio_list *bl) #define bio_integrity(bio) (bio->bi_integrity != NULL) +extern struct bio_integrity_payload *bio_integrity_alloc_bioset(struct bio *, gfp_t, unsigned int, struct bio_set *); extern struct bio_integrity_payload *bio_integrity_alloc(struct bio *, gfp_t, unsigned int); -extern void bio_integrity_free(struct bio *); +extern void bio_integrity_free(struct bio *, struct bio_set *); extern int bio_integrity_add_page(struct bio *, struct page *, unsigned int, unsigned int); extern int bio_integrity_enabled(struct bio *bio); extern int bio_integrity_set_tag(struct bio *, void *, unsigned int); @@ -645,21 +649,27 @@ extern void bio_integrity_endio(struct bio *, int); extern void bio_integrity_advance(struct bio *, unsigned int); extern void bio_integrity_trim(struct bio *, unsigned int, unsigned int); extern void bio_integrity_split(struct bio *, struct bio_pair *, int); -extern int bio_integrity_clone(struct bio *, struct bio *, gfp_t); +extern int bio_integrity_clone(struct bio *, struct bio *, gfp_t, struct bio_set *); +extern int bioset_integrity_create(struct bio_set *, int); +extern void bioset_integrity_free(struct bio_set *); +extern void bio_integrity_init(void); #else /* CONFIG_BLK_DEV_INTEGRITY */ #define bio_integrity(a) (0) +#define bioset_integrity_create(a, b) (0) #define bio_integrity_prep(a) (0) #define bio_integrity_enabled(a) (0) -#define bio_integrity_clone(a, b, c) (0) -#define bio_integrity_free(a) do { } while (0) +#define bio_integrity_clone(a, b, c, d) (0) +#define bioset_integrity_free(a) do { } while (0) +#define bio_integrity_free(a, b) do { } while (0) #define bio_integrity_endio(a, b) do { } while (0) #define bio_integrity_advance(a, b) do { } while (0) #define bio_integrity_trim(a, b, c) do { } while (0) #define bio_integrity_split(a, b, c) do { } while (0) #define bio_integrity_set_tag(a, b, c) do { } while (0) #define bio_integrity_get_tag(a, b, c) do { } while (0) +#define bio_integrity_init(a) do { } while (0) #endif /* CONFIG_BLK_DEV_INTEGRITY */ -- cgit v1.2.3 From 018e0446890661504783f92388ecce7138c1566d Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Fri, 26 Jun 2009 16:27:10 +0200 Subject: block: get rid of queue-private command filter The initial patches to support this through sysfs export were broken and have been if 0'ed out in any release. So lets just kill the code and reclaim some space in struct request_queue, if anyone would later like to fixup the sysfs bits, the git history can easily restore the removed bits. Signed-off-by: Jens Axboe --- block/Makefile | 2 +- block/blk-core.c | 2 - block/bsg.c | 2 +- block/cmd-filter.c | 233 ------------------------------------------------- block/scsi_ioctl.c | 43 +++++++-- drivers/scsi/sg.c | 4 +- include/linux/blkdev.h | 15 +--- 7 files changed, 42 insertions(+), 259 deletions(-) delete mode 100644 block/cmd-filter.c (limited to 'include') diff --git a/block/Makefile b/block/Makefile index e9fa4dd690f2..6c54ed0ff755 100644 --- a/block/Makefile +++ b/block/Makefile @@ -5,7 +5,7 @@ obj-$(CONFIG_BLOCK) := elevator.o blk-core.o blk-tag.o blk-sysfs.o \ blk-barrier.o blk-settings.o blk-ioc.o blk-map.o \ blk-exec.o blk-merge.o blk-softirq.o blk-timeout.o \ - ioctl.o genhd.o scsi_ioctl.o cmd-filter.o + ioctl.o genhd.o scsi_ioctl.o obj-$(CONFIG_BLK_DEV_BSG) += bsg.o obj-$(CONFIG_IOSCHED_NOOP) += noop-iosched.o diff --git a/block/blk-core.c b/block/blk-core.c index 345d99da8d41..02b87134a167 100644 --- a/block/blk-core.c +++ b/block/blk-core.c @@ -595,8 +595,6 @@ blk_init_queue_node(request_fn_proc *rfn, spinlock_t *lock, int node_id) q->sg_reserved_size = INT_MAX; - blk_set_cmd_filter_defaults(&q->cmd_filter); - /* * all done */ diff --git a/block/bsg.c b/block/bsg.c index e7d475254248..5f184bb3ff9e 100644 --- a/block/bsg.c +++ b/block/bsg.c @@ -186,7 +186,7 @@ static int blk_fill_sgv4_hdr_rq(struct request_queue *q, struct request *rq, return -EFAULT; if (hdr->subprotocol == BSG_SUB_PROTOCOL_SCSI_CMD) { - if (blk_verify_command(&q->cmd_filter, rq->cmd, has_write_perm)) + if (blk_verify_command(rq->cmd, has_write_perm)) return -EPERM; } else if (!capable(CAP_SYS_RAWIO)) return -EPERM; diff --git a/block/cmd-filter.c b/block/cmd-filter.c deleted file mode 100644 index 572bbc2f900d..000000000000 --- a/block/cmd-filter.c +++ /dev/null @@ -1,233 +0,0 @@ -/* - * Copyright 2004 Peter M. Jones - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License version 2 as - * published by the Free Software Foundation. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public Licens - * along with this program; if not, write to the Free Software - * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111- - * - */ - -#include -#include -#include -#include -#include -#include - -#include -#include - -int blk_verify_command(struct blk_cmd_filter *filter, - unsigned char *cmd, fmode_t has_write_perm) -{ - /* root can do any command. */ - if (capable(CAP_SYS_RAWIO)) - return 0; - - /* if there's no filter set, assume we're filtering everything out */ - if (!filter) - return -EPERM; - - /* Anybody who can open the device can do a read-safe command */ - if (test_bit(cmd[0], filter->read_ok)) - return 0; - - /* Write-safe commands require a writable open */ - if (test_bit(cmd[0], filter->write_ok) && has_write_perm) - return 0; - - return -EPERM; -} -EXPORT_SYMBOL(blk_verify_command); - -#if 0 -/* and now, the sysfs stuff */ -static ssize_t rcf_cmds_show(struct blk_cmd_filter *filter, char *page, - int rw) -{ - char *npage = page; - unsigned long *okbits; - int i; - - if (rw == READ) - okbits = filter->read_ok; - else - okbits = filter->write_ok; - - for (i = 0; i < BLK_SCSI_MAX_CMDS; i++) { - if (test_bit(i, okbits)) { - npage += sprintf(npage, "0x%02x", i); - if (i < BLK_SCSI_MAX_CMDS - 1) - sprintf(npage++, " "); - } - } - - if (npage != page) - npage += sprintf(npage, "\n"); - - return npage - page; -} - -static ssize_t rcf_readcmds_show(struct blk_cmd_filter *filter, char *page) -{ - return rcf_cmds_show(filter, page, READ); -} - -static ssize_t rcf_writecmds_show(struct blk_cmd_filter *filter, - char *page) -{ - return rcf_cmds_show(filter, page, WRITE); -} - -static ssize_t rcf_cmds_store(struct blk_cmd_filter *filter, - const char *page, size_t count, int rw) -{ - unsigned long okbits[BLK_SCSI_CMD_PER_LONG], *target_okbits; - int cmd, set; - char *p, *status; - - if (rw == READ) { - memcpy(&okbits, filter->read_ok, sizeof(okbits)); - target_okbits = filter->read_ok; - } else { - memcpy(&okbits, filter->write_ok, sizeof(okbits)); - target_okbits = filter->write_ok; - } - - while ((p = strsep((char **)&page, " ")) != NULL) { - set = 1; - - if (p[0] == '+') { - p++; - } else if (p[0] == '-') { - set = 0; - p++; - } - - cmd = simple_strtol(p, &status, 16); - - /* either of these cases means invalid input, so do nothing. */ - if ((status == p) || cmd >= BLK_SCSI_MAX_CMDS) - return -EINVAL; - - if (set) - __set_bit(cmd, okbits); - else - __clear_bit(cmd, okbits); - } - - memcpy(target_okbits, okbits, sizeof(okbits)); - return count; -} - -static ssize_t rcf_readcmds_store(struct blk_cmd_filter *filter, - const char *page, size_t count) -{ - return rcf_cmds_store(filter, page, count, READ); -} - -static ssize_t rcf_writecmds_store(struct blk_cmd_filter *filter, - const char *page, size_t count) -{ - return rcf_cmds_store(filter, page, count, WRITE); -} - -struct rcf_sysfs_entry { - struct attribute attr; - ssize_t (*show)(struct blk_cmd_filter *, char *); - ssize_t (*store)(struct blk_cmd_filter *, const char *, size_t); -}; - -static struct rcf_sysfs_entry rcf_readcmds_entry = { - .attr = { .name = "read_table", .mode = S_IRUGO | S_IWUSR }, - .show = rcf_readcmds_show, - .store = rcf_readcmds_store, -}; - -static struct rcf_sysfs_entry rcf_writecmds_entry = { - .attr = {.name = "write_table", .mode = S_IRUGO | S_IWUSR }, - .show = rcf_writecmds_show, - .store = rcf_writecmds_store, -}; - -static struct attribute *default_attrs[] = { - &rcf_readcmds_entry.attr, - &rcf_writecmds_entry.attr, - NULL, -}; - -#define to_rcf(atr) container_of((atr), struct rcf_sysfs_entry, attr) - -static ssize_t -rcf_attr_show(struct kobject *kobj, struct attribute *attr, char *page) -{ - struct rcf_sysfs_entry *entry = to_rcf(attr); - struct blk_cmd_filter *filter; - - filter = container_of(kobj, struct blk_cmd_filter, kobj); - if (entry->show) - return entry->show(filter, page); - - return 0; -} - -static ssize_t -rcf_attr_store(struct kobject *kobj, struct attribute *attr, - const char *page, size_t length) -{ - struct rcf_sysfs_entry *entry = to_rcf(attr); - struct blk_cmd_filter *filter; - - if (!capable(CAP_SYS_RAWIO)) - return -EPERM; - - if (!entry->store) - return -EINVAL; - - filter = container_of(kobj, struct blk_cmd_filter, kobj); - return entry->store(filter, page, length); -} - -static struct sysfs_ops rcf_sysfs_ops = { - .show = rcf_attr_show, - .store = rcf_attr_store, -}; - -static struct kobj_type rcf_ktype = { - .sysfs_ops = &rcf_sysfs_ops, - .default_attrs = default_attrs, -}; - -int blk_register_filter(struct gendisk *disk) -{ - int ret; - struct blk_cmd_filter *filter = &disk->queue->cmd_filter; - - ret = kobject_init_and_add(&filter->kobj, &rcf_ktype, - &disk_to_dev(disk)->kobj, - "%s", "cmd_filter"); - if (ret < 0) - return ret; - - return 0; -} -EXPORT_SYMBOL(blk_register_filter); - -void blk_unregister_filter(struct gendisk *disk) -{ - struct blk_cmd_filter *filter = &disk->queue->cmd_filter; - - kobject_put(&filter->kobj); -} -EXPORT_SYMBOL(blk_unregister_filter); -#endif diff --git a/block/scsi_ioctl.c b/block/scsi_ioctl.c index 5f8e798ede4e..f0e0ce0a607d 100644 --- a/block/scsi_ioctl.c +++ b/block/scsi_ioctl.c @@ -32,6 +32,11 @@ #include #include +struct blk_cmd_filter { + unsigned long read_ok[BLK_SCSI_CMD_PER_LONG]; + unsigned long write_ok[BLK_SCSI_CMD_PER_LONG]; +} blk_default_cmd_filter; + /* Command group 3 is reserved and should never be used. */ const unsigned char scsi_command_size_tbl[8] = { @@ -105,7 +110,7 @@ static int sg_emulated_host(struct request_queue *q, int __user *p) return put_user(1, p); } -void blk_set_cmd_filter_defaults(struct blk_cmd_filter *filter) +static void blk_set_cmd_filter_defaults(struct blk_cmd_filter *filter) { /* Basic read-only commands */ __set_bit(TEST_UNIT_READY, filter->read_ok); @@ -187,14 +192,37 @@ void blk_set_cmd_filter_defaults(struct blk_cmd_filter *filter) __set_bit(GPCMD_SET_STREAMING, filter->write_ok); __set_bit(GPCMD_SET_READ_AHEAD, filter->write_ok); } -EXPORT_SYMBOL_GPL(blk_set_cmd_filter_defaults); + +int blk_verify_command(unsigned char *cmd, fmode_t has_write_perm) +{ + struct blk_cmd_filter *filter = &blk_default_cmd_filter; + + /* root can do any command. */ + if (capable(CAP_SYS_RAWIO)) + return 0; + + /* if there's no filter set, assume we're filtering everything out */ + if (!filter) + return -EPERM; + + /* Anybody who can open the device can do a read-safe command */ + if (test_bit(cmd[0], filter->read_ok)) + return 0; + + /* Write-safe commands require a writable open */ + if (test_bit(cmd[0], filter->write_ok) && has_write_perm) + return 0; + + return -EPERM; +} +EXPORT_SYMBOL(blk_verify_command); static int blk_fill_sghdr_rq(struct request_queue *q, struct request *rq, struct sg_io_hdr *hdr, fmode_t mode) { if (copy_from_user(rq->cmd, hdr->cmdp, hdr->cmd_len)) return -EFAULT; - if (blk_verify_command(&q->cmd_filter, rq->cmd, mode & FMODE_WRITE)) + if (blk_verify_command(rq->cmd, mode & FMODE_WRITE)) return -EPERM; /* @@ -427,7 +455,7 @@ int sg_scsi_ioctl(struct request_queue *q, struct gendisk *disk, fmode_t mode, if (in_len && copy_from_user(buffer, sic->data + cmdlen, in_len)) goto error; - err = blk_verify_command(&q->cmd_filter, rq->cmd, mode & FMODE_WRITE); + err = blk_verify_command(rq->cmd, mode & FMODE_WRITE); if (err) goto error; @@ -645,5 +673,10 @@ int scsi_cmd_ioctl(struct request_queue *q, struct gendisk *bd_disk, fmode_t mod blk_put_queue(q); return err; } - EXPORT_SYMBOL(scsi_cmd_ioctl); + +int __init blk_scsi_ioctl_init(void) +{ + blk_set_cmd_filter_defaults(&blk_default_cmd_filter); + return 0; +} diff --git a/drivers/scsi/sg.c b/drivers/scsi/sg.c index 8201387b4daa..ef142fd47a83 100644 --- a/drivers/scsi/sg.c +++ b/drivers/scsi/sg.c @@ -210,13 +210,11 @@ static void sg_put_dev(Sg_device *sdp); static int sg_allow_access(struct file *filp, unsigned char *cmd) { struct sg_fd *sfp = (struct sg_fd *)filp->private_data; - struct request_queue *q = sfp->parentdp->device->request_queue; if (sfp->parentdp->device->type == TYPE_SCANNER) return 0; - return blk_verify_command(&q->cmd_filter, - cmd, filp->f_mode & FMODE_WRITE); + return blk_verify_command(cmd, filp->f_mode & FMODE_WRITE); } static int diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index 8963d9149b5f..49ae07951d55 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -301,12 +301,6 @@ struct blk_queue_tag { #define BLK_SCSI_MAX_CMDS (256) #define BLK_SCSI_CMD_PER_LONG (BLK_SCSI_MAX_CMDS / (sizeof(long) * 8)) -struct blk_cmd_filter { - unsigned long read_ok[BLK_SCSI_CMD_PER_LONG]; - unsigned long write_ok[BLK_SCSI_CMD_PER_LONG]; - struct kobject kobj; -}; - struct queue_limits { unsigned long bounce_pfn; unsigned long seg_boundary_mask; @@ -445,7 +439,6 @@ struct request_queue #if defined(CONFIG_BLK_DEV_BSG) struct bsg_class_device bsg_dev; #endif - struct blk_cmd_filter cmd_filter; }; #define QUEUE_FLAG_CLUSTER 0 /* cluster several segments into 1 */ @@ -998,13 +991,7 @@ static inline int sb_issue_discard(struct super_block *sb, return blkdev_issue_discard(sb->s_bdev, block, nr_blocks, GFP_KERNEL); } -/* -* command filter functions -*/ -extern int blk_verify_command(struct blk_cmd_filter *filter, - unsigned char *cmd, fmode_t has_write_perm); -extern void blk_unregister_filter(struct gendisk *disk); -extern void blk_set_cmd_filter_defaults(struct blk_cmd_filter *filter); +extern int blk_verify_command(unsigned char *cmd, fmode_t has_write_perm); #define MAX_PHYS_SEGMENTS 128 #define MAX_HW_SEGMENTS 128 -- cgit v1.2.3 From ee8076ed3e1cdd0cd1e61318386932669c90b92f Mon Sep 17 00:00:00 2001 From: Andres Salomon Date: Thu, 2 Jul 2009 09:45:18 -0400 Subject: power_supply: Add a charge_type property, and use it for olpc driver This adds a new sysfs file called 'charge_type' which displays the type of charging (unknown, n/a, trickle charge, or fast charging). This allows things like battery diagnostics to determine what the battery/EC is doing without resorting to changing the 'status' sysfs output. Signed-off-by: Andres Salomon Acked-by: Mark Brown Signed-off-by: Anton Vorontsov --- Documentation/power/power_supply_class.txt | 5 +++++ drivers/power/olpc_battery.c | 9 +++++++++ drivers/power/power_supply_sysfs.c | 6 ++++++ include/linux/power_supply.h | 8 ++++++++ 4 files changed, 28 insertions(+) (limited to 'include') diff --git a/Documentation/power/power_supply_class.txt b/Documentation/power/power_supply_class.txt index 709d95571d7b..9f16c5178b66 100644 --- a/Documentation/power/power_supply_class.txt +++ b/Documentation/power/power_supply_class.txt @@ -76,6 +76,11 @@ STATUS - this attribute represents operating status (charging, full, discharging (i.e. powering a load), etc.). This corresponds to BATTERY_STATUS_* values, as defined in battery.h. +CHARGE_TYPE - batteries can typically charge at different rates. +This defines trickle and fast charges. For batteries that +are already charged or discharging, 'n/a' can be displayed (or +'unknown', if the status is not known). + HEALTH - represents health of the battery, values corresponds to POWER_SUPPLY_HEALTH_*, defined in battery.h. diff --git a/drivers/power/olpc_battery.c b/drivers/power/olpc_battery.c index 602bbd008f78..8fefe5a73558 100644 --- a/drivers/power/olpc_battery.c +++ b/drivers/power/olpc_battery.c @@ -233,6 +233,14 @@ static int olpc_bat_get_property(struct power_supply *psy, if (ret) return ret; break; + case POWER_SUPPLY_PROP_CHARGE_TYPE: + if (ec_byte & BAT_STAT_TRICKLE) + val->intval = POWER_SUPPLY_CHARGE_TYPE_TRICKLE; + else if (ec_byte & BAT_STAT_CHARGING) + val->intval = POWER_SUPPLY_CHARGE_TYPE_FAST; + else + val->intval = POWER_SUPPLY_CHARGE_TYPE_NONE; + break; case POWER_SUPPLY_PROP_PRESENT: val->intval = !!(ec_byte & (BAT_STAT_PRESENT | BAT_STAT_TRICKLE)); @@ -325,6 +333,7 @@ static int olpc_bat_get_property(struct power_supply *psy, static enum power_supply_property olpc_bat_props[] = { POWER_SUPPLY_PROP_STATUS, + POWER_SUPPLY_PROP_CHARGE_TYPE, POWER_SUPPLY_PROP_PRESENT, POWER_SUPPLY_PROP_HEALTH, POWER_SUPPLY_PROP_TECHNOLOGY, diff --git a/drivers/power/power_supply_sysfs.c b/drivers/power/power_supply_sysfs.c index 9deabbde6fd6..08144393d64b 100644 --- a/drivers/power/power_supply_sysfs.c +++ b/drivers/power/power_supply_sysfs.c @@ -43,6 +43,9 @@ static ssize_t power_supply_show_property(struct device *dev, static char *status_text[] = { "Unknown", "Charging", "Discharging", "Not charging", "Full" }; + static char *charge_type[] = { + "Unknown", "N/A", "Trickle", "Fast" + }; static char *health_text[] = { "Unknown", "Good", "Overheat", "Dead", "Over voltage", "Unspecified failure", "Cold", @@ -70,6 +73,8 @@ static ssize_t power_supply_show_property(struct device *dev, if (off == POWER_SUPPLY_PROP_STATUS) return sprintf(buf, "%s\n", status_text[value.intval]); + else if (off == POWER_SUPPLY_PROP_CHARGE_TYPE) + return sprintf(buf, "%s\n", charge_type[value.intval]); else if (off == POWER_SUPPLY_PROP_HEALTH) return sprintf(buf, "%s\n", health_text[value.intval]); else if (off == POWER_SUPPLY_PROP_TECHNOLOGY) @@ -86,6 +91,7 @@ static ssize_t power_supply_show_property(struct device *dev, static struct device_attribute power_supply_attrs[] = { /* Properties of type `int' */ POWER_SUPPLY_ATTR(status), + POWER_SUPPLY_ATTR(charge_type), POWER_SUPPLY_ATTR(health), POWER_SUPPLY_ATTR(present), POWER_SUPPLY_ATTR(online), diff --git a/include/linux/power_supply.h b/include/linux/power_supply.h index 0ab6aa171241..4c7c6fc35487 100644 --- a/include/linux/power_supply.h +++ b/include/linux/power_supply.h @@ -38,6 +38,13 @@ enum { POWER_SUPPLY_STATUS_FULL, }; +enum { + POWER_SUPPLY_CHARGE_TYPE_UNKNOWN = 0, + POWER_SUPPLY_CHARGE_TYPE_NONE, + POWER_SUPPLY_CHARGE_TYPE_TRICKLE, + POWER_SUPPLY_CHARGE_TYPE_FAST, +}; + enum { POWER_SUPPLY_HEALTH_UNKNOWN = 0, POWER_SUPPLY_HEALTH_GOOD, @@ -70,6 +77,7 @@ enum { enum power_supply_property { /* Properties of type `int' */ POWER_SUPPLY_PROP_STATUS = 0, + POWER_SUPPLY_PROP_CHARGE_TYPE, POWER_SUPPLY_PROP_HEALTH, POWER_SUPPLY_PROP_PRESENT, POWER_SUPPLY_PROP_ONLINE, -- cgit v1.2.3 From 887b5ea3683ce04966e35fa2e5fe215bedcde73c Mon Sep 17 00:00:00 2001 From: Alexander Duyck Date: Wed, 1 Jul 2009 13:38:26 +0000 Subject: if_ether: add define for 1588 aka Timesync This patch adds ETH_P_1588 protocol ID define. Signed-off-by: Alexander Duyck Signed-off-by: David S. Miller --- include/linux/if_ether.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include') diff --git a/include/linux/if_ether.h b/include/linux/if_ether.h index ae3a1871413d..70fdba2bbf71 100644 --- a/include/linux/if_ether.h +++ b/include/linux/if_ether.h @@ -78,6 +78,7 @@ #define ETH_P_PAE 0x888E /* Port Access Entity (IEEE 802.1X) */ #define ETH_P_AOE 0x88A2 /* ATA over Ethernet */ #define ETH_P_TIPC 0x88CA /* TIPC */ +#define ETH_P_1588 0x88F7 /* IEEE 1588 Timesync */ #define ETH_P_FCOE 0x8906 /* Fibre Channel over Ethernet */ #define ETH_P_FIP 0x8914 /* FCoE Initialization Protocol */ #define ETH_P_EDSA 0xDADA /* Ethertype DSA [ NOT AN OFFICIALLY REGISTERED ID ] */ -- cgit v1.2.3 From 03b042bf1dc14a268a3d65d38b4ec2a4261e8477 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Thu, 25 Jun 2009 09:08:16 -0700 Subject: rcu: Add synchronize_sched_expedited() primitive This adds the synchronize_sched_expedited() primitive that implements the "big hammer" expedited RCU grace periods. This primitive is placed in kernel/sched.c rather than kernel/rcupdate.c due to its need to interact closely with the migration_thread() kthread. The idea is to wake up this kthread with req->task set to NULL, in response to which the kthread reports the quiescent state resulting from the kthread having been scheduled. Because this patch needs to fallback to the slow versions of the primitives in response to some races with CPU onlining and offlining, a new synchronize_rcu_bh() primitive is added as well. Signed-off-by: Paul E. McKenney Cc: akpm@linux-foundation.org Cc: torvalds@linux-foundation.org Cc: davem@davemloft.net Cc: dada1@cosmosbay.com Cc: zbr@ioremap.net Cc: jeff.chua.linux@gmail.com Cc: paulus@samba.org Cc: laijs@cn.fujitsu.com Cc: jengelh@medozas.de Cc: r000n@r000n.net Cc: benh@kernel.crashing.org Cc: mathieu.desnoyers@polymtl.ca LKML-Reference: <12459460982947-git-send-email-> Signed-off-by: Ingo Molnar --- include/linux/rcupdate.h | 25 ++++----- include/linux/rcupreempt.h | 10 ++++ include/linux/rcutree.h | 12 ++++- kernel/rcupdate.c | 25 +++++++++ kernel/sched.c | 129 ++++++++++++++++++++++++++++++++++++++++++++- 5 files changed, 186 insertions(+), 15 deletions(-) (limited to 'include') diff --git a/include/linux/rcupdate.h b/include/linux/rcupdate.h index 0cdfdb622faa..3c89d6a2591f 100644 --- a/include/linux/rcupdate.h +++ b/include/linux/rcupdate.h @@ -51,7 +51,19 @@ struct rcu_head { void (*func)(struct rcu_head *head); }; -/* Internal to kernel, but needed by rcupreempt.h. */ +/* Exported common interfaces */ +extern void synchronize_rcu(void); +extern void synchronize_rcu_bh(void); +extern void rcu_barrier(void); +extern void rcu_barrier_bh(void); +extern void rcu_barrier_sched(void); +extern void synchronize_sched_expedited(void); +extern int sched_expedited_torture_stats(char *page); + +/* Internal to kernel */ +extern void rcu_init(void); +extern void rcu_scheduler_starting(void); +extern int rcu_needs_cpu(int cpu); extern int rcu_scheduler_active; #if defined(CONFIG_TREE_RCU) @@ -257,15 +269,4 @@ extern void call_rcu(struct rcu_head *head, extern void call_rcu_bh(struct rcu_head *head, void (*func)(struct rcu_head *head)); -/* Exported common interfaces */ -extern void synchronize_rcu(void); -extern void rcu_barrier(void); -extern void rcu_barrier_bh(void); -extern void rcu_barrier_sched(void); - -/* Internal to kernel */ -extern void rcu_init(void); -extern void rcu_scheduler_starting(void); -extern int rcu_needs_cpu(int cpu); - #endif /* __LINUX_RCUPDATE_H */ diff --git a/include/linux/rcupreempt.h b/include/linux/rcupreempt.h index fce522782ffa..f164ac9b7807 100644 --- a/include/linux/rcupreempt.h +++ b/include/linux/rcupreempt.h @@ -74,6 +74,16 @@ extern int rcu_needs_cpu(int cpu); extern void __synchronize_sched(void); +static inline void synchronize_rcu_expedited(void) +{ + synchronize_rcu(); /* Placeholder for new rcupreempt implementation. */ +} + +static inline void synchronize_rcu_bh_expedited(void) +{ + synchronize_rcu_bh(); /* Placeholder for new rcupreempt impl. */ +} + extern void __rcu_init(void); extern void rcu_init_sched(void); extern void rcu_check_callbacks(int cpu, int user); diff --git a/include/linux/rcutree.h b/include/linux/rcutree.h index 5a5153806c42..d4dfd2489633 100644 --- a/include/linux/rcutree.h +++ b/include/linux/rcutree.h @@ -286,8 +286,14 @@ static inline void __rcu_read_unlock_bh(void) #define call_rcu_sched(head, func) call_rcu(head, func) -static inline void rcu_init_sched(void) +static inline void synchronize_rcu_expedited(void) +{ + synchronize_sched_expedited(); +} + +static inline void synchronize_rcu_bh_expedited(void) { + synchronize_sched_expedited(); } extern void __rcu_init(void); @@ -297,6 +303,10 @@ extern void rcu_restart_cpu(int cpu); extern long rcu_batches_completed(void); extern long rcu_batches_completed_bh(void); +static inline void rcu_init_sched(void) +{ +} + #ifdef CONFIG_NO_HZ void rcu_enter_nohz(void); void rcu_exit_nohz(void); diff --git a/kernel/rcupdate.c b/kernel/rcupdate.c index a967c9feb90a..eae29c25fb14 100644 --- a/kernel/rcupdate.c +++ b/kernel/rcupdate.c @@ -98,6 +98,30 @@ void synchronize_rcu(void) } EXPORT_SYMBOL_GPL(synchronize_rcu); +/** + * synchronize_rcu_bh - wait until an rcu_bh grace period has elapsed. + * + * Control will return to the caller some time after a full rcu_bh grace + * period has elapsed, in other words after all currently executing rcu_bh + * read-side critical sections have completed. RCU read-side critical + * sections are delimited by rcu_read_lock_bh() and rcu_read_unlock_bh(), + * and may be nested. + */ +void synchronize_rcu_bh(void) +{ + struct rcu_synchronize rcu; + + if (rcu_blocking_is_gp()) + return; + + init_completion(&rcu.completion); + /* Will wake me after RCU finished. */ + call_rcu_bh(&rcu.head, wakeme_after_rcu); + /* Wait for it. */ + wait_for_completion(&rcu.completion); +} +EXPORT_SYMBOL_GPL(synchronize_rcu_bh); + static void rcu_barrier_callback(struct rcu_head *notused) { if (atomic_dec_and_test(&rcu_barrier_cpu_count)) @@ -129,6 +153,7 @@ static void rcu_barrier_func(void *type) static inline void wait_migrated_callbacks(void) { wait_event(rcu_migrate_wq, !atomic_read(&rcu_migrate_type_count)); + smp_mb(); /* In case we didn't sleep. */ } /* diff --git a/kernel/sched.c b/kernel/sched.c index 7c9098d186e6..9ae80bec1c1e 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -7024,6 +7024,11 @@ fail: return ret; } +#define RCU_MIGRATION_IDLE 0 +#define RCU_MIGRATION_NEED_QS 1 +#define RCU_MIGRATION_GOT_QS 2 +#define RCU_MIGRATION_MUST_SYNC 3 + /* * migration_thread - this is a highprio system thread that performs * thread migration by bumping thread off CPU then 'pushing' onto @@ -7031,6 +7036,7 @@ fail: */ static int migration_thread(void *data) { + int badcpu; int cpu = (long)data; struct rq *rq; @@ -7065,8 +7071,17 @@ static int migration_thread(void *data) req = list_entry(head->next, struct migration_req, list); list_del_init(head->next); - spin_unlock(&rq->lock); - __migrate_task(req->task, cpu, req->dest_cpu); + if (req->task != NULL) { + spin_unlock(&rq->lock); + __migrate_task(req->task, cpu, req->dest_cpu); + } else if (likely(cpu == (badcpu = smp_processor_id()))) { + req->dest_cpu = RCU_MIGRATION_GOT_QS; + spin_unlock(&rq->lock); + } else { + req->dest_cpu = RCU_MIGRATION_MUST_SYNC; + spin_unlock(&rq->lock); + WARN_ONCE(1, "migration_thread() on CPU %d, expected %d\n", badcpu, cpu); + } local_irq_enable(); complete(&req->done); @@ -10554,3 +10569,113 @@ struct cgroup_subsys cpuacct_subsys = { .subsys_id = cpuacct_subsys_id, }; #endif /* CONFIG_CGROUP_CPUACCT */ + +#ifndef CONFIG_SMP + +int rcu_expedited_torture_stats(char *page) +{ + return 0; +} +EXPORT_SYMBOL_GPL(rcu_expedited_torture_stats); + +void synchronize_sched_expedited(void) +{ +} +EXPORT_SYMBOL_GPL(synchronize_sched_expedited); + +#else /* #ifndef CONFIG_SMP */ + +static DEFINE_PER_CPU(struct migration_req, rcu_migration_req); +static DEFINE_MUTEX(rcu_sched_expedited_mutex); + +#define RCU_EXPEDITED_STATE_POST -2 +#define RCU_EXPEDITED_STATE_IDLE -1 + +static int rcu_expedited_state = RCU_EXPEDITED_STATE_IDLE; + +int rcu_expedited_torture_stats(char *page) +{ + int cnt = 0; + int cpu; + + cnt += sprintf(&page[cnt], "state: %d /", rcu_expedited_state); + for_each_online_cpu(cpu) { + cnt += sprintf(&page[cnt], " %d:%d", + cpu, per_cpu(rcu_migration_req, cpu).dest_cpu); + } + cnt += sprintf(&page[cnt], "\n"); + return cnt; +} +EXPORT_SYMBOL_GPL(rcu_expedited_torture_stats); + +static long synchronize_sched_expedited_count; + +/* + * Wait for an rcu-sched grace period to elapse, but use "big hammer" + * approach to force grace period to end quickly. This consumes + * significant time on all CPUs, and is thus not recommended for + * any sort of common-case code. + * + * Note that it is illegal to call this function while holding any + * lock that is acquired by a CPU-hotplug notifier. Failing to + * observe this restriction will result in deadlock. + */ +void synchronize_sched_expedited(void) +{ + int cpu; + unsigned long flags; + bool need_full_sync = 0; + struct rq *rq; + struct migration_req *req; + long snap; + int trycount = 0; + + smp_mb(); /* ensure prior mod happens before capturing snap. */ + snap = ACCESS_ONCE(synchronize_sched_expedited_count) + 1; + get_online_cpus(); + while (!mutex_trylock(&rcu_sched_expedited_mutex)) { + put_online_cpus(); + if (trycount++ < 10) + udelay(trycount * num_online_cpus()); + else { + synchronize_sched(); + return; + } + if (ACCESS_ONCE(synchronize_sched_expedited_count) - snap > 0) { + smp_mb(); /* ensure test happens before caller kfree */ + return; + } + get_online_cpus(); + } + rcu_expedited_state = RCU_EXPEDITED_STATE_POST; + for_each_online_cpu(cpu) { + rq = cpu_rq(cpu); + req = &per_cpu(rcu_migration_req, cpu); + init_completion(&req->done); + req->task = NULL; + req->dest_cpu = RCU_MIGRATION_NEED_QS; + spin_lock_irqsave(&rq->lock, flags); + list_add(&req->list, &rq->migration_queue); + spin_unlock_irqrestore(&rq->lock, flags); + wake_up_process(rq->migration_thread); + } + for_each_online_cpu(cpu) { + rcu_expedited_state = cpu; + req = &per_cpu(rcu_migration_req, cpu); + rq = cpu_rq(cpu); + wait_for_completion(&req->done); + spin_lock_irqsave(&rq->lock, flags); + if (unlikely(req->dest_cpu == RCU_MIGRATION_MUST_SYNC)) + need_full_sync = 1; + req->dest_cpu = RCU_MIGRATION_IDLE; + spin_unlock_irqrestore(&rq->lock, flags); + } + rcu_expedited_state = RCU_EXPEDITED_STATE_IDLE; + mutex_unlock(&rcu_sched_expedited_mutex); + put_online_cpus(); + if (need_full_sync) + synchronize_sched(); +} +EXPORT_SYMBOL_GPL(synchronize_sched_expedited); + +#endif /* #else #ifndef CONFIG_SMP */ -- cgit v1.2.3 From 788e5abc5441e9046dd91c995c6f1f75bbd144bf Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Sat, 4 Jul 2009 08:10:58 +0900 Subject: percpu: drop @unit_size from embed first chunk allocator The only extra feature @unit_size provides is making dead space at the end of the first chunk which doesn't have any valid usecase. Drop the parameter. This will increase consistency with generalized 4k allocator. James Bottomley spotted missing conversion for the default setup_per_cpu_areas() which caused build breakage on all arcsh which use it. [ Impact: drop unused code path ] Signed-off-by: Tejun Heo Cc: James Bottomley Cc: Ingo Molnar --- arch/x86/kernel/setup_percpu.c | 2 +- include/linux/percpu.h | 2 +- mm/percpu.c | 18 ++++++------------ 3 files changed, 8 insertions(+), 14 deletions(-) (limited to 'include') diff --git a/arch/x86/kernel/setup_percpu.c b/arch/x86/kernel/setup_percpu.c index 29a3eef7cf4a..14728206fb52 100644 --- a/arch/x86/kernel/setup_percpu.c +++ b/arch/x86/kernel/setup_percpu.c @@ -342,7 +342,7 @@ static ssize_t __init setup_pcpu_embed(size_t static_size, bool chosen) return -EINVAL; return pcpu_embed_first_chunk(static_size, PERCPU_FIRST_CHUNK_RESERVE, - reserve - PERCPU_FIRST_CHUNK_RESERVE, -1); + reserve - PERCPU_FIRST_CHUNK_RESERVE); } /* diff --git a/include/linux/percpu.h b/include/linux/percpu.h index e5000343dd61..83bff053bd1c 100644 --- a/include/linux/percpu.h +++ b/include/linux/percpu.h @@ -69,7 +69,7 @@ extern size_t __init pcpu_setup_first_chunk(pcpu_get_page_fn_t get_page_fn, extern ssize_t __init pcpu_embed_first_chunk( size_t static_size, size_t reserved_size, - ssize_t dyn_size, ssize_t unit_size); + ssize_t dyn_size); /* * Use this to get to a cpu's version of the per-cpu object diff --git a/mm/percpu.c b/mm/percpu.c index 19dd83b5cbdc..fc6babe6e554 100644 --- a/mm/percpu.c +++ b/mm/percpu.c @@ -1207,7 +1207,6 @@ static struct page * __init pcpue_get_page(unsigned int cpu, int pageno) * @static_size: the size of static percpu area in bytes * @reserved_size: the size of reserved percpu area in bytes * @dyn_size: free size for dynamic allocation in bytes, -1 for auto - * @unit_size: unit size in bytes, must be multiple of PAGE_SIZE, -1 for auto * * This is a helper to ease setting up embedded first percpu chunk and * can be called where pcpu_setup_first_chunk() is expected. @@ -1219,9 +1218,9 @@ static struct page * __init pcpue_get_page(unsigned int cpu, int pageno) * page size. * * When @dyn_size is positive, dynamic area might be larger than - * specified to fill page alignment. Also, when @dyn_size is auto, - * @dyn_size does not fill the whole first chunk but only what's - * necessary for page alignment after static and reserved areas. + * specified to fill page alignment. When @dyn_size is auto, + * @dyn_size is just big enough to fill page alignment after static + * and reserved areas. * * If the needed size is smaller than the minimum or specified unit * size, the leftover is returned to the bootmem allocator. @@ -1231,7 +1230,7 @@ static struct page * __init pcpue_get_page(unsigned int cpu, int pageno) * percpu access on success, -errno on failure. */ ssize_t __init pcpu_embed_first_chunk(size_t static_size, size_t reserved_size, - ssize_t dyn_size, ssize_t unit_size) + ssize_t dyn_size) { size_t chunk_size; unsigned int cpu; @@ -1242,12 +1241,7 @@ ssize_t __init pcpu_embed_first_chunk(size_t static_size, size_t reserved_size, if (dyn_size != 0) dyn_size = pcpue_size - static_size - reserved_size; - if (unit_size >= 0) { - BUG_ON(unit_size < pcpue_size); - pcpue_unit_size = unit_size; - } else - pcpue_unit_size = max_t(size_t, pcpue_size, PCPU_MIN_UNIT_SIZE); - + pcpue_unit_size = max_t(size_t, pcpue_size, PCPU_MIN_UNIT_SIZE); chunk_size = pcpue_unit_size * num_possible_cpus(); pcpue_ptr = __alloc_bootmem_nopanic(chunk_size, PAGE_SIZE, @@ -1304,7 +1298,7 @@ void __init setup_per_cpu_areas(void) * what the legacy allocator did. */ unit_size = pcpu_embed_first_chunk(static_size, PERCPU_MODULE_RESERVE, - PERCPU_DYNAMIC_RESERVE, -1); + PERCPU_DYNAMIC_RESERVE); if (unit_size < 0) panic("Failed to initialized percpu areas."); -- cgit v1.2.3 From d4b95f80399471e4bce5e992700ff7f06ef91f6a Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Sat, 4 Jul 2009 08:10:59 +0900 Subject: x86,percpu: generalize 4k first chunk allocator Generalize and move x86 setup_pcpu_4k() into pcpu_4k_first_chunk(). setup_pcpu_4k() now is a simple wrapper around the generalized version. Other than taking size parameters and using arch supplied callbacks to allocate/free memory, pcpu_4k_first_chunk() is identical to the original implementation. This simplifies arch code and will help converting more archs to dynamic percpu allocator. While at it, s/pcpu_populate_pte_fn_t/pcpu_fc_populate_pte_fn_t/ for consistency. [ Impact: code reorganization and generalization ] Signed-off-by: Tejun Heo Cc: Ingo Molnar --- arch/x86/kernel/setup_percpu.c | 78 ++++++++++---------------------------- include/linux/percpu.h | 12 +++++- mm/percpu.c | 85 +++++++++++++++++++++++++++++++++++++++++- 3 files changed, 113 insertions(+), 62 deletions(-) (limited to 'include') diff --git a/arch/x86/kernel/setup_percpu.c b/arch/x86/kernel/setup_percpu.c index 14728206fb52..ab896b31e80b 100644 --- a/arch/x86/kernel/setup_percpu.c +++ b/arch/x86/kernel/setup_percpu.c @@ -123,6 +123,19 @@ static void * __init pcpu_alloc_bootmem(unsigned int cpu, unsigned long size, #endif } +/* + * Helpers for first chunk memory allocation + */ +static void * __init pcpu_fc_alloc(unsigned int cpu, size_t size) +{ + return pcpu_alloc_bootmem(cpu, size, size); +} + +static void __init pcpu_fc_free(void *ptr, size_t size) +{ + free_bootmem(__pa(ptr), size); +} + /* * Large page remap allocator * @@ -346,22 +359,11 @@ static ssize_t __init setup_pcpu_embed(size_t static_size, bool chosen) } /* - * 4k page allocator + * 4k allocator * - * This is the basic allocator. Static percpu area is allocated - * page-by-page and most of initialization is done by the generic - * setup function. + * Boring fallback 4k allocator. This allocator puts more pressure on + * PTE TLBs but other than that behaves nicely on both UMA and NUMA. */ -static struct page **pcpu4k_pages __initdata; -static int pcpu4k_nr_static_pages __initdata; - -static struct page * __init pcpu4k_get_page(unsigned int cpu, int pageno) -{ - if (pageno < pcpu4k_nr_static_pages) - return pcpu4k_pages[cpu * pcpu4k_nr_static_pages + pageno]; - return NULL; -} - static void __init pcpu4k_populate_pte(unsigned long addr) { populate_extra_pte(addr); @@ -369,51 +371,9 @@ static void __init pcpu4k_populate_pte(unsigned long addr) static ssize_t __init setup_pcpu_4k(size_t static_size) { - size_t pages_size; - unsigned int cpu; - int i, j; - ssize_t ret; - - pcpu4k_nr_static_pages = PFN_UP(static_size); - - /* unaligned allocations can't be freed, round up to page size */ - pages_size = PFN_ALIGN(pcpu4k_nr_static_pages * num_possible_cpus() - * sizeof(pcpu4k_pages[0])); - pcpu4k_pages = alloc_bootmem(pages_size); - - /* allocate and copy */ - j = 0; - for_each_possible_cpu(cpu) - for (i = 0; i < pcpu4k_nr_static_pages; i++) { - void *ptr; - - ptr = pcpu_alloc_bootmem(cpu, PAGE_SIZE, PAGE_SIZE); - if (!ptr) { - pr_warning("PERCPU: failed to allocate " - "4k page for cpu%u\n", cpu); - goto enomem; - } - - memcpy(ptr, __per_cpu_load + i * PAGE_SIZE, PAGE_SIZE); - pcpu4k_pages[j++] = virt_to_page(ptr); - } - - /* we're ready, commit */ - pr_info("PERCPU: Allocated %d 4k pages, static data %zu bytes\n", - pcpu4k_nr_static_pages, static_size); - - ret = pcpu_setup_first_chunk(pcpu4k_get_page, static_size, - PERCPU_FIRST_CHUNK_RESERVE, -1, - -1, NULL, pcpu4k_populate_pte); - goto out_free_ar; - -enomem: - while (--j >= 0) - free_bootmem(__pa(page_address(pcpu4k_pages[j])), PAGE_SIZE); - ret = -ENOMEM; -out_free_ar: - free_bootmem(__pa(pcpu4k_pages), pages_size); - return ret; + return pcpu_4k_first_chunk(static_size, PERCPU_FIRST_CHUNK_RESERVE, + pcpu_fc_alloc, pcpu_fc_free, + pcpu4k_populate_pte); } /* for explicit first chunk allocator selection */ diff --git a/include/linux/percpu.h b/include/linux/percpu.h index 83bff053bd1c..41b5bfab4195 100644 --- a/include/linux/percpu.h +++ b/include/linux/percpu.h @@ -59,18 +59,26 @@ extern void *pcpu_base_addr; typedef struct page * (*pcpu_get_page_fn_t)(unsigned int cpu, int pageno); -typedef void (*pcpu_populate_pte_fn_t)(unsigned long addr); +typedef void * (*pcpu_fc_alloc_fn_t)(unsigned int cpu, size_t size); +typedef void (*pcpu_fc_free_fn_t)(void *ptr, size_t size); +typedef void (*pcpu_fc_populate_pte_fn_t)(unsigned long addr); extern size_t __init pcpu_setup_first_chunk(pcpu_get_page_fn_t get_page_fn, size_t static_size, size_t reserved_size, ssize_t dyn_size, ssize_t unit_size, void *base_addr, - pcpu_populate_pte_fn_t populate_pte_fn); + pcpu_fc_populate_pte_fn_t populate_pte_fn); extern ssize_t __init pcpu_embed_first_chunk( size_t static_size, size_t reserved_size, ssize_t dyn_size); +extern ssize_t __init pcpu_4k_first_chunk( + size_t static_size, size_t reserved_size, + pcpu_fc_alloc_fn_t alloc_fn, + pcpu_fc_free_fn_t free_fn, + pcpu_fc_populate_pte_fn_t populate_pte_fn); + /* * Use this to get to a cpu's version of the per-cpu object * dynamically allocated. Non-atomic access to the current CPU's diff --git a/mm/percpu.c b/mm/percpu.c index fc6babe6e554..27b0f40a3ea8 100644 --- a/mm/percpu.c +++ b/mm/percpu.c @@ -1037,7 +1037,7 @@ size_t __init pcpu_setup_first_chunk(pcpu_get_page_fn_t get_page_fn, size_t static_size, size_t reserved_size, ssize_t dyn_size, ssize_t unit_size, void *base_addr, - pcpu_populate_pte_fn_t populate_pte_fn) + pcpu_fc_populate_pte_fn_t populate_pte_fn) { static struct vm_struct first_vm; static int smap[2], dmap[2]; @@ -1270,6 +1270,89 @@ ssize_t __init pcpu_embed_first_chunk(size_t static_size, size_t reserved_size, pcpue_unit_size, pcpue_ptr, NULL); } +/* + * 4k page first chunk setup helper. + */ +static struct page **pcpu4k_pages __initdata; +static int pcpu4k_nr_static_pages __initdata; + +static struct page * __init pcpu4k_get_page(unsigned int cpu, int pageno) +{ + if (pageno < pcpu4k_nr_static_pages) + return pcpu4k_pages[cpu * pcpu4k_nr_static_pages + pageno]; + return NULL; +} + +/** + * pcpu_4k_first_chunk - map the first chunk using PAGE_SIZE pages + * @static_size: the size of static percpu area in bytes + * @reserved_size: the size of reserved percpu area in bytes + * @alloc_fn: function to allocate percpu page, always called with PAGE_SIZE + * @free_fn: funtion to free percpu page, always called with PAGE_SIZE + * @populate_pte_fn: function to populate pte + * + * This is a helper to ease setting up embedded first percpu chunk and + * can be called where pcpu_setup_first_chunk() is expected. + * + * This is the basic allocator. Static percpu area is allocated + * page-by-page into vmalloc area. + * + * RETURNS: + * The determined pcpu_unit_size which can be used to initialize + * percpu access on success, -errno on failure. + */ +ssize_t __init pcpu_4k_first_chunk(size_t static_size, size_t reserved_size, + pcpu_fc_alloc_fn_t alloc_fn, + pcpu_fc_free_fn_t free_fn, + pcpu_fc_populate_pte_fn_t populate_pte_fn) +{ + size_t pages_size; + unsigned int cpu; + int i, j; + ssize_t ret; + + pcpu4k_nr_static_pages = PFN_UP(static_size); + + /* unaligned allocations can't be freed, round up to page size */ + pages_size = PFN_ALIGN(pcpu4k_nr_static_pages * num_possible_cpus() * + sizeof(pcpu4k_pages[0])); + pcpu4k_pages = alloc_bootmem(pages_size); + + /* allocate and copy */ + j = 0; + for_each_possible_cpu(cpu) + for (i = 0; i < pcpu4k_nr_static_pages; i++) { + void *ptr; + + ptr = alloc_fn(cpu, PAGE_SIZE); + if (!ptr) { + pr_warning("PERCPU: failed to allocate " + "4k page for cpu%u\n", cpu); + goto enomem; + } + + memcpy(ptr, __per_cpu_load + i * PAGE_SIZE, PAGE_SIZE); + pcpu4k_pages[j++] = virt_to_page(ptr); + } + + /* we're ready, commit */ + pr_info("PERCPU: Allocated %d 4k pages, static data %zu bytes\n", + pcpu4k_nr_static_pages, static_size); + + ret = pcpu_setup_first_chunk(pcpu4k_get_page, static_size, + reserved_size, -1, + -1, NULL, populate_pte_fn); + goto out_free_ar; + +enomem: + while (--j >= 0) + free_fn(page_address(pcpu4k_pages[j]), PAGE_SIZE); + ret = -ENOMEM; +out_free_ar: + free_bootmem(__pa(pcpu4k_pages), pages_size); + return ret; +} + /* * Generic percpu area setup. * -- cgit v1.2.3 From 8c4bfc6e8801616ab2e01c38140b2159b388d2ff Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Sat, 4 Jul 2009 08:10:59 +0900 Subject: x86,percpu: generalize lpage first chunk allocator Generalize and move x86 setup_pcpu_lpage() into pcpu_lpage_first_chunk(). setup_pcpu_lpage() now is a simple wrapper around the generalized version. Other than taking size parameters and using arch supplied callbacks to allocate/free/map memory, pcpu_lpage_first_chunk() is identical to the original implementation. This simplifies arch code and will help converting more archs to dynamic percpu allocator. While at it, factor out pcpu_calc_fc_sizes() which is common to pcpu_embed_first_chunk() and pcpu_lpage_first_chunk(). [ Impact: code reorganization and generalization ] Signed-off-by: Tejun Heo Cc: Ingo Molnar --- arch/x86/include/asm/percpu.h | 9 -- arch/x86/kernel/setup_percpu.c | 169 +++------------------------------ arch/x86/mm/pageattr.c | 1 + include/linux/percpu.h | 27 ++++++ mm/percpu.c | 209 ++++++++++++++++++++++++++++++++++++++++- 5 files changed, 244 insertions(+), 171 deletions(-) (limited to 'include') diff --git a/arch/x86/include/asm/percpu.h b/arch/x86/include/asm/percpu.h index 103f1ddb0d85..a18c038a3079 100644 --- a/arch/x86/include/asm/percpu.h +++ b/arch/x86/include/asm/percpu.h @@ -156,15 +156,6 @@ do { \ /* We can use this directly for local CPU (faster). */ DECLARE_PER_CPU(unsigned long, this_cpu_off); -#ifdef CONFIG_NEED_MULTIPLE_NODES -void *pcpu_lpage_remapped(void *kaddr); -#else -static inline void *pcpu_lpage_remapped(void *kaddr) -{ - return NULL; -} -#endif - #endif /* !__ASSEMBLY__ */ #ifdef CONFIG_SMP diff --git a/arch/x86/kernel/setup_percpu.c b/arch/x86/kernel/setup_percpu.c index ab896b31e80b..4f2e0ac9130b 100644 --- a/arch/x86/kernel/setup_percpu.c +++ b/arch/x86/kernel/setup_percpu.c @@ -137,44 +137,21 @@ static void __init pcpu_fc_free(void *ptr, size_t size) } /* - * Large page remap allocator - * - * This allocator uses PMD page as unit. A PMD page is allocated for - * each cpu and each is remapped into vmalloc area using PMD mapping. - * As PMD page is quite large, only part of it is used for the first - * chunk. Unused part is returned to the bootmem allocator. - * - * So, the PMD pages are mapped twice - once to the physical mapping - * and to the vmalloc area for the first percpu chunk. The double - * mapping does add one more PMD TLB entry pressure but still is much - * better than only using 4k mappings while still being NUMA friendly. + * Large page remapping allocator */ #ifdef CONFIG_NEED_MULTIPLE_NODES -struct pcpul_ent { - unsigned int cpu; - void *ptr; -}; - -static size_t pcpul_size; -static struct pcpul_ent *pcpul_map; -static struct vm_struct pcpul_vm; - -static struct page * __init pcpul_get_page(unsigned int cpu, int pageno) +static void __init pcpul_map(void *ptr, size_t size, void *addr) { - size_t off = (size_t)pageno << PAGE_SHIFT; + pmd_t *pmd, pmd_v; - if (off >= pcpul_size) - return NULL; - - return virt_to_page(pcpul_map[cpu].ptr + off); + pmd = populate_extra_pmd((unsigned long)addr); + pmd_v = pfn_pmd(page_to_pfn(virt_to_page(ptr)), PAGE_KERNEL_LARGE); + set_pmd(pmd, pmd_v); } static ssize_t __init setup_pcpu_lpage(size_t static_size, bool chosen) { - size_t map_size, dyn_size; - unsigned int cpu; - int i, j; - ssize_t ret; + size_t reserve = PERCPU_MODULE_RESERVE + PERCPU_DYNAMIC_RESERVE; if (!chosen) { size_t vm_size = VMALLOC_END - VMALLOC_START; @@ -198,134 +175,10 @@ static ssize_t __init setup_pcpu_lpage(size_t static_size, bool chosen) return -EINVAL; } - /* - * Currently supports only single page. Supporting multiple - * pages won't be too difficult if it ever becomes necessary. - */ - pcpul_size = PFN_ALIGN(static_size + PERCPU_MODULE_RESERVE + - PERCPU_DYNAMIC_RESERVE); - if (pcpul_size > PMD_SIZE) { - pr_warning("PERCPU: static data is larger than large page, " - "can't use large page\n"); - return -EINVAL; - } - dyn_size = pcpul_size - static_size - PERCPU_FIRST_CHUNK_RESERVE; - - /* allocate pointer array and alloc large pages */ - map_size = PFN_ALIGN(num_possible_cpus() * sizeof(pcpul_map[0])); - pcpul_map = alloc_bootmem(map_size); - - for_each_possible_cpu(cpu) { - pcpul_map[cpu].cpu = cpu; - pcpul_map[cpu].ptr = pcpu_alloc_bootmem(cpu, PMD_SIZE, - PMD_SIZE); - if (!pcpul_map[cpu].ptr) { - pr_warning("PERCPU: failed to allocate large page " - "for cpu%u\n", cpu); - goto enomem; - } - - /* - * Only use pcpul_size bytes and give back the rest. - * - * Ingo: The 2MB up-rounding bootmem is needed to make - * sure the partial 2MB page is still fully RAM - it's - * not well-specified to have a PAT-incompatible area - * (unmapped RAM, device memory, etc.) in that hole. - */ - free_bootmem(__pa(pcpul_map[cpu].ptr + pcpul_size), - PMD_SIZE - pcpul_size); - - memcpy(pcpul_map[cpu].ptr, __per_cpu_load, static_size); - } - - /* allocate address and map */ - pcpul_vm.flags = VM_ALLOC; - pcpul_vm.size = num_possible_cpus() * PMD_SIZE; - vm_area_register_early(&pcpul_vm, PMD_SIZE); - - for_each_possible_cpu(cpu) { - pmd_t *pmd, pmd_v; - - pmd = populate_extra_pmd((unsigned long)pcpul_vm.addr + - cpu * PMD_SIZE); - pmd_v = pfn_pmd(page_to_pfn(virt_to_page(pcpul_map[cpu].ptr)), - PAGE_KERNEL_LARGE); - set_pmd(pmd, pmd_v); - } - - /* we're ready, commit */ - pr_info("PERCPU: Remapped at %p with large pages, static data " - "%zu bytes\n", pcpul_vm.addr, static_size); - - ret = pcpu_setup_first_chunk(pcpul_get_page, static_size, - PERCPU_FIRST_CHUNK_RESERVE, dyn_size, - PMD_SIZE, pcpul_vm.addr, NULL); - - /* sort pcpul_map array for pcpu_lpage_remapped() */ - for (i = 0; i < num_possible_cpus() - 1; i++) - for (j = i + 1; j < num_possible_cpus(); j++) - if (pcpul_map[i].ptr > pcpul_map[j].ptr) { - struct pcpul_ent tmp = pcpul_map[i]; - pcpul_map[i] = pcpul_map[j]; - pcpul_map[j] = tmp; - } - - return ret; - -enomem: - for_each_possible_cpu(cpu) - if (pcpul_map[cpu].ptr) - free_bootmem(__pa(pcpul_map[cpu].ptr), pcpul_size); - free_bootmem(__pa(pcpul_map), map_size); - return -ENOMEM; -} - -/** - * pcpu_lpage_remapped - determine whether a kaddr is in pcpul recycled area - * @kaddr: the kernel address in question - * - * Determine whether @kaddr falls in the pcpul recycled area. This is - * used by pageattr to detect VM aliases and break up the pcpu PMD - * mapping such that the same physical page is not mapped under - * different attributes. - * - * The recycled area is always at the tail of a partially used PMD - * page. - * - * RETURNS: - * Address of corresponding remapped pcpu address if match is found; - * otherwise, NULL. - */ -void *pcpu_lpage_remapped(void *kaddr) -{ - void *pmd_addr = (void *)((unsigned long)kaddr & PMD_MASK); - unsigned long offset = (unsigned long)kaddr & ~PMD_MASK; - int left = 0, right = num_possible_cpus() - 1; - int pos; - - /* pcpul in use at all? */ - if (!pcpul_map) - return NULL; - - /* okay, perform binary search */ - while (left <= right) { - pos = (left + right) / 2; - - if (pcpul_map[pos].ptr < pmd_addr) - left = pos + 1; - else if (pcpul_map[pos].ptr > pmd_addr) - right = pos - 1; - else { - /* it shouldn't be in the area for the first chunk */ - WARN_ON(offset < pcpul_size); - - return pcpul_vm.addr + - pcpul_map[pos].cpu * PMD_SIZE + offset; - } - } - - return NULL; + return pcpu_lpage_first_chunk(static_size, PERCPU_FIRST_CHUNK_RESERVE, + reserve - PERCPU_FIRST_CHUNK_RESERVE, + PMD_SIZE, + pcpu_fc_alloc, pcpu_fc_free, pcpul_map); } #else static ssize_t __init setup_pcpu_lpage(size_t static_size, bool chosen) diff --git a/arch/x86/mm/pageattr.c b/arch/x86/mm/pageattr.c index 1b734d7a8966..c106f7852424 100644 --- a/arch/x86/mm/pageattr.c +++ b/arch/x86/mm/pageattr.c @@ -12,6 +12,7 @@ #include #include #include +#include #include #include diff --git a/include/linux/percpu.h b/include/linux/percpu.h index 41b5bfab4195..9f6bfd7d4b92 100644 --- a/include/linux/percpu.h +++ b/include/linux/percpu.h @@ -62,6 +62,7 @@ typedef struct page * (*pcpu_get_page_fn_t)(unsigned int cpu, int pageno); typedef void * (*pcpu_fc_alloc_fn_t)(unsigned int cpu, size_t size); typedef void (*pcpu_fc_free_fn_t)(void *ptr, size_t size); typedef void (*pcpu_fc_populate_pte_fn_t)(unsigned long addr); +typedef void (*pcpu_fc_map_fn_t)(void *ptr, size_t size, void *addr); extern size_t __init pcpu_setup_first_chunk(pcpu_get_page_fn_t get_page_fn, size_t static_size, size_t reserved_size, @@ -79,6 +80,32 @@ extern ssize_t __init pcpu_4k_first_chunk( pcpu_fc_free_fn_t free_fn, pcpu_fc_populate_pte_fn_t populate_pte_fn); +#ifdef CONFIG_NEED_MULTIPLE_NODES +extern ssize_t __init pcpu_lpage_first_chunk( + size_t static_size, size_t reserved_size, + ssize_t dyn_size, size_t lpage_size, + pcpu_fc_alloc_fn_t alloc_fn, + pcpu_fc_free_fn_t free_fn, + pcpu_fc_map_fn_t map_fn); + +extern void *pcpu_lpage_remapped(void *kaddr); +#else +static inline ssize_t __init pcpu_lpage_first_chunk( + size_t static_size, size_t reserved_size, + ssize_t dyn_size, size_t lpage_size, + pcpu_fc_alloc_fn_t alloc_fn, + pcpu_fc_free_fn_t free_fn, + pcpu_fc_map_fn_t map_fn) +{ + return -EINVAL; +} + +static inline void *pcpu_lpage_remapped(void *kaddr) +{ + return NULL; +} +#endif + /* * Use this to get to a cpu's version of the per-cpu object * dynamically allocated. Non-atomic access to the current CPU's diff --git a/mm/percpu.c b/mm/percpu.c index f3fe7bc7378f..17db527ee2e2 100644 --- a/mm/percpu.c +++ b/mm/percpu.c @@ -1190,6 +1190,19 @@ size_t __init pcpu_setup_first_chunk(pcpu_get_page_fn_t get_page_fn, return pcpu_unit_size; } +static size_t pcpu_calc_fc_sizes(size_t static_size, size_t reserved_size, + ssize_t *dyn_sizep) +{ + size_t size_sum; + + size_sum = PFN_ALIGN(static_size + reserved_size + + (*dyn_sizep >= 0 ? *dyn_sizep : 0)); + if (*dyn_sizep != 0) + *dyn_sizep = size_sum - static_size - reserved_size; + + return size_sum; +} + /* * Embedding first chunk setup helper. */ @@ -1241,10 +1254,7 @@ ssize_t __init pcpu_embed_first_chunk(size_t static_size, size_t reserved_size, unsigned int cpu; /* determine parameters and allocate */ - pcpue_size = PFN_ALIGN(static_size + reserved_size + - (dyn_size >= 0 ? dyn_size : 0)); - if (dyn_size != 0) - dyn_size = pcpue_size - static_size - reserved_size; + pcpue_size = pcpu_calc_fc_sizes(static_size, reserved_size, &dyn_size); pcpue_unit_size = max_t(size_t, pcpue_size, PCPU_MIN_UNIT_SIZE); chunk_size = pcpue_unit_size * num_possible_cpus(); @@ -1390,6 +1400,197 @@ out_free_ar: return ret; } +/* + * Large page remapping first chunk setup helper + */ +#ifdef CONFIG_NEED_MULTIPLE_NODES +struct pcpul_ent { + unsigned int cpu; + void *ptr; +}; + +static size_t pcpul_size; +static size_t pcpul_unit_size; +static struct pcpul_ent *pcpul_map; +static struct vm_struct pcpul_vm; + +static struct page * __init pcpul_get_page(unsigned int cpu, int pageno) +{ + size_t off = (size_t)pageno << PAGE_SHIFT; + + if (off >= pcpul_size) + return NULL; + + return virt_to_page(pcpul_map[cpu].ptr + off); +} + +/** + * pcpu_lpage_first_chunk - remap the first percpu chunk using large page + * @static_size: the size of static percpu area in bytes + * @reserved_size: the size of reserved percpu area in bytes + * @dyn_size: free size for dynamic allocation in bytes, -1 for auto + * @lpage_size: the size of a large page + * @alloc_fn: function to allocate percpu lpage, always called with lpage_size + * @free_fn: function to free percpu memory, @size <= lpage_size + * @map_fn: function to map percpu lpage, always called with lpage_size + * + * This allocator uses large page as unit. A large page is allocated + * for each cpu and each is remapped into vmalloc area using large + * page mapping. As large page can be quite large, only part of it is + * used for the first chunk. Unused part is returned to the bootmem + * allocator. + * + * So, the large pages are mapped twice - once to the physical mapping + * and to the vmalloc area for the first percpu chunk. The double + * mapping does add one more large TLB entry pressure but still is + * much better than only using 4k mappings while still being NUMA + * friendly. + * + * RETURNS: + * The determined pcpu_unit_size which can be used to initialize + * percpu access on success, -errno on failure. + */ +ssize_t __init pcpu_lpage_first_chunk(size_t static_size, size_t reserved_size, + ssize_t dyn_size, size_t lpage_size, + pcpu_fc_alloc_fn_t alloc_fn, + pcpu_fc_free_fn_t free_fn, + pcpu_fc_map_fn_t map_fn) +{ + size_t size_sum; + size_t map_size; + unsigned int cpu; + int i, j; + ssize_t ret; + + /* + * Currently supports only single page. Supporting multiple + * pages won't be too difficult if it ever becomes necessary. + */ + size_sum = pcpu_calc_fc_sizes(static_size, reserved_size, &dyn_size); + + pcpul_unit_size = lpage_size; + pcpul_size = max_t(size_t, size_sum, PCPU_MIN_UNIT_SIZE); + if (pcpul_size > pcpul_unit_size) { + pr_warning("PERCPU: static data is larger than large page, " + "can't use large page\n"); + return -EINVAL; + } + + /* allocate pointer array and alloc large pages */ + map_size = PFN_ALIGN(num_possible_cpus() * sizeof(pcpul_map[0])); + pcpul_map = alloc_bootmem(map_size); + + for_each_possible_cpu(cpu) { + void *ptr; + + ptr = alloc_fn(cpu, lpage_size); + if (!ptr) { + pr_warning("PERCPU: failed to allocate large page " + "for cpu%u\n", cpu); + goto enomem; + } + + /* + * Only use pcpul_size bytes and give back the rest. + * + * Ingo: The lpage_size up-rounding bootmem is needed + * to make sure the partial lpage is still fully RAM - + * it's not well-specified to have a incompatible area + * (unmapped RAM, device memory, etc.) in that hole. + */ + free_fn(ptr + pcpul_size, lpage_size - pcpul_size); + + pcpul_map[cpu].cpu = cpu; + pcpul_map[cpu].ptr = ptr; + + memcpy(ptr, __per_cpu_load, static_size); + } + + /* allocate address and map */ + pcpul_vm.flags = VM_ALLOC; + pcpul_vm.size = num_possible_cpus() * pcpul_unit_size; + vm_area_register_early(&pcpul_vm, pcpul_unit_size); + + for_each_possible_cpu(cpu) + map_fn(pcpul_map[cpu].ptr, pcpul_unit_size, + pcpul_vm.addr + cpu * pcpul_unit_size); + + /* we're ready, commit */ + pr_info("PERCPU: Remapped at %p with large pages, static data " + "%zu bytes\n", pcpul_vm.addr, static_size); + + ret = pcpu_setup_first_chunk(pcpul_get_page, static_size, + reserved_size, dyn_size, pcpul_unit_size, + pcpul_vm.addr, NULL); + + /* sort pcpul_map array for pcpu_lpage_remapped() */ + for (i = 0; i < num_possible_cpus() - 1; i++) + for (j = i + 1; j < num_possible_cpus(); j++) + if (pcpul_map[i].ptr > pcpul_map[j].ptr) { + struct pcpul_ent tmp = pcpul_map[i]; + pcpul_map[i] = pcpul_map[j]; + pcpul_map[j] = tmp; + } + + return ret; + +enomem: + for_each_possible_cpu(cpu) + if (pcpul_map[cpu].ptr) + free_fn(pcpul_map[cpu].ptr, pcpul_size); + free_bootmem(__pa(pcpul_map), map_size); + return -ENOMEM; +} + +/** + * pcpu_lpage_remapped - determine whether a kaddr is in pcpul recycled area + * @kaddr: the kernel address in question + * + * Determine whether @kaddr falls in the pcpul recycled area. This is + * used by pageattr to detect VM aliases and break up the pcpu large + * page mapping such that the same physical page is not mapped under + * different attributes. + * + * The recycled area is always at the tail of a partially used large + * page. + * + * RETURNS: + * Address of corresponding remapped pcpu address if match is found; + * otherwise, NULL. + */ +void *pcpu_lpage_remapped(void *kaddr) +{ + unsigned long unit_mask = pcpul_unit_size - 1; + void *lpage_addr = (void *)((unsigned long)kaddr & ~unit_mask); + unsigned long offset = (unsigned long)kaddr & unit_mask; + int left = 0, right = num_possible_cpus() - 1; + int pos; + + /* pcpul in use at all? */ + if (!pcpul_map) + return NULL; + + /* okay, perform binary search */ + while (left <= right) { + pos = (left + right) / 2; + + if (pcpul_map[pos].ptr < lpage_addr) + left = pos + 1; + else if (pcpul_map[pos].ptr > lpage_addr) + right = pos - 1; + else { + /* it shouldn't be in the area for the first chunk */ + WARN_ON(offset < pcpul_size); + + return pcpul_vm.addr + + pcpul_map[pos].cpu * pcpul_unit_size + offset; + } + } + + return NULL; +} +#endif + /* * Generic percpu area setup. * -- cgit v1.2.3 From 38a6be525460f52ac6f2de1c3f73c5615a8853cd Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Sat, 4 Jul 2009 08:10:59 +0900 Subject: percpu: simplify pcpu_setup_first_chunk() Now that all first chunk allocator helpers allocate and map the first chunk themselves, there's no need to have optional default alloc/map in pcpu_setup_first_chunk(). Drop @populate_pte_fn and only leave @dyn_size optional and make all other params mandatory. This makes it much easier to follow what pcpu_setup_first_chunk() is doing and what actual differences tweaking each parameter results in. [ Impact: drop unused code path ] Signed-off-by: Tejun Heo Cc: Ingo Molnar --- arch/sparc/kernel/smp_64.c | 2 +- include/linux/percpu.h | 5 +-- mm/percpu.c | 104 +++++++++++++-------------------------------- 3 files changed, 33 insertions(+), 78 deletions(-) (limited to 'include') diff --git a/arch/sparc/kernel/smp_64.c b/arch/sparc/kernel/smp_64.c index fa44eaf8d897..ccad7b20ae75 100644 --- a/arch/sparc/kernel/smp_64.c +++ b/arch/sparc/kernel/smp_64.c @@ -1528,7 +1528,7 @@ void __init setup_per_cpu_areas(void) pcpu_unit_size = pcpu_setup_first_chunk(pcpur_get_page, static_size, PERCPU_MODULE_RESERVE, dyn_size, - PCPU_CHUNK_SIZE, vm.addr, NULL); + PCPU_CHUNK_SIZE, vm.addr); free_bootmem(__pa(pcpur_ptrs), ptrs_size); diff --git a/include/linux/percpu.h b/include/linux/percpu.h index 9f6bfd7d4b92..ec64357e1762 100644 --- a/include/linux/percpu.h +++ b/include/linux/percpu.h @@ -66,9 +66,8 @@ typedef void (*pcpu_fc_map_fn_t)(void *ptr, size_t size, void *addr); extern size_t __init pcpu_setup_first_chunk(pcpu_get_page_fn_t get_page_fn, size_t static_size, size_t reserved_size, - ssize_t dyn_size, ssize_t unit_size, - void *base_addr, - pcpu_fc_populate_pte_fn_t populate_pte_fn); + ssize_t dyn_size, size_t unit_size, + void *base_addr); extern ssize_t __init pcpu_embed_first_chunk( size_t static_size, size_t reserved_size, diff --git a/mm/percpu.c b/mm/percpu.c index 17db527ee2e2..21d938a10662 100644 --- a/mm/percpu.c +++ b/mm/percpu.c @@ -983,24 +983,22 @@ EXPORT_SYMBOL_GPL(free_percpu); * pcpu_setup_first_chunk - initialize the first percpu chunk * @get_page_fn: callback to fetch page pointer * @static_size: the size of static percpu area in bytes - * @reserved_size: the size of reserved percpu area in bytes + * @reserved_size: the size of reserved percpu area in bytes, 0 for none * @dyn_size: free size for dynamic allocation in bytes, -1 for auto - * @unit_size: unit size in bytes, must be multiple of PAGE_SIZE, -1 for auto - * @base_addr: mapped address, NULL for auto - * @populate_pte_fn: callback to allocate pagetable, NULL if unnecessary + * @unit_size: unit size in bytes, must be multiple of PAGE_SIZE + * @base_addr: mapped address * * Initialize the first percpu chunk which contains the kernel static * perpcu area. This function is to be called from arch percpu area - * setup path. The first two parameters are mandatory. The rest are - * optional. + * setup path. * * @get_page_fn() should return pointer to percpu page given cpu * number and page number. It should at least return enough pages to * cover the static area. The returned pages for static area should - * have been initialized with valid data. If @unit_size is specified, - * it can also return pages after the static area. NULL return - * indicates end of pages for the cpu. Note that @get_page_fn() must - * return the same number of pages for all cpus. + * have been initialized with valid data. It can also return pages + * after the static area. NULL return indicates end of pages for the + * cpu. Note that @get_page_fn() must return the same number of pages + * for all cpus. * * @reserved_size, if non-zero, specifies the amount of bytes to * reserve after the static area in the first chunk. This reserves @@ -1015,17 +1013,12 @@ EXPORT_SYMBOL_GPL(free_percpu); * non-negative value makes percpu leave alone the area beyond * @static_size + @reserved_size + @dyn_size. * - * @unit_size, if non-negative, specifies unit size and must be - * aligned to PAGE_SIZE and equal to or larger than @static_size + - * @reserved_size + if non-negative, @dyn_size. - * - * Non-null @base_addr means that the caller already allocated virtual - * region for the first chunk and mapped it. percpu must not mess - * with the chunk. Note that @base_addr with 0 @unit_size or non-NULL - * @populate_pte_fn doesn't make any sense. + * @unit_size specifies unit size and must be aligned to PAGE_SIZE and + * equal to or larger than @static_size + @reserved_size + if + * non-negative, @dyn_size. * - * @populate_pte_fn is used to populate the pagetable. NULL means the - * caller already populated the pagetable. + * The caller should have mapped the first chunk at @base_addr and + * copied static data to each unit. * * If the first chunk ends up with both reserved and dynamic areas, it * is served by two chunks - one to serve the core static and reserved @@ -1040,9 +1033,8 @@ EXPORT_SYMBOL_GPL(free_percpu); */ size_t __init pcpu_setup_first_chunk(pcpu_get_page_fn_t get_page_fn, size_t static_size, size_t reserved_size, - ssize_t dyn_size, ssize_t unit_size, - void *base_addr, - pcpu_fc_populate_pte_fn_t populate_pte_fn) + ssize_t dyn_size, size_t unit_size, + void *base_addr) { static struct vm_struct first_vm; static int smap[2], dmap[2]; @@ -1050,27 +1042,18 @@ size_t __init pcpu_setup_first_chunk(pcpu_get_page_fn_t get_page_fn, (dyn_size >= 0 ? dyn_size : 0); struct pcpu_chunk *schunk, *dchunk = NULL; unsigned int cpu; - int nr_pages; - int err, i; + int i, nr_pages; /* santiy checks */ BUILD_BUG_ON(ARRAY_SIZE(smap) >= PCPU_DFL_MAP_ALLOC || ARRAY_SIZE(dmap) >= PCPU_DFL_MAP_ALLOC); BUG_ON(!static_size); - if (unit_size >= 0) { - BUG_ON(unit_size < size_sum); - BUG_ON(unit_size & ~PAGE_MASK); - BUG_ON(unit_size < PCPU_MIN_UNIT_SIZE); - } else - BUG_ON(base_addr); - BUG_ON(base_addr && populate_pte_fn); - - if (unit_size >= 0) - pcpu_unit_pages = unit_size >> PAGE_SHIFT; - else - pcpu_unit_pages = max_t(int, PCPU_MIN_UNIT_SIZE >> PAGE_SHIFT, - PFN_UP(size_sum)); + BUG_ON(!base_addr); + BUG_ON(unit_size < size_sum); + BUG_ON(unit_size & ~PAGE_MASK); + BUG_ON(unit_size < PCPU_MIN_UNIT_SIZE); + pcpu_unit_pages = unit_size >> PAGE_SHIFT; pcpu_unit_size = pcpu_unit_pages << PAGE_SHIFT; pcpu_chunk_size = num_possible_cpus() * pcpu_unit_size; pcpu_chunk_struct_size = sizeof(struct pcpu_chunk) @@ -1079,6 +1062,10 @@ size_t __init pcpu_setup_first_chunk(pcpu_get_page_fn_t get_page_fn, if (dyn_size < 0) dyn_size = pcpu_unit_size - static_size - reserved_size; + first_vm.flags = VM_ALLOC; + first_vm.size = pcpu_chunk_size; + first_vm.addr = base_addr; + /* * Allocate chunk slots. The additional last slot is for * empty chunks. @@ -1101,6 +1088,7 @@ size_t __init pcpu_setup_first_chunk(pcpu_get_page_fn_t get_page_fn, schunk->map = smap; schunk->map_alloc = ARRAY_SIZE(smap); schunk->page = schunk->page_ar; + schunk->immutable = true; if (reserved_size) { schunk->free_size = reserved_size; @@ -1124,31 +1112,13 @@ size_t __init pcpu_setup_first_chunk(pcpu_get_page_fn_t get_page_fn, dchunk->map = dmap; dchunk->map_alloc = ARRAY_SIZE(dmap); dchunk->page = schunk->page_ar; /* share page map with schunk */ + dchunk->immutable = true; dchunk->contig_hint = dchunk->free_size = dyn_size; dchunk->map[dchunk->map_used++] = -pcpu_reserved_chunk_limit; dchunk->map[dchunk->map_used++] = dchunk->free_size; } - /* allocate vm address */ - first_vm.flags = VM_ALLOC; - first_vm.size = pcpu_chunk_size; - - if (!base_addr) - vm_area_register_early(&first_vm, PAGE_SIZE); - else { - /* - * Pages already mapped. No need to remap into - * vmalloc area. In this case the first chunks can't - * be mapped or unmapped by percpu and are marked - * immutable. - */ - first_vm.addr = base_addr; - schunk->immutable = true; - if (dchunk) - dchunk->immutable = true; - } - /* assign pages */ nr_pages = -1; for_each_possible_cpu(cpu) { @@ -1168,19 +1138,6 @@ size_t __init pcpu_setup_first_chunk(pcpu_get_page_fn_t get_page_fn, BUG_ON(nr_pages != i); } - /* map them */ - if (populate_pte_fn) { - for_each_possible_cpu(cpu) - for (i = 0; i < nr_pages; i++) - populate_pte_fn(pcpu_chunk_addr(schunk, - cpu, i)); - - err = pcpu_map(schunk, 0, nr_pages); - if (err) - panic("failed to setup static percpu area, err=%d\n", - err); - } - /* link the first chunk in */ pcpu_first_chunk = dchunk ?: schunk; pcpu_chunk_relocate(pcpu_first_chunk, -1); @@ -1282,7 +1239,7 @@ ssize_t __init pcpu_embed_first_chunk(size_t static_size, size_t reserved_size, return pcpu_setup_first_chunk(pcpue_get_page, static_size, reserved_size, dyn_size, - pcpue_unit_size, pcpue_ptr, NULL); + pcpue_unit_size, pcpue_ptr); } /* @@ -1387,8 +1344,7 @@ ssize_t __init pcpu_4k_first_chunk(size_t static_size, size_t reserved_size, ret = pcpu_setup_first_chunk(pcpu4k_get_page, static_size, reserved_size, -1, - pcpu4k_unit_pages << PAGE_SHIFT, vm.addr, - NULL); + pcpu4k_unit_pages << PAGE_SHIFT, vm.addr); goto out_free_ar; enomem: @@ -1521,7 +1477,7 @@ ssize_t __init pcpu_lpage_first_chunk(size_t static_size, size_t reserved_size, ret = pcpu_setup_first_chunk(pcpul_get_page, static_size, reserved_size, dyn_size, pcpul_unit_size, - pcpul_vm.addr, NULL); + pcpul_vm.addr); /* sort pcpul_map array for pcpu_lpage_remapped() */ for (i = 0; i < num_possible_cpus() - 1; i++) -- cgit v1.2.3 From ce3141a277ff6cc37e51008b8888dc2cb7456ef1 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Sat, 4 Jul 2009 08:11:00 +0900 Subject: percpu: drop pcpu_chunk->page[] percpu core doesn't need to tack all the allocated pages. It needs to know whether certain pages are populated and a way to reverse map address to page when freeing. This patch drops pcpu_chunk->page[] and use populated bitmap and vmalloc_to_page() lookup instead. Using vmalloc_to_page() exclusively is also possible but complicates first chunk handling, inflates cache footprint and prevents non-standard memory allocation for percpu memory. pcpu_chunk->page[] was used to track each page's allocation and allowed asymmetric population which happens during failure path; however, with single bitmap for all units, this is no longer possible. Bite the bullet and rewrite (de)populate functions so that things are done in clearly separated steps such that asymmetric population doesn't happen. This makes the (de)population process much more modular and will also ease implementing non-standard memory usage in the future (e.g. large pages). This makes @get_page_fn parameter to pcpu_setup_first_chunk() unnecessary. The parameter is dropped and all first chunk helpers are updated accordingly. Please note that despite the volume most changes to first chunk helpers are symbol renames for variables which don't need to be referenced outside of the helper anymore. This change reduces memory usage and cache footprint of pcpu_chunk. Now only #unit_pages bits are necessary per chunk. [ Impact: reduced memory usage and cache footprint for bookkeeping ] Signed-off-by: Tejun Heo Cc: Ingo Molnar Cc: David Miller --- arch/sparc/kernel/smp_64.c | 42 ++-- include/linux/percpu.h | 3 +- mm/percpu.c | 604 ++++++++++++++++++++++++++++----------------- 3 files changed, 400 insertions(+), 249 deletions(-) (limited to 'include') diff --git a/arch/sparc/kernel/smp_64.c b/arch/sparc/kernel/smp_64.c index ccad7b20ae75..f2f22ee97a7a 100644 --- a/arch/sparc/kernel/smp_64.c +++ b/arch/sparc/kernel/smp_64.c @@ -1415,19 +1415,6 @@ static void * __init pcpu_alloc_bootmem(unsigned int cpu, unsigned long size, #endif } -static size_t pcpur_size __initdata; -static void **pcpur_ptrs __initdata; - -static struct page * __init pcpur_get_page(unsigned int cpu, int pageno) -{ - size_t off = (size_t)pageno << PAGE_SHIFT; - - if (off >= pcpur_size) - return NULL; - - return virt_to_page(pcpur_ptrs[cpu] + off); -} - #define PCPU_CHUNK_SIZE (4UL * 1024UL * 1024UL) static void __init pcpu_map_range(unsigned long start, unsigned long end, @@ -1491,25 +1478,26 @@ void __init setup_per_cpu_areas(void) size_t dyn_size, static_size = __per_cpu_end - __per_cpu_start; static struct vm_struct vm; unsigned long delta, cpu; - size_t pcpu_unit_size; + size_t size_sum, pcpu_unit_size; size_t ptrs_size; + void **ptrs; - pcpur_size = PFN_ALIGN(static_size + PERCPU_MODULE_RESERVE + - PERCPU_DYNAMIC_RESERVE); - dyn_size = pcpur_size - static_size - PERCPU_MODULE_RESERVE; + size_sum = PFN_ALIGN(static_size + PERCPU_MODULE_RESERVE + + PERCPU_DYNAMIC_RESERVE); + dyn_size = size_sum - static_size - PERCPU_MODULE_RESERVE; - ptrs_size = PFN_ALIGN(num_possible_cpus() * sizeof(pcpur_ptrs[0])); - pcpur_ptrs = alloc_bootmem(ptrs_size); + ptrs_size = PFN_ALIGN(num_possible_cpus() * sizeof(ptrs[0])); + ptrs = alloc_bootmem(ptrs_size); for_each_possible_cpu(cpu) { - pcpur_ptrs[cpu] = pcpu_alloc_bootmem(cpu, PCPU_CHUNK_SIZE, - PCPU_CHUNK_SIZE); + ptrs[cpu] = pcpu_alloc_bootmem(cpu, PCPU_CHUNK_SIZE, + PCPU_CHUNK_SIZE); - free_bootmem(__pa(pcpur_ptrs[cpu] + pcpur_size), - PCPU_CHUNK_SIZE - pcpur_size); + free_bootmem(__pa(ptrs[cpu] + size_sum), + PCPU_CHUNK_SIZE - size_sum); - memcpy(pcpur_ptrs[cpu], __per_cpu_load, static_size); + memcpy(ptrs[cpu], __per_cpu_load, static_size); } /* allocate address and map */ @@ -1523,14 +1511,14 @@ void __init setup_per_cpu_areas(void) start += cpu * PCPU_CHUNK_SIZE; end = start + PCPU_CHUNK_SIZE; - pcpu_map_range(start, end, virt_to_page(pcpur_ptrs[cpu])); + pcpu_map_range(start, end, virt_to_page(ptrs[cpu])); } - pcpu_unit_size = pcpu_setup_first_chunk(pcpur_get_page, static_size, + pcpu_unit_size = pcpu_setup_first_chunk(static_size, PERCPU_MODULE_RESERVE, dyn_size, PCPU_CHUNK_SIZE, vm.addr); - free_bootmem(__pa(pcpur_ptrs), ptrs_size); + free_bootmem(__pa(ptrs), ptrs_size); delta = (unsigned long)pcpu_base_addr - (unsigned long)__per_cpu_start; for_each_possible_cpu(cpu) { diff --git a/include/linux/percpu.h b/include/linux/percpu.h index ec64357e1762..63c8b7a23e66 100644 --- a/include/linux/percpu.h +++ b/include/linux/percpu.h @@ -58,13 +58,12 @@ extern void *pcpu_base_addr; -typedef struct page * (*pcpu_get_page_fn_t)(unsigned int cpu, int pageno); typedef void * (*pcpu_fc_alloc_fn_t)(unsigned int cpu, size_t size); typedef void (*pcpu_fc_free_fn_t)(void *ptr, size_t size); typedef void (*pcpu_fc_populate_pte_fn_t)(unsigned long addr); typedef void (*pcpu_fc_map_fn_t)(void *ptr, size_t size, void *addr); -extern size_t __init pcpu_setup_first_chunk(pcpu_get_page_fn_t get_page_fn, +extern size_t __init pcpu_setup_first_chunk( size_t static_size, size_t reserved_size, ssize_t dyn_size, size_t unit_size, void *base_addr); diff --git a/mm/percpu.c b/mm/percpu.c index 639fce4d2caf..21756814d99f 100644 --- a/mm/percpu.c +++ b/mm/percpu.c @@ -94,8 +94,7 @@ struct pcpu_chunk { int map_alloc; /* # of map entries allocated */ int *map; /* allocation map */ bool immutable; /* no [de]population allowed */ - struct page **page; /* points to page array */ - struct page *page_ar[]; /* #cpus * UNIT_PAGES */ + unsigned long populated[]; /* populated bitmap */ }; static int pcpu_unit_pages __read_mostly; @@ -129,9 +128,9 @@ static int pcpu_reserved_chunk_limit; * Synchronization rules. * * There are two locks - pcpu_alloc_mutex and pcpu_lock. The former - * protects allocation/reclaim paths, chunks and chunk->page arrays. - * The latter is a spinlock and protects the index data structures - - * chunk slots, chunks and area maps in chunks. + * protects allocation/reclaim paths, chunks, populated bitmap and + * vmalloc mapping. The latter is a spinlock and protects the index + * data structures - chunk slots, chunks and area maps in chunks. * * During allocation, pcpu_alloc_mutex is kept locked all the time and * pcpu_lock is grabbed and released as necessary. All actual memory @@ -188,16 +187,13 @@ static unsigned long pcpu_chunk_addr(struct pcpu_chunk *chunk, (pcpu_page_idx(cpu, page_idx) << PAGE_SHIFT); } -static struct page **pcpu_chunk_pagep(struct pcpu_chunk *chunk, - unsigned int cpu, int page_idx) +static struct page *pcpu_chunk_page(struct pcpu_chunk *chunk, + unsigned int cpu, int page_idx) { - return &chunk->page[pcpu_page_idx(cpu, page_idx)]; -} + /* must not be used on pre-mapped chunk */ + WARN_ON(chunk->immutable); -static bool pcpu_chunk_page_occupied(struct pcpu_chunk *chunk, - int page_idx) -{ - return *pcpu_chunk_pagep(chunk, 0, page_idx) != NULL; + return vmalloc_to_page((void *)pcpu_chunk_addr(chunk, cpu, page_idx)); } /* set the pointer to a chunk in a page struct */ @@ -212,6 +208,34 @@ static struct pcpu_chunk *pcpu_get_page_chunk(struct page *page) return (struct pcpu_chunk *)page->index; } +static void pcpu_next_unpop(struct pcpu_chunk *chunk, int *rs, int *re, int end) +{ + *rs = find_next_zero_bit(chunk->populated, end, *rs); + *re = find_next_bit(chunk->populated, end, *rs + 1); +} + +static void pcpu_next_pop(struct pcpu_chunk *chunk, int *rs, int *re, int end) +{ + *rs = find_next_bit(chunk->populated, end, *rs); + *re = find_next_zero_bit(chunk->populated, end, *rs + 1); +} + +/* + * (Un)populated page region iterators. Iterate over (un)populated + * page regions betwen @start and @end in @chunk. @rs and @re should + * be integer variables and will be set to start and end page index of + * the current region. + */ +#define pcpu_for_each_unpop_region(chunk, rs, re, start, end) \ + for ((rs) = (start), pcpu_next_unpop((chunk), &(rs), &(re), (end)); \ + (rs) < (re); \ + (rs) = (re) + 1, pcpu_next_unpop((chunk), &(rs), &(re), (end))) + +#define pcpu_for_each_pop_region(chunk, rs, re, start, end) \ + for ((rs) = (start), pcpu_next_pop((chunk), &(rs), &(re), (end)); \ + (rs) < (re); \ + (rs) = (re) + 1, pcpu_next_pop((chunk), &(rs), &(re), (end))) + /** * pcpu_mem_alloc - allocate memory * @size: bytes to allocate @@ -545,42 +569,197 @@ static void pcpu_free_area(struct pcpu_chunk *chunk, int freeme) } /** - * pcpu_unmap - unmap pages out of a pcpu_chunk + * pcpu_get_pages_and_bitmap - get temp pages array and bitmap + * @chunk: chunk of interest + * @bitmapp: output parameter for bitmap + * @may_alloc: may allocate the array + * + * Returns pointer to array of pointers to struct page and bitmap, + * both of which can be indexed with pcpu_page_idx(). The returned + * array is cleared to zero and *@bitmapp is copied from + * @chunk->populated. Note that there is only one array and bitmap + * and access exclusion is the caller's responsibility. + * + * CONTEXT: + * pcpu_alloc_mutex and does GFP_KERNEL allocation if @may_alloc. + * Otherwise, don't care. + * + * RETURNS: + * Pointer to temp pages array on success, NULL on failure. + */ +static struct page **pcpu_get_pages_and_bitmap(struct pcpu_chunk *chunk, + unsigned long **bitmapp, + bool may_alloc) +{ + static struct page **pages; + static unsigned long *bitmap; + size_t pages_size = num_possible_cpus() * pcpu_unit_pages * + sizeof(pages[0]); + size_t bitmap_size = BITS_TO_LONGS(pcpu_unit_pages) * + sizeof(unsigned long); + + if (!pages || !bitmap) { + if (may_alloc && !pages) + pages = pcpu_mem_alloc(pages_size); + if (may_alloc && !bitmap) + bitmap = pcpu_mem_alloc(bitmap_size); + if (!pages || !bitmap) + return NULL; + } + + memset(pages, 0, pages_size); + bitmap_copy(bitmap, chunk->populated, pcpu_unit_pages); + + *bitmapp = bitmap; + return pages; +} + +/** + * pcpu_free_pages - free pages which were allocated for @chunk + * @chunk: chunk pages were allocated for + * @pages: array of pages to be freed, indexed by pcpu_page_idx() + * @populated: populated bitmap + * @page_start: page index of the first page to be freed + * @page_end: page index of the last page to be freed + 1 + * + * Free pages [@page_start and @page_end) in @pages for all units. + * The pages were allocated for @chunk. + */ +static void pcpu_free_pages(struct pcpu_chunk *chunk, + struct page **pages, unsigned long *populated, + int page_start, int page_end) +{ + unsigned int cpu; + int i; + + for_each_possible_cpu(cpu) { + for (i = page_start; i < page_end; i++) { + struct page *page = pages[pcpu_page_idx(cpu, i)]; + + if (page) + __free_page(page); + } + } +} + +/** + * pcpu_alloc_pages - allocates pages for @chunk + * @chunk: target chunk + * @pages: array to put the allocated pages into, indexed by pcpu_page_idx() + * @populated: populated bitmap + * @page_start: page index of the first page to be allocated + * @page_end: page index of the last page to be allocated + 1 + * + * Allocate pages [@page_start,@page_end) into @pages for all units. + * The allocation is for @chunk. Percpu core doesn't care about the + * content of @pages and will pass it verbatim to pcpu_map_pages(). + */ +static int pcpu_alloc_pages(struct pcpu_chunk *chunk, + struct page **pages, unsigned long *populated, + int page_start, int page_end) +{ + const gfp_t gfp = GFP_KERNEL | __GFP_HIGHMEM | __GFP_COLD; + unsigned int cpu; + int i; + + for_each_possible_cpu(cpu) { + for (i = page_start; i < page_end; i++) { + struct page **pagep = &pages[pcpu_page_idx(cpu, i)]; + + *pagep = alloc_pages_node(cpu_to_node(cpu), gfp, 0); + if (!*pagep) { + pcpu_free_pages(chunk, pages, populated, + page_start, page_end); + return -ENOMEM; + } + } + } + return 0; +} + +/** + * pcpu_pre_unmap_flush - flush cache prior to unmapping + * @chunk: chunk the regions to be flushed belongs to + * @page_start: page index of the first page to be flushed + * @page_end: page index of the last page to be flushed + 1 + * + * Pages in [@page_start,@page_end) of @chunk are about to be + * unmapped. Flush cache. As each flushing trial can be very + * expensive, issue flush on the whole region at once rather than + * doing it for each cpu. This could be an overkill but is more + * scalable. + */ +static void pcpu_pre_unmap_flush(struct pcpu_chunk *chunk, + int page_start, int page_end) +{ + unsigned int last = num_possible_cpus() - 1; + + flush_cache_vunmap(pcpu_chunk_addr(chunk, 0, page_start), + pcpu_chunk_addr(chunk, last, page_end)); +} + +static void __pcpu_unmap_pages(unsigned long addr, int nr_pages) +{ + unmap_kernel_range_noflush(addr, nr_pages << PAGE_SHIFT); +} + +/** + * pcpu_unmap_pages - unmap pages out of a pcpu_chunk * @chunk: chunk of interest + * @pages: pages array which can be used to pass information to free + * @populated: populated bitmap * @page_start: page index of the first page to unmap * @page_end: page index of the last page to unmap + 1 - * @flush_tlb: whether to flush tlb or not * * For each cpu, unmap pages [@page_start,@page_end) out of @chunk. - * If @flush is true, vcache is flushed before unmapping and tlb - * after. + * Corresponding elements in @pages were cleared by the caller and can + * be used to carry information to pcpu_free_pages() which will be + * called after all unmaps are finished. The caller should call + * proper pre/post flush functions. */ -static void pcpu_unmap(struct pcpu_chunk *chunk, int page_start, int page_end, - bool flush_tlb) +static void pcpu_unmap_pages(struct pcpu_chunk *chunk, + struct page **pages, unsigned long *populated, + int page_start, int page_end) { - unsigned int last = num_possible_cpus() - 1; unsigned int cpu; + int i; - /* unmap must not be done on immutable chunk */ - WARN_ON(chunk->immutable); + for_each_possible_cpu(cpu) { + for (i = page_start; i < page_end; i++) { + struct page *page; - /* - * Each flushing trial can be very expensive, issue flush on - * the whole region at once rather than doing it for each cpu. - * This could be an overkill but is more scalable. - */ - flush_cache_vunmap(pcpu_chunk_addr(chunk, 0, page_start), - pcpu_chunk_addr(chunk, last, page_end)); + page = pcpu_chunk_page(chunk, cpu, i); + WARN_ON(!page); + pages[pcpu_page_idx(cpu, i)] = page; + } + __pcpu_unmap_pages(pcpu_chunk_addr(chunk, cpu, page_start), + page_end - page_start); + } - for_each_possible_cpu(cpu) - unmap_kernel_range_noflush( - pcpu_chunk_addr(chunk, cpu, page_start), - (page_end - page_start) << PAGE_SHIFT); - - /* ditto as flush_cache_vunmap() */ - if (flush_tlb) - flush_tlb_kernel_range(pcpu_chunk_addr(chunk, 0, page_start), - pcpu_chunk_addr(chunk, last, page_end)); + for (i = page_start; i < page_end; i++) + __clear_bit(i, populated); +} + +/** + * pcpu_post_unmap_tlb_flush - flush TLB after unmapping + * @chunk: pcpu_chunk the regions to be flushed belong to + * @page_start: page index of the first page to be flushed + * @page_end: page index of the last page to be flushed + 1 + * + * Pages [@page_start,@page_end) of @chunk have been unmapped. Flush + * TLB for the regions. This can be skipped if the area is to be + * returned to vmalloc as vmalloc will handle TLB flushing lazily. + * + * As with pcpu_pre_unmap_flush(), TLB flushing also is done at once + * for the whole region. + */ +static void pcpu_post_unmap_tlb_flush(struct pcpu_chunk *chunk, + int page_start, int page_end) +{ + unsigned int last = num_possible_cpus() - 1; + + flush_tlb_kernel_range(pcpu_chunk_addr(chunk, 0, page_start), + pcpu_chunk_addr(chunk, last, page_end)); } static int __pcpu_map_pages(unsigned long addr, struct page **pages, @@ -591,35 +770,76 @@ static int __pcpu_map_pages(unsigned long addr, struct page **pages, } /** - * pcpu_map - map pages into a pcpu_chunk + * pcpu_map_pages - map pages into a pcpu_chunk * @chunk: chunk of interest + * @pages: pages array containing pages to be mapped + * @populated: populated bitmap * @page_start: page index of the first page to map * @page_end: page index of the last page to map + 1 * - * For each cpu, map pages [@page_start,@page_end) into @chunk. - * vcache is flushed afterwards. + * For each cpu, map pages [@page_start,@page_end) into @chunk. The + * caller is responsible for calling pcpu_post_map_flush() after all + * mappings are complete. + * + * This function is responsible for setting corresponding bits in + * @chunk->populated bitmap and whatever is necessary for reverse + * lookup (addr -> chunk). */ -static int pcpu_map(struct pcpu_chunk *chunk, int page_start, int page_end) +static int pcpu_map_pages(struct pcpu_chunk *chunk, + struct page **pages, unsigned long *populated, + int page_start, int page_end) { - unsigned int last = num_possible_cpus() - 1; - unsigned int cpu; - int err; - - /* map must not be done on immutable chunk */ - WARN_ON(chunk->immutable); + unsigned int cpu, tcpu; + int i, err; for_each_possible_cpu(cpu) { err = __pcpu_map_pages(pcpu_chunk_addr(chunk, cpu, page_start), - pcpu_chunk_pagep(chunk, cpu, page_start), + &pages[pcpu_page_idx(cpu, page_start)], page_end - page_start); if (err < 0) - return err; + goto err; } + /* mapping successful, link chunk and mark populated */ + for (i = page_start; i < page_end; i++) { + for_each_possible_cpu(cpu) + pcpu_set_page_chunk(pages[pcpu_page_idx(cpu, i)], + chunk); + __set_bit(i, populated); + } + + return 0; + +err: + for_each_possible_cpu(tcpu) { + if (tcpu == cpu) + break; + __pcpu_unmap_pages(pcpu_chunk_addr(chunk, tcpu, page_start), + page_end - page_start); + } + return err; +} + +/** + * pcpu_post_map_flush - flush cache after mapping + * @chunk: pcpu_chunk the regions to be flushed belong to + * @page_start: page index of the first page to be flushed + * @page_end: page index of the last page to be flushed + 1 + * + * Pages [@page_start,@page_end) of @chunk have been mapped. Flush + * cache. + * + * As with pcpu_pre_unmap_flush(), TLB flushing also is done at once + * for the whole region. + */ +static void pcpu_post_map_flush(struct pcpu_chunk *chunk, + int page_start, int page_end) +{ + unsigned int last = num_possible_cpus() - 1; + /* flush at once, please read comments in pcpu_unmap() */ flush_cache_vmap(pcpu_chunk_addr(chunk, 0, page_start), pcpu_chunk_addr(chunk, last, page_end)); - return 0; } /** @@ -636,39 +856,45 @@ static int pcpu_map(struct pcpu_chunk *chunk, int page_start, int page_end) * CONTEXT: * pcpu_alloc_mutex. */ -static void pcpu_depopulate_chunk(struct pcpu_chunk *chunk, int off, int size, - bool flush) +static void pcpu_depopulate_chunk(struct pcpu_chunk *chunk, int off, int size) { int page_start = PFN_DOWN(off); int page_end = PFN_UP(off + size); - int unmap_start = -1; - int uninitialized_var(unmap_end); - unsigned int cpu; - int i; + struct page **pages; + unsigned long *populated; + int rs, re; + + /* quick path, check whether it's empty already */ + pcpu_for_each_unpop_region(chunk, rs, re, page_start, page_end) { + if (rs == page_start && re == page_end) + return; + break; + } - for (i = page_start; i < page_end; i++) { - for_each_possible_cpu(cpu) { - struct page **pagep = pcpu_chunk_pagep(chunk, cpu, i); + /* immutable chunks can't be depopulated */ + WARN_ON(chunk->immutable); - if (!*pagep) - continue; + /* + * If control reaches here, there must have been at least one + * successful population attempt so the temp pages array must + * be available now. + */ + pages = pcpu_get_pages_and_bitmap(chunk, &populated, false); + BUG_ON(!pages); - __free_page(*pagep); + /* unmap and free */ + pcpu_pre_unmap_flush(chunk, page_start, page_end); - /* - * If it's partial depopulation, it might get - * populated or depopulated again. Mark the - * page gone. - */ - *pagep = NULL; + pcpu_for_each_pop_region(chunk, rs, re, page_start, page_end) + pcpu_unmap_pages(chunk, pages, populated, rs, re); - unmap_start = unmap_start < 0 ? i : unmap_start; - unmap_end = i + 1; - } - } + /* no need to flush tlb, vmalloc will handle it lazily */ + + pcpu_for_each_pop_region(chunk, rs, re, page_start, page_end) + pcpu_free_pages(chunk, pages, populated, rs, re); - if (unmap_start >= 0) - pcpu_unmap(chunk, unmap_start, unmap_end, flush); + /* commit new bitmap */ + bitmap_copy(chunk->populated, populated, pcpu_unit_pages); } /** @@ -685,50 +911,61 @@ static void pcpu_depopulate_chunk(struct pcpu_chunk *chunk, int off, int size, */ static int pcpu_populate_chunk(struct pcpu_chunk *chunk, int off, int size) { - const gfp_t alloc_mask = GFP_KERNEL | __GFP_HIGHMEM | __GFP_COLD; int page_start = PFN_DOWN(off); int page_end = PFN_UP(off + size); - int map_start = -1; - int uninitialized_var(map_end); + int free_end = page_start, unmap_end = page_start; + struct page **pages; + unsigned long *populated; unsigned int cpu; - int i; + int rs, re, rc; - for (i = page_start; i < page_end; i++) { - if (pcpu_chunk_page_occupied(chunk, i)) { - if (map_start >= 0) { - if (pcpu_map(chunk, map_start, map_end)) - goto err; - map_start = -1; - } - continue; - } + /* quick path, check whether all pages are already there */ + pcpu_for_each_pop_region(chunk, rs, re, page_start, page_end) { + if (rs == page_start && re == page_end) + goto clear; + break; + } - map_start = map_start < 0 ? i : map_start; - map_end = i + 1; + /* need to allocate and map pages, this chunk can't be immutable */ + WARN_ON(chunk->immutable); - for_each_possible_cpu(cpu) { - struct page **pagep = pcpu_chunk_pagep(chunk, cpu, i); + pages = pcpu_get_pages_and_bitmap(chunk, &populated, true); + if (!pages) + return -ENOMEM; - *pagep = alloc_pages_node(cpu_to_node(cpu), - alloc_mask, 0); - if (!*pagep) - goto err; - pcpu_set_page_chunk(*pagep, chunk); - } + /* alloc and map */ + pcpu_for_each_unpop_region(chunk, rs, re, page_start, page_end) { + rc = pcpu_alloc_pages(chunk, pages, populated, rs, re); + if (rc) + goto err_free; + free_end = re; } - if (map_start >= 0 && pcpu_map(chunk, map_start, map_end)) - goto err; + pcpu_for_each_unpop_region(chunk, rs, re, page_start, page_end) { + rc = pcpu_map_pages(chunk, pages, populated, rs, re); + if (rc) + goto err_unmap; + unmap_end = re; + } + pcpu_post_map_flush(chunk, page_start, page_end); + /* commit new bitmap */ + bitmap_copy(chunk->populated, populated, pcpu_unit_pages); +clear: for_each_possible_cpu(cpu) memset(chunk->vm->addr + cpu * pcpu_unit_size + off, 0, size); - return 0; -err: - /* likely under heavy memory pressure, give memory back */ - pcpu_depopulate_chunk(chunk, off, size, true); - return -ENOMEM; + +err_unmap: + pcpu_pre_unmap_flush(chunk, page_start, unmap_end); + pcpu_for_each_unpop_region(chunk, rs, re, page_start, unmap_end) + pcpu_unmap_pages(chunk, pages, populated, rs, re); + pcpu_post_unmap_tlb_flush(chunk, page_start, unmap_end); +err_free: + pcpu_for_each_unpop_region(chunk, rs, re, page_start, free_end) + pcpu_free_pages(chunk, pages, populated, rs, re); + return rc; } static void free_pcpu_chunk(struct pcpu_chunk *chunk) @@ -752,7 +989,6 @@ static struct pcpu_chunk *alloc_pcpu_chunk(void) chunk->map = pcpu_mem_alloc(PCPU_DFL_MAP_ALLOC * sizeof(chunk->map[0])); chunk->map_alloc = PCPU_DFL_MAP_ALLOC; chunk->map[chunk->map_used++] = pcpu_unit_size; - chunk->page = chunk->page_ar; chunk->vm = get_vm_area(pcpu_chunk_size, GFP_KERNEL); if (!chunk->vm) { @@ -933,7 +1169,7 @@ static void pcpu_reclaim(struct work_struct *work) mutex_unlock(&pcpu_alloc_mutex); list_for_each_entry_safe(chunk, next, &todo, list) { - pcpu_depopulate_chunk(chunk, 0, pcpu_unit_size, false); + pcpu_depopulate_chunk(chunk, 0, pcpu_unit_size); free_pcpu_chunk(chunk); } } @@ -981,7 +1217,6 @@ EXPORT_SYMBOL_GPL(free_percpu); /** * pcpu_setup_first_chunk - initialize the first percpu chunk - * @get_page_fn: callback to fetch page pointer * @static_size: the size of static percpu area in bytes * @reserved_size: the size of reserved percpu area in bytes, 0 for none * @dyn_size: free size for dynamic allocation in bytes, -1 for auto @@ -992,14 +1227,6 @@ EXPORT_SYMBOL_GPL(free_percpu); * perpcu area. This function is to be called from arch percpu area * setup path. * - * @get_page_fn() should return pointer to percpu page given cpu - * number and page number. It should at least return enough pages to - * cover the static area. The returned pages for static area should - * have been initialized with valid data. It can also return pages - * after the static area. NULL return indicates end of pages for the - * cpu. Note that @get_page_fn() must return the same number of pages - * for all cpus. - * * @reserved_size, if non-zero, specifies the amount of bytes to * reserve after the static area in the first chunk. This reserves * the first chunk such that it's available only through reserved @@ -1031,8 +1258,7 @@ EXPORT_SYMBOL_GPL(free_percpu); * The determined pcpu_unit_size which can be used to initialize * percpu access. */ -size_t __init pcpu_setup_first_chunk(pcpu_get_page_fn_t get_page_fn, - size_t static_size, size_t reserved_size, +size_t __init pcpu_setup_first_chunk(size_t static_size, size_t reserved_size, ssize_t dyn_size, size_t unit_size, void *base_addr) { @@ -1041,8 +1267,7 @@ size_t __init pcpu_setup_first_chunk(pcpu_get_page_fn_t get_page_fn, size_t size_sum = static_size + reserved_size + (dyn_size >= 0 ? dyn_size : 0); struct pcpu_chunk *schunk, *dchunk = NULL; - unsigned int cpu; - int i, nr_pages; + int i; /* santiy checks */ BUILD_BUG_ON(ARRAY_SIZE(smap) >= PCPU_DFL_MAP_ALLOC || @@ -1056,8 +1281,8 @@ size_t __init pcpu_setup_first_chunk(pcpu_get_page_fn_t get_page_fn, pcpu_unit_pages = unit_size >> PAGE_SHIFT; pcpu_unit_size = pcpu_unit_pages << PAGE_SHIFT; pcpu_chunk_size = num_possible_cpus() * pcpu_unit_size; - pcpu_chunk_struct_size = sizeof(struct pcpu_chunk) - + num_possible_cpus() * pcpu_unit_pages * sizeof(struct page *); + pcpu_chunk_struct_size = sizeof(struct pcpu_chunk) + + BITS_TO_LONGS(pcpu_unit_pages) * sizeof(unsigned long); if (dyn_size < 0) dyn_size = pcpu_unit_size - static_size - reserved_size; @@ -1087,8 +1312,8 @@ size_t __init pcpu_setup_first_chunk(pcpu_get_page_fn_t get_page_fn, schunk->vm = &first_vm; schunk->map = smap; schunk->map_alloc = ARRAY_SIZE(smap); - schunk->page = schunk->page_ar; schunk->immutable = true; + bitmap_fill(schunk->populated, pcpu_unit_pages); if (reserved_size) { schunk->free_size = reserved_size; @@ -1106,38 +1331,19 @@ size_t __init pcpu_setup_first_chunk(pcpu_get_page_fn_t get_page_fn, /* init dynamic chunk if necessary */ if (dyn_size) { - dchunk = alloc_bootmem(sizeof(struct pcpu_chunk)); + dchunk = alloc_bootmem(pcpu_chunk_struct_size); INIT_LIST_HEAD(&dchunk->list); dchunk->vm = &first_vm; dchunk->map = dmap; dchunk->map_alloc = ARRAY_SIZE(dmap); - dchunk->page = schunk->page_ar; /* share page map with schunk */ dchunk->immutable = true; + bitmap_fill(dchunk->populated, pcpu_unit_pages); dchunk->contig_hint = dchunk->free_size = dyn_size; dchunk->map[dchunk->map_used++] = -pcpu_reserved_chunk_limit; dchunk->map[dchunk->map_used++] = dchunk->free_size; } - /* assign pages */ - nr_pages = -1; - for_each_possible_cpu(cpu) { - for (i = 0; i < pcpu_unit_pages; i++) { - struct page *page = get_page_fn(cpu, i); - - if (!page) - break; - *pcpu_chunk_pagep(schunk, cpu, i) = page; - } - - BUG_ON(i < PFN_UP(static_size)); - - if (nr_pages < 0) - nr_pages = i; - else - BUG_ON(nr_pages != i); - } - /* link the first chunk in */ pcpu_first_chunk = dchunk ?: schunk; pcpu_chunk_relocate(pcpu_first_chunk, -1); @@ -1160,23 +1366,6 @@ static size_t pcpu_calc_fc_sizes(size_t static_size, size_t reserved_size, return size_sum; } -/* - * Embedding first chunk setup helper. - */ -static void *pcpue_ptr __initdata; -static size_t pcpue_size __initdata; -static size_t pcpue_unit_size __initdata; - -static struct page * __init pcpue_get_page(unsigned int cpu, int pageno) -{ - size_t off = (size_t)pageno << PAGE_SHIFT; - - if (off >= pcpue_size) - return NULL; - - return virt_to_page(pcpue_ptr + cpu * pcpue_unit_size + off); -} - /** * pcpu_embed_first_chunk - embed the first percpu chunk into bootmem * @static_size: the size of static percpu area in bytes @@ -1207,18 +1396,19 @@ static struct page * __init pcpue_get_page(unsigned int cpu, int pageno) ssize_t __init pcpu_embed_first_chunk(size_t static_size, size_t reserved_size, ssize_t dyn_size) { - size_t chunk_size; + size_t size_sum, unit_size, chunk_size; + void *base; unsigned int cpu; /* determine parameters and allocate */ - pcpue_size = pcpu_calc_fc_sizes(static_size, reserved_size, &dyn_size); + size_sum = pcpu_calc_fc_sizes(static_size, reserved_size, &dyn_size); - pcpue_unit_size = max_t(size_t, pcpue_size, PCPU_MIN_UNIT_SIZE); - chunk_size = pcpue_unit_size * num_possible_cpus(); + unit_size = max_t(size_t, size_sum, PCPU_MIN_UNIT_SIZE); + chunk_size = unit_size * num_possible_cpus(); - pcpue_ptr = __alloc_bootmem_nopanic(chunk_size, PAGE_SIZE, - __pa(MAX_DMA_ADDRESS)); - if (!pcpue_ptr) { + base = __alloc_bootmem_nopanic(chunk_size, PAGE_SIZE, + __pa(MAX_DMA_ADDRESS)); + if (!base) { pr_warning("PERCPU: failed to allocate %zu bytes for " "embedding\n", chunk_size); return -ENOMEM; @@ -1226,33 +1416,18 @@ ssize_t __init pcpu_embed_first_chunk(size_t static_size, size_t reserved_size, /* return the leftover and copy */ for_each_possible_cpu(cpu) { - void *ptr = pcpue_ptr + cpu * pcpue_unit_size; + void *ptr = base + cpu * unit_size; - free_bootmem(__pa(ptr + pcpue_size), - pcpue_unit_size - pcpue_size); + free_bootmem(__pa(ptr + size_sum), unit_size - size_sum); memcpy(ptr, __per_cpu_load, static_size); } /* we're ready, commit */ pr_info("PERCPU: Embedded %zu pages at %p, static data %zu bytes\n", - pcpue_size >> PAGE_SHIFT, pcpue_ptr, static_size); + size_sum >> PAGE_SHIFT, base, static_size); - return pcpu_setup_first_chunk(pcpue_get_page, static_size, - reserved_size, dyn_size, - pcpue_unit_size, pcpue_ptr); -} - -/* - * 4k page first chunk setup helper. - */ -static struct page **pcpu4k_pages __initdata; -static int pcpu4k_unit_pages __initdata; - -static struct page * __init pcpu4k_get_page(unsigned int cpu, int pageno) -{ - if (pageno < pcpu4k_unit_pages) - return pcpu4k_pages[cpu * pcpu4k_unit_pages + pageno]; - return NULL; + return pcpu_setup_first_chunk(static_size, reserved_size, dyn_size, + unit_size, base); } /** @@ -1279,23 +1454,25 @@ ssize_t __init pcpu_4k_first_chunk(size_t static_size, size_t reserved_size, pcpu_fc_populate_pte_fn_t populate_pte_fn) { static struct vm_struct vm; + int unit_pages; size_t pages_size; + struct page **pages; unsigned int cpu; int i, j; ssize_t ret; - pcpu4k_unit_pages = PFN_UP(max_t(size_t, static_size + reserved_size, - PCPU_MIN_UNIT_SIZE)); + unit_pages = PFN_UP(max_t(size_t, static_size + reserved_size, + PCPU_MIN_UNIT_SIZE)); /* unaligned allocations can't be freed, round up to page size */ - pages_size = PFN_ALIGN(pcpu4k_unit_pages * num_possible_cpus() * - sizeof(pcpu4k_pages[0])); - pcpu4k_pages = alloc_bootmem(pages_size); + pages_size = PFN_ALIGN(unit_pages * num_possible_cpus() * + sizeof(pages[0])); + pages = alloc_bootmem(pages_size); /* allocate pages */ j = 0; for_each_possible_cpu(cpu) - for (i = 0; i < pcpu4k_unit_pages; i++) { + for (i = 0; i < unit_pages; i++) { void *ptr; ptr = alloc_fn(cpu, PAGE_SIZE); @@ -1304,25 +1481,24 @@ ssize_t __init pcpu_4k_first_chunk(size_t static_size, size_t reserved_size, "4k page for cpu%u\n", cpu); goto enomem; } - pcpu4k_pages[j++] = virt_to_page(ptr); + pages[j++] = virt_to_page(ptr); } /* allocate vm area, map the pages and copy static data */ vm.flags = VM_ALLOC; - vm.size = num_possible_cpus() * pcpu4k_unit_pages << PAGE_SHIFT; + vm.size = num_possible_cpus() * unit_pages << PAGE_SHIFT; vm_area_register_early(&vm, PAGE_SIZE); for_each_possible_cpu(cpu) { unsigned long unit_addr = (unsigned long)vm.addr + - (cpu * pcpu4k_unit_pages << PAGE_SHIFT); + (cpu * unit_pages << PAGE_SHIFT); - for (i = 0; i < pcpu4k_unit_pages; i++) + for (i = 0; i < unit_pages; i++) populate_pte_fn(unit_addr + (i << PAGE_SHIFT)); /* pte already populated, the following shouldn't fail */ - ret = __pcpu_map_pages(unit_addr, - &pcpu4k_pages[cpu * pcpu4k_unit_pages], - pcpu4k_unit_pages); + ret = __pcpu_map_pages(unit_addr, &pages[cpu * unit_pages], + unit_pages); if (ret < 0) panic("failed to map percpu area, err=%zd\n", ret); @@ -1340,19 +1516,18 @@ ssize_t __init pcpu_4k_first_chunk(size_t static_size, size_t reserved_size, /* we're ready, commit */ pr_info("PERCPU: %d 4k pages per cpu, static data %zu bytes\n", - pcpu4k_unit_pages, static_size); + unit_pages, static_size); - ret = pcpu_setup_first_chunk(pcpu4k_get_page, static_size, - reserved_size, -1, - pcpu4k_unit_pages << PAGE_SHIFT, vm.addr); + ret = pcpu_setup_first_chunk(static_size, reserved_size, -1, + unit_pages << PAGE_SHIFT, vm.addr); goto out_free_ar; enomem: while (--j >= 0) - free_fn(page_address(pcpu4k_pages[j]), PAGE_SIZE); + free_fn(page_address(pages[j]), PAGE_SIZE); ret = -ENOMEM; out_free_ar: - free_bootmem(__pa(pcpu4k_pages), pages_size); + free_bootmem(__pa(pages), pages_size); return ret; } @@ -1370,16 +1545,6 @@ static size_t pcpul_unit_size; static struct pcpul_ent *pcpul_map; static struct vm_struct pcpul_vm; -static struct page * __init pcpul_get_page(unsigned int cpu, int pageno) -{ - size_t off = (size_t)pageno << PAGE_SHIFT; - - if (off >= pcpul_size) - return NULL; - - return virt_to_page(pcpul_map[cpu].ptr + off); -} - /** * pcpu_lpage_first_chunk - remap the first percpu chunk using large page * @static_size: the size of static percpu area in bytes @@ -1475,9 +1640,8 @@ ssize_t __init pcpu_lpage_first_chunk(size_t static_size, size_t reserved_size, pr_info("PERCPU: Remapped at %p with large pages, static data " "%zu bytes\n", pcpul_vm.addr, static_size); - ret = pcpu_setup_first_chunk(pcpul_get_page, static_size, - reserved_size, dyn_size, pcpul_unit_size, - pcpul_vm.addr); + ret = pcpu_setup_first_chunk(static_size, reserved_size, dyn_size, + pcpul_unit_size, pcpul_vm.addr); /* sort pcpul_map array for pcpu_lpage_remapped() */ for (i = 0; i < num_possible_cpus() - 1; i++) -- cgit v1.2.3 From 2f39e637ea240efb74cf807d31c93a71a0b89174 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Sat, 4 Jul 2009 08:11:00 +0900 Subject: percpu: allow non-linear / sparse cpu -> unit mapping Currently cpu and unit are always identity mapped. To allow more efficient large page support on NUMA and lazy allocation for possible but offline cpus, cpu -> unit mapping needs to be non-linear and/or sparse. This can be easily implemented by adding a cpu -> unit mapping array and using it whenever looking up the matching unit for a cpu. The only unusal conversion is in pcpu_chunk_addr_search(). The passed in address is unit0 based and unit0 might not be in use so it needs to be converted to address of an in-use unit. This is easily done by adding the unit offset for the current processor. [ Impact: allows non-linear/sparse cpu -> unit mapping, no visible change yet ] Signed-off-by: Tejun Heo Cc: Ingo Molnar Cc: David Miller --- arch/sparc/kernel/smp_64.c | 2 +- include/linux/percpu.h | 3 +- mm/percpu.c | 129 +++++++++++++++++++++++++++++++++------------ 3 files changed, 97 insertions(+), 37 deletions(-) (limited to 'include') diff --git a/arch/sparc/kernel/smp_64.c b/arch/sparc/kernel/smp_64.c index f2f22ee97a7a..6970333b48b8 100644 --- a/arch/sparc/kernel/smp_64.c +++ b/arch/sparc/kernel/smp_64.c @@ -1516,7 +1516,7 @@ void __init setup_per_cpu_areas(void) pcpu_unit_size = pcpu_setup_first_chunk(static_size, PERCPU_MODULE_RESERVE, dyn_size, - PCPU_CHUNK_SIZE, vm.addr); + PCPU_CHUNK_SIZE, vm.addr, NULL); free_bootmem(__pa(ptrs), ptrs_size); diff --git a/include/linux/percpu.h b/include/linux/percpu.h index 63c8b7a23e66..1e0e8878dc2a 100644 --- a/include/linux/percpu.h +++ b/include/linux/percpu.h @@ -57,6 +57,7 @@ #endif extern void *pcpu_base_addr; +extern const int *pcpu_unit_map; typedef void * (*pcpu_fc_alloc_fn_t)(unsigned int cpu, size_t size); typedef void (*pcpu_fc_free_fn_t)(void *ptr, size_t size); @@ -66,7 +67,7 @@ typedef void (*pcpu_fc_map_fn_t)(void *ptr, size_t size, void *addr); extern size_t __init pcpu_setup_first_chunk( size_t static_size, size_t reserved_size, ssize_t dyn_size, size_t unit_size, - void *base_addr); + void *base_addr, const int *unit_map); extern ssize_t __init pcpu_embed_first_chunk( size_t static_size, size_t reserved_size, diff --git a/mm/percpu.c b/mm/percpu.c index 21756814d99f..2196fae24f00 100644 --- a/mm/percpu.c +++ b/mm/percpu.c @@ -8,12 +8,13 @@ * * This is percpu allocator which can handle both static and dynamic * areas. Percpu areas are allocated in chunks in vmalloc area. Each - * chunk is consisted of num_possible_cpus() units and the first chunk - * is used for static percpu variables in the kernel image (special - * boot time alloc/init handling necessary as these areas need to be - * brought up before allocation services are running). Unit grows as - * necessary and all units grow or shrink in unison. When a chunk is - * filled up, another chunk is allocated. ie. in vmalloc area + * chunk is consisted of boot-time determined number of units and the + * first chunk is used for static percpu variables in the kernel image + * (special boot time alloc/init handling necessary as these areas + * need to be brought up before allocation services are running). + * Unit grows as necessary and all units grow or shrink in unison. + * When a chunk is filled up, another chunk is allocated. ie. in + * vmalloc area * * c0 c1 c2 * ------------------- ------------------- ------------ @@ -22,11 +23,13 @@ * * Allocation is done in offset-size areas of single unit space. Ie, * an area of 512 bytes at 6k in c1 occupies 512 bytes at 6k of c1:u0, - * c1:u1, c1:u2 and c1:u3. Percpu access can be done by configuring - * percpu base registers pcpu_unit_size apart. + * c1:u1, c1:u2 and c1:u3. On UMA, units corresponds directly to + * cpus. On NUMA, the mapping can be non-linear and even sparse. + * Percpu access can be done by configuring percpu base registers + * according to cpu to unit mapping and pcpu_unit_size. * - * There are usually many small percpu allocations many of them as - * small as 4 bytes. The allocator organizes chunks into lists + * There are usually many small percpu allocations many of them being + * as small as 4 bytes. The allocator organizes chunks into lists * according to free size and tries to allocate from the fullest one. * Each chunk keeps the maximum contiguous area size hint which is * guaranteed to be eqaul to or larger than the maximum contiguous @@ -99,14 +102,22 @@ struct pcpu_chunk { static int pcpu_unit_pages __read_mostly; static int pcpu_unit_size __read_mostly; +static int pcpu_nr_units __read_mostly; static int pcpu_chunk_size __read_mostly; static int pcpu_nr_slots __read_mostly; static size_t pcpu_chunk_struct_size __read_mostly; +/* cpus with the lowest and highest unit numbers */ +static unsigned int pcpu_first_unit_cpu __read_mostly; +static unsigned int pcpu_last_unit_cpu __read_mostly; + /* the address of the first chunk which starts with the kernel static area */ void *pcpu_base_addr __read_mostly; EXPORT_SYMBOL_GPL(pcpu_base_addr); +/* cpu -> unit map */ +const int *pcpu_unit_map __read_mostly; + /* * The first chunk which always exists. Note that unlike other * chunks, this one can be allocated and mapped in several different @@ -177,7 +188,7 @@ static int pcpu_chunk_slot(const struct pcpu_chunk *chunk) static int pcpu_page_idx(unsigned int cpu, int page_idx) { - return cpu * pcpu_unit_pages + page_idx; + return pcpu_unit_map[cpu] * pcpu_unit_pages + page_idx; } static unsigned long pcpu_chunk_addr(struct pcpu_chunk *chunk, @@ -321,6 +332,14 @@ static struct pcpu_chunk *pcpu_chunk_addr_search(void *addr) return pcpu_first_chunk; } + /* + * The address is relative to unit0 which might be unused and + * thus unmapped. Offset the address to the unit space of the + * current processor before looking it up in the vmalloc + * space. Note that any possible cpu id can be used here, so + * there's no need to worry about preemption or cpu hotplug. + */ + addr += pcpu_unit_map[smp_processor_id()] * pcpu_unit_size; return pcpu_get_page_chunk(vmalloc_to_page(addr)); } @@ -593,8 +612,7 @@ static struct page **pcpu_get_pages_and_bitmap(struct pcpu_chunk *chunk, { static struct page **pages; static unsigned long *bitmap; - size_t pages_size = num_possible_cpus() * pcpu_unit_pages * - sizeof(pages[0]); + size_t pages_size = pcpu_nr_units * pcpu_unit_pages * sizeof(pages[0]); size_t bitmap_size = BITS_TO_LONGS(pcpu_unit_pages) * sizeof(unsigned long); @@ -692,10 +710,9 @@ static int pcpu_alloc_pages(struct pcpu_chunk *chunk, static void pcpu_pre_unmap_flush(struct pcpu_chunk *chunk, int page_start, int page_end) { - unsigned int last = num_possible_cpus() - 1; - - flush_cache_vunmap(pcpu_chunk_addr(chunk, 0, page_start), - pcpu_chunk_addr(chunk, last, page_end)); + flush_cache_vunmap( + pcpu_chunk_addr(chunk, pcpu_first_unit_cpu, page_start), + pcpu_chunk_addr(chunk, pcpu_last_unit_cpu, page_end)); } static void __pcpu_unmap_pages(unsigned long addr, int nr_pages) @@ -756,10 +773,9 @@ static void pcpu_unmap_pages(struct pcpu_chunk *chunk, static void pcpu_post_unmap_tlb_flush(struct pcpu_chunk *chunk, int page_start, int page_end) { - unsigned int last = num_possible_cpus() - 1; - - flush_tlb_kernel_range(pcpu_chunk_addr(chunk, 0, page_start), - pcpu_chunk_addr(chunk, last, page_end)); + flush_tlb_kernel_range( + pcpu_chunk_addr(chunk, pcpu_first_unit_cpu, page_start), + pcpu_chunk_addr(chunk, pcpu_last_unit_cpu, page_end)); } static int __pcpu_map_pages(unsigned long addr, struct page **pages, @@ -835,11 +851,9 @@ err: static void pcpu_post_map_flush(struct pcpu_chunk *chunk, int page_start, int page_end) { - unsigned int last = num_possible_cpus() - 1; - - /* flush at once, please read comments in pcpu_unmap() */ - flush_cache_vmap(pcpu_chunk_addr(chunk, 0, page_start), - pcpu_chunk_addr(chunk, last, page_end)); + flush_cache_vmap( + pcpu_chunk_addr(chunk, pcpu_first_unit_cpu, page_start), + pcpu_chunk_addr(chunk, pcpu_last_unit_cpu, page_end)); } /** @@ -953,8 +967,7 @@ static int pcpu_populate_chunk(struct pcpu_chunk *chunk, int off, int size) bitmap_copy(chunk->populated, populated, pcpu_unit_pages); clear: for_each_possible_cpu(cpu) - memset(chunk->vm->addr + cpu * pcpu_unit_size + off, 0, - size); + memset((void *)pcpu_chunk_addr(chunk, cpu, 0) + off, 0, size); return 0; err_unmap: @@ -1088,6 +1101,7 @@ area_found: mutex_unlock(&pcpu_alloc_mutex); + /* return address relative to unit0 */ return __addr_to_pcpu_ptr(chunk->vm->addr + off); fail_unlock: @@ -1222,6 +1236,7 @@ EXPORT_SYMBOL_GPL(free_percpu); * @dyn_size: free size for dynamic allocation in bytes, -1 for auto * @unit_size: unit size in bytes, must be multiple of PAGE_SIZE * @base_addr: mapped address + * @unit_map: cpu -> unit map, NULL for sequential mapping * * Initialize the first percpu chunk which contains the kernel static * perpcu area. This function is to be called from arch percpu area @@ -1260,16 +1275,17 @@ EXPORT_SYMBOL_GPL(free_percpu); */ size_t __init pcpu_setup_first_chunk(size_t static_size, size_t reserved_size, ssize_t dyn_size, size_t unit_size, - void *base_addr) + void *base_addr, const int *unit_map) { static struct vm_struct first_vm; static int smap[2], dmap[2]; size_t size_sum = static_size + reserved_size + (dyn_size >= 0 ? dyn_size : 0); struct pcpu_chunk *schunk, *dchunk = NULL; + unsigned int cpu, tcpu; int i; - /* santiy checks */ + /* sanity checks */ BUILD_BUG_ON(ARRAY_SIZE(smap) >= PCPU_DFL_MAP_ALLOC || ARRAY_SIZE(dmap) >= PCPU_DFL_MAP_ALLOC); BUG_ON(!static_size); @@ -1278,9 +1294,52 @@ size_t __init pcpu_setup_first_chunk(size_t static_size, size_t reserved_size, BUG_ON(unit_size & ~PAGE_MASK); BUG_ON(unit_size < PCPU_MIN_UNIT_SIZE); + /* determine number of units and verify and initialize pcpu_unit_map */ + if (unit_map) { + int first_unit = INT_MAX, last_unit = INT_MIN; + + for_each_possible_cpu(cpu) { + int unit = unit_map[cpu]; + + BUG_ON(unit < 0); + for_each_possible_cpu(tcpu) { + if (tcpu == cpu) + break; + /* the mapping should be one-to-one */ + BUG_ON(unit_map[tcpu] == unit); + } + + if (unit < first_unit) { + pcpu_first_unit_cpu = cpu; + first_unit = unit; + } + if (unit > last_unit) { + pcpu_last_unit_cpu = cpu; + last_unit = unit; + } + } + pcpu_nr_units = last_unit + 1; + pcpu_unit_map = unit_map; + } else { + int *identity_map; + + /* #units == #cpus, identity mapped */ + identity_map = alloc_bootmem(num_possible_cpus() * + sizeof(identity_map[0])); + + for_each_possible_cpu(cpu) + identity_map[cpu] = cpu; + + pcpu_first_unit_cpu = 0; + pcpu_last_unit_cpu = pcpu_nr_units - 1; + pcpu_nr_units = num_possible_cpus(); + pcpu_unit_map = identity_map; + } + + /* determine basic parameters */ pcpu_unit_pages = unit_size >> PAGE_SHIFT; pcpu_unit_size = pcpu_unit_pages << PAGE_SHIFT; - pcpu_chunk_size = num_possible_cpus() * pcpu_unit_size; + pcpu_chunk_size = pcpu_nr_units * pcpu_unit_size; pcpu_chunk_struct_size = sizeof(struct pcpu_chunk) + BITS_TO_LONGS(pcpu_unit_pages) * sizeof(unsigned long); @@ -1349,7 +1408,7 @@ size_t __init pcpu_setup_first_chunk(size_t static_size, size_t reserved_size, pcpu_chunk_relocate(pcpu_first_chunk, -1); /* we're done */ - pcpu_base_addr = (void *)pcpu_chunk_addr(schunk, 0, 0); + pcpu_base_addr = schunk->vm->addr; return pcpu_unit_size; } @@ -1427,7 +1486,7 @@ ssize_t __init pcpu_embed_first_chunk(size_t static_size, size_t reserved_size, size_sum >> PAGE_SHIFT, base, static_size); return pcpu_setup_first_chunk(static_size, reserved_size, dyn_size, - unit_size, base); + unit_size, base, NULL); } /** @@ -1519,7 +1578,7 @@ ssize_t __init pcpu_4k_first_chunk(size_t static_size, size_t reserved_size, unit_pages, static_size); ret = pcpu_setup_first_chunk(static_size, reserved_size, -1, - unit_pages << PAGE_SHIFT, vm.addr); + unit_pages << PAGE_SHIFT, vm.addr, NULL); goto out_free_ar; enomem: @@ -1641,7 +1700,7 @@ ssize_t __init pcpu_lpage_first_chunk(size_t static_size, size_t reserved_size, "%zu bytes\n", pcpul_vm.addr, static_size); ret = pcpu_setup_first_chunk(static_size, reserved_size, dyn_size, - pcpul_unit_size, pcpul_vm.addr); + pcpul_unit_size, pcpul_vm.addr, NULL); /* sort pcpul_map array for pcpu_lpage_remapped() */ for (i = 0; i < num_possible_cpus() - 1; i++) -- cgit v1.2.3 From a530b7958612bafe2027e21359083dba84f0b3b4 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Sat, 4 Jul 2009 08:11:00 +0900 Subject: percpu: teach large page allocator about NUMA Large page first chunk allocator is primarily used for NUMA machines; however, its NUMA handling is extremely simplistic. Regardless of their proximity, each cpu is put into separate large page just to return most of the allocated space back wasting large amount of vmalloc space and increasing cache footprint. This patch teachs NUMA details to large page allocator. Given processor proximity information, pcpu_lpage_build_unit_map() will find fitting cpu -> unit mapping in which cpus in LOCAL_DISTANCE share the same large page and not too much virtual address space is wasted. This greatly reduces the unit and thus chunk size and wastes much less address space for the first chunk. For example, on 4/4 NUMA machine, the original code occupied 16MB of virtual space for the first chunk while the new code only uses 4MB - one 2MB page for each node. [ Impact: much better space efficiency on NUMA machines ] Signed-off-by: Tejun Heo Cc: Ingo Molnar Cc: Jan Beulich Cc: Andi Kleen Cc: David Miller --- arch/x86/kernel/setup_percpu.c | 72 +++++++-- include/linux/percpu.h | 24 ++- mm/percpu.c | 358 ++++++++++++++++++++++++++++++++--------- 3 files changed, 359 insertions(+), 95 deletions(-) (limited to 'include') diff --git a/arch/x86/kernel/setup_percpu.c b/arch/x86/kernel/setup_percpu.c index 4f2e0ac9130b..7501bb14bd51 100644 --- a/arch/x86/kernel/setup_percpu.c +++ b/arch/x86/kernel/setup_percpu.c @@ -149,36 +149,73 @@ static void __init pcpul_map(void *ptr, size_t size, void *addr) set_pmd(pmd, pmd_v); } +static int pcpu_lpage_cpu_distance(unsigned int from, unsigned int to) +{ + if (early_cpu_to_node(from) == early_cpu_to_node(to)) + return LOCAL_DISTANCE; + else + return REMOTE_DISTANCE; +} + static ssize_t __init setup_pcpu_lpage(size_t static_size, bool chosen) { size_t reserve = PERCPU_MODULE_RESERVE + PERCPU_DYNAMIC_RESERVE; + size_t dyn_size = reserve - PERCPU_FIRST_CHUNK_RESERVE; + size_t unit_map_size, unit_size; + int *unit_map; + int nr_units; + ssize_t ret; + + /* on non-NUMA, embedding is better */ + if (!chosen && !pcpu_need_numa()) + return -EINVAL; + + /* need PSE */ + if (!cpu_has_pse) { + pr_warning("PERCPU: lpage allocator requires PSE\n"); + return -EINVAL; + } + /* allocate and build unit_map */ + unit_map_size = num_possible_cpus() * sizeof(int); + unit_map = alloc_bootmem_nopanic(unit_map_size); + if (!unit_map) { + pr_warning("PERCPU: failed to allocate unit_map\n"); + return -ENOMEM; + } + + ret = pcpu_lpage_build_unit_map(static_size, + PERCPU_FIRST_CHUNK_RESERVE, + &dyn_size, &unit_size, PMD_SIZE, + unit_map, pcpu_lpage_cpu_distance); + if (ret < 0) { + pr_warning("PERCPU: failed to build unit_map\n"); + goto out_free; + } + nr_units = ret; + + /* do the parameters look okay? */ if (!chosen) { size_t vm_size = VMALLOC_END - VMALLOC_START; - size_t tot_size = num_possible_cpus() * PMD_SIZE; - - /* on non-NUMA, embedding is better */ - if (!pcpu_need_numa()) - return -EINVAL; + size_t tot_size = nr_units * unit_size; /* don't consume more than 20% of vmalloc area */ if (tot_size > vm_size / 5) { pr_info("PERCPU: too large chunk size %zuMB for " "large page remap\n", tot_size >> 20); - return -EINVAL; + ret = -EINVAL; + goto out_free; } } - /* need PSE */ - if (!cpu_has_pse) { - pr_warning("PERCPU: lpage allocator requires PSE\n"); - return -EINVAL; - } - - return pcpu_lpage_first_chunk(static_size, PERCPU_FIRST_CHUNK_RESERVE, - reserve - PERCPU_FIRST_CHUNK_RESERVE, - PMD_SIZE, - pcpu_fc_alloc, pcpu_fc_free, pcpul_map); + ret = pcpu_lpage_first_chunk(static_size, PERCPU_FIRST_CHUNK_RESERVE, + dyn_size, unit_size, PMD_SIZE, + unit_map, nr_units, + pcpu_fc_alloc, pcpu_fc_free, pcpul_map); +out_free: + if (ret < 0) + free_bootmem(__pa(unit_map), unit_map_size); + return ret; } #else static ssize_t __init setup_pcpu_lpage(size_t static_size, bool chosen) @@ -299,7 +336,8 @@ void __init setup_per_cpu_areas(void) /* alrighty, percpu areas up and running */ delta = (unsigned long)pcpu_base_addr - (unsigned long)__per_cpu_start; for_each_possible_cpu(cpu) { - per_cpu_offset(cpu) = delta + cpu * pcpu_unit_size; + per_cpu_offset(cpu) = + delta + pcpu_unit_map[cpu] * pcpu_unit_size; per_cpu(this_cpu_off, cpu) = per_cpu_offset(cpu); per_cpu(cpu_number, cpu) = cpu; setup_percpu_segment(cpu); diff --git a/include/linux/percpu.h b/include/linux/percpu.h index 1e0e8878dc2a..8ce91af4aa19 100644 --- a/include/linux/percpu.h +++ b/include/linux/percpu.h @@ -62,6 +62,7 @@ extern const int *pcpu_unit_map; typedef void * (*pcpu_fc_alloc_fn_t)(unsigned int cpu, size_t size); typedef void (*pcpu_fc_free_fn_t)(void *ptr, size_t size); typedef void (*pcpu_fc_populate_pte_fn_t)(unsigned long addr); +typedef int (pcpu_fc_cpu_distance_fn_t)(unsigned int from, unsigned int to); typedef void (*pcpu_fc_map_fn_t)(void *ptr, size_t size, void *addr); extern size_t __init pcpu_setup_first_chunk( @@ -80,18 +81,37 @@ extern ssize_t __init pcpu_4k_first_chunk( pcpu_fc_populate_pte_fn_t populate_pte_fn); #ifdef CONFIG_NEED_MULTIPLE_NODES +extern int __init pcpu_lpage_build_unit_map( + size_t static_size, size_t reserved_size, + ssize_t *dyn_sizep, size_t *unit_sizep, + size_t lpage_size, int *unit_map, + pcpu_fc_cpu_distance_fn_t cpu_distance_fn); + extern ssize_t __init pcpu_lpage_first_chunk( size_t static_size, size_t reserved_size, - ssize_t dyn_size, size_t lpage_size, + size_t dyn_size, size_t unit_size, + size_t lpage_size, const int *unit_map, + int nr_units, pcpu_fc_alloc_fn_t alloc_fn, pcpu_fc_free_fn_t free_fn, pcpu_fc_map_fn_t map_fn); extern void *pcpu_lpage_remapped(void *kaddr); #else +static inline int pcpu_lpage_build_unit_map( + size_t static_size, size_t reserved_size, + ssize_t *dyn_sizep, size_t *unit_sizep, + size_t lpage_size, int *unit_map, + pcpu_fc_cpu_distance_fn_t cpu_distance_fn) +{ + return -EINVAL; +} + static inline ssize_t __init pcpu_lpage_first_chunk( size_t static_size, size_t reserved_size, - ssize_t dyn_size, size_t lpage_size, + size_t dyn_size, size_t unit_size, + size_t lpage_size, const int *unit_map, + int nr_units, pcpu_fc_alloc_fn_t alloc_fn, pcpu_fc_free_fn_t free_fn, pcpu_fc_map_fn_t map_fn) diff --git a/mm/percpu.c b/mm/percpu.c index 2196fae24f00..b3d0bcff8c7c 100644 --- a/mm/percpu.c +++ b/mm/percpu.c @@ -59,6 +59,7 @@ #include #include #include +#include #include #include #include @@ -1594,75 +1595,259 @@ out_free_ar: * Large page remapping first chunk setup helper */ #ifdef CONFIG_NEED_MULTIPLE_NODES + +/** + * pcpu_lpage_build_unit_map - build unit_map for large page remapping + * @static_size: the size of static percpu area in bytes + * @reserved_size: the size of reserved percpu area in bytes + * @dyn_sizep: in/out parameter for dynamic size, -1 for auto + * @unit_sizep: out parameter for unit size + * @unit_map: unit_map to be filled + * @cpu_distance_fn: callback to determine distance between cpus + * + * This function builds cpu -> unit map and determine other parameters + * considering needed percpu size, large page size and distances + * between CPUs in NUMA. + * + * CPUs which are of LOCAL_DISTANCE both ways are grouped together and + * may share units in the same large page. The returned configuration + * is guaranteed to have CPUs on different nodes on different large + * pages and >=75% usage of allocated virtual address space. + * + * RETURNS: + * On success, fills in @unit_map, sets *@dyn_sizep, *@unit_sizep and + * returns the number of units to be allocated. -errno on failure. + */ +int __init pcpu_lpage_build_unit_map(size_t static_size, size_t reserved_size, + ssize_t *dyn_sizep, size_t *unit_sizep, + size_t lpage_size, int *unit_map, + pcpu_fc_cpu_distance_fn_t cpu_distance_fn) +{ + static int group_map[NR_CPUS] __initdata; + static int group_cnt[NR_CPUS] __initdata; + int group_cnt_max = 0; + size_t size_sum, min_unit_size, alloc_size; + int upa, max_upa, uninitialized_var(best_upa); /* units_per_alloc */ + int last_allocs; + unsigned int cpu, tcpu; + int group, unit; + + /* + * Determine min_unit_size, alloc_size and max_upa such that + * alloc_size is multiple of lpage_size and is the smallest + * which can accomodate 4k aligned segments which are equal to + * or larger than min_unit_size. + */ + size_sum = pcpu_calc_fc_sizes(static_size, reserved_size, dyn_sizep); + min_unit_size = max_t(size_t, size_sum, PCPU_MIN_UNIT_SIZE); + + alloc_size = roundup(min_unit_size, lpage_size); + upa = alloc_size / min_unit_size; + while (alloc_size % upa || ((alloc_size / upa) & ~PAGE_MASK)) + upa--; + max_upa = upa; + + /* group cpus according to their proximity */ + for_each_possible_cpu(cpu) { + group = 0; + next_group: + for_each_possible_cpu(tcpu) { + if (cpu == tcpu) + break; + if (group_map[tcpu] == group && + (cpu_distance_fn(cpu, tcpu) > LOCAL_DISTANCE || + cpu_distance_fn(tcpu, cpu) > LOCAL_DISTANCE)) { + group++; + goto next_group; + } + } + group_map[cpu] = group; + group_cnt[group]++; + group_cnt_max = max(group_cnt_max, group_cnt[group]); + } + + /* + * Expand unit size until address space usage goes over 75% + * and then as much as possible without using more address + * space. + */ + last_allocs = INT_MAX; + for (upa = max_upa; upa; upa--) { + int allocs = 0, wasted = 0; + + if (alloc_size % upa || ((alloc_size / upa) & ~PAGE_MASK)) + continue; + + for (group = 0; group_cnt[group]; group++) { + int this_allocs = DIV_ROUND_UP(group_cnt[group], upa); + allocs += this_allocs; + wasted += this_allocs * upa - group_cnt[group]; + } + + /* + * Don't accept if wastage is over 25%. The + * greater-than comparison ensures upa==1 always + * passes the following check. + */ + if (wasted > num_possible_cpus() / 3) + continue; + + /* and then don't consume more memory */ + if (allocs > last_allocs) + break; + last_allocs = allocs; + best_upa = upa; + } + *unit_sizep = alloc_size / best_upa; + + /* assign units to cpus accordingly */ + unit = 0; + for (group = 0; group_cnt[group]; group++) { + for_each_possible_cpu(cpu) + if (group_map[cpu] == group) + unit_map[cpu] = unit++; + unit = roundup(unit, best_upa); + } + + return unit; /* unit contains aligned number of units */ +} + struct pcpul_ent { - unsigned int cpu; void *ptr; + void *map_addr; }; static size_t pcpul_size; -static size_t pcpul_unit_size; +static size_t pcpul_lpage_size; +static int pcpul_nr_lpages; static struct pcpul_ent *pcpul_map; -static struct vm_struct pcpul_vm; + +static bool __init pcpul_unit_to_cpu(int unit, const int *unit_map, + unsigned int *cpup) +{ + unsigned int cpu; + + for_each_possible_cpu(cpu) + if (unit_map[cpu] == unit) { + if (cpup) + *cpup = cpu; + return true; + } + + return false; +} + +static void __init pcpul_lpage_dump_cfg(const char *lvl, size_t static_size, + size_t reserved_size, size_t dyn_size, + size_t unit_size, size_t lpage_size, + const int *unit_map, int nr_units) +{ + int width = 1, v = nr_units; + char empty_str[] = "--------"; + int upl, lpl; /* units per lpage, lpage per line */ + unsigned int cpu; + int lpage, unit; + + while (v /= 10) + width++; + empty_str[min_t(int, width, sizeof(empty_str) - 1)] = '\0'; + + upl = max_t(int, lpage_size / unit_size, 1); + lpl = rounddown_pow_of_two(max_t(int, 60 / (upl * (width + 1) + 2), 1)); + + printk("%spcpu-lpage: sta/res/dyn=%zu/%zu/%zu unit=%zu lpage=%zu", lvl, + static_size, reserved_size, dyn_size, unit_size, lpage_size); + + for (lpage = 0, unit = 0; unit < nr_units; unit++) { + if (!(unit % upl)) { + if (!(lpage++ % lpl)) { + printk("\n"); + printk("%spcpu-lpage: ", lvl); + } else + printk("| "); + } + if (pcpul_unit_to_cpu(unit, unit_map, &cpu)) + printk("%0*d ", width, cpu); + else + printk("%s ", empty_str); + } + printk("\n"); +} /** * pcpu_lpage_first_chunk - remap the first percpu chunk using large page * @static_size: the size of static percpu area in bytes * @reserved_size: the size of reserved percpu area in bytes - * @dyn_size: free size for dynamic allocation in bytes, -1 for auto + * @dyn_size: free size for dynamic allocation in bytes + * @unit_size: unit size in bytes * @lpage_size: the size of a large page + * @unit_map: cpu -> unit mapping + * @nr_units: the number of units * @alloc_fn: function to allocate percpu lpage, always called with lpage_size * @free_fn: function to free percpu memory, @size <= lpage_size * @map_fn: function to map percpu lpage, always called with lpage_size * - * This allocator uses large page as unit. A large page is allocated - * for each cpu and each is remapped into vmalloc area using large - * page mapping. As large page can be quite large, only part of it is - * used for the first chunk. Unused part is returned to the bootmem - * allocator. - * - * So, the large pages are mapped twice - once to the physical mapping - * and to the vmalloc area for the first percpu chunk. The double - * mapping does add one more large TLB entry pressure but still is - * much better than only using 4k mappings while still being NUMA - * friendly. + * This allocator uses large page to build and map the first chunk. + * Unlike other helpers, the caller should always specify @dyn_size + * and @unit_size. These parameters along with @unit_map and + * @nr_units can be determined using pcpu_lpage_build_unit_map(). + * This two stage initialization is to allow arch code to evaluate the + * parameters before committing to it. + * + * Large pages are allocated as directed by @unit_map and other + * parameters and mapped to vmalloc space. Unused holes are returned + * to the page allocator. Note that these holes end up being actively + * mapped twice - once to the physical mapping and to the vmalloc area + * for the first percpu chunk. Depending on architecture, this might + * cause problem when changing page attributes of the returned area. + * These double mapped areas can be detected using + * pcpu_lpage_remapped(). * * RETURNS: * The determined pcpu_unit_size which can be used to initialize * percpu access on success, -errno on failure. */ ssize_t __init pcpu_lpage_first_chunk(size_t static_size, size_t reserved_size, - ssize_t dyn_size, size_t lpage_size, + size_t dyn_size, size_t unit_size, + size_t lpage_size, const int *unit_map, + int nr_units, pcpu_fc_alloc_fn_t alloc_fn, pcpu_fc_free_fn_t free_fn, pcpu_fc_map_fn_t map_fn) { - size_t size_sum; + static struct vm_struct vm; + size_t chunk_size = unit_size * nr_units; size_t map_size; unsigned int cpu; - int i, j; ssize_t ret; + int i, j, unit; - /* - * Currently supports only single page. Supporting multiple - * pages won't be too difficult if it ever becomes necessary. - */ - size_sum = pcpu_calc_fc_sizes(static_size, reserved_size, &dyn_size); + pcpul_lpage_dump_cfg(KERN_DEBUG, static_size, reserved_size, dyn_size, + unit_size, lpage_size, unit_map, nr_units); - pcpul_unit_size = lpage_size; - pcpul_size = max_t(size_t, size_sum, PCPU_MIN_UNIT_SIZE); - if (pcpul_size > pcpul_unit_size) { - pr_warning("PERCPU: static data is larger than large page, " - "can't use large page\n"); - return -EINVAL; - } + BUG_ON(chunk_size % lpage_size); + + pcpul_size = static_size + reserved_size + dyn_size; + pcpul_lpage_size = lpage_size; + pcpul_nr_lpages = chunk_size / lpage_size; /* allocate pointer array and alloc large pages */ - map_size = PFN_ALIGN(num_possible_cpus() * sizeof(pcpul_map[0])); + map_size = pcpul_nr_lpages * sizeof(pcpul_map[0]); pcpul_map = alloc_bootmem(map_size); - for_each_possible_cpu(cpu) { + /* allocate all pages */ + for (i = 0; i < pcpul_nr_lpages; i++) { + size_t offset = i * lpage_size; + int first_unit = offset / unit_size; + int last_unit = (offset + lpage_size - 1) / unit_size; void *ptr; + /* find out which cpu is mapped to this unit */ + for (unit = first_unit; unit <= last_unit; unit++) + if (pcpul_unit_to_cpu(unit, unit_map, &cpu)) + goto found; + continue; + found: ptr = alloc_fn(cpu, lpage_size); if (!ptr) { pr_warning("PERCPU: failed to allocate large page " @@ -1670,53 +1855,79 @@ ssize_t __init pcpu_lpage_first_chunk(size_t static_size, size_t reserved_size, goto enomem; } - /* - * Only use pcpul_size bytes and give back the rest. - * - * Ingo: The lpage_size up-rounding bootmem is needed - * to make sure the partial lpage is still fully RAM - - * it's not well-specified to have a incompatible area - * (unmapped RAM, device memory, etc.) in that hole. - */ - free_fn(ptr + pcpul_size, lpage_size - pcpul_size); - - pcpul_map[cpu].cpu = cpu; - pcpul_map[cpu].ptr = ptr; + pcpul_map[i].ptr = ptr; + } - memcpy(ptr, __per_cpu_load, static_size); + /* return unused holes */ + for (unit = 0; unit < nr_units; unit++) { + size_t start = unit * unit_size; + size_t end = start + unit_size; + size_t off, next; + + /* don't free used part of occupied unit */ + if (pcpul_unit_to_cpu(unit, unit_map, NULL)) + start += pcpul_size; + + /* unit can span more than one page, punch the holes */ + for (off = start; off < end; off = next) { + void *ptr = pcpul_map[off / lpage_size].ptr; + next = min(roundup(off + 1, lpage_size), end); + if (ptr) + free_fn(ptr + off % lpage_size, next - off); + } } - /* allocate address and map */ - pcpul_vm.flags = VM_ALLOC; - pcpul_vm.size = num_possible_cpus() * pcpul_unit_size; - vm_area_register_early(&pcpul_vm, pcpul_unit_size); + /* allocate address, map and copy */ + vm.flags = VM_ALLOC; + vm.size = chunk_size; + vm_area_register_early(&vm, unit_size); + + for (i = 0; i < pcpul_nr_lpages; i++) { + if (!pcpul_map[i].ptr) + continue; + pcpul_map[i].map_addr = vm.addr + i * lpage_size; + map_fn(pcpul_map[i].ptr, lpage_size, pcpul_map[i].map_addr); + } for_each_possible_cpu(cpu) - map_fn(pcpul_map[cpu].ptr, pcpul_unit_size, - pcpul_vm.addr + cpu * pcpul_unit_size); + memcpy(vm.addr + unit_map[cpu] * unit_size, __per_cpu_load, + static_size); /* we're ready, commit */ pr_info("PERCPU: Remapped at %p with large pages, static data " - "%zu bytes\n", pcpul_vm.addr, static_size); + "%zu bytes\n", vm.addr, static_size); ret = pcpu_setup_first_chunk(static_size, reserved_size, dyn_size, - pcpul_unit_size, pcpul_vm.addr, NULL); - - /* sort pcpul_map array for pcpu_lpage_remapped() */ - for (i = 0; i < num_possible_cpus() - 1; i++) - for (j = i + 1; j < num_possible_cpus(); j++) - if (pcpul_map[i].ptr > pcpul_map[j].ptr) { - struct pcpul_ent tmp = pcpul_map[i]; - pcpul_map[i] = pcpul_map[j]; - pcpul_map[j] = tmp; - } + unit_size, vm.addr, unit_map); + + /* + * Sort pcpul_map array for pcpu_lpage_remapped(). Unmapped + * lpages are pushed to the end and trimmed. + */ + for (i = 0; i < pcpul_nr_lpages - 1; i++) + for (j = i + 1; j < pcpul_nr_lpages; j++) { + struct pcpul_ent tmp; + + if (!pcpul_map[j].ptr) + continue; + if (pcpul_map[i].ptr && + pcpul_map[i].ptr < pcpul_map[j].ptr) + continue; + + tmp = pcpul_map[i]; + pcpul_map[i] = pcpul_map[j]; + pcpul_map[j] = tmp; + } + + while (pcpul_nr_lpages && !pcpul_map[pcpul_nr_lpages - 1].ptr) + pcpul_nr_lpages--; return ret; enomem: - for_each_possible_cpu(cpu) - if (pcpul_map[cpu].ptr) - free_fn(pcpul_map[cpu].ptr, pcpul_size); + for (i = 0; i < pcpul_nr_lpages; i++) + if (pcpul_map[i].ptr) + free_fn(pcpul_map[i].ptr, lpage_size); free_bootmem(__pa(pcpul_map), map_size); return -ENOMEM; } @@ -1739,10 +1950,10 @@ enomem: */ void *pcpu_lpage_remapped(void *kaddr) { - unsigned long unit_mask = pcpul_unit_size - 1; - void *lpage_addr = (void *)((unsigned long)kaddr & ~unit_mask); - unsigned long offset = (unsigned long)kaddr & unit_mask; - int left = 0, right = num_possible_cpus() - 1; + unsigned long lpage_mask = pcpul_lpage_size - 1; + void *lpage_addr = (void *)((unsigned long)kaddr & ~lpage_mask); + unsigned long offset = (unsigned long)kaddr & lpage_mask; + int left = 0, right = pcpul_nr_lpages - 1; int pos; /* pcpul in use at all? */ @@ -1757,13 +1968,8 @@ void *pcpu_lpage_remapped(void *kaddr) left = pos + 1; else if (pcpul_map[pos].ptr > lpage_addr) right = pos - 1; - else { - /* it shouldn't be in the area for the first chunk */ - WARN_ON(offset < pcpul_size); - - return pcpul_vm.addr + - pcpul_map[pos].cpu * pcpul_unit_size + offset; - } + else + return pcpul_map[pos].map_addr + offset; } return NULL; -- cgit v1.2.3 From 96ccd4a43a4d80c80be636cd025a69959cf47424 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Sun, 5 Jul 2009 12:47:52 +0200 Subject: genirq: Remove obsolete defines and typedefs The defines and typedefs (hw_interrupt_type, no_irq_type, irq_desc_t) have been kept around for migration reasons. The last users are gone, remove them. Signed-off-by: Thomas Gleixner --- Documentation/feature-removal-schedule.txt | 9 --------- include/linux/irq.h | 7 ------- 2 files changed, 16 deletions(-) (limited to 'include') diff --git a/Documentation/feature-removal-schedule.txt b/Documentation/feature-removal-schedule.txt index f8cd450be9aa..2af78126b053 100644 --- a/Documentation/feature-removal-schedule.txt +++ b/Documentation/feature-removal-schedule.txt @@ -394,15 +394,6 @@ Who: Thomas Gleixner ----------------------------- -What: obsolete generic irq defines and typedefs -When: 2.6.30 -Why: The defines and typedefs (hw_interrupt_type, no_irq_type, irq_desc_t) - have been kept around for migration reasons. After more than two years - it's time to remove them finally -Who: Thomas Gleixner - ---------------------------- - What: fakephp and associated sysfs files in /sys/bus/pci/slots/ When: 2011 Why: In 2.6.27, the semantics of /sys/bus/pci/slots was redefined to diff --git a/include/linux/irq.h b/include/linux/irq.h index cb2e77a3f7f7..6956df9961ab 100644 --- a/include/linux/irq.h +++ b/include/linux/irq.h @@ -219,13 +219,6 @@ static inline struct irq_desc *move_irq_desc(struct irq_desc *desc, int node) extern struct irq_desc *irq_to_desc_alloc_node(unsigned int irq, int node); -/* - * Migration helpers for obsolete names, they will go away: - */ -#define hw_interrupt_type irq_chip -#define no_irq_type no_irq_chip -typedef struct irq_desc irq_desc_t; - /* * Pick up the arch-dependent methods: */ -- cgit v1.2.3 From 509dd025a4ac6f32921211557cfd434b81d8d7a7 Mon Sep 17 00:00:00 2001 From: Mauro Carvalho Chehab Date: Tue, 30 Jun 2009 16:07:26 -0300 Subject: V4L/DVB (12148): move V4L2_PIX_FMT_SGRBG8 to the proper place Instead of defining a new pif format on an internal header, move it to the V4L2 API header. Signed-off-by: Mauro Carvalho Chehab --- drivers/media/video/gspca/stv06xx/stv06xx.h | 4 ---- include/linux/videodev2.h | 2 ++ 2 files changed, 2 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/drivers/media/video/gspca/stv06xx/stv06xx.h b/drivers/media/video/gspca/stv06xx/stv06xx.h index 9df7137fe67e..992ce530f138 100644 --- a/drivers/media/video/gspca/stv06xx/stv06xx.h +++ b/drivers/media/video/gspca/stv06xx/stv06xx.h @@ -36,10 +36,6 @@ #define STV_ISOC_ENDPOINT_ADDR 0x81 -#ifndef V4L2_PIX_FMT_SGRBG8 -#define V4L2_PIX_FMT_SGRBG8 v4l2_fourcc('G', 'R', 'B', 'G') -#endif - #define STV_REG23 0x0423 /* Control registers of the STV0600 ASIC */ diff --git a/include/linux/videodev2.h b/include/linux/videodev2.h index 8a025d510904..95846d988011 100644 --- a/include/linux/videodev2.h +++ b/include/linux/videodev2.h @@ -318,6 +318,8 @@ struct v4l2_pix_format { /* see http://www.siliconimaging.com/RGB%20Bayer.htm */ #define V4L2_PIX_FMT_SBGGR8 v4l2_fourcc('B', 'A', '8', '1') /* 8 BGBG.. GRGR.. */ #define V4L2_PIX_FMT_SGBRG8 v4l2_fourcc('G', 'B', 'R', 'G') /* 8 GBGB.. RGRG.. */ +#define V4L2_PIX_FMT_SGRBG8 v4l2_fourcc('G', 'R', 'B', 'G') /* 8 GRGR.. BGBG.. */ + /* * 10bit raw bayer, expanded to 16 bits * xxxxrrrrrrrrrrxxxxgggggggggg xxxxggggggggggxxxxbbbbbbbbbb... -- cgit v1.2.3 From 7dfba00d05f3c7db9510f3b54a472981cf1521af Mon Sep 17 00:00:00 2001 From: Mauro Carvalho Chehab Date: Mon, 29 Jun 2009 05:41:26 -0300 Subject: V4L/DVB (12135): Add a driver for mt9v011 sensor Adds driver for mt9v011 based on its datasheet, available at: http://download.micron.com/pdf/datasheets/imaging/MT9V011.pdf The driver was tested with a webcam that will be added on a next patch. Signed-off-by: Mauro Carvalho Chehab --- drivers/media/video/Kconfig | 8 + drivers/media/video/Makefile | 1 + drivers/media/video/mt9v011.c | 355 ++++++++++++++++++++++++++++++++++++++++ drivers/media/video/mt9v011.h | 35 ++++ include/media/v4l2-chip-ident.h | 3 + 5 files changed, 402 insertions(+) create mode 100644 drivers/media/video/mt9v011.c create mode 100644 drivers/media/video/mt9v011.h (limited to 'include') diff --git a/drivers/media/video/Kconfig b/drivers/media/video/Kconfig index 061e147f6f26..84b6fc15519d 100644 --- a/drivers/media/video/Kconfig +++ b/drivers/media/video/Kconfig @@ -312,6 +312,14 @@ config VIDEO_OV7670 OV7670 VGA camera. It currently only works with the M88ALP01 controller. +config VIDEO_MT9V011 + tristate "Micron mt9v011 sensor support" + depends on I2C && VIDEO_V4L2 + ---help--- + This is a Video4Linux2 sensor-level driver for the Micron + mt0v011 1.3 Mpixel camera. It currently only works with the + em28xx driver. + config VIDEO_TCM825X tristate "TCM825x camera sensor support" depends on I2C && VIDEO_V4L2 diff --git a/drivers/media/video/Makefile b/drivers/media/video/Makefile index 7fb3add1b387..9f2e3214a482 100644 --- a/drivers/media/video/Makefile +++ b/drivers/media/video/Makefile @@ -69,6 +69,7 @@ obj-$(CONFIG_VIDEO_UPD64083) += upd64083.o obj-$(CONFIG_VIDEO_OV7670) += ov7670.o obj-$(CONFIG_VIDEO_TCM825X) += tcm825x.o obj-$(CONFIG_VIDEO_TVEEPROM) += tveeprom.o +obj-$(CONFIG_VIDEO_MT9V011) += mt9v011.o obj-$(CONFIG_SOC_CAMERA_MT9M001) += mt9m001.o obj-$(CONFIG_SOC_CAMERA_MT9M111) += mt9m111.o diff --git a/drivers/media/video/mt9v011.c b/drivers/media/video/mt9v011.c new file mode 100644 index 000000000000..4389197cedd7 --- /dev/null +++ b/drivers/media/video/mt9v011.c @@ -0,0 +1,355 @@ +/* + * mt9v011 -Micron 1/4-Inch VGA Digital Image Sensor + * + * Copyright (c) 2009 Mauro Carvalho Chehab (mchehab@redhat.com) + * This code is placed under the terms of the GNU General Public License v2 + */ + +#include +#include +#include +#include +#include "mt9v011.h" +#include +#include + +MODULE_DESCRIPTION("Micron mt9v011 sensor driver"); +MODULE_AUTHOR("Mauro Carvalho Chehab "); +MODULE_LICENSE("GPL"); + + +static int debug; +module_param(debug, int, 0); +MODULE_PARM_DESC(debug, "Debug level (0-2)"); + +/* supported controls */ +static struct v4l2_queryctrl mt9v011_qctrl[] = { + { + .id = V4L2_CID_GAIN, + .type = V4L2_CTRL_TYPE_INTEGER, + .name = "Gain", + .minimum = 0, + .maximum = (1 << 10) - 1, + .step = 1, + .default_value = 0x0020, + .flags = 0, + }, { + .id = V4L2_CID_RED_BALANCE, + .type = V4L2_CTRL_TYPE_INTEGER, + .name = "Red Balance", + .minimum = -1 << 9, + .maximum = (1 << 9) - 1, + .step = 1, + .default_value = 0, + .flags = 0, + }, { + .id = V4L2_CID_BLUE_BALANCE, + .type = V4L2_CTRL_TYPE_INTEGER, + .name = "Blue Balance", + .minimum = -1 << 9, + .maximum = (1 << 9) - 1, + .step = 1, + .default_value = 0, + .flags = 0, + }, +}; + +struct mt9v011 { + struct v4l2_subdev sd; + + u16 global_gain, red_bal, blue_bal; +}; + +static inline struct mt9v011 *to_mt9v011(struct v4l2_subdev *sd) +{ + return container_of(sd, struct mt9v011, sd); +} + +static int mt9v011_read(struct v4l2_subdev *sd, unsigned char addr) +{ + struct i2c_client *c = v4l2_get_subdevdata(sd); + __be16 buffer; + int rc, val; + + if (1 != (rc = i2c_master_send(c, &addr, 1))) + v4l2_dbg(0, debug, sd, + "i2c i/o error: rc == %d (should be 1)\n", rc); + + msleep(10); + + if (2 != (rc = i2c_master_recv(c, (char *)&buffer, 2))) + v4l2_dbg(0, debug, sd, + "i2c i/o error: rc == %d (should be 1)\n", rc); + + val = be16_to_cpu(buffer); + + v4l2_dbg(2, debug, sd, "mt9v011: read 0x%02x = 0x%04x\n", addr, val); + + return val; +} + +static void mt9v011_write(struct v4l2_subdev *sd, unsigned char addr, + u16 value) +{ + struct i2c_client *c = v4l2_get_subdevdata(sd); + unsigned char buffer[3]; + int rc; + + buffer[0] = addr; + buffer[1] = value >> 8; + buffer[2] = value & 0xff; + + v4l2_dbg(2, debug, sd, + "mt9v011: writing 0x%02x 0x%04x\n", buffer[0], value); + if (3 != (rc = i2c_master_send(c, buffer, 3))) + v4l2_dbg(0, debug, sd, + "i2c i/o error: rc == %d (should be 3)\n", rc); +} + + +struct i2c_reg_value { + unsigned char reg; + u16 value; +}; + +/* + * Values used at the original driver + * Some values are marked as Reserved at the datasheet + */ +static const struct i2c_reg_value mt9v011_init_default[] = { + /* guessed meaning - as mt9m111 */ + { R0D_MT9V011_RESET, 0x0001 }, + { R0D_MT9V011_RESET, 0x0000 }, + { R01_MT9V011_ROWSTART, 0x0008 }, + { R02_MT9V011_COLSTART, 0x0014 }, + { R03_MT9V011_HEIGHT, 0x01e0 }, + { R04_MT9V011_WIDTH, 0x0280 }, + { R05_MT9V011_HBLANK, 0x0001 }, + { R05_MT9V011_HBLANK, 0x0001 }, + { R0A_MT9V011_CLK_SPEED, 0x0000 }, + { R05_MT9V011_HBLANK, 0x000a }, + { 0x30, 0x0005 }, + { 0x34, 0x0100 }, + { 0x3d, 0x068f }, + { 0x40, 0x01e0 }, + { 0x52, 0x0100 }, + { 0x58, 0x0038 }, /* Datasheet default 0x0078 */ + { 0x59, 0x0723 }, /* Datasheet default 0x0703 */ + { 0x62, 0x041a }, /* Datasheet default 0x0418 */ + { R09_MT9V011_SHUTTER_WIDTH, 0x0418 }, + { R20_MT9V011_READ_MODE, 0x1100 }, +}; + +static void set_balance(struct v4l2_subdev *sd) +{ + struct mt9v011 *core = to_mt9v011(sd); + u16 green1_gain, green2_gain, blue_gain, red_gain; + + green1_gain = core->global_gain; + green2_gain = core->global_gain; + + blue_gain = core->global_gain + + core->global_gain * core->blue_bal / (1 << 9); + + red_gain = core->global_gain + + core->global_gain * core->blue_bal / (1 << 9); + + mt9v011_write(sd, R2B_MT9V011_GREEN_1_GAIN, green1_gain); + mt9v011_write(sd, R2E_MT9V011_GREEN_2_GAIN, green1_gain); + mt9v011_write(sd, R2C_MT9V011_BLUE_GAIN, blue_gain); + mt9v011_write(sd, R2D_MT9V011_RED_GAIN, red_gain); +} + +static int mt9v011_reset(struct v4l2_subdev *sd, u32 val) +{ + u16 version; + int i; + + version = mt9v011_read(sd, R00_MT9V011_CHIP_VERSION); + + if (version != MT9V011_VERSION) { + v4l2_info(sd, "*** unknown micron chip detected (0x%04x.\n", + version); + return -EINVAL; + + } + + for (i = 0; i < ARRAY_SIZE(mt9v011_init_default); i++) + mt9v011_write(sd, mt9v011_init_default[i].reg, + mt9v011_init_default[i].value); + + set_balance(sd); + + return 0; +}; + +static int mt9v011_g_ctrl(struct v4l2_subdev *sd, struct v4l2_control *ctrl) +{ + struct mt9v011 *core = to_mt9v011(sd); + + v4l2_dbg(1, debug, sd, "g_ctrl called\n"); + + switch (ctrl->id) { + case V4L2_CID_GAIN: + ctrl->value = core->global_gain; + return 0; + case V4L2_CID_RED_BALANCE: + ctrl->value = core->red_bal; + return 0; + case V4L2_CID_BLUE_BALANCE: + ctrl->value = core->blue_bal; + return 0; + } + return -EINVAL; +} + +static int mt9v011_s_ctrl(struct v4l2_subdev *sd, struct v4l2_control *ctrl) +{ + struct mt9v011 *core = to_mt9v011(sd); + u8 i, n; + n = ARRAY_SIZE(mt9v011_qctrl); + + for (i = 0; i < n; i++) { + if (ctrl->id != mt9v011_qctrl[i].id) + continue; + if (ctrl->value < mt9v011_qctrl[i].minimum || + ctrl->value > mt9v011_qctrl[i].maximum) + return -ERANGE; + v4l2_dbg(1, debug, sd, "s_ctrl: id=%d, value=%d\n", + ctrl->id, ctrl->value); + break; + } + + switch (ctrl->id) { + case V4L2_CID_GAIN: + core->global_gain = ctrl->value; + break; + case V4L2_CID_RED_BALANCE: + core->red_bal = ctrl->value; + break; + case V4L2_CID_BLUE_BALANCE: + core->blue_bal = ctrl->value; + break; + default: + return -EINVAL; + } + + set_balance(sd); + + return 0; +} + +#ifdef CONFIG_VIDEO_ADV_DEBUG +static int mt9v011_g_register(struct v4l2_subdev *sd, + struct v4l2_dbg_register *reg) +{ + struct i2c_client *client = v4l2_get_subdevdata(sd); + + if (!v4l2_chip_match_i2c_client(client, ®->match)) + return -EINVAL; + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + + reg->val = mt9v011_read(sd, reg->reg & 0xff); + reg->size = 2; + + return 0; +} + +static int mt9v011_s_register(struct v4l2_subdev *sd, + struct v4l2_dbg_register *reg) +{ + struct i2c_client *client = v4l2_get_subdevdata(sd); + + if (!v4l2_chip_match_i2c_client(client, ®->match)) + return -EINVAL; + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + + mt9v011_write(sd, reg->reg & 0xff, reg->val & 0xffff); + + return 0; +} +#endif + +static int mt9v011_g_chip_ident(struct v4l2_subdev *sd, + struct v4l2_dbg_chip_ident *chip) +{ + struct i2c_client *client = v4l2_get_subdevdata(sd); + + return v4l2_chip_ident_i2c_client(client, chip, V4L2_IDENT_MT9V011, + MT9V011_VERSION); +} + +static const struct v4l2_subdev_core_ops mt9v011_core_ops = { + .g_ctrl = mt9v011_g_ctrl, + .s_ctrl = mt9v011_s_ctrl, + .reset = mt9v011_reset, + .g_chip_ident = mt9v011_g_chip_ident, +#ifdef CONFIG_VIDEO_ADV_DEBUG + .g_register = mt9v011_g_register, + .s_register = mt9v011_s_register, +#endif +}; + +static const struct v4l2_subdev_ops mt9v011_ops = { + .core = &mt9v011_core_ops, +}; + + +/**************************************************************************** + I2C Client & Driver + ****************************************************************************/ + +static int mt9v011_probe(struct i2c_client *c, + const struct i2c_device_id *id) +{ + struct mt9v011 *core; + struct v4l2_subdev *sd; + + /* Check if the adapter supports the needed features */ + if (!i2c_check_functionality(c->adapter, + I2C_FUNC_SMBUS_READ_BYTE | I2C_FUNC_SMBUS_WRITE_BYTE_DATA)) + return -EIO; + + core = kzalloc(sizeof(struct mt9v011), GFP_KERNEL); + if (!core) + return -ENOMEM; + + core->global_gain = 0x0024; + + sd = &core->sd; + v4l2_i2c_subdev_init(sd, c, &mt9v011_ops); + v4l_info(c, "chip found @ 0x%02x (%s)\n", + c->addr << 1, c->adapter->name); + + return 0; +} + +static int mt9v011_remove(struct i2c_client *c) +{ + struct v4l2_subdev *sd = i2c_get_clientdata(c); + + v4l2_dbg(1, debug, sd, + "mt9v011.c: removing mt9v011 adapter on address 0x%x\n", + c->addr << 1); + + v4l2_device_unregister_subdev(sd); + kfree(to_mt9v011(sd)); + return 0; +} + +/* ----------------------------------------------------------------------- */ + +static const struct i2c_device_id mt9v011_id[] = { + { "mt9v011", 0 }, + { } +}; +MODULE_DEVICE_TABLE(i2c, mt9v011_id); + +static struct v4l2_i2c_driver_data v4l2_i2c_data = { + .name = "mt9v011", + .probe = mt9v011_probe, + .remove = mt9v011_remove, + .id_table = mt9v011_id, +}; diff --git a/drivers/media/video/mt9v011.h b/drivers/media/video/mt9v011.h new file mode 100644 index 000000000000..9e443ee30558 --- /dev/null +++ b/drivers/media/video/mt9v011.h @@ -0,0 +1,35 @@ +/* + * mt9v011 -Micron 1/4-Inch VGA Digital Image Sensor + * + * Copyright (c) 2009 Mauro Carvalho Chehab (mchehab@redhat.com) + * This code is placed under the terms of the GNU General Public License v2 + */ + +#ifndef MT9V011_H_ +#define MT9V011_H_ + +#define R00_MT9V011_CHIP_VERSION 0x00 +#define R01_MT9V011_ROWSTART 0x01 +#define R02_MT9V011_COLSTART 0x02 +#define R03_MT9V011_HEIGHT 0x03 +#define R04_MT9V011_WIDTH 0x04 +#define R05_MT9V011_HBLANK 0x05 +#define R06_MT9V011_VBLANK 0x06 +#define R07_MT9V011_OUT_CTRL 0x07 +#define R09_MT9V011_SHUTTER_WIDTH 0x09 +#define R0A_MT9V011_CLK_SPEED 0x0a +#define R0B_MT9V011_RESTART 0x0b +#define R0C_MT9V011_SHUTTER_DELAY 0x0c +#define R0D_MT9V011_RESET 0x0d +#define R1E_MT9V011_DIGITAL_ZOOM 0x1e +#define R20_MT9V011_READ_MODE 0x20 +#define R2B_MT9V011_GREEN_1_GAIN 0x2b +#define R2C_MT9V011_BLUE_GAIN 0x2c +#define R2D_MT9V011_RED_GAIN 0x2d +#define R2E_MT9V011_GREEN_2_GAIN 0x2e +#define R35_MT9V011_GLOBAL_GAIN 0x35 +#define RF1_MT9V011_CHIP_ENABLE 0xf1 + +#define MT9V011_VERSION 0x8243 + +#endif diff --git a/include/media/v4l2-chip-ident.h b/include/media/v4l2-chip-ident.h index 4d7e2272c42f..11a4a2d3e364 100644 --- a/include/media/v4l2-chip-ident.h +++ b/include/media/v4l2-chip-ident.h @@ -155,6 +155,9 @@ enum { /* module cafe_ccic, just ident 8801 */ V4L2_IDENT_CAFE = 8801, + /* module mt9v011, just ident 8243 */ + V4L2_IDENT_MT9V011 = 8243, + /* module tw9910: just ident 9910 */ V4L2_IDENT_TW9910 = 9910, -- cgit v1.2.3 From 0e8635a8e1f2d4a9e1bfc6c3b21419a5921e674f Mon Sep 17 00:00:00 2001 From: Florian Westphal Date: Sat, 20 Jun 2009 00:53:25 +0000 Subject: net: remove NET_RX_BAD and NET_RX_CN* defines almost no users in the tree; and the few that use them treat them like NET_RX_DROP. Signed-off-by: Florian Westphal Signed-off-by: David S. Miller --- drivers/net/rionet.c | 5 ----- drivers/net/wan/farsync.c | 19 ------------------- drivers/staging/otus/wrap_pkt.c | 3 --- include/linux/netdevice.h | 4 ---- net/decnet/dn_route.c | 2 +- net/lapb/lapb_iface.c | 2 +- 6 files changed, 2 insertions(+), 33 deletions(-) (limited to 'include') diff --git a/drivers/net/rionet.c b/drivers/net/rionet.c index 8702e7acdee6..74cdb6f8f84f 100644 --- a/drivers/net/rionet.c +++ b/drivers/net/rionet.c @@ -114,11 +114,6 @@ static int rionet_rx_clean(struct net_device *ndev) if (error == NET_RX_DROP) { ndev->stats.rx_dropped++; - } else if (error == NET_RX_BAD) { - if (netif_msg_rx_err(rnet)) - printk(KERN_WARNING "%s: bad rx packet\n", - DRV_NAME); - ndev->stats.rx_errors++; } else { ndev->stats.rx_packets++; ndev->stats.rx_bytes += RIO_MAX_MSG_SIZE; diff --git a/drivers/net/wan/farsync.c b/drivers/net/wan/farsync.c index 25c9ef6a1815..90c0a317d9d3 100644 --- a/drivers/net/wan/farsync.c +++ b/drivers/net/wan/farsync.c @@ -792,25 +792,6 @@ fst_process_rx_status(int rx_status, char *name) */ break; } - - case NET_RX_CN_LOW: - { - dbg(DBG_ASS, "%s: Receive Low Congestion\n", name); - break; - } - - case NET_RX_CN_MOD: - { - dbg(DBG_ASS, "%s: Receive Moderate Congestion\n", name); - break; - } - - case NET_RX_CN_HIGH: - { - dbg(DBG_ASS, "%s: Receive High Congestion\n", name); - break; - } - case NET_RX_DROP: { dbg(DBG_ASS, "%s: Received packet dropped\n", name); diff --git a/drivers/staging/otus/wrap_pkt.c b/drivers/staging/otus/wrap_pkt.c index 5db0004c8739..89a6b92f5972 100644 --- a/drivers/staging/otus/wrap_pkt.c +++ b/drivers/staging/otus/wrap_pkt.c @@ -156,10 +156,7 @@ void zfLnxRecvEth(zdev_t* dev, zbuf_t* buf, u16_t port) switch(netif_rx(buf)) #endif { - case NET_RX_BAD: case NET_RX_DROP: - case NET_RX_CN_MOD: - case NET_RX_CN_HIGH: break; default: macp->drv_stats.net_stats.rx_packets++; diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index d4a4d9867794..9f25ab2899de 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -72,10 +72,6 @@ struct wireless_dev; /* Backlog congestion levels */ #define NET_RX_SUCCESS 0 /* keep 'em coming, baby */ #define NET_RX_DROP 1 /* packet dropped */ -#define NET_RX_CN_LOW 2 /* storm alert, just in case */ -#define NET_RX_CN_MOD 3 /* Storm on its way! */ -#define NET_RX_CN_HIGH 4 /* The storm is here */ -#define NET_RX_BAD 5 /* packet dropped due to kernel error */ /* NET_XMIT_CN is special. It does not guarantee that this packet is lost. It * indicates that the device will soon be dropping packets, or already drops diff --git a/net/decnet/dn_route.c b/net/decnet/dn_route.c index 1d6ca8a98dc6..9383d3e5a1ab 100644 --- a/net/decnet/dn_route.c +++ b/net/decnet/dn_route.c @@ -774,7 +774,7 @@ static int dn_rt_bug(struct sk_buff *skb) kfree_skb(skb); - return NET_RX_BAD; + return NET_RX_DROP; } static int dn_rt_set_next_hop(struct dn_route *rt, struct dn_fib_res *res) diff --git a/net/lapb/lapb_iface.c b/net/lapb/lapb_iface.c index 2ba1bc4f3c3a..bda96d18fd98 100644 --- a/net/lapb/lapb_iface.c +++ b/net/lapb/lapb_iface.c @@ -407,7 +407,7 @@ int lapb_data_indication(struct lapb_cb *lapb, struct sk_buff *skb) return lapb->callbacks.data_indication(lapb->dev, skb); kfree_skb(skb); - return NET_RX_CN_HIGH; /* For now; must be != NET_RX_DROP */ + return NET_RX_SUCCESS; /* For now; must be != NET_RX_DROP */ } int lapb_data_transmit(struct lapb_cb *lapb, struct sk_buff *skb) -- cgit v1.2.3 From af794c74240d8d788058bdfee339512e7ac029b2 Mon Sep 17 00:00:00 2001 From: Rami Rosen Date: Thu, 25 Jun 2009 04:42:19 +0000 Subject: cleanup: remove unused member in scm_cookie. This patch removes an unused member (seq) scm_cookie; besides initialized to 0 in the header file, it is not used. Signed-off-by: Rami Rosen Signed-off-by: David S. Miller --- include/net/scm.h | 2 -- 1 file changed, 2 deletions(-) (limited to 'include') diff --git a/include/net/scm.h b/include/net/scm.h index f45bb6eca7d4..cf48c800e926 100644 --- a/include/net/scm.h +++ b/include/net/scm.h @@ -26,7 +26,6 @@ struct scm_cookie #ifdef CONFIG_SECURITY_NETWORK u32 secid; /* Passed security ID */ #endif - unsigned long seq; /* Connection seqno */ }; extern void scm_detach_fds(struct msghdr *msg, struct scm_cookie *scm); @@ -59,7 +58,6 @@ static __inline__ int scm_send(struct socket *sock, struct msghdr *msg, scm->creds.gid = current_gid(); scm->creds.pid = task_tgid_vnr(p); scm->fp = NULL; - scm->seq = 0; unix_get_peersec_dgram(sock, scm); if (msg->msg_controllen <= 0) return 0; -- cgit v1.2.3 From 6650613d3387dcc30685e2781818ea7d0f840027 Mon Sep 17 00:00:00 2001 From: "oscar.medina@motorola.com" Date: Tue, 30 Jun 2009 03:25:39 +0000 Subject: tipc: Add socket options to get number of queued messages This patch allows a TIPC application to determine the number of messages currently waiting in a socket's receive queue (TIPC_SOCK_RECVQ_DEPTH) or in all TIPC socket receive queues (TIPC_NODE_RECVQ_DEPTH). Signed-off-by: Oscar Medina Signed-off-by: Allan Stephens Signed-off-by: David S. Miller --- include/linux/tipc.h | 2 ++ net/tipc/socket.c | 6 ++++++ 2 files changed, 8 insertions(+) (limited to 'include') diff --git a/include/linux/tipc.h b/include/linux/tipc.h index bea469455a0c..3d92396639de 100644 --- a/include/linux/tipc.h +++ b/include/linux/tipc.h @@ -209,5 +209,7 @@ struct sockaddr_tipc { #define TIPC_SRC_DROPPABLE 128 /* Default: 0 (resend congested msg) */ #define TIPC_DEST_DROPPABLE 129 /* Default: based on socket type */ #define TIPC_CONN_TIMEOUT 130 /* Default: 8000 (ms) */ +#define TIPC_NODE_RECVQ_DEPTH 131 /* Default: none (read only) */ +#define TIPC_SOCK_RECVQ_DEPTH 132 /* Default: none (read only) */ #endif diff --git a/net/tipc/socket.c b/net/tipc/socket.c index 1848693ebb82..e8254e809b79 100644 --- a/net/tipc/socket.c +++ b/net/tipc/socket.c @@ -1748,6 +1748,12 @@ static int getsockopt(struct socket *sock, value = jiffies_to_msecs(sk->sk_rcvtimeo); /* no need to set "res", since already 0 at this point */ break; + case TIPC_NODE_RECVQ_DEPTH: + value = (u32)atomic_read(&tipc_queue_size); + break; + case TIPC_SOCK_RECVQ_DEPTH: + value = skb_queue_len(&sk->sk_receive_queue); + break; default: res = -EINVAL; } -- cgit v1.2.3 From e04af024b2e74249990587e76ec98220028c01c3 Mon Sep 17 00:00:00 2001 From: Cyrill Gorcunov Date: Fri, 3 Jul 2009 20:11:58 +0000 Subject: net, netns_xt: shrink netns_xt members In case if kernel was compiled without ebtables support there is no need to keep ebt_table pointers in netns_xt structure. Make it config dependent. Signed-off-by: Cyrill Gorcunov Signed-off-by: David S. Miller --- include/net/netns/x_tables.h | 3 +++ 1 file changed, 3 insertions(+) (limited to 'include') diff --git a/include/net/netns/x_tables.h b/include/net/netns/x_tables.h index 9554a644a8f8..591db7d657a3 100644 --- a/include/net/netns/x_tables.h +++ b/include/net/netns/x_tables.h @@ -8,8 +8,11 @@ struct ebt_table; struct netns_xt { struct list_head tables[NFPROTO_NUMPROTO]; +#if defined(CONFIG_BRIDGE_NF_EBTABLES) || \ + defined(CONFIG_BRIDGE_NF_EBTABLES_MODULE) struct ebt_table *broute_table; struct ebt_table *frame_filter; struct ebt_table *frame_nat; +#endif }; #endif -- cgit v1.2.3 From a65e7bfcd74e4c0939f235d2bf9f48ddb3a57991 Mon Sep 17 00:00:00 2001 From: Hui Zhu Date: Sun, 5 Jul 2009 12:08:15 -0700 Subject: elf: fix multithreaded program core dumping on arm Fix the multithread program core thread message error. This issue affects arches with neither has CORE_DUMP_USE_REGSET nor ELF_CORE_COPY_TASK_REGS, ARM is one of them. The thread message of core file is generated in elf_dump_thread_status. The register values is set by elf_core_copy_task_regs in this function. If an arch doesn't define ELF_CORE_COPY_TASK_REGS, elf_core_copy_task_regs() will do nothing. Then the core file will not have the register message of thread. So add elf_core_copy_regs to set regiser values if ELF_CORE_COPY_TASK_REGS doesn't define. The following is how to reproduce this issue: cat 1.c #include #include #include void td1(void * i) { while (1) { printf ("1\n"); sleep (1); } return; } void td2(void * i) { while (1) { printf ("2\n"); sleep (1); } return; } int main(int argc,char *argv[],char *envp[]) { pthread_t t1,t2; pthread_create(&t1, NULL, (void*)td1, NULL); pthread_create(&t2, NULL, (void*)td2, NULL); sleep (10); assert(0); return (0); } arm-xxx-gcc -g -lpthread 1.c -o 1 copy 1.c and 1 to a arm board. Goto this board. ulimit -c 1800000 ./1 # ./1 1 2 1 ... ... 1 1: 1.c:37: main: Assertion `0' failed. Aborted (core dumped) Then you can get a core file. gdb 1 core.xxx Without the patch: (gdb) info threads 3 process 909 0x00000000 in ?? () 2 process 908 0x00000000 in ?? () * 1 process 907 0x4a6e2238 in raise () from /lib/libc.so.6 You can found that the pc of 909 and 908 is 0x00000000. With the patch: (gdb) info threads 3 process 885 0x4a749974 in nanosleep () from /lib/libc.so.6 2 process 884 0x4a749974 in nanosleep () from /lib/libc.so.6 * 1 process 883 0x4a6e2238 in raise () from /lib/libc.so.6 The pc of 885 and 884 is right. Signed-off-by: Hui Zhu Cc: Amerigo Wang Cc: Al Viro Cc: David Howells Cc: Roland McGrath Cc: Jakub Jelinek Cc: Russell King Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/elfcore.h | 2 ++ 1 file changed, 2 insertions(+) (limited to 'include') diff --git a/include/linux/elfcore.h b/include/linux/elfcore.h index 7605c5e9589f..03ec16779802 100644 --- a/include/linux/elfcore.h +++ b/include/linux/elfcore.h @@ -125,6 +125,8 @@ static inline int elf_core_copy_task_regs(struct task_struct *t, elf_gregset_t* #ifdef ELF_CORE_COPY_TASK_REGS return ELF_CORE_COPY_TASK_REGS(t, elfregs); +#else + elf_core_copy_regs(elfregs, task_pt_regs(t)); #endif return 0; } -- cgit v1.2.3 From 82e3310ace59794ecf0f531eca94647b2863dfda Mon Sep 17 00:00:00 2001 From: Tobias Doerffel Date: Sun, 5 Jul 2009 12:08:23 -0700 Subject: linux/sysrq.h needs linux/errno.h In include/linux/sysrq.h the constant EINVAL is being used but is undefined if include/linux/errno.h is not included before. Fix this by adding #include at the beginning. Signed-off-by: Tobias Doerffel Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/sysrq.h | 2 ++ 1 file changed, 2 insertions(+) (limited to 'include') diff --git a/include/linux/sysrq.h b/include/linux/sysrq.h index 98a1d8cfb73d..99adcdc0d3ca 100644 --- a/include/linux/sysrq.h +++ b/include/linux/sysrq.h @@ -14,6 +14,8 @@ #ifndef _LINUX_SYSRQ_H #define _LINUX_SYSRQ_H +#include + struct pt_regs; struct tty_struct; -- cgit v1.2.3 From f2ac72e8268d9559c3114d5a22679f91f80a2238 Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Tue, 7 Jul 2009 12:30:33 +0800 Subject: crypto: api - Add new template create function This patch introduces the template->create function intended to replace the existing alloc function. The intention is for create to handle the registration directly, whereas currently the caller of alloc has to handle the registration. This allows type-specific code to be run prior to registration. Signed-off-by: Herbert Xu --- crypto/algboss.c | 5 +++++ include/crypto/algapi.h | 1 + 2 files changed, 6 insertions(+) (limited to 'include') diff --git a/crypto/algboss.c b/crypto/algboss.c index 9908dd830c26..412241ce4cfa 100644 --- a/crypto/algboss.c +++ b/crypto/algboss.c @@ -68,6 +68,11 @@ static int cryptomgr_probe(void *data) goto err; do { + if (tmpl->create) { + err = tmpl->create(tmpl, param->tb); + continue; + } + inst = tmpl->alloc(param->tb); if (IS_ERR(inst)) err = PTR_ERR(inst); diff --git a/include/crypto/algapi.h b/include/crypto/algapi.h index 010545436efa..ce010a346420 100644 --- a/include/crypto/algapi.h +++ b/include/crypto/algapi.h @@ -52,6 +52,7 @@ struct crypto_template { struct crypto_instance *(*alloc)(struct rtattr **tb); void (*free)(struct crypto_instance *inst); + int (*create)(struct crypto_template *tmpl, struct rtattr **tb); char name[CRYPTO_MAX_ALG_NAME]; }; -- cgit v1.2.3 From 7afdbf23c3acdec3eaf1b94f87132fff3d81ce73 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Tue, 7 Jul 2009 10:23:43 +0200 Subject: signals: declare sys_rt_tgsigqueueinfo in syscalls.h sys_rt_tgsigqueueinfo needs to be declared in linux/syscalls.h so that architectures defining the system call table in C can reference it. Signed-off-by: Arnd Bergmann LKML-Reference: <200907071023.44008.arnd@arndb.de> Signed-off-by: Thomas Gleixner --- include/linux/syscalls.h | 2 ++ 1 file changed, 2 insertions(+) (limited to 'include') diff --git a/include/linux/syscalls.h b/include/linux/syscalls.h index fa4242cdade8..80de7003d8c2 100644 --- a/include/linux/syscalls.h +++ b/include/linux/syscalls.h @@ -321,6 +321,8 @@ asmlinkage long sys_rt_sigtimedwait(const sigset_t __user *uthese, siginfo_t __user *uinfo, const struct timespec __user *uts, size_t sigsetsize); +asmlinkage long sys_rt_tgsigqueueinfo(pid_t tgid, pid_t pid, int sig, + siginfo_t __user *uinfo); asmlinkage long sys_kill(int pid, int sig); asmlinkage long sys_tgkill(int tgid, int pid, int sig); asmlinkage long sys_tkill(int pid, int sig); -- cgit v1.2.3 From 70ec7bb91ad0d6cce84c8e17f8cbb608dda7b18c Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Tue, 7 Jul 2009 14:07:37 +0800 Subject: crypto: api - Add crypto_alloc_instance2 This patch adds a new argument to crypto_alloc_instance which sets aside some space before the instance for use by algorithms such as shash that place type-specific data before crypto_alg. For compatibility the function has been renamed so that existing users aren't affected. Signed-off-by: Herbert Xu --- crypto/algapi.c | 37 +++++++++++++++++++++++++++++++------ include/crypto/algapi.h | 2 ++ 2 files changed, 33 insertions(+), 6 deletions(-) (limited to 'include') diff --git a/crypto/algapi.c b/crypto/algapi.c index 56c62e2858d5..429f1a003b06 100644 --- a/crypto/algapi.c +++ b/crypto/algapi.c @@ -627,17 +627,20 @@ int crypto_attr_u32(struct rtattr *rta, u32 *num) } EXPORT_SYMBOL_GPL(crypto_attr_u32); -struct crypto_instance *crypto_alloc_instance(const char *name, - struct crypto_alg *alg) +void *crypto_alloc_instance2(const char *name, struct crypto_alg *alg, + unsigned int head) { struct crypto_instance *inst; - struct crypto_spawn *spawn; + char *p; int err; - inst = kzalloc(sizeof(*inst) + sizeof(*spawn), GFP_KERNEL); - if (!inst) + p = kzalloc(head + sizeof(*inst) + sizeof(struct crypto_spawn), + GFP_KERNEL); + if (!p) return ERR_PTR(-ENOMEM); + inst = (void *)(p + head); + err = -ENAMETOOLONG; if (snprintf(inst->alg.cra_name, CRYPTO_MAX_ALG_NAME, "%s(%s)", name, alg->cra_name) >= CRYPTO_MAX_ALG_NAME) @@ -647,6 +650,25 @@ struct crypto_instance *crypto_alloc_instance(const char *name, name, alg->cra_driver_name) >= CRYPTO_MAX_ALG_NAME) goto err_free_inst; + return p; + +err_free_inst: + kfree(p); + return ERR_PTR(err); +} +EXPORT_SYMBOL_GPL(crypto_alloc_instance2); + +struct crypto_instance *crypto_alloc_instance(const char *name, + struct crypto_alg *alg) +{ + struct crypto_instance *inst; + struct crypto_spawn *spawn; + int err; + + inst = crypto_alloc_instance2(name, alg, 0); + if (IS_ERR(inst)) + goto out; + spawn = crypto_instance_ctx(inst); err = crypto_init_spawn(spawn, alg, inst, CRYPTO_ALG_TYPE_MASK | CRYPTO_ALG_ASYNC); @@ -658,7 +680,10 @@ struct crypto_instance *crypto_alloc_instance(const char *name, err_free_inst: kfree(inst); - return ERR_PTR(err); + inst = ERR_PTR(err); + +out: + return inst; } EXPORT_SYMBOL_GPL(crypto_alloc_instance); diff --git a/include/crypto/algapi.h b/include/crypto/algapi.h index ce010a346420..99bb29723130 100644 --- a/include/crypto/algapi.h +++ b/include/crypto/algapi.h @@ -132,6 +132,8 @@ int crypto_check_attr_type(struct rtattr **tb, u32 type); const char *crypto_attr_alg_name(struct rtattr *rta); struct crypto_alg *crypto_attr_alg(struct rtattr *rta, u32 type, u32 mask); int crypto_attr_u32(struct rtattr *rta, u32 *num); +void *crypto_alloc_instance2(const char *name, struct crypto_alg *alg, + unsigned int head); struct crypto_instance *crypto_alloc_instance(const char *name, struct crypto_alg *alg); -- cgit v1.2.3 From 7a6d3c8b3049d07123628f2bf57127bba2cc878f Mon Sep 17 00:00:00 2001 From: Csaba Henk Date: Wed, 1 Jul 2009 17:28:41 -0700 Subject: fuse: make the number of max background requests and congestion threshold tunable The practical values for these limits depend on the design of the filesystem server so let userspace set them at initialization time. Signed-off-by: Csaba Henk Signed-off-by: Miklos Szeredi --- fs/fuse/dev.c | 10 +++++----- fs/fuse/fuse_i.h | 12 ++++++------ fs/fuse/inode.c | 14 ++++++++++++++ include/linux/fuse.h | 9 +++++++-- 4 files changed, 32 insertions(+), 13 deletions(-) (limited to 'include') diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index f58ecbc416c8..b152761c1bf6 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -250,7 +250,7 @@ static void queue_request(struct fuse_conn *fc, struct fuse_req *req) static void flush_bg_queue(struct fuse_conn *fc) { - while (fc->active_background < FUSE_MAX_BACKGROUND && + while (fc->active_background < fc->max_background && !list_empty(&fc->bg_queue)) { struct fuse_req *req; @@ -280,11 +280,11 @@ __releases(&fc->lock) list_del(&req->intr_entry); req->state = FUSE_REQ_FINISHED; if (req->background) { - if (fc->num_background == FUSE_MAX_BACKGROUND) { + if (fc->num_background == fc->max_background) { fc->blocked = 0; wake_up_all(&fc->blocked_waitq); } - if (fc->num_background == FUSE_CONGESTION_THRESHOLD && + if (fc->num_background == fc->congestion_threshold && fc->connected && fc->bdi_initialized) { clear_bdi_congested(&fc->bdi, READ); clear_bdi_congested(&fc->bdi, WRITE); @@ -410,9 +410,9 @@ static void fuse_request_send_nowait_locked(struct fuse_conn *fc, { req->background = 1; fc->num_background++; - if (fc->num_background == FUSE_MAX_BACKGROUND) + if (fc->num_background == fc->max_background) fc->blocked = 1; - if (fc->num_background == FUSE_CONGESTION_THRESHOLD && + if (fc->num_background == fc->congestion_threshold && fc->bdi_initialized) { set_bdi_congested(&fc->bdi, READ); set_bdi_congested(&fc->bdi, WRITE); diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h index 52b641fc0faf..6bcfab04396f 100644 --- a/fs/fuse/fuse_i.h +++ b/fs/fuse/fuse_i.h @@ -25,12 +25,6 @@ /** Max number of pages that can be used in a single read request */ #define FUSE_MAX_PAGES_PER_REQ 32 -/** Maximum number of outstanding background requests */ -#define FUSE_MAX_BACKGROUND 12 - -/** Congestion starts at 75% of maximum */ -#define FUSE_CONGESTION_THRESHOLD (FUSE_MAX_BACKGROUND * 75 / 100) - /** Bias for fi->writectr, meaning new writepages must not be sent */ #define FUSE_NOWRITE INT_MIN @@ -349,6 +343,12 @@ struct fuse_conn { /** rbtree of fuse_files waiting for poll events indexed by ph */ struct rb_root polled_files; + /** Maximum number of outstanding background requests */ + unsigned max_background; + + /** Number of background requests at which congestion starts */ + unsigned congestion_threshold; + /** Number of requests currently in the background */ unsigned num_background; diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c index f91ccc4a189d..9aa6f46d0c32 100644 --- a/fs/fuse/inode.c +++ b/fs/fuse/inode.c @@ -32,6 +32,12 @@ DEFINE_MUTEX(fuse_mutex); #define FUSE_DEFAULT_BLKSIZE 512 +/** Maximum number of outstanding background requests */ +#define FUSE_DEFAULT_MAX_BACKGROUND 12 + +/** Congestion starts at 75% of maximum */ +#define FUSE_DEFAULT_CONGESTION_THRESHOLD (FUSE_DEFAULT_MAX_BACKGROUND * 3 / 4) + struct fuse_mount_data { int fd; unsigned rootmode; @@ -517,6 +523,8 @@ void fuse_conn_init(struct fuse_conn *fc) INIT_LIST_HEAD(&fc->bg_queue); INIT_LIST_HEAD(&fc->entry); atomic_set(&fc->num_waiting, 0); + fc->max_background = FUSE_DEFAULT_MAX_BACKGROUND; + fc->congestion_threshold = FUSE_DEFAULT_CONGESTION_THRESHOLD; fc->khctr = 0; fc->polled_files = RB_ROOT; fc->reqctr = 0; @@ -736,6 +744,12 @@ static void process_init_reply(struct fuse_conn *fc, struct fuse_req *req) else { unsigned long ra_pages; + if (arg->minor >= 13) { + if (arg->max_background) + fc->max_background = arg->max_background; + if (arg->congestion_threshold) + fc->congestion_threshold = arg->congestion_threshold; + } if (arg->minor >= 6) { ra_pages = arg->max_readahead / PAGE_CACHE_SIZE; if (arg->flags & FUSE_ASYNC_READ) diff --git a/include/linux/fuse.h b/include/linux/fuse.h index cf593bf9fd32..b3700f0ac268 100644 --- a/include/linux/fuse.h +++ b/include/linux/fuse.h @@ -30,6 +30,10 @@ * - add umask flag to input argument of open, mknod and mkdir * - add notification messages for invalidation of inodes and * directory entries + * + * 7.13 + * - make max number of background requests and congestion threshold + * tunables */ #ifndef _LINUX_FUSE_H @@ -41,7 +45,7 @@ #define FUSE_KERNEL_VERSION 7 /** Minor version number of this interface */ -#define FUSE_KERNEL_MINOR_VERSION 12 +#define FUSE_KERNEL_MINOR_VERSION 13 /** The node ID of the root inode */ #define FUSE_ROOT_ID 1 @@ -427,7 +431,8 @@ struct fuse_init_out { __u32 minor; __u32 max_readahead; __u32 flags; - __u32 unused; + __u16 max_background; + __u16 congestion_threshold; __u32 max_write; }; -- cgit v1.2.3 From 77ae365eca895061c8bf2b2e3ae1d9ea62869739 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Fri, 27 Mar 2009 11:00:29 -0400 Subject: ring-buffer: make lockless This patch converts the ring buffers into a completely lockless buffer recording system. The read side still takes locks since we still serialize readers. But the writers are the ones that must be lockless (those can happen in NMIs). The main change is to the "head_page" pointer. We write to the tail, and read from the head. The "head_page" pointer in the cpu buffer is now just a reference to where to look. The real head page is now kept in the head_page->list->prev->next pointer. That is, in the list head of the previous page we set flags. The list pages are allocated to be aligned such that the lowest significant bits are always zero pointing to the list. This gives us play to put in flags to their pointers. bit 0: set when the page is a head page bit 1: set when the writer is moving the page (for overwrite mode) cmpxchg is used to update the pointer. When the writer wraps the buffer and the tail meets the head, in overwrite mode, the writer must move the head page forward. It first uses cmpxchg to change the pointer flag from 1 to 2. Once this is done, the reader on another CPU will not take the page from the buffer. The writers need to protect against interrupts (we don't bother with disabling interrupts because NMIs are allowed to write too). After the writer sets the pointer flag to 2, it takes care to manage interrupts coming in. This is discribed in detail within the comments of the code. Changes in version 2: - Let reader reset entries value of header page. - Fix tail page passing commit page on reader page test. - Always increment entries and write counter in rb_tail_page_update - Add safety check in rb_set_commit_to_write to break out of infinite loop - add mask in rb_is_reader_page [ Impact: lock free writing to the ring buffer ] Signed-off-by: Steven Rostedt --- include/linux/ring_buffer.h | 1 - kernel/trace/ring_buffer.c | 886 ++++++++++++++++++++++++++++++++++++-------- kernel/trace/trace.c | 3 - 3 files changed, 738 insertions(+), 152 deletions(-) (limited to 'include') diff --git a/include/linux/ring_buffer.h b/include/linux/ring_buffer.h index 29f8599e6bea..7fca71693ae7 100644 --- a/include/linux/ring_buffer.h +++ b/include/linux/ring_buffer.h @@ -170,7 +170,6 @@ unsigned long ring_buffer_overruns(struct ring_buffer *buffer); unsigned long ring_buffer_entries_cpu(struct ring_buffer *buffer, int cpu); unsigned long ring_buffer_overrun_cpu(struct ring_buffer *buffer, int cpu); unsigned long ring_buffer_commit_overrun_cpu(struct ring_buffer *buffer, int cpu); -unsigned long ring_buffer_nmi_dropped_cpu(struct ring_buffer *buffer, int cpu); u64 ring_buffer_time_stamp(struct ring_buffer *buffer, int cpu); void ring_buffer_normalize_time_stamp(struct ring_buffer *buffer, diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c index 7c0168ad6d51..e648ba4f70e0 100644 --- a/kernel/trace/ring_buffer.c +++ b/kernel/trace/ring_buffer.c @@ -322,6 +322,14 @@ struct buffer_data_page { unsigned char data[]; /* data of buffer page */ }; +/* + * Note, the buffer_page list must be first. The buffer pages + * are allocated in cache lines, which means that each buffer + * page will be at the beginning of a cache line, and thus + * the least significant bits will be zero. We use this to + * add flags in the list struct pointers, to make the ring buffer + * lockless. + */ struct buffer_page { struct list_head list; /* list of buffer pages */ local_t write; /* index for next write */ @@ -330,6 +338,21 @@ struct buffer_page { struct buffer_data_page *page; /* Actual data page */ }; +/* + * The buffer page counters, write and entries, must be reset + * atomically when crossing page boundaries. To synchronize this + * update, two counters are inserted into the number. One is + * the actual counter for the write position or count on the page. + * + * The other is a counter of updaters. Before an update happens + * the update partition of the counter is incremented. This will + * allow the updater to update the counter atomically. + * + * The counter is 20 bits, and the state data is 12. + */ +#define RB_WRITE_MASK 0xfffff +#define RB_WRITE_INTCNT (1 << 20) + static void rb_init_page(struct buffer_data_page *bpage) { local_set(&bpage->commit, 0); @@ -403,7 +426,7 @@ int ring_buffer_print_page_header(struct trace_seq *s) struct ring_buffer_per_cpu { int cpu; struct ring_buffer *buffer; - spinlock_t reader_lock; /* serialize readers */ + spinlock_t reader_lock; /* serialize readers */ raw_spinlock_t lock; struct lock_class_key lock_key; struct list_head *pages; @@ -411,13 +434,12 @@ struct ring_buffer_per_cpu { struct buffer_page *tail_page; /* write to tail */ struct buffer_page *commit_page; /* committed pages */ struct buffer_page *reader_page; - unsigned long nmi_dropped; - unsigned long commit_overrun; - unsigned long overrun; - unsigned long read; + local_t commit_overrun; + local_t overrun; local_t entries; local_t committing; local_t commits; + unsigned long read; u64 write_stamp; u64 read_stamp; atomic_t record_disabled; @@ -489,6 +511,385 @@ void ring_buffer_normalize_time_stamp(struct ring_buffer *buffer, } EXPORT_SYMBOL_GPL(ring_buffer_normalize_time_stamp); +/* + * Making the ring buffer lockless makes things tricky. + * Although writes only happen on the CPU that they are on, + * and they only need to worry about interrupts. Reads can + * happen on any CPU. + * + * The reader page is always off the ring buffer, but when the + * reader finishes with a page, it needs to swap its page with + * a new one from the buffer. The reader needs to take from + * the head (writes go to the tail). But if a writer is in overwrite + * mode and wraps, it must push the head page forward. + * + * Here lies the problem. + * + * The reader must be careful to replace only the head page, and + * not another one. As described at the top of the file in the + * ASCII art, the reader sets its old page to point to the next + * page after head. It then sets the page after head to point to + * the old reader page. But if the writer moves the head page + * during this operation, the reader could end up with the tail. + * + * We use cmpxchg to help prevent this race. We also do something + * special with the page before head. We set the LSB to 1. + * + * When the writer must push the page forward, it will clear the + * bit that points to the head page, move the head, and then set + * the bit that points to the new head page. + * + * We also don't want an interrupt coming in and moving the head + * page on another writer. Thus we use the second LSB to catch + * that too. Thus: + * + * head->list->prev->next bit 1 bit 0 + * ------- ------- + * Normal page 0 0 + * Points to head page 0 1 + * New head page 1 0 + * + * Note we can not trust the prev pointer of the head page, because: + * + * +----+ +-----+ +-----+ + * | |------>| T |---X--->| N | + * | |<------| | | | + * +----+ +-----+ +-----+ + * ^ ^ | + * | +-----+ | | + * +----------| R |----------+ | + * | |<-----------+ + * +-----+ + * + * Key: ---X--> HEAD flag set in pointer + * T Tail page + * R Reader page + * N Next page + * + * (see __rb_reserve_next() to see where this happens) + * + * What the above shows is that the reader just swapped out + * the reader page with a page in the buffer, but before it + * could make the new header point back to the new page added + * it was preempted by a writer. The writer moved forward onto + * the new page added by the reader and is about to move forward + * again. + * + * You can see, it is legitimate for the previous pointer of + * the head (or any page) not to point back to itself. But only + * temporarially. + */ + +#define RB_PAGE_NORMAL 0UL +#define RB_PAGE_HEAD 1UL +#define RB_PAGE_UPDATE 2UL + + +#define RB_FLAG_MASK 3UL + +/* PAGE_MOVED is not part of the mask */ +#define RB_PAGE_MOVED 4UL + +/* + * rb_list_head - remove any bit + */ +static struct list_head *rb_list_head(struct list_head *list) +{ + unsigned long val = (unsigned long)list; + + return (struct list_head *)(val & ~RB_FLAG_MASK); +} + +/* + * rb_is_head_page - test if the give page is the head page + * + * Because the reader may move the head_page pointer, we can + * not trust what the head page is (it may be pointing to + * the reader page). But if the next page is a header page, + * its flags will be non zero. + */ +static int inline +rb_is_head_page(struct ring_buffer_per_cpu *cpu_buffer, + struct buffer_page *page, struct list_head *list) +{ + unsigned long val; + + val = (unsigned long)list->next; + + if ((val & ~RB_FLAG_MASK) != (unsigned long)&page->list) + return RB_PAGE_MOVED; + + return val & RB_FLAG_MASK; +} + +/* + * rb_is_reader_page + * + * The unique thing about the reader page, is that, if the + * writer is ever on it, the previous pointer never points + * back to the reader page. + */ +static int rb_is_reader_page(struct buffer_page *page) +{ + struct list_head *list = page->list.prev; + + return rb_list_head(list->next) != &page->list; +} + +/* + * rb_set_list_to_head - set a list_head to be pointing to head. + */ +static void rb_set_list_to_head(struct ring_buffer_per_cpu *cpu_buffer, + struct list_head *list) +{ + unsigned long *ptr; + + ptr = (unsigned long *)&list->next; + *ptr |= RB_PAGE_HEAD; + *ptr &= ~RB_PAGE_UPDATE; +} + +/* + * rb_head_page_activate - sets up head page + */ +static void rb_head_page_activate(struct ring_buffer_per_cpu *cpu_buffer) +{ + struct buffer_page *head; + + head = cpu_buffer->head_page; + if (!head) + return; + + /* + * Set the previous list pointer to have the HEAD flag. + */ + rb_set_list_to_head(cpu_buffer, head->list.prev); +} + +static void rb_list_head_clear(struct list_head *list) +{ + unsigned long *ptr = (unsigned long *)&list->next; + + *ptr &= ~RB_FLAG_MASK; +} + +/* + * rb_head_page_dactivate - clears head page ptr (for free list) + */ +static void +rb_head_page_deactivate(struct ring_buffer_per_cpu *cpu_buffer) +{ + struct list_head *hd; + + /* Go through the whole list and clear any pointers found. */ + rb_list_head_clear(cpu_buffer->pages); + + list_for_each(hd, cpu_buffer->pages) + rb_list_head_clear(hd); +} + +static int rb_head_page_set(struct ring_buffer_per_cpu *cpu_buffer, + struct buffer_page *head, + struct buffer_page *prev, + int old_flag, int new_flag) +{ + struct list_head *list; + unsigned long val = (unsigned long)&head->list; + unsigned long ret; + + list = &prev->list; + + val &= ~RB_FLAG_MASK; + + ret = (unsigned long)cmpxchg(&list->next, + val | old_flag, val | new_flag); + + /* check if the reader took the page */ + if ((ret & ~RB_FLAG_MASK) != val) + return RB_PAGE_MOVED; + + return ret & RB_FLAG_MASK; +} + +static int rb_head_page_set_update(struct ring_buffer_per_cpu *cpu_buffer, + struct buffer_page *head, + struct buffer_page *prev, + int old_flag) +{ + return rb_head_page_set(cpu_buffer, head, prev, + old_flag, RB_PAGE_UPDATE); +} + +static int rb_head_page_set_head(struct ring_buffer_per_cpu *cpu_buffer, + struct buffer_page *head, + struct buffer_page *prev, + int old_flag) +{ + return rb_head_page_set(cpu_buffer, head, prev, + old_flag, RB_PAGE_HEAD); +} + +static int rb_head_page_set_normal(struct ring_buffer_per_cpu *cpu_buffer, + struct buffer_page *head, + struct buffer_page *prev, + int old_flag) +{ + return rb_head_page_set(cpu_buffer, head, prev, + old_flag, RB_PAGE_NORMAL); +} + +static inline void rb_inc_page(struct ring_buffer_per_cpu *cpu_buffer, + struct buffer_page **bpage) +{ + struct list_head *p = rb_list_head((*bpage)->list.next); + + *bpage = list_entry(p, struct buffer_page, list); +} + +static struct buffer_page * +rb_set_head_page(struct ring_buffer_per_cpu *cpu_buffer) +{ + struct buffer_page *head; + struct buffer_page *page; + struct list_head *list; + int i; + + if (RB_WARN_ON(cpu_buffer, !cpu_buffer->head_page)) + return NULL; + + /* sanity check */ + list = cpu_buffer->pages; + if (RB_WARN_ON(cpu_buffer, rb_list_head(list->prev->next) != list)) + return NULL; + + page = head = cpu_buffer->head_page; + /* + * It is possible that the writer moves the header behind + * where we started, and we miss in one loop. + * A second loop should grab the header, but we'll do + * three loops just because I'm paranoid. + */ + for (i = 0; i < 3; i++) { + do { + if (rb_is_head_page(cpu_buffer, page, page->list.prev)) { + cpu_buffer->head_page = page; + return page; + } + rb_inc_page(cpu_buffer, &page); + } while (page != head); + } + + RB_WARN_ON(cpu_buffer, 1); + + return NULL; +} + +static int rb_head_page_replace(struct buffer_page *old, + struct buffer_page *new) +{ + unsigned long *ptr = (unsigned long *)&old->list.prev->next; + unsigned long val; + unsigned long ret; + + val = *ptr & ~RB_FLAG_MASK; + val |= RB_PAGE_HEAD; + + ret = cmpxchg(ptr, val, &new->list); + + return ret == val; +} + +/* + * rb_tail_page_update - move the tail page forward + * + * Returns 1 if moved tail page, 0 if someone else did. + */ +static int rb_tail_page_update(struct ring_buffer_per_cpu *cpu_buffer, + struct buffer_page *tail_page, + struct buffer_page *next_page) +{ + struct buffer_page *old_tail; + unsigned long old_entries; + unsigned long old_write; + int ret = 0; + + /* + * The tail page now needs to be moved forward. + * + * We need to reset the tail page, but without messing + * with possible erasing of data brought in by interrupts + * that have moved the tail page and are currently on it. + * + * We add a counter to the write field to denote this. + */ + old_write = local_add_return(RB_WRITE_INTCNT, &next_page->write); + old_entries = local_add_return(RB_WRITE_INTCNT, &next_page->entries); + + /* + * Just make sure we have seen our old_write and synchronize + * with any interrupts that come in. + */ + barrier(); + + /* + * If the tail page is still the same as what we think + * it is, then it is up to us to update the tail + * pointer. + */ + if (tail_page == cpu_buffer->tail_page) { + /* Zero the write counter */ + unsigned long val = old_write & ~RB_WRITE_MASK; + unsigned long eval = old_entries & ~RB_WRITE_MASK; + + /* + * This will only succeed if an interrupt did + * not come in and change it. In which case, we + * do not want to modify it. + */ + local_cmpxchg(&next_page->write, old_write, val); + local_cmpxchg(&next_page->entries, old_entries, eval); + + /* + * No need to worry about races with clearing out the commit. + * it only can increment when a commit takes place. But that + * only happens in the outer most nested commit. + */ + local_set(&next_page->page->commit, 0); + + old_tail = cmpxchg(&cpu_buffer->tail_page, + tail_page, next_page); + + if (old_tail == tail_page) + ret = 1; + } + + return ret; +} + +static int rb_check_bpage(struct ring_buffer_per_cpu *cpu_buffer, + struct buffer_page *bpage) +{ + unsigned long val = (unsigned long)bpage; + + if (RB_WARN_ON(cpu_buffer, val & RB_FLAG_MASK)) + return 1; + + return 0; +} + +/** + * rb_check_list - make sure a pointer to a list has the last bits zero + */ +static int rb_check_list(struct ring_buffer_per_cpu *cpu_buffer, + struct list_head *list) +{ + if (RB_WARN_ON(cpu_buffer, rb_list_head(list->prev) != list->prev)) + return 1; + if (RB_WARN_ON(cpu_buffer, rb_list_head(list->next) != list->next)) + return 1; + return 0; +} + /** * check_pages - integrity check of buffer pages * @cpu_buffer: CPU buffer with pages to test @@ -501,11 +902,16 @@ static int rb_check_pages(struct ring_buffer_per_cpu *cpu_buffer) struct list_head *head = cpu_buffer->pages; struct buffer_page *bpage, *tmp; + rb_head_page_deactivate(cpu_buffer); + if (RB_WARN_ON(cpu_buffer, head->next->prev != head)) return -1; if (RB_WARN_ON(cpu_buffer, head->prev->next != head)) return -1; + if (rb_check_list(cpu_buffer, head)) + return -1; + list_for_each_entry_safe(bpage, tmp, head, list) { if (RB_WARN_ON(cpu_buffer, bpage->list.next->prev != &bpage->list)) @@ -513,8 +919,12 @@ static int rb_check_pages(struct ring_buffer_per_cpu *cpu_buffer) if (RB_WARN_ON(cpu_buffer, bpage->list.prev->next != &bpage->list)) return -1; + if (rb_check_list(cpu_buffer, &bpage->list)) + return -1; } + rb_head_page_activate(cpu_buffer); + return 0; } @@ -533,6 +943,9 @@ static int rb_allocate_pages(struct ring_buffer_per_cpu *cpu_buffer, GFP_KERNEL, cpu_to_node(cpu_buffer->cpu)); if (!bpage) goto free_pages; + + rb_check_bpage(cpu_buffer, bpage); + list_add(&bpage->list, &pages); addr = __get_free_page(GFP_KERNEL); @@ -586,6 +999,8 @@ rb_allocate_cpu_buffer(struct ring_buffer *buffer, int cpu) if (!bpage) goto fail_free_buffer; + rb_check_bpage(cpu_buffer, bpage); + cpu_buffer->reader_page = bpage; addr = __get_free_page(GFP_KERNEL); if (!addr) @@ -603,6 +1018,8 @@ rb_allocate_cpu_buffer(struct ring_buffer *buffer, int cpu) = list_entry(cpu_buffer->pages, struct buffer_page, list); cpu_buffer->tail_page = cpu_buffer->commit_page = cpu_buffer->head_page; + rb_head_page_activate(cpu_buffer); + return cpu_buffer; fail_free_reader: @@ -620,6 +1037,8 @@ static void rb_free_cpu_buffer(struct ring_buffer_per_cpu *cpu_buffer) free_buffer_page(cpu_buffer->reader_page); + rb_head_page_deactivate(cpu_buffer); + if (head) { list_for_each_entry_safe(bpage, tmp, head, list) { list_del_init(&bpage->list); @@ -770,6 +1189,8 @@ rb_remove_pages(struct ring_buffer_per_cpu *cpu_buffer, unsigned nr_pages) atomic_inc(&cpu_buffer->record_disabled); synchronize_sched(); + rb_head_page_deactivate(cpu_buffer); + for (i = 0; i < nr_pages; i++) { if (RB_WARN_ON(cpu_buffer, list_empty(cpu_buffer->pages))) return; @@ -800,6 +1221,9 @@ rb_insert_pages(struct ring_buffer_per_cpu *cpu_buffer, atomic_inc(&cpu_buffer->record_disabled); synchronize_sched(); + spin_lock_irq(&cpu_buffer->reader_lock); + rb_head_page_deactivate(cpu_buffer); + for (i = 0; i < nr_pages; i++) { if (RB_WARN_ON(cpu_buffer, list_empty(pages))) return; @@ -809,6 +1233,7 @@ rb_insert_pages(struct ring_buffer_per_cpu *cpu_buffer, list_add_tail(&bpage->list, cpu_buffer->pages); } rb_reset_cpu(cpu_buffer); + spin_unlock_irq(&cpu_buffer->reader_lock); rb_check_pages(cpu_buffer); @@ -958,22 +1383,15 @@ rb_reader_event(struct ring_buffer_per_cpu *cpu_buffer) cpu_buffer->reader_page->read); } -static inline struct ring_buffer_event * -rb_head_event(struct ring_buffer_per_cpu *cpu_buffer) -{ - return __rb_page_index(cpu_buffer->head_page, - cpu_buffer->head_page->read); -} - static inline struct ring_buffer_event * rb_iter_head_event(struct ring_buffer_iter *iter) { return __rb_page_index(iter->head_page, iter->head); } -static inline unsigned rb_page_write(struct buffer_page *bpage) +static inline unsigned long rb_page_write(struct buffer_page *bpage) { - return local_read(&bpage->write); + return local_read(&bpage->write) & RB_WRITE_MASK; } static inline unsigned rb_page_commit(struct buffer_page *bpage) @@ -981,6 +1399,11 @@ static inline unsigned rb_page_commit(struct buffer_page *bpage) return local_read(&bpage->page->commit); } +static inline unsigned long rb_page_entries(struct buffer_page *bpage) +{ + return local_read(&bpage->entries) & RB_WRITE_MASK; +} + /* Size is determined by what has been commited */ static inline unsigned rb_page_size(struct buffer_page *bpage) { @@ -993,19 +1416,6 @@ rb_commit_index(struct ring_buffer_per_cpu *cpu_buffer) return rb_page_commit(cpu_buffer->commit_page); } -static inline unsigned rb_head_size(struct ring_buffer_per_cpu *cpu_buffer) -{ - return rb_page_commit(cpu_buffer->head_page); -} - -static inline void rb_inc_page(struct ring_buffer_per_cpu *cpu_buffer, - struct buffer_page **bpage) -{ - struct list_head *p = (*bpage)->list.next; - - *bpage = list_entry(p, struct buffer_page, list); -} - static inline unsigned rb_event_index(struct ring_buffer_event *event) { @@ -1031,6 +1441,8 @@ rb_event_is_commit(struct ring_buffer_per_cpu *cpu_buffer, static void rb_set_commit_to_write(struct ring_buffer_per_cpu *cpu_buffer) { + unsigned long max_count; + /* * We only race with interrupts and NMIs on this CPU. * If we own the commit event, then we can commit @@ -1040,9 +1452,16 @@ rb_set_commit_to_write(struct ring_buffer_per_cpu *cpu_buffer) * assign the commit to the tail. */ again: + max_count = cpu_buffer->buffer->pages * 100; + while (cpu_buffer->commit_page != cpu_buffer->tail_page) { - cpu_buffer->commit_page->page->commit = - cpu_buffer->commit_page->write; + if (RB_WARN_ON(cpu_buffer, !(--max_count))) + return; + if (RB_WARN_ON(cpu_buffer, + rb_is_reader_page(cpu_buffer->tail_page))) + return; + local_set(&cpu_buffer->commit_page->page->commit, + rb_page_write(cpu_buffer->commit_page)); rb_inc_page(cpu_buffer, &cpu_buffer->commit_page); cpu_buffer->write_stamp = cpu_buffer->commit_page->page->time_stamp; @@ -1051,8 +1470,12 @@ rb_set_commit_to_write(struct ring_buffer_per_cpu *cpu_buffer) } while (rb_commit_index(cpu_buffer) != rb_page_write(cpu_buffer->commit_page)) { - cpu_buffer->commit_page->page->commit = - cpu_buffer->commit_page->write; + + local_set(&cpu_buffer->commit_page->page->commit, + rb_page_write(cpu_buffer->commit_page)); + RB_WARN_ON(cpu_buffer, + local_read(&cpu_buffer->commit_page->page->commit) & + ~RB_WRITE_MASK); barrier(); } @@ -1085,7 +1508,7 @@ static void rb_inc_iter(struct ring_buffer_iter *iter) * to the head page instead of next. */ if (iter->head_page == cpu_buffer->reader_page) - iter->head_page = cpu_buffer->head_page; + iter->head_page = rb_set_head_page(cpu_buffer); else rb_inc_page(cpu_buffer, &iter->head_page); @@ -1129,6 +1552,163 @@ rb_update_event(struct ring_buffer_event *event, } } +/* + * rb_handle_head_page - writer hit the head page + * + * Returns: +1 to retry page + * 0 to continue + * -1 on error + */ +static int +rb_handle_head_page(struct ring_buffer_per_cpu *cpu_buffer, + struct buffer_page *tail_page, + struct buffer_page *next_page) +{ + struct buffer_page *new_head; + int entries; + int type; + int ret; + + entries = rb_page_entries(next_page); + + /* + * The hard part is here. We need to move the head + * forward, and protect against both readers on + * other CPUs and writers coming in via interrupts. + */ + type = rb_head_page_set_update(cpu_buffer, next_page, tail_page, + RB_PAGE_HEAD); + + /* + * type can be one of four: + * NORMAL - an interrupt already moved it for us + * HEAD - we are the first to get here. + * UPDATE - we are the interrupt interrupting + * a current move. + * MOVED - a reader on another CPU moved the next + * pointer to its reader page. Give up + * and try again. + */ + + switch (type) { + case RB_PAGE_HEAD: + /* + * We changed the head to UPDATE, thus + * it is our responsibility to update + * the counters. + */ + local_add(entries, &cpu_buffer->overrun); + + /* + * The entries will be zeroed out when we move the + * tail page. + */ + + /* still more to do */ + break; + + case RB_PAGE_UPDATE: + /* + * This is an interrupt that interrupt the + * previous update. Still more to do. + */ + break; + case RB_PAGE_NORMAL: + /* + * An interrupt came in before the update + * and processed this for us. + * Nothing left to do. + */ + return 1; + case RB_PAGE_MOVED: + /* + * The reader is on another CPU and just did + * a swap with our next_page. + * Try again. + */ + return 1; + default: + RB_WARN_ON(cpu_buffer, 1); /* WTF??? */ + return -1; + } + + /* + * Now that we are here, the old head pointer is + * set to UPDATE. This will keep the reader from + * swapping the head page with the reader page. + * The reader (on another CPU) will spin till + * we are finished. + * + * We just need to protect against interrupts + * doing the job. We will set the next pointer + * to HEAD. After that, we set the old pointer + * to NORMAL, but only if it was HEAD before. + * otherwise we are an interrupt, and only + * want the outer most commit to reset it. + */ + new_head = next_page; + rb_inc_page(cpu_buffer, &new_head); + + ret = rb_head_page_set_head(cpu_buffer, new_head, next_page, + RB_PAGE_NORMAL); + + /* + * Valid returns are: + * HEAD - an interrupt came in and already set it. + * NORMAL - One of two things: + * 1) We really set it. + * 2) A bunch of interrupts came in and moved + * the page forward again. + */ + switch (ret) { + case RB_PAGE_HEAD: + case RB_PAGE_NORMAL: + /* OK */ + break; + default: + RB_WARN_ON(cpu_buffer, 1); + return -1; + } + + /* + * It is possible that an interrupt came in, + * set the head up, then more interrupts came in + * and moved it again. When we get back here, + * the page would have been set to NORMAL but we + * just set it back to HEAD. + * + * How do you detect this? Well, if that happened + * the tail page would have moved. + */ + if (ret == RB_PAGE_NORMAL) { + /* + * If the tail had moved passed next, then we need + * to reset the pointer. + */ + if (cpu_buffer->tail_page != tail_page && + cpu_buffer->tail_page != next_page) + rb_head_page_set_normal(cpu_buffer, new_head, + next_page, + RB_PAGE_HEAD); + } + + /* + * If this was the outer most commit (the one that + * changed the original pointer from HEAD to UPDATE), + * then it is up to us to reset it to NORMAL. + */ + if (type == RB_PAGE_HEAD) { + ret = rb_head_page_set_normal(cpu_buffer, next_page, + tail_page, + RB_PAGE_UPDATE); + if (RB_WARN_ON(cpu_buffer, + ret != RB_PAGE_UPDATE)) + return -1; + } + + return 0; +} + static unsigned rb_calculate_event_length(unsigned length) { struct ring_buffer_event event; /* Used only for sizeof array */ @@ -1207,96 +1787,93 @@ rb_move_tail(struct ring_buffer_per_cpu *cpu_buffer, struct buffer_page *commit_page, struct buffer_page *tail_page, u64 *ts) { - struct buffer_page *next_page, *head_page, *reader_page; struct ring_buffer *buffer = cpu_buffer->buffer; - bool lock_taken = false; - unsigned long flags; + struct buffer_page *next_page; + int ret; next_page = tail_page; - local_irq_save(flags); - /* - * Since the write to the buffer is still not - * fully lockless, we must be careful with NMIs. - * The locks in the writers are taken when a write - * crosses to a new page. The locks protect against - * races with the readers (this will soon be fixed - * with a lockless solution). - * - * Because we can not protect against NMIs, and we - * want to keep traces reentrant, we need to manage - * what happens when we are in an NMI. - * - * NMIs can happen after we take the lock. - * If we are in an NMI, only take the lock - * if it is not already taken. Otherwise - * simply fail. - */ - if (unlikely(in_nmi())) { - if (!__raw_spin_trylock(&cpu_buffer->lock)) { - cpu_buffer->nmi_dropped++; - goto out_reset; - } - } else - __raw_spin_lock(&cpu_buffer->lock); - - lock_taken = true; - rb_inc_page(cpu_buffer, &next_page); - head_page = cpu_buffer->head_page; - reader_page = cpu_buffer->reader_page; - - /* we grabbed the lock before incrementing */ - if (RB_WARN_ON(cpu_buffer, next_page == reader_page)) - goto out_reset; - /* * If for some reason, we had an interrupt storm that made * it all the way around the buffer, bail, and warn * about it. */ if (unlikely(next_page == commit_page)) { - cpu_buffer->commit_overrun++; + local_inc(&cpu_buffer->commit_overrun); goto out_reset; } - if (next_page == head_page) { - if (!(buffer->flags & RB_FL_OVERWRITE)) - goto out_reset; - - /* tail_page has not moved yet? */ - if (tail_page == cpu_buffer->tail_page) { - /* count overflows */ - cpu_buffer->overrun += - local_read(&head_page->entries); + /* + * This is where the fun begins! + * + * We are fighting against races between a reader that + * could be on another CPU trying to swap its reader + * page with the buffer head. + * + * We are also fighting against interrupts coming in and + * moving the head or tail on us as well. + * + * If the next page is the head page then we have filled + * the buffer, unless the commit page is still on the + * reader page. + */ + if (rb_is_head_page(cpu_buffer, next_page, &tail_page->list)) { - rb_inc_page(cpu_buffer, &head_page); - cpu_buffer->head_page = head_page; - cpu_buffer->head_page->read = 0; + /* + * If the commit is not on the reader page, then + * move the header page. + */ + if (!rb_is_reader_page(cpu_buffer->commit_page)) { + /* + * If we are not in overwrite mode, + * this is easy, just stop here. + */ + if (!(buffer->flags & RB_FL_OVERWRITE)) + goto out_reset; + + ret = rb_handle_head_page(cpu_buffer, + tail_page, + next_page); + if (ret < 0) + goto out_reset; + if (ret) + goto out_again; + } else { + /* + * We need to be careful here too. The + * commit page could still be on the reader + * page. We could have a small buffer, and + * have filled up the buffer with events + * from interrupts and such, and wrapped. + * + * Note, if the tail page is also the on the + * reader_page, we let it move out. + */ + if (unlikely((cpu_buffer->commit_page != + cpu_buffer->tail_page) && + (cpu_buffer->commit_page == + cpu_buffer->reader_page))) { + local_inc(&cpu_buffer->commit_overrun); + goto out_reset; + } } } - /* - * If the tail page is still the same as what we think - * it is, then it is up to us to update the tail - * pointer. - */ - if (tail_page == cpu_buffer->tail_page) { - local_set(&next_page->write, 0); - local_set(&next_page->entries, 0); - local_set(&next_page->page->commit, 0); - cpu_buffer->tail_page = next_page; - - /* reread the time stamp */ + ret = rb_tail_page_update(cpu_buffer, tail_page, next_page); + if (ret) { + /* + * Nested commits always have zero deltas, so + * just reread the time stamp + */ *ts = rb_time_stamp(buffer, cpu_buffer->cpu); - cpu_buffer->tail_page->page->time_stamp = *ts; + next_page->page->time_stamp = *ts; } - rb_reset_tail(cpu_buffer, tail_page, tail, length); + out_again: - __raw_spin_unlock(&cpu_buffer->lock); - local_irq_restore(flags); + rb_reset_tail(cpu_buffer, tail_page, tail, length); /* fail and let the caller try again */ return ERR_PTR(-EAGAIN); @@ -1305,9 +1882,6 @@ rb_move_tail(struct ring_buffer_per_cpu *cpu_buffer, /* reset write */ rb_reset_tail(cpu_buffer, tail_page, tail, length); - if (likely(lock_taken)) - __raw_spin_unlock(&cpu_buffer->lock); - local_irq_restore(flags); return NULL; } @@ -1324,6 +1898,9 @@ __rb_reserve_next(struct ring_buffer_per_cpu *cpu_buffer, barrier(); tail_page = cpu_buffer->tail_page; write = local_add_return(length, &tail_page->write); + + /* set write to only the index of the write */ + write &= RB_WRITE_MASK; tail = write - length; /* See if we shot pass the end of this buffer page */ @@ -1368,12 +1945,16 @@ rb_try_to_discard(struct ring_buffer_per_cpu *cpu_buffer, bpage = cpu_buffer->tail_page; if (bpage->page == (void *)addr && rb_page_write(bpage) == old_index) { + unsigned long write_mask = + local_read(&bpage->write) & ~RB_WRITE_MASK; /* * This is on the tail page. It is possible that * a write could come in and move the tail page * and write to the next page. That is fine * because we just shorten what is on this page. */ + old_index += write_mask; + new_index += write_mask; index = local_cmpxchg(&bpage->write, old_index, new_index); if (index == old_index) return 1; @@ -1882,9 +2463,13 @@ EXPORT_SYMBOL_GPL(ring_buffer_write); static int rb_per_cpu_empty(struct ring_buffer_per_cpu *cpu_buffer) { struct buffer_page *reader = cpu_buffer->reader_page; - struct buffer_page *head = cpu_buffer->head_page; + struct buffer_page *head = rb_set_head_page(cpu_buffer); struct buffer_page *commit = cpu_buffer->commit_page; + /* In case of error, head will be NULL */ + if (unlikely(!head)) + return 1; + return reader->read == rb_page_commit(reader) && (commit == reader || (commit == head && @@ -1975,7 +2560,7 @@ unsigned long ring_buffer_entries_cpu(struct ring_buffer *buffer, int cpu) return 0; cpu_buffer = buffer->buffers[cpu]; - ret = (local_read(&cpu_buffer->entries) - cpu_buffer->overrun) + ret = (local_read(&cpu_buffer->entries) - local_read(&cpu_buffer->overrun)) - cpu_buffer->read; return ret; @@ -1996,32 +2581,12 @@ unsigned long ring_buffer_overrun_cpu(struct ring_buffer *buffer, int cpu) return 0; cpu_buffer = buffer->buffers[cpu]; - ret = cpu_buffer->overrun; + ret = local_read(&cpu_buffer->overrun); return ret; } EXPORT_SYMBOL_GPL(ring_buffer_overrun_cpu); -/** - * ring_buffer_nmi_dropped_cpu - get the number of nmis that were dropped - * @buffer: The ring buffer - * @cpu: The per CPU buffer to get the number of overruns from - */ -unsigned long ring_buffer_nmi_dropped_cpu(struct ring_buffer *buffer, int cpu) -{ - struct ring_buffer_per_cpu *cpu_buffer; - unsigned long ret; - - if (!cpumask_test_cpu(cpu, buffer->cpumask)) - return 0; - - cpu_buffer = buffer->buffers[cpu]; - ret = cpu_buffer->nmi_dropped; - - return ret; -} -EXPORT_SYMBOL_GPL(ring_buffer_nmi_dropped_cpu); - /** * ring_buffer_commit_overrun_cpu - get the number of overruns caused by commits * @buffer: The ring buffer @@ -2037,7 +2602,7 @@ ring_buffer_commit_overrun_cpu(struct ring_buffer *buffer, int cpu) return 0; cpu_buffer = buffer->buffers[cpu]; - ret = cpu_buffer->commit_overrun; + ret = local_read(&cpu_buffer->commit_overrun); return ret; } @@ -2060,7 +2625,7 @@ unsigned long ring_buffer_entries(struct ring_buffer *buffer) for_each_buffer_cpu(buffer, cpu) { cpu_buffer = buffer->buffers[cpu]; entries += (local_read(&cpu_buffer->entries) - - cpu_buffer->overrun) - cpu_buffer->read; + local_read(&cpu_buffer->overrun)) - cpu_buffer->read; } return entries; @@ -2083,7 +2648,7 @@ unsigned long ring_buffer_overruns(struct ring_buffer *buffer) /* if you care about this being correct, lock the buffer */ for_each_buffer_cpu(buffer, cpu) { cpu_buffer = buffer->buffers[cpu]; - overruns += cpu_buffer->overrun; + overruns += local_read(&cpu_buffer->overrun); } return overruns; @@ -2096,8 +2661,10 @@ static void rb_iter_reset(struct ring_buffer_iter *iter) /* Iterator usage is expected to have record disabled */ if (list_empty(&cpu_buffer->reader_page->list)) { - iter->head_page = cpu_buffer->head_page; - iter->head = cpu_buffer->head_page->read; + iter->head_page = rb_set_head_page(cpu_buffer); + if (unlikely(!iter->head_page)) + return; + iter->head = iter->head_page->read; } else { iter->head_page = cpu_buffer->reader_page; iter->head = cpu_buffer->reader_page->read; @@ -2214,6 +2781,7 @@ rb_get_reader_page(struct ring_buffer_per_cpu *cpu_buffer) struct buffer_page *reader = NULL; unsigned long flags; int nr_loops = 0; + int ret; local_irq_save(flags); __raw_spin_lock(&cpu_buffer->lock); @@ -2247,11 +2815,17 @@ rb_get_reader_page(struct ring_buffer_per_cpu *cpu_buffer) goto out; /* - * Splice the empty reader page into the list around the head. * Reset the reader page to size zero. */ + local_set(&cpu_buffer->reader_page->write, 0); + local_set(&cpu_buffer->reader_page->entries, 0); + local_set(&cpu_buffer->reader_page->page->commit, 0); - reader = cpu_buffer->head_page; + spin: + /* + * Splice the empty reader page into the list around the head. + */ + reader = rb_set_head_page(cpu_buffer); cpu_buffer->reader_page->list.next = reader->list.next; cpu_buffer->reader_page->list.prev = reader->list.prev; @@ -2262,22 +2836,35 @@ rb_get_reader_page(struct ring_buffer_per_cpu *cpu_buffer) */ cpu_buffer->pages = reader->list.prev; - local_set(&cpu_buffer->reader_page->write, 0); - local_set(&cpu_buffer->reader_page->entries, 0); - local_set(&cpu_buffer->reader_page->page->commit, 0); + /* The reader page will be pointing to the new head */ + rb_set_list_to_head(cpu_buffer, &cpu_buffer->reader_page->list); - /* Make the reader page now replace the head */ - reader->list.prev->next = &cpu_buffer->reader_page->list; - reader->list.next->prev = &cpu_buffer->reader_page->list; + /* + * Here's the tricky part. + * + * We need to move the pointer past the header page. + * But we can only do that if a writer is not currently + * moving it. The page before the header page has the + * flag bit '1' set if it is pointing to the page we want. + * but if the writer is in the process of moving it + * than it will be '2' or already moved '0'. + */ + + ret = rb_head_page_replace(reader, cpu_buffer->reader_page); /* - * If the tail is on the reader, then we must set the head - * to the inserted page, otherwise we set it one before. + * If we did not convert it, then we must try again. */ - cpu_buffer->head_page = cpu_buffer->reader_page; + if (!ret) + goto spin; - if (cpu_buffer->commit_page != reader) - rb_inc_page(cpu_buffer, &cpu_buffer->head_page); + /* + * Yeah! We succeeded in replacing the page. + * + * Now make the new head point back to the reader page. + */ + reader->list.next->prev = &cpu_buffer->reader_page->list; + rb_inc_page(cpu_buffer, &cpu_buffer->head_page); /* Finally update the reader page to the new head */ cpu_buffer->reader_page = reader; @@ -2733,6 +3320,8 @@ EXPORT_SYMBOL_GPL(ring_buffer_size); static void rb_reset_cpu(struct ring_buffer_per_cpu *cpu_buffer) { + rb_head_page_deactivate(cpu_buffer); + cpu_buffer->head_page = list_entry(cpu_buffer->pages, struct buffer_page, list); local_set(&cpu_buffer->head_page->write, 0); @@ -2750,16 +3339,17 @@ rb_reset_cpu(struct ring_buffer_per_cpu *cpu_buffer) local_set(&cpu_buffer->reader_page->page->commit, 0); cpu_buffer->reader_page->read = 0; - cpu_buffer->nmi_dropped = 0; - cpu_buffer->commit_overrun = 0; - cpu_buffer->overrun = 0; - cpu_buffer->read = 0; + local_set(&cpu_buffer->commit_overrun, 0); + local_set(&cpu_buffer->overrun, 0); local_set(&cpu_buffer->entries, 0); local_set(&cpu_buffer->committing, 0); local_set(&cpu_buffer->commits, 0); + cpu_buffer->read = 0; cpu_buffer->write_stamp = 0; cpu_buffer->read_stamp = 0; + + rb_head_page_activate(cpu_buffer); } /** @@ -3107,7 +3697,7 @@ int ring_buffer_read_page(struct ring_buffer *buffer, read = 0; } else { /* update the entry counter */ - cpu_buffer->read += local_read(&reader->entries); + cpu_buffer->read += rb_page_entries(reader); /* swap the pages */ rb_init_page(bpage); diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index bdb3afc8b306..b591f7a1bd7b 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -3630,9 +3630,6 @@ tracing_stats_read(struct file *filp, char __user *ubuf, cnt = ring_buffer_commit_overrun_cpu(tr->buffer, cpu); trace_seq_printf(s, "commit overrun: %ld\n", cnt); - cnt = ring_buffer_nmi_dropped_cpu(tr->buffer, cpu); - trace_seq_printf(s, "nmi dropped: %ld\n", cnt); - count = simple_read_from_buffer(ubuf, count, ppos, s->buffer, s->len); kfree(s); -- cgit v1.2.3 From 2e4fddd8e420e8f531a34e7a97f9cdb851a6ad13 Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Tue, 7 Jul 2009 15:17:12 +0800 Subject: crypto: shash - Add shash_instance This patch adds shash_instance and the associated alloc/free functions. This is meant to be an instance that with a shash algorithm under it. Note that the instance itself doesn't have to be shash. Signed-off-by: Herbert Xu --- crypto/shash.c | 7 +++++++ include/crypto/internal/hash.h | 26 ++++++++++++++++++++++++++ 2 files changed, 33 insertions(+) (limited to 'include') diff --git a/crypto/shash.c b/crypto/shash.c index 2ccc8b0076ce..8f9d8831e293 100644 --- a/crypto/shash.c +++ b/crypto/shash.c @@ -502,5 +502,12 @@ int crypto_unregister_shash(struct shash_alg *alg) } EXPORT_SYMBOL_GPL(crypto_unregister_shash); +void shash_free_instance(struct crypto_instance *inst) +{ + crypto_drop_spawn(crypto_instance_ctx(inst)); + kfree(shash_instance(inst)); +} +EXPORT_SYMBOL_GPL(shash_free_instance); + MODULE_LICENSE("GPL"); MODULE_DESCRIPTION("Synchronous cryptographic hash type"); diff --git a/include/crypto/internal/hash.h b/include/crypto/internal/hash.h index 82b70564bcab..44d3bcdce6e2 100644 --- a/include/crypto/internal/hash.h +++ b/include/crypto/internal/hash.h @@ -34,6 +34,10 @@ struct crypto_hash_walk { unsigned int flags; }; +struct shash_instance { + struct shash_alg alg; +}; + extern const struct crypto_type crypto_ahash_type; int crypto_hash_walk_done(struct crypto_hash_walk *walk, int err); @@ -46,6 +50,8 @@ int crypto_hash_walk_first_compat(struct hash_desc *hdesc, int crypto_register_shash(struct shash_alg *alg); int crypto_unregister_shash(struct shash_alg *alg); +void shash_free_instance(struct crypto_instance *inst); + static inline void *crypto_ahash_ctx(struct crypto_ahash *tfm) { return crypto_tfm_ctx(&tfm->base); @@ -80,5 +86,25 @@ static inline void *crypto_shash_ctx(struct crypto_shash *tfm) return crypto_tfm_ctx(&tfm->base); } +static inline struct crypto_instance *shash_crypto_instance( + struct shash_instance *inst) +{ + return container_of(&inst->alg.base, struct crypto_instance, alg); +} + +static inline struct shash_instance *shash_instance( + struct crypto_instance *inst) +{ + return container_of(__crypto_shash_alg(&inst->alg), + struct shash_instance, alg); +} + +static inline struct shash_instance *shash_alloc_instance( + const char *name, struct crypto_alg *alg) +{ + return crypto_alloc_instance2(name, alg, + sizeof(struct shash_alg) - sizeof(*alg)); +} + #endif /* _CRYPTO_INTERNAL_HASH_H */ -- cgit v1.2.3 From 97eedce1a64a57648ac5e39f03825528c47ba72e Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Wed, 8 Jul 2009 15:55:52 +0800 Subject: crypto: api - Add new style spawn support This patch modifies the spawn infrastructure to support new style algorithms like shash. In particular, this means storing the frontend type in the spawn and using crypto_create_tfm to allocate the tfm. Signed-off-by: Herbert Xu --- crypto/algapi.c | 55 ++++++++++++++++++++++++++++++++++++++++++++++--- include/crypto/algapi.h | 6 ++++++ 2 files changed, 58 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/crypto/algapi.c b/crypto/algapi.c index 429f1a003b06..17a5fff8f777 100644 --- a/crypto/algapi.c +++ b/crypto/algapi.c @@ -488,6 +488,23 @@ int crypto_init_spawn(struct crypto_spawn *spawn, struct crypto_alg *alg, } EXPORT_SYMBOL_GPL(crypto_init_spawn); +int crypto_init_spawn2(struct crypto_spawn *spawn, struct crypto_alg *alg, + struct crypto_instance *inst, + const struct crypto_type *frontend) +{ + int err = -EINVAL; + + if (frontend && (alg->cra_flags ^ frontend->type) & frontend->maskset) + goto out; + + spawn->frontend = frontend; + err = crypto_init_spawn(spawn, alg, inst, frontend->maskset); + +out: + return err; +} +EXPORT_SYMBOL_GPL(crypto_init_spawn2); + void crypto_drop_spawn(struct crypto_spawn *spawn) { down_write(&crypto_alg_sem); @@ -496,12 +513,10 @@ void crypto_drop_spawn(struct crypto_spawn *spawn) } EXPORT_SYMBOL_GPL(crypto_drop_spawn); -struct crypto_tfm *crypto_spawn_tfm(struct crypto_spawn *spawn, u32 type, - u32 mask) +static struct crypto_alg *crypto_spawn_alg(struct crypto_spawn *spawn) { struct crypto_alg *alg; struct crypto_alg *alg2; - struct crypto_tfm *tfm; down_read(&crypto_alg_sem); alg = spawn->alg; @@ -516,6 +531,19 @@ struct crypto_tfm *crypto_spawn_tfm(struct crypto_spawn *spawn, u32 type, return ERR_PTR(-EAGAIN); } + return alg; +} + +struct crypto_tfm *crypto_spawn_tfm(struct crypto_spawn *spawn, u32 type, + u32 mask) +{ + struct crypto_alg *alg; + struct crypto_tfm *tfm; + + alg = crypto_spawn_alg(spawn); + if (IS_ERR(alg)) + return ERR_CAST(alg); + tfm = ERR_PTR(-EINVAL); if (unlikely((alg->cra_flags ^ type) & mask)) goto out_put_alg; @@ -532,6 +560,27 @@ out_put_alg: } EXPORT_SYMBOL_GPL(crypto_spawn_tfm); +void *crypto_spawn_tfm2(struct crypto_spawn *spawn) +{ + struct crypto_alg *alg; + struct crypto_tfm *tfm; + + alg = crypto_spawn_alg(spawn); + if (IS_ERR(alg)) + return ERR_CAST(alg); + + tfm = crypto_create_tfm(alg, spawn->frontend); + if (IS_ERR(tfm)) + goto out_put_alg; + + return tfm; + +out_put_alg: + crypto_mod_put(alg); + return tfm; +} +EXPORT_SYMBOL_GPL(crypto_spawn_tfm2); + int crypto_register_notifier(struct notifier_block *nb) { return blocking_notifier_chain_register(&crypto_chain, nb); diff --git a/include/crypto/algapi.h b/include/crypto/algapi.h index 99bb29723130..c9bff92a9e0e 100644 --- a/include/crypto/algapi.h +++ b/include/crypto/algapi.h @@ -61,6 +61,7 @@ struct crypto_spawn { struct list_head list; struct crypto_alg *alg; struct crypto_instance *inst; + const struct crypto_type *frontend; u32 mask; }; @@ -117,9 +118,14 @@ struct crypto_template *crypto_lookup_template(const char *name); int crypto_init_spawn(struct crypto_spawn *spawn, struct crypto_alg *alg, struct crypto_instance *inst, u32 mask); +int crypto_init_spawn2(struct crypto_spawn *spawn, struct crypto_alg *alg, + struct crypto_instance *inst, + const struct crypto_type *frontend); + void crypto_drop_spawn(struct crypto_spawn *spawn); struct crypto_tfm *crypto_spawn_tfm(struct crypto_spawn *spawn, u32 type, u32 mask); +void *crypto_spawn_tfm2(struct crypto_spawn *spawn); static inline void crypto_set_spawn(struct crypto_spawn *spawn, struct crypto_instance *inst) -- cgit v1.2.3 From 942969992d86330c9700e2cd9afe8a6bea42df78 Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Wed, 8 Jul 2009 17:21:37 +0800 Subject: crypto: shash - Add spawn support This patch adds the functions needed to create and use shash spawns, i.e., to use shash algorithms in a template. Signed-off-by: Herbert Xu --- crypto/shash.c | 9 +++++++++ include/crypto/internal/hash.h | 14 ++++++++++++++ 2 files changed, 23 insertions(+) (limited to 'include') diff --git a/crypto/shash.c b/crypto/shash.c index 8f9d8831e293..97f6c8ba103a 100644 --- a/crypto/shash.c +++ b/crypto/shash.c @@ -509,5 +509,14 @@ void shash_free_instance(struct crypto_instance *inst) } EXPORT_SYMBOL_GPL(shash_free_instance); +int crypto_init_shash_spawn(struct crypto_shash_spawn *spawn, + struct shash_alg *alg, + struct crypto_instance *inst) +{ + return crypto_init_spawn2(&spawn->base, &alg->base, inst, + &crypto_shash_type); +} +EXPORT_SYMBOL_GPL(crypto_init_shash_spawn); + MODULE_LICENSE("GPL"); MODULE_DESCRIPTION("Synchronous cryptographic hash type"); diff --git a/include/crypto/internal/hash.h b/include/crypto/internal/hash.h index 44d3bcdce6e2..6cc824b79331 100644 --- a/include/crypto/internal/hash.h +++ b/include/crypto/internal/hash.h @@ -38,6 +38,10 @@ struct shash_instance { struct shash_alg alg; }; +struct crypto_shash_spawn { + struct crypto_spawn base; +}; + extern const struct crypto_type crypto_ahash_type; int crypto_hash_walk_done(struct crypto_hash_walk *walk, int err); @@ -52,6 +56,10 @@ int crypto_unregister_shash(struct shash_alg *alg); void shash_free_instance(struct crypto_instance *inst); +int crypto_init_shash_spawn(struct crypto_shash_spawn *spawn, + struct shash_alg *alg, + struct crypto_instance *inst); + static inline void *crypto_ahash_ctx(struct crypto_ahash *tfm) { return crypto_tfm_ctx(&tfm->base); @@ -106,5 +114,11 @@ static inline struct shash_instance *shash_alloc_instance( sizeof(struct shash_alg) - sizeof(*alg)); } +static inline struct crypto_shash *crypto_spawn_shash( + struct crypto_shash_spawn *spawn) +{ + return crypto_spawn_tfm2(&spawn->base); +} + #endif /* _CRYPTO_INTERNAL_HASH_H */ -- cgit v1.2.3 From d06854f0243d91badabaab14503f7f3bb770061d Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Wed, 8 Jul 2009 17:53:16 +0800 Subject: crypto: api - Add crypto_attr_alg2 helper This patch adds the helper crypto_attr_alg2 which is similar to crypto_attr_alg but takes an extra frontend argument. This is intended to be used by new style algorithm types such as shash. Signed-off-by: Herbert Xu --- crypto/algapi.c | 8 +++++--- crypto/api.c | 31 ++++++++++++++++++++++--------- crypto/internal.h | 3 +++ include/crypto/algapi.h | 11 ++++++++++- 4 files changed, 40 insertions(+), 13 deletions(-) (limited to 'include') diff --git a/crypto/algapi.c b/crypto/algapi.c index 17a5fff8f777..2154815201a7 100644 --- a/crypto/algapi.c +++ b/crypto/algapi.c @@ -644,7 +644,9 @@ const char *crypto_attr_alg_name(struct rtattr *rta) } EXPORT_SYMBOL_GPL(crypto_attr_alg_name); -struct crypto_alg *crypto_attr_alg(struct rtattr *rta, u32 type, u32 mask) +struct crypto_alg *crypto_attr_alg2(struct rtattr *rta, + const struct crypto_type *frontend, + u32 type, u32 mask) { const char *name; int err; @@ -654,9 +656,9 @@ struct crypto_alg *crypto_attr_alg(struct rtattr *rta, u32 type, u32 mask) if (IS_ERR(name)) return ERR_PTR(err); - return crypto_alg_mod_lookup(name, type, mask); + return crypto_find_alg(name, frontend, type, mask); } -EXPORT_SYMBOL_GPL(crypto_attr_alg); +EXPORT_SYMBOL_GPL(crypto_attr_alg2); int crypto_attr_u32(struct rtattr *rta, u32 *num) { diff --git a/crypto/api.c b/crypto/api.c index d5944f92b416..ba81221b3bef 100644 --- a/crypto/api.c +++ b/crypto/api.c @@ -503,6 +503,27 @@ out: } EXPORT_SYMBOL_GPL(crypto_create_tfm); +struct crypto_alg *crypto_find_alg(const char *alg_name, + const struct crypto_type *frontend, + u32 type, u32 mask) +{ + struct crypto_alg *(*lookup)(const char *name, u32 type, u32 mask) = + crypto_alg_mod_lookup; + + if (frontend) { + type &= frontend->maskclear; + mask &= frontend->maskclear; + type |= frontend->type; + mask |= frontend->maskset; + + if (frontend->lookup) + lookup = frontend->lookup; + } + + return lookup(alg_name, type, mask); +} +EXPORT_SYMBOL_GPL(crypto_find_alg); + /* * crypto_alloc_tfm - Locate algorithm and allocate transform * @alg_name: Name of algorithm @@ -526,21 +547,13 @@ EXPORT_SYMBOL_GPL(crypto_create_tfm); void *crypto_alloc_tfm(const char *alg_name, const struct crypto_type *frontend, u32 type, u32 mask) { - struct crypto_alg *(*lookup)(const char *name, u32 type, u32 mask); void *tfm; int err; - type &= frontend->maskclear; - mask &= frontend->maskclear; - type |= frontend->type; - mask |= frontend->maskset; - - lookup = frontend->lookup ?: crypto_alg_mod_lookup; - for (;;) { struct crypto_alg *alg; - alg = lookup(alg_name, type, mask); + alg = crypto_find_alg(alg_name, frontend, type, mask); if (IS_ERR(alg)) { err = PTR_ERR(alg); goto err; diff --git a/crypto/internal.h b/crypto/internal.h index 95baaea21fbc..7efa4d0533ff 100644 --- a/crypto/internal.h +++ b/crypto/internal.h @@ -106,6 +106,9 @@ struct crypto_tfm *__crypto_alloc_tfm(struct crypto_alg *alg, u32 type, u32 mask); void *crypto_create_tfm(struct crypto_alg *alg, const struct crypto_type *frontend); +struct crypto_alg *crypto_find_alg(const char *alg_name, + const struct crypto_type *frontend, + u32 type, u32 mask); void *crypto_alloc_tfm(const char *alg_name, const struct crypto_type *frontend, u32 type, u32 mask); diff --git a/include/crypto/algapi.h b/include/crypto/algapi.h index c9bff92a9e0e..1d15a926041e 100644 --- a/include/crypto/algapi.h +++ b/include/crypto/algapi.h @@ -136,7 +136,16 @@ static inline void crypto_set_spawn(struct crypto_spawn *spawn, struct crypto_attr_type *crypto_get_attr_type(struct rtattr **tb); int crypto_check_attr_type(struct rtattr **tb, u32 type); const char *crypto_attr_alg_name(struct rtattr *rta); -struct crypto_alg *crypto_attr_alg(struct rtattr *rta, u32 type, u32 mask); +struct crypto_alg *crypto_attr_alg2(struct rtattr *rta, + const struct crypto_type *frontend, + u32 type, u32 mask); + +static inline struct crypto_alg *crypto_attr_alg(struct rtattr *rta, + u32 type, u32 mask) +{ + return crypto_attr_alg2(rta, NULL, type, mask); +} + int crypto_attr_u32(struct rtattr *rta, u32 *num); void *crypto_alloc_instance2(const char *name, struct crypto_alg *alg, unsigned int head); -- cgit v1.2.3 From 7d6f56400a695af497a8b7c23ea0ff9c3d9d99f4 Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Wed, 8 Jul 2009 17:56:28 +0800 Subject: crypto: shash - Add shash_attr_alg2 helper This patch adds the helper shash_attr_alg2 which locates a shash algorithm based on the information in the given attribute. Signed-off-by: Herbert Xu --- crypto/shash.c | 10 ++++++++++ include/crypto/internal/hash.h | 2 ++ 2 files changed, 12 insertions(+) (limited to 'include') diff --git a/crypto/shash.c b/crypto/shash.c index 97f6c8ba103a..21bcff6be5b0 100644 --- a/crypto/shash.c +++ b/crypto/shash.c @@ -518,5 +518,15 @@ int crypto_init_shash_spawn(struct crypto_shash_spawn *spawn, } EXPORT_SYMBOL_GPL(crypto_init_shash_spawn); +struct shash_alg *shash_attr_alg(struct rtattr *rta, u32 type, u32 mask) +{ + struct crypto_alg *alg; + + alg = crypto_attr_alg2(rta, &crypto_shash_type, type, mask); + return IS_ERR(alg) ? ERR_CAST(alg) : + container_of(alg, struct shash_alg, base); +} +EXPORT_SYMBOL_GPL(shash_attr_alg); + MODULE_LICENSE("GPL"); MODULE_DESCRIPTION("Synchronous cryptographic hash type"); diff --git a/include/crypto/internal/hash.h b/include/crypto/internal/hash.h index 6cc824b79331..e2ab35a74c62 100644 --- a/include/crypto/internal/hash.h +++ b/include/crypto/internal/hash.h @@ -60,6 +60,8 @@ int crypto_init_shash_spawn(struct crypto_shash_spawn *spawn, struct shash_alg *alg, struct crypto_instance *inst); +struct shash_alg *shash_attr_alg(struct rtattr *rta, u32 type, u32 mask); + static inline void *crypto_ahash_ctx(struct crypto_ahash *tfm) { return crypto_tfm_ctx(&tfm->base); -- cgit v1.2.3 From 619a6ebd2547f3a8ec2fbc5245daaa1f2056eb32 Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Wed, 8 Jul 2009 18:46:23 +0800 Subject: crypto: shash - Add shash_register_instance This patch adds shash_register_instance so that shash instances can be registered without bypassing the shash checks applied to normal algorithms. Signed-off-by: Herbert Xu --- crypto/shash.c | 26 +++++++++++++++++++++++++- include/crypto/internal/hash.h | 3 ++- 2 files changed, 27 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/crypto/shash.c b/crypto/shash.c index 21bcff6be5b0..3d242425d692 100644 --- a/crypto/shash.c +++ b/crypto/shash.c @@ -480,7 +480,7 @@ struct crypto_shash *crypto_alloc_shash(const char *alg_name, u32 type, } EXPORT_SYMBOL_GPL(crypto_alloc_shash); -int crypto_register_shash(struct shash_alg *alg) +static int shash_prepare_alg(struct shash_alg *alg) { struct crypto_alg *base = &alg->base; @@ -491,6 +491,17 @@ int crypto_register_shash(struct shash_alg *alg) base->cra_type = &crypto_shash_type; base->cra_flags &= ~CRYPTO_ALG_TYPE_MASK; base->cra_flags |= CRYPTO_ALG_TYPE_SHASH; + return 0; +} + +int crypto_register_shash(struct shash_alg *alg) +{ + struct crypto_alg *base = &alg->base; + int err; + + err = shash_prepare_alg(alg); + if (err) + return err; return crypto_register_alg(base); } @@ -502,6 +513,19 @@ int crypto_unregister_shash(struct shash_alg *alg) } EXPORT_SYMBOL_GPL(crypto_unregister_shash); +int shash_register_instance(struct crypto_template *tmpl, + struct shash_instance *inst) +{ + int err; + + err = shash_prepare_alg(&inst->alg); + if (err) + return err; + + return crypto_register_instance(tmpl, shash_crypto_instance(inst)); +} +EXPORT_SYMBOL_GPL(shash_register_instance); + void shash_free_instance(struct crypto_instance *inst) { crypto_drop_spawn(crypto_instance_ctx(inst)); diff --git a/include/crypto/internal/hash.h b/include/crypto/internal/hash.h index e2ab35a74c62..fa5c9fb7ce5a 100644 --- a/include/crypto/internal/hash.h +++ b/include/crypto/internal/hash.h @@ -53,7 +53,8 @@ int crypto_hash_walk_first_compat(struct hash_desc *hdesc, int crypto_register_shash(struct shash_alg *alg); int crypto_unregister_shash(struct shash_alg *alg); - +int shash_register_instance(struct crypto_template *tmpl, + struct shash_instance *inst); void shash_free_instance(struct crypto_instance *inst); int crypto_init_shash_spawn(struct crypto_shash_spawn *spawn, -- cgit v1.2.3 From cc6a8acdeee932f6911d8b236d2c7d6bcc4616f6 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Tue, 17 Jun 2008 16:39:06 +0200 Subject: ALSA: Fix SG-buffer DMA with non-coherent architectures Using SG-buffers with dma_alloc_coherent() is often very inefficient on non-coherent architectures because a tracking record could be allocated in addition for each dma_alloc_coherent() call. Instead, simply disable SG-buffers but just allocate normal continuous buffers on non-supported (currently all but x86) architectures. Signed-off-by: Takashi Iwai --- include/sound/memalloc.h | 6 ++++++ include/sound/pcm.h | 23 +++++++++++++++++++++++ sound/core/Kconfig | 4 ++++ sound/core/Makefile | 2 +- sound/core/memalloc.c | 4 ++++ sound/core/pcm_memory.c | 2 ++ 6 files changed, 40 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/sound/memalloc.h b/include/sound/memalloc.h index 7ccce94a5255..c42506212649 100644 --- a/include/sound/memalloc.h +++ b/include/sound/memalloc.h @@ -47,7 +47,11 @@ struct snd_dma_device { #define SNDRV_DMA_TYPE_UNKNOWN 0 /* not defined */ #define SNDRV_DMA_TYPE_CONTINUOUS 1 /* continuous no-DMA memory */ #define SNDRV_DMA_TYPE_DEV 2 /* generic device continuous */ +#ifdef CONFIG_SND_DMA_SGBUF #define SNDRV_DMA_TYPE_DEV_SG 3 /* generic device SG-buffer */ +#else +#define SNDRV_DMA_TYPE_DEV_SG SNDRV_DMA_TYPE_DEV /* no SG-buf support */ +#endif /* * info for buffer allocation @@ -60,6 +64,7 @@ struct snd_dma_buffer { void *private_data; /* private for allocator; don't touch */ }; +#ifdef CONFIG_SND_DMA_SGBUF /* * Scatter-Gather generic device pages */ @@ -107,6 +112,7 @@ static inline void *snd_sgbuf_get_ptr(struct snd_sg_buf *sgbuf, size_t offset) { return sgbuf->table[offset >> PAGE_SHIFT].buf + offset % PAGE_SIZE; } +#endif /* CONFIG_SND_DMA_SGBUF */ /* allocate/release a buffer */ int snd_dma_alloc_pages(int type, struct device *dev, size_t size, diff --git a/include/sound/pcm.h b/include/sound/pcm.h index 23893523dc8c..1691c7fe35af 100644 --- a/include/sound/pcm.h +++ b/include/sound/pcm.h @@ -902,6 +902,7 @@ int snd_pcm_lib_preallocate_pages_for_all(struct snd_pcm *pcm, int snd_pcm_lib_malloc_pages(struct snd_pcm_substream *substream, size_t size); int snd_pcm_lib_free_pages(struct snd_pcm_substream *substream); +#ifdef CONFIG_SND_DMA_SGBUF /* * SG-buffer handling */ @@ -927,6 +928,28 @@ struct page *snd_pcm_sgbuf_ops_page(struct snd_pcm_substream *substream, unsigned int snd_pcm_sgbuf_get_chunk_size(struct snd_pcm_substream *substream, unsigned int ofs, unsigned int size); +#else /* !SND_DMA_SGBUF */ +/* + * fake using a continuous buffer + */ +static inline dma_addr_t +snd_pcm_sgbuf_get_addr(struct snd_pcm_substream *substream, unsigned int ofs) +{ + return substream->runtime->dma_addr + ofs; +} + +static inline void * +snd_pcm_sgbuf_get_ptr(struct snd_pcm_substream *substream, unsigned int ofs) +{ + return substream->runtime->dma_area + ofs; +} + +#define snd_pcm_sgbuf_ops_page NULL + +#define snd_pcm_sgbuf_get_chunk_size(subs, ofs, size) (size) + +#endif /* SND_DMA_SGBUF */ + /* handle mmap counter - PCM mmap callback should handle this counter properly */ static inline void snd_pcm_mmap_data_open(struct vm_area_struct *area) { diff --git a/sound/core/Kconfig b/sound/core/Kconfig index 6061fb5f4e1c..c15682a2f9db 100644 --- a/sound/core/Kconfig +++ b/sound/core/Kconfig @@ -206,4 +206,8 @@ config SND_PCM_XRUN_DEBUG config SND_VMASTER bool +config SND_DMA_SGBUF + def_bool y + depends on X86 + source "sound/core/seq/Kconfig" diff --git a/sound/core/Makefile b/sound/core/Makefile index 4229052e7b91..350a08d277f4 100644 --- a/sound/core/Makefile +++ b/sound/core/Makefile @@ -13,7 +13,7 @@ snd-pcm-objs := pcm.o pcm_native.o pcm_lib.o pcm_timer.o pcm_misc.o \ pcm_memory.o snd-page-alloc-y := memalloc.o -snd-page-alloc-$(CONFIG_HAS_DMA) += sgbuf.o +snd-page-alloc-$(CONFIG_SND_DMA_SGBUF) += sgbuf.o snd-rawmidi-objs := rawmidi.o snd-timer-objs := timer.o diff --git a/sound/core/memalloc.c b/sound/core/memalloc.c index 1b3534d67686..9e92441f9b78 100644 --- a/sound/core/memalloc.c +++ b/sound/core/memalloc.c @@ -199,6 +199,8 @@ int snd_dma_alloc_pages(int type, struct device *device, size_t size, case SNDRV_DMA_TYPE_DEV: dmab->area = snd_malloc_dev_pages(device, size, &dmab->addr); break; +#endif +#ifdef CONFIG_SND_DMA_SGBUF case SNDRV_DMA_TYPE_DEV_SG: snd_malloc_sgbuf_pages(device, size, dmab, NULL); break; @@ -269,6 +271,8 @@ void snd_dma_free_pages(struct snd_dma_buffer *dmab) case SNDRV_DMA_TYPE_DEV: snd_free_dev_pages(dmab->dev.dev, dmab->bytes, dmab->area, dmab->addr); break; +#endif +#ifdef CONFIG_SND_DMA_SGBUF case SNDRV_DMA_TYPE_DEV_SG: snd_free_sgbuf_pages(dmab); break; diff --git a/sound/core/pcm_memory.c b/sound/core/pcm_memory.c index a6d42808828c..caa7796bc2f5 100644 --- a/sound/core/pcm_memory.c +++ b/sound/core/pcm_memory.c @@ -304,6 +304,7 @@ int snd_pcm_lib_preallocate_pages_for_all(struct snd_pcm *pcm, EXPORT_SYMBOL(snd_pcm_lib_preallocate_pages_for_all); +#ifdef CONFIG_SND_DMA_SGBUF /** * snd_pcm_sgbuf_ops_page - get the page struct at the given offset * @substream: the pcm substream instance @@ -349,6 +350,7 @@ unsigned int snd_pcm_sgbuf_get_chunk_size(struct snd_pcm_substream *substream, return size; } EXPORT_SYMBOL(snd_pcm_sgbuf_get_chunk_size); +#endif /* CONFIG_SND_DMA_SGBUF */ /** * snd_pcm_lib_malloc_pages - allocate the DMA buffer -- cgit v1.2.3 From e4f7c0b44a8ac8935f223195af9ea637d0c08091 Mon Sep 17 00:00:00 2001 From: Catalin Marinas Date: Tue, 7 Jul 2009 10:32:59 +0100 Subject: kmemleak: Trace the kmalloc_large* functions in slub The kmalloc_large() and kmalloc_large_node() functions were missed when adding the kmemleak hooks to the slub allocator. However, they should be traced to avoid false positives. Signed-off-by: Catalin Marinas Cc: Christoph Lameter Acked-by: Pekka Enberg --- include/linux/slub_def.h | 2 ++ mm/slub.c | 10 ++++++---- 2 files changed, 8 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/include/linux/slub_def.h b/include/linux/slub_def.h index 4dcbc2c71491..c1c862b1d01a 100644 --- a/include/linux/slub_def.h +++ b/include/linux/slub_def.h @@ -11,6 +11,7 @@ #include #include #include +#include enum stat_item { ALLOC_FASTPATH, /* Allocation from cpu slab */ @@ -233,6 +234,7 @@ static __always_inline void *kmalloc_large(size_t size, gfp_t flags) unsigned int order = get_order(size); void *ret = (void *) __get_free_pages(flags | __GFP_COMP, order); + kmemleak_alloc(ret, size, 1, flags); trace_kmalloc(_THIS_IP_, ret, size, PAGE_SIZE << order, flags); return ret; diff --git a/mm/slub.c b/mm/slub.c index a9201d83178b..b9f1491a58a1 100644 --- a/mm/slub.c +++ b/mm/slub.c @@ -21,7 +21,6 @@ #include #include #include -#include #include #include #include @@ -2835,13 +2834,15 @@ EXPORT_SYMBOL(__kmalloc); static void *kmalloc_large_node(size_t size, gfp_t flags, int node) { struct page *page; + void *ptr = NULL; flags |= __GFP_COMP | __GFP_NOTRACK; page = alloc_pages_node(node, flags, get_order(size)); if (page) - return page_address(page); - else - return NULL; + ptr = page_address(page); + + kmemleak_alloc(ptr, size, 1, flags); + return ptr; } #ifdef CONFIG_NUMA @@ -2926,6 +2927,7 @@ void kfree(const void *x) page = virt_to_head_page(x); if (unlikely(!PageSlab(page))) { BUG_ON(!PageCompound(page)); + kmemleak_free(x); put_page(page); return; } -- cgit v1.2.3 From 53238a60dd4a679f6fe5613a7ed46899587205cf Mon Sep 17 00:00:00 2001 From: Catalin Marinas Date: Tue, 7 Jul 2009 10:33:00 +0100 Subject: kmemleak: Allow partial freeing of memory blocks Functions like free_bootmem() are allowed to free only part of a memory block. This patch adds support for this via the kmemleak_free_part() callback which removes the original object and creates one or two additional objects as a result of the memory block split. Signed-off-by: Catalin Marinas Cc: Ingo Molnar Acked-by: Pekka Enberg --- include/linux/kmemleak.h | 4 ++ mm/kmemleak.c | 95 +++++++++++++++++++++++++++++++++++++++++------- 2 files changed, 85 insertions(+), 14 deletions(-) (limited to 'include') diff --git a/include/linux/kmemleak.h b/include/linux/kmemleak.h index 7796aed6cdd5..6a63807f714e 100644 --- a/include/linux/kmemleak.h +++ b/include/linux/kmemleak.h @@ -27,6 +27,7 @@ extern void kmemleak_init(void); extern void kmemleak_alloc(const void *ptr, size_t size, int min_count, gfp_t gfp); extern void kmemleak_free(const void *ptr); +extern void kmemleak_free_part(const void *ptr, size_t size); extern void kmemleak_padding(const void *ptr, unsigned long offset, size_t size); extern void kmemleak_not_leak(const void *ptr); @@ -71,6 +72,9 @@ static inline void kmemleak_alloc_recursive(const void *ptr, size_t size, static inline void kmemleak_free(const void *ptr) { } +static inline void kmemleak_free_part(const void *ptr, size_t size) +{ +} static inline void kmemleak_free_recursive(const void *ptr, unsigned long flags) { } diff --git a/mm/kmemleak.c b/mm/kmemleak.c index 466d39007264..5aabd41ffb8f 100644 --- a/mm/kmemleak.c +++ b/mm/kmemleak.c @@ -210,6 +210,7 @@ static DEFINE_MUTEX(scan_mutex); enum { KMEMLEAK_ALLOC, KMEMLEAK_FREE, + KMEMLEAK_FREE_PART, KMEMLEAK_NOT_LEAK, KMEMLEAK_IGNORE, KMEMLEAK_SCAN_AREA, @@ -523,27 +524,17 @@ out: * Remove the metadata (struct kmemleak_object) for a memory block from the * object_list and object_tree_root and decrement its use_count. */ -static void delete_object(unsigned long ptr) +static void __delete_object(struct kmemleak_object *object) { unsigned long flags; - struct kmemleak_object *object; write_lock_irqsave(&kmemleak_lock, flags); - object = lookup_object(ptr, 0); - if (!object) { -#ifdef DEBUG - kmemleak_warn("Freeing unknown object at 0x%08lx\n", - ptr); -#endif - write_unlock_irqrestore(&kmemleak_lock, flags); - return; - } prio_tree_remove(&object_tree_root, &object->tree_node); list_del_rcu(&object->object_list); write_unlock_irqrestore(&kmemleak_lock, flags); WARN_ON(!(object->flags & OBJECT_ALLOCATED)); - WARN_ON(atomic_read(&object->use_count) < 1); + WARN_ON(atomic_read(&object->use_count) < 2); /* * Locking here also ensures that the corresponding memory block @@ -555,6 +546,64 @@ static void delete_object(unsigned long ptr) put_object(object); } +/* + * Look up the metadata (struct kmemleak_object) corresponding to ptr and + * delete it. + */ +static void delete_object_full(unsigned long ptr) +{ + struct kmemleak_object *object; + + object = find_and_get_object(ptr, 0); + if (!object) { +#ifdef DEBUG + kmemleak_warn("Freeing unknown object at 0x%08lx\n", + ptr); +#endif + return; + } + __delete_object(object); + put_object(object); +} + +/* + * Look up the metadata (struct kmemleak_object) corresponding to ptr and + * delete it. If the memory block is partially freed, the function may create + * additional metadata for the remaining parts of the block. + */ +static void delete_object_part(unsigned long ptr, size_t size) +{ + struct kmemleak_object *object; + unsigned long start, end; + + object = find_and_get_object(ptr, 1); + if (!object) { +#ifdef DEBUG + kmemleak_warn("Partially freeing unknown object at 0x%08lx " + "(size %zu)\n", ptr, size); +#endif + return; + } + __delete_object(object); + + /* + * Create one or two objects that may result from the memory block + * split. Note that partial freeing is only done by free_bootmem() and + * this happens before kmemleak_init() is called. The path below is + * only executed during early log recording in kmemleak_init(), so + * GFP_KERNEL is enough. + */ + start = object->pointer; + end = object->pointer + object->size; + if (ptr > start) + create_object(start, ptr - start, object->min_count, + GFP_KERNEL); + if (ptr + size < end) + create_object(ptr + size, end - ptr - size, object->min_count, + GFP_KERNEL); + + put_object(object); +} /* * Make a object permanently as gray-colored so that it can no longer be * reported as a leak. This is used in general to mark a false positive. @@ -719,12 +768,27 @@ void kmemleak_free(const void *ptr) pr_debug("%s(0x%p)\n", __func__, ptr); if (atomic_read(&kmemleak_enabled) && ptr && !IS_ERR(ptr)) - delete_object((unsigned long)ptr); + delete_object_full((unsigned long)ptr); else if (atomic_read(&kmemleak_early_log)) log_early(KMEMLEAK_FREE, ptr, 0, 0, 0, 0); } EXPORT_SYMBOL_GPL(kmemleak_free); +/* + * Partial memory freeing function callback. This function is usually called + * from bootmem allocator when (part of) a memory block is freed. + */ +void kmemleak_free_part(const void *ptr, size_t size) +{ + pr_debug("%s(0x%p)\n", __func__, ptr); + + if (atomic_read(&kmemleak_enabled) && ptr && !IS_ERR(ptr)) + delete_object_part((unsigned long)ptr, size); + else if (atomic_read(&kmemleak_early_log)) + log_early(KMEMLEAK_FREE_PART, ptr, size, 0, 0, 0); +} +EXPORT_SYMBOL_GPL(kmemleak_free_part); + /* * Mark an already allocated memory block as a false positive. This will cause * the block to no longer be reported as leak and always be scanned. @@ -1318,7 +1382,7 @@ static int kmemleak_cleanup_thread(void *arg) rcu_read_lock(); list_for_each_entry_rcu(object, &object_list, object_list) - delete_object(object->pointer); + delete_object_full(object->pointer); rcu_read_unlock(); mutex_unlock(&scan_mutex); @@ -1413,6 +1477,9 @@ void __init kmemleak_init(void) case KMEMLEAK_FREE: kmemleak_free(log->ptr); break; + case KMEMLEAK_FREE_PART: + kmemleak_free_part(log->ptr, log->size); + break; case KMEMLEAK_NOT_LEAK: kmemleak_not_leak(log->ptr); break; -- cgit v1.2.3 From cde6263fa954dfc03ebe169aa3f7f71176d7901b Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Wed, 8 Jul 2009 22:32:07 +0800 Subject: crypto: shash - Add crypto_shash_ctx_aligned This patch adds crypto_shash_ctx_aligned which will be needed by hmac after its conversion to shash. Signed-off-by: Herbert Xu --- include/crypto/internal/hash.h | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'include') diff --git a/include/crypto/internal/hash.h b/include/crypto/internal/hash.h index fa5c9fb7ce5a..f1041140d3d9 100644 --- a/include/crypto/internal/hash.h +++ b/include/crypto/internal/hash.h @@ -123,5 +123,10 @@ static inline struct crypto_shash *crypto_spawn_shash( return crypto_spawn_tfm2(&spawn->base); } +static inline void *crypto_shash_ctx_aligned(struct crypto_shash *tfm) +{ + return crypto_tfm_ctx_aligned(&tfm->base); +} + #endif /* _CRYPTO_INTERNAL_HASH_H */ -- cgit v1.2.3 From 0390e6aecf571ddac934e6c0644bb4038167b698 Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Wed, 8 Jul 2009 22:36:36 +0800 Subject: crypto: shash - Add __crypto_shash_cast This patch adds __crypto_shash_cast which turns a crypto_tfm into crypto_shash. It's analogous to the other __crypto_*_cast functions. It hasn't been needed until now since no existing shash algorithms have had an init function. Signed-off-by: Herbert Xu --- include/crypto/internal/hash.h | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'include') diff --git a/include/crypto/internal/hash.h b/include/crypto/internal/hash.h index f1041140d3d9..3af34ca8e8bc 100644 --- a/include/crypto/internal/hash.h +++ b/include/crypto/internal/hash.h @@ -128,5 +128,10 @@ static inline void *crypto_shash_ctx_aligned(struct crypto_shash *tfm) return crypto_tfm_ctx_aligned(&tfm->base); } +static inline struct crypto_shash *__crypto_shash_cast(struct crypto_tfm *tfm) +{ + return (struct crypto_shash *)tfm; +} + #endif /* _CRYPTO_INTERNAL_HASH_H */ -- cgit v1.2.3 From 37d217f029a56a6d385f99773fb27dfcb51f9a46 Mon Sep 17 00:00:00 2001 From: Miklos Szeredi Date: Wed, 8 Jul 2009 18:17:58 +0200 Subject: fuse: document protocol version negotiation Clarify how the protocol version should be negotiated between kernel and userspace. Notably libfuse didn't correctly handle the case when the supported major versions didn't match. Signed-off-by: Miklos Szeredi --- include/linux/fuse.h | 20 ++++++++++++++++++++ 1 file changed, 20 insertions(+) (limited to 'include') diff --git a/include/linux/fuse.h b/include/linux/fuse.h index b3700f0ac268..3e2925a34bf0 100644 --- a/include/linux/fuse.h +++ b/include/linux/fuse.h @@ -41,6 +41,26 @@ #include +/* + * Version negotiation: + * + * Both the kernel and userspace send the version they support in the + * INIT request and reply respectively. + * + * If the major versions match then both shall use the smallest + * of the two minor versions for communication. + * + * If the kernel supports a larger major version, then userspace shall + * reply with the major version it supports, ignore the rest of the + * INIT message and expect a new INIT message from the kernel with a + * matching major version. + * + * If the library supports a larger major version, then it shall fall + * back to the major protocol version sent by the kernel for + * communication and reply with that major version (and an arbitrary + * supported minor version). + */ + /** Version number of this interface */ #define FUSE_KERNEL_VERSION 7 -- cgit v1.2.3 From e9bf0cc7cbfbf3952cdf8028aa0d348d09ecdba1 Mon Sep 17 00:00:00 2001 From: Parag Warudkar Date: Wed, 8 Jul 2009 11:46:02 -0400 Subject: elfcore.h : Fix UML build breakage Commit a65e7bfcd74e4c0939f235d2bf9f48ddb3a57991 broke the UML build with the following error - In file included from fs/proc/kcore.c:17: include/linux/elfcore.h: In function 'elf_core_copy_task_regs': include/linux/elfcore.h:129: error: implicit declaration of function 'task_pt_regs' Fix this by restoring the previous behavior of returning 0 for all arches like UML that don't define task_pt_regs. Signed-off-by: Parag Warudkar Acked-by: Amerigo Wang Signed-off-by: Linus Torvalds --- include/linux/elfcore.h | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/include/linux/elfcore.h b/include/linux/elfcore.h index 03ec16779802..00d6a68d0421 100644 --- a/include/linux/elfcore.h +++ b/include/linux/elfcore.h @@ -122,10 +122,9 @@ static inline void elf_core_copy_kernel_regs(elf_gregset_t *elfregs, struct pt_r static inline int elf_core_copy_task_regs(struct task_struct *t, elf_gregset_t* elfregs) { -#ifdef ELF_CORE_COPY_TASK_REGS - +#if defined (ELF_CORE_COPY_TASK_REGS) return ELF_CORE_COPY_TASK_REGS(t, elfregs); -#else +#elif defined (task_pt_regs) elf_core_copy_regs(elfregs, task_pt_regs(t)); #endif return 0; -- cgit v1.2.3 From b43f3cbd21ffbd719fd4fa6642bfe6af255ded34 Mon Sep 17 00:00:00 2001 From: Alexey Dobriyan Date: Wed, 8 Jul 2009 01:54:37 +0400 Subject: headers: mnt_namespace.h redux Fix various silly problems wrt mnt_namespace.h: - exit_mnt_ns() isn't used, remove it - done that, sched.h and nsproxy.h inclusions aren't needed - mount.h inclusion was need for vfsmount_lock, but no longer - remove mnt_namespace.h inclusion from files which don't use anything from mnt_namespace.h Signed-off-by: Alexey Dobriyan Signed-off-by: Linus Torvalds --- fs/afs/mntpt.c | 1 - fs/namespace.c | 1 + fs/nfs/getroot.c | 1 - fs/reiserfs/super.c | 1 - include/linux/mnt_namespace.h | 13 ++----------- kernel/exit.c | 1 - kernel/fork.c | 1 - kernel/kmod.c | 1 - 8 files changed, 3 insertions(+), 17 deletions(-) (limited to 'include') diff --git a/fs/afs/mntpt.c b/fs/afs/mntpt.c index c52be53f6946..5ffb570cd3a8 100644 --- a/fs/afs/mntpt.c +++ b/fs/afs/mntpt.c @@ -17,7 +17,6 @@ #include #include #include -#include #include "internal.h" diff --git a/fs/namespace.c b/fs/namespace.c index 3dc283fd4716..277c28a63ead 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -22,6 +22,7 @@ #include #include #include +#include #include #include #include diff --git a/fs/nfs/getroot.c b/fs/nfs/getroot.c index 46177cb87064..b35d2a616066 100644 --- a/fs/nfs/getroot.c +++ b/fs/nfs/getroot.c @@ -30,7 +30,6 @@ #include #include #include -#include #include #include diff --git a/fs/reiserfs/super.c b/fs/reiserfs/super.c index d3aeb061612b..7adea74d6a8a 100644 --- a/fs/reiserfs/super.c +++ b/fs/reiserfs/super.c @@ -24,7 +24,6 @@ #include #include #include -#include #include #include #include diff --git a/include/linux/mnt_namespace.h b/include/linux/mnt_namespace.h index 3beb2592b03f..d74785c2393a 100644 --- a/include/linux/mnt_namespace.h +++ b/include/linux/mnt_namespace.h @@ -2,10 +2,9 @@ #define _NAMESPACE_H_ #ifdef __KERNEL__ -#include -#include -#include +#include #include +#include struct mnt_namespace { atomic_t count; @@ -28,14 +27,6 @@ extern struct mnt_namespace *create_mnt_ns(struct vfsmount *mnt); extern struct mnt_namespace *copy_mnt_ns(unsigned long, struct mnt_namespace *, struct fs_struct *); extern void put_mnt_ns(struct mnt_namespace *ns); - -static inline void exit_mnt_ns(struct task_struct *p) -{ - struct mnt_namespace *ns = p->nsproxy->mnt_ns; - if (ns) - put_mnt_ns(ns); -} - static inline void get_mnt_ns(struct mnt_namespace *ns) { atomic_inc(&ns->count); diff --git a/kernel/exit.c b/kernel/exit.c index 628d41f0dd54..869dc221733e 100644 --- a/kernel/exit.c +++ b/kernel/exit.c @@ -12,7 +12,6 @@ #include #include #include -#include #include #include #include diff --git a/kernel/fork.c b/kernel/fork.c index 467746b3f0aa..bd2959228871 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -17,7 +17,6 @@ #include #include #include -#include #include #include #include diff --git a/kernel/kmod.c b/kernel/kmod.c index 7e95bedb2bfc..385c31a1bdbf 100644 --- a/kernel/kmod.c +++ b/kernel/kmod.c @@ -24,7 +24,6 @@ #include #include #include -#include #include #include #include -- cgit v1.2.3 From 1ce822fa04fd6878f079461a4b8affe4bb5ec27b Mon Sep 17 00:00:00 2001 From: Jaswinder Singh Rajput Date: Wed, 8 Jul 2009 21:25:54 +0530 Subject: includecheck fix: include/linux, rfkill.h fix the following 'make includecheck' warning: include/linux/rfkill.h: linux/types.h is included more than once. Signed-off-by: Jaswinder Singh Rajput Signed-off-by: John W. Linville --- include/linux/rfkill.h | 1 - 1 file changed, 1 deletion(-) (limited to 'include') diff --git a/include/linux/rfkill.h b/include/linux/rfkill.h index e73e2429a1b1..2ce29831feb6 100644 --- a/include/linux/rfkill.h +++ b/include/linux/rfkill.h @@ -99,7 +99,6 @@ enum rfkill_user_states { #undef RFKILL_STATE_UNBLOCKED #undef RFKILL_STATE_HARD_BLOCKED -#include #include #include #include -- cgit v1.2.3 From 023bf6f1b8bf58dc4da7f0dc1cf4787b0d5297c1 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Thu, 9 Jul 2009 11:27:40 +0900 Subject: linker script: unify usage of discard definition Discarded sections in different archs share some commonality but have considerable differences. This led to linker script for each arch implementing its own /DISCARD/ definition, which makes maintaining tedious and adding new entries error-prone. This patch makes all linker scripts to move discard definitions to the end of the linker script and use the common DISCARDS macro. As ld uses the first matching section definition, archs can include default discarded sections by including them earlier in the linker script. ia64 is notable because it first throws away some ia64 specific subsections and then include the rest of the sections into the final image, so those sections must be discarded before the inclusion. defconfig compile tested for x86, x86-64, powerpc, powerpc64, ia64, alpha, sparc, sparc64 and s390. Michal Simek tested microblaze. Signed-off-by: Tejun Heo Acked-by: Paul Mundt Acked-by: Mike Frysinger Tested-by: Michal Simek Cc: linux-arch@vger.kernel.org Cc: Michal Simek Cc: microblaze-uclinux@itee.uq.edu.au Cc: Sam Ravnborg Cc: Tony Luck --- arch/alpha/kernel/vmlinux.lds.S | 10 ++-------- arch/avr32/kernel/vmlinux.lds.S | 10 +++------- arch/blackfin/kernel/vmlinux.lds.S | 6 +----- arch/cris/kernel/vmlinux.lds.S | 10 ++-------- arch/frv/kernel/vmlinux.lds.S | 2 +- arch/h8300/kernel/vmlinux.lds.S | 6 ++---- arch/ia64/kernel/vmlinux.lds.S | 17 ++++++++--------- arch/m32r/kernel/vmlinux.lds.S | 11 +++-------- arch/m68k/kernel/vmlinux-std.lds | 11 +++-------- arch/m68k/kernel/vmlinux-sun3.lds | 10 ++-------- arch/m68knommu/kernel/vmlinux.lds.S | 8 +------- arch/microblaze/kernel/vmlinux.lds.S | 2 +- arch/mips/kernel/vmlinux.lds.S | 22 ++++++++++------------ arch/mn10300/kernel/vmlinux.lds.S | 9 +++------ arch/parisc/kernel/vmlinux.lds.S | 9 ++++----- arch/powerpc/kernel/vmlinux.lds.S | 10 +++------- arch/s390/kernel/vmlinux.lds.S | 10 +++------- arch/sh/kernel/vmlinux.lds.S | 11 ++++------- arch/sparc/kernel/vmlinux.lds.S | 9 ++------- arch/um/include/asm/common.lds.S | 5 ----- arch/um/kernel/dyn.lds.S | 2 +- arch/um/kernel/uml.lds.S | 2 +- arch/x86/kernel/vmlinux.lds.S | 11 ++++------- arch/xtensa/kernel/vmlinux.lds.S | 14 ++++---------- include/asm-generic/vmlinux.lds.h | 18 ++++++++++++------ 25 files changed, 80 insertions(+), 155 deletions(-) (limited to 'include') diff --git a/arch/alpha/kernel/vmlinux.lds.S b/arch/alpha/kernel/vmlinux.lds.S index 75fe1d6877e9..6dc03c35caa0 100644 --- a/arch/alpha/kernel/vmlinux.lds.S +++ b/arch/alpha/kernel/vmlinux.lds.S @@ -134,14 +134,6 @@ SECTIONS __bss_stop = .; _end = .; - /* Sections to be discarded */ - /DISCARD/ : { - EXIT_TEXT - EXIT_DATA - *(.exitcall.exit) - *(.discard) - } - .mdebug 0 : { *(.mdebug) } @@ -151,4 +143,6 @@ SECTIONS STABS_DEBUG DWARF_DEBUG + + DISCARDS } diff --git a/arch/avr32/kernel/vmlinux.lds.S b/arch/avr32/kernel/vmlinux.lds.S index b8324608ec0c..c4b56654349a 100644 --- a/arch/avr32/kernel/vmlinux.lds.S +++ b/arch/avr32/kernel/vmlinux.lds.S @@ -124,15 +124,11 @@ SECTIONS _end = .; } + DWARF_DEBUG + /* When something in the kernel is NOT compiled as a module, the module * cleanup code and data are put into these segments. Both can then be * thrown away, as cleanup code is never called unless it's a module. */ - /DISCARD/ : { - EXIT_DATA - *(.exitcall.exit) - *(.discard) - } - - DWARF_DEBUG + DISCARDS } diff --git a/arch/blackfin/kernel/vmlinux.lds.S b/arch/blackfin/kernel/vmlinux.lds.S index 6e8eabd8f0a6..d7ffe299b979 100644 --- a/arch/blackfin/kernel/vmlinux.lds.S +++ b/arch/blackfin/kernel/vmlinux.lds.S @@ -277,9 +277,5 @@ SECTIONS DWARF_DEBUG - /DISCARD/ : - { - *(.exitcall.exit) - *(.discard) - } + DISCARDS } diff --git a/arch/cris/kernel/vmlinux.lds.S b/arch/cris/kernel/vmlinux.lds.S index a3175ebb38cc..6c81836b9229 100644 --- a/arch/cris/kernel/vmlinux.lds.S +++ b/arch/cris/kernel/vmlinux.lds.S @@ -140,13 +140,7 @@ SECTIONS _end = .; __end = .; - /* Sections to be discarded */ - /DISCARD/ : { - EXIT_TEXT - EXIT_DATA - *(.exitcall.exit) - *(.discard) - } - dram_end = dram_start + (CONFIG_ETRAX_DRAM_SIZE - __CONFIG_ETRAX_VMEM_SIZE)*1024*1024; + + DISCARDS } diff --git a/arch/frv/kernel/vmlinux.lds.S b/arch/frv/kernel/vmlinux.lds.S index 64b5a5e4d35e..7dbf41f68b52 100644 --- a/arch/frv/kernel/vmlinux.lds.S +++ b/arch/frv/kernel/vmlinux.lds.S @@ -178,7 +178,7 @@ SECTIONS .comment 0 : { *(.comment) } - /DISCARD/ : { *(.discard) } + DISCARDS } __kernel_image_size_no_bss = __bss_start - __kernel_image_start; diff --git a/arch/h8300/kernel/vmlinux.lds.S b/arch/h8300/kernel/vmlinux.lds.S index 03d6c0df33db..662b02ecb86e 100644 --- a/arch/h8300/kernel/vmlinux.lds.S +++ b/arch/h8300/kernel/vmlinux.lds.S @@ -152,10 +152,6 @@ SECTIONS __end = . ; __ramstart = .; } - /DISCARD/ : { - *(.exitcall.exit) - *(.discard) - } .romfs : { *(.romfs*) @@ -166,4 +162,6 @@ SECTIONS COMMAND_START = . - 0x200 ; __ramend = . ; } + + DISCARDS } diff --git a/arch/ia64/kernel/vmlinux.lds.S b/arch/ia64/kernel/vmlinux.lds.S index 13d958975874..eb4214d1c5af 100644 --- a/arch/ia64/kernel/vmlinux.lds.S +++ b/arch/ia64/kernel/vmlinux.lds.S @@ -24,15 +24,14 @@ PHDRS { } SECTIONS { - /* Sections to be discarded */ + /* unwind exit sections must be discarded before the rest of the + sections get included. */ /DISCARD/ : { - EXIT_TEXT - EXIT_DATA - *(.exitcall.exit) - *(.discard) *(.IA_64.unwind.exit.text) *(.IA_64.unwind_info.exit.text) - } + *(.comment) + *(.note) + } v = PAGE_OFFSET; /* this symbol is here to make debugging easier... */ phys_start = _start - LOAD_OFFSET; @@ -317,7 +316,7 @@ SECTIONS .debug_funcnames 0 : { *(.debug_funcnames) } .debug_typenames 0 : { *(.debug_typenames) } .debug_varnames 0 : { *(.debug_varnames) } - /* These must appear regardless of . */ - /DISCARD/ : { *(.comment) } - /DISCARD/ : { *(.note) } + + /* Default discards */ + DISCARDS } diff --git a/arch/m32r/kernel/vmlinux.lds.S b/arch/m32r/kernel/vmlinux.lds.S index 480a49944cfd..de5e21cca6a5 100644 --- a/arch/m32r/kernel/vmlinux.lds.S +++ b/arch/m32r/kernel/vmlinux.lds.S @@ -120,14 +120,6 @@ SECTIONS _end = . ; - /* Sections to be discarded */ - /DISCARD/ : { - EXIT_TEXT - EXIT_DATA - *(.exitcall.exit) - *(.discard) - } - /* Stabs debugging sections. */ .stab 0 : { *(.stab) } .stabstr 0 : { *(.stabstr) } @@ -136,4 +128,7 @@ SECTIONS .stab.index 0 : { *(.stab.index) } .stab.indexstr 0 : { *(.stab.indexstr) } .comment 0 : { *(.comment) } + + /* Sections to be discarded */ + DISCARDS } diff --git a/arch/m68k/kernel/vmlinux-std.lds b/arch/m68k/kernel/vmlinux-std.lds index 905a797ada93..47eac19e8f61 100644 --- a/arch/m68k/kernel/vmlinux-std.lds +++ b/arch/m68k/kernel/vmlinux-std.lds @@ -82,14 +82,6 @@ SECTIONS _end = . ; - /* Sections to be discarded */ - /DISCARD/ : { - EXIT_TEXT - EXIT_DATA - *(.exitcall.exit) - *(.discard) - } - /* Stabs debugging sections. */ .stab 0 : { *(.stab) } .stabstr 0 : { *(.stabstr) } @@ -98,4 +90,7 @@ SECTIONS .stab.index 0 : { *(.stab.index) } .stab.indexstr 0 : { *(.stab.indexstr) } .comment 0 : { *(.comment) } + + /* Sections to be discarded */ + DISCARDS } diff --git a/arch/m68k/kernel/vmlinux-sun3.lds b/arch/m68k/kernel/vmlinux-sun3.lds index 47d04be322aa..03efaf04d7d7 100644 --- a/arch/m68k/kernel/vmlinux-sun3.lds +++ b/arch/m68k/kernel/vmlinux-sun3.lds @@ -77,14 +77,6 @@ __init_begin = .; _end = . ; - /* Sections to be discarded */ - /DISCARD/ : { - EXIT_TEXT - EXIT_DATA - *(.exitcall.exit) - *(.discard) - } - .crap : { /* Stabs debugging sections. */ *(.stab) @@ -97,4 +89,6 @@ __init_begin = .; *(.note) } + /* Sections to be discarded */ + DISCARDS } diff --git a/arch/m68knommu/kernel/vmlinux.lds.S b/arch/m68knommu/kernel/vmlinux.lds.S index 68111a61a77f..2736a5e309c0 100644 --- a/arch/m68knommu/kernel/vmlinux.lds.S +++ b/arch/m68knommu/kernel/vmlinux.lds.S @@ -184,13 +184,6 @@ SECTIONS { __init_end = .; } > INIT - /DISCARD/ : { - EXIT_TEXT - EXIT_DATA - *(.exitcall.exit) - *(.discard) - } - .bss : { . = ALIGN(4); _sbss = . ; @@ -201,5 +194,6 @@ SECTIONS { _end = . ; } > BSS + DISCARDS } diff --git a/arch/microblaze/kernel/vmlinux.lds.S b/arch/microblaze/kernel/vmlinux.lds.S index 81bebdcb18fe..ec5fa91a48d8 100644 --- a/arch/microblaze/kernel/vmlinux.lds.S +++ b/arch/microblaze/kernel/vmlinux.lds.S @@ -163,5 +163,5 @@ SECTIONS { . = ALIGN(4096); _end = .; - /DISCARD/ : { *(.discard) } + DISCARDS } diff --git a/arch/mips/kernel/vmlinux.lds.S b/arch/mips/kernel/vmlinux.lds.S index 45901609b741..1474c18fb777 100644 --- a/arch/mips/kernel/vmlinux.lds.S +++ b/arch/mips/kernel/vmlinux.lds.S @@ -176,18 +176,6 @@ SECTIONS _end = . ; - /* Sections to be discarded */ - /DISCARD/ : { - *(.exitcall.exit) - *(.discard) - - /* ABI crap starts here */ - *(.MIPS.options) - *(.options) - *(.pdr) - *(.reginfo) - } - /* These mark the ABI of the kernel for debuggers. */ .mdebug.abi32 : { KEEP(*(.mdebug.abi32)) @@ -213,4 +201,14 @@ SECTIONS *(.gptab.bss) *(.gptab.sbss) } + + /* Sections to be discarded */ + DISCARDS + /DISCARD/ : { + /* ABI crap starts here */ + *(.MIPS.options) + *(.options) + *(.pdr) + *(.reginfo) + } } diff --git a/arch/mn10300/kernel/vmlinux.lds.S b/arch/mn10300/kernel/vmlinux.lds.S index 5609d4962a55..8fcd0f1e21de 100644 --- a/arch/mn10300/kernel/vmlinux.lds.S +++ b/arch/mn10300/kernel/vmlinux.lds.S @@ -115,13 +115,10 @@ SECTIONS . = ALIGN(PAGE_SIZE); pg0 = .; - /* Sections to be discarded */ - /DISCARD/ : { - EXIT_CALL - *(.discard) - } - STABS_DEBUG DWARF_DEBUG + + /* Sections to be discarded */ + DISCARDS } diff --git a/arch/parisc/kernel/vmlinux.lds.S b/arch/parisc/kernel/vmlinux.lds.S index ccf58341845a..aea1784edbd1 100644 --- a/arch/parisc/kernel/vmlinux.lds.S +++ b/arch/parisc/kernel/vmlinux.lds.S @@ -237,10 +237,12 @@ SECTIONS /* freed after init ends here */ _end = . ; + STABS_DEBUG + .note 0 : { *(.note) } + /* Sections to be discarded */ + DISCARDS /DISCARD/ : { - *(.exitcall.exit) - *(.discard) #ifdef CONFIG_64BIT /* temporary hack until binutils is fixed to not emit these * for static binaries @@ -253,7 +255,4 @@ SECTIONS *(.gnu.hash) #endif } - - STABS_DEBUG - .note 0 : { *(.note) } } diff --git a/arch/powerpc/kernel/vmlinux.lds.S b/arch/powerpc/kernel/vmlinux.lds.S index 7fca9355fd3d..244e3658983c 100644 --- a/arch/powerpc/kernel/vmlinux.lds.S +++ b/arch/powerpc/kernel/vmlinux.lds.S @@ -37,13 +37,6 @@ jiffies = jiffies_64 + 4; #endif SECTIONS { - /* Sections to be discarded. */ - /DISCARD/ : { - *(.exitcall.exit) - *(.discard) - EXIT_DATA - } - . = KERNELBASE; /* @@ -299,4 +292,7 @@ SECTIONS . = ALIGN(PAGE_SIZE); _end = . ; PROVIDE32 (end = .); + + /* Sections to be discarded. */ + DISCARDS } diff --git a/arch/s390/kernel/vmlinux.lds.S b/arch/s390/kernel/vmlinux.lds.S index 98867dfea469..82415c75b996 100644 --- a/arch/s390/kernel/vmlinux.lds.S +++ b/arch/s390/kernel/vmlinux.lds.S @@ -157,14 +157,10 @@ SECTIONS _end = . ; - /* Sections to be discarded */ - /DISCARD/ : { - EXIT_DATA - *(.exitcall.exit) - *(.discard) - } - /* Debugging sections. */ STABS_DEBUG DWARF_DEBUG + + /* Sections to be discarded */ + DISCARDS } diff --git a/arch/sh/kernel/vmlinux.lds.S b/arch/sh/kernel/vmlinux.lds.S index 766976d27b21..0ce254bca92f 100644 --- a/arch/sh/kernel/vmlinux.lds.S +++ b/arch/sh/kernel/vmlinux.lds.S @@ -163,17 +163,14 @@ SECTIONS _end = . ; } + STABS_DEBUG + DWARF_DEBUG + /* * When something in the kernel is NOT compiled as a module, the * module cleanup code and data are put into these segments. Both * can then be thrown away, as cleanup code is never called unless * it's a module. */ - /DISCARD/ : { - *(.exitcall.exit) - *(.discard) - } - - STABS_DEBUG - DWARF_DEBUG + DISCARDS } diff --git a/arch/sparc/kernel/vmlinux.lds.S b/arch/sparc/kernel/vmlinux.lds.S index d63cf914667d..866390feb683 100644 --- a/arch/sparc/kernel/vmlinux.lds.S +++ b/arch/sparc/kernel/vmlinux.lds.S @@ -171,13 +171,8 @@ SECTIONS } _end = . ; - /DISCARD/ : { - EXIT_TEXT - EXIT_DATA - *(.exitcall.exit) - *(.discard) - } - STABS_DEBUG DWARF_DEBUG + + DISCARDS } diff --git a/arch/um/include/asm/common.lds.S b/arch/um/include/asm/common.lds.S index cb0248616d49..37ecc5577a9a 100644 --- a/arch/um/include/asm/common.lds.S +++ b/arch/um/include/asm/common.lds.S @@ -123,8 +123,3 @@ __initramfs_end = .; } - /* Sections to be discarded */ - /DISCARD/ : { - *(.exitcall.exit) - } - diff --git a/arch/um/kernel/dyn.lds.S b/arch/um/kernel/dyn.lds.S index 2916d6eadffd..715a188c0472 100644 --- a/arch/um/kernel/dyn.lds.S +++ b/arch/um/kernel/dyn.lds.S @@ -157,5 +157,5 @@ SECTIONS DWARF_DEBUG - /DISCARD/ : { *(.discard) } + DISCARDS } diff --git a/arch/um/kernel/uml.lds.S b/arch/um/kernel/uml.lds.S index 1f8a622cabe1..2ebd39765db8 100644 --- a/arch/um/kernel/uml.lds.S +++ b/arch/um/kernel/uml.lds.S @@ -101,5 +101,5 @@ SECTIONS DWARF_DEBUG - /DISCARD/ : { *(.discard) } + DISCARDS } diff --git a/arch/x86/kernel/vmlinux.lds.S b/arch/x86/kernel/vmlinux.lds.S index 367e87882041..b600c843710b 100644 --- a/arch/x86/kernel/vmlinux.lds.S +++ b/arch/x86/kernel/vmlinux.lds.S @@ -387,15 +387,12 @@ SECTIONS _end = .; } - /* Sections to be discarded */ - /DISCARD/ : { - *(.exitcall.exit) - *(.eh_frame) - *(.discard) - } - STABS_DEBUG DWARF_DEBUG + + /* Sections to be discarded */ + DISCARDS + /DISCARD/ : { *(.eh_frame) } } diff --git a/arch/xtensa/kernel/vmlinux.lds.S b/arch/xtensa/kernel/vmlinux.lds.S index b1e24638acd7..921b6ff3b645 100644 --- a/arch/xtensa/kernel/vmlinux.lds.S +++ b/arch/xtensa/kernel/vmlinux.lds.S @@ -280,16 +280,6 @@ SECTIONS *(.ResetVector.text) } - /* Sections to be discarded */ - /DISCARD/ : - { - *(.exit.literal) - EXIT_TEXT - EXIT_DATA - *(.exitcall.exit) - *(.discard) - } - .xt.lit : { *(.xt.lit) } .xt.prop : { *(.xt.prop) } @@ -322,4 +312,8 @@ SECTIONS *(.xt.lit) *(.gnu.linkonce.p*) } + + /* Sections to be discarded */ + DISCARDS + /DISCARD/ : { *(.exit.literal) } } diff --git a/include/asm-generic/vmlinux.lds.h b/include/asm-generic/vmlinux.lds.h index c5c18ac878ab..ab8ea9b7741e 100644 --- a/include/asm-generic/vmlinux.lds.h +++ b/include/asm-generic/vmlinux.lds.h @@ -35,13 +35,10 @@ * __bss_stop = .; * _end = .; * - * /DISCARD/ : { - * EXIT_TEXT - * EXIT_DATA - * EXIT_CALL - * } * STABS_DEBUG * DWARF_DEBUG + * + * DISCARDS // must be the last * } * * [__init_begin, __init_end] is the init section that may be freed after init @@ -629,11 +626,20 @@ #define INIT_RAM_FS #endif +/* + * Default discarded sections. + * + * Some archs want to discard exit text/data at runtime rather than + * link time due to cross-section references such as alt instructions, + * bug table, eh_frame, etc. DISCARDS must be the last of output + * section definitions so that such archs put those in earlier section + * definitions. + */ #define DISCARDS \ /DISCARD/ : { \ EXIT_TEXT \ EXIT_DATA \ - *(.exitcall.exit) \ + EXIT_CALL \ *(.discard) \ } -- cgit v1.2.3 From ef5d590c9b9a00b017683006fe4cf959b9532336 Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Thu, 9 Jul 2009 11:32:55 +0800 Subject: crypto: shash - Add shash_instance_ctx This patch adds the helper shash_instance_ctx which is the shash analogue of crypto_instance_ctx. Signed-off-by: Herbert Xu --- include/crypto/internal/hash.h | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'include') diff --git a/include/crypto/internal/hash.h b/include/crypto/internal/hash.h index 3af34ca8e8bc..069b93e1a8ec 100644 --- a/include/crypto/internal/hash.h +++ b/include/crypto/internal/hash.h @@ -110,6 +110,11 @@ static inline struct shash_instance *shash_instance( struct shash_instance, alg); } +static inline void *shash_instance_ctx(struct shash_instance *inst) +{ + return crypto_instance_ctx(shash_crypto_instance(inst)); +} + static inline struct shash_instance *shash_alloc_instance( const char *name, struct crypto_alg *alg) { -- cgit v1.2.3 From ed5215a21460f63d6bdc118cb55a9e6d1b433f35 Mon Sep 17 00:00:00 2001 From: Thomas Liu Date: Thu, 9 Jul 2009 10:00:29 -0400 Subject: Move variable function in lsm_audit.h into SMACK private space Moved variable function in include/linux/lsm_audit.h into the smack_audit_data struct since it is never used outside of it. Also removed setting of function in the COMMON_AUDIT_DATA_INIT macro because that variable is now private to SMACK. Signed-off-by: Thomas Liu Acked-by: Eric Paris I-dont-see-any-problems-with-it: Casey Schaufler Signed-off-by: James Morris --- include/linux/lsm_audit.h | 4 ++-- security/smack/smack.h | 2 +- security/smack/smack_access.c | 7 ++++--- 3 files changed, 7 insertions(+), 6 deletions(-) (limited to 'include') diff --git a/include/linux/lsm_audit.h b/include/linux/lsm_audit.h index e461b2c3d711..68f7bce572b0 100644 --- a/include/linux/lsm_audit.h +++ b/include/linux/lsm_audit.h @@ -66,11 +66,11 @@ struct common_audit_data { } key_struct; #endif } u; - const char *function; /* this union contains LSM specific data */ union { /* SMACK data */ struct smack_audit_data { + const char *function; char *subject; char *object; char *request; @@ -104,7 +104,7 @@ int ipv6_skb_to_auditdata(struct sk_buff *skb, /* Initialize an LSM audit data structure. */ #define COMMON_AUDIT_DATA_INIT(_d, _t) \ { memset((_d), 0, sizeof(struct common_audit_data)); \ - (_d)->type = LSM_AUDIT_DATA_##_t; (_d)->function = __func__; } + (_d)->type = LSM_AUDIT_DATA_##_t; } void common_lsm_audit(struct common_audit_data *a); diff --git a/security/smack/smack.h b/security/smack/smack.h index 243bec175be0..ff180ede3e47 100644 --- a/security/smack/smack.h +++ b/security/smack/smack.h @@ -275,7 +275,7 @@ static inline void smk_ad_init(struct smk_audit_info *a, const char *func, { memset(a, 0, sizeof(*a)); a->a.type = type; - a->a.function = func; + a->a.lsm_priv.smack_audit_data.function = func; } static inline void smk_ad_setfield_u_tsk(struct smk_audit_info *a, diff --git a/security/smack/smack_access.c b/security/smack/smack_access.c index 513dc1aa16dd..dd84877dff30 100644 --- a/security/smack/smack_access.c +++ b/security/smack/smack_access.c @@ -241,7 +241,8 @@ static void smack_log_callback(struct audit_buffer *ab, void *a) { struct common_audit_data *ad = a; struct smack_audit_data *sad = &ad->lsm_priv.smack_audit_data; - audit_log_format(ab, "lsm=SMACK fn=%s action=%s", ad->function, + audit_log_format(ab, "lsm=SMACK fn=%s action=%s", + ad->lsm_priv.smack_audit_data.function, sad->result ? "denied" : "granted"); audit_log_format(ab, " subject="); audit_log_untrustedstring(ab, sad->subject); @@ -274,8 +275,8 @@ void smack_log(char *subject_label, char *object_label, int request, if (result == 0 && (log_policy & SMACK_AUDIT_ACCEPT) == 0) return; - if (a->function == NULL) - a->function = "unknown"; + if (a->lsm_priv.smack_audit_data.function == NULL) + a->lsm_priv.smack_audit_data.function = "unknown"; /* end preparing the audit data */ sad = &a->lsm_priv.smack_audit_data; -- cgit v1.2.3 From d4131ded4d4c1a5c1363ddd93ca104ed97dd0458 Mon Sep 17 00:00:00 2001 From: Thomas Liu Date: Thu, 9 Jul 2009 10:00:30 -0400 Subject: security: Make lsm_priv union in lsm_audit.h anonymous Made the lsm_priv union in include/linux/lsm_audit.h anonymous. Signed-off-by: Thomas Liu Acked-by: Eric Paris Signed-off-by: James Morris --- include/linux/lsm_audit.h | 2 +- security/smack/smack.h | 2 +- security/smack/smack_access.c | 10 +++++----- 3 files changed, 7 insertions(+), 7 deletions(-) (limited to 'include') diff --git a/include/linux/lsm_audit.h b/include/linux/lsm_audit.h index 68f7bce572b0..40d1b84f2a3c 100644 --- a/include/linux/lsm_audit.h +++ b/include/linux/lsm_audit.h @@ -86,7 +86,7 @@ struct common_audit_data { struct av_decision *avd; int result; } selinux_audit_data; - } lsm_priv; + }; /* these callback will be implemented by a specific LSM */ void (*lsm_pre_audit)(struct audit_buffer *, void *); void (*lsm_post_audit)(struct audit_buffer *, void *); diff --git a/security/smack/smack.h b/security/smack/smack.h index ff180ede3e47..c6e9acae72e4 100644 --- a/security/smack/smack.h +++ b/security/smack/smack.h @@ -275,7 +275,7 @@ static inline void smk_ad_init(struct smk_audit_info *a, const char *func, { memset(a, 0, sizeof(*a)); a->a.type = type; - a->a.lsm_priv.smack_audit_data.function = func; + a->a.smack_audit_data.function = func; } static inline void smk_ad_setfield_u_tsk(struct smk_audit_info *a, diff --git a/security/smack/smack_access.c b/security/smack/smack_access.c index dd84877dff30..0f9ac8146900 100644 --- a/security/smack/smack_access.c +++ b/security/smack/smack_access.c @@ -240,9 +240,9 @@ static inline void smack_str_from_perm(char *string, int access) static void smack_log_callback(struct audit_buffer *ab, void *a) { struct common_audit_data *ad = a; - struct smack_audit_data *sad = &ad->lsm_priv.smack_audit_data; + struct smack_audit_data *sad = &ad->smack_audit_data; audit_log_format(ab, "lsm=SMACK fn=%s action=%s", - ad->lsm_priv.smack_audit_data.function, + ad->smack_audit_data.function, sad->result ? "denied" : "granted"); audit_log_format(ab, " subject="); audit_log_untrustedstring(ab, sad->subject); @@ -275,11 +275,11 @@ void smack_log(char *subject_label, char *object_label, int request, if (result == 0 && (log_policy & SMACK_AUDIT_ACCEPT) == 0) return; - if (a->lsm_priv.smack_audit_data.function == NULL) - a->lsm_priv.smack_audit_data.function = "unknown"; + if (a->smack_audit_data.function == NULL) + a->smack_audit_data.function = "unknown"; /* end preparing the audit data */ - sad = &a->lsm_priv.smack_audit_data; + sad = &a->smack_audit_data; smack_str_from_perm(request_buffer, request); sad->subject = subject_label; sad->object = object_label; -- cgit v1.2.3 From 65c3f0a2d0f72d210c879e4974c2d222b7951321 Mon Sep 17 00:00:00 2001 From: Thomas Liu Date: Thu, 9 Jul 2009 10:00:31 -0400 Subject: security: Wrap SMACK and SELINUX audit data structs in ifdefs Wrapped the smack_audit_data and selinux_audit_data structs in include/linux/lsm_audit.h in ifdefs so that the union will always be the correct size. Signed-off-by: Thomas Liu Acked-by: Eric Paris Signed-off-by: James Morris --- include/linux/lsm_audit.h | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'include') diff --git a/include/linux/lsm_audit.h b/include/linux/lsm_audit.h index 40d1b84f2a3c..a5514a3a4f17 100644 --- a/include/linux/lsm_audit.h +++ b/include/linux/lsm_audit.h @@ -68,6 +68,7 @@ struct common_audit_data { } u; /* this union contains LSM specific data */ union { +#ifdef CONFIG_SECURITY_SMACK /* SMACK data */ struct smack_audit_data { const char *function; @@ -76,6 +77,8 @@ struct common_audit_data { char *request; int result; } smack_audit_data; +#endif +#ifdef CONFIG_SECURITY_SELINUX /* SELinux data */ struct { u32 ssid; @@ -86,6 +89,7 @@ struct common_audit_data { struct av_decision *avd; int result; } selinux_audit_data; +#endif }; /* these callback will be implemented by a specific LSM */ void (*lsm_pre_audit)(struct audit_buffer *, void *); -- cgit v1.2.3 From a57de0b4336e48db2811a2030bb68dba8dd09d88 Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Wed, 8 Jul 2009 12:09:13 +0000 Subject: net: adding memory barrier to the poll and receive callbacks Adding memory barrier after the poll_wait function, paired with receive callbacks. Adding fuctions sock_poll_wait and sk_has_sleeper to wrap the memory barrier. Without the memory barrier, following race can happen. The race fires, when following code paths meet, and the tp->rcv_nxt and __add_wait_queue updates stay in CPU caches. CPU1 CPU2 sys_select receive packet ... ... __add_wait_queue update tp->rcv_nxt ... ... tp->rcv_nxt check sock_def_readable ... { schedule ... if (sk->sk_sleep && waitqueue_active(sk->sk_sleep)) wake_up_interruptible(sk->sk_sleep) ... } If there was no cache the code would work ok, since the wait_queue and rcv_nxt are opposit to each other. Meaning that once tp->rcv_nxt is updated by CPU2, the CPU1 either already passed the tp->rcv_nxt check and sleeps, or will get the new value for tp->rcv_nxt and will return with new data mask. In both cases the process (CPU1) is being added to the wait queue, so the waitqueue_active (CPU2) call cannot miss and will wake up CPU1. The bad case is when the __add_wait_queue changes done by CPU1 stay in its cache, and so does the tp->rcv_nxt update on CPU2 side. The CPU1 will then endup calling schedule and sleep forever if there are no more data on the socket. Calls to poll_wait in following modules were ommited: net/bluetooth/af_bluetooth.c net/irda/af_irda.c net/irda/irnet/irnet_ppp.c net/mac80211/rc80211_pid_debugfs.c net/phonet/socket.c net/rds/af_rds.c net/rfkill/core.c net/sunrpc/cache.c net/sunrpc/rpc_pipe.c net/tipc/socket.c Signed-off-by: Jiri Olsa Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- include/net/sock.h | 66 ++++++++++++++++++++++++++++++++++++++++++++++++++++ net/atm/common.c | 6 ++--- net/core/datagram.c | 2 +- net/core/sock.c | 8 +++---- net/dccp/output.c | 2 +- net/dccp/proto.c | 2 +- net/ipv4/tcp.c | 2 +- net/iucv/af_iucv.c | 4 ++-- net/rxrpc/af_rxrpc.c | 4 ++-- net/unix/af_unix.c | 8 +++---- 10 files changed, 85 insertions(+), 19 deletions(-) (limited to 'include') diff --git a/include/net/sock.h b/include/net/sock.h index 352f06bbd7a9..4eb8409249f6 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -54,6 +54,7 @@ #include #include +#include #include #include @@ -1241,6 +1242,71 @@ static inline int sk_has_allocations(const struct sock *sk) return sk_wmem_alloc_get(sk) || sk_rmem_alloc_get(sk); } +/** + * sk_has_sleeper - check if there are any waiting processes + * @sk: socket + * + * Returns true if socket has waiting processes + * + * The purpose of the sk_has_sleeper and sock_poll_wait is to wrap the memory + * barrier call. They were added due to the race found within the tcp code. + * + * Consider following tcp code paths: + * + * CPU1 CPU2 + * + * sys_select receive packet + * ... ... + * __add_wait_queue update tp->rcv_nxt + * ... ... + * tp->rcv_nxt check sock_def_readable + * ... { + * schedule ... + * if (sk->sk_sleep && waitqueue_active(sk->sk_sleep)) + * wake_up_interruptible(sk->sk_sleep) + * ... + * } + * + * The race for tcp fires when the __add_wait_queue changes done by CPU1 stay + * in its cache, and so does the tp->rcv_nxt update on CPU2 side. The CPU1 + * could then endup calling schedule and sleep forever if there are no more + * data on the socket. + */ +static inline int sk_has_sleeper(struct sock *sk) +{ + /* + * We need to be sure we are in sync with the + * add_wait_queue modifications to the wait queue. + * + * This memory barrier is paired in the sock_poll_wait. + */ + smp_mb(); + return sk->sk_sleep && waitqueue_active(sk->sk_sleep); +} + +/** + * sock_poll_wait - place memory barrier behind the poll_wait call. + * @filp: file + * @wait_address: socket wait queue + * @p: poll_table + * + * See the comments in the sk_has_sleeper function. + */ +static inline void sock_poll_wait(struct file *filp, + wait_queue_head_t *wait_address, poll_table *p) +{ + if (p && wait_address) { + poll_wait(filp, wait_address, p); + /* + * We need to be sure we are in sync with the + * socket flags modification. + * + * This memory barrier is paired in the sk_has_sleeper. + */ + smp_mb(); + } +} + /* * Queue a received datagram if it will fit. Stream and sequenced * protocols can't normally use this as they need to fit buffers in diff --git a/net/atm/common.c b/net/atm/common.c index c1c97936192c..8c4d843eb17f 100644 --- a/net/atm/common.c +++ b/net/atm/common.c @@ -92,7 +92,7 @@ static void vcc_sock_destruct(struct sock *sk) static void vcc_def_wakeup(struct sock *sk) { read_lock(&sk->sk_callback_lock); - if (sk->sk_sleep && waitqueue_active(sk->sk_sleep)) + if (sk_has_sleeper(sk)) wake_up(sk->sk_sleep); read_unlock(&sk->sk_callback_lock); } @@ -110,7 +110,7 @@ static void vcc_write_space(struct sock *sk) read_lock(&sk->sk_callback_lock); if (vcc_writable(sk)) { - if (sk->sk_sleep && waitqueue_active(sk->sk_sleep)) + if (sk_has_sleeper(sk)) wake_up_interruptible(sk->sk_sleep); sk_wake_async(sk, SOCK_WAKE_SPACE, POLL_OUT); @@ -594,7 +594,7 @@ unsigned int vcc_poll(struct file *file, struct socket *sock, poll_table *wait) struct atm_vcc *vcc; unsigned int mask; - poll_wait(file, sk->sk_sleep, wait); + sock_poll_wait(file, sk->sk_sleep, wait); mask = 0; vcc = ATM_SD(sock); diff --git a/net/core/datagram.c b/net/core/datagram.c index 58abee1f1df1..b0fe69211eef 100644 --- a/net/core/datagram.c +++ b/net/core/datagram.c @@ -712,7 +712,7 @@ unsigned int datagram_poll(struct file *file, struct socket *sock, struct sock *sk = sock->sk; unsigned int mask; - poll_wait(file, sk->sk_sleep, wait); + sock_poll_wait(file, sk->sk_sleep, wait); mask = 0; /* exceptional events? */ diff --git a/net/core/sock.c b/net/core/sock.c index b0ba569bc973..6354863b1c68 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -1715,7 +1715,7 @@ EXPORT_SYMBOL(sock_no_sendpage); static void sock_def_wakeup(struct sock *sk) { read_lock(&sk->sk_callback_lock); - if (sk->sk_sleep && waitqueue_active(sk->sk_sleep)) + if (sk_has_sleeper(sk)) wake_up_interruptible_all(sk->sk_sleep); read_unlock(&sk->sk_callback_lock); } @@ -1723,7 +1723,7 @@ static void sock_def_wakeup(struct sock *sk) static void sock_def_error_report(struct sock *sk) { read_lock(&sk->sk_callback_lock); - if (sk->sk_sleep && waitqueue_active(sk->sk_sleep)) + if (sk_has_sleeper(sk)) wake_up_interruptible_poll(sk->sk_sleep, POLLERR); sk_wake_async(sk, SOCK_WAKE_IO, POLL_ERR); read_unlock(&sk->sk_callback_lock); @@ -1732,7 +1732,7 @@ static void sock_def_error_report(struct sock *sk) static void sock_def_readable(struct sock *sk, int len) { read_lock(&sk->sk_callback_lock); - if (sk->sk_sleep && waitqueue_active(sk->sk_sleep)) + if (sk_has_sleeper(sk)) wake_up_interruptible_sync_poll(sk->sk_sleep, POLLIN | POLLRDNORM | POLLRDBAND); sk_wake_async(sk, SOCK_WAKE_WAITD, POLL_IN); @@ -1747,7 +1747,7 @@ static void sock_def_write_space(struct sock *sk) * progress. --DaveM */ if ((atomic_read(&sk->sk_wmem_alloc) << 1) <= sk->sk_sndbuf) { - if (sk->sk_sleep && waitqueue_active(sk->sk_sleep)) + if (sk_has_sleeper(sk)) wake_up_interruptible_sync_poll(sk->sk_sleep, POLLOUT | POLLWRNORM | POLLWRBAND); diff --git a/net/dccp/output.c b/net/dccp/output.c index c0e88c16d088..c96119fda688 100644 --- a/net/dccp/output.c +++ b/net/dccp/output.c @@ -196,7 +196,7 @@ void dccp_write_space(struct sock *sk) { read_lock(&sk->sk_callback_lock); - if (sk->sk_sleep && waitqueue_active(sk->sk_sleep)) + if (sk_has_sleeper(sk)) wake_up_interruptible(sk->sk_sleep); /* Should agree with poll, otherwise some programs break */ if (sock_writeable(sk)) diff --git a/net/dccp/proto.c b/net/dccp/proto.c index 314a1b5c033c..94ca8eaace7d 100644 --- a/net/dccp/proto.c +++ b/net/dccp/proto.c @@ -311,7 +311,7 @@ unsigned int dccp_poll(struct file *file, struct socket *sock, unsigned int mask; struct sock *sk = sock->sk; - poll_wait(file, sk->sk_sleep, wait); + sock_poll_wait(file, sk->sk_sleep, wait); if (sk->sk_state == DCCP_LISTEN) return inet_csk_listen_poll(sk); diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c index 7870a535dac6..91145244ea63 100644 --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@ -339,7 +339,7 @@ unsigned int tcp_poll(struct file *file, struct socket *sock, poll_table *wait) struct sock *sk = sock->sk; struct tcp_sock *tp = tcp_sk(sk); - poll_wait(file, sk->sk_sleep, wait); + sock_poll_wait(file, sk->sk_sleep, wait); if (sk->sk_state == TCP_LISTEN) return inet_csk_listen_poll(sk); diff --git a/net/iucv/af_iucv.c b/net/iucv/af_iucv.c index 6be5f92d1094..49c15b48408e 100644 --- a/net/iucv/af_iucv.c +++ b/net/iucv/af_iucv.c @@ -306,7 +306,7 @@ static inline int iucv_below_msglim(struct sock *sk) static void iucv_sock_wake_msglim(struct sock *sk) { read_lock(&sk->sk_callback_lock); - if (sk->sk_sleep && waitqueue_active(sk->sk_sleep)) + if (sk_has_sleeper(sk)) wake_up_interruptible_all(sk->sk_sleep); sk_wake_async(sk, SOCK_WAKE_SPACE, POLL_OUT); read_unlock(&sk->sk_callback_lock); @@ -1256,7 +1256,7 @@ unsigned int iucv_sock_poll(struct file *file, struct socket *sock, struct sock *sk = sock->sk; unsigned int mask = 0; - poll_wait(file, sk->sk_sleep, wait); + sock_poll_wait(file, sk->sk_sleep, wait); if (sk->sk_state == IUCV_LISTEN) return iucv_accept_poll(sk); diff --git a/net/rxrpc/af_rxrpc.c b/net/rxrpc/af_rxrpc.c index eac5e7bb7365..bfe493ebf27c 100644 --- a/net/rxrpc/af_rxrpc.c +++ b/net/rxrpc/af_rxrpc.c @@ -63,7 +63,7 @@ static void rxrpc_write_space(struct sock *sk) _enter("%p", sk); read_lock(&sk->sk_callback_lock); if (rxrpc_writable(sk)) { - if (sk->sk_sleep && waitqueue_active(sk->sk_sleep)) + if (sk_has_sleeper(sk)) wake_up_interruptible(sk->sk_sleep); sk_wake_async(sk, SOCK_WAKE_SPACE, POLL_OUT); } @@ -588,7 +588,7 @@ static unsigned int rxrpc_poll(struct file *file, struct socket *sock, unsigned int mask; struct sock *sk = sock->sk; - poll_wait(file, sk->sk_sleep, wait); + sock_poll_wait(file, sk->sk_sleep, wait); mask = 0; /* the socket is readable if there are any messages waiting on the Rx diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c index 36d4e44d6233..fc3ebb906911 100644 --- a/net/unix/af_unix.c +++ b/net/unix/af_unix.c @@ -315,7 +315,7 @@ static void unix_write_space(struct sock *sk) { read_lock(&sk->sk_callback_lock); if (unix_writable(sk)) { - if (sk->sk_sleep && waitqueue_active(sk->sk_sleep)) + if (sk_has_sleeper(sk)) wake_up_interruptible_sync(sk->sk_sleep); sk_wake_async(sk, SOCK_WAKE_SPACE, POLL_OUT); } @@ -1985,7 +1985,7 @@ static unsigned int unix_poll(struct file *file, struct socket *sock, poll_table struct sock *sk = sock->sk; unsigned int mask; - poll_wait(file, sk->sk_sleep, wait); + sock_poll_wait(file, sk->sk_sleep, wait); mask = 0; /* exceptional events? */ @@ -2022,7 +2022,7 @@ static unsigned int unix_dgram_poll(struct file *file, struct socket *sock, struct sock *sk = sock->sk, *other; unsigned int mask, writable; - poll_wait(file, sk->sk_sleep, wait); + sock_poll_wait(file, sk->sk_sleep, wait); mask = 0; /* exceptional events? */ @@ -2053,7 +2053,7 @@ static unsigned int unix_dgram_poll(struct file *file, struct socket *sock, other = unix_peer_get(sk); if (other) { if (unix_peer(other) != sk) { - poll_wait(file, &unix_sk(other)->peer_wait, + sock_poll_wait(file, &unix_sk(other)->peer_wait, wait); if (unix_recvq_full(other)) writable = 0; -- cgit v1.2.3 From ad46276952f1af34cd91d46d49ba13d347d56367 Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Wed, 8 Jul 2009 12:10:31 +0000 Subject: memory barrier: adding smp_mb__after_lock Adding smp_mb__after_lock define to be used as a smp_mb call after a lock. Making it nop for x86, since {read|write|spin}_lock() on x86 are full memory barriers. Signed-off-by: Jiri Olsa Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- arch/x86/include/asm/spinlock.h | 4 ++++ include/linux/spinlock.h | 5 +++++ include/net/sock.h | 5 ++++- 3 files changed, 13 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/arch/x86/include/asm/spinlock.h b/arch/x86/include/asm/spinlock.h index b7e5db876399..4e77853321db 100644 --- a/arch/x86/include/asm/spinlock.h +++ b/arch/x86/include/asm/spinlock.h @@ -302,4 +302,8 @@ static inline void __raw_write_unlock(raw_rwlock_t *rw) #define _raw_read_relax(lock) cpu_relax() #define _raw_write_relax(lock) cpu_relax() +/* The {read|write|spin}_lock() on x86 are full memory barriers. */ +static inline void smp_mb__after_lock(void) { } +#define ARCH_HAS_SMP_MB_AFTER_LOCK + #endif /* _ASM_X86_SPINLOCK_H */ diff --git a/include/linux/spinlock.h b/include/linux/spinlock.h index 252b245cfcf4..4be57ab03478 100644 --- a/include/linux/spinlock.h +++ b/include/linux/spinlock.h @@ -132,6 +132,11 @@ do { \ #endif /*__raw_spin_is_contended*/ #endif +/* The lock does not imply full memory barrier. */ +#ifndef ARCH_HAS_SMP_MB_AFTER_LOCK +static inline void smp_mb__after_lock(void) { smp_mb(); } +#endif + /** * spin_unlock_wait - wait until the spinlock gets unlocked * @lock: the spinlock in question. diff --git a/include/net/sock.h b/include/net/sock.h index 4eb8409249f6..2c0da9239b95 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -1271,6 +1271,9 @@ static inline int sk_has_allocations(const struct sock *sk) * in its cache, and so does the tp->rcv_nxt update on CPU2 side. The CPU1 * could then endup calling schedule and sleep forever if there are no more * data on the socket. + * + * The sk_has_sleeper is always called right after a call to read_lock, so we + * can use smp_mb__after_lock barrier. */ static inline int sk_has_sleeper(struct sock *sk) { @@ -1280,7 +1283,7 @@ static inline int sk_has_sleeper(struct sock *sk) * * This memory barrier is paired in the sock_poll_wait. */ - smp_mb(); + smp_mb__after_lock(); return sk->sk_sleep && waitqueue_active(sk->sk_sleep); } -- cgit v1.2.3 From 6ff7041dbfeb3bd7dfe9aa67275c21199ef760d6 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Fri, 10 Jul 2009 14:57:05 +0200 Subject: hrtimer: Fix migration expiry check The timer migration expiry check should prevent the migration of a timer to another CPU when the timer expires before the next event is scheduled on the other CPU. Migrating the timer might delay it because we can not reprogram the clock event device on the other CPU. But the code implementing that check has two flaws: - for !HIGHRES the check compares the expiry value with the clock events device expiry value which is wrong for CLOCK_REALTIME based timers. - the check is racy. It holds the hrtimer base lock of the target CPU, but the clock event device expiry value can be modified nevertheless, e.g. by an timer interrupt firing. The !HIGHRES case is easy to fix as we can enqueue the timer on the cpu which was selected by the load balancer. It runs the idle balancing code once per jiffy anyway. So the maximum delay for the timer is the same as when we keep the tick on the current cpu going. In the HIGHRES case we can get the next expiry value from the hrtimer cpu_base of the target CPU and serialize the update with the cpu_base lock. This moves the lock section in hrtimer_interrupt() so we can set next_event to KTIME_MAX while we are handling the expired timers and set it to the next expiry value after we handled the timers under the base lock. While the expired timers are processed timer migration is blocked because the expiry time of the timer is always <= KTIME_MAX. Also remove the now useless clockevents_get_next_event() function. Signed-off-by: Thomas Gleixner --- include/linux/clockchips.h | 9 ---- kernel/hrtimer.c | 121 ++++++++++++++++++++++++--------------------- kernel/time/clockevents.c | 11 ----- 3 files changed, 64 insertions(+), 77 deletions(-) (limited to 'include') diff --git a/include/linux/clockchips.h b/include/linux/clockchips.h index 20a100fe2b4f..3a1dbba4d3ae 100644 --- a/include/linux/clockchips.h +++ b/include/linux/clockchips.h @@ -143,12 +143,3 @@ extern void clockevents_notify(unsigned long reason, void *arg); #endif #endif - -#ifdef CONFIG_GENERIC_CLOCKEVENTS -extern ktime_t clockevents_get_next_event(int cpu); -#else -static inline ktime_t clockevents_get_next_event(int cpu) -{ - return (ktime_t) { .tv64 = KTIME_MAX }; -} -#endif diff --git a/kernel/hrtimer.c b/kernel/hrtimer.c index 126b9808f287..49da79ab8486 100644 --- a/kernel/hrtimer.c +++ b/kernel/hrtimer.c @@ -191,6 +191,46 @@ struct hrtimer_clock_base *lock_hrtimer_base(const struct hrtimer *timer, } } + +/* + * Get the preferred target CPU for NOHZ + */ +static int hrtimer_get_target(int this_cpu, int pinned) +{ +#ifdef CONFIG_NO_HZ + if (!pinned && get_sysctl_timer_migration() && idle_cpu(this_cpu)) { + int preferred_cpu = get_nohz_load_balancer(); + + if (preferred_cpu >= 0) + return preferred_cpu; + } +#endif + return this_cpu; +} + +/* + * With HIGHRES=y we do not migrate the timer when it is expiring + * before the next event on the target cpu because we cannot reprogram + * the target cpu hardware and we would cause it to fire late. + * + * Called with cpu_base->lock of target cpu held. + */ +static int +hrtimer_check_target(struct hrtimer *timer, struct hrtimer_clock_base *new_base) +{ +#ifdef CONFIG_HIGH_RES_TIMERS + ktime_t expires; + + if (!new_base->cpu_base->hres_active) + return 0; + + expires = ktime_sub(hrtimer_get_expires(timer), new_base->offset); + return expires.tv64 <= new_base->cpu_base->expires_next.tv64; +#else + return 0; +#endif +} + /* * Switch the timer base to the current CPU when possible. */ @@ -200,27 +240,8 @@ switch_hrtimer_base(struct hrtimer *timer, struct hrtimer_clock_base *base, { struct hrtimer_clock_base *new_base; struct hrtimer_cpu_base *new_cpu_base; - int cpu, preferred_cpu = -1; - - cpu = smp_processor_id(); -#if defined(CONFIG_NO_HZ) && defined(CONFIG_SMP) - if (!pinned && get_sysctl_timer_migration() && idle_cpu(cpu)) { - preferred_cpu = get_nohz_load_balancer(); - if (preferred_cpu >= 0) { - /* - * We must not check the expiry value when - * preferred_cpu is the current cpu. If base - * != new_base we would loop forever when the - * timer expires before the current programmed - * next timer event. - */ - if (preferred_cpu != cpu) - cpu = preferred_cpu; - else - preferred_cpu = -1; - } - } -#endif + int this_cpu = smp_processor_id(); + int cpu = hrtimer_get_target(this_cpu, pinned); again: new_cpu_base = &per_cpu(hrtimer_bases, cpu); @@ -228,7 +249,7 @@ again: if (base != new_base) { /* - * We are trying to schedule the timer on the local CPU. + * We are trying to move timer to new_base. * However we can't change timer's base while it is running, * so we keep it on the same CPU. No hassle vs. reprogramming * the event source in the high resolution case. The softirq @@ -244,38 +265,12 @@ again: spin_unlock(&base->cpu_base->lock); spin_lock(&new_base->cpu_base->lock); - /* Optimized away for NOHZ=n SMP=n */ - if (cpu == preferred_cpu) { - /* Calculate clock monotonic expiry time */ -#ifdef CONFIG_HIGH_RES_TIMERS - ktime_t expires = ktime_sub(hrtimer_get_expires(timer), - new_base->offset); -#else - ktime_t expires = hrtimer_get_expires(timer); -#endif - - /* - * Get the next event on target cpu from the - * clock events layer. - * This covers the highres=off nohz=on case as well. - */ - ktime_t next = clockevents_get_next_event(cpu); - - ktime_t delta = ktime_sub(expires, next); - - /* - * We do not migrate the timer when it is expiring - * before the next event on the target cpu because - * we cannot reprogram the target cpu hardware and - * we would cause it to fire late. - */ - if (delta.tv64 < 0) { - cpu = smp_processor_id(); - spin_unlock(&new_base->cpu_base->lock); - spin_lock(&base->cpu_base->lock); - timer->base = base; - goto again; - } + if (cpu != this_cpu && hrtimer_check_target(timer, new_base)) { + cpu = this_cpu; + spin_unlock(&new_base->cpu_base->lock); + spin_lock(&base->cpu_base->lock); + timer->base = base; + goto again; } timer->base = new_base; } @@ -1287,14 +1282,22 @@ void hrtimer_interrupt(struct clock_event_device *dev) expires_next.tv64 = KTIME_MAX; + spin_lock(&cpu_base->lock); + /* + * We set expires_next to KTIME_MAX here with cpu_base->lock + * held to prevent that a timer is enqueued in our queue via + * the migration code. This does not affect enqueueing of + * timers which run their callback and need to be requeued on + * this CPU. + */ + cpu_base->expires_next.tv64 = KTIME_MAX; + base = cpu_base->clock_base; for (i = 0; i < HRTIMER_MAX_CLOCK_BASES; i++) { ktime_t basenow; struct rb_node *node; - spin_lock(&cpu_base->lock); - basenow = ktime_add(now, base->offset); while ((node = base->first)) { @@ -1327,11 +1330,15 @@ void hrtimer_interrupt(struct clock_event_device *dev) __run_hrtimer(timer); } - spin_unlock(&cpu_base->lock); base++; } + /* + * Store the new expiry value so the migration code can verify + * against it. + */ cpu_base->expires_next = expires_next; + spin_unlock(&cpu_base->lock); /* Reprogramming necessary ? */ if (expires_next.tv64 != KTIME_MAX) { diff --git a/kernel/time/clockevents.c b/kernel/time/clockevents.c index 1ad6dd461119..a6dcd67b041d 100644 --- a/kernel/time/clockevents.c +++ b/kernel/time/clockevents.c @@ -254,15 +254,4 @@ void clockevents_notify(unsigned long reason, void *arg) spin_unlock(&clockevents_lock); } EXPORT_SYMBOL_GPL(clockevents_notify); - -ktime_t clockevents_get_next_event(int cpu) -{ - struct tick_device *td; - struct clock_event_device *dev; - - td = &per_cpu(tick_cpu_device, cpu); - dev = td->evtdev; - - return dev->next_event; -} #endif -- cgit v1.2.3 From f9f868dbcca961ed62f1df1d114abd0c38c47dce Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Thu, 9 Jul 2009 11:24:26 +0200 Subject: timer stats: fix quick check optimization git commit 507e1231 "timer stats: Optimize by adding quick check to avoid function calls" added one wrong check so that one unnecessary function call isn't elimated. time_stats_account_hrtimer() checks if timer->start_pid isn't initialized in order to find out if timer_stats_update_stats() should be called. However start_pid is initialized with -1 instead of 0, so that the function call always happens. Check timer->start_site like in timer_stats_account_timer() to fix this. Signed-off-by: Heiko Carstens Signed-off-by: Thomas Gleixner --- include/linux/hrtimer.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/hrtimer.h b/include/linux/hrtimer.h index 54648e625efd..4759917adc71 100644 --- a/include/linux/hrtimer.h +++ b/include/linux/hrtimer.h @@ -448,7 +448,7 @@ extern void timer_stats_update_stats(void *timer, pid_t pid, void *startf, static inline void timer_stats_account_hrtimer(struct hrtimer *timer) { - if (likely(!timer->start_pid)) + if (likely(!timer->start_site)) return; timer_stats_update_stats(timer, timer->start_pid, timer->start_site, timer->function, timer->start_comm, 0); -- cgit v1.2.3 From 8aa7e847d834ed937a9ad37a0f2ad5b8584c1ab0 Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Thu, 9 Jul 2009 14:52:32 +0200 Subject: Fix congestion_wait() sync/async vs read/write confusion Commit 1faa16d22877f4839bd433547d770c676d1d964c accidentally broke the bdi congestion wait queue logic, causing us to wait on congestion for WRITE (== 1) when we really wanted BLK_RW_ASYNC (== 0) instead. Signed-off-by: Jens Axboe --- arch/x86/lib/usercopy_32.c | 2 +- drivers/block/pktcdvd.c | 10 ++++++---- drivers/md/dm-crypt.c | 2 +- fs/fat/file.c | 2 +- fs/fuse/dev.c | 8 ++++---- fs/nfs/write.c | 8 +++++--- fs/reiserfs/journal.c | 2 +- fs/xfs/linux-2.6/kmem.c | 4 ++-- fs/xfs/linux-2.6/xfs_buf.c | 2 +- include/linux/backing-dev.h | 6 +++--- include/linux/blkdev.h | 8 ++++---- mm/backing-dev.c | 7 +++---- mm/memcontrol.c | 2 +- mm/page-writeback.c | 8 ++++---- mm/page_alloc.c | 4 ++-- mm/vmscan.c | 8 ++++---- 16 files changed, 43 insertions(+), 40 deletions(-) (limited to 'include') diff --git a/arch/x86/lib/usercopy_32.c b/arch/x86/lib/usercopy_32.c index 7c8ca91bb9ec..1f118d462acc 100644 --- a/arch/x86/lib/usercopy_32.c +++ b/arch/x86/lib/usercopy_32.c @@ -751,7 +751,7 @@ survive: if (retval == -ENOMEM && is_global_init(current)) { up_read(¤t->mm->mmap_sem); - congestion_wait(WRITE, HZ/50); + congestion_wait(BLK_RW_ASYNC, HZ/50); goto survive; } diff --git a/drivers/block/pktcdvd.c b/drivers/block/pktcdvd.c index 83650e00632d..99a506f619b7 100644 --- a/drivers/block/pktcdvd.c +++ b/drivers/block/pktcdvd.c @@ -1372,8 +1372,10 @@ try_next_bio: wakeup = (pd->write_congestion_on > 0 && pd->bio_queue_size <= pd->write_congestion_off); spin_unlock(&pd->lock); - if (wakeup) - clear_bdi_congested(&pd->disk->queue->backing_dev_info, WRITE); + if (wakeup) { + clear_bdi_congested(&pd->disk->queue->backing_dev_info, + BLK_RW_ASYNC); + } pkt->sleep_time = max(PACKET_WAIT_TIME, 1); pkt_set_state(pkt, PACKET_WAITING_STATE); @@ -2592,10 +2594,10 @@ static int pkt_make_request(struct request_queue *q, struct bio *bio) spin_lock(&pd->lock); if (pd->write_congestion_on > 0 && pd->bio_queue_size >= pd->write_congestion_on) { - set_bdi_congested(&q->backing_dev_info, WRITE); + set_bdi_congested(&q->backing_dev_info, BLK_RW_ASYNC); do { spin_unlock(&pd->lock); - congestion_wait(WRITE, HZ); + congestion_wait(BLK_RW_ASYNC, HZ); spin_lock(&pd->lock); } while(pd->bio_queue_size > pd->write_congestion_off); } diff --git a/drivers/md/dm-crypt.c b/drivers/md/dm-crypt.c index 9933eb861c71..529e2ba505c3 100644 --- a/drivers/md/dm-crypt.c +++ b/drivers/md/dm-crypt.c @@ -776,7 +776,7 @@ static void kcryptd_crypt_write_convert(struct dm_crypt_io *io) * But don't wait if split was due to the io size restriction */ if (unlikely(out_of_pages)) - congestion_wait(WRITE, HZ/100); + congestion_wait(BLK_RW_ASYNC, HZ/100); /* * With async crypto it is unsafe to share the crypto context diff --git a/fs/fat/file.c b/fs/fat/file.c index b28ea646ff60..f042b965c95c 100644 --- a/fs/fat/file.c +++ b/fs/fat/file.c @@ -134,7 +134,7 @@ static int fat_file_release(struct inode *inode, struct file *filp) if ((filp->f_mode & FMODE_WRITE) && MSDOS_SB(inode->i_sb)->options.flush) { fat_flush_inodes(inode->i_sb, inode, NULL); - congestion_wait(WRITE, HZ/10); + congestion_wait(BLK_RW_ASYNC, HZ/10); } return 0; } diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index f58ecbc416c8..6484eb75acd6 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -286,8 +286,8 @@ __releases(&fc->lock) } if (fc->num_background == FUSE_CONGESTION_THRESHOLD && fc->connected && fc->bdi_initialized) { - clear_bdi_congested(&fc->bdi, READ); - clear_bdi_congested(&fc->bdi, WRITE); + clear_bdi_congested(&fc->bdi, BLK_RW_SYNC); + clear_bdi_congested(&fc->bdi, BLK_RW_ASYNC); } fc->num_background--; fc->active_background--; @@ -414,8 +414,8 @@ static void fuse_request_send_nowait_locked(struct fuse_conn *fc, fc->blocked = 1; if (fc->num_background == FUSE_CONGESTION_THRESHOLD && fc->bdi_initialized) { - set_bdi_congested(&fc->bdi, READ); - set_bdi_congested(&fc->bdi, WRITE); + set_bdi_congested(&fc->bdi, BLK_RW_SYNC); + set_bdi_congested(&fc->bdi, BLK_RW_ASYNC); } list_add_tail(&req->list, &fc->bg_queue); flush_bg_queue(fc); diff --git a/fs/nfs/write.c b/fs/nfs/write.c index ce728829f79a..0a0a2ff767c3 100644 --- a/fs/nfs/write.c +++ b/fs/nfs/write.c @@ -202,8 +202,10 @@ static int nfs_set_page_writeback(struct page *page) struct nfs_server *nfss = NFS_SERVER(inode); if (atomic_long_inc_return(&nfss->writeback) > - NFS_CONGESTION_ON_THRESH) - set_bdi_congested(&nfss->backing_dev_info, WRITE); + NFS_CONGESTION_ON_THRESH) { + set_bdi_congested(&nfss->backing_dev_info, + BLK_RW_ASYNC); + } } return ret; } @@ -215,7 +217,7 @@ static void nfs_end_page_writeback(struct page *page) end_page_writeback(page); if (atomic_long_dec_return(&nfss->writeback) < NFS_CONGESTION_OFF_THRESH) - clear_bdi_congested(&nfss->backing_dev_info, WRITE); + clear_bdi_congested(&nfss->backing_dev_info, BLK_RW_ASYNC); } /* diff --git a/fs/reiserfs/journal.c b/fs/reiserfs/journal.c index 77f5bb746bf0..90622200b39c 100644 --- a/fs/reiserfs/journal.c +++ b/fs/reiserfs/journal.c @@ -997,7 +997,7 @@ static int reiserfs_async_progress_wait(struct super_block *s) DEFINE_WAIT(wait); struct reiserfs_journal *j = SB_JOURNAL(s); if (atomic_read(&j->j_async_throttle)) - congestion_wait(WRITE, HZ / 10); + congestion_wait(BLK_RW_ASYNC, HZ / 10); return 0; } diff --git a/fs/xfs/linux-2.6/kmem.c b/fs/xfs/linux-2.6/kmem.c index 1cd3b55ee3d2..2d3f90afe5f1 100644 --- a/fs/xfs/linux-2.6/kmem.c +++ b/fs/xfs/linux-2.6/kmem.c @@ -53,7 +53,7 @@ kmem_alloc(size_t size, unsigned int __nocast flags) printk(KERN_ERR "XFS: possible memory allocation " "deadlock in %s (mode:0x%x)\n", __func__, lflags); - congestion_wait(WRITE, HZ/50); + congestion_wait(BLK_RW_ASYNC, HZ/50); } while (1); } @@ -130,7 +130,7 @@ kmem_zone_alloc(kmem_zone_t *zone, unsigned int __nocast flags) printk(KERN_ERR "XFS: possible memory allocation " "deadlock in %s (mode:0x%x)\n", __func__, lflags); - congestion_wait(WRITE, HZ/50); + congestion_wait(BLK_RW_ASYNC, HZ/50); } while (1); } diff --git a/fs/xfs/linux-2.6/xfs_buf.c b/fs/xfs/linux-2.6/xfs_buf.c index 1418b916fc27..0c93c7ef3d18 100644 --- a/fs/xfs/linux-2.6/xfs_buf.c +++ b/fs/xfs/linux-2.6/xfs_buf.c @@ -412,7 +412,7 @@ _xfs_buf_lookup_pages( XFS_STATS_INC(xb_page_retries); xfsbufd_wakeup(0, gfp_mask); - congestion_wait(WRITE, HZ/50); + congestion_wait(BLK_RW_ASYNC, HZ/50); goto retry; } diff --git a/include/linux/backing-dev.h b/include/linux/backing-dev.h index 0ec2c594868e..3a52a63c1351 100644 --- a/include/linux/backing-dev.h +++ b/include/linux/backing-dev.h @@ -229,9 +229,9 @@ static inline int bdi_rw_congested(struct backing_dev_info *bdi) (1 << BDI_async_congested)); } -void clear_bdi_congested(struct backing_dev_info *bdi, int rw); -void set_bdi_congested(struct backing_dev_info *bdi, int rw); -long congestion_wait(int rw, long timeout); +void clear_bdi_congested(struct backing_dev_info *bdi, int sync); +void set_bdi_congested(struct backing_dev_info *bdi, int sync); +long congestion_wait(int sync, long timeout); static inline bool bdi_cap_writeback_dirty(struct backing_dev_info *bdi) diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index 49ae07951d55..bb3d39978701 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -779,18 +779,18 @@ extern int sg_scsi_ioctl(struct request_queue *, struct gendisk *, fmode_t, * congested queues, and wake up anyone who was waiting for requests to be * put back. */ -static inline void blk_clear_queue_congested(struct request_queue *q, int rw) +static inline void blk_clear_queue_congested(struct request_queue *q, int sync) { - clear_bdi_congested(&q->backing_dev_info, rw); + clear_bdi_congested(&q->backing_dev_info, sync); } /* * A queue has just entered congestion. Flag that in the queue's VM-visible * state flags and increment the global gounter of congested queues. */ -static inline void blk_set_queue_congested(struct request_queue *q, int rw) +static inline void blk_set_queue_congested(struct request_queue *q, int sync) { - set_bdi_congested(&q->backing_dev_info, rw); + set_bdi_congested(&q->backing_dev_info, sync); } extern void blk_start_queue(struct request_queue *q); diff --git a/mm/backing-dev.c b/mm/backing-dev.c index 493b468a5035..c86edd244294 100644 --- a/mm/backing-dev.c +++ b/mm/backing-dev.c @@ -283,7 +283,6 @@ static wait_queue_head_t congestion_wqh[2] = { __WAIT_QUEUE_HEAD_INITIALIZER(congestion_wqh[1]) }; - void clear_bdi_congested(struct backing_dev_info *bdi, int sync) { enum bdi_state bit; @@ -308,18 +307,18 @@ EXPORT_SYMBOL(set_bdi_congested); /** * congestion_wait - wait for a backing_dev to become uncongested - * @rw: READ or WRITE + * @sync: SYNC or ASYNC IO * @timeout: timeout in jiffies * * Waits for up to @timeout jiffies for a backing_dev (any backing_dev) to exit * write congestion. If no backing_devs are congested then just wait for the * next write to be completed. */ -long congestion_wait(int rw, long timeout) +long congestion_wait(int sync, long timeout) { long ret; DEFINE_WAIT(wait); - wait_queue_head_t *wqh = &congestion_wqh[rw]; + wait_queue_head_t *wqh = &congestion_wqh[sync]; prepare_to_wait(wqh, &wait, TASK_UNINTERRUPTIBLE); ret = io_schedule_timeout(timeout); diff --git a/mm/memcontrol.c b/mm/memcontrol.c index e2fa20dadf40..e717964cb5a0 100644 --- a/mm/memcontrol.c +++ b/mm/memcontrol.c @@ -1973,7 +1973,7 @@ try_to_free: if (!progress) { nr_retries--; /* maybe some writeback is necessary */ - congestion_wait(WRITE, HZ/10); + congestion_wait(BLK_RW_ASYNC, HZ/10); } } diff --git a/mm/page-writeback.c b/mm/page-writeback.c index 7687879253b9..81627ebcd313 100644 --- a/mm/page-writeback.c +++ b/mm/page-writeback.c @@ -575,7 +575,7 @@ static void balance_dirty_pages(struct address_space *mapping) if (pages_written >= write_chunk) break; /* We've done our duty */ - congestion_wait(WRITE, HZ/10); + congestion_wait(BLK_RW_ASYNC, HZ/10); } if (bdi_nr_reclaimable + bdi_nr_writeback < bdi_thresh && @@ -669,7 +669,7 @@ void throttle_vm_writeout(gfp_t gfp_mask) if (global_page_state(NR_UNSTABLE_NFS) + global_page_state(NR_WRITEBACK) <= dirty_thresh) break; - congestion_wait(WRITE, HZ/10); + congestion_wait(BLK_RW_ASYNC, HZ/10); /* * The caller might hold locks which can prevent IO completion @@ -715,7 +715,7 @@ static void background_writeout(unsigned long _min_pages) if (wbc.nr_to_write > 0 || wbc.pages_skipped > 0) { /* Wrote less than expected */ if (wbc.encountered_congestion || wbc.more_io) - congestion_wait(WRITE, HZ/10); + congestion_wait(BLK_RW_ASYNC, HZ/10); else break; } @@ -787,7 +787,7 @@ static void wb_kupdate(unsigned long arg) writeback_inodes(&wbc); if (wbc.nr_to_write > 0) { if (wbc.encountered_congestion || wbc.more_io) - congestion_wait(WRITE, HZ/10); + congestion_wait(BLK_RW_ASYNC, HZ/10); else break; /* All the old data is written */ } diff --git a/mm/page_alloc.c b/mm/page_alloc.c index ad7cd1c56b07..a35eeab2724c 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -1666,7 +1666,7 @@ __alloc_pages_high_priority(gfp_t gfp_mask, unsigned int order, preferred_zone, migratetype); if (!page && gfp_mask & __GFP_NOFAIL) - congestion_wait(WRITE, HZ/50); + congestion_wait(BLK_RW_ASYNC, HZ/50); } while (!page && (gfp_mask & __GFP_NOFAIL)); return page; @@ -1831,7 +1831,7 @@ rebalance: pages_reclaimed += did_some_progress; if (should_alloc_retry(gfp_mask, order, pages_reclaimed)) { /* Wait for some write requests to complete then retry */ - congestion_wait(WRITE, HZ/50); + congestion_wait(BLK_RW_ASYNC, HZ/50); goto rebalance; } diff --git a/mm/vmscan.c b/mm/vmscan.c index 54155268dfca..dea7abd31098 100644 --- a/mm/vmscan.c +++ b/mm/vmscan.c @@ -1104,7 +1104,7 @@ static unsigned long shrink_inactive_list(unsigned long max_scan, */ if (nr_freed < nr_taken && !current_is_kswapd() && lumpy_reclaim) { - congestion_wait(WRITE, HZ/10); + congestion_wait(BLK_RW_ASYNC, HZ/10); /* * The attempt at page out may have made some @@ -1721,7 +1721,7 @@ static unsigned long do_try_to_free_pages(struct zonelist *zonelist, /* Take a nap, wait for some writeback to complete */ if (sc->nr_scanned && priority < DEF_PRIORITY - 2) - congestion_wait(WRITE, HZ/10); + congestion_wait(BLK_RW_ASYNC, HZ/10); } /* top priority shrink_zones still had more to do? don't OOM, then */ if (!sc->all_unreclaimable && scanning_global_lru(sc)) @@ -1960,7 +1960,7 @@ loop_again: * another pass across the zones. */ if (total_scanned && priority < DEF_PRIORITY - 2) - congestion_wait(WRITE, HZ/10); + congestion_wait(BLK_RW_ASYNC, HZ/10); /* * We do this so kswapd doesn't build up large priorities for @@ -2233,7 +2233,7 @@ unsigned long shrink_all_memory(unsigned long nr_pages) goto out; if (sc.nr_scanned && prio < DEF_PRIORITY - 2) - congestion_wait(WRITE, HZ / 10); + congestion_wait(BLK_RW_ASYNC, HZ / 10); } } -- cgit v1.2.3 From ecb554a846f8e9d2a58f6d6c118168a63ac065aa Mon Sep 17 00:00:00 2001 From: FUJITA Tomonori Date: Thu, 9 Jul 2009 14:46:53 +0200 Subject: block: fix sg SG_DXFER_TO_FROM_DEV regression I overlooked SG_DXFER_TO_FROM_DEV support when I converted sg to use the block layer mapping API (2.6.28). Douglas Gilbert explained SG_DXFER_TO_FROM_DEV: http://www.spinics.net/lists/linux-scsi/msg37135.html = The semantics of SG_DXFER_TO_FROM_DEV were: - copy user space buffer to kernel (LLD) buffer - do SCSI command which is assumed to be of the DATA_IN (data from device) variety. This would overwrite some or all of the kernel buffer - copy kernel (LLD) buffer back to the user space. The idea was to detect short reads by filling the original user space buffer with some marker bytes ("0xec" it would seem in this report). The "resid" value is a better way of detecting short reads but that was only added this century and requires co-operation from the LLD. = This patch changes the block layer mapping API to support this semantics. This simply adds another field to struct rq_map_data and enables __bio_copy_iov() to copy data from user space even with READ requests. It's better to add the flags field and kills null_mapped and the new from_user fields in struct rq_map_data but that approach makes it difficult to send this patch to stable trees because st and osst drivers use struct rq_map_data (they were converted to use the block layer in 2.6.29 and 2.6.30). Well, I should clean up the block layer mapping API. zhou sf reported this regiression and tested this patch: http://www.spinics.net/lists/linux-scsi/msg37128.html http://www.spinics.net/lists/linux-scsi/msg37168.html Reported-by: zhou sf Tested-by: zhou sf Cc: stable@kernel.org Signed-off-by: FUJITA Tomonori Signed-off-by: Jens Axboe --- drivers/scsi/sg.c | 4 ++++ fs/bio.c | 22 ++++++++++++---------- include/linux/blkdev.h | 1 + 3 files changed, 17 insertions(+), 10 deletions(-) (limited to 'include') diff --git a/drivers/scsi/sg.c b/drivers/scsi/sg.c index 4d6f2fe1cfe9..9230402c45af 100644 --- a/drivers/scsi/sg.c +++ b/drivers/scsi/sg.c @@ -1656,6 +1656,10 @@ static int sg_start_req(Sg_request *srp, unsigned char *cmd) md->nr_entries = req_schp->k_use_sg; md->offset = 0; md->null_mapped = hp->dxferp ? 0 : 1; + if (dxfer_dir == SG_DXFER_TO_FROM_DEV) + md->from_user = 1; + else + md->from_user = 0; } if (iov_count) { diff --git a/fs/bio.c b/fs/bio.c index 1486b19fc431..76738005c8e8 100644 --- a/fs/bio.c +++ b/fs/bio.c @@ -705,14 +705,13 @@ static struct bio_map_data *bio_alloc_map_data(int nr_segs, int iov_count, } static int __bio_copy_iov(struct bio *bio, struct bio_vec *iovecs, - struct sg_iovec *iov, int iov_count, int uncopy, - int do_free_page) + struct sg_iovec *iov, int iov_count, + int to_user, int from_user, int do_free_page) { int ret = 0, i; struct bio_vec *bvec; int iov_idx = 0; unsigned int iov_off = 0; - int read = bio_data_dir(bio) == READ; __bio_for_each_segment(bvec, bio, i, 0) { char *bv_addr = page_address(bvec->bv_page); @@ -727,13 +726,14 @@ static int __bio_copy_iov(struct bio *bio, struct bio_vec *iovecs, iov_addr = iov[iov_idx].iov_base + iov_off; if (!ret) { - if (!read && !uncopy) - ret = copy_from_user(bv_addr, iov_addr, - bytes); - if (read && uncopy) + if (to_user) ret = copy_to_user(iov_addr, bv_addr, bytes); + if (from_user) + ret = copy_from_user(bv_addr, iov_addr, + bytes); + if (ret) ret = -EFAULT; } @@ -770,7 +770,8 @@ int bio_uncopy_user(struct bio *bio) if (!bio_flagged(bio, BIO_NULL_MAPPED)) ret = __bio_copy_iov(bio, bmd->iovecs, bmd->sgvecs, - bmd->nr_sgvecs, 1, bmd->is_our_pages); + bmd->nr_sgvecs, bio_data_dir(bio) == READ, + 0, bmd->is_our_pages); bio_free_map_data(bmd); bio_put(bio); return ret; @@ -875,8 +876,9 @@ struct bio *bio_copy_user_iov(struct request_queue *q, /* * success */ - if (!write_to_vm && (!map_data || !map_data->null_mapped)) { - ret = __bio_copy_iov(bio, bio->bi_io_vec, iov, iov_count, 0, 0); + if ((!write_to_vm && (!map_data || !map_data->null_mapped)) || + (map_data && map_data->from_user)) { + ret = __bio_copy_iov(bio, bio->bi_io_vec, iov, iov_count, 0, 1, 0); if (ret) goto cleanup; } diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index bb3d39978701..0146e0fecf1a 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -723,6 +723,7 @@ struct rq_map_data { int nr_entries; unsigned long offset; int null_mapped; + int from_user; }; struct req_iterator { -- cgit v1.2.3 From e36d56b64808aec54b68b4e9976180c1da0933b2 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Tue, 9 Jun 2009 21:04:43 +0200 Subject: cfg80211: pass netdev to change_virtual_intf If there was a reason I'm passing the ifidx I cannot remember it any more and don't see one now, so let's just pass the pointer itself. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- drivers/net/wireless/iwmc3200wifi/cfg80211.c | 9 ++------- drivers/net/wireless/rndis_wlan.c | 15 +++++---------- include/net/cfg80211.h | 3 ++- net/mac80211/cfg.c | 9 ++------- net/wireless/nl80211.c | 12 +++++------- net/wireless/wext-compat.c | 2 +- 6 files changed, 17 insertions(+), 33 deletions(-) (limited to 'include') diff --git a/drivers/net/wireless/iwmc3200wifi/cfg80211.c b/drivers/net/wireless/iwmc3200wifi/cfg80211.c index 96f714e6e12b..bc89b1ef0cad 100644 --- a/drivers/net/wireless/iwmc3200wifi/cfg80211.c +++ b/drivers/net/wireless/iwmc3200wifi/cfg80211.c @@ -167,20 +167,15 @@ int iwm_cfg80211_inform_bss(struct iwm_priv *iwm) return 0; } -static int iwm_cfg80211_change_iface(struct wiphy *wiphy, int ifindex, +static int iwm_cfg80211_change_iface(struct wiphy *wiphy, + struct net_device *ndev, enum nl80211_iftype type, u32 *flags, struct vif_params *params) { - struct net_device *ndev; struct wireless_dev *wdev; struct iwm_priv *iwm; u32 old_mode; - /* we're under RTNL */ - ndev = __dev_get_by_index(&init_net, ifindex); - if (!ndev) - return -ENODEV; - wdev = ndev->ieee80211_ptr; iwm = ndev_to_iwm(ndev); old_mode = iwm->conf.mode; diff --git a/drivers/net/wireless/rndis_wlan.c b/drivers/net/wireless/rndis_wlan.c index 3bec3dbd3450..e8b0793b030a 100644 --- a/drivers/net/wireless/rndis_wlan.c +++ b/drivers/net/wireless/rndis_wlan.c @@ -420,7 +420,8 @@ struct rndis_wlan_private { /* * cfg80211 ops */ -static int rndis_change_virtual_intf(struct wiphy *wiphy, int ifindex, +static int rndis_change_virtual_intf(struct wiphy *wiphy, + struct net_device *dev, enum nl80211_iftype type, u32 *flags, struct vif_params *params); @@ -1222,20 +1223,14 @@ static void set_multicast_list(struct usbnet *usbdev) /* * cfg80211 ops */ -static int rndis_change_virtual_intf(struct wiphy *wiphy, int ifindex, +static int rndis_change_virtual_intf(struct wiphy *wiphy, + struct net_device *dev, enum nl80211_iftype type, u32 *flags, struct vif_params *params) { - struct net_device *dev; - struct usbnet *usbdev; + struct usbnet *usbdev = netdev_priv(dev); int mode; - /* we're under RTNL */ - dev = __dev_get_by_index(&init_net, ifindex); - if (!dev) - return -ENODEV; - usbdev = netdev_priv(dev); - switch (type) { case NL80211_IFTYPE_ADHOC: mode = NDIS_80211_INFRA_ADHOC; diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index 1a21895b732b..90f9bfa3bfc2 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -866,7 +866,8 @@ struct cfg80211_ops { enum nl80211_iftype type, u32 *flags, struct vif_params *params); int (*del_virtual_intf)(struct wiphy *wiphy, int ifindex); - int (*change_virtual_intf)(struct wiphy *wiphy, int ifindex, + int (*change_virtual_intf)(struct wiphy *wiphy, + struct net_device *dev, enum nl80211_iftype type, u32 *flags, struct vif_params *params); diff --git a/net/mac80211/cfg.c b/net/mac80211/cfg.c index 3f47276caeb8..eb93eb6a9cc7 100644 --- a/net/mac80211/cfg.c +++ b/net/mac80211/cfg.c @@ -74,19 +74,14 @@ static int ieee80211_del_iface(struct wiphy *wiphy, int ifindex) return 0; } -static int ieee80211_change_iface(struct wiphy *wiphy, int ifindex, +static int ieee80211_change_iface(struct wiphy *wiphy, + struct net_device *dev, enum nl80211_iftype type, u32 *flags, struct vif_params *params) { - struct net_device *dev; struct ieee80211_sub_if_data *sdata; int ret; - /* we're under RTNL */ - dev = __dev_get_by_index(&init_net, ifindex); - if (!dev) - return -ENODEV; - if (!nl80211_type_check(type)) return -EINVAL; diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index 43bdb1372cae..f91e5d472c60 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -767,7 +767,7 @@ static int nl80211_set_interface(struct sk_buff *skb, struct genl_info *info) { struct cfg80211_registered_device *drv; struct vif_params params; - int err, ifindex; + int err; enum nl80211_iftype otype, ntype; struct net_device *dev; u32 _flags, *flags = NULL; @@ -781,9 +781,7 @@ static int nl80211_set_interface(struct sk_buff *skb, struct genl_info *info) if (err) goto unlock_rtnl; - ifindex = dev->ifindex; otype = ntype = dev->ieee80211_ptr->iftype; - dev_put(dev); if (info->attrs[NL80211_ATTR_IFTYPE]) { ntype = nla_get_u32(info->attrs[NL80211_ATTR_IFTYPE]); @@ -826,20 +824,20 @@ static int nl80211_set_interface(struct sk_buff *skb, struct genl_info *info) } if (change) - err = drv->ops->change_virtual_intf(&drv->wiphy, ifindex, + err = drv->ops->change_virtual_intf(&drv->wiphy, dev, ntype, flags, ¶ms); else err = 0; - dev = __dev_get_by_index(&init_net, ifindex); - WARN_ON(!dev || (!err && dev->ieee80211_ptr->iftype != ntype)); + WARN_ON(!err && dev->ieee80211_ptr->iftype != ntype); - if (dev && !err && (ntype != otype)) { + if (!err && (ntype != otype)) { if (otype == NL80211_IFTYPE_ADHOC) cfg80211_clear_ibss(dev, false); } unlock: + dev_put(dev); cfg80211_put_dev(drv); unlock_rtnl: rtnl_unlock(); diff --git a/net/wireless/wext-compat.c b/net/wireless/wext-compat.c index d030c5315672..9e56f3569e3e 100644 --- a/net/wireless/wext-compat.c +++ b/net/wireless/wext-compat.c @@ -103,7 +103,7 @@ int cfg80211_wext_siwmode(struct net_device *dev, struct iw_request_info *info, memset(&vifparams, 0, sizeof(vifparams)); - ret = rdev->ops->change_virtual_intf(wdev->wiphy, dev->ifindex, type, + ret = rdev->ops->change_virtual_intf(wdev->wiphy, dev, type, NULL, &vifparams); WARN_ON(!ret && wdev->iftype != type); -- cgit v1.2.3 From a33e9e7f35ef6dcab528e0327f29188475f60691 Mon Sep 17 00:00:00 2001 From: Jussi Kivilinna Date: Tue, 16 Jun 2009 17:17:27 +0300 Subject: usbnet: Add stop function pointer to 'struct rndis_data'. Allow minidriver to know that netdev has stopped. This is to let wireless turn off radio when usbnet dev is stopped. Signed-off-by: Jussi Kivilinna Acked-by: David Brownell Acked-by: David S. Miller Signed-off-by: John W. Linville --- drivers/net/usb/usbnet.c | 14 ++++++++++++++ include/linux/usb/usbnet.h | 3 +++ 2 files changed, 17 insertions(+) (limited to 'include') diff --git a/drivers/net/usb/usbnet.c b/drivers/net/usb/usbnet.c index edfd9e10ceba..25e435c49040 100644 --- a/drivers/net/usb/usbnet.c +++ b/drivers/net/usb/usbnet.c @@ -575,7 +575,9 @@ EXPORT_SYMBOL_GPL(usbnet_unlink_rx_urbs); int usbnet_stop (struct net_device *net) { struct usbnet *dev = netdev_priv(net); + struct driver_info *info = dev->driver_info; int temp; + int retval; DECLARE_WAIT_QUEUE_HEAD_ONSTACK (unlink_wakeup); DECLARE_WAITQUEUE (wait, current); @@ -587,6 +589,18 @@ int usbnet_stop (struct net_device *net) net->stats.rx_errors, net->stats.tx_errors ); + /* allow minidriver to stop correctly (wireless devices to turn off + * radio etc) */ + if (info->stop) { + retval = info->stop(dev); + if (retval < 0 && netif_msg_ifdown(dev)) + devinfo(dev, + "stop fail (%d) usbnet usb-%s-%s, %s", + retval, + dev->udev->bus->bus_name, dev->udev->devpath, + info->description); + } + // ensure there are no more active urbs add_wait_queue (&unlink_wakeup, &wait); dev->wait = &unlink_wakeup; diff --git a/include/linux/usb/usbnet.h b/include/linux/usb/usbnet.h index 310e18a880ff..7c17b2efba86 100644 --- a/include/linux/usb/usbnet.h +++ b/include/linux/usb/usbnet.h @@ -97,6 +97,9 @@ struct driver_info { /* reset device ... can sleep */ int (*reset)(struct usbnet *); + /* stop device ... can sleep */ + int (*stop)(struct usbnet *); + /* see if peer is connected ... can sleep */ int (*check_connect)(struct usbnet *); -- cgit v1.2.3 From f1d58c2521eb160178b2151d6326d8dc5d7c8560 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Wed, 17 Jun 2009 13:13:00 +0200 Subject: mac80211: push rx status into skb->cb Within mac80211, we often need to copy the rx status into skb->cb. This is wasteful, as drivers could be building it in there to start with. This patch changes the API so that drivers are expected to pass the RX status in skb->cb, now accessible as IEEE80211_SKB_RXCB(skb). It also updates all drivers to pass the rx status in there, but only by making them memcpy() it into place before the call to the receive function (ieee80211_rx(_irqsafe)). Each driver can now be optimised on its own schedule. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- drivers/net/wireless/adm8211.c | 3 +- drivers/net/wireless/at76c50x-usb.c | 3 +- drivers/net/wireless/ath/ar9170/main.c | 6 ++- drivers/net/wireless/ath/ath5k/base.c | 3 +- drivers/net/wireless/ath/ath9k/recv.c | 13 ++++-- drivers/net/wireless/b43/xmit.c | 3 +- drivers/net/wireless/b43legacy/xmit.c | 3 +- drivers/net/wireless/iwlwifi/iwl-3945.c | 3 +- drivers/net/wireless/iwlwifi/iwl-rx.c | 3 +- drivers/net/wireless/libertas_tf/main.c | 3 +- drivers/net/wireless/mac80211_hwsim.c | 3 +- drivers/net/wireless/mwl8k.c | 3 +- drivers/net/wireless/p54/p54common.c | 3 +- drivers/net/wireless/rt2x00/rt2x00dev.c | 3 +- drivers/net/wireless/rtl818x/rtl8180_dev.c | 3 +- drivers/net/wireless/rtl818x/rtl8187_dev.c | 3 +- drivers/net/wireless/wl12xx/wl1251_rx.c | 3 +- drivers/net/wireless/zd1211rw/zd_mac.c | 3 +- drivers/staging/agnx/xmit.c | 3 +- drivers/staging/stlc45xx/stlc45xx.c | 3 +- drivers/staging/winbond/wb35rx.c | 3 +- include/net/mac80211.h | 29 ++++++------ net/mac80211/ibss.c | 6 +-- net/mac80211/ieee80211_i.h | 10 ++--- net/mac80211/main.c | 5 +-- net/mac80211/mesh.c | 6 +-- net/mac80211/mesh.h | 3 +- net/mac80211/mlme.c | 4 +- net/mac80211/rx.c | 71 +++++++++++++----------------- net/mac80211/scan.c | 4 +- 30 files changed, 108 insertions(+), 106 deletions(-) (limited to 'include') diff --git a/drivers/net/wireless/adm8211.c b/drivers/net/wireless/adm8211.c index 2b9e379994a1..ecc93834533f 100644 --- a/drivers/net/wireless/adm8211.c +++ b/drivers/net/wireless/adm8211.c @@ -452,7 +452,8 @@ static void adm8211_interrupt_rci(struct ieee80211_hw *dev) rx_status.freq = adm8211_channels[priv->channel - 1].center_freq; rx_status.band = IEEE80211_BAND_2GHZ; - ieee80211_rx_irqsafe(dev, skb, &rx_status); + memcpy(IEEE80211_SKB_RXCB(skb), &rx_status, sizeof(rx_status)); + ieee80211_rx_irqsafe(dev, skb); } entry = (++priv->cur_rx) % priv->rx_ring_size; diff --git a/drivers/net/wireless/at76c50x-usb.c b/drivers/net/wireless/at76c50x-usb.c index 4efbdbe6d6bf..13303fa34734 100644 --- a/drivers/net/wireless/at76c50x-usb.c +++ b/drivers/net/wireless/at76c50x-usb.c @@ -1568,7 +1568,8 @@ static void at76_rx_tasklet(unsigned long param) at76_dbg(DBG_MAC80211, "calling ieee80211_rx_irqsafe(): %d/%d", priv->rx_skb->len, priv->rx_skb->data_len); - ieee80211_rx_irqsafe(priv->hw, priv->rx_skb, &rx_status); + memcpy(IEEE80211_SKB_RXCB(priv->rx_skb), &rx_status, sizeof(rx_status)); + ieee80211_rx_irqsafe(priv->hw, priv->rx_skb); /* Use a new skb for the next receive */ priv->rx_skb = NULL; diff --git a/drivers/net/wireless/ath/ar9170/main.c b/drivers/net/wireless/ath/ar9170/main.c index 9d38cf60a0db..51753ed1b8ba 100644 --- a/drivers/net/wireless/ath/ar9170/main.c +++ b/drivers/net/wireless/ath/ar9170/main.c @@ -917,8 +917,10 @@ static void ar9170_handle_mpdu(struct ar9170 *ar, u8 *buf, int len) ar9170_rx_phy_status(ar, phy, &status); skb = ar9170_rx_copy_data(buf, mpdu_len); - if (likely(skb)) - ieee80211_rx_irqsafe(ar->hw, skb, &status); + if (likely(skb)) { + memcpy(IEEE80211_SKB_RXCB(skb), &status, sizeof(status)); + ieee80211_rx_irqsafe(ar->hw, skb); + } } void ar9170_rx(struct ar9170 *ar, struct sk_buff *skb) diff --git a/drivers/net/wireless/ath/ath5k/base.c b/drivers/net/wireless/ath/ath5k/base.c index f26a68960622..c6e70919435c 100644 --- a/drivers/net/wireless/ath/ath5k/base.c +++ b/drivers/net/wireless/ath/ath5k/base.c @@ -1905,7 +1905,8 @@ accept: if (sc->opmode == NL80211_IFTYPE_ADHOC) ath5k_check_ibss_tsf(sc, skb, &rxs); - __ieee80211_rx(sc->hw, skb, &rxs); + memcpy(IEEE80211_SKB_RXCB(skb), &rxs, sizeof(rxs)); + ieee80211_rx(sc->hw, skb); bf->skb = next_skb; bf->skbaddr = next_skb_addr; diff --git a/drivers/net/wireless/ath/ath9k/recv.c b/drivers/net/wireless/ath/ath9k/recv.c index cece1c4c6bda..c00b9051bb53 100644 --- a/drivers/net/wireless/ath/ath9k/recv.c +++ b/drivers/net/wireless/ath/ath9k/recv.c @@ -619,13 +619,18 @@ static void ath_rx_send_to_mac80211(struct ath_softc *sc, struct sk_buff *skb, if (aphy == NULL) continue; nskb = skb_copy(skb, GFP_ATOMIC); - if (nskb) - __ieee80211_rx(aphy->hw, nskb, rx_status); + if (nskb) { + memcpy(IEEE80211_SKB_RXCB(nskb), rx_status, + sizeof(*rx_status)); + ieee80211_rx(aphy->hw, nskb); + } } - __ieee80211_rx(sc->hw, skb, rx_status); + memcpy(IEEE80211_SKB_RXCB(skb), rx_status, sizeof(*rx_status)); + ieee80211_rx(sc->hw, skb); } else { /* Deliver unicast frames based on receiver address */ - __ieee80211_rx(ath_get_virt_hw(sc, hdr), skb, rx_status); + memcpy(IEEE80211_SKB_RXCB(skb), rx_status, sizeof(*rx_status)); + ieee80211_rx(ath_get_virt_hw(sc, hdr), skb); } } diff --git a/drivers/net/wireless/b43/xmit.c b/drivers/net/wireless/b43/xmit.c index 55f36a7254d9..5b85e7d73592 100644 --- a/drivers/net/wireless/b43/xmit.c +++ b/drivers/net/wireless/b43/xmit.c @@ -670,7 +670,8 @@ void b43_rx(struct b43_wldev *dev, struct sk_buff *skb, const void *_rxhdr) goto drop; } - ieee80211_rx_irqsafe(dev->wl->hw, skb, &status); + memcpy(IEEE80211_SKB_RXCB(skb), &status, sizeof(status)); + ieee80211_rx_irqsafe(dev->wl->hw, skb); return; drop: diff --git a/drivers/net/wireless/b43legacy/xmit.c b/drivers/net/wireless/b43legacy/xmit.c index b8e39dd06e99..f79cee82601b 100644 --- a/drivers/net/wireless/b43legacy/xmit.c +++ b/drivers/net/wireless/b43legacy/xmit.c @@ -591,7 +591,8 @@ void b43legacy_rx(struct b43legacy_wldev *dev, } dev->stats.last_rx = jiffies; - ieee80211_rx_irqsafe(dev->wl->hw, skb, &status); + memcpy(IEEE80211_SKB_RXCB(skb), &status, sizeof(status)); + ieee80211_rx_irqsafe(dev->wl->hw, skb); return; drop: diff --git a/drivers/net/wireless/iwlwifi/iwl-3945.c b/drivers/net/wireless/iwlwifi/iwl-3945.c index 46288e724889..777c09534cec 100644 --- a/drivers/net/wireless/iwlwifi/iwl-3945.c +++ b/drivers/net/wireless/iwlwifi/iwl-3945.c @@ -577,7 +577,8 @@ static void iwl3945_pass_packet_to_mac80211(struct iwl_priv *priv, if (ieee80211_is_data(hdr->frame_control)) priv->rxtxpackets += len; #endif - ieee80211_rx_irqsafe(priv->hw, rxb->skb, stats); + memcpy(IEEE80211_SKB_RXCB(rxb->skb), stats, sizeof(*stats)); + ieee80211_rx_irqsafe(priv->hw, rxb->skb); rxb->skb = NULL; } diff --git a/drivers/net/wireless/iwlwifi/iwl-rx.c b/drivers/net/wireless/iwlwifi/iwl-rx.c index 2b8d40b37a1c..2160795ed015 100644 --- a/drivers/net/wireless/iwlwifi/iwl-rx.c +++ b/drivers/net/wireless/iwlwifi/iwl-rx.c @@ -932,7 +932,8 @@ static void iwl_pass_packet_to_mac80211(struct iwl_priv *priv, return; iwl_update_rx_stats(priv, le16_to_cpu(hdr->frame_control), len); - ieee80211_rx_irqsafe(priv->hw, rxb->skb, stats); + memcpy(IEEE80211_SKB_RXCB(rxb->skb), stats, sizeof(*stats)); + ieee80211_rx_irqsafe(priv->hw, rxb->skb); priv->alloc_rxb_skb--; rxb->skb = NULL; } diff --git a/drivers/net/wireless/libertas_tf/main.c b/drivers/net/wireless/libertas_tf/main.c index 10a99e26d392..4872345a2f61 100644 --- a/drivers/net/wireless/libertas_tf/main.c +++ b/drivers/net/wireless/libertas_tf/main.c @@ -503,7 +503,8 @@ int lbtf_rx(struct lbtf_private *priv, struct sk_buff *skb) skb_reserve(skb, 2); } - ieee80211_rx_irqsafe(priv->hw, skb, &stats); + memcpy(IEEE80211_SKB_RXCB(skb), &stats, sizeof(stats)); + ieee80211_rx_irqsafe(priv->hw, skb); return 0; } EXPORT_SYMBOL_GPL(lbtf_rx); diff --git a/drivers/net/wireless/mac80211_hwsim.c b/drivers/net/wireless/mac80211_hwsim.c index c47ef48f31c5..b1e4baec29f4 100644 --- a/drivers/net/wireless/mac80211_hwsim.c +++ b/drivers/net/wireless/mac80211_hwsim.c @@ -430,7 +430,8 @@ static bool mac80211_hwsim_tx_frame(struct ieee80211_hw *hw, if (memcmp(hdr->addr1, data2->hw->wiphy->perm_addr, ETH_ALEN) == 0) ack = true; - ieee80211_rx_irqsafe(data2->hw, nskb, &rx_status); + memcpy(IEEE80211_SKB_RXCB(nskb), &rx_status, sizeof(rx_status)); + ieee80211_rx_irqsafe(data2->hw, nskb); } spin_unlock(&hwsim_radio_lock); diff --git a/drivers/net/wireless/mwl8k.c b/drivers/net/wireless/mwl8k.c index a263d5c84c08..b9eded88c322 100644 --- a/drivers/net/wireless/mwl8k.c +++ b/drivers/net/wireless/mwl8k.c @@ -1047,7 +1047,8 @@ static int rxq_process(struct ieee80211_hw *hw, int index, int limit) status.flag = 0; status.band = IEEE80211_BAND_2GHZ; status.freq = ieee80211_channel_to_frequency(rx_desc->channel); - ieee80211_rx_irqsafe(hw, skb, &status); + memcpy(IEEE80211_SKB_RXCB(skb), &status, sizeof(status)); + ieee80211_rx_irqsafe(hw, skb); processed++; } diff --git a/drivers/net/wireless/p54/p54common.c b/drivers/net/wireless/p54/p54common.c index 22ca122bd798..1b15f9e0b861 100644 --- a/drivers/net/wireless/p54/p54common.c +++ b/drivers/net/wireless/p54/p54common.c @@ -794,7 +794,8 @@ static int p54_rx_data(struct ieee80211_hw *dev, struct sk_buff *skb) skb_pull(skb, header_len); skb_trim(skb, le16_to_cpu(hdr->len)); - ieee80211_rx_irqsafe(dev, skb, &rx_status); + memcpy(IEEE80211_SKB_RXCB(skb), &rx_status, sizeof(rx_status)); + ieee80211_rx_irqsafe(dev, skb); queue_delayed_work(dev->workqueue, &priv->work, msecs_to_jiffies(P54_STATISTICS_UPDATE)); diff --git a/drivers/net/wireless/rt2x00/rt2x00dev.c b/drivers/net/wireless/rt2x00/rt2x00dev.c index 57813e72c808..41e33798adb5 100644 --- a/drivers/net/wireless/rt2x00/rt2x00dev.c +++ b/drivers/net/wireless/rt2x00/rt2x00dev.c @@ -449,7 +449,8 @@ void rt2x00lib_rxdone(struct rt2x00_dev *rt2x00dev, * mac80211 will clean up the skb structure. */ rt2x00debug_dump_frame(rt2x00dev, DUMP_FRAME_RXDONE, entry->skb); - ieee80211_rx_irqsafe(rt2x00dev->hw, entry->skb, rx_status); + memcpy(IEEE80211_SKB_RXCB(entry->skb), rx_status, sizeof(*rx_status)); + ieee80211_rx_irqsafe(rt2x00dev->hw, entry->skb); /* * Replace the skb with the freshly allocated one. diff --git a/drivers/net/wireless/rtl818x/rtl8180_dev.c b/drivers/net/wireless/rtl818x/rtl8180_dev.c index 7e65d7c31802..47521c5b6f91 100644 --- a/drivers/net/wireless/rtl818x/rtl8180_dev.c +++ b/drivers/net/wireless/rtl818x/rtl8180_dev.c @@ -143,7 +143,8 @@ static void rtl8180_handle_rx(struct ieee80211_hw *dev) if (flags & RTL818X_RX_DESC_FLAG_CRC32_ERR) rx_status.flag |= RX_FLAG_FAILED_FCS_CRC; - ieee80211_rx_irqsafe(dev, skb, &rx_status); + memcpy(IEEE80211_SKB_RXCB(skb), &rx_status, sizeof(rx_status)); + ieee80211_rx_irqsafe(dev, skb); skb = new_skb; priv->rx_buf[priv->rx_idx] = skb; diff --git a/drivers/net/wireless/rtl818x/rtl8187_dev.c b/drivers/net/wireless/rtl818x/rtl8187_dev.c index 294250e294dd..c9b9dbe584c6 100644 --- a/drivers/net/wireless/rtl818x/rtl8187_dev.c +++ b/drivers/net/wireless/rtl818x/rtl8187_dev.c @@ -380,7 +380,8 @@ static void rtl8187_rx_cb(struct urb *urb) rx_status.flag |= RX_FLAG_TSFT; if (flags & RTL818X_RX_DESC_FLAG_CRC32_ERR) rx_status.flag |= RX_FLAG_FAILED_FCS_CRC; - ieee80211_rx_irqsafe(dev, skb, &rx_status); + memcpy(IEEE80211_SKB_RXCB(skb), &rx_status, sizeof(rx_status)); + ieee80211_rx_irqsafe(dev, skb); skb = dev_alloc_skb(RTL8187_MAX_RX); if (unlikely(!skb)) { diff --git a/drivers/net/wireless/wl12xx/wl1251_rx.c b/drivers/net/wireless/wl12xx/wl1251_rx.c index 48fa39ea17ed..0dbb483a0973 100644 --- a/drivers/net/wireless/wl12xx/wl1251_rx.c +++ b/drivers/net/wireless/wl12xx/wl1251_rx.c @@ -151,7 +151,8 @@ static void wl1251_rx_body(struct wl1251 *wl, wl1251_debug(DEBUG_RX, "rx skb 0x%p: %d B %s", skb, skb->len, beacon ? "beacon" : ""); - ieee80211_rx(wl->hw, skb, &status); + memcpy(IEEE80211_SKB_RXCB(skb), &status, sizeof(status)); + ieee80211_rx(wl->hw, skb); } static void wl1251_rx_ack(struct wl1251 *wl) diff --git a/drivers/net/wireless/zd1211rw/zd_mac.c b/drivers/net/wireless/zd1211rw/zd_mac.c index 40b07b988224..9600b72495da 100644 --- a/drivers/net/wireless/zd1211rw/zd_mac.c +++ b/drivers/net/wireless/zd1211rw/zd_mac.c @@ -711,7 +711,8 @@ int zd_mac_rx(struct ieee80211_hw *hw, const u8 *buffer, unsigned int length) memcpy(skb_put(skb, length), buffer, length); - ieee80211_rx_irqsafe(hw, skb, &stats); + memcpy(IEEE80211_SKB_RXCB(skb), &stats, sizeof(stats)); + ieee80211_rx_irqsafe(hw, skb); return 0; } diff --git a/drivers/staging/agnx/xmit.c b/drivers/staging/agnx/xmit.c index 0e034081f3a5..42db41070cf0 100644 --- a/drivers/staging/agnx/xmit.c +++ b/drivers/staging/agnx/xmit.c @@ -384,7 +384,8 @@ void handle_rx_irq(struct agnx_priv *priv) /* dump_ieee80211_hdr((struct ieee80211_hdr *)skb->data, "RX G"); */ } else agnx_bug("Unknown packets type"); - ieee80211_rx_irqsafe(priv->hw, skb, &status); + memcpy(IEEE80211_SKB_RXCB(skb), &status, sizeof(status)); + ieee80211_rx_irqsafe(priv->hw, skb); rx_desc_reinit(priv, i); } while (priv->rx.idx++); diff --git a/drivers/staging/stlc45xx/stlc45xx.c b/drivers/staging/stlc45xx/stlc45xx.c index cfdaac9b747e..52744faedbff 100644 --- a/drivers/staging/stlc45xx/stlc45xx.c +++ b/drivers/staging/stlc45xx/stlc45xx.c @@ -1429,7 +1429,8 @@ static int stlc45xx_rx_data(struct stlc45xx *stlc, struct sk_buff *skb) stlc45xx_debug(DEBUG_RX, "rx data 0x%p %d B", skb->data, skb->len); stlc45xx_dump(DEBUG_RX_CONTENT, skb->data, skb->len); - ieee80211_rx(stlc->hw, skb, &status); + memcpy(IEEE80211_SKB_RXCB(skb), &status, sizeof(status)); + ieee80211_rx(stlc->hw, skb); return 0; } diff --git a/drivers/staging/winbond/wb35rx.c b/drivers/staging/winbond/wb35rx.c index 3e8cf08b87e6..b905e7b43a19 100644 --- a/drivers/staging/winbond/wb35rx.c +++ b/drivers/staging/winbond/wb35rx.c @@ -40,7 +40,8 @@ static void packet_came(struct ieee80211_hw *hw, char *pRxBufferAddress, int Pac rx_status.phymode = MODE_IEEE80211B; */ - ieee80211_rx_irqsafe(hw, skb, &rx_status); + memcpy(IEEE80211_SKB_RXCB(skb), &rx_status, sizeof(rx_status)); + ieee80211_rx_irqsafe(hw, skb); } static void Wb35Rx_adjust(PDESCRIPTOR pRxDes) diff --git a/include/net/mac80211.h b/include/net/mac80211.h index c06104476973..fe80771d95f1 100644 --- a/include/net/mac80211.h +++ b/include/net/mac80211.h @@ -397,6 +397,11 @@ static inline struct ieee80211_tx_info *IEEE80211_SKB_CB(struct sk_buff *skb) return (struct ieee80211_tx_info *)skb->cb; } +static inline struct ieee80211_rx_status *IEEE80211_SKB_RXCB(struct sk_buff *skb) +{ + return (struct ieee80211_rx_status *)skb->cb; +} + /** * ieee80211_tx_info_clear_status - clear TX status * @@ -478,7 +483,7 @@ enum mac80211_rx_flags { * * The low-level driver should provide this information (the subset * supported by hardware) to the 802.11 code with each received - * frame. + * frame, in the skb's control buffer (cb). * * @mactime: value in microseconds of the 64-bit Time Synchronization Function * (TSF) timer when the first data symbol (MPDU) arrived at the hardware. @@ -1606,9 +1611,11 @@ void ieee80211_free_hw(struct ieee80211_hw *hw); */ void ieee80211_restart_hw(struct ieee80211_hw *hw); -/* trick to avoid symbol clashes with the ieee80211 subsystem */ -void __ieee80211_rx(struct ieee80211_hw *hw, struct sk_buff *skb, - struct ieee80211_rx_status *status); +/* + * trick to avoid symbol clashes with the ieee80211 subsystem, + * use the inline below instead + */ +void __ieee80211_rx(struct ieee80211_hw *hw, struct sk_buff *skb); /** * ieee80211_rx - receive frame @@ -1624,13 +1631,10 @@ void __ieee80211_rx(struct ieee80211_hw *hw, struct sk_buff *skb, * * @hw: the hardware this frame came in on * @skb: the buffer to receive, owned by mac80211 after this call - * @status: status of this frame; the status pointer need not be valid - * after this function returns */ -static inline void ieee80211_rx(struct ieee80211_hw *hw, struct sk_buff *skb, - struct ieee80211_rx_status *status) +static inline void ieee80211_rx(struct ieee80211_hw *hw, struct sk_buff *skb) { - __ieee80211_rx(hw, skb, status); + __ieee80211_rx(hw, skb); } /** @@ -1644,13 +1648,8 @@ static inline void ieee80211_rx(struct ieee80211_hw *hw, struct sk_buff *skb, * * @hw: the hardware this frame came in on * @skb: the buffer to receive, owned by mac80211 after this call - * @status: status of this frame; the status pointer need not be valid - * after this function returns and is not freed by mac80211, - * it is recommended that it points to a stack area */ -void ieee80211_rx_irqsafe(struct ieee80211_hw *hw, - struct sk_buff *skb, - struct ieee80211_rx_status *status); +void ieee80211_rx_irqsafe(struct ieee80211_hw *hw, struct sk_buff *skb); /** * ieee80211_tx_status - transmit status callback diff --git a/net/mac80211/ibss.c b/net/mac80211/ibss.c index 0b30277eb366..15d5a53b59a8 100644 --- a/net/mac80211/ibss.c +++ b/net/mac80211/ibss.c @@ -705,7 +705,7 @@ static void ieee80211_ibss_rx_queued_mgmt(struct ieee80211_sub_if_data *sdata, struct ieee80211_mgmt *mgmt; u16 fc; - rx_status = (struct ieee80211_rx_status *) skb->cb; + rx_status = IEEE80211_SKB_RXCB(skb); mgmt = (struct ieee80211_mgmt *) skb->data; fc = le16_to_cpu(mgmt->frame_control); @@ -836,8 +836,7 @@ void ieee80211_ibss_notify_scan_completed(struct ieee80211_local *local) } ieee80211_rx_result -ieee80211_ibss_rx_mgmt(struct ieee80211_sub_if_data *sdata, struct sk_buff *skb, - struct ieee80211_rx_status *rx_status) +ieee80211_ibss_rx_mgmt(struct ieee80211_sub_if_data *sdata, struct sk_buff *skb) { struct ieee80211_local *local = sdata->local; struct ieee80211_mgmt *mgmt; @@ -852,7 +851,6 @@ ieee80211_ibss_rx_mgmt(struct ieee80211_sub_if_data *sdata, struct sk_buff *skb, switch (fc & IEEE80211_FCTL_STYPE) { case IEEE80211_STYPE_PROBE_RESP: case IEEE80211_STYPE_BEACON: - memcpy(skb->cb, rx_status, sizeof(*rx_status)); case IEEE80211_STYPE_PROBE_REQ: case IEEE80211_STYPE_AUTH: skb_queue_tail(&sdata->u.ibss.skb_queue, skb); diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h index 68eb5052179a..c65c65a9e697 100644 --- a/net/mac80211/ieee80211_i.h +++ b/net/mac80211/ieee80211_i.h @@ -943,8 +943,7 @@ extern const struct iw_handler_def ieee80211_iw_handler_def; /* STA code */ void ieee80211_sta_setup_sdata(struct ieee80211_sub_if_data *sdata); ieee80211_rx_result ieee80211_sta_rx_mgmt(struct ieee80211_sub_if_data *sdata, - struct sk_buff *skb, - struct ieee80211_rx_status *rx_status); + struct sk_buff *skb); int ieee80211_sta_commit(struct ieee80211_sub_if_data *sdata); int ieee80211_sta_set_ssid(struct ieee80211_sub_if_data *sdata, char *ssid, size_t len); int ieee80211_sta_get_ssid(struct ieee80211_sub_if_data *sdata, char *ssid, size_t *len); @@ -967,8 +966,7 @@ void ieee80211_sta_restart(struct ieee80211_sub_if_data *sdata); void ieee80211_ibss_notify_scan_completed(struct ieee80211_local *local); void ieee80211_ibss_setup_sdata(struct ieee80211_sub_if_data *sdata); ieee80211_rx_result -ieee80211_ibss_rx_mgmt(struct ieee80211_sub_if_data *sdata, struct sk_buff *skb, - struct ieee80211_rx_status *rx_status); +ieee80211_ibss_rx_mgmt(struct ieee80211_sub_if_data *sdata, struct sk_buff *skb); struct sta_info *ieee80211_ibss_add_sta(struct ieee80211_sub_if_data *sdata, u8 *bssid, u8 *addr, u32 supp_rates); int ieee80211_ibss_join(struct ieee80211_sub_if_data *sdata, @@ -988,9 +986,7 @@ int ieee80211_scan_results(struct ieee80211_local *local, char *buf, size_t len); void ieee80211_scan_cancel(struct ieee80211_local *local); ieee80211_rx_result -ieee80211_scan_rx(struct ieee80211_sub_if_data *sdata, - struct sk_buff *skb, - struct ieee80211_rx_status *rx_status); +ieee80211_scan_rx(struct ieee80211_sub_if_data *sdata, struct sk_buff *skb); int ieee80211_sta_set_extra_ie(struct ieee80211_sub_if_data *sdata, const char *ie, size_t len); diff --git a/net/mac80211/main.c b/net/mac80211/main.c index 092a017b237e..5b69f5f07299 100644 --- a/net/mac80211/main.c +++ b/net/mac80211/main.c @@ -330,19 +330,16 @@ static void ieee80211_tasklet_handler(unsigned long data) { struct ieee80211_local *local = (struct ieee80211_local *) data; struct sk_buff *skb; - struct ieee80211_rx_status rx_status; struct ieee80211_ra_tid *ra_tid; while ((skb = skb_dequeue(&local->skb_queue)) || (skb = skb_dequeue(&local->skb_queue_unreliable))) { switch (skb->pkt_type) { case IEEE80211_RX_MSG: - /* status is in skb->cb */ - memcpy(&rx_status, skb->cb, sizeof(rx_status)); /* Clear skb->pkt_type in order to not confuse kernel * netstack. */ skb->pkt_type = 0; - __ieee80211_rx(local_to_hw(local), skb, &rx_status); + ieee80211_rx(local_to_hw(local), skb); break; case IEEE80211_TX_STATUS_MSG: skb->pkt_type = 0; diff --git a/net/mac80211/mesh.c b/net/mac80211/mesh.c index 11cf45bce38a..542ea025494e 100644 --- a/net/mac80211/mesh.c +++ b/net/mac80211/mesh.c @@ -568,7 +568,7 @@ static void ieee80211_mesh_rx_queued_mgmt(struct ieee80211_sub_if_data *sdata, ifmsh = &sdata->u.mesh; - rx_status = (struct ieee80211_rx_status *) skb->cb; + rx_status = IEEE80211_SKB_RXCB(skb); mgmt = (struct ieee80211_mgmt *) skb->data; stype = le16_to_cpu(mgmt->frame_control) & IEEE80211_FCTL_STYPE; @@ -671,8 +671,7 @@ void ieee80211_mesh_init_sdata(struct ieee80211_sub_if_data *sdata) } ieee80211_rx_result -ieee80211_mesh_rx_mgmt(struct ieee80211_sub_if_data *sdata, struct sk_buff *skb, - struct ieee80211_rx_status *rx_status) +ieee80211_mesh_rx_mgmt(struct ieee80211_sub_if_data *sdata, struct sk_buff *skb) { struct ieee80211_local *local = sdata->local; struct ieee80211_if_mesh *ifmsh = &sdata->u.mesh; @@ -689,7 +688,6 @@ ieee80211_mesh_rx_mgmt(struct ieee80211_sub_if_data *sdata, struct sk_buff *skb, case IEEE80211_STYPE_PROBE_RESP: case IEEE80211_STYPE_BEACON: case IEEE80211_STYPE_ACTION: - memcpy(skb->cb, rx_status, sizeof(*rx_status)); skb_queue_tail(&ifmsh->skb_queue, skb); queue_work(local->hw.workqueue, &ifmsh->work); return RX_QUEUED; diff --git a/net/mac80211/mesh.h b/net/mac80211/mesh.h index c7d72819cdd2..2a2ed182cb7e 100644 --- a/net/mac80211/mesh.h +++ b/net/mac80211/mesh.h @@ -208,8 +208,7 @@ void ieee80211s_init(void); void ieee80211s_stop(void); void ieee80211_mesh_init_sdata(struct ieee80211_sub_if_data *sdata); ieee80211_rx_result -ieee80211_mesh_rx_mgmt(struct ieee80211_sub_if_data *sdata, struct sk_buff *skb, - struct ieee80211_rx_status *rx_status); +ieee80211_mesh_rx_mgmt(struct ieee80211_sub_if_data *sdata, struct sk_buff *skb); void ieee80211_start_mesh(struct ieee80211_sub_if_data *sdata); void ieee80211_stop_mesh(struct ieee80211_sub_if_data *sdata); diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index aca22b00b6a3..5e25d320deae 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -2063,8 +2063,7 @@ static void ieee80211_rx_mgmt_beacon(struct ieee80211_sub_if_data *sdata, } ieee80211_rx_result ieee80211_sta_rx_mgmt(struct ieee80211_sub_if_data *sdata, - struct sk_buff *skb, - struct ieee80211_rx_status *rx_status) + struct sk_buff *skb) { struct ieee80211_local *local = sdata->local; struct ieee80211_mgmt *mgmt; @@ -2080,7 +2079,6 @@ ieee80211_rx_result ieee80211_sta_rx_mgmt(struct ieee80211_sub_if_data *sdata, case IEEE80211_STYPE_PROBE_REQ: case IEEE80211_STYPE_PROBE_RESP: case IEEE80211_STYPE_BEACON: - memcpy(skb->cb, rx_status, sizeof(*rx_status)); case IEEE80211_STYPE_AUTH: case IEEE80211_STYPE_ASSOC_RESP: case IEEE80211_STYPE_REASSOC_RESP: diff --git a/net/mac80211/rx.c b/net/mac80211/rx.c index de5bba7f910a..0563b6969a21 100644 --- a/net/mac80211/rx.c +++ b/net/mac80211/rx.c @@ -30,7 +30,6 @@ static u8 ieee80211_sta_manage_reorder_buf(struct ieee80211_hw *hw, struct tid_ampdu_rx *tid_agg_rx, struct sk_buff *skb, - struct ieee80211_rx_status *status, u16 mpdu_seq_num, int bar_req); /* @@ -59,11 +58,11 @@ static struct sk_buff *remove_monitor_info(struct ieee80211_local *local, return skb; } -static inline int should_drop_frame(struct ieee80211_rx_status *status, - struct sk_buff *skb, +static inline int should_drop_frame(struct sk_buff *skb, int present_fcs_len, int radiotap_len) { + struct ieee80211_rx_status *status = IEEE80211_SKB_RXCB(skb); struct ieee80211_hdr *hdr = (struct ieee80211_hdr *)skb->data; if (status->flag & (RX_FLAG_FAILED_FCS_CRC | RX_FLAG_FAILED_PLCP_CRC)) @@ -111,10 +110,10 @@ ieee80211_rx_radiotap_len(struct ieee80211_local *local, static void ieee80211_add_rx_radiotap_header(struct ieee80211_local *local, struct sk_buff *skb, - struct ieee80211_rx_status *status, struct ieee80211_rate *rate, int rtap_len) { + struct ieee80211_rx_status *status = IEEE80211_SKB_RXCB(skb); struct ieee80211_radiotap_header *rthdr; unsigned char *pos; @@ -220,9 +219,9 @@ ieee80211_add_rx_radiotap_header(struct ieee80211_local *local, */ static struct sk_buff * ieee80211_rx_monitor(struct ieee80211_local *local, struct sk_buff *origskb, - struct ieee80211_rx_status *status, struct ieee80211_rate *rate) { + struct ieee80211_rx_status *status = IEEE80211_SKB_RXCB(origskb); struct ieee80211_sub_if_data *sdata; int needed_headroom = 0; struct sk_buff *skb, *skb2; @@ -248,8 +247,7 @@ ieee80211_rx_monitor(struct ieee80211_local *local, struct sk_buff *origskb, present_fcs_len = FCS_LEN; if (!local->monitors) { - if (should_drop_frame(status, origskb, present_fcs_len, - rtap_len)) { + if (should_drop_frame(origskb, present_fcs_len, rtap_len)) { dev_kfree_skb(origskb); return NULL; } @@ -257,7 +255,7 @@ ieee80211_rx_monitor(struct ieee80211_local *local, struct sk_buff *origskb, return remove_monitor_info(local, origskb, rtap_len); } - if (should_drop_frame(status, origskb, present_fcs_len, rtap_len)) { + if (should_drop_frame(origskb, present_fcs_len, rtap_len)) { /* only need to expand headroom if necessary */ skb = origskb; origskb = NULL; @@ -289,7 +287,7 @@ ieee80211_rx_monitor(struct ieee80211_local *local, struct sk_buff *origskb, /* if necessary, prepend radiotap information */ if (!(status->flag & RX_FLAG_RADIOTAP)) - ieee80211_add_rx_radiotap_header(local, skb, status, rate, + ieee80211_add_rx_radiotap_header(local, skb, rate, needed_headroom); skb_reset_mac_header(skb); @@ -421,12 +419,11 @@ ieee80211_rx_h_passive_scan(struct ieee80211_rx_data *rx) struct sk_buff *skb = rx->skb; if (unlikely(local->hw_scanning)) - return ieee80211_scan_rx(rx->sdata, skb, rx->status); + return ieee80211_scan_rx(rx->sdata, skb); if (unlikely(local->sw_scanning)) { /* drop all the other packets during a software scan anyway */ - if (ieee80211_scan_rx(rx->sdata, skb, rx->status) - != RX_QUEUED) + if (ieee80211_scan_rx(rx->sdata, skb) != RX_QUEUED) dev_kfree_skb(skb); return RX_QUEUED; } @@ -1620,7 +1617,7 @@ ieee80211_rx_h_ctrl(struct ieee80211_rx_data *rx) /* manage reordering buffer according to requested */ /* sequence number */ rcu_read_lock(); - ieee80211_sta_manage_reorder_buf(hw, tid_agg_rx, NULL, NULL, + ieee80211_sta_manage_reorder_buf(hw, tid_agg_rx, NULL, start_seq_num, 1); rcu_read_unlock(); return RX_DROP_UNUSABLE; @@ -1817,13 +1814,13 @@ ieee80211_rx_h_mgmt(struct ieee80211_rx_data *rx) return RX_DROP_MONITOR; if (ieee80211_vif_is_mesh(&sdata->vif)) - return ieee80211_mesh_rx_mgmt(sdata, rx->skb, rx->status); + return ieee80211_mesh_rx_mgmt(sdata, rx->skb); if (sdata->vif.type == NL80211_IFTYPE_ADHOC) - return ieee80211_ibss_rx_mgmt(sdata, rx->skb, rx->status); + return ieee80211_ibss_rx_mgmt(sdata, rx->skb); if (sdata->vif.type == NL80211_IFTYPE_STATION) - return ieee80211_sta_rx_mgmt(sdata, rx->skb, rx->status); + return ieee80211_sta_rx_mgmt(sdata, rx->skb); return RX_DROP_MONITOR; } @@ -2114,9 +2111,9 @@ static int prepare_for_handlers(struct ieee80211_sub_if_data *sdata, */ static void __ieee80211_rx_handle_packet(struct ieee80211_hw *hw, struct sk_buff *skb, - struct ieee80211_rx_status *status, struct ieee80211_rate *rate) { + struct ieee80211_rx_status *status = IEEE80211_SKB_RXCB(skb); struct ieee80211_local *local = hw_to_local(hw); struct ieee80211_sub_if_data *sdata; struct ieee80211_hdr *hdr; @@ -2227,20 +2224,21 @@ static void ieee80211_release_reorder_frame(struct ieee80211_hw *hw, { struct ieee80211_supported_band *sband; struct ieee80211_rate *rate; - struct ieee80211_rx_status status; + struct sk_buff *skb = tid_agg_rx->reorder_buf[index]; + struct ieee80211_rx_status *status; - if (!tid_agg_rx->reorder_buf[index]) + if (!skb) goto no_frame; + status = IEEE80211_SKB_RXCB(skb); + /* release the reordered frames to stack */ - memcpy(&status, tid_agg_rx->reorder_buf[index]->cb, sizeof(status)); - sband = hw->wiphy->bands[status.band]; - if (status.flag & RX_FLAG_HT) + sband = hw->wiphy->bands[status->band]; + if (status->flag & RX_FLAG_HT) rate = sband->bitrates; /* TODO: HT rates */ else - rate = &sband->bitrates[status.rate_idx]; - __ieee80211_rx_handle_packet(hw, tid_agg_rx->reorder_buf[index], - &status, rate); + rate = &sband->bitrates[status->rate_idx]; + __ieee80211_rx_handle_packet(hw, skb, rate); tid_agg_rx->stored_mpdu_num--; tid_agg_rx->reorder_buf[index] = NULL; @@ -2265,7 +2263,6 @@ no_frame: static u8 ieee80211_sta_manage_reorder_buf(struct ieee80211_hw *hw, struct tid_ampdu_rx *tid_agg_rx, struct sk_buff *skb, - struct ieee80211_rx_status *rxstatus, u16 mpdu_seq_num, int bar_req) { @@ -2324,8 +2321,6 @@ static u8 ieee80211_sta_manage_reorder_buf(struct ieee80211_hw *hw, /* put the frame in the reordering buffer */ tid_agg_rx->reorder_buf[index] = skb; tid_agg_rx->reorder_time[index] = jiffies; - memcpy(tid_agg_rx->reorder_buf[index]->cb, rxstatus, - sizeof(*rxstatus)); tid_agg_rx->stored_mpdu_num++; /* release the buffer until next missing frame */ index = seq_sub(tid_agg_rx->head_seq_num, tid_agg_rx->ssn) @@ -2374,8 +2369,7 @@ static u8 ieee80211_sta_manage_reorder_buf(struct ieee80211_hw *hw, } static u8 ieee80211_rx_reorder_ampdu(struct ieee80211_local *local, - struct sk_buff *skb, - struct ieee80211_rx_status *status) + struct sk_buff *skb) { struct ieee80211_hw *hw = &local->hw; struct ieee80211_hdr *hdr = (struct ieee80211_hdr *) skb->data; @@ -2424,7 +2418,7 @@ static u8 ieee80211_rx_reorder_ampdu(struct ieee80211_local *local, /* according to mpdu sequence number deal with reordering buffer */ mpdu_seq_num = (sc & IEEE80211_SCTL_SEQ) >> 4; - ret = ieee80211_sta_manage_reorder_buf(hw, tid_agg_rx, skb, status, + ret = ieee80211_sta_manage_reorder_buf(hw, tid_agg_rx, skb, mpdu_seq_num, 0); end_reorder: return ret; @@ -2434,12 +2428,12 @@ static u8 ieee80211_rx_reorder_ampdu(struct ieee80211_local *local, * This is the receive path handler. It is called by a low level driver when an * 802.11 MPDU is received from the hardware. */ -void __ieee80211_rx(struct ieee80211_hw *hw, struct sk_buff *skb, - struct ieee80211_rx_status *status) +void __ieee80211_rx(struct ieee80211_hw *hw, struct sk_buff *skb) { struct ieee80211_local *local = hw_to_local(hw); struct ieee80211_rate *rate = NULL; struct ieee80211_supported_band *sband; + struct ieee80211_rx_status *status = IEEE80211_SKB_RXCB(skb); if (status->band < 0 || status->band >= IEEE80211_NUM_BANDS) { @@ -2482,7 +2476,7 @@ void __ieee80211_rx(struct ieee80211_hw *hw, struct sk_buff *skb, * if it was previously present. * Also, frames with less than 16 bytes are dropped. */ - skb = ieee80211_rx_monitor(local, skb, status, rate); + skb = ieee80211_rx_monitor(local, skb, rate); if (!skb) { rcu_read_unlock(); return; @@ -2500,8 +2494,8 @@ void __ieee80211_rx(struct ieee80211_hw *hw, struct sk_buff *skb, * frames from other than operational channel), but that should not * happen in normal networks. */ - if (!ieee80211_rx_reorder_ampdu(local, skb, status)) - __ieee80211_rx_handle_packet(hw, skb, status, rate); + if (!ieee80211_rx_reorder_ampdu(local, skb)) + __ieee80211_rx_handle_packet(hw, skb, rate); rcu_read_unlock(); } @@ -2509,16 +2503,13 @@ EXPORT_SYMBOL(__ieee80211_rx); /* This is a version of the rx handler that can be called from hard irq * context. Post the skb on the queue and schedule the tasklet */ -void ieee80211_rx_irqsafe(struct ieee80211_hw *hw, struct sk_buff *skb, - struct ieee80211_rx_status *status) +void ieee80211_rx_irqsafe(struct ieee80211_hw *hw, struct sk_buff *skb) { struct ieee80211_local *local = hw_to_local(hw); BUILD_BUG_ON(sizeof(struct ieee80211_rx_status) > sizeof(skb->cb)); skb->dev = local->mdev; - /* copy status into skb->cb for use by tasklet */ - memcpy(skb->cb, status, sizeof(*status)); skb->pkt_type = IEEE80211_RX_MSG; skb_queue_tail(&local->skb_queue, skb); tasklet_schedule(&local->tasklet); diff --git a/net/mac80211/scan.c b/net/mac80211/scan.c index 2a8d09ad17ff..8b2416c77a6e 100644 --- a/net/mac80211/scan.c +++ b/net/mac80211/scan.c @@ -135,9 +135,9 @@ void ieee80211_rx_bss_remove(struct ieee80211_sub_if_data *sdata, u8 *bssid, } ieee80211_rx_result -ieee80211_scan_rx(struct ieee80211_sub_if_data *sdata, struct sk_buff *skb, - struct ieee80211_rx_status *rx_status) +ieee80211_scan_rx(struct ieee80211_sub_if_data *sdata, struct sk_buff *skb) { + struct ieee80211_rx_status *rx_status = IEEE80211_SKB_RXCB(skb); struct ieee80211_mgmt *mgmt; struct ieee80211_bss *bss; u8 *elements; -- cgit v1.2.3 From f1f74825fe01ac77204ca34e3240dec50a8207c2 Mon Sep 17 00:00:00 2001 From: David Kilroy Date: Thu, 18 Jun 2009 23:21:13 +0100 Subject: cfg80211: add wrapper function to get wiphy from priv pointer Signed-off-by: David Kilroy Acked-by: Johannes Berg Signed-off-by: John W. Linville --- include/net/cfg80211.h | 11 +++++++++++ 1 file changed, 11 insertions(+) (limited to 'include') diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index 90f9bfa3bfc2..dba7874d1963 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -1058,6 +1058,17 @@ static inline void *wiphy_priv(struct wiphy *wiphy) return &wiphy->priv; } +/** + * priv_to_wiphy - return the wiphy containing the priv + * + * @priv: a pointer previously returned by wiphy_priv + */ +static inline struct wiphy *priv_to_wiphy(void *priv) +{ + BUG_ON(!priv); + return container_of(priv, struct wiphy, priv); +} + /** * set_wiphy_dev - set device pointer for wiphy * -- cgit v1.2.3 From e6d6e3420d511cd7552a95d1f04bd4c80a9ddb34 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Wed, 1 Jul 2009 21:26:47 +0200 Subject: cfg80211: use proper allocation flags Instead of hardcoding GFP_ATOMIC everywhere, add a new function parameter that gets the flags from the caller. Obviously then I need to update all callers (all of them in mac80211), and it turns out that now it's ok to use GFP_KERNEL in almost all places. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- include/net/cfg80211.h | 21 +++++++++++++------- net/mac80211/event.c | 5 +++-- net/mac80211/ieee80211_i.h | 3 ++- net/mac80211/mlme.c | 30 ++++++++++++++++++----------- net/mac80211/rx.c | 3 ++- net/mac80211/wpa.c | 3 ++- net/wireless/mlme.c | 30 ++++++++++++++--------------- net/wireless/nl80211.c | 48 ++++++++++++++++++++++++++-------------------- net/wireless/nl80211.h | 14 +++++++------- 9 files changed, 91 insertions(+), 66 deletions(-) (limited to 'include') diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index dba7874d1963..1696ff647a08 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -1572,64 +1572,70 @@ void cfg80211_unlink_bss(struct wiphy *wiphy, struct cfg80211_bss *bss); * @dev: network device * @buf: authentication frame (header + body) * @len: length of the frame data + * @gfp: allocation flags * * This function is called whenever an authentication has been processed in * station mode. The driver is required to call either this function or * cfg80211_send_auth_timeout() to indicate the result of cfg80211_ops::auth() * call. */ -void cfg80211_send_rx_auth(struct net_device *dev, const u8 *buf, size_t len); +void cfg80211_send_rx_auth(struct net_device *dev, const u8 *buf, size_t len, gfp_t gfp); /** * cfg80211_send_auth_timeout - notification of timed out authentication * @dev: network device * @addr: The MAC address of the device with which the authentication timed out + * @gfp: allocation flags */ -void cfg80211_send_auth_timeout(struct net_device *dev, const u8 *addr); +void cfg80211_send_auth_timeout(struct net_device *dev, const u8 *addr, gfp_t gfp); /** * cfg80211_send_rx_assoc - notification of processed association * @dev: network device * @buf: (re)association response frame (header + body) * @len: length of the frame data + * @gfp: allocation flags * * This function is called whenever a (re)association response has been * processed in station mode. The driver is required to call either this * function or cfg80211_send_assoc_timeout() to indicate the result of * cfg80211_ops::assoc() call. */ -void cfg80211_send_rx_assoc(struct net_device *dev, const u8 *buf, size_t len); +void cfg80211_send_rx_assoc(struct net_device *dev, const u8 *buf, size_t len, gfp_t gfp); /** * cfg80211_send_assoc_timeout - notification of timed out association * @dev: network device * @addr: The MAC address of the device with which the association timed out + * @gfp: allocation flags */ -void cfg80211_send_assoc_timeout(struct net_device *dev, const u8 *addr); +void cfg80211_send_assoc_timeout(struct net_device *dev, const u8 *addr, gfp_t gfp); /** * cfg80211_send_deauth - notification of processed deauthentication * @dev: network device * @buf: deauthentication frame (header + body) * @len: length of the frame data + * @gfp: allocation flags * * This function is called whenever deauthentication has been processed in * station mode. This includes both received deauthentication frames and * locally generated ones. */ -void cfg80211_send_deauth(struct net_device *dev, const u8 *buf, size_t len); +void cfg80211_send_deauth(struct net_device *dev, const u8 *buf, size_t len, gfp_t gfp); /** * cfg80211_send_disassoc - notification of processed disassociation * @dev: network device * @buf: disassociation response frame (header + body) * @len: length of the frame data + * @gfp: allocation flags * * This function is called whenever disassociation has been processed in * station mode. This includes both received disassociation frames and locally * generated ones. */ -void cfg80211_send_disassoc(struct net_device *dev, const u8 *buf, size_t len); +void cfg80211_send_disassoc(struct net_device *dev, const u8 *buf, size_t len, gfp_t gfp); /** * cfg80211_hold_bss - exclude bss from expiration @@ -1655,6 +1661,7 @@ void cfg80211_unhold_bss(struct cfg80211_bss *bss); * @key_type: The key type that the received frame used * @key_id: Key identifier (0..3) * @tsc: The TSC value of the frame that generated the MIC failure (6 octets) + * @gfp: allocation flags * * This function is called whenever the local MAC detects a MIC failure in a * received frame. This matches with MLME-MICHAELMICFAILURE.indication() @@ -1662,7 +1669,7 @@ void cfg80211_unhold_bss(struct cfg80211_bss *bss); */ void cfg80211_michael_mic_failure(struct net_device *dev, const u8 *addr, enum nl80211_key_type key_type, int key_id, - const u8 *tsc); + const u8 *tsc, gfp_t gfp); /** * cfg80211_ibss_joined - notify cfg80211 that device joined an IBSS diff --git a/net/mac80211/event.c b/net/mac80211/event.c index 3ac636285fb1..01ae759518f6 100644 --- a/net/mac80211/event.c +++ b/net/mac80211/event.c @@ -16,11 +16,12 @@ * driver or is still in the frame), it should provide that information. */ void mac80211_ev_michael_mic_failure(struct ieee80211_sub_if_data *sdata, int keyidx, - struct ieee80211_hdr *hdr, const u8 *tsc) + struct ieee80211_hdr *hdr, const u8 *tsc, + gfp_t gfp) { cfg80211_michael_mic_failure(sdata->dev, hdr->addr2, (hdr->addr1[0] & 0x01) ? NL80211_KEYTYPE_GROUP : NL80211_KEYTYPE_PAIRWISE, - keyidx, tsc); + keyidx, tsc, gfp); } diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h index c65c65a9e697..e0323e540a03 100644 --- a/net/mac80211/ieee80211_i.h +++ b/net/mac80211/ieee80211_i.h @@ -1088,7 +1088,8 @@ u8 *ieee80211_get_bssid(struct ieee80211_hdr *hdr, size_t len, int ieee80211_frame_duration(struct ieee80211_local *local, size_t len, int rate, int erp, int short_preamble); void mac80211_ev_michael_mic_failure(struct ieee80211_sub_if_data *sdata, int keyidx, - struct ieee80211_hdr *hdr, const u8 *tsc); + struct ieee80211_hdr *hdr, const u8 *tsc, + gfp_t gfp); void ieee80211_set_wmm_default(struct ieee80211_sub_if_data *sdata); void ieee80211_tx_skb(struct ieee80211_sub_if_data *sdata, struct sk_buff *skb, int encrypt); diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index 5e25d320deae..383392b04282 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -419,9 +419,11 @@ static void ieee80211_send_deauth_disassoc(struct ieee80211_sub_if_data *sdata, mgmt->u.deauth.reason_code = cpu_to_le16(reason); if (stype == IEEE80211_STYPE_DEAUTH) - cfg80211_send_deauth(sdata->dev, (u8 *) mgmt, skb->len); + cfg80211_send_deauth(sdata->dev, (u8 *) mgmt, skb->len, + GFP_KERNEL); else - cfg80211_send_disassoc(sdata->dev, (u8 *) mgmt, skb->len); + cfg80211_send_disassoc(sdata->dev, (u8 *) mgmt, skb->len, + GFP_KERNEL); ieee80211_tx_skb(sdata, skb, ifmgd->flags & IEEE80211_STA_MFP_ENABLED); } @@ -1006,7 +1008,8 @@ static void ieee80211_direct_probe(struct ieee80211_sub_if_data *sdata) sdata->dev->name, ifmgd->bssid); ifmgd->state = IEEE80211_STA_MLME_DISABLED; ieee80211_recalc_idle(local); - cfg80211_send_auth_timeout(sdata->dev, ifmgd->bssid); + cfg80211_send_auth_timeout(sdata->dev, ifmgd->bssid, + GFP_KERNEL); /* * Most likely AP is not in the range so remove the @@ -1055,7 +1058,8 @@ static void ieee80211_authenticate(struct ieee80211_sub_if_data *sdata) sdata->dev->name, ifmgd->bssid); ifmgd->state = IEEE80211_STA_MLME_DISABLED; ieee80211_recalc_idle(local); - cfg80211_send_auth_timeout(sdata->dev, ifmgd->bssid); + cfg80211_send_auth_timeout(sdata->dev, ifmgd->bssid, + GFP_KERNEL); ieee80211_rx_bss_remove(sdata, ifmgd->bssid, sdata->local->hw.conf.channel->center_freq, ifmgd->ssid, ifmgd->ssid_len); @@ -1243,7 +1247,8 @@ static void ieee80211_associate(struct ieee80211_sub_if_data *sdata) sdata->dev->name, ifmgd->bssid); ifmgd->state = IEEE80211_STA_MLME_DISABLED; ieee80211_recalc_idle(local); - cfg80211_send_assoc_timeout(sdata->dev, ifmgd->bssid); + cfg80211_send_assoc_timeout(sdata->dev, ifmgd->bssid, + GFP_KERNEL); ieee80211_rx_bss_remove(sdata, ifmgd->bssid, sdata->local->hw.conf.channel->center_freq, ifmgd->ssid, ifmgd->ssid_len); @@ -1517,12 +1522,14 @@ static void ieee80211_rx_mgmt_auth(struct ieee80211_sub_if_data *sdata, case WLAN_AUTH_LEAP: case WLAN_AUTH_FT: ieee80211_auth_completed(sdata); - cfg80211_send_rx_auth(sdata->dev, (u8 *) mgmt, len); + cfg80211_send_rx_auth(sdata->dev, (u8 *) mgmt, len, + GFP_KERNEL); break; case WLAN_AUTH_SHARED_KEY: if (ifmgd->auth_transaction == 4) { ieee80211_auth_completed(sdata); - cfg80211_send_rx_auth(sdata->dev, (u8 *) mgmt, len); + cfg80211_send_rx_auth(sdata->dev, (u8 *) mgmt, len, + GFP_KERNEL); } else ieee80211_auth_challenge(sdata, mgmt, len); break; @@ -1560,7 +1567,7 @@ static void ieee80211_rx_mgmt_deauth(struct ieee80211_sub_if_data *sdata, ieee80211_set_disassoc(sdata, true, false, 0); ifmgd->flags &= ~IEEE80211_STA_AUTHENTICATED; - cfg80211_send_deauth(sdata->dev, (u8 *) mgmt, len); + cfg80211_send_deauth(sdata->dev, (u8 *) mgmt, len, GFP_KERNEL); } @@ -1591,7 +1598,7 @@ static void ieee80211_rx_mgmt_disassoc(struct ieee80211_sub_if_data *sdata, } ieee80211_set_disassoc(sdata, false, false, reason_code); - cfg80211_send_disassoc(sdata->dev, (u8 *) mgmt, len); + cfg80211_send_disassoc(sdata->dev, (u8 *) mgmt, len, GFP_KERNEL); } @@ -1660,7 +1667,8 @@ static void ieee80211_rx_mgmt_assoc_resp(struct ieee80211_sub_if_data *sdata, * association next time. This works around some broken APs * which do not correctly reject reassociation requests. */ ifmgd->flags &= ~IEEE80211_STA_PREV_BSSID_SET; - cfg80211_send_rx_assoc(sdata->dev, (u8 *) mgmt, len); + cfg80211_send_rx_assoc(sdata->dev, (u8 *) mgmt, len, + GFP_KERNEL); if (ifmgd->flags & IEEE80211_STA_EXT_SME) { /* Wait for SME to decide what to do next */ ifmgd->state = IEEE80211_STA_MLME_DISABLED; @@ -1823,7 +1831,7 @@ static void ieee80211_rx_mgmt_assoc_resp(struct ieee80211_sub_if_data *sdata, ifmgd->last_beacon = jiffies; ieee80211_associated(sdata); - cfg80211_send_rx_assoc(sdata->dev, (u8 *) mgmt, len); + cfg80211_send_rx_assoc(sdata->dev, (u8 *) mgmt, len, GFP_KERNEL); } diff --git a/net/mac80211/rx.c b/net/mac80211/rx.c index 0563b6969a21..ec5acc6dc021 100644 --- a/net/mac80211/rx.c +++ b/net/mac80211/rx.c @@ -1863,7 +1863,8 @@ static void ieee80211_rx_michael_mic_report(struct net_device *dev, !ieee80211_is_auth(hdr->frame_control)) goto ignore; - mac80211_ev_michael_mic_failure(rx->sdata, keyidx, hdr, NULL); + mac80211_ev_michael_mic_failure(rx->sdata, keyidx, hdr, NULL, + GFP_ATOMIC); ignore: dev_kfree_skb(rx->skb); rx->skb = NULL; diff --git a/net/mac80211/wpa.c b/net/mac80211/wpa.c index dcfae8884b86..70778694877b 100644 --- a/net/mac80211/wpa.c +++ b/net/mac80211/wpa.c @@ -122,7 +122,8 @@ ieee80211_rx_h_michael_mic_verify(struct ieee80211_rx_data *rx) return RX_DROP_UNUSABLE; mac80211_ev_michael_mic_failure(rx->sdata, rx->key->conf.keyidx, - (void *) skb->data, NULL); + (void *) skb->data, NULL, + GFP_ATOMIC); return RX_DROP_UNUSABLE; } diff --git a/net/wireless/mlme.c b/net/wireless/mlme.c index e56bbea10fc8..c4e6d4b84a46 100644 --- a/net/wireless/mlme.c +++ b/net/wireless/mlme.c @@ -12,35 +12,35 @@ #include "core.h" #include "nl80211.h" -void cfg80211_send_rx_auth(struct net_device *dev, const u8 *buf, size_t len) +void cfg80211_send_rx_auth(struct net_device *dev, const u8 *buf, size_t len, gfp_t gfp) { struct wiphy *wiphy = dev->ieee80211_ptr->wiphy; struct cfg80211_registered_device *rdev = wiphy_to_dev(wiphy); - nl80211_send_rx_auth(rdev, dev, buf, len); + nl80211_send_rx_auth(rdev, dev, buf, len, gfp); } EXPORT_SYMBOL(cfg80211_send_rx_auth); -void cfg80211_send_rx_assoc(struct net_device *dev, const u8 *buf, size_t len) +void cfg80211_send_rx_assoc(struct net_device *dev, const u8 *buf, size_t len, gfp_t gfp) { struct wiphy *wiphy = dev->ieee80211_ptr->wiphy; struct cfg80211_registered_device *rdev = wiphy_to_dev(wiphy); - nl80211_send_rx_assoc(rdev, dev, buf, len); + nl80211_send_rx_assoc(rdev, dev, buf, len, gfp); } EXPORT_SYMBOL(cfg80211_send_rx_assoc); -void cfg80211_send_deauth(struct net_device *dev, const u8 *buf, size_t len) +void cfg80211_send_deauth(struct net_device *dev, const u8 *buf, size_t len, gfp_t gfp) { struct wiphy *wiphy = dev->ieee80211_ptr->wiphy; struct cfg80211_registered_device *rdev = wiphy_to_dev(wiphy); - nl80211_send_deauth(rdev, dev, buf, len); + nl80211_send_deauth(rdev, dev, buf, len, gfp); } EXPORT_SYMBOL(cfg80211_send_deauth); -void cfg80211_send_disassoc(struct net_device *dev, const u8 *buf, size_t len) +void cfg80211_send_disassoc(struct net_device *dev, const u8 *buf, size_t len, gfp_t gfp) { struct wiphy *wiphy = dev->ieee80211_ptr->wiphy; struct cfg80211_registered_device *rdev = wiphy_to_dev(wiphy); - nl80211_send_disassoc(rdev, dev, buf, len); + nl80211_send_disassoc(rdev, dev, buf, len, gfp); } EXPORT_SYMBOL(cfg80211_send_disassoc); @@ -53,33 +53,33 @@ static void cfg80211_wext_disconnected(struct net_device *dev) #endif } -void cfg80211_send_auth_timeout(struct net_device *dev, const u8 *addr) +void cfg80211_send_auth_timeout(struct net_device *dev, const u8 *addr, gfp_t gfp) { struct wiphy *wiphy = dev->ieee80211_ptr->wiphy; struct cfg80211_registered_device *rdev = wiphy_to_dev(wiphy); - nl80211_send_auth_timeout(rdev, dev, addr); + nl80211_send_auth_timeout(rdev, dev, addr, gfp); cfg80211_wext_disconnected(dev); } EXPORT_SYMBOL(cfg80211_send_auth_timeout); -void cfg80211_send_assoc_timeout(struct net_device *dev, const u8 *addr) +void cfg80211_send_assoc_timeout(struct net_device *dev, const u8 *addr, gfp_t gfp) { struct wiphy *wiphy = dev->ieee80211_ptr->wiphy; struct cfg80211_registered_device *rdev = wiphy_to_dev(wiphy); - nl80211_send_assoc_timeout(rdev, dev, addr); + nl80211_send_assoc_timeout(rdev, dev, addr, gfp); cfg80211_wext_disconnected(dev); } EXPORT_SYMBOL(cfg80211_send_assoc_timeout); void cfg80211_michael_mic_failure(struct net_device *dev, const u8 *addr, enum nl80211_key_type key_type, int key_id, - const u8 *tsc) + const u8 *tsc, gfp_t gfp) { struct wiphy *wiphy = dev->ieee80211_ptr->wiphy; struct cfg80211_registered_device *rdev = wiphy_to_dev(wiphy); #ifdef CONFIG_WIRELESS_EXT union iwreq_data wrqu; - char *buf = kmalloc(128, GFP_ATOMIC); + char *buf = kmalloc(128, gfp); if (buf) { sprintf(buf, "MLME-MICHAELMICFAILURE.indication(" @@ -93,6 +93,6 @@ void cfg80211_michael_mic_failure(struct net_device *dev, const u8 *addr, } #endif - nl80211_michael_mic_failure(rdev, dev, addr, key_type, key_id, tsc); + nl80211_michael_mic_failure(rdev, dev, addr, key_type, key_id, tsc, gfp); } EXPORT_SYMBOL(cfg80211_michael_mic_failure); diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index 7946b82c5716..01523ba81baf 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -3832,12 +3832,12 @@ nla_put_failure: static void nl80211_send_mlme_event(struct cfg80211_registered_device *rdev, struct net_device *netdev, const u8 *buf, size_t len, - enum nl80211_commands cmd) + enum nl80211_commands cmd, gfp_t gfp) { struct sk_buff *msg; void *hdr; - msg = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_ATOMIC); + msg = nlmsg_new(NLMSG_DEFAULT_SIZE, gfp); if (!msg) return; @@ -3856,7 +3856,7 @@ static void nl80211_send_mlme_event(struct cfg80211_registered_device *rdev, return; } - genlmsg_multicast(msg, 0, nl80211_mlme_mcgrp.id, GFP_ATOMIC); + genlmsg_multicast(msg, 0, nl80211_mlme_mcgrp.id, gfp); return; nla_put_failure: @@ -3865,42 +3865,45 @@ static void nl80211_send_mlme_event(struct cfg80211_registered_device *rdev, } void nl80211_send_rx_auth(struct cfg80211_registered_device *rdev, - struct net_device *netdev, const u8 *buf, size_t len) + struct net_device *netdev, const u8 *buf, + size_t len, gfp_t gfp) { nl80211_send_mlme_event(rdev, netdev, buf, len, - NL80211_CMD_AUTHENTICATE); + NL80211_CMD_AUTHENTICATE, gfp); } void nl80211_send_rx_assoc(struct cfg80211_registered_device *rdev, struct net_device *netdev, const u8 *buf, - size_t len) + size_t len, gfp_t gfp) { - nl80211_send_mlme_event(rdev, netdev, buf, len, NL80211_CMD_ASSOCIATE); + nl80211_send_mlme_event(rdev, netdev, buf, len, + NL80211_CMD_ASSOCIATE, gfp); } void nl80211_send_deauth(struct cfg80211_registered_device *rdev, - struct net_device *netdev, const u8 *buf, size_t len) + struct net_device *netdev, const u8 *buf, + size_t len, gfp_t gfp) { nl80211_send_mlme_event(rdev, netdev, buf, len, - NL80211_CMD_DEAUTHENTICATE); + NL80211_CMD_DEAUTHENTICATE, gfp); } void nl80211_send_disassoc(struct cfg80211_registered_device *rdev, struct net_device *netdev, const u8 *buf, - size_t len) + size_t len, gfp_t gfp) { nl80211_send_mlme_event(rdev, netdev, buf, len, - NL80211_CMD_DISASSOCIATE); + NL80211_CMD_DISASSOCIATE, gfp); } static void nl80211_send_mlme_timeout(struct cfg80211_registered_device *rdev, struct net_device *netdev, int cmd, - const u8 *addr) + const u8 *addr, gfp_t gfp) { struct sk_buff *msg; void *hdr; - msg = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_ATOMIC); + msg = nlmsg_new(NLMSG_DEFAULT_SIZE, gfp); if (!msg) return; @@ -3920,7 +3923,7 @@ static void nl80211_send_mlme_timeout(struct cfg80211_registered_device *rdev, return; } - genlmsg_multicast(msg, 0, nl80211_mlme_mcgrp.id, GFP_ATOMIC); + genlmsg_multicast(msg, 0, nl80211_mlme_mcgrp.id, gfp); return; nla_put_failure: @@ -3929,16 +3932,19 @@ static void nl80211_send_mlme_timeout(struct cfg80211_registered_device *rdev, } void nl80211_send_auth_timeout(struct cfg80211_registered_device *rdev, - struct net_device *netdev, const u8 *addr) + struct net_device *netdev, const u8 *addr, + gfp_t gfp) { nl80211_send_mlme_timeout(rdev, netdev, NL80211_CMD_AUTHENTICATE, - addr); + addr, gfp); } void nl80211_send_assoc_timeout(struct cfg80211_registered_device *rdev, - struct net_device *netdev, const u8 *addr) + struct net_device *netdev, const u8 *addr, + gfp_t gfp) { - nl80211_send_mlme_timeout(rdev, netdev, NL80211_CMD_ASSOCIATE, addr); + nl80211_send_mlme_timeout(rdev, netdev, NL80211_CMD_ASSOCIATE, + addr, gfp); } void nl80211_send_ibss_bssid(struct cfg80211_registered_device *rdev, @@ -3978,12 +3984,12 @@ void nl80211_send_ibss_bssid(struct cfg80211_registered_device *rdev, void nl80211_michael_mic_failure(struct cfg80211_registered_device *rdev, struct net_device *netdev, const u8 *addr, enum nl80211_key_type key_type, int key_id, - const u8 *tsc) + const u8 *tsc, gfp_t gfp) { struct sk_buff *msg; void *hdr; - msg = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_ATOMIC); + msg = nlmsg_new(NLMSG_DEFAULT_SIZE, gfp); if (!msg) return; @@ -4007,7 +4013,7 @@ void nl80211_michael_mic_failure(struct cfg80211_registered_device *rdev, return; } - genlmsg_multicast(msg, 0, nl80211_mlme_mcgrp.id, GFP_ATOMIC); + genlmsg_multicast(msg, 0, nl80211_mlme_mcgrp.id, gfp); return; nla_put_failure: diff --git a/net/wireless/nl80211.h b/net/wireless/nl80211.h index cf0d271f7e1f..662c216e8d4f 100644 --- a/net/wireless/nl80211.h +++ b/net/wireless/nl80211.h @@ -15,27 +15,27 @@ void nl80211_send_scan_aborted(struct cfg80211_registered_device *rdev, void nl80211_send_reg_change_event(struct regulatory_request *request); void nl80211_send_rx_auth(struct cfg80211_registered_device *rdev, struct net_device *netdev, - const u8 *buf, size_t len); + const u8 *buf, size_t len, gfp_t gfp); void nl80211_send_rx_assoc(struct cfg80211_registered_device *rdev, struct net_device *netdev, - const u8 *buf, size_t len); + const u8 *buf, size_t len, gfp_t gfp); void nl80211_send_deauth(struct cfg80211_registered_device *rdev, struct net_device *netdev, - const u8 *buf, size_t len); + const u8 *buf, size_t len, gfp_t gfp); void nl80211_send_disassoc(struct cfg80211_registered_device *rdev, struct net_device *netdev, - const u8 *buf, size_t len); + const u8 *buf, size_t len, gfp_t gfp); void nl80211_send_auth_timeout(struct cfg80211_registered_device *rdev, struct net_device *netdev, - const u8 *addr); + const u8 *addr, gfp_t gfp); void nl80211_send_assoc_timeout(struct cfg80211_registered_device *rdev, struct net_device *netdev, - const u8 *addr); + const u8 *addr, gfp_t gfp); void nl80211_michael_mic_failure(struct cfg80211_registered_device *rdev, struct net_device *netdev, const u8 *addr, enum nl80211_key_type key_type, - int key_id, const u8 *tsc); + int key_id, const u8 *tsc, gfp_t gfp); void nl80211_send_beacon_hint_event(struct wiphy *wiphy, -- cgit v1.2.3 From 7ebbe6bd51a259e16608b3fd7b578f5dd1292a45 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Wed, 1 Jul 2009 21:26:48 +0200 Subject: cfg80211: remove wireless_dev->bssid This variable isn't necessary -- the wext code keeps track of the BSSID itself, and otherwise we have current_bss. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- include/net/cfg80211.h | 1 - net/wireless/ibss.c | 13 +++---------- 2 files changed, 3 insertions(+), 11 deletions(-) (limited to 'include') diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index 1696ff647a08..10eb53e2bc9f 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -1170,7 +1170,6 @@ struct wireless_dev { /* currently used for IBSS - might be rearranged in the future */ struct cfg80211_bss *current_bss; - u8 bssid[ETH_ALEN]; u8 ssid[IEEE80211_MAX_SSID_LEN]; u8 ssid_len; diff --git a/net/wireless/ibss.c b/net/wireless/ibss.c index a4a1c3498ff2..34b11eae30c8 100644 --- a/net/wireless/ibss.c +++ b/net/wireless/ibss.c @@ -24,9 +24,6 @@ void cfg80211_ibss_joined(struct net_device *dev, const u8 *bssid, gfp_t gfp) if (WARN_ON(!wdev->ssid_len)) return; - if (memcmp(bssid, wdev->bssid, ETH_ALEN) == 0) - return; - bss = cfg80211_get_bss(wdev->wiphy, NULL, bssid, wdev->ssid, wdev->ssid_len, WLAN_CAPABILITY_IBSS, WLAN_CAPABILITY_IBSS); @@ -41,7 +38,6 @@ void cfg80211_ibss_joined(struct net_device *dev, const u8 *bssid, gfp_t gfp) cfg80211_hold_bss(bss); wdev->current_bss = bss; - memcpy(wdev->bssid, bssid, ETH_ALEN); nl80211_send_ibss_bssid(wiphy_to_dev(wdev->wiphy), dev, bssid, gfp); #ifdef CONFIG_WIRELESS_EXT @@ -87,7 +83,6 @@ void cfg80211_clear_ibss(struct net_device *dev, bool nowext) wdev->current_bss = NULL; wdev->ssid_len = 0; - memset(wdev->bssid, 0, ETH_ALEN); #ifdef CONFIG_WIRELESS_EXT if (!nowext) wdev->wext.ibss.ssid_len = 0; @@ -356,12 +351,10 @@ int cfg80211_ibss_wext_giwap(struct net_device *dev, ap_addr->sa_family = ARPHRD_ETHER; - if (wdev->wext.ibss.bssid) { + if (wdev->current_bss) + memcpy(ap_addr->sa_data, wdev->current_bss->bssid, ETH_ALEN); + else memcpy(ap_addr->sa_data, wdev->wext.ibss.bssid, ETH_ALEN); - return 0; - } - - memcpy(ap_addr->sa_data, wdev->bssid, ETH_ALEN); return 0; } /* temporary symbol - mark GPL - in the future the handler won't be */ -- cgit v1.2.3 From 5121ea0481f9cea1dfd958f18d7b4ac78778cd40 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Wed, 1 Jul 2009 21:26:50 +0200 Subject: wext: constify extra argument to wireless_send_event This is never changed by the function, so can be marked const. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- include/net/iw_handler.h | 2 +- net/wireless/wext.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/include/net/iw_handler.h b/include/net/iw_handler.h index 51b9a37de991..2b3fbbb8669e 100644 --- a/include/net/iw_handler.h +++ b/include/net/iw_handler.h @@ -443,7 +443,7 @@ extern int dev_get_wireless_info(char * buffer, char **start, off_t offset, extern void wireless_send_event(struct net_device * dev, unsigned int cmd, union iwreq_data * wrqu, - char * extra); + const char * extra); /* We may need a function to send a stream of events to user space. * More on that later... */ diff --git a/net/wireless/wext.c b/net/wireless/wext.c index 5da07e079c20..425f7d58b961 100644 --- a/net/wireless/wext.c +++ b/net/wireless/wext.c @@ -1376,7 +1376,7 @@ static void rtmsg_iwinfo(struct net_device *dev, char *event, int event_len) void wireless_send_event(struct net_device * dev, unsigned int cmd, union iwreq_data * wrqu, - char * extra) + const char * extra) { const struct iw_ioctl_description * descr = NULL; int extra_len = 0; -- cgit v1.2.3 From aff89a9b9084931e51b89d8f3ee3c547bea6c422 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Wed, 1 Jul 2009 21:26:51 +0200 Subject: cfg80211: introduce nl80211 testmode command This introduces a new NL80211_CMD_TESTMODE for testing and calibration use with nl80211. There's no multiplexing like like iwpriv had, and the command is not available by default, it needs to be explicitly enabled in Kconfig and shouldn't be enabled in most kernels. The command requires a wiphy index or interface index to identify the device to operate on, and the new TESTDATA attribute. There also is API for sending replies to the command, and testmode multicast messages (on a testmode multicast group). I've also updated mac80211 to be able to pass through the command to the driver, since it itself doesn't implement the testmode command. Additionally, to give people an idea of how to use the command, I've added a little code to hwsim that makes use of the new command to set the powersave mode, this is currently done via debugfs and should remain there, and the testmode command only serves as an example of how to use this best -- with nested netlink attributes in the TESTDATA attribute. A hwsim testmode tool can be found at http://git.sipsolutions.net/hwsim.git/. This tool is BSD licensed so people can easily use it as a basis for their own internal fabrication and validation tools. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- drivers/net/wireless/mac80211_hwsim.c | 68 +++++++++++++++++ include/linux/nl80211.h | 11 +++ include/net/cfg80211.h | 83 +++++++++++++++++++++ include/net/mac80211.h | 5 ++ net/mac80211/cfg.c | 13 ++++ net/wireless/Kconfig | 15 ++++ net/wireless/core.h | 4 + net/wireless/nl80211.c | 136 ++++++++++++++++++++++++++++++++++ 8 files changed, 335 insertions(+) (limited to 'include') diff --git a/drivers/net/wireless/mac80211_hwsim.c b/drivers/net/wireless/mac80211_hwsim.c index 93c1c4a73e6c..6ac8565072e9 100644 --- a/drivers/net/wireless/mac80211_hwsim.c +++ b/drivers/net/wireless/mac80211_hwsim.c @@ -700,6 +700,73 @@ static int mac80211_hwsim_conf_tx( return 0; } +#ifdef CONFIG_NL80211_TESTMODE +/* + * This section contains example code for using netlink + * attributes with the testmode command in nl80211. + */ + +/* These enums need to be kept in sync with userspace */ +enum hwsim_testmode_attr { + __HWSIM_TM_ATTR_INVALID = 0, + HWSIM_TM_ATTR_CMD = 1, + HWSIM_TM_ATTR_PS = 2, + + /* keep last */ + __HWSIM_TM_ATTR_AFTER_LAST, + HWSIM_TM_ATTR_MAX = __HWSIM_TM_ATTR_AFTER_LAST - 1 +}; + +enum hwsim_testmode_cmd { + HWSIM_TM_CMD_SET_PS = 0, + HWSIM_TM_CMD_GET_PS = 1, +}; + +static const struct nla_policy hwsim_testmode_policy[HWSIM_TM_ATTR_MAX + 1] = { + [HWSIM_TM_ATTR_CMD] = { .type = NLA_U32 }, + [HWSIM_TM_ATTR_PS] = { .type = NLA_U32 }, +}; + +static int hwsim_fops_ps_write(void *dat, u64 val); + +int mac80211_hwsim_testmode_cmd(struct ieee80211_hw *hw, void *data, int len) +{ + struct mac80211_hwsim_data *hwsim = hw->priv; + struct nlattr *tb[HWSIM_TM_ATTR_MAX + 1]; + struct sk_buff *skb; + int err, ps; + + err = nla_parse(tb, HWSIM_TM_ATTR_MAX, data, len, + hwsim_testmode_policy); + if (err) + return err; + + if (!tb[HWSIM_TM_ATTR_CMD]) + return -EINVAL; + + switch (nla_get_u32(tb[HWSIM_TM_ATTR_CMD])) { + case HWSIM_TM_CMD_SET_PS: + if (!tb[HWSIM_TM_ATTR_PS]) + return -EINVAL; + ps = nla_get_u32(tb[HWSIM_TM_ATTR_PS]); + return hwsim_fops_ps_write(hwsim, ps); + case HWSIM_TM_CMD_GET_PS: + skb = cfg80211_testmode_alloc_reply_skb(hw->wiphy, + nla_total_size(sizeof(u32))); + if (!skb) + return -ENOMEM; + NLA_PUT_U32(skb, HWSIM_TM_ATTR_PS, hwsim->ps); + return cfg80211_testmode_reply(skb); + default: + return -EOPNOTSUPP; + } + + nla_put_failure: + kfree_skb(skb); + return -ENOBUFS; +} +#endif + static const struct ieee80211_ops mac80211_hwsim_ops = { .tx = mac80211_hwsim_tx, @@ -713,6 +780,7 @@ static const struct ieee80211_ops mac80211_hwsim_ops = .sta_notify = mac80211_hwsim_sta_notify, .set_tim = mac80211_hwsim_set_tim, .conf_tx = mac80211_hwsim_conf_tx, + CFG80211_TESTMODE_CMD(mac80211_hwsim_testmode_cmd) }; diff --git a/include/linux/nl80211.h b/include/linux/nl80211.h index dbea93b694e5..651b18839088 100644 --- a/include/linux/nl80211.h +++ b/include/linux/nl80211.h @@ -242,6 +242,10 @@ * @NL80211_CMD_LEAVE_IBSS: Leave the IBSS -- no special arguments, the IBSS is * determined by the network interface. * + * @NL80211_CMD_TESTMODE: testmode command, takes a wiphy (or ifindex) attribute + * to identify the device, and the TESTDATA blob attribute to pass through + * to the driver. + * * @NL80211_CMD_MAX: highest used command number * @__NL80211_CMD_AFTER_LAST: internal use */ @@ -310,6 +314,8 @@ enum nl80211_commands { NL80211_CMD_JOIN_IBSS, NL80211_CMD_LEAVE_IBSS, + NL80211_CMD_TESTMODE, + /* add new commands above here */ /* used to define NL80211_CMD_MAX below */ @@ -511,6 +517,9 @@ enum nl80211_commands { * authorized by user space. Otherwise, port is marked authorized by * default in station mode. * + * @NL80211_ATTR_TESTDATA: Testmode data blob, passed through to the driver. + * We recommend using nested, driver-specific attributes within this. + * * @NL80211_ATTR_MAX: highest attribute number currently defined * @__NL80211_ATTR_AFTER_LAST: internal use */ @@ -619,6 +628,8 @@ enum nl80211_attrs { NL80211_ATTR_CONTROL_PORT, + NL80211_ATTR_TESTDATA, + /* add attributes here, update the policy in nl80211.c */ __NL80211_ATTR_AFTER_LAST, diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index 10eb53e2bc9f..885d4e5bc4b5 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -857,6 +857,8 @@ enum tx_power_setting { * * @rfkill_poll: polls the hw rfkill line, use cfg80211 reporting * functions to adjust rfkill hw state + * + * @testmode_cmd: run a test mode command */ struct cfg80211_ops { int (*suspend)(struct wiphy *wiphy); @@ -955,6 +957,10 @@ struct cfg80211_ops { int (*get_tx_power)(struct wiphy *wiphy, int *dbm); void (*rfkill_poll)(struct wiphy *wiphy); + +#ifdef CONFIG_NL80211_TESTMODE + int (*testmode_cmd)(struct wiphy *wiphy, void *data, int len); +#endif }; /* @@ -1705,4 +1711,81 @@ void wiphy_rfkill_start_polling(struct wiphy *wiphy); */ void wiphy_rfkill_stop_polling(struct wiphy *wiphy); +#ifdef CONFIG_NL80211_TESTMODE +/** + * cfg80211_testmode_alloc_reply_skb - allocate testmode reply + * @wiphy: the wiphy + * @approxlen: an upper bound of the length of the data that will + * be put into the skb + * + * This function allocates and pre-fills an skb for a reply to + * the testmode command. Since it is intended for a reply, calling + * it outside of the @testmode_cmd operation is invalid. + * + * The returned skb (or %NULL if any errors happen) is pre-filled + * with the wiphy index and set up in a way that any data that is + * put into the skb (with skb_put(), nla_put() or similar) will end + * up being within the %NL80211_ATTR_TESTDATA attribute, so all that + * needs to be done with the skb is adding data for the corresponding + * userspace tool which can then read that data out of the testdata + * attribute. You must not modify the skb in any other way. + * + * When done, call cfg80211_testmode_reply() with the skb and return + * its error code as the result of the @testmode_cmd operation. + */ +struct sk_buff *cfg80211_testmode_alloc_reply_skb(struct wiphy *wiphy, + int approxlen); + +/** + * cfg80211_testmode_reply - send the reply skb + * @skb: The skb, must have been allocated with + * cfg80211_testmode_alloc_reply_skb() + * + * Returns an error code or 0 on success, since calling this + * function will usually be the last thing before returning + * from the @testmode_cmd you should return the error code. + * Note that this function consumes the skb regardless of the + * return value. + */ +int cfg80211_testmode_reply(struct sk_buff *skb); + +/** + * cfg80211_testmode_alloc_event_skb - allocate testmode event + * @wiphy: the wiphy + * @approxlen: an upper bound of the length of the data that will + * be put into the skb + * @gfp: allocation flags + * + * This function allocates and pre-fills an skb for an event on the + * testmode multicast group. + * + * The returned skb (or %NULL if any errors happen) is set up in the + * same way as with cfg80211_testmode_alloc_reply_skb() but prepared + * for an event. As there, you should simply add data to it that will + * then end up in the %NL80211_ATTR_TESTDATA attribute. Again, you must + * not modify the skb in any other way. + * + * When done filling the skb, call cfg80211_testmode_event() with the + * skb to send the event. + */ +struct sk_buff *cfg80211_testmode_alloc_event_skb(struct wiphy *wiphy, + int approxlen, gfp_t gfp); + +/** + * cfg80211_testmode_event - send the event + * @skb: The skb, must have been allocated with + * cfg80211_testmode_alloc_event_skb() + * @gfp: allocation flags + * + * This function sends the given @skb, which must have been allocated + * by cfg80211_testmode_alloc_event_skb(), as an event. It always + * consumes it. + */ +void cfg80211_testmode_event(struct sk_buff *skb, gfp_t gfp); + +#define CFG80211_TESTMODE_CMD(cmd) .testmode_cmd = (cmd), +#else +#define CFG80211_TESTMODE_CMD(cmd) +#endif + #endif /* __NET_CFG80211_H */ diff --git a/include/net/mac80211.h b/include/net/mac80211.h index fe80771d95f1..ce7cb1b5d453 100644 --- a/include/net/mac80211.h +++ b/include/net/mac80211.h @@ -1416,6 +1416,8 @@ enum ieee80211_ampdu_mlme_action { * @rfkill_poll: Poll rfkill hardware state. If you need this, you also * need to set wiphy->rfkill_poll to %true before registration, * and need to call wiphy_rfkill_set_hw_state() in the callback. + * + * @testmode_cmd: Implement a cfg80211 test mode command. */ struct ieee80211_ops { int (*tx)(struct ieee80211_hw *hw, struct sk_buff *skb); @@ -1466,6 +1468,9 @@ struct ieee80211_ops { struct ieee80211_sta *sta, u16 tid, u16 *ssn); void (*rfkill_poll)(struct ieee80211_hw *hw); +#ifdef CONFIG_NL80211_TESTMODE + int (*testmode_cmd)(struct ieee80211_hw *hw, void *data, int len); +#endif }; /** diff --git a/net/mac80211/cfg.c b/net/mac80211/cfg.c index eb93eb6a9cc7..c34c1a41019a 100644 --- a/net/mac80211/cfg.c +++ b/net/mac80211/cfg.c @@ -1376,6 +1376,18 @@ static void ieee80211_rfkill_poll(struct wiphy *wiphy) drv_rfkill_poll(local); } +#ifdef CONFIG_NL80211_TESTMODE +int ieee80211_testmode_cmd(struct wiphy *wiphy, void *data, int len) +{ + struct ieee80211_local *local = wiphy_priv(wiphy); + + if (!local->ops->testmode_cmd) + return -EOPNOTSUPP; + + return local->ops->testmode_cmd(&local->hw, data, len); +} +#endif + struct cfg80211_ops mac80211_config_ops = { .add_virtual_intf = ieee80211_add_iface, .del_virtual_intf = ieee80211_del_iface, @@ -1418,4 +1430,5 @@ struct cfg80211_ops mac80211_config_ops = { .set_tx_power = ieee80211_set_tx_power, .get_tx_power = ieee80211_get_tx_power, .rfkill_poll = ieee80211_rfkill_poll, + CFG80211_TESTMODE_CMD(ieee80211_testmode_cmd) }; diff --git a/net/wireless/Kconfig b/net/wireless/Kconfig index ec64571c4c23..040263118a20 100644 --- a/net/wireless/Kconfig +++ b/net/wireless/Kconfig @@ -2,6 +2,21 @@ config CFG80211 tristate "Improved wireless configuration API" depends on RFKILL || !RFKILL +config NL80211_TESTMODE + bool "nl80211 testmode command" + depends on CFG80211 + help + The nl80211 testmode command helps implementing things like + factory calibration or validation tools for wireless chips. + + Select this option ONLY for kernels that are specifically + built for such purposes. + + Debugging tools that are supposed to end up in the hands of + users should better be implemented with debugfs. + + Say N. + config CFG80211_REG_DEBUG bool "cfg80211 regulatory debugging" depends on CFG80211 diff --git a/net/wireless/core.h b/net/wireless/core.h index bfa340c7abb5..bc084b68865c 100644 --- a/net/wireless/core.h +++ b/net/wireless/core.h @@ -58,6 +58,10 @@ struct cfg80211_registered_device { struct cfg80211_scan_request *scan_req; /* protected by RTNL */ unsigned long suspend_at; +#ifdef CONFIG_NL80211_TESTMODE + struct genl_info *testmode_info; +#endif + #ifdef CONFIG_CFG80211_DEBUGFS /* Debugfs entries */ struct wiphy_debugfsdentries { diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index 01523ba81baf..bb8de268a6bf 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -3416,6 +3416,128 @@ unlock_rtnl: return err; } +#ifdef CONFIG_NL80211_TESTMODE +static struct genl_multicast_group nl80211_testmode_mcgrp = { + .name = "testmode", +}; + +static int nl80211_testmode_do(struct sk_buff *skb, struct genl_info *info) +{ + struct cfg80211_registered_device *rdev; + int err; + + if (!info->attrs[NL80211_ATTR_TESTDATA]) + return -EINVAL; + + rtnl_lock(); + + rdev = cfg80211_get_dev_from_info(info); + if (IS_ERR(rdev)) { + err = PTR_ERR(rdev); + goto unlock_rtnl; + } + + err = -EOPNOTSUPP; + if (rdev->ops->testmode_cmd) { + rdev->testmode_info = info; + err = rdev->ops->testmode_cmd(&rdev->wiphy, + nla_data(info->attrs[NL80211_ATTR_TESTDATA]), + nla_len(info->attrs[NL80211_ATTR_TESTDATA])); + rdev->testmode_info = NULL; + } + + cfg80211_put_dev(rdev); + + unlock_rtnl: + rtnl_unlock(); + return err; +} + +static struct sk_buff * +__cfg80211_testmode_alloc_skb(struct cfg80211_registered_device *rdev, + int approxlen, u32 pid, u32 seq, gfp_t gfp) +{ + struct sk_buff *skb; + void *hdr; + struct nlattr *data; + + skb = nlmsg_new(approxlen + 100, gfp); + if (!skb) + return NULL; + + hdr = nl80211hdr_put(skb, pid, seq, 0, NL80211_CMD_TESTMODE); + if (!hdr) { + kfree_skb(skb); + return NULL; + } + + NLA_PUT_U32(skb, NL80211_ATTR_WIPHY, rdev->wiphy_idx); + data = nla_nest_start(skb, NL80211_ATTR_TESTDATA); + + ((void **)skb->cb)[0] = rdev; + ((void **)skb->cb)[1] = hdr; + ((void **)skb->cb)[2] = data; + + return skb; + + nla_put_failure: + kfree_skb(skb); + return NULL; +} + +struct sk_buff *cfg80211_testmode_alloc_reply_skb(struct wiphy *wiphy, + int approxlen) +{ + struct cfg80211_registered_device *rdev = wiphy_to_dev(wiphy); + + if (WARN_ON(!rdev->testmode_info)) + return NULL; + + return __cfg80211_testmode_alloc_skb(rdev, approxlen, + rdev->testmode_info->snd_pid, + rdev->testmode_info->snd_seq, + GFP_KERNEL); +} +EXPORT_SYMBOL(cfg80211_testmode_alloc_reply_skb); + +int cfg80211_testmode_reply(struct sk_buff *skb) +{ + struct cfg80211_registered_device *rdev = ((void **)skb->cb)[0]; + void *hdr = ((void **)skb->cb)[1]; + struct nlattr *data = ((void **)skb->cb)[2]; + + if (WARN_ON(!rdev->testmode_info)) { + kfree_skb(skb); + return -EINVAL; + } + + nla_nest_end(skb, data); + genlmsg_end(skb, hdr); + return genlmsg_reply(skb, rdev->testmode_info); +} +EXPORT_SYMBOL(cfg80211_testmode_reply); + +struct sk_buff *cfg80211_testmode_alloc_event_skb(struct wiphy *wiphy, + int approxlen, gfp_t gfp) +{ + struct cfg80211_registered_device *rdev = wiphy_to_dev(wiphy); + + return __cfg80211_testmode_alloc_skb(rdev, approxlen, 0, 0, gfp); +} +EXPORT_SYMBOL(cfg80211_testmode_alloc_event_skb); + +void cfg80211_testmode_event(struct sk_buff *skb, gfp_t gfp) +{ + void *hdr = ((void **)skb->cb)[1]; + struct nlattr *data = ((void **)skb->cb)[2]; + + nla_nest_end(skb, data); + genlmsg_end(skb, hdr); + genlmsg_multicast(skb, 0, nl80211_testmode_mcgrp.id, gfp); +} +EXPORT_SYMBOL(cfg80211_testmode_event); +#endif + static struct genl_ops nl80211_ops[] = { { .cmd = NL80211_CMD_GET_WIPHY, @@ -3629,6 +3751,14 @@ static struct genl_ops nl80211_ops[] = { .policy = nl80211_policy, .flags = GENL_ADMIN_PERM, }, +#ifdef CONFIG_NL80211_TESTMODE + { + .cmd = NL80211_CMD_TESTMODE, + .doit = nl80211_testmode_do, + .policy = nl80211_policy, + .flags = GENL_ADMIN_PERM, + }, +#endif }; static struct genl_multicast_group nl80211_mlme_mcgrp = { .name = "mlme", @@ -4102,6 +4232,12 @@ int nl80211_init(void) if (err) goto err_out; +#ifdef CONFIG_NL80211_TESTMODE + err = genl_register_mc_group(&nl80211_fam, &nl80211_testmode_mcgrp); + if (err) + goto err_out; +#endif + return 0; err_out: genl_unregister_family(&nl80211_fam); -- cgit v1.2.3 From 6a669e65c5ec393a650362874e13f7d3365a7827 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Wed, 1 Jul 2009 21:26:53 +0200 Subject: wireless: define AKM suites We'll need these values for some drivers using connect API and for wext compat code, so let's define them. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- include/linux/ieee80211.h | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'include') diff --git a/include/linux/ieee80211.h b/include/linux/ieee80211.h index a9173d5434d1..23343ab0bb03 100644 --- a/include/linux/ieee80211.h +++ b/include/linux/ieee80211.h @@ -1196,6 +1196,10 @@ enum ieee80211_sa_query_action { #define WLAN_CIPHER_SUITE_WEP104 0x000FAC05 #define WLAN_CIPHER_SUITE_AES_CMAC 0x000FAC06 +/* AKM suite selectors */ +#define WLAN_AKM_SUITE_8021X 0x000FAC01 +#define WLAN_AKM_SUITE_PSK 0x000FAC02 + #define WLAN_MAX_KEY_LEN 32 /** -- cgit v1.2.3 From b23aa676ab9d54469cda9f7151f51a2851c6f36e Mon Sep 17 00:00:00 2001 From: Samuel Ortiz Date: Wed, 1 Jul 2009 21:26:54 +0200 Subject: cfg80211: connect/disconnect API This patch introduces the cfg80211 connect/disconnect API. The goal here is to run the AUTH and ASSOC steps in one call. This is needed for some fullmac cards that run both steps directly from the target, after the host driver sends a connect command. Additionally, all the new crypto parameters for connect() are now also valid for associate() -- although associate requires the IEs to be used, the information can be useful for drivers and should be given. Signed-off-by: Samuel Ortiz Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- include/linux/nl80211.h | 80 +++++++++++ include/net/cfg80211.h | 135 +++++++++++++++++- net/mac80211/cfg.c | 2 +- net/wireless/Makefile | 2 +- net/wireless/core.c | 16 ++- net/wireless/core.h | 7 + net/wireless/nl80211.c | 368 +++++++++++++++++++++++++++++++++++++++++++++++- net/wireless/nl80211.h | 13 ++ net/wireless/sme.c | 224 +++++++++++++++++++++++++++++ 9 files changed, 829 insertions(+), 18 deletions(-) create mode 100644 net/wireless/sme.c (limited to 'include') diff --git a/include/linux/nl80211.h b/include/linux/nl80211.h index 651b18839088..b34c17f52f3e 100644 --- a/include/linux/nl80211.h +++ b/include/linux/nl80211.h @@ -246,6 +246,22 @@ * to identify the device, and the TESTDATA blob attribute to pass through * to the driver. * + * @NL80211_CMD_CONNECT: connection request and notification; this command + * requests to connect to a specified network but without separating + * auth and assoc steps. For this, you need to specify the SSID in a + * %NL80211_ATTR_SSID attribute, and can optionally specify the association + * IEs in %NL80211_ATTR_IE, %NL80211_ATTR_AUTH_TYPE, %NL80211_ATTR_MAC, + * %NL80211_ATTR_WIPHY_FREQ and %NL80211_ATTR_CONTROL_PORT. + * It is also sent as an event, with the BSSID and response IEs when the + * connection is established or failed to be established. This can be + * determined by the STATUS_CODE attribute. + * @NL80211_CMD_ROAM: request that the card roam (currently not implemented), + * sent as an event when the card/driver roamed by itself. + * @NL80211_CMD_DISCONNECT: drop a given connection; also used to notify + * userspace that a connection was dropped by the AP or due to other + * reasons, for this the %NL80211_ATTR_DISCONNECTED_BY_AP and + * %NL80211_ATTR_REASON_CODE attributes are used. + * * @NL80211_CMD_MAX: highest used command number * @__NL80211_CMD_AFTER_LAST: internal use */ @@ -316,6 +332,10 @@ enum nl80211_commands { NL80211_CMD_TESTMODE, + NL80211_CMD_CONNECT, + NL80211_CMD_ROAM, + NL80211_CMD_DISCONNECT, + /* add new commands above here */ /* used to define NL80211_CMD_MAX below */ @@ -520,6 +540,30 @@ enum nl80211_commands { * @NL80211_ATTR_TESTDATA: Testmode data blob, passed through to the driver. * We recommend using nested, driver-specific attributes within this. * + * @NL80211_ATTR_DISCONNECTED_BY_AP: A flag indicating that the DISCONNECT + * event was due to the AP disconnecting the station, and not due to + * a local disconnect request. + * @NL80211_ATTR_STATUS_CODE: StatusCode for the %NL80211_CMD_CONNECT + * event (u16) + * @NL80211_ATTR_PRIVACY: Flag attribute, used with connect(), indicating + * that protected APs should be used. + * + * @NL80211_ATTR_CIPHERS_PAIRWISE: Used with CONNECT and ASSOCIATE to + * indicate which unicast key ciphers will be used with the connection + * (an array of u32). + * @NL80211_ATTR_CIPHER_GROUP: Used with CONNECT and ASSOCIATE to indicate + * which group key cipher will be used with the connection (a u32). + * @NL80211_ATTR_WPA_VERSIONS: Used with CONNECT and ASSOCIATE to indicate + * which WPA version(s) the AP we want to associate with is using + * (a u32 with flags from &enum nl80211_wpa_versions). + * @NL80211_ATTR_AKM_SUITES: Used with CONNECT and ASSOCIATE to indicate + * which key management algorithm(s) to use (an array of u32). + * + * @NL80211_ATTR_REQ_IE: (Re)association request information elements as + * sent out by the card, for ROAM and successful CONNECT events. + * @NL80211_ATTR_RESP_IE: (Re)association response information elements as + * sent by peer, for ROAM and successful CONNECT events. + * * @NL80211_ATTR_MAX: highest attribute number currently defined * @__NL80211_ATTR_AFTER_LAST: internal use */ @@ -630,6 +674,19 @@ enum nl80211_attrs { NL80211_ATTR_TESTDATA, + NL80211_ATTR_PRIVACY, + + NL80211_ATTR_DISCONNECTED_BY_AP, + NL80211_ATTR_STATUS_CODE, + + NL80211_ATTR_CIPHER_SUITES_PAIRWISE, + NL80211_ATTR_CIPHER_SUITE_GROUP, + NL80211_ATTR_WPA_VERSIONS, + NL80211_ATTR_AKM_SUITES, + + NL80211_ATTR_REQ_IE, + NL80211_ATTR_RESP_IE, + /* add attributes here, update the policy in nl80211.c */ __NL80211_ATTR_AFTER_LAST, @@ -640,6 +697,7 @@ enum nl80211_attrs { * Allow user space programs to use #ifdef on new attributes by defining them * here */ +#define NL80211_CMD_CONNECT NL80211_CMD_CONNECT #define NL80211_ATTR_HT_CAPABILITY NL80211_ATTR_HT_CAPABILITY #define NL80211_ATTR_BSS_BASIC_RATES NL80211_ATTR_BSS_BASIC_RATES #define NL80211_ATTR_WIPHY_TXQ_PARAMS NL80211_ATTR_WIPHY_TXQ_PARAMS @@ -653,6 +711,10 @@ enum nl80211_attrs { #define NL80211_ATTR_SSID NL80211_ATTR_SSID #define NL80211_ATTR_AUTH_TYPE NL80211_ATTR_AUTH_TYPE #define NL80211_ATTR_REASON_CODE NL80211_ATTR_REASON_CODE +#define NL80211_ATTR_CIPHER_SUITES_PAIRWISE NL80211_ATTR_CIPHER_SUITES_PAIRWISE +#define NL80211_ATTR_CIPHER_SUITE_GROUP NL80211_ATTR_CIPHER_SUITE_GROUP +#define NL80211_ATTR_WPA_VERSIONS NL80211_ATTR_WPA_VERSIONS +#define NL80211_ATTR_AKM_SUITES NL80211_ATTR_AKM_SUITES #define NL80211_MAX_SUPP_RATES 32 #define NL80211_MAX_SUPP_REG_RULES 32 @@ -661,6 +723,9 @@ enum nl80211_attrs { #define NL80211_TKIP_DATA_OFFSET_RX_MIC_KEY 24 #define NL80211_HT_CAPABILITY_LEN 26 +#define NL80211_MAX_NR_CIPHER_SUITES 5 +#define NL80211_MAX_NR_AKM_SUITES 2 + /** * enum nl80211_iftype - (virtual) interface types * @@ -1205,12 +1270,22 @@ enum nl80211_bss { * @NL80211_AUTHTYPE_SHARED_KEY: Shared Key authentication (WEP only) * @NL80211_AUTHTYPE_FT: Fast BSS Transition (IEEE 802.11r) * @NL80211_AUTHTYPE_NETWORK_EAP: Network EAP (some Cisco APs and mainly LEAP) + * @__NL80211_AUTHTYPE_NUM: internal + * @NL80211_AUTHTYPE_MAX: maximum valid auth algorithm + * @NL80211_AUTHTYPE_AUTOMATIC: determine automatically (if necessary by + * trying multiple times); this is invalid in netlink -- leave out + * the attribute for this on CONNECT commands. */ enum nl80211_auth_type { NL80211_AUTHTYPE_OPEN_SYSTEM, NL80211_AUTHTYPE_SHARED_KEY, NL80211_AUTHTYPE_FT, NL80211_AUTHTYPE_NETWORK_EAP, + + /* keep last */ + __NL80211_AUTHTYPE_NUM, + NL80211_AUTHTYPE_MAX = __NL80211_AUTHTYPE_NUM - 1, + NL80211_AUTHTYPE_AUTOMATIC }; /** @@ -1235,4 +1310,9 @@ enum nl80211_mfp { NL80211_MFP_REQUIRED, }; +enum nl80211_wpa_versions { + NL80211_WPA_VERSION_1 = 1 << 0, + NL80211_WPA_VERSION_2 = 1 << 1, +}; + #endif /* __LINUX_NL80211_H */ diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index 885d4e5bc4b5..68e11321ed74 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -604,6 +604,30 @@ struct cfg80211_bss { u8 priv[0] __attribute__((__aligned__(sizeof(void *)))); }; +/** + * struct cfg80211_crypto_settings - Crypto settings + * @wpa_versions: indicates which, if any, WPA versions are enabled + * (from enum nl80211_wpa_versions) + * @cipher_group: group key cipher suite (or 0 if unset) + * @n_ciphers_pairwise: number of AP supported unicast ciphers + * @ciphers_pairwise: unicast key cipher suites + * @n_akm_suites: number of AKM suites + * @akm_suites: AKM suites + * @control_port: Whether user space controls IEEE 802.1X port, i.e., + * sets/clears %NL80211_STA_FLAG_AUTHORIZED. If true, the driver is + * required to assume that the port is unauthorized until authorized by + * user space. Otherwise, port is marked authorized by default. + */ +struct cfg80211_crypto_settings { + u32 wpa_versions; + u32 cipher_group; + int n_ciphers_pairwise; + u32 ciphers_pairwise[NL80211_MAX_NR_CIPHER_SUITES]; + int n_akm_suites; + u32 akm_suites[NL80211_MAX_NR_AKM_SUITES]; + bool control_port; +}; + /** * struct cfg80211_auth_request - Authentication request data * @@ -658,10 +682,7 @@ struct cfg80211_auth_request { * @ie: Extra IEs to add to (Re)Association Request frame or %NULL * @ie_len: Length of ie buffer in octets * @use_mfp: Use management frame protection (IEEE 802.11w) in this association - * @control_port: Whether user space controls IEEE 802.1X port, i.e., - * sets/clears %NL80211_STA_FLAG_AUTHORIZED. If true, the driver is - * required to assume that the port is unauthorized until authorized by - * user space. Otherwise, port is marked authorized by default. + * @crypto: crypto settings */ struct cfg80211_assoc_request { struct ieee80211_channel *chan; @@ -671,7 +692,7 @@ struct cfg80211_assoc_request { const u8 *ie; size_t ie_len; bool use_mfp; - bool control_port; + struct cfg80211_crypto_settings crypto; }; /** @@ -737,6 +758,36 @@ struct cfg80211_ibss_params { bool channel_fixed; }; +/** + * struct cfg80211_connect_params - Connection parameters + * + * This structure provides information needed to complete IEEE 802.11 + * authentication and association. + * + * @channel: The channel to use or %NULL if not specified (auto-select based + * on scan results) + * @bssid: The AP BSSID or %NULL if not specified (auto-select based on scan + * results) + * @ssid: SSID + * @ssid_len: Length of ssid in octets + * @auth_type: Authentication type (algorithm) + * @assoc_ie: IEs for association request + * @assoc_ie_len: Length of assoc_ie in octets + * @privacy: indicates whether privacy-enabled APs should be used + * @crypto: crypto settings + */ +struct cfg80211_connect_params { + struct ieee80211_channel *channel; + u8 *bssid; + u8 *ssid; + size_t ssid_len; + enum nl80211_auth_type auth_type; + u8 *ie; + size_t ie_len; + bool privacy; + struct cfg80211_crypto_settings crypto; +}; + /** * enum wiphy_params_flags - set_wiphy_params bitfield values * WIPHY_PARAM_RETRY_SHORT: wiphy->retry_short has changed @@ -841,6 +892,12 @@ enum tx_power_setting { * @deauth: Request to deauthenticate from the specified peer * @disassoc: Request to disassociate from the specified peer * + * @connect: Connect to the ESS with the specified parameters. When connected, + * call cfg80211_connect_result() with status code %WLAN_STATUS_SUCCESS. + * If the connection fails for some reason, call cfg80211_connect_result() + * with the status from the AP. + * @disconnect: Disconnect from the BSS/ESS. + * * @join_ibss: Join the specified IBSS (or create if necessary). Once done, call * cfg80211_ibss_joined(), also call that function when changing BSSID due * to a merge. @@ -946,6 +1003,11 @@ struct cfg80211_ops { int (*disassoc)(struct wiphy *wiphy, struct net_device *dev, struct cfg80211_disassoc_request *req); + int (*connect)(struct wiphy *wiphy, struct net_device *dev, + struct cfg80211_connect_params *sme); + int (*disconnect)(struct wiphy *wiphy, struct net_device *dev, + u16 reason_code); + int (*join_ibss)(struct wiphy *wiphy, struct net_device *dev, struct cfg80211_ibss_params *params); int (*leave_ibss)(struct wiphy *wiphy, struct net_device *dev); @@ -1174,10 +1236,15 @@ struct wireless_dev { struct list_head list; struct net_device *netdev; - /* currently used for IBSS - might be rearranged in the future */ + /* currently used for IBSS and SME - might be rearranged later */ struct cfg80211_bss *current_bss; u8 ssid[IEEE80211_MAX_SSID_LEN]; u8 ssid_len; + enum { + CFG80211_SME_IDLE, + CFG80211_SME_CONNECTING, /* ->connect called */ + CFG80211_SME_CONNECTED, + } sme_state; #ifdef CONFIG_WIRELESS_EXT /* wext data */ @@ -1788,4 +1855,60 @@ void cfg80211_testmode_event(struct sk_buff *skb, gfp_t gfp); #define CFG80211_TESTMODE_CMD(cmd) #endif +/** + * cfg80211_connect_result - notify cfg80211 of connection result + * + * @dev: network device + * @bssid: the BSSID of the AP + * @req_ie: association request IEs (maybe be %NULL) + * @req_ie_len: association request IEs length + * @resp_ie: association response IEs (may be %NULL) + * @resp_ie_len: assoc response IEs length + * @status: status code, 0 for successful connection, use + * %WLAN_STATUS_UNSPECIFIED_FAILURE if your device cannot give you + * the real status code for failures. + * @gfp: allocation flags + * + * It should be called by the underlying driver whenever connect() has + * succeeded. + */ +void cfg80211_connect_result(struct net_device *dev, const u8 *bssid, + const u8 *req_ie, size_t req_ie_len, + const u8 *resp_ie, size_t resp_ie_len, + u16 status, gfp_t gfp); + +/** + * cfg80211_roamed - notify cfg80211 of roaming + * + * @dev: network device + * @bssid: the BSSID of the new AP + * @req_ie: association request IEs (maybe be %NULL) + * @req_ie_len: association request IEs length + * @resp_ie: association response IEs (may be %NULL) + * @resp_ie_len: assoc response IEs length + * @gfp: allocation flags + * + * It should be called by the underlying driver whenever it roamed + * from one AP to another while connected. + */ +void cfg80211_roamed(struct net_device *dev, const u8 *bssid, + const u8 *req_ie, size_t req_ie_len, + const u8 *resp_ie, size_t resp_ie_len, gfp_t gfp); + +/** + * cfg80211_disconnected - notify cfg80211 that connection was dropped + * + * @dev: network device + * @ie: information elements of the deauth/disassoc frame (may be %NULL) + * @ie_len: length of IEs + * @reason: reason code for the disconnection, set it to 0 if unknown + * @gfp: allocation flags + * + * After it calls this function, the driver should enter an idle state + * and not try to connect to any AP any more. + */ +void cfg80211_disconnected(struct net_device *dev, u16 reason, + u8 *ie, size_t ie_len, gfp_t gfp); + + #endif /* __NET_CFG80211_H */ diff --git a/net/mac80211/cfg.c b/net/mac80211/cfg.c index c34c1a41019a..03de4024597a 100644 --- a/net/mac80211/cfg.c +++ b/net/mac80211/cfg.c @@ -1262,7 +1262,7 @@ static int ieee80211_assoc(struct wiphy *wiphy, struct net_device *dev, sdata->u.mgd.flags &= ~IEEE80211_STA_MFP_ENABLED; } - if (req->control_port) + if (req->crypto.control_port) sdata->u.mgd.flags |= IEEE80211_STA_CONTROL_PORT; else sdata->u.mgd.flags &= ~IEEE80211_STA_CONTROL_PORT; diff --git a/net/wireless/Makefile b/net/wireless/Makefile index f78c4832a9ca..750c08e31b10 100644 --- a/net/wireless/Makefile +++ b/net/wireless/Makefile @@ -5,7 +5,7 @@ obj-$(CONFIG_LIB80211_CRYPT_WEP) += lib80211_crypt_wep.o obj-$(CONFIG_LIB80211_CRYPT_CCMP) += lib80211_crypt_ccmp.o obj-$(CONFIG_LIB80211_CRYPT_TKIP) += lib80211_crypt_tkip.o -cfg80211-y += core.o sysfs.o radiotap.o util.o reg.o scan.o nl80211.o mlme.o ibss.o +cfg80211-y += core.o sysfs.o radiotap.o util.o reg.o scan.o nl80211.o mlme.o ibss.o sme.o cfg80211-$(CONFIG_CFG80211_DEBUGFS) += debugfs.o cfg80211-$(CONFIG_WIRELESS_EXT) += wext-compat.o diff --git a/net/wireless/core.c b/net/wireless/core.c index d41b7412b212..314e00f70e3b 100644 --- a/net/wireless/core.c +++ b/net/wireless/core.c @@ -546,6 +546,7 @@ static int cfg80211_netdev_notifier_call(struct notifier_block * nb, "symlink to netdev!\n"); } wdev->netdev = dev; + wdev->sme_state = CFG80211_SME_IDLE; #ifdef CONFIG_WIRELESS_EXT wdev->wext.default_key = -1; wdev->wext.default_mgmt_key = -1; @@ -553,11 +554,20 @@ static int cfg80211_netdev_notifier_call(struct notifier_block * nb, mutex_unlock(&rdev->devlist_mtx); break; case NETDEV_GOING_DOWN: - if (wdev->iftype != NL80211_IFTYPE_ADHOC) - break; if (!wdev->ssid_len) break; - cfg80211_leave_ibss(rdev, dev, true); + + switch (wdev->iftype) { + case NL80211_IFTYPE_ADHOC: + cfg80211_leave_ibss(rdev, dev, true); + break; + case NL80211_IFTYPE_STATION: + cfg80211_disconnect(rdev, dev, + WLAN_REASON_DEAUTH_LEAVING); + break; + default: + break; + } break; case NETDEV_UP: #ifdef CONFIG_WIRELESS_EXT diff --git a/net/wireless/core.h b/net/wireless/core.h index bc084b68865c..f93f96f85d2b 100644 --- a/net/wireless/core.h +++ b/net/wireless/core.h @@ -174,6 +174,13 @@ void cfg80211_clear_ibss(struct net_device *dev, bool nowext); int cfg80211_leave_ibss(struct cfg80211_registered_device *rdev, struct net_device *dev, bool nowext); +/* SME */ +int cfg80211_connect(struct cfg80211_registered_device *rdev, + struct net_device *dev, + struct cfg80211_connect_params *connect); +int cfg80211_disconnect(struct cfg80211_registered_device *rdev, + struct net_device *dev, u16 reason); + /* internal helpers */ int cfg80211_validate_key_settings(struct key_params *params, int key_idx, const u8 *mac_addr); diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index bb8de268a6bf..89dd3793e03c 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -128,6 +128,9 @@ static struct nla_policy nl80211_policy[NL80211_ATTR_MAX+1] __read_mostly = { .len = sizeof(struct nl80211_sta_flag_update), }, [NL80211_ATTR_CONTROL_PORT] = { .type = NLA_FLAG }, + [NL80211_ATTR_PRIVACY] = { .type = NLA_FLAG }, + [NL80211_ATTR_CIPHER_SUITE_GROUP] = { .type = NLA_U32 }, + [NL80211_ATTR_WPA_VERSIONS] = { .type = NLA_U32 }, }; /* IE validation */ @@ -347,6 +350,17 @@ static int nl80211_send_wiphy(struct sk_buff *msg, u32 pid, u32 seq, int flags, CMD(join_ibss, JOIN_IBSS); #undef CMD + + if (dev->ops->connect) { + i++; + NLA_PUT_U32(msg, i, NL80211_CMD_CONNECT); + } + + if (dev->ops->disconnect) { + i++; + NLA_PUT_U32(msg, i, NL80211_CMD_DISCONNECT); + } + nla_nest_end(msg, nl_cmds); return genlmsg_end(msg, hdr); @@ -3001,12 +3015,31 @@ static int nl80211_dump_scan(struct sk_buff *skb, static bool nl80211_valid_auth_type(enum nl80211_auth_type auth_type) { - return auth_type == NL80211_AUTHTYPE_OPEN_SYSTEM || - auth_type == NL80211_AUTHTYPE_SHARED_KEY || - auth_type == NL80211_AUTHTYPE_FT || - auth_type == NL80211_AUTHTYPE_NETWORK_EAP; + return auth_type <= NL80211_AUTHTYPE_MAX; +} + +static bool nl80211_valid_wpa_versions(u32 wpa_versions) +{ + return !(wpa_versions & ~(NL80211_WPA_VERSION_1 | + NL80211_WPA_VERSION_2)); +} + +static bool nl80211_valid_akm_suite(u32 akm) +{ + return akm == WLAN_AKM_SUITE_8021X || + akm == WLAN_AKM_SUITE_PSK; +} + +static bool nl80211_valid_cipher_suite(u32 cipher) +{ + return cipher == WLAN_CIPHER_SUITE_WEP40 || + cipher == WLAN_CIPHER_SUITE_WEP104 || + cipher == WLAN_CIPHER_SUITE_TKIP || + cipher == WLAN_CIPHER_SUITE_CCMP || + cipher == WLAN_CIPHER_SUITE_AES_CMAC; } + static int nl80211_authenticate(struct sk_buff *skb, struct genl_info *info) { struct cfg80211_registered_device *drv; @@ -3086,6 +3119,68 @@ unlock_rtnl: return err; } +static int nl80211_crypto_settings(struct genl_info *info, + struct cfg80211_crypto_settings *settings) +{ + settings->control_port = info->attrs[NL80211_ATTR_CONTROL_PORT]; + + if (info->attrs[NL80211_ATTR_CIPHER_SUITES_PAIRWISE]) { + void *data; + int len, i; + + data = nla_data(info->attrs[NL80211_ATTR_CIPHER_SUITES_PAIRWISE]); + len = nla_len(info->attrs[NL80211_ATTR_CIPHER_SUITES_PAIRWISE]); + settings->n_ciphers_pairwise = len / sizeof(u32); + + if (len % sizeof(u32)) + return -EINVAL; + + if (settings->n_ciphers_pairwise > NL80211_MAX_NR_CIPHER_SUITES) + return -EINVAL; + + memcpy(settings->ciphers_pairwise, data, len); + + for (i = 0; i < settings->n_ciphers_pairwise; i++) + if (!nl80211_valid_cipher_suite( + settings->ciphers_pairwise[i])) + return -EINVAL; + } + + if (info->attrs[NL80211_ATTR_CIPHER_SUITE_GROUP]) { + settings->cipher_group = + nla_get_u32(info->attrs[NL80211_ATTR_CIPHER_SUITE_GROUP]); + if (!nl80211_valid_cipher_suite(settings->cipher_group)) + return -EINVAL; + } + + if (info->attrs[NL80211_ATTR_WPA_VERSIONS]) { + settings->wpa_versions = + nla_get_u32(info->attrs[NL80211_ATTR_WPA_VERSIONS]); + if (!nl80211_valid_wpa_versions(settings->wpa_versions)) + return -EINVAL; + } + + if (info->attrs[NL80211_ATTR_AKM_SUITES]) { + void *data; + int len, i; + + data = nla_data(info->attrs[NL80211_ATTR_AKM_SUITES]); + len = nla_len(info->attrs[NL80211_ATTR_AKM_SUITES]); + settings->n_akm_suites = len / sizeof(u32); + + if (len % sizeof(u32)) + return -EINVAL; + + memcpy(settings->akm_suites, data, len); + + for (i = 0; i < settings->n_ciphers_pairwise; i++) + if (!nl80211_valid_akm_suite(settings->akm_suites[i])) + return -EINVAL; + } + + return 0; +} + static int nl80211_associate(struct sk_buff *skb, struct genl_info *info) { struct cfg80211_registered_device *drv; @@ -3156,9 +3251,9 @@ static int nl80211_associate(struct sk_buff *skb, struct genl_info *info) } } - req.control_port = info->attrs[NL80211_ATTR_CONTROL_PORT]; - - err = drv->ops->assoc(&drv->wiphy, dev, &req); + err = nl80211_crypto_settings(info, &req.crypto); + if (!err) + err = drv->ops->assoc(&drv->wiphy, dev, &req); out: cfg80211_put_dev(drv); @@ -3538,6 +3633,130 @@ void cfg80211_testmode_event(struct sk_buff *skb, gfp_t gfp) EXPORT_SYMBOL(cfg80211_testmode_event); #endif +static int nl80211_connect(struct sk_buff *skb, struct genl_info *info) +{ + struct cfg80211_registered_device *drv; + struct net_device *dev; + struct cfg80211_connect_params connect; + struct wiphy *wiphy; + int err; + + memset(&connect, 0, sizeof(connect)); + + if (!is_valid_ie_attr(info->attrs[NL80211_ATTR_IE])) + return -EINVAL; + + if (!info->attrs[NL80211_ATTR_SSID] || + !nla_len(info->attrs[NL80211_ATTR_SSID])) + return -EINVAL; + + if (info->attrs[NL80211_ATTR_AUTH_TYPE]) { + connect.auth_type = + nla_get_u32(info->attrs[NL80211_ATTR_AUTH_TYPE]); + if (!nl80211_valid_auth_type(connect.auth_type)) + return -EINVAL; + } else + connect.auth_type = NL80211_AUTHTYPE_AUTOMATIC; + + connect.privacy = info->attrs[NL80211_ATTR_PRIVACY]; + + err = nl80211_crypto_settings(info, &connect.crypto); + if (err) + return err; + rtnl_lock(); + + err = get_drv_dev_by_info_ifindex(info->attrs, &drv, &dev); + if (err) + goto unlock_rtnl; + + if (dev->ieee80211_ptr->iftype != NL80211_IFTYPE_STATION) { + err = -EOPNOTSUPP; + goto out; + } + + if (!netif_running(dev)) { + err = -ENETDOWN; + goto out; + } + + wiphy = &drv->wiphy; + + connect.bssid = NULL; + connect.channel = NULL; + connect.auth_type = NL80211_AUTHTYPE_OPEN_SYSTEM; + + if (info->attrs[NL80211_ATTR_MAC]) + connect.bssid = nla_data(info->attrs[NL80211_ATTR_MAC]); + connect.ssid = nla_data(info->attrs[NL80211_ATTR_SSID]); + connect.ssid_len = nla_len(info->attrs[NL80211_ATTR_SSID]); + + if (info->attrs[NL80211_ATTR_IE]) { + connect.ie = nla_data(info->attrs[NL80211_ATTR_IE]); + connect.ie_len = nla_len(info->attrs[NL80211_ATTR_IE]); + } + + if (info->attrs[NL80211_ATTR_WIPHY_FREQ]) { + connect.channel = + ieee80211_get_channel(wiphy, + nla_get_u32(info->attrs[NL80211_ATTR_WIPHY_FREQ])); + if (!connect.channel || + connect.channel->flags & IEEE80211_CHAN_DISABLED) { + err = -EINVAL; + goto out; + } + } + + err = cfg80211_connect(drv, dev, &connect); + +out: + cfg80211_put_dev(drv); + dev_put(dev); +unlock_rtnl: + rtnl_unlock(); + return err; +} + +static int nl80211_disconnect(struct sk_buff *skb, struct genl_info *info) +{ + struct cfg80211_registered_device *drv; + struct net_device *dev; + int err; + u16 reason; + + if (!info->attrs[NL80211_ATTR_REASON_CODE]) + reason = WLAN_REASON_DEAUTH_LEAVING; + else + reason = nla_get_u16(info->attrs[NL80211_ATTR_REASON_CODE]); + + if (reason == 0) + return -EINVAL; + + rtnl_lock(); + + err = get_drv_dev_by_info_ifindex(info->attrs, &drv, &dev); + if (err) + goto unlock_rtnl; + + if (dev->ieee80211_ptr->iftype != NL80211_IFTYPE_STATION) { + err = -EOPNOTSUPP; + goto out; + } + + if (!netif_running(dev)) { + err = -ENETDOWN; + goto out; + } + + err = cfg80211_disconnect(drv, dev, reason); + +out: + cfg80211_put_dev(drv); + dev_put(dev); +unlock_rtnl: + rtnl_unlock(); + return err; +} + static struct genl_ops nl80211_ops[] = { { .cmd = NL80211_CMD_GET_WIPHY, @@ -3759,6 +3978,18 @@ static struct genl_ops nl80211_ops[] = { .flags = GENL_ADMIN_PERM, }, #endif + { + .cmd = NL80211_CMD_CONNECT, + .doit = nl80211_connect, + .policy = nl80211_policy, + .flags = GENL_ADMIN_PERM, + }, + { + .cmd = NL80211_CMD_DISCONNECT, + .doit = nl80211_disconnect, + .policy = nl80211_policy, + .flags = GENL_ADMIN_PERM, + }, }; static struct genl_multicast_group nl80211_mlme_mcgrp = { .name = "mlme", @@ -4077,6 +4308,129 @@ void nl80211_send_assoc_timeout(struct cfg80211_registered_device *rdev, addr, gfp); } +void nl80211_send_connect_result(struct cfg80211_registered_device *rdev, + struct net_device *netdev, const u8 *bssid, + const u8 *req_ie, size_t req_ie_len, + const u8 *resp_ie, size_t resp_ie_len, + u16 status, gfp_t gfp) +{ + struct sk_buff *msg; + void *hdr; + + msg = nlmsg_new(NLMSG_GOODSIZE, gfp); + if (!msg) + return; + + hdr = nl80211hdr_put(msg, 0, 0, 0, NL80211_CMD_CONNECT); + if (!hdr) { + nlmsg_free(msg); + return; + } + + NLA_PUT_U32(msg, NL80211_ATTR_WIPHY, rdev->wiphy_idx); + NLA_PUT_U32(msg, NL80211_ATTR_IFINDEX, netdev->ifindex); + if (bssid) + NLA_PUT(msg, NL80211_ATTR_MAC, ETH_ALEN, bssid); + NLA_PUT_U16(msg, NL80211_ATTR_STATUS_CODE, status); + if (req_ie) + NLA_PUT(msg, NL80211_ATTR_REQ_IE, req_ie_len, req_ie); + if (resp_ie) + NLA_PUT(msg, NL80211_ATTR_RESP_IE, resp_ie_len, resp_ie); + + if (genlmsg_end(msg, hdr) < 0) { + nlmsg_free(msg); + return; + } + + genlmsg_multicast(msg, 0, nl80211_mlme_mcgrp.id, gfp); + return; + + nla_put_failure: + genlmsg_cancel(msg, hdr); + nlmsg_free(msg); + +} + +void nl80211_send_roamed(struct cfg80211_registered_device *rdev, + struct net_device *netdev, const u8 *bssid, + const u8 *req_ie, size_t req_ie_len, + const u8 *resp_ie, size_t resp_ie_len, gfp_t gfp) +{ + struct sk_buff *msg; + void *hdr; + + msg = nlmsg_new(NLMSG_GOODSIZE, gfp); + if (!msg) + return; + + hdr = nl80211hdr_put(msg, 0, 0, 0, NL80211_CMD_ROAM); + if (!hdr) { + nlmsg_free(msg); + return; + } + + NLA_PUT_U32(msg, NL80211_ATTR_WIPHY, rdev->wiphy_idx); + NLA_PUT_U32(msg, NL80211_ATTR_IFINDEX, netdev->ifindex); + NLA_PUT(msg, NL80211_ATTR_MAC, ETH_ALEN, bssid); + if (req_ie) + NLA_PUT(msg, NL80211_ATTR_REQ_IE, req_ie_len, req_ie); + if (resp_ie) + NLA_PUT(msg, NL80211_ATTR_RESP_IE, resp_ie_len, resp_ie); + + if (genlmsg_end(msg, hdr) < 0) { + nlmsg_free(msg); + return; + } + + genlmsg_multicast(msg, 0, nl80211_mlme_mcgrp.id, gfp); + return; + + nla_put_failure: + genlmsg_cancel(msg, hdr); + nlmsg_free(msg); + +} + +void nl80211_send_disconnected(struct cfg80211_registered_device *rdev, + struct net_device *netdev, u16 reason, + u8 *ie, size_t ie_len, bool from_ap, gfp_t gfp) +{ + struct sk_buff *msg; + void *hdr; + + msg = nlmsg_new(NLMSG_GOODSIZE, gfp); + if (!msg) + return; + + hdr = nl80211hdr_put(msg, 0, 0, 0, NL80211_CMD_DISCONNECT); + if (!hdr) { + nlmsg_free(msg); + return; + } + + NLA_PUT_U32(msg, NL80211_ATTR_WIPHY, rdev->wiphy_idx); + NLA_PUT_U32(msg, NL80211_ATTR_IFINDEX, netdev->ifindex); + if (from_ap && reason) + NLA_PUT_U16(msg, NL80211_ATTR_REASON_CODE, reason); + if (from_ap) + NLA_PUT_FLAG(msg, NL80211_ATTR_DISCONNECTED_BY_AP); + if (ie) + NLA_PUT(msg, NL80211_ATTR_IE, ie_len, ie); + + if (genlmsg_end(msg, hdr) < 0) { + nlmsg_free(msg); + return; + } + + genlmsg_multicast(msg, 0, nl80211_mlme_mcgrp.id, gfp); + return; + + nla_put_failure: + genlmsg_cancel(msg, hdr); + nlmsg_free(msg); + +} + void nl80211_send_ibss_bssid(struct cfg80211_registered_device *rdev, struct net_device *netdev, const u8 *bssid, gfp_t gfp) diff --git a/net/wireless/nl80211.h b/net/wireless/nl80211.h index 662c216e8d4f..cf3708b48c29 100644 --- a/net/wireless/nl80211.h +++ b/net/wireless/nl80211.h @@ -31,6 +31,19 @@ void nl80211_send_auth_timeout(struct cfg80211_registered_device *rdev, void nl80211_send_assoc_timeout(struct cfg80211_registered_device *rdev, struct net_device *netdev, const u8 *addr, gfp_t gfp); +void nl80211_send_connect_result(struct cfg80211_registered_device *rdev, + struct net_device *netdev, const u8 *bssid, + const u8 *req_ie, size_t req_ie_len, + const u8 *resp_ie, size_t resp_ie_len, + u16 status, gfp_t gfp); +void nl80211_send_roamed(struct cfg80211_registered_device *rdev, + struct net_device *netdev, const u8 *bssid, + const u8 *req_ie, size_t req_ie_len, + const u8 *resp_ie, size_t resp_ie_len, gfp_t gfp); +void nl80211_send_disconnected(struct cfg80211_registered_device *rdev, + struct net_device *netdev, u16 reason, + u8 *ie, size_t ie_len, bool from_ap, gfp_t gfp); + void nl80211_michael_mic_failure(struct cfg80211_registered_device *rdev, struct net_device *netdev, const u8 *addr, diff --git a/net/wireless/sme.c b/net/wireless/sme.c new file mode 100644 index 000000000000..fc117031d0bb --- /dev/null +++ b/net/wireless/sme.c @@ -0,0 +1,224 @@ +/* + * SME code for cfg80211's connect emulation. + * + * Copyright 2009 Johannes Berg + * Copyright (C) 2009 Intel Corporation. All rights reserved. + */ + +#include +#include +#include +#include +#include +#include "nl80211.h" + + +void cfg80211_connect_result(struct net_device *dev, const u8 *bssid, + const u8 *req_ie, size_t req_ie_len, + const u8 *resp_ie, size_t resp_ie_len, + u16 status, gfp_t gfp) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + struct cfg80211_bss *bss; +#ifdef CONFIG_WIRELESS_EXT + union iwreq_data wrqu; +#endif + + if (WARN_ON(wdev->iftype != NL80211_IFTYPE_STATION)) + return; + + if (WARN_ON(wdev->sme_state != CFG80211_SME_CONNECTING)) + return; + + if (wdev->current_bss) { + cfg80211_unhold_bss(wdev->current_bss); + cfg80211_put_bss(wdev->current_bss); + wdev->current_bss = NULL; + } + + if (status == WLAN_STATUS_SUCCESS) { + bss = cfg80211_get_bss(wdev->wiphy, NULL, bssid, + wdev->ssid, wdev->ssid_len, + WLAN_CAPABILITY_ESS, + WLAN_CAPABILITY_ESS); + + if (WARN_ON(!bss)) + return; + + cfg80211_hold_bss(bss); + wdev->current_bss = bss; + + wdev->sme_state = CFG80211_SME_CONNECTED; + } else { + wdev->sme_state = CFG80211_SME_IDLE; + } + + nl80211_send_connect_result(wiphy_to_dev(wdev->wiphy), dev, bssid, + req_ie, req_ie_len, resp_ie, resp_ie_len, + status, gfp); + +#ifdef CONFIG_WIRELESS_EXT + if (req_ie && status == WLAN_STATUS_SUCCESS) { + memset(&wrqu, 0, sizeof(wrqu)); + wrqu.data.length = req_ie_len; + wireless_send_event(dev, IWEVASSOCRESPIE, &wrqu, req_ie); + } + + if (resp_ie && status == WLAN_STATUS_SUCCESS) { + memset(&wrqu, 0, sizeof(wrqu)); + wrqu.data.length = resp_ie_len; + wireless_send_event(dev, IWEVASSOCRESPIE, &wrqu, resp_ie); + } + + memset(&wrqu, 0, sizeof(wrqu)); + wrqu.ap_addr.sa_family = ARPHRD_ETHER; + if (bssid) + memcpy(wrqu.ap_addr.sa_data, bssid, ETH_ALEN); + wireless_send_event(dev, SIOCGIWAP, &wrqu, NULL); +#endif +} +EXPORT_SYMBOL(cfg80211_connect_result); + +void cfg80211_roamed(struct net_device *dev, const u8 *bssid, + const u8 *req_ie, size_t req_ie_len, + const u8 *resp_ie, size_t resp_ie_len, gfp_t gfp) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + struct cfg80211_bss *bss; +#ifdef CONFIG_WIRELESS_EXT + union iwreq_data wrqu; +#endif + + if (WARN_ON(wdev->iftype != NL80211_IFTYPE_STATION)) + return; + + if (WARN_ON(wdev->sme_state != CFG80211_SME_CONNECTED)) + return; + + /* internal error -- how did we get to CONNECTED w/o BSS? */ + if (WARN_ON(!wdev->current_bss)) { + return; + } + + cfg80211_unhold_bss(wdev->current_bss); + cfg80211_put_bss(wdev->current_bss); + wdev->current_bss = NULL; + + bss = cfg80211_get_bss(wdev->wiphy, NULL, bssid, + wdev->ssid, wdev->ssid_len, + WLAN_CAPABILITY_ESS, WLAN_CAPABILITY_ESS); + + if (WARN_ON(!bss)) + return; + + cfg80211_hold_bss(bss); + wdev->current_bss = bss; + + nl80211_send_roamed(wiphy_to_dev(wdev->wiphy), dev, bssid, + req_ie, req_ie_len, resp_ie, resp_ie_len, gfp); + +#ifdef CONFIG_WIRELESS_EXT + if (req_ie) { + memset(&wrqu, 0, sizeof(wrqu)); + wrqu.data.length = req_ie_len; + wireless_send_event(dev, IWEVASSOCRESPIE, &wrqu, req_ie); + } + + if (resp_ie) { + memset(&wrqu, 0, sizeof(wrqu)); + wrqu.data.length = resp_ie_len; + wireless_send_event(dev, IWEVASSOCRESPIE, &wrqu, resp_ie); + } + + memset(&wrqu, 0, sizeof(wrqu)); + wrqu.ap_addr.sa_family = ARPHRD_ETHER; + memcpy(wrqu.ap_addr.sa_data, bssid, ETH_ALEN); + wireless_send_event(dev, SIOCGIWAP, &wrqu, NULL); +#endif +} +EXPORT_SYMBOL(cfg80211_roamed); + +static void __cfg80211_disconnected(struct net_device *dev, gfp_t gfp, + u8 *ie, size_t ie_len, u16 reason, + bool from_ap) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; +#ifdef CONFIG_WIRELESS_EXT + union iwreq_data wrqu; +#endif + + if (WARN_ON(wdev->iftype != NL80211_IFTYPE_STATION)) + return; + + if (WARN_ON(wdev->sme_state != CFG80211_SME_CONNECTED)) + return; + + if (wdev->current_bss) { + cfg80211_unhold_bss(wdev->current_bss); + cfg80211_put_bss(wdev->current_bss); + } + + wdev->current_bss = NULL; + wdev->sme_state = CFG80211_SME_IDLE; + + nl80211_send_disconnected(wiphy_to_dev(wdev->wiphy), dev, + reason, ie, ie_len, from_ap, gfp); + +#ifdef CONFIG_WIRELESS_EXT + memset(&wrqu, 0, sizeof(wrqu)); + wrqu.ap_addr.sa_family = ARPHRD_ETHER; + wireless_send_event(dev, SIOCGIWAP, &wrqu, NULL); +#endif +} + +void cfg80211_disconnected(struct net_device *dev, u16 reason, + u8 *ie, size_t ie_len, gfp_t gfp) +{ + __cfg80211_disconnected(dev, reason, ie, ie_len, true, gfp); +} +EXPORT_SYMBOL(cfg80211_disconnected); + +int cfg80211_connect(struct cfg80211_registered_device *rdev, + struct net_device *dev, + struct cfg80211_connect_params *connect) +{ + int err; + struct wireless_dev *wdev = dev->ieee80211_ptr; + + if (wdev->sme_state != CFG80211_SME_IDLE) + return -EALREADY; + + if (!rdev->ops->connect) { + return -EOPNOTSUPP; + } else { + wdev->sme_state = CFG80211_SME_CONNECTING; + err = rdev->ops->connect(&rdev->wiphy, dev, connect); + if (err) { + wdev->sme_state = CFG80211_SME_IDLE; + return err; + } + } + + memcpy(wdev->ssid, connect->ssid, connect->ssid_len); + wdev->ssid_len = connect->ssid_len; + + return 0; +} + +int cfg80211_disconnect(struct cfg80211_registered_device *rdev, + struct net_device *dev, u16 reason) +{ + int err; + + if (!rdev->ops->disconnect) { + return -EOPNOTSUPP; + } else { + err = rdev->ops->disconnect(&rdev->wiphy, dev, reason); + if (err) + return err; + } + + __cfg80211_disconnected(dev, 0, NULL, 0, false, GFP_KERNEL); + + return 0; +} -- cgit v1.2.3 From 6829c878ecd24ff0ae41b4668c7e9d0f11b66942 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Thu, 2 Jul 2009 09:13:27 +0200 Subject: cfg80211: emulate connect with auth/assoc This adds code to cfg80211 so that drivers (mac80211 right now) that don't implement connect but rather auth/assoc can still be used with the nl80211 connect command. This will also be necessary for the wext compat code. Signed-off-by: Samuel Ortiz Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- include/net/cfg80211.h | 6 +- net/wireless/core.c | 7 + net/wireless/core.h | 8 + net/wireless/mlme.c | 79 ++++++++-- net/wireless/nl80211.c | 4 +- net/wireless/scan.c | 7 + net/wireless/sme.c | 395 +++++++++++++++++++++++++++++++++++++++++++++++-- 7 files changed, 474 insertions(+), 32 deletions(-) (limited to 'include') diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index 68e11321ed74..24fab439d415 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -1209,6 +1209,9 @@ extern void wiphy_unregister(struct wiphy *wiphy); */ extern void wiphy_free(struct wiphy *wiphy); +/* internal struct */ +struct cfg80211_conn; + /** * struct wireless_dev - wireless per-netdev state * @@ -1242,9 +1245,10 @@ struct wireless_dev { u8 ssid_len; enum { CFG80211_SME_IDLE, - CFG80211_SME_CONNECTING, /* ->connect called */ + CFG80211_SME_CONNECTING, CFG80211_SME_CONNECTED, } sme_state; + struct cfg80211_conn *conn; #ifdef CONFIG_WIRELESS_EXT /* wext data */ diff --git a/net/wireless/core.c b/net/wireless/core.c index 314e00f70e3b..a0a679704612 100644 --- a/net/wireless/core.c +++ b/net/wireless/core.c @@ -321,6 +321,7 @@ struct wiphy *wiphy_new(const struct cfg80211_ops *ops, int sizeof_priv) } INIT_WORK(&drv->rfkill_sync, cfg80211_rfkill_sync_work); + INIT_WORK(&drv->conn_work, cfg80211_conn_work); /* * Initialize wiphy parameters to IEEE 802.11 MIB default values. @@ -481,6 +482,8 @@ void wiphy_unregister(struct wiphy *wiphy) /* unlock again before freeing */ mutex_unlock(&drv->mtx); + cancel_work_sync(&drv->conn_work); + cfg80211_debugfs_drv_del(drv); /* If this device got a regulatory hint tell core its @@ -569,6 +572,10 @@ static int cfg80211_netdev_notifier_call(struct notifier_block * nb, break; } break; + case NETDEV_DOWN: + kfree(wdev->conn); + wdev->conn = NULL; + break; case NETDEV_UP: #ifdef CONFIG_WIRELESS_EXT if (wdev->iftype != NL80211_IFTYPE_ADHOC) diff --git a/net/wireless/core.h b/net/wireless/core.h index f93f96f85d2b..2c0f64252f3d 100644 --- a/net/wireless/core.h +++ b/net/wireless/core.h @@ -62,6 +62,8 @@ struct cfg80211_registered_device { struct genl_info *testmode_info; #endif + struct work_struct conn_work; + #ifdef CONFIG_CFG80211_DEBUGFS /* Debugfs entries */ struct wiphy_debugfsdentries { @@ -181,8 +183,14 @@ int cfg80211_connect(struct cfg80211_registered_device *rdev, int cfg80211_disconnect(struct cfg80211_registered_device *rdev, struct net_device *dev, u16 reason); +void cfg80211_conn_work(struct work_struct *work); + /* internal helpers */ int cfg80211_validate_key_settings(struct key_params *params, int key_idx, const u8 *mac_addr); +void __cfg80211_disconnected(struct net_device *dev, gfp_t gfp, u8 *ie, + size_t ie_len, u16 reason, bool from_ap); +void cfg80211_sme_scan_done(struct net_device *dev); +void cfg80211_sme_rx_auth(struct net_device *dev, const u8 *buf, size_t len); #endif /* __NET_WIRELESS_CORE_H */ diff --git a/net/wireless/mlme.c b/net/wireless/mlme.c index c4e6d4b84a46..3427fe73d3c3 100644 --- a/net/wireless/mlme.c +++ b/net/wireless/mlme.c @@ -16,58 +16,105 @@ void cfg80211_send_rx_auth(struct net_device *dev, const u8 *buf, size_t len, gf { struct wiphy *wiphy = dev->ieee80211_ptr->wiphy; struct cfg80211_registered_device *rdev = wiphy_to_dev(wiphy); + nl80211_send_rx_auth(rdev, dev, buf, len, gfp); + cfg80211_sme_rx_auth(dev, buf, len); } EXPORT_SYMBOL(cfg80211_send_rx_auth); void cfg80211_send_rx_assoc(struct net_device *dev, const u8 *buf, size_t len, gfp_t gfp) { - struct wiphy *wiphy = dev->ieee80211_ptr->wiphy; + u16 status_code; + struct wireless_dev *wdev = dev->ieee80211_ptr; + struct wiphy *wiphy = wdev->wiphy; struct cfg80211_registered_device *rdev = wiphy_to_dev(wiphy); + struct ieee80211_mgmt *mgmt = (struct ieee80211_mgmt *)buf; + u8 *ie = mgmt->u.assoc_resp.variable; + int ieoffs = offsetof(struct ieee80211_mgmt, u.assoc_resp.variable); + + status_code = le16_to_cpu(mgmt->u.assoc_resp.status_code); + nl80211_send_rx_assoc(rdev, dev, buf, len, gfp); + + cfg80211_connect_result(dev, mgmt->bssid, NULL, 0, ie, len - ieoffs, + status_code, gfp); } EXPORT_SYMBOL(cfg80211_send_rx_assoc); void cfg80211_send_deauth(struct net_device *dev, const u8 *buf, size_t len, gfp_t gfp) { - struct wiphy *wiphy = dev->ieee80211_ptr->wiphy; + struct wireless_dev *wdev = dev->ieee80211_ptr; + struct wiphy *wiphy = wdev->wiphy; struct cfg80211_registered_device *rdev = wiphy_to_dev(wiphy); + struct ieee80211_mgmt *mgmt = (struct ieee80211_mgmt *)buf; + nl80211_send_deauth(rdev, dev, buf, len, gfp); + + if (wdev->sme_state == CFG80211_SME_CONNECTED) { + u16 reason_code; + bool from_ap; + + reason_code = le16_to_cpu(mgmt->u.deauth.reason_code); + + from_ap = memcmp(mgmt->da, dev->dev_addr, ETH_ALEN) == 0; + __cfg80211_disconnected(dev, gfp, NULL, 0, + reason_code, from_ap); + + wdev->sme_state = CFG80211_SME_IDLE; + } else if (wdev->sme_state == CFG80211_SME_CONNECTING) { + cfg80211_connect_result(dev, mgmt->bssid, NULL, 0, NULL, 0, + WLAN_STATUS_UNSPECIFIED_FAILURE, gfp); + } } EXPORT_SYMBOL(cfg80211_send_deauth); void cfg80211_send_disassoc(struct net_device *dev, const u8 *buf, size_t len, gfp_t gfp) { - struct wiphy *wiphy = dev->ieee80211_ptr->wiphy; + struct wireless_dev *wdev = dev->ieee80211_ptr; + struct wiphy *wiphy = wdev->wiphy; struct cfg80211_registered_device *rdev = wiphy_to_dev(wiphy); + struct ieee80211_mgmt *mgmt = (struct ieee80211_mgmt *)buf; + nl80211_send_disassoc(rdev, dev, buf, len, gfp); -} -EXPORT_SYMBOL(cfg80211_send_disassoc); -static void cfg80211_wext_disconnected(struct net_device *dev) -{ -#ifdef CONFIG_WIRELESS_EXT - union iwreq_data wrqu; - memset(&wrqu, 0, sizeof(wrqu)); - wireless_send_event(dev, SIOCGIWAP, &wrqu, NULL); -#endif + if (wdev->sme_state == CFG80211_SME_CONNECTED) { + u16 reason_code; + bool from_ap; + + reason_code = le16_to_cpu(mgmt->u.disassoc.reason_code); + + from_ap = memcmp(mgmt->da, dev->dev_addr, ETH_ALEN) == 0; + __cfg80211_disconnected(dev, gfp, NULL, 0, + reason_code, from_ap); + + wdev->sme_state = CFG80211_SME_IDLE; + } } +EXPORT_SYMBOL(cfg80211_send_disassoc); void cfg80211_send_auth_timeout(struct net_device *dev, const u8 *addr, gfp_t gfp) { - struct wiphy *wiphy = dev->ieee80211_ptr->wiphy; + struct wireless_dev *wdev = dev->ieee80211_ptr; + struct wiphy *wiphy = wdev->wiphy; struct cfg80211_registered_device *rdev = wiphy_to_dev(wiphy); nl80211_send_auth_timeout(rdev, dev, addr, gfp); - cfg80211_wext_disconnected(dev); + if (wdev->sme_state == CFG80211_SME_CONNECTING) + cfg80211_connect_result(dev, addr, NULL, 0, NULL, 0, + WLAN_STATUS_UNSPECIFIED_FAILURE, gfp); + wdev->sme_state = CFG80211_SME_IDLE; } EXPORT_SYMBOL(cfg80211_send_auth_timeout); void cfg80211_send_assoc_timeout(struct net_device *dev, const u8 *addr, gfp_t gfp) { - struct wiphy *wiphy = dev->ieee80211_ptr->wiphy; + struct wireless_dev *wdev = dev->ieee80211_ptr; + struct wiphy *wiphy = wdev->wiphy; struct cfg80211_registered_device *rdev = wiphy_to_dev(wiphy); nl80211_send_assoc_timeout(rdev, dev, addr, gfp); - cfg80211_wext_disconnected(dev); + if (wdev->sme_state == CFG80211_SME_CONNECTING) + cfg80211_connect_result(dev, addr, NULL, 0, NULL, 0, + WLAN_STATUS_UNSPECIFIED_FAILURE, gfp); + wdev->sme_state = CFG80211_SME_IDLE; } EXPORT_SYMBOL(cfg80211_send_assoc_timeout); diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index 89dd3793e03c..89aa9e781d10 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -351,12 +351,12 @@ static int nl80211_send_wiphy(struct sk_buff *msg, u32 pid, u32 seq, int flags, #undef CMD - if (dev->ops->connect) { + if (dev->ops->connect || dev->ops->auth) { i++; NLA_PUT_U32(msg, i, NL80211_CMD_CONNECT); } - if (dev->ops->disconnect) { + if (dev->ops->disconnect || dev->ops->deauth) { i++; NLA_PUT_U32(msg, i, NL80211_CMD_DISCONNECT); } diff --git a/net/wireless/scan.c b/net/wireless/scan.c index 261a06386822..82b33e708488 100644 --- a/net/wireless/scan.c +++ b/net/wireless/scan.c @@ -30,6 +30,13 @@ void cfg80211_scan_done(struct cfg80211_scan_request *request, bool aborted) WARN_ON(request != wiphy_to_dev(request->wiphy)->scan_req); + /* + * This must be before sending the other events! + * Otherwise, wpa_supplicant gets completely confused with + * wext events. + */ + cfg80211_sme_scan_done(dev); + if (aborted) nl80211_send_scan_aborted(wiphy_to_dev(request->wiphy), dev); else diff --git a/net/wireless/sme.c b/net/wireless/sme.c index fc117031d0bb..3abb04729873 100644 --- a/net/wireless/sme.c +++ b/net/wireless/sme.c @@ -12,6 +12,266 @@ #include #include "nl80211.h" +struct cfg80211_conn { + struct cfg80211_connect_params params; + /* these are sub-states of the _CONNECTING sme_state */ + enum { + CFG80211_CONN_IDLE, + CFG80211_CONN_SCANNING, + CFG80211_CONN_SCAN_AGAIN, + CFG80211_CONN_AUTHENTICATE_NEXT, + CFG80211_CONN_AUTHENTICATING, + CFG80211_CONN_ASSOCIATE_NEXT, + CFG80211_CONN_ASSOCIATING, + } state; + u8 bssid[ETH_ALEN]; + u8 *ie; + size_t ie_len; + bool auto_auth; +}; + + +static int cfg80211_conn_scan(struct wireless_dev *wdev) +{ + struct cfg80211_registered_device *drv = wiphy_to_dev(wdev->wiphy); + struct cfg80211_scan_request *request; + int n_channels, err; + + ASSERT_RTNL(); + + if (drv->scan_req) + return -EBUSY; + + if (wdev->conn->params.channel) { + n_channels = 1; + } else { + enum ieee80211_band band; + n_channels = 0; + + for (band = 0; band < IEEE80211_NUM_BANDS; band++) { + if (!wdev->wiphy->bands[band]) + continue; + n_channels += wdev->wiphy->bands[band]->n_channels; + } + } + request = kzalloc(sizeof(*request) + sizeof(request->ssids[0]) + + sizeof(request->channels[0]) * n_channels, + GFP_KERNEL); + if (!request) + return -ENOMEM; + + request->channels = (void *)((char *)request + sizeof(*request)); + if (wdev->conn->params.channel) + request->channels[0] = wdev->conn->params.channel; + else { + int i = 0, j; + enum ieee80211_band band; + + for (band = 0; band < IEEE80211_NUM_BANDS; band++) { + if (!wdev->wiphy->bands[band]) + continue; + for (j = 0; j < wdev->wiphy->bands[band]->n_channels; + i++, j++) + request->channels[i] = + &wdev->wiphy->bands[band]->channels[j]; + } + } + request->n_channels = n_channels; + request->ssids = (void *)(request->channels + n_channels); + request->n_ssids = 1; + + memcpy(request->ssids[0].ssid, wdev->conn->params.ssid, + wdev->conn->params.ssid_len); + request->ssids[0].ssid_len = wdev->conn->params.ssid_len; + + request->ifidx = wdev->netdev->ifindex; + request->wiphy = &drv->wiphy; + + drv->scan_req = request; + + err = drv->ops->scan(wdev->wiphy, wdev->netdev, request); + if (!err) { + wdev->conn->state = CFG80211_CONN_SCANNING; + nl80211_send_scan_start(drv, wdev->netdev); + } else { + drv->scan_req = NULL; + kfree(request); + } + return err; +} + +static int cfg80211_conn_do_work(struct wireless_dev *wdev) +{ + struct cfg80211_registered_device *drv = wiphy_to_dev(wdev->wiphy); + union { + struct cfg80211_auth_request auth_req; + struct cfg80211_assoc_request assoc_req; + } u; + + memset(&u, 0, sizeof(u)); + + if (!wdev->conn) + return 0; + + switch (wdev->conn->state) { + case CFG80211_CONN_SCAN_AGAIN: + return cfg80211_conn_scan(wdev); + case CFG80211_CONN_AUTHENTICATE_NEXT: + u.auth_req.chan = wdev->conn->params.channel; + u.auth_req.peer_addr = wdev->conn->params.bssid; + u.auth_req.ssid = wdev->conn->params.ssid; + u.auth_req.ssid_len = wdev->conn->params.ssid_len; + u.auth_req.auth_type = wdev->conn->params.auth_type; + u.auth_req.ie = NULL; + u.auth_req.ie_len = 0; + wdev->conn->state = CFG80211_CONN_AUTHENTICATING; + BUG_ON(!drv->ops->auth); + return drv->ops->auth(wdev->wiphy, wdev->netdev, &u.auth_req); + case CFG80211_CONN_ASSOCIATE_NEXT: + u.assoc_req.chan = wdev->conn->params.channel; + u.assoc_req.peer_addr = wdev->conn->params.bssid; + u.assoc_req.ssid = wdev->conn->params.ssid; + u.assoc_req.ssid_len = wdev->conn->params.ssid_len; + u.assoc_req.ie = wdev->conn->params.ie; + u.assoc_req.ie_len = wdev->conn->params.ie_len; + u.assoc_req.use_mfp = false; + memcpy(&u.assoc_req.crypto, &wdev->conn->params.crypto, + sizeof(u.assoc_req.crypto)); + wdev->conn->state = CFG80211_CONN_ASSOCIATING; + BUG_ON(!drv->ops->assoc); + return drv->ops->assoc(wdev->wiphy, wdev->netdev, + &u.assoc_req); + default: + return 0; + } +} + +void cfg80211_conn_work(struct work_struct *work) +{ + struct cfg80211_registered_device *drv = + container_of(work, struct cfg80211_registered_device, conn_work); + struct wireless_dev *wdev; + + rtnl_lock(); + mutex_lock(&drv->devlist_mtx); + + list_for_each_entry(wdev, &drv->netdev_list, list) { + if (!netif_running(wdev->netdev)) + continue; + if (wdev->sme_state != CFG80211_SME_CONNECTING) + continue; + if (cfg80211_conn_do_work(wdev)) + cfg80211_connect_result(wdev->netdev, + wdev->conn->params.bssid, + NULL, 0, NULL, 0, + WLAN_STATUS_UNSPECIFIED_FAILURE, + GFP_ATOMIC); + } + + mutex_unlock(&drv->devlist_mtx); + rtnl_unlock(); +} + +static bool cfg80211_get_conn_bss(struct wireless_dev *wdev) +{ + struct cfg80211_registered_device *drv = wiphy_to_dev(wdev->wiphy); + struct cfg80211_bss *bss; + u16 capa = WLAN_CAPABILITY_ESS; + + if (wdev->conn->params.privacy) + capa |= WLAN_CAPABILITY_PRIVACY; + + bss = cfg80211_get_bss(wdev->wiphy, NULL, wdev->conn->params.bssid, + wdev->conn->params.ssid, + wdev->conn->params.ssid_len, + WLAN_CAPABILITY_ESS | WLAN_CAPABILITY_PRIVACY, + capa); + + if (!bss) + return false; + + memcpy(wdev->conn->bssid, bss->bssid, ETH_ALEN); + wdev->conn->params.bssid = wdev->conn->bssid; + wdev->conn->params.channel = bss->channel; + wdev->conn->state = CFG80211_CONN_AUTHENTICATE_NEXT; + schedule_work(&drv->conn_work); + + cfg80211_put_bss(bss); + return true; +} + +void cfg80211_sme_scan_done(struct net_device *dev) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + struct cfg80211_registered_device *drv = wiphy_to_dev(wdev->wiphy); + + if (wdev->sme_state != CFG80211_SME_CONNECTING) + return; + + if (WARN_ON(!wdev->conn)) + return; + + if (wdev->conn->state != CFG80211_CONN_SCANNING && + wdev->conn->state != CFG80211_CONN_SCAN_AGAIN) + return; + + if (!cfg80211_get_conn_bss(wdev)) { + /* not found */ + if (wdev->conn->state == CFG80211_CONN_SCAN_AGAIN) + schedule_work(&drv->conn_work); + else + cfg80211_connect_result(dev, wdev->conn->params.bssid, + NULL, 0, NULL, 0, + WLAN_STATUS_UNSPECIFIED_FAILURE, + GFP_ATOMIC); + return; + } +} + +void cfg80211_sme_rx_auth(struct net_device *dev, const u8 *buf, size_t len) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + struct wiphy *wiphy = wdev->wiphy; + struct cfg80211_registered_device *rdev = wiphy_to_dev(wiphy); + struct ieee80211_mgmt *mgmt = (struct ieee80211_mgmt *)buf; + u16 status_code = le16_to_cpu(mgmt->u.auth.status_code); + + /* should only RX auth frames when connecting */ + if (wdev->sme_state != CFG80211_SME_CONNECTING) + return; + + if (WARN_ON(!wdev->conn)) + return; + + if (status_code == WLAN_STATUS_NOT_SUPPORTED_AUTH_ALG && + wdev->conn->auto_auth && + wdev->conn->params.auth_type != NL80211_AUTHTYPE_NETWORK_EAP) { + /* select automatically between only open, shared, leap */ + switch (wdev->conn->params.auth_type) { + case NL80211_AUTHTYPE_OPEN_SYSTEM: + wdev->conn->params.auth_type = + NL80211_AUTHTYPE_SHARED_KEY; + break; + case NL80211_AUTHTYPE_SHARED_KEY: + wdev->conn->params.auth_type = + NL80211_AUTHTYPE_NETWORK_EAP; + break; + default: + /* huh? */ + wdev->conn->params.auth_type = + NL80211_AUTHTYPE_OPEN_SYSTEM; + break; + } + wdev->conn->state = CFG80211_CONN_AUTHENTICATE_NEXT; + schedule_work(&rdev->conn_work); + } else if (status_code != WLAN_STATUS_SUCCESS) + wdev->sme_state = CFG80211_SME_IDLE; + else if (wdev->sme_state == CFG80211_SME_CONNECTING && + wdev->conn->state == CFG80211_CONN_AUTHENTICATING) { + wdev->conn->state = CFG80211_CONN_ASSOCIATE_NEXT; + schedule_work(&rdev->conn_work); + } +} void cfg80211_connect_result(struct net_device *dev, const u8 *bssid, const u8 *req_ie, size_t req_ie_len, @@ -27,7 +287,7 @@ void cfg80211_connect_result(struct net_device *dev, const u8 *bssid, if (WARN_ON(wdev->iftype != NL80211_IFTYPE_STATION)) return; - if (WARN_ON(wdev->sme_state != CFG80211_SME_CONNECTING)) + if (wdev->sme_state != CFG80211_SME_CONNECTING) return; if (wdev->current_bss) { @@ -53,6 +313,9 @@ void cfg80211_connect_result(struct net_device *dev, const u8 *bssid, wdev->sme_state = CFG80211_SME_IDLE; } + if (wdev->conn) + wdev->conn->state = CFG80211_CONN_IDLE; + nl80211_send_connect_result(wiphy_to_dev(wdev->wiphy), dev, bssid, req_ie, req_ie_len, resp_ie, resp_ie_len, status, gfp); @@ -72,7 +335,7 @@ void cfg80211_connect_result(struct net_device *dev, const u8 *bssid, memset(&wrqu, 0, sizeof(wrqu)); wrqu.ap_addr.sa_family = ARPHRD_ETHER; - if (bssid) + if (bssid && status == WLAN_STATUS_SUCCESS) memcpy(wrqu.ap_addr.sa_data, bssid, ETH_ALEN); wireless_send_event(dev, SIOCGIWAP, &wrqu, NULL); #endif @@ -138,9 +401,8 @@ void cfg80211_roamed(struct net_device *dev, const u8 *bssid, } EXPORT_SYMBOL(cfg80211_roamed); -static void __cfg80211_disconnected(struct net_device *dev, gfp_t gfp, - u8 *ie, size_t ie_len, u16 reason, - bool from_ap) +void __cfg80211_disconnected(struct net_device *dev, gfp_t gfp, u8 *ie, + size_t ie_len, u16 reason, bool from_ap) { struct wireless_dev *wdev = dev->ieee80211_ptr; #ifdef CONFIG_WIRELESS_EXT @@ -161,6 +423,11 @@ static void __cfg80211_disconnected(struct net_device *dev, gfp_t gfp, wdev->current_bss = NULL; wdev->sme_state = CFG80211_SME_IDLE; + if (wdev->conn) { + kfree(wdev->conn->ie); + wdev->conn->ie = NULL; + } + nl80211_send_disconnected(wiphy_to_dev(wdev->wiphy), dev, reason, ie, ie_len, from_ap, gfp); @@ -174,7 +441,7 @@ static void __cfg80211_disconnected(struct net_device *dev, gfp_t gfp, void cfg80211_disconnected(struct net_device *dev, u16 reason, u8 *ie, size_t ie_len, gfp_t gfp) { - __cfg80211_disconnected(dev, reason, ie, ie_len, true, gfp); + __cfg80211_disconnected(dev, gfp, ie, ie_len, reason, true); } EXPORT_SYMBOL(cfg80211_disconnected); @@ -189,7 +456,74 @@ int cfg80211_connect(struct cfg80211_registered_device *rdev, return -EALREADY; if (!rdev->ops->connect) { - return -EOPNOTSUPP; + if (!rdev->ops->auth || !rdev->ops->assoc) + return -EOPNOTSUPP; + + if (!wdev->conn) { + wdev->conn = kzalloc(sizeof(*wdev->conn), GFP_KERNEL); + if (!wdev->conn) + return -ENOMEM; + } else + memset(wdev->conn, 0, sizeof(*wdev->conn)); + + /* + * Copy all parameters, and treat explicitly IEs, BSSID, SSID. + */ + memcpy(&wdev->conn->params, connect, sizeof(*connect)); + if (connect->bssid) { + wdev->conn->params.bssid = wdev->conn->bssid; + memcpy(wdev->conn->bssid, connect->bssid, ETH_ALEN); + } + + if (connect->ie) { + wdev->conn->ie = kmemdup(connect->ie, connect->ie_len, + GFP_KERNEL); + wdev->conn->params.ie = wdev->conn->ie; + if (!wdev->conn->ie) + return -ENOMEM; + } + + if (connect->auth_type == NL80211_AUTHTYPE_AUTOMATIC) { + wdev->conn->auto_auth = true; + /* start with open system ... should mostly work */ + wdev->conn->params.auth_type = + NL80211_AUTHTYPE_OPEN_SYSTEM; + } else { + wdev->conn->auto_auth = false; + } + + memcpy(wdev->ssid, connect->ssid, connect->ssid_len); + wdev->ssid_len = connect->ssid_len; + wdev->conn->params.ssid = wdev->ssid; + wdev->conn->params.ssid_len = connect->ssid_len; + + /* don't care about result -- but fill bssid & channel */ + if (!wdev->conn->params.bssid || !wdev->conn->params.channel) + cfg80211_get_conn_bss(wdev); + + wdev->sme_state = CFG80211_SME_CONNECTING; + + /* we're good if we have both BSSID and channel */ + if (wdev->conn->params.bssid && wdev->conn->params.channel) { + wdev->conn->state = CFG80211_CONN_AUTHENTICATE_NEXT; + err = cfg80211_conn_do_work(wdev); + } else { + /* otherwise we'll need to scan for the AP first */ + err = cfg80211_conn_scan(wdev); + /* + * If we can't scan right now, then we need to scan again + * after the current scan finished, since the parameters + * changed (unless we find a good AP anyway). + */ + if (err == -EBUSY) { + err = 0; + wdev->conn->state = CFG80211_CONN_SCAN_AGAIN; + } + } + if (err) + wdev->sme_state = CFG80211_SME_IDLE; + + return err; } else { wdev->sme_state = CFG80211_SME_CONNECTING; err = rdev->ops->connect(&rdev->wiphy, dev, connect); @@ -197,28 +531,63 @@ int cfg80211_connect(struct cfg80211_registered_device *rdev, wdev->sme_state = CFG80211_SME_IDLE; return err; } - } - memcpy(wdev->ssid, connect->ssid, connect->ssid_len); - wdev->ssid_len = connect->ssid_len; + memcpy(wdev->ssid, connect->ssid, connect->ssid_len); + wdev->ssid_len = connect->ssid_len; - return 0; + return 0; + } } int cfg80211_disconnect(struct cfg80211_registered_device *rdev, struct net_device *dev, u16 reason) { + struct wireless_dev *wdev = dev->ieee80211_ptr; int err; + if (wdev->sme_state == CFG80211_SME_IDLE) + return -EINVAL; + if (!rdev->ops->disconnect) { - return -EOPNOTSUPP; + struct cfg80211_deauth_request deauth; + u8 bssid[ETH_ALEN]; + + /* internal bug. */ + if (WARN_ON(!wdev->conn)) + return -EINVAL; + + if (wdev->sme_state == CFG80211_SME_CONNECTING && + (wdev->conn->state == CFG80211_CONN_SCANNING || + wdev->conn->state == CFG80211_CONN_SCAN_AGAIN)) { + wdev->sme_state = CFG80211_SME_IDLE; + return 0; + } + + if (!rdev->ops->deauth) + return -EOPNOTSUPP; + + memset(&deauth, 0, sizeof(deauth)); + + /* wdev->conn->params.bssid must be set if > SCANNING */ + memcpy(bssid, wdev->conn->params.bssid, ETH_ALEN); + deauth.peer_addr = bssid; + deauth.reason_code = reason; + + err = rdev->ops->deauth(&rdev->wiphy, dev, &deauth); + if (err) + return err; } else { err = rdev->ops->disconnect(&rdev->wiphy, dev, reason); if (err) return err; } - __cfg80211_disconnected(dev, 0, NULL, 0, false, GFP_KERNEL); + if (wdev->sme_state == CFG80211_SME_CONNECTED) + __cfg80211_disconnected(dev, GFP_KERNEL, NULL, 0, 0, false); + else if (wdev->sme_state == CFG80211_SME_CONNECTING) + cfg80211_connect_result(dev, NULL, NULL, 0, NULL, 0, + WLAN_STATUS_UNSPECIFIED_FAILURE, + GFP_KERNEL); return 0; } -- cgit v1.2.3 From f21293549f60f88c74fcb9944737f11048896dc4 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Wed, 1 Jul 2009 21:26:56 +0200 Subject: cfg80211: managed mode wext compatibility This adds code to make it possible to use the cfg80211 connect() API with wireless extensions, and because the previous patch added emulation of that API with auth() and assoc(), by extension also supports wext on that. At the same time, removes code from mac80211 for wext, but doesn't yet clean up mac80211's mlme code more. Signed-off-by: Samuel Ortiz Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- include/net/cfg80211.h | 36 ++++- net/mac80211/mlme.c | 69 ---------- net/mac80211/wext.c | 215 +++-------------------------- net/wireless/Makefile | 2 +- net/wireless/core.c | 23 +++- net/wireless/core.h | 3 +- net/wireless/nl80211.c | 2 +- net/wireless/sme.c | 55 +++++--- net/wireless/wext-compat.c | 229 +++++++++++++++++++++++++------ net/wireless/wext-sme.c | 329 +++++++++++++++++++++++++++++++++++++++++++++ 10 files changed, 625 insertions(+), 338 deletions(-) create mode 100644 net/wireless/wext-sme.c (limited to 'include') diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index 24fab439d415..07085216532d 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -1253,8 +1253,14 @@ struct wireless_dev { #ifdef CONFIG_WIRELESS_EXT /* wext data */ struct { - struct cfg80211_ibss_params ibss; + union { + struct cfg80211_ibss_params ibss; + struct cfg80211_connect_params connect; + }; + u8 *ie; + size_t ie_len; u8 bssid[ETH_ALEN]; + u8 ssid[IEEE80211_MAX_SSID_LEN]; s8 default_key, default_mgmt_key; } wext; #endif @@ -1535,6 +1541,34 @@ int cfg80211_ibss_wext_giwap(struct net_device *dev, struct iw_request_info *info, struct sockaddr *ap_addr, char *extra); +int cfg80211_mgd_wext_siwfreq(struct net_device *dev, + struct iw_request_info *info, + struct iw_freq *freq, char *extra); +int cfg80211_mgd_wext_giwfreq(struct net_device *dev, + struct iw_request_info *info, + struct iw_freq *freq, char *extra); +int cfg80211_mgd_wext_siwessid(struct net_device *dev, + struct iw_request_info *info, + struct iw_point *data, char *ssid); +int cfg80211_mgd_wext_giwessid(struct net_device *dev, + struct iw_request_info *info, + struct iw_point *data, char *ssid); +int cfg80211_mgd_wext_siwap(struct net_device *dev, + struct iw_request_info *info, + struct sockaddr *ap_addr, char *extra); +int cfg80211_mgd_wext_giwap(struct net_device *dev, + struct iw_request_info *info, + struct sockaddr *ap_addr, char *extra); +int cfg80211_wext_siwgenie(struct net_device *dev, + struct iw_request_info *info, + struct iw_point *data, char *extra); +int cfg80211_wext_siwauth(struct net_device *dev, + struct iw_request_info *info, + struct iw_param *data, char *extra); +int cfg80211_wext_giwauth(struct net_device *dev, + struct iw_request_info *info, + struct iw_param *data, char *extra); + struct ieee80211_channel *cfg80211_wext_freq(struct wiphy *wiphy, struct iw_freq *freq); diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index 58135a5096af..fbb93a70ddc7 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -870,70 +870,6 @@ static u32 ieee80211_handle_bss_capability(struct ieee80211_sub_if_data *sdata, return changed; } -static void ieee80211_sta_send_apinfo(struct ieee80211_sub_if_data *sdata) -{ - union iwreq_data wrqu; - - memset(&wrqu, 0, sizeof(wrqu)); - if (sdata->u.mgd.flags & IEEE80211_STA_ASSOCIATED) - memcpy(wrqu.ap_addr.sa_data, sdata->u.mgd.bssid, ETH_ALEN); - wrqu.ap_addr.sa_family = ARPHRD_ETHER; - wireless_send_event(sdata->dev, SIOCGIWAP, &wrqu, NULL); -} - -static void ieee80211_sta_send_associnfo(struct ieee80211_sub_if_data *sdata) -{ - struct ieee80211_if_managed *ifmgd = &sdata->u.mgd; - char *buf; - size_t len; - int i; - union iwreq_data wrqu; - - if (!ifmgd->assocreq_ies && !ifmgd->assocresp_ies) - return; - - buf = kmalloc(50 + 2 * (ifmgd->assocreq_ies_len + - ifmgd->assocresp_ies_len), GFP_KERNEL); - if (!buf) - return; - - len = sprintf(buf, "ASSOCINFO("); - if (ifmgd->assocreq_ies) { - len += sprintf(buf + len, "ReqIEs="); - for (i = 0; i < ifmgd->assocreq_ies_len; i++) { - len += sprintf(buf + len, "%02x", - ifmgd->assocreq_ies[i]); - } - } - if (ifmgd->assocresp_ies) { - if (ifmgd->assocreq_ies) - len += sprintf(buf + len, " "); - len += sprintf(buf + len, "RespIEs="); - for (i = 0; i < ifmgd->assocresp_ies_len; i++) { - len += sprintf(buf + len, "%02x", - ifmgd->assocresp_ies[i]); - } - } - len += sprintf(buf + len, ")"); - - if (len > IW_CUSTOM_MAX) { - len = sprintf(buf, "ASSOCRESPIE="); - for (i = 0; i < ifmgd->assocresp_ies_len; i++) { - len += sprintf(buf + len, "%02x", - ifmgd->assocresp_ies[i]); - } - } - - if (len <= IW_CUSTOM_MAX) { - memset(&wrqu, 0, sizeof(wrqu)); - wrqu.data.length = len; - wireless_send_event(sdata->dev, IWEVCUSTOM, &wrqu, buf); - } - - kfree(buf); -} - - static void ieee80211_set_associated(struct ieee80211_sub_if_data *sdata, u32 bss_info_changed) { @@ -966,7 +902,6 @@ static void ieee80211_set_associated(struct ieee80211_sub_if_data *sdata, ifmgd->flags |= IEEE80211_STA_PREV_BSSID_SET; memcpy(ifmgd->prev_bssid, sdata->u.mgd.bssid, ETH_ALEN); - ieee80211_sta_send_associnfo(sdata); ifmgd->last_probe = jiffies; ieee80211_led_assoc(local, 1); @@ -993,8 +928,6 @@ static void ieee80211_set_associated(struct ieee80211_sub_if_data *sdata, netif_tx_start_all_queues(sdata->dev); netif_carrier_on(sdata->dev); - - ieee80211_sta_send_apinfo(sdata); } static void ieee80211_direct_probe(struct ieee80211_sub_if_data *sdata) @@ -1147,8 +1080,6 @@ static void ieee80211_set_disassoc(struct ieee80211_sub_if_data *sdata, changed |= BSS_CHANGED_ASSOC; sdata->vif.bss_conf.assoc = false; - ieee80211_sta_send_apinfo(sdata); - if (self_disconnected || reason == WLAN_REASON_DISASSOC_STA_HAS_LEFT) { ifmgd->state = IEEE80211_STA_MLME_DISABLED; ieee80211_rx_bss_remove(sdata, ifmgd->bssid, diff --git a/net/mac80211/wext.c b/net/mac80211/wext.c index 1da81f456744..d4e61dc903e8 100644 --- a/net/mac80211/wext.c +++ b/net/mac80211/wext.c @@ -27,29 +27,6 @@ #include "aes_ccm.h" -static int ieee80211_ioctl_siwgenie(struct net_device *dev, - struct iw_request_info *info, - struct iw_point *data, char *extra) -{ - struct ieee80211_sub_if_data *sdata; - - sdata = IEEE80211_DEV_TO_SUB_IF(dev); - - if (sdata->vif.type == NL80211_IFTYPE_STATION) { - int ret = ieee80211_sta_set_extra_ie(sdata, extra, data->length); - if (ret && ret != -EALREADY) - return ret; - sdata->u.mgd.flags &= ~IEEE80211_STA_AUTO_BSSID_SEL; - sdata->u.mgd.flags &= ~IEEE80211_STA_EXT_SME; - sdata->u.mgd.flags &= ~IEEE80211_STA_CONTROL_PORT; - if (ret != -EALREADY) - ieee80211_sta_req_auth(sdata); - return 0; - } - - return -EOPNOTSUPP; -} - static int ieee80211_ioctl_siwfreq(struct net_device *dev, struct iw_request_info *info, struct iw_freq *freq, char *extra) @@ -61,16 +38,13 @@ static int ieee80211_ioctl_siwfreq(struct net_device *dev, if (sdata->vif.type == NL80211_IFTYPE_ADHOC) return cfg80211_ibss_wext_siwfreq(dev, info, freq, extra); else if (sdata->vif.type == NL80211_IFTYPE_STATION) - sdata->u.mgd.flags &= ~IEEE80211_STA_AUTO_CHANNEL_SEL; + return cfg80211_mgd_wext_siwfreq(dev, info, freq, extra); /* freq->e == 0: freq->m = channel; otherwise freq = m * 10^e */ if (freq->e == 0) { - if (freq->m < 0) { - if (sdata->vif.type == NL80211_IFTYPE_STATION) - sdata->u.mgd.flags |= - IEEE80211_STA_AUTO_CHANNEL_SEL; - return 0; - } else + if (freq->m < 0) + return -EINVAL; + else chan = ieee80211_get_channel(local->hw.wiphy, ieee80211_channel_to_frequency(freq->m)); } else { @@ -95,9 +69,6 @@ static int ieee80211_ioctl_siwfreq(struct net_device *dev, if (local->oper_channel == chan) return 0; - if (sdata->vif.type == NL80211_IFTYPE_STATION) - ieee80211_sta_req_auth(sdata); - local->oper_channel = chan; local->oper_channel_type = NL80211_CHAN_NO_HT; ieee80211_hw_config(local, 0); @@ -115,6 +86,8 @@ static int ieee80211_ioctl_giwfreq(struct net_device *dev, if (sdata->vif.type == NL80211_IFTYPE_ADHOC) return cfg80211_ibss_wext_giwfreq(dev, info, freq, extra); + else if (sdata->vif.type == NL80211_IFTYPE_STATION) + return cfg80211_mgd_wext_giwfreq(dev, info, freq, extra); freq->m = local->oper_channel->center_freq; freq->e = 6; @@ -128,31 +101,11 @@ static int ieee80211_ioctl_siwessid(struct net_device *dev, struct iw_point *data, char *ssid) { struct ieee80211_sub_if_data *sdata = IEEE80211_DEV_TO_SUB_IF(dev); - size_t len = data->length; - int ret; if (sdata->vif.type == NL80211_IFTYPE_ADHOC) return cfg80211_ibss_wext_siwessid(dev, info, data, ssid); - - /* iwconfig uses nul termination in SSID.. */ - if (len > 0 && ssid[len - 1] == '\0') - len--; - - if (sdata->vif.type == NL80211_IFTYPE_STATION) { - if (data->flags) - sdata->u.mgd.flags &= ~IEEE80211_STA_AUTO_SSID_SEL; - else - sdata->u.mgd.flags |= IEEE80211_STA_AUTO_SSID_SEL; - - ret = ieee80211_sta_set_ssid(sdata, ssid, len); - if (ret) - return ret; - - sdata->u.mgd.flags &= ~IEEE80211_STA_EXT_SME; - sdata->u.mgd.flags &= ~IEEE80211_STA_CONTROL_PORT; - ieee80211_sta_req_auth(sdata); - return 0; - } + else if (sdata->vif.type == NL80211_IFTYPE_STATION) + return cfg80211_mgd_wext_siwessid(dev, info, data, ssid); return -EOPNOTSUPP; } @@ -162,23 +115,14 @@ static int ieee80211_ioctl_giwessid(struct net_device *dev, struct iw_request_info *info, struct iw_point *data, char *ssid) { - size_t len; struct ieee80211_sub_if_data *sdata; sdata = IEEE80211_DEV_TO_SUB_IF(dev); if (sdata->vif.type == NL80211_IFTYPE_ADHOC) return cfg80211_ibss_wext_giwessid(dev, info, data, ssid); - - if (sdata->vif.type == NL80211_IFTYPE_STATION) { - int res = ieee80211_sta_get_ssid(sdata, ssid, &len); - if (res == 0) { - data->length = len; - data->flags = 1; - } else - data->flags = 0; - return res; - } + else if (sdata->vif.type == NL80211_IFTYPE_STATION) + return cfg80211_mgd_wext_giwessid(dev, info, data, ssid); return -EOPNOTSUPP; } @@ -193,24 +137,10 @@ static int ieee80211_ioctl_siwap(struct net_device *dev, if (sdata->vif.type == NL80211_IFTYPE_ADHOC) return cfg80211_ibss_wext_siwap(dev, info, ap_addr, extra); - if (sdata->vif.type == NL80211_IFTYPE_STATION) { - int ret; + if (sdata->vif.type == NL80211_IFTYPE_STATION) + return cfg80211_mgd_wext_siwap(dev, info, ap_addr, extra); - if (is_zero_ether_addr((u8 *) &ap_addr->sa_data)) - sdata->u.mgd.flags |= IEEE80211_STA_AUTO_BSSID_SEL | - IEEE80211_STA_AUTO_CHANNEL_SEL; - else if (is_broadcast_ether_addr((u8 *) &ap_addr->sa_data)) - sdata->u.mgd.flags |= IEEE80211_STA_AUTO_BSSID_SEL; - else - sdata->u.mgd.flags &= ~IEEE80211_STA_AUTO_BSSID_SEL; - ret = ieee80211_sta_set_bssid(sdata, (u8 *) &ap_addr->sa_data); - if (ret) - return ret; - sdata->u.mgd.flags &= ~IEEE80211_STA_EXT_SME; - sdata->u.mgd.flags &= ~IEEE80211_STA_CONTROL_PORT; - ieee80211_sta_req_auth(sdata); - return 0; - } else if (sdata->vif.type == NL80211_IFTYPE_WDS) { + if (sdata->vif.type == NL80211_IFTYPE_WDS) { /* * If it is necessary to update the WDS peer address * while the interface is running, then we need to do @@ -240,14 +170,10 @@ static int ieee80211_ioctl_giwap(struct net_device *dev, if (sdata->vif.type == NL80211_IFTYPE_ADHOC) return cfg80211_ibss_wext_giwap(dev, info, ap_addr, extra); - if (sdata->vif.type == NL80211_IFTYPE_STATION) { - if (sdata->u.mgd.state == IEEE80211_STA_MLME_ASSOCIATED) { - ap_addr->sa_family = ARPHRD_ETHER; - memcpy(&ap_addr->sa_data, sdata->u.mgd.bssid, ETH_ALEN); - } else - memset(&ap_addr->sa_data, 0, ETH_ALEN); - return 0; - } else if (sdata->vif.type == NL80211_IFTYPE_WDS) { + if (sdata->vif.type == NL80211_IFTYPE_STATION) + return cfg80211_mgd_wext_giwap(dev, info, ap_addr, extra); + + if (sdata->vif.type == NL80211_IFTYPE_WDS) { ap_addr->sa_family = ARPHRD_ETHER; memcpy(&ap_addr->sa_data, sdata->u.wds.remote_addr, ETH_ALEN); return 0; @@ -395,85 +321,6 @@ static int ieee80211_ioctl_giwpower(struct net_device *dev, return 0; } -static int ieee80211_ioctl_siwauth(struct net_device *dev, - struct iw_request_info *info, - struct iw_param *data, char *extra) -{ - struct ieee80211_sub_if_data *sdata = IEEE80211_DEV_TO_SUB_IF(dev); - int ret = 0; - - switch (data->flags & IW_AUTH_INDEX) { - case IW_AUTH_WPA_VERSION: - case IW_AUTH_CIPHER_GROUP: - case IW_AUTH_WPA_ENABLED: - case IW_AUTH_RX_UNENCRYPTED_EAPOL: - case IW_AUTH_KEY_MGMT: - case IW_AUTH_CIPHER_GROUP_MGMT: - break; - case IW_AUTH_CIPHER_PAIRWISE: - if (sdata->vif.type == NL80211_IFTYPE_STATION) { - if (data->value & (IW_AUTH_CIPHER_WEP40 | - IW_AUTH_CIPHER_WEP104 | IW_AUTH_CIPHER_TKIP)) - sdata->u.mgd.flags |= - IEEE80211_STA_TKIP_WEP_USED; - else - sdata->u.mgd.flags &= - ~IEEE80211_STA_TKIP_WEP_USED; - } - break; - case IW_AUTH_DROP_UNENCRYPTED: - sdata->drop_unencrypted = !!data->value; - break; - case IW_AUTH_PRIVACY_INVOKED: - if (sdata->vif.type != NL80211_IFTYPE_STATION) - ret = -EINVAL; - else { - sdata->u.mgd.flags &= ~IEEE80211_STA_PRIVACY_INVOKED; - /* - * Privacy invoked by wpa_supplicant, store the - * value and allow associating to a protected - * network without having a key up front. - */ - if (data->value) - sdata->u.mgd.flags |= - IEEE80211_STA_PRIVACY_INVOKED; - } - break; - case IW_AUTH_80211_AUTH_ALG: - if (sdata->vif.type == NL80211_IFTYPE_STATION) - sdata->u.mgd.auth_algs = data->value; - else - ret = -EOPNOTSUPP; - break; - case IW_AUTH_MFP: - if (!(sdata->local->hw.flags & IEEE80211_HW_MFP_CAPABLE)) { - ret = -EOPNOTSUPP; - break; - } - if (sdata->vif.type == NL80211_IFTYPE_STATION) { - switch (data->value) { - case IW_AUTH_MFP_DISABLED: - sdata->u.mgd.mfp = IEEE80211_MFP_DISABLED; - break; - case IW_AUTH_MFP_OPTIONAL: - sdata->u.mgd.mfp = IEEE80211_MFP_OPTIONAL; - break; - case IW_AUTH_MFP_REQUIRED: - sdata->u.mgd.mfp = IEEE80211_MFP_REQUIRED; - break; - default: - ret = -EINVAL; - } - } else - ret = -EOPNOTSUPP; - break; - default: - ret = -EOPNOTSUPP; - break; - } - return ret; -} - /* Get wireless statistics. Called by /proc/net/wireless and by SIOCGIWSTATS */ static struct iw_statistics *ieee80211_get_wireless_stats(struct net_device *dev) { @@ -541,28 +388,6 @@ static struct iw_statistics *ieee80211_get_wireless_stats(struct net_device *dev return wstats; } -static int ieee80211_ioctl_giwauth(struct net_device *dev, - struct iw_request_info *info, - struct iw_param *data, char *extra) -{ - struct ieee80211_sub_if_data *sdata = IEEE80211_DEV_TO_SUB_IF(dev); - int ret = 0; - - switch (data->flags & IW_AUTH_INDEX) { - case IW_AUTH_80211_AUTH_ALG: - if (sdata->vif.type == NL80211_IFTYPE_STATION) - data->value = sdata->u.mgd.auth_algs; - else - ret = -EOPNOTSUPP; - break; - default: - ret = -EOPNOTSUPP; - break; - } - return ret; -} - - /* Structures to export the Wireless Handlers */ static const iw_handler ieee80211_handler[] = @@ -615,10 +440,10 @@ static const iw_handler ieee80211_handler[] = (iw_handler) ieee80211_ioctl_giwpower, /* SIOCGIWPOWER */ (iw_handler) NULL, /* -- hole -- */ (iw_handler) NULL, /* -- hole -- */ - (iw_handler) ieee80211_ioctl_siwgenie, /* SIOCSIWGENIE */ + (iw_handler) cfg80211_wext_siwgenie, /* SIOCSIWGENIE */ (iw_handler) NULL, /* SIOCGIWGENIE */ - (iw_handler) ieee80211_ioctl_siwauth, /* SIOCSIWAUTH */ - (iw_handler) ieee80211_ioctl_giwauth, /* SIOCGIWAUTH */ + (iw_handler) cfg80211_wext_siwauth, /* SIOCSIWAUTH */ + (iw_handler) cfg80211_wext_giwauth, /* SIOCGIWAUTH */ (iw_handler) cfg80211_wext_siwencodeext, /* SIOCSIWENCODEEXT */ (iw_handler) NULL, /* SIOCGIWENCODEEXT */ (iw_handler) NULL, /* SIOCSIWPMKSA */ diff --git a/net/wireless/Makefile b/net/wireless/Makefile index 750c08e31b10..d74cc77fa57a 100644 --- a/net/wireless/Makefile +++ b/net/wireless/Makefile @@ -7,6 +7,6 @@ obj-$(CONFIG_LIB80211_CRYPT_TKIP) += lib80211_crypt_tkip.o cfg80211-y += core.o sysfs.o radiotap.o util.o reg.o scan.o nl80211.o mlme.o ibss.o sme.o cfg80211-$(CONFIG_CFG80211_DEBUGFS) += debugfs.o -cfg80211-$(CONFIG_WIRELESS_EXT) += wext-compat.o +cfg80211-$(CONFIG_WIRELESS_EXT) += wext-compat.o wext-sme.o ccflags-y += -D__CHECK_ENDIAN__ diff --git a/net/wireless/core.c b/net/wireless/core.c index a0a679704612..e2f80dd0e4a6 100644 --- a/net/wireless/core.c +++ b/net/wireless/core.c @@ -553,6 +553,7 @@ static int cfg80211_netdev_notifier_call(struct notifier_block * nb, #ifdef CONFIG_WIRELESS_EXT wdev->wext.default_key = -1; wdev->wext.default_mgmt_key = -1; + wdev->wext.connect.auth_type = NL80211_AUTHTYPE_AUTOMATIC; #endif mutex_unlock(&rdev->devlist_mtx); break; @@ -565,8 +566,13 @@ static int cfg80211_netdev_notifier_call(struct notifier_block * nb, cfg80211_leave_ibss(rdev, dev, true); break; case NL80211_IFTYPE_STATION: +#ifdef CONFIG_WIRELESS_EXT + kfree(wdev->wext.ie); + wdev->wext.ie = NULL; + wdev->wext.ie_len = 0; +#endif cfg80211_disconnect(rdev, dev, - WLAN_REASON_DEAUTH_LEAVING); + WLAN_REASON_DEAUTH_LEAVING, true); break; default: break; @@ -578,11 +584,20 @@ static int cfg80211_netdev_notifier_call(struct notifier_block * nb, break; case NETDEV_UP: #ifdef CONFIG_WIRELESS_EXT - if (wdev->iftype != NL80211_IFTYPE_ADHOC) + switch (wdev->iftype) { + case NL80211_IFTYPE_ADHOC: + if (wdev->wext.ibss.ssid_len) + cfg80211_join_ibss(rdev, dev, + &wdev->wext.ibss); break; - if (!wdev->wext.ibss.ssid_len) + case NL80211_IFTYPE_STATION: + if (wdev->wext.connect.ssid_len) + cfg80211_connect(rdev, dev, + &wdev->wext.connect); + break; + default: break; - cfg80211_join_ibss(rdev, dev, &wdev->wext.ibss); + } #endif break; case NETDEV_UNREGISTER: diff --git a/net/wireless/core.h b/net/wireless/core.h index 2c0f64252f3d..5209acb0ff7e 100644 --- a/net/wireless/core.h +++ b/net/wireless/core.h @@ -181,7 +181,8 @@ int cfg80211_connect(struct cfg80211_registered_device *rdev, struct net_device *dev, struct cfg80211_connect_params *connect); int cfg80211_disconnect(struct cfg80211_registered_device *rdev, - struct net_device *dev, u16 reason); + struct net_device *dev, u16 reason, + bool wextev); void cfg80211_conn_work(struct work_struct *work); diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index 89aa9e781d10..0008144b354b 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -3747,7 +3747,7 @@ static int nl80211_disconnect(struct sk_buff *skb, struct genl_info *info) goto out; } - err = cfg80211_disconnect(drv, dev, reason); + err = cfg80211_disconnect(drv, dev, reason, true); out: cfg80211_put_dev(drv); diff --git a/net/wireless/sme.c b/net/wireless/sme.c index 3abb04729873..f272ebf94303 100644 --- a/net/wireless/sme.c +++ b/net/wireless/sme.c @@ -273,10 +273,10 @@ void cfg80211_sme_rx_auth(struct net_device *dev, const u8 *buf, size_t len) } } -void cfg80211_connect_result(struct net_device *dev, const u8 *bssid, - const u8 *req_ie, size_t req_ie_len, - const u8 *resp_ie, size_t resp_ie_len, - u16 status, gfp_t gfp) +static void __cfg80211_connect_result(struct net_device *dev, const u8 *bssid, + const u8 *req_ie, size_t req_ie_len, + const u8 *resp_ie, size_t resp_ie_len, + u16 status, bool wextev, gfp_t gfp) { struct wireless_dev *wdev = dev->ieee80211_ptr; struct cfg80211_bss *bss; @@ -321,25 +321,36 @@ void cfg80211_connect_result(struct net_device *dev, const u8 *bssid, status, gfp); #ifdef CONFIG_WIRELESS_EXT - if (req_ie && status == WLAN_STATUS_SUCCESS) { - memset(&wrqu, 0, sizeof(wrqu)); - wrqu.data.length = req_ie_len; - wireless_send_event(dev, IWEVASSOCRESPIE, &wrqu, req_ie); - } + if (wextev) { + if (req_ie && status == WLAN_STATUS_SUCCESS) { + memset(&wrqu, 0, sizeof(wrqu)); + wrqu.data.length = req_ie_len; + wireless_send_event(dev, IWEVASSOCRESPIE, &wrqu, req_ie); + } + + if (resp_ie && status == WLAN_STATUS_SUCCESS) { + memset(&wrqu, 0, sizeof(wrqu)); + wrqu.data.length = resp_ie_len; + wireless_send_event(dev, IWEVASSOCRESPIE, &wrqu, resp_ie); + } - if (resp_ie && status == WLAN_STATUS_SUCCESS) { memset(&wrqu, 0, sizeof(wrqu)); - wrqu.data.length = resp_ie_len; - wireless_send_event(dev, IWEVASSOCRESPIE, &wrqu, resp_ie); + wrqu.ap_addr.sa_family = ARPHRD_ETHER; + if (bssid && status == WLAN_STATUS_SUCCESS) + memcpy(wrqu.ap_addr.sa_data, bssid, ETH_ALEN); + wireless_send_event(dev, SIOCGIWAP, &wrqu, NULL); } - - memset(&wrqu, 0, sizeof(wrqu)); - wrqu.ap_addr.sa_family = ARPHRD_ETHER; - if (bssid && status == WLAN_STATUS_SUCCESS) - memcpy(wrqu.ap_addr.sa_data, bssid, ETH_ALEN); - wireless_send_event(dev, SIOCGIWAP, &wrqu, NULL); #endif } + +void cfg80211_connect_result(struct net_device *dev, const u8 *bssid, + const u8 *req_ie, size_t req_ie_len, + const u8 *resp_ie, size_t resp_ie_len, + u16 status, gfp_t gfp) +{ + bool wextev = status == WLAN_STATUS_SUCCESS; + __cfg80211_connect_result(dev, bssid, req_ie, req_ie_len, resp_ie, resp_ie_len, status, wextev, gfp); +} EXPORT_SYMBOL(cfg80211_connect_result); void cfg80211_roamed(struct net_device *dev, const u8 *bssid, @@ -540,7 +551,7 @@ int cfg80211_connect(struct cfg80211_registered_device *rdev, } int cfg80211_disconnect(struct cfg80211_registered_device *rdev, - struct net_device *dev, u16 reason) + struct net_device *dev, u16 reason, bool wextev) { struct wireless_dev *wdev = dev->ieee80211_ptr; int err; @@ -585,9 +596,9 @@ int cfg80211_disconnect(struct cfg80211_registered_device *rdev, if (wdev->sme_state == CFG80211_SME_CONNECTED) __cfg80211_disconnected(dev, GFP_KERNEL, NULL, 0, 0, false); else if (wdev->sme_state == CFG80211_SME_CONNECTING) - cfg80211_connect_result(dev, NULL, NULL, 0, NULL, 0, - WLAN_STATUS_UNSPECIFIED_FAILURE, - GFP_KERNEL); + __cfg80211_connect_result(dev, NULL, NULL, 0, NULL, 0, + WLAN_STATUS_UNSPECIFIED_FAILURE, + wextev, GFP_KERNEL); return 0; } diff --git a/net/wireless/wext-compat.c b/net/wireless/wext-compat.c index cae3b52fba7f..02f052fc1808 100644 --- a/net/wireless/wext-compat.c +++ b/net/wireless/wext-compat.c @@ -261,50 +261,6 @@ int cfg80211_wext_giwrange(struct net_device *dev, } EXPORT_SYMBOL_GPL(cfg80211_wext_giwrange); -int cfg80211_wext_siwmlme(struct net_device *dev, - struct iw_request_info *info, - struct iw_point *data, char *extra) -{ - struct wireless_dev *wdev = dev->ieee80211_ptr; - struct iw_mlme *mlme = (struct iw_mlme *)extra; - struct cfg80211_registered_device *rdev; - union { - struct cfg80211_disassoc_request disassoc; - struct cfg80211_deauth_request deauth; - } cmd; - - if (!wdev) - return -EOPNOTSUPP; - - rdev = wiphy_to_dev(wdev->wiphy); - - if (wdev->iftype != NL80211_IFTYPE_STATION) - return -EINVAL; - - if (mlme->addr.sa_family != ARPHRD_ETHER) - return -EINVAL; - - memset(&cmd, 0, sizeof(cmd)); - - switch (mlme->cmd) { - case IW_MLME_DEAUTH: - if (!rdev->ops->deauth) - return -EOPNOTSUPP; - cmd.deauth.peer_addr = mlme->addr.sa_data; - cmd.deauth.reason_code = mlme->reason_code; - return rdev->ops->deauth(wdev->wiphy, dev, &cmd.deauth); - case IW_MLME_DISASSOC: - if (!rdev->ops->disassoc) - return -EOPNOTSUPP; - cmd.disassoc.peer_addr = mlme->addr.sa_data; - cmd.disassoc.reason_code = mlme->reason_code; - return rdev->ops->disassoc(wdev->wiphy, dev, &cmd.disassoc); - default: - return -EOPNOTSUPP; - } -} -EXPORT_SYMBOL_GPL(cfg80211_wext_siwmlme); - /** * cfg80211_wext_freq - get wext frequency for non-"auto" @@ -846,3 +802,188 @@ int cfg80211_wext_giwtxpower(struct net_device *dev, return 0; } EXPORT_SYMBOL_GPL(cfg80211_wext_giwtxpower); + +static int cfg80211_set_auth_alg(struct wireless_dev *wdev, + s32 auth_alg) +{ + int nr_alg = 0; + + if (!auth_alg) + return -EINVAL; + + if (auth_alg & ~(IW_AUTH_ALG_OPEN_SYSTEM | + IW_AUTH_ALG_SHARED_KEY | + IW_AUTH_ALG_LEAP)) + return -EINVAL; + + if (auth_alg & IW_AUTH_ALG_OPEN_SYSTEM) { + nr_alg++; + wdev->wext.connect.auth_type = NL80211_AUTHTYPE_OPEN_SYSTEM; + } + + if (auth_alg & IW_AUTH_ALG_SHARED_KEY) { + nr_alg++; + wdev->wext.connect.auth_type = NL80211_AUTHTYPE_SHARED_KEY; + } + + if (auth_alg & IW_AUTH_ALG_LEAP) { + nr_alg++; + wdev->wext.connect.auth_type = NL80211_AUTHTYPE_NETWORK_EAP; + } + + if (nr_alg > 1) + wdev->wext.connect.auth_type = NL80211_AUTHTYPE_AUTOMATIC; + + return 0; +} + +static int cfg80211_set_wpa_version(struct wireless_dev *wdev, u32 wpa_versions) +{ + wdev->wext.connect.crypto.wpa_versions = 0; + + if (wpa_versions & ~(IW_AUTH_WPA_VERSION_WPA | + IW_AUTH_WPA_VERSION_WPA2)) + return -EINVAL; + + if (wpa_versions & IW_AUTH_WPA_VERSION_WPA) + wdev->wext.connect.crypto.wpa_versions |= + NL80211_WPA_VERSION_1; + + if (wpa_versions & IW_AUTH_WPA_VERSION_WPA2) + wdev->wext.connect.crypto.wpa_versions |= + NL80211_WPA_VERSION_2; + + return 0; +} + +int cfg80211_set_cipher_group(struct wireless_dev *wdev, u32 cipher) +{ + wdev->wext.connect.crypto.cipher_group = 0; + + if (cipher & IW_AUTH_CIPHER_WEP40) + wdev->wext.connect.crypto.cipher_group = + WLAN_CIPHER_SUITE_WEP40; + else if (cipher & IW_AUTH_CIPHER_WEP104) + wdev->wext.connect.crypto.cipher_group = + WLAN_CIPHER_SUITE_WEP104; + else if (cipher & IW_AUTH_CIPHER_TKIP) + wdev->wext.connect.crypto.cipher_group = + WLAN_CIPHER_SUITE_TKIP; + else if (cipher & IW_AUTH_CIPHER_CCMP) + wdev->wext.connect.crypto.cipher_group = + WLAN_CIPHER_SUITE_CCMP; + else if (cipher & IW_AUTH_CIPHER_AES_CMAC) + wdev->wext.connect.crypto.cipher_group = + WLAN_CIPHER_SUITE_AES_CMAC; + else + return -EINVAL; + + return 0; +} + +int cfg80211_set_cipher_pairwise(struct wireless_dev *wdev, u32 cipher) +{ + int nr_ciphers = 0; + u32 *ciphers_pairwise = wdev->wext.connect.crypto.ciphers_pairwise; + + if (cipher & IW_AUTH_CIPHER_WEP40) { + ciphers_pairwise[nr_ciphers] = WLAN_CIPHER_SUITE_WEP40; + nr_ciphers++; + } + + if (cipher & IW_AUTH_CIPHER_WEP104) { + ciphers_pairwise[nr_ciphers] = WLAN_CIPHER_SUITE_WEP104; + nr_ciphers++; + } + + if (cipher & IW_AUTH_CIPHER_TKIP) { + ciphers_pairwise[nr_ciphers] = WLAN_CIPHER_SUITE_TKIP; + nr_ciphers++; + } + + if (cipher & IW_AUTH_CIPHER_CCMP) { + ciphers_pairwise[nr_ciphers] = WLAN_CIPHER_SUITE_CCMP; + nr_ciphers++; + } + + if (cipher & IW_AUTH_CIPHER_AES_CMAC) { + ciphers_pairwise[nr_ciphers] = WLAN_CIPHER_SUITE_AES_CMAC; + nr_ciphers++; + } + + BUILD_BUG_ON(NL80211_MAX_NR_CIPHER_SUITES < 5); + + wdev->wext.connect.crypto.n_ciphers_pairwise = nr_ciphers; + + return 0; +} + + +int cfg80211_set_key_mgt(struct wireless_dev *wdev, u32 key_mgt) +{ + int nr_akm_suites = 0; + + if (key_mgt & ~(IW_AUTH_KEY_MGMT_802_1X | + IW_AUTH_KEY_MGMT_PSK)) + return -EINVAL; + + if (key_mgt & IW_AUTH_KEY_MGMT_802_1X) { + wdev->wext.connect.crypto.akm_suites[nr_akm_suites] = + WLAN_AKM_SUITE_8021X; + nr_akm_suites++; + } + + if (key_mgt & IW_AUTH_KEY_MGMT_PSK) { + wdev->wext.connect.crypto.akm_suites[nr_akm_suites] = + WLAN_AKM_SUITE_PSK; + nr_akm_suites++; + } + + wdev->wext.connect.crypto.n_akm_suites = nr_akm_suites; + + return 0; +} + +int cfg80211_wext_siwauth(struct net_device *dev, + struct iw_request_info *info, + struct iw_param *data, char *extra) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + + if (wdev->iftype != NL80211_IFTYPE_STATION) + return -EOPNOTSUPP; + + switch (data->flags & IW_AUTH_INDEX) { + case IW_AUTH_PRIVACY_INVOKED: + wdev->wext.connect.privacy = data->value; + return 0; + case IW_AUTH_WPA_VERSION: + return cfg80211_set_wpa_version(wdev, data->value); + case IW_AUTH_CIPHER_GROUP: + return cfg80211_set_cipher_group(wdev, data->value); + case IW_AUTH_KEY_MGMT: + return cfg80211_set_key_mgt(wdev, data->value); + case IW_AUTH_CIPHER_PAIRWISE: + return cfg80211_set_cipher_pairwise(wdev, data->value); + case IW_AUTH_80211_AUTH_ALG: + return cfg80211_set_auth_alg(wdev, data->value); + case IW_AUTH_WPA_ENABLED: + case IW_AUTH_RX_UNENCRYPTED_EAPOL: + case IW_AUTH_DROP_UNENCRYPTED: + case IW_AUTH_MFP: + return 0; + default: + return -EOPNOTSUPP; + } +} +EXPORT_SYMBOL_GPL(cfg80211_wext_siwauth); + +int cfg80211_wext_giwauth(struct net_device *dev, + struct iw_request_info *info, + struct iw_param *data, char *extra) +{ + /* XXX: what do we need? */ + + return -EOPNOTSUPP; +} +EXPORT_SYMBOL_GPL(cfg80211_wext_giwauth); diff --git a/net/wireless/wext-sme.c b/net/wireless/wext-sme.c new file mode 100644 index 000000000000..3b531d572b69 --- /dev/null +++ b/net/wireless/wext-sme.c @@ -0,0 +1,329 @@ +/* + * cfg80211 wext compat for managed mode. + * + * Copyright 2009 Johannes Berg + * Copyright (C) 2009 Intel Corporation. All rights reserved. + */ + +#include +#include +#include +#include "nl80211.h" + +static int cfg80211_mgd_wext_connect(struct cfg80211_registered_device *rdev, + struct wireless_dev *wdev) +{ + int err; + + if (!netif_running(wdev->netdev)) + return 0; + + wdev->wext.connect.ie = wdev->wext.ie; + wdev->wext.connect.ie_len = wdev->wext.ie_len; + wdev->wext.connect.privacy = wdev->wext.default_key != -1; + + err = 0; + if (wdev->wext.connect.ssid_len != 0) + err = cfg80211_connect(rdev, wdev->netdev, + &wdev->wext.connect); + + return err; +} + +int cfg80211_mgd_wext_siwfreq(struct net_device *dev, + struct iw_request_info *info, + struct iw_freq *freq, char *extra) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + struct cfg80211_registered_device *rdev = wiphy_to_dev(wdev->wiphy); + struct ieee80211_channel *chan; + int err; + + /* call only for station! */ + if (WARN_ON(wdev->iftype != NL80211_IFTYPE_STATION)) + return -EINVAL; + + chan = cfg80211_wext_freq(wdev->wiphy, freq); + if (chan && IS_ERR(chan)) + return PTR_ERR(chan); + + if (chan && (chan->flags & IEEE80211_CHAN_DISABLED)) + return -EINVAL; + + if (wdev->wext.connect.channel == chan) + return 0; + + if (wdev->sme_state != CFG80211_SME_IDLE) { + bool event = true; + /* if SSID set, we'll try right again, avoid event */ + if (wdev->wext.connect.ssid_len) + event = false; + err = cfg80211_disconnect(wiphy_to_dev(wdev->wiphy), + dev, WLAN_REASON_DEAUTH_LEAVING, + event); + if (err) + return err; + } + + wdev->wext.connect.channel = chan; + + /* SSID is not set, we just want to switch channel */ + if (wdev->wext.connect.ssid_len && chan) { + if (!rdev->ops->set_channel) + return -EOPNOTSUPP; + + return rdev->ops->set_channel(wdev->wiphy, chan, + NL80211_CHAN_NO_HT); + } + + return cfg80211_mgd_wext_connect(wiphy_to_dev(wdev->wiphy), wdev); +} +/* temporary symbol - mark GPL - in the future the handler won't be */ +EXPORT_SYMBOL_GPL(cfg80211_mgd_wext_siwfreq); + +int cfg80211_mgd_wext_giwfreq(struct net_device *dev, + struct iw_request_info *info, + struct iw_freq *freq, char *extra) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + struct ieee80211_channel *chan = NULL; + + /* call only for station! */ + if (WARN_ON(wdev->iftype != NL80211_IFTYPE_STATION)) + return -EINVAL; + + if (wdev->current_bss) + chan = wdev->current_bss->channel; + else if (wdev->wext.connect.channel) + chan = wdev->wext.connect.channel; + + if (chan) { + freq->m = chan->center_freq; + freq->e = 6; + return 0; + } + + /* no channel if not joining */ + return -EINVAL; +} +/* temporary symbol - mark GPL - in the future the handler won't be */ +EXPORT_SYMBOL_GPL(cfg80211_mgd_wext_giwfreq); + +int cfg80211_mgd_wext_siwessid(struct net_device *dev, + struct iw_request_info *info, + struct iw_point *data, char *ssid) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + size_t len = data->length; + int err; + + /* call only for station! */ + if (WARN_ON(wdev->iftype != NL80211_IFTYPE_STATION)) + return -EINVAL; + + if (!data->flags) + len = 0; + + /* iwconfig uses nul termination in SSID.. */ + if (len > 0 && ssid[len - 1] == '\0') + len--; + + if (wdev->wext.connect.ssid && len && + len == wdev->wext.connect.ssid_len && + memcmp(wdev->wext.connect.ssid, ssid, len)) + return 0; + + if (wdev->sme_state != CFG80211_SME_IDLE) { + bool event = true; + /* if SSID set now, we'll try to connect, avoid event */ + if (len) + event = false; + err = cfg80211_disconnect(wiphy_to_dev(wdev->wiphy), + dev, WLAN_REASON_DEAUTH_LEAVING, + event); + if (err) + return err; + } + + wdev->wext.connect.ssid = wdev->wext.ssid; + memcpy(wdev->wext.ssid, ssid, len); + wdev->wext.connect.ssid_len = len; + + wdev->wext.connect.crypto.control_port = false; + + return cfg80211_mgd_wext_connect(wiphy_to_dev(wdev->wiphy), wdev); +} +/* temporary symbol - mark GPL - in the future the handler won't be */ +EXPORT_SYMBOL_GPL(cfg80211_mgd_wext_siwessid); + +int cfg80211_mgd_wext_giwessid(struct net_device *dev, + struct iw_request_info *info, + struct iw_point *data, char *ssid) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + + /* call only for station! */ + if (WARN_ON(wdev->iftype != NL80211_IFTYPE_STATION)) + return -EINVAL; + + data->flags = 0; + + if (wdev->ssid_len) { + data->flags = 1; + data->length = wdev->ssid_len; + memcpy(ssid, wdev->ssid, data->length); + } else if (wdev->wext.connect.ssid && wdev->wext.connect.ssid_len) { + data->flags = 1; + data->length = wdev->wext.connect.ssid_len; + memcpy(ssid, wdev->wext.connect.ssid, data->length); + } else + data->flags = 0; + + return 0; +} +/* temporary symbol - mark GPL - in the future the handler won't be */ +EXPORT_SYMBOL_GPL(cfg80211_mgd_wext_giwessid); + +int cfg80211_mgd_wext_siwap(struct net_device *dev, + struct iw_request_info *info, + struct sockaddr *ap_addr, char *extra) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + u8 *bssid = ap_addr->sa_data; + int err; + + /* call only for station! */ + if (WARN_ON(wdev->iftype != NL80211_IFTYPE_STATION)) + return -EINVAL; + + if (ap_addr->sa_family != ARPHRD_ETHER) + return -EINVAL; + + /* automatic mode */ + if (is_zero_ether_addr(bssid) || is_broadcast_ether_addr(bssid)) + bssid = NULL; + + /* both automatic */ + if (!bssid && !wdev->wext.connect.bssid) + return 0; + + /* fixed already - and no change */ + if (wdev->wext.connect.bssid && bssid && + compare_ether_addr(bssid, wdev->wext.connect.bssid) == 0) + return 0; + + if (wdev->sme_state != CFG80211_SME_IDLE) { + err = cfg80211_disconnect(wiphy_to_dev(wdev->wiphy), + dev, WLAN_REASON_DEAUTH_LEAVING, + false); + if (err) + return err; + } + + if (bssid) { + memcpy(wdev->wext.bssid, bssid, ETH_ALEN); + wdev->wext.connect.bssid = wdev->wext.bssid; + } else + wdev->wext.connect.bssid = NULL; + + return cfg80211_mgd_wext_connect(wiphy_to_dev(wdev->wiphy), wdev); +} +/* temporary symbol - mark GPL - in the future the handler won't be */ +EXPORT_SYMBOL_GPL(cfg80211_mgd_wext_siwap); + +int cfg80211_mgd_wext_giwap(struct net_device *dev, + struct iw_request_info *info, + struct sockaddr *ap_addr, char *extra) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + + /* call only for station! */ + if (WARN_ON(wdev->iftype != NL80211_IFTYPE_STATION)) + return -EINVAL; + + ap_addr->sa_family = ARPHRD_ETHER; + + if (wdev->current_bss) + memcpy(ap_addr->sa_data, wdev->current_bss->bssid, ETH_ALEN); + else if (wdev->wext.connect.bssid) + memcpy(ap_addr->sa_data, wdev->wext.connect.bssid, ETH_ALEN); + else + memset(ap_addr->sa_data, 0, ETH_ALEN); + + return 0; +} +/* temporary symbol - mark GPL - in the future the handler won't be */ +EXPORT_SYMBOL_GPL(cfg80211_mgd_wext_giwap); + +int cfg80211_wext_siwgenie(struct net_device *dev, + struct iw_request_info *info, + struct iw_point *data, char *extra) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + struct cfg80211_registered_device *rdev = wiphy_to_dev(wdev->wiphy); + u8 *ie = extra; + int ie_len = data->length, err; + + if (wdev->iftype != NL80211_IFTYPE_STATION) + return -EOPNOTSUPP; + + if (!ie_len) + ie = NULL; + + /* no change */ + if (wdev->wext.ie_len == ie_len && + memcmp(wdev->wext.ie, ie, ie_len) == 0) + return 0; + + if (ie_len) { + ie = kmemdup(extra, ie_len, GFP_KERNEL); + if (!ie) + return -ENOMEM; + } else + ie = NULL; + + kfree(wdev->wext.ie); + wdev->wext.ie = ie; + wdev->wext.ie_len = ie_len; + + if (wdev->sme_state != CFG80211_SME_IDLE) { + err = cfg80211_disconnect(rdev, dev, + WLAN_REASON_DEAUTH_LEAVING, false); + if (err) + return err; + } + + /* userspace better not think we'll reconnect */ + return 0; +} +EXPORT_SYMBOL_GPL(cfg80211_wext_siwgenie); + +int cfg80211_wext_siwmlme(struct net_device *dev, + struct iw_request_info *info, + struct iw_point *data, char *extra) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + struct iw_mlme *mlme = (struct iw_mlme *)extra; + struct cfg80211_registered_device *rdev; + + if (!wdev) + return -EOPNOTSUPP; + + rdev = wiphy_to_dev(wdev->wiphy); + + if (wdev->iftype != NL80211_IFTYPE_STATION) + return -EINVAL; + + if (mlme->addr.sa_family != ARPHRD_ETHER) + return -EINVAL; + + switch (mlme->cmd) { + case IW_MLME_DEAUTH: + case IW_MLME_DISASSOC: + return cfg80211_disconnect(rdev, dev, mlme->reason_code, + true); + default: + return -EOPNOTSUPP; + } +} +EXPORT_SYMBOL_GPL(cfg80211_wext_siwmlme); -- cgit v1.2.3 From bc92afd92088ab41223383cc6863ab4792533c54 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Wed, 1 Jul 2009 21:26:57 +0200 Subject: cfg80211: implement iwpower Just on/off and timeout, and with a hacky cfg80211 method until we figure out what we want, though this is probably sufficient as we want to use pm_qos for wifi everywhere. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- drivers/net/wireless/iwmc3200wifi/cfg80211.c | 22 +++++++++ drivers/net/wireless/iwmc3200wifi/wext.c | 47 +------------------ include/net/cfg80211.h | 13 ++++++ net/mac80211/Kconfig | 16 ------- net/mac80211/cfg.c | 26 +++++++++++ net/mac80211/mlme.c | 5 -- net/mac80211/wext.c | 70 +--------------------------- net/wireless/Kconfig | 16 +++++++ net/wireless/core.c | 11 ++++- net/wireless/wext-compat.c | 60 ++++++++++++++++++++++++ 10 files changed, 151 insertions(+), 135 deletions(-) (limited to 'include') diff --git a/drivers/net/wireless/iwmc3200wifi/cfg80211.c b/drivers/net/wireless/iwmc3200wifi/cfg80211.c index d0629d4757d7..54bebba8e27e 100644 --- a/drivers/net/wireless/iwmc3200wifi/cfg80211.c +++ b/drivers/net/wireless/iwmc3200wifi/cfg80211.c @@ -522,6 +522,27 @@ static int iwm_cfg80211_get_txpower(struct wiphy *wiphy, int *dbm) return 0; } +static int iwm_cfg80211_set_power_mgmt(struct wiphy *wiphy, + struct net_device *dev, + bool enabled, int timeout) +{ + struct iwm_priv *iwm = wiphy_to_iwm(wiphy); + u32 power_index; + + if (enabled) + power_index = IWM_POWER_INDEX_DEFAULT; + else + power_index = IWM_POWER_INDEX_MIN; + + if (power_index == iwm->conf.power_index) + return 0; + + iwm->conf.power_index = power_index; + + return iwm_umac_set_config_fix(iwm, UMAC_PARAM_TBL_CFG_FIX, + CFG_POWER_INDEX, iwm->conf.power_index); +} + static struct cfg80211_ops iwm_cfg80211_ops = { .change_virtual_intf = iwm_cfg80211_change_iface, .add_key = iwm_cfg80211_add_key, @@ -534,6 +555,7 @@ static struct cfg80211_ops iwm_cfg80211_ops = { .leave_ibss = iwm_cfg80211_leave_ibss, .set_tx_power = iwm_cfg80211_set_txpower, .get_tx_power = iwm_cfg80211_get_txpower, + .set_power_mgmt = iwm_cfg80211_set_power_mgmt, }; struct wireless_dev *iwm_wdev_alloc(int sizeof_bus, struct device *dev) diff --git a/drivers/net/wireless/iwmc3200wifi/wext.c b/drivers/net/wireless/iwmc3200wifi/wext.c index 973457383c11..2e7eaf96cf93 100644 --- a/drivers/net/wireless/iwmc3200wifi/wext.c +++ b/drivers/net/wireless/iwmc3200wifi/wext.c @@ -238,49 +238,6 @@ static int iwm_set_wpa_version(struct iwm_priv *iwm, u8 wpa_version) return 0; } -static int iwm_wext_siwpower(struct net_device *dev, - struct iw_request_info *info, - struct iw_param *wrq, char *extra) -{ - struct iwm_priv *iwm = ndev_to_iwm(dev); - u32 power_index; - - if (wrq->disabled) { - power_index = IWM_POWER_INDEX_MIN; - goto set; - } else - power_index = IWM_POWER_INDEX_DEFAULT; - - switch (wrq->flags & IW_POWER_MODE) { - case IW_POWER_ON: - case IW_POWER_MODE: - case IW_POWER_ALL_R: - break; - default: - return -EINVAL; - } - - set: - if (power_index == iwm->conf.power_index) - return 0; - - iwm->conf.power_index = power_index; - - return iwm_umac_set_config_fix(iwm, UMAC_PARAM_TBL_CFG_FIX, - CFG_POWER_INDEX, iwm->conf.power_index); -} - -static int iwm_wext_giwpower(struct net_device *dev, - struct iw_request_info *info, - union iwreq_data *wrqu, char *extra) -{ - struct iwm_priv *iwm = ndev_to_iwm(dev); - - wrqu->power.disabled = (iwm->conf.power_index == IWM_POWER_INDEX_MIN); - - return 0; -} - static int iwm_set_key_mgt(struct iwm_priv *iwm, u8 key_mgt) { u8 *auth_type = &iwm->umac_profile->sec.auth_type; @@ -458,8 +415,8 @@ static const iw_handler iwm_handlers[] = (iw_handler) NULL, /* SIOCGIWRETRY */ (iw_handler) cfg80211_wext_siwencode, /* SIOCSIWENCODE */ (iw_handler) cfg80211_wext_giwencode, /* SIOCGIWENCODE */ - (iw_handler) iwm_wext_siwpower, /* SIOCSIWPOWER */ - (iw_handler) iwm_wext_giwpower, /* SIOCGIWPOWER */ + (iw_handler) cfg80211_wext_siwpower, /* SIOCSIWPOWER */ + (iw_handler) cfg80211_wext_giwpower, /* SIOCGIWPOWER */ (iw_handler) NULL, /* -- hole -- */ (iw_handler) NULL, /* -- hole -- */ (iw_handler) NULL, /* SIOCSIWGENIE */ diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index 07085216532d..82b7d804f6da 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -1023,6 +1023,10 @@ struct cfg80211_ops { #ifdef CONFIG_NL80211_TESTMODE int (*testmode_cmd)(struct wiphy *wiphy, void *data, int len); #endif + + /* some temporary stuff to finish wext */ + int (*set_power_mgmt)(struct wiphy *wiphy, struct net_device *dev, + bool enabled, int timeout); }; /* @@ -1262,6 +1266,8 @@ struct wireless_dev { u8 bssid[ETH_ALEN]; u8 ssid[IEEE80211_MAX_SSID_LEN]; s8 default_key, default_mgmt_key; + bool ps; + int ps_timeout; } wext; #endif }; @@ -1606,6 +1612,13 @@ int cfg80211_wext_giwtxpower(struct net_device *dev, struct iw_request_info *info, union iwreq_data *data, char *keybuf); +int cfg80211_wext_siwpower(struct net_device *dev, + struct iw_request_info *info, + struct iw_param *wrq, char *extra); +int cfg80211_wext_giwpower(struct net_device *dev, + struct iw_request_info *info, + struct iw_param *wrq, char *extra); + /* * callbacks for asynchronous cfg80211 methods, notification * functions and BSS handling helpers diff --git a/net/mac80211/Kconfig b/net/mac80211/Kconfig index ba2643a43c73..41a32cd919af 100644 --- a/net/mac80211/Kconfig +++ b/net/mac80211/Kconfig @@ -14,22 +14,6 @@ config MAC80211 comment "CFG80211 needs to be enabled for MAC80211" depends on CFG80211=n -config MAC80211_DEFAULT_PS - bool "enable powersave by default" - depends on MAC80211 - default y - help - This option enables powersave mode by default. - - If this causes your applications to misbehave you should fix your - applications instead -- they need to register their network - latency requirement, see Documentation/power/pm_qos_interface.txt. - -config MAC80211_DEFAULT_PS_VALUE - int - default 1 if MAC80211_DEFAULT_PS - default 0 - menu "Rate control algorithm selection" depends on MAC80211 != n diff --git a/net/mac80211/cfg.c b/net/mac80211/cfg.c index 03de4024597a..8c7b2cdbeeda 100644 --- a/net/mac80211/cfg.c +++ b/net/mac80211/cfg.c @@ -1388,6 +1388,31 @@ int ieee80211_testmode_cmd(struct wiphy *wiphy, void *data, int len) } #endif +static int ieee80211_set_power_mgmt(struct wiphy *wiphy, struct net_device *dev, + bool enabled, int timeout) +{ + struct ieee80211_sub_if_data *sdata = IEEE80211_DEV_TO_SUB_IF(dev); + struct ieee80211_local *local = wdev_priv(dev->ieee80211_ptr); + struct ieee80211_conf *conf = &local->hw.conf; + + if (!(local->hw.flags & IEEE80211_HW_SUPPORTS_PS)) + return -EOPNOTSUPP; + + if (enabled == sdata->u.mgd.powersave && + timeout == conf->dynamic_ps_timeout) + return 0; + + sdata->u.mgd.powersave = enabled; + conf->dynamic_ps_timeout = timeout; + + if (local->hw.flags & IEEE80211_HW_SUPPORTS_DYNAMIC_PS) + ieee80211_hw_config(local, IEEE80211_CONF_CHANGE_PS); + + ieee80211_recalc_ps(local, -1); + + return 0; +} + struct cfg80211_ops mac80211_config_ops = { .add_virtual_intf = ieee80211_add_iface, .del_virtual_intf = ieee80211_del_iface, @@ -1431,4 +1456,5 @@ struct cfg80211_ops mac80211_config_ops = { .get_tx_power = ieee80211_get_tx_power, .rfkill_poll = ieee80211_rfkill_poll, CFG80211_TESTMODE_CMD(ieee80211_testmode_cmd) + .set_power_mgmt = ieee80211_set_power_mgmt, }; diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index fbb93a70ddc7..2a7860009f9c 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -2360,11 +2360,6 @@ void ieee80211_sta_setup_sdata(struct ieee80211_sub_if_data *sdata) ifmgd->flags |= IEEE80211_STA_WMM_ENABLED; hw_flags = sdata->local->hw.flags; - - if (hw_flags & IEEE80211_HW_SUPPORTS_PS) { - ifmgd->powersave = CONFIG_MAC80211_DEFAULT_PS_VALUE; - sdata->local->hw.conf.dynamic_ps_timeout = 500; - } } /* configuration hooks */ diff --git a/net/mac80211/wext.c b/net/mac80211/wext.c index d4e61dc903e8..f77929802c7a 100644 --- a/net/mac80211/wext.c +++ b/net/mac80211/wext.c @@ -255,72 +255,6 @@ static int ieee80211_ioctl_giwrate(struct net_device *dev, return 0; } -static int ieee80211_ioctl_siwpower(struct net_device *dev, - struct iw_request_info *info, - struct iw_param *wrq, - char *extra) -{ - struct ieee80211_sub_if_data *sdata = IEEE80211_DEV_TO_SUB_IF(dev); - struct ieee80211_local *local = wdev_priv(dev->ieee80211_ptr); - struct ieee80211_conf *conf = &local->hw.conf; - int timeout = 0; - bool ps; - - if (!(local->hw.flags & IEEE80211_HW_SUPPORTS_PS)) - return -EOPNOTSUPP; - - if (sdata->vif.type != NL80211_IFTYPE_STATION) - return -EINVAL; - - if (wrq->disabled) { - ps = false; - timeout = 0; - goto set; - } - - switch (wrq->flags & IW_POWER_MODE) { - case IW_POWER_ON: /* If not specified */ - case IW_POWER_MODE: /* If set all mask */ - case IW_POWER_ALL_R: /* If explicitely state all */ - ps = true; - break; - default: /* Otherwise we ignore */ - return -EINVAL; - } - - if (wrq->flags & ~(IW_POWER_MODE | IW_POWER_TIMEOUT)) - return -EINVAL; - - if (wrq->flags & IW_POWER_TIMEOUT) - timeout = wrq->value / 1000; - - set: - if (ps == sdata->u.mgd.powersave && timeout == conf->dynamic_ps_timeout) - return 0; - - sdata->u.mgd.powersave = ps; - conf->dynamic_ps_timeout = timeout; - - if (local->hw.flags & IEEE80211_HW_SUPPORTS_DYNAMIC_PS) - ieee80211_hw_config(local, IEEE80211_CONF_CHANGE_PS); - - ieee80211_recalc_ps(local, -1); - - return 0; -} - -static int ieee80211_ioctl_giwpower(struct net_device *dev, - struct iw_request_info *info, - union iwreq_data *wrqu, - char *extra) -{ - struct ieee80211_sub_if_data *sdata = IEEE80211_DEV_TO_SUB_IF(dev); - - wrqu->power.disabled = !sdata->u.mgd.powersave; - - return 0; -} - /* Get wireless statistics. Called by /proc/net/wireless and by SIOCGIWSTATS */ static struct iw_statistics *ieee80211_get_wireless_stats(struct net_device *dev) { @@ -436,8 +370,8 @@ static const iw_handler ieee80211_handler[] = (iw_handler) cfg80211_wext_giwretry, /* SIOCGIWRETRY */ (iw_handler) cfg80211_wext_siwencode, /* SIOCSIWENCODE */ (iw_handler) cfg80211_wext_giwencode, /* SIOCGIWENCODE */ - (iw_handler) ieee80211_ioctl_siwpower, /* SIOCSIWPOWER */ - (iw_handler) ieee80211_ioctl_giwpower, /* SIOCGIWPOWER */ + (iw_handler) cfg80211_wext_siwpower, /* SIOCSIWPOWER */ + (iw_handler) cfg80211_wext_giwpower, /* SIOCGIWPOWER */ (iw_handler) NULL, /* -- hole -- */ (iw_handler) NULL, /* -- hole -- */ (iw_handler) cfg80211_wext_siwgenie, /* SIOCSIWGENIE */ diff --git a/net/wireless/Kconfig b/net/wireless/Kconfig index 040263118a20..c6031d5b135f 100644 --- a/net/wireless/Kconfig +++ b/net/wireless/Kconfig @@ -26,6 +26,22 @@ config CFG80211_REG_DEBUG If unsure, say N. +config CFG80211_DEFAULT_PS + bool "enable powersave by default" + depends on CFG80211 + default y + help + This option enables powersave mode by default. + + If this causes your applications to misbehave you should fix your + applications instead -- they need to register their network + latency requirement, see Documentation/power/pm_qos_interface.txt. + +config CFG80211_DEFAULT_PS_VALUE + int + default 1 if CFG80211_DEFAULT_PS + default 0 + config CFG80211_DEBUGFS bool "cfg80211 DebugFS entries" depends on CFG80211 && DEBUG_FS diff --git a/net/wireless/core.c b/net/wireless/core.c index e2f80dd0e4a6..413d291d07d7 100644 --- a/net/wireless/core.c +++ b/net/wireless/core.c @@ -550,12 +550,21 @@ static int cfg80211_netdev_notifier_call(struct notifier_block * nb, } wdev->netdev = dev; wdev->sme_state = CFG80211_SME_IDLE; + mutex_unlock(&rdev->devlist_mtx); #ifdef CONFIG_WIRELESS_EXT wdev->wext.default_key = -1; wdev->wext.default_mgmt_key = -1; wdev->wext.connect.auth_type = NL80211_AUTHTYPE_AUTOMATIC; + wdev->wext.ps = CONFIG_CFG80211_DEFAULT_PS_VALUE; + wdev->wext.ps_timeout = 500; + if (rdev->ops->set_power_mgmt) + if (rdev->ops->set_power_mgmt(wdev->wiphy, dev, + wdev->wext.ps, + wdev->wext.ps_timeout)) { + /* assume this means it's off */ + wdev->wext.ps = false; + } #endif - mutex_unlock(&rdev->devlist_mtx); break; case NETDEV_GOING_DOWN: if (!wdev->ssid_len) diff --git a/net/wireless/wext-compat.c b/net/wireless/wext-compat.c index 02f052fc1808..2e1ab78fb0d7 100644 --- a/net/wireless/wext-compat.c +++ b/net/wireless/wext-compat.c @@ -987,3 +987,63 @@ int cfg80211_wext_giwauth(struct net_device *dev, return -EOPNOTSUPP; } EXPORT_SYMBOL_GPL(cfg80211_wext_giwauth); + +int cfg80211_wext_siwpower(struct net_device *dev, + struct iw_request_info *info, + struct iw_param *wrq, char *extra) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + struct cfg80211_registered_device *rdev = wiphy_to_dev(wdev->wiphy); + bool ps = wdev->wext.ps; + int timeout = wdev->wext.ps_timeout; + int err; + + if (wdev->iftype != NL80211_IFTYPE_STATION) + return -EINVAL; + + if (!rdev->ops->set_power_mgmt) + return -EOPNOTSUPP; + + if (wrq->disabled) { + ps = false; + } else { + switch (wrq->flags & IW_POWER_MODE) { + case IW_POWER_ON: /* If not specified */ + case IW_POWER_MODE: /* If set all mask */ + case IW_POWER_ALL_R: /* If explicitely state all */ + ps = true; + break; + default: /* Otherwise we ignore */ + return -EINVAL; + } + + if (wrq->flags & ~(IW_POWER_MODE | IW_POWER_TIMEOUT)) + return -EINVAL; + + if (wrq->flags & IW_POWER_TIMEOUT) + timeout = wrq->value / 1000; + } + + err = rdev->ops->set_power_mgmt(wdev->wiphy, dev, ps, timeout); + if (err) + return err; + + wdev->wext.ps = ps; + wdev->wext.ps_timeout = timeout; + + return 0; + +} +EXPORT_SYMBOL_GPL(cfg80211_wext_siwpower); + +int cfg80211_wext_giwpower(struct net_device *dev, + struct iw_request_info *info, + struct iw_param *wrq, char *extra) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + + wrq->disabled = !wdev->wext.ps; + + return 0; +} +EXPORT_SYMBOL_GPL(cfg80211_wext_giwpower); -- cgit v1.2.3 From ab737a4f7dbe57b12b73f482a7b973bf00b41942 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Wed, 1 Jul 2009 21:26:58 +0200 Subject: cfg80211: implement IWAP for WDS This implements siocsiwap/giwap for WDS mode. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- include/net/cfg80211.h | 10 ++++++++++ net/mac80211/cfg.c | 11 +++++++++++ net/mac80211/wext.c | 26 ++++---------------------- net/wireless/wext-compat.c | 46 ++++++++++++++++++++++++++++++++++++++++++++++ 4 files changed, 71 insertions(+), 22 deletions(-) (limited to 'include') diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index 82b7d804f6da..b396d11564bc 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -1018,6 +1018,9 @@ struct cfg80211_ops { enum tx_power_setting type, int dbm); int (*get_tx_power)(struct wiphy *wiphy, int *dbm); + int (*set_wds_peer)(struct wiphy *wiphy, struct net_device *dev, + u8 *addr); + void (*rfkill_poll)(struct wiphy *wiphy); #ifdef CONFIG_NL80211_TESTMODE @@ -1619,6 +1622,13 @@ int cfg80211_wext_giwpower(struct net_device *dev, struct iw_request_info *info, struct iw_param *wrq, char *extra); +int cfg80211_wds_wext_siwap(struct net_device *dev, + struct iw_request_info *info, + struct sockaddr *addr, char *extra); +int cfg80211_wds_wext_giwap(struct net_device *dev, + struct iw_request_info *info, + struct sockaddr *addr, char *extra); + /* * callbacks for asynchronous cfg80211 methods, notification * functions and BSS handling helpers diff --git a/net/mac80211/cfg.c b/net/mac80211/cfg.c index 8c7b2cdbeeda..2cf5bf6378e4 100644 --- a/net/mac80211/cfg.c +++ b/net/mac80211/cfg.c @@ -1369,6 +1369,16 @@ static int ieee80211_get_tx_power(struct wiphy *wiphy, int *dbm) return 0; } +static int ieee80211_set_wds_peer(struct wiphy *wiphy, struct net_device *dev, + u8 *addr) +{ + struct ieee80211_sub_if_data *sdata = IEEE80211_DEV_TO_SUB_IF(dev); + + memcpy(&sdata->u.wds.remote_addr, addr, ETH_ALEN); + + return 0; +} + static void ieee80211_rfkill_poll(struct wiphy *wiphy) { struct ieee80211_local *local = wiphy_priv(wiphy); @@ -1454,6 +1464,7 @@ struct cfg80211_ops mac80211_config_ops = { .set_wiphy_params = ieee80211_set_wiphy_params, .set_tx_power = ieee80211_set_tx_power, .get_tx_power = ieee80211_get_tx_power, + .set_wds_peer = ieee80211_set_wds_peer, .rfkill_poll = ieee80211_rfkill_poll, CFG80211_TESTMODE_CMD(ieee80211_testmode_cmd) .set_power_mgmt = ieee80211_set_power_mgmt, diff --git a/net/mac80211/wext.c b/net/mac80211/wext.c index f77929802c7a..4053d766af2d 100644 --- a/net/mac80211/wext.c +++ b/net/mac80211/wext.c @@ -140,23 +140,8 @@ static int ieee80211_ioctl_siwap(struct net_device *dev, if (sdata->vif.type == NL80211_IFTYPE_STATION) return cfg80211_mgd_wext_siwap(dev, info, ap_addr, extra); - if (sdata->vif.type == NL80211_IFTYPE_WDS) { - /* - * If it is necessary to update the WDS peer address - * while the interface is running, then we need to do - * more work here, namely if it is running we need to - * add a new and remove the old STA entry, this is - * normally handled by _open() and _stop(). - */ - if (netif_running(dev)) - return -EBUSY; - - memcpy(&sdata->u.wds.remote_addr, (u8 *) &ap_addr->sa_data, - ETH_ALEN); - - return 0; - } - + if (sdata->vif.type == NL80211_IFTYPE_WDS) + return cfg80211_wds_wext_siwap(dev, info, ap_addr, extra); return -EOPNOTSUPP; } @@ -173,11 +158,8 @@ static int ieee80211_ioctl_giwap(struct net_device *dev, if (sdata->vif.type == NL80211_IFTYPE_STATION) return cfg80211_mgd_wext_giwap(dev, info, ap_addr, extra); - if (sdata->vif.type == NL80211_IFTYPE_WDS) { - ap_addr->sa_family = ARPHRD_ETHER; - memcpy(&ap_addr->sa_data, sdata->u.wds.remote_addr, ETH_ALEN); - return 0; - } + if (sdata->vif.type == NL80211_IFTYPE_WDS) + return cfg80211_wds_wext_giwap(dev, info, ap_addr, extra); return -EOPNOTSUPP; } diff --git a/net/wireless/wext-compat.c b/net/wireless/wext-compat.c index 2e1ab78fb0d7..2f72dae2634f 100644 --- a/net/wireless/wext-compat.c +++ b/net/wireless/wext-compat.c @@ -1047,3 +1047,49 @@ int cfg80211_wext_giwpower(struct net_device *dev, return 0; } EXPORT_SYMBOL_GPL(cfg80211_wext_giwpower); + +int cfg80211_wds_wext_siwap(struct net_device *dev, + struct iw_request_info *info, + struct sockaddr *addr, char *extra) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + struct cfg80211_registered_device *rdev = wiphy_to_dev(wdev->wiphy); + int err; + + if (WARN_ON(wdev->iftype != NL80211_IFTYPE_WDS)) + return -EINVAL; + + if (addr->sa_family != ARPHRD_ETHER) + return -EINVAL; + + if (netif_running(dev)) + return -EBUSY; + + if (!rdev->ops->set_wds_peer) + return -EOPNOTSUPP; + + err = rdev->ops->set_wds_peer(wdev->wiphy, dev, (u8 *) &addr->sa_data); + if (err) + return err; + + memcpy(&wdev->wext.bssid, (u8 *) &addr->sa_data, ETH_ALEN); + + return 0; +} +EXPORT_SYMBOL_GPL(cfg80211_wds_wext_siwap); + +int cfg80211_wds_wext_giwap(struct net_device *dev, + struct iw_request_info *info, + struct sockaddr *addr, char *extra) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + + if (WARN_ON(wdev->iftype != NL80211_IFTYPE_WDS)) + return -EINVAL; + + addr->sa_family = ARPHRD_ETHER; + memcpy(&addr->sa_data, wdev->wext.bssid, ETH_ALEN); + + return 0; +} +EXPORT_SYMBOL_GPL(cfg80211_wds_wext_giwap); -- cgit v1.2.3 From 9930380f0bd8405fa6a51d644f3de88c30666519 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Wed, 1 Jul 2009 21:26:59 +0200 Subject: cfg80211: implement IWRATE For now, let's implement that using a very hackish way: simply mirror the wext API in the cfg80211 API. This will have to be changed later when we implement proper bitrate API. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- include/net/cfg80211.h | 32 +++++++++++++++++++ net/mac80211/cfg.c | 43 ++++++++++++++++++++++++++ net/mac80211/wext.c | 76 ++-------------------------------------------- net/wireless/wext-compat.c | 63 ++++++++++++++++++++++++++++++++++++++ 4 files changed, 140 insertions(+), 74 deletions(-) (limited to 'include') diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index b396d11564bc..579085564883 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -815,6 +815,26 @@ enum tx_power_setting { TX_POWER_FIXED, }; +/* + * cfg80211_bitrate_mask - masks for bitrate control + */ +struct cfg80211_bitrate_mask { +/* + * As discussed in Berlin, this struct really + * should look like this: + + struct { + u32 legacy; + u8 mcs[IEEE80211_HT_MCS_MASK_LEN]; + } control[IEEE80211_NUM_BANDS]; + + * Since we can always fix in-kernel users, let's keep + * it simpler for now: + */ + u32 fixed; /* fixed bitrate, 0 == not fixed */ + u32 maxrate; /* in kbps, 0 == no limit */ +}; + /** * struct cfg80211_ops - backend description for wireless configuration * @@ -1027,6 +1047,11 @@ struct cfg80211_ops { int (*testmode_cmd)(struct wiphy *wiphy, void *data, int len); #endif + int (*set_bitrate_mask)(struct wiphy *wiphy, + struct net_device *dev, + const u8 *peer, + const struct cfg80211_bitrate_mask *mask); + /* some temporary stuff to finish wext */ int (*set_power_mgmt)(struct wiphy *wiphy, struct net_device *dev, bool enabled, int timeout); @@ -1581,6 +1606,13 @@ int cfg80211_wext_giwauth(struct net_device *dev, struct ieee80211_channel *cfg80211_wext_freq(struct wiphy *wiphy, struct iw_freq *freq); +int cfg80211_wext_siwrate(struct net_device *dev, + struct iw_request_info *info, + struct iw_param *rate, char *extra); +int cfg80211_wext_giwrate(struct net_device *dev, + struct iw_request_info *info, + struct iw_param *rate, char *extra); + int cfg80211_wext_siwrts(struct net_device *dev, struct iw_request_info *info, struct iw_param *rts, char *extra); diff --git a/net/mac80211/cfg.c b/net/mac80211/cfg.c index 2cf5bf6378e4..028f6430879d 100644 --- a/net/mac80211/cfg.c +++ b/net/mac80211/cfg.c @@ -1423,6 +1423,48 @@ static int ieee80211_set_power_mgmt(struct wiphy *wiphy, struct net_device *dev, return 0; } +static int ieee80211_set_bitrate_mask(struct wiphy *wiphy, + struct net_device *dev, + const u8 *addr, + const struct cfg80211_bitrate_mask *mask) +{ + struct ieee80211_sub_if_data *sdata = IEEE80211_DEV_TO_SUB_IF(dev); + struct ieee80211_local *local = wdev_priv(dev->ieee80211_ptr); + int i, err = -EINVAL; + u32 target_rate; + struct ieee80211_supported_band *sband; + + sband = local->hw.wiphy->bands[local->hw.conf.channel->band]; + + /* target_rate = -1, rate->fixed = 0 means auto only, so use all rates + * target_rate = X, rate->fixed = 1 means only rate X + * target_rate = X, rate->fixed = 0 means all rates <= X */ + sdata->max_ratectrl_rateidx = -1; + sdata->force_unicast_rateidx = -1; + + if (mask->fixed) + target_rate = mask->fixed / 100; + else if (mask->maxrate) + target_rate = mask->maxrate / 100; + else + return 0; + + for (i=0; i< sband->n_bitrates; i++) { + struct ieee80211_rate *brate = &sband->bitrates[i]; + int this_rate = brate->bitrate; + + if (target_rate == this_rate) { + sdata->max_ratectrl_rateidx = i; + if (mask->fixed) + sdata->force_unicast_rateidx = i; + err = 0; + break; + } + } + + return err; +} + struct cfg80211_ops mac80211_config_ops = { .add_virtual_intf = ieee80211_add_iface, .del_virtual_intf = ieee80211_del_iface, @@ -1468,4 +1510,5 @@ struct cfg80211_ops mac80211_config_ops = { .rfkill_poll = ieee80211_rfkill_poll, CFG80211_TESTMODE_CMD(ieee80211_testmode_cmd) .set_power_mgmt = ieee80211_set_power_mgmt, + .set_bitrate_mask = ieee80211_set_bitrate_mask, }; diff --git a/net/mac80211/wext.c b/net/mac80211/wext.c index 4053d766af2d..244d830f5cfb 100644 --- a/net/mac80211/wext.c +++ b/net/mac80211/wext.c @@ -165,78 +165,6 @@ static int ieee80211_ioctl_giwap(struct net_device *dev, } -static int ieee80211_ioctl_siwrate(struct net_device *dev, - struct iw_request_info *info, - struct iw_param *rate, char *extra) -{ - struct ieee80211_local *local = wdev_priv(dev->ieee80211_ptr); - int i, err = -EINVAL; - u32 target_rate = rate->value / 100000; - struct ieee80211_sub_if_data *sdata; - struct ieee80211_supported_band *sband; - - sdata = IEEE80211_DEV_TO_SUB_IF(dev); - - sband = local->hw.wiphy->bands[local->hw.conf.channel->band]; - - /* target_rate = -1, rate->fixed = 0 means auto only, so use all rates - * target_rate = X, rate->fixed = 1 means only rate X - * target_rate = X, rate->fixed = 0 means all rates <= X */ - sdata->max_ratectrl_rateidx = -1; - sdata->force_unicast_rateidx = -1; - if (rate->value < 0) - return 0; - - for (i=0; i< sband->n_bitrates; i++) { - struct ieee80211_rate *brate = &sband->bitrates[i]; - int this_rate = brate->bitrate; - - if (target_rate == this_rate) { - sdata->max_ratectrl_rateidx = i; - if (rate->fixed) - sdata->force_unicast_rateidx = i; - err = 0; - break; - } - } - return err; -} - -static int ieee80211_ioctl_giwrate(struct net_device *dev, - struct iw_request_info *info, - struct iw_param *rate, char *extra) -{ - struct ieee80211_local *local = wdev_priv(dev->ieee80211_ptr); - struct sta_info *sta; - struct ieee80211_sub_if_data *sdata; - struct ieee80211_supported_band *sband; - - sdata = IEEE80211_DEV_TO_SUB_IF(dev); - - if (sdata->vif.type != NL80211_IFTYPE_STATION) - return -EOPNOTSUPP; - - sband = local->hw.wiphy->bands[local->hw.conf.channel->band]; - - rcu_read_lock(); - - sta = sta_info_get(local, sdata->u.mgd.bssid); - - if (sta && !(sta->last_tx_rate.flags & IEEE80211_TX_RC_MCS)) - rate->value = sband->bitrates[sta->last_tx_rate.idx].bitrate; - else - rate->value = 0; - - rcu_read_unlock(); - - if (!sta) - return -ENODEV; - - rate->value *= 100000; - - return 0; -} - /* Get wireless statistics. Called by /proc/net/wireless and by SIOCGIWSTATS */ static struct iw_statistics *ieee80211_get_wireless_stats(struct net_device *dev) { @@ -340,8 +268,8 @@ static const iw_handler ieee80211_handler[] = (iw_handler) NULL, /* SIOCGIWNICKN */ (iw_handler) NULL, /* -- hole -- */ (iw_handler) NULL, /* -- hole -- */ - (iw_handler) ieee80211_ioctl_siwrate, /* SIOCSIWRATE */ - (iw_handler) ieee80211_ioctl_giwrate, /* SIOCGIWRATE */ + (iw_handler) cfg80211_wext_siwrate, /* SIOCSIWRATE */ + (iw_handler) cfg80211_wext_giwrate, /* SIOCGIWRATE */ (iw_handler) cfg80211_wext_siwrts, /* SIOCSIWRTS */ (iw_handler) cfg80211_wext_giwrts, /* SIOCGIWRTS */ (iw_handler) cfg80211_wext_siwfrag, /* SIOCSIWFRAG */ diff --git a/net/wireless/wext-compat.c b/net/wireless/wext-compat.c index 2f72dae2634f..3a5f999703f1 100644 --- a/net/wireless/wext-compat.c +++ b/net/wireless/wext-compat.c @@ -1093,3 +1093,66 @@ int cfg80211_wds_wext_giwap(struct net_device *dev, return 0; } EXPORT_SYMBOL_GPL(cfg80211_wds_wext_giwap); + +int cfg80211_wext_siwrate(struct net_device *dev, + struct iw_request_info *info, + struct iw_param *rate, char *extra) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + struct cfg80211_registered_device *rdev = wiphy_to_dev(wdev->wiphy); + struct cfg80211_bitrate_mask mask; + + if (!rdev->ops->set_bitrate_mask) + return -EOPNOTSUPP; + + mask.fixed = 0; + mask.maxrate = 0; + + if (rate->value < 0) { + /* nothing */ + } else if (rate->fixed) { + mask.fixed = rate->value / 1000; /* kbps */ + } else { + mask.maxrate = rate->value / 1000; /* kbps */ + } + + return rdev->ops->set_bitrate_mask(wdev->wiphy, dev, NULL, &mask); +} +EXPORT_SYMBOL_GPL(cfg80211_wext_siwrate); + +int cfg80211_wext_giwrate(struct net_device *dev, + struct iw_request_info *info, + struct iw_param *rate, char *extra) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + struct cfg80211_registered_device *rdev = wiphy_to_dev(wdev->wiphy); + /* we are under RTNL - globally locked - so can use a static struct */ + static struct station_info sinfo; + u8 *addr; + int err; + + if (wdev->iftype != NL80211_IFTYPE_STATION) + return -EOPNOTSUPP; + + if (!rdev->ops->get_station) + return -EOPNOTSUPP; + + addr = wdev->wext.connect.bssid; + if (!addr) + return -EOPNOTSUPP; + + err = rdev->ops->get_station(&rdev->wiphy, dev, addr, &sinfo); + if (err) + return err; + + if (!(sinfo.filled & STATION_INFO_TX_BITRATE)) + return -EOPNOTSUPP; + + rate->value = 0; + + if (!(sinfo.txrate.flags & RATE_INFO_FLAGS_MCS)) + rate->value = 100000 * sinfo.txrate.legacy; + + return 0; +} +EXPORT_SYMBOL_GPL(cfg80211_wext_giwrate); -- cgit v1.2.3 From 8990646d2fafeacfacba4a4b1073a4216662089a Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Wed, 1 Jul 2009 21:27:00 +0200 Subject: cfg80211: implement get_wireless_stats By dropping the noise reporting, we can implement wireless stats in cfg80211. We also make the handler return NULL if we have no information, which is possible thanks to the recent wext change. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- include/net/cfg80211.h | 1 + net/mac80211/wext.c | 69 +--------------------------------------------- net/wireless/wext-compat.c | 59 +++++++++++++++++++++++++++++++++++++++ 3 files changed, 61 insertions(+), 68 deletions(-) (limited to 'include') diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index 579085564883..fe87819954a5 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -1646,6 +1646,7 @@ int cfg80211_wext_siwtxpower(struct net_device *dev, int cfg80211_wext_giwtxpower(struct net_device *dev, struct iw_request_info *info, union iwreq_data *data, char *keybuf); +struct iw_statistics *cfg80211_wireless_stats(struct net_device *dev); int cfg80211_wext_siwpower(struct net_device *dev, struct iw_request_info *info, diff --git a/net/mac80211/wext.c b/net/mac80211/wext.c index 244d830f5cfb..5acb8140ee58 100644 --- a/net/mac80211/wext.c +++ b/net/mac80211/wext.c @@ -165,73 +165,6 @@ static int ieee80211_ioctl_giwap(struct net_device *dev, } -/* Get wireless statistics. Called by /proc/net/wireless and by SIOCGIWSTATS */ -static struct iw_statistics *ieee80211_get_wireless_stats(struct net_device *dev) -{ - struct ieee80211_local *local = wdev_priv(dev->ieee80211_ptr); - struct iw_statistics *wstats = &local->wstats; - struct ieee80211_sub_if_data *sdata = IEEE80211_DEV_TO_SUB_IF(dev); - struct sta_info *sta = NULL; - - rcu_read_lock(); - - if (sdata->vif.type == NL80211_IFTYPE_STATION) - sta = sta_info_get(local, sdata->u.mgd.bssid); - - if (!sta) { - wstats->discard.fragment = 0; - wstats->discard.misc = 0; - wstats->qual.qual = 0; - wstats->qual.level = 0; - wstats->qual.noise = 0; - wstats->qual.updated = IW_QUAL_ALL_INVALID; - } else { - wstats->qual.updated = 0; - /* - * mirror what cfg80211 does for iwrange/scan results, - * otherwise userspace gets confused. - */ - if (local->hw.flags & (IEEE80211_HW_SIGNAL_UNSPEC | - IEEE80211_HW_SIGNAL_DBM)) { - wstats->qual.updated |= IW_QUAL_LEVEL_UPDATED; - wstats->qual.updated |= IW_QUAL_QUAL_UPDATED; - } else { - wstats->qual.updated |= IW_QUAL_LEVEL_INVALID; - wstats->qual.updated |= IW_QUAL_QUAL_INVALID; - } - - if (local->hw.flags & IEEE80211_HW_SIGNAL_UNSPEC) { - wstats->qual.level = sta->last_signal; - wstats->qual.qual = sta->last_signal; - } else if (local->hw.flags & IEEE80211_HW_SIGNAL_DBM) { - int sig = sta->last_signal; - - wstats->qual.updated |= IW_QUAL_DBM; - wstats->qual.level = sig; - if (sig < -110) - sig = -110; - else if (sig > -40) - sig = -40; - wstats->qual.qual = sig + 110; - } - - if (local->hw.flags & IEEE80211_HW_NOISE_DBM) { - /* - * This assumes that if driver reports noise, it also - * reports signal in dBm. - */ - wstats->qual.noise = sta->last_noise; - wstats->qual.updated |= IW_QUAL_NOISE_UPDATED; - } else { - wstats->qual.updated |= IW_QUAL_NOISE_INVALID; - } - } - - rcu_read_unlock(); - - return wstats; -} - /* Structures to export the Wireless Handlers */ static const iw_handler ieee80211_handler[] = @@ -298,5 +231,5 @@ const struct iw_handler_def ieee80211_iw_handler_def = { .num_standard = ARRAY_SIZE(ieee80211_handler), .standard = (iw_handler *) ieee80211_handler, - .get_wireless_stats = ieee80211_get_wireless_stats, + .get_wireless_stats = cfg80211_wireless_stats, }; diff --git a/net/wireless/wext-compat.c b/net/wireless/wext-compat.c index 3a5f999703f1..226cf8609079 100644 --- a/net/wireless/wext-compat.c +++ b/net/wireless/wext-compat.c @@ -1156,3 +1156,62 @@ int cfg80211_wext_giwrate(struct net_device *dev, return 0; } EXPORT_SYMBOL_GPL(cfg80211_wext_giwrate); + +/* Get wireless statistics. Called by /proc/net/wireless and by SIOCGIWSTATS */ +struct iw_statistics *cfg80211_wireless_stats(struct net_device *dev) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + struct cfg80211_registered_device *rdev = wiphy_to_dev(wdev->wiphy); + /* we are under RTNL - globally locked - so can use static structs */ + static struct iw_statistics wstats; + static struct station_info sinfo; + u8 *addr; + + if (dev->ieee80211_ptr->iftype != NL80211_IFTYPE_STATION) + return NULL; + + if (!rdev->ops->get_station) + return NULL; + + addr = wdev->wext.connect.bssid; + if (!addr) + return NULL; + + if (rdev->ops->get_station(&rdev->wiphy, dev, addr, &sinfo)) + return NULL; + + memset(&wstats, 0, sizeof(wstats)); + + switch (rdev->wiphy.signal_type) { + case CFG80211_SIGNAL_TYPE_MBM: + if (sinfo.filled & STATION_INFO_SIGNAL) { + int sig = sinfo.signal; + wstats.qual.updated |= IW_QUAL_LEVEL_UPDATED; + wstats.qual.updated |= IW_QUAL_QUAL_UPDATED; + wstats.qual.updated |= IW_QUAL_DBM; + wstats.qual.level = sig; + if (sig < -110) + sig = -110; + else if (sig > -40) + sig = -40; + wstats.qual.qual = sig + 110; + break; + } + case CFG80211_SIGNAL_TYPE_UNSPEC: + if (sinfo.filled & STATION_INFO_SIGNAL) { + wstats.qual.updated |= IW_QUAL_LEVEL_UPDATED; + wstats.qual.updated |= IW_QUAL_QUAL_UPDATED; + wstats.qual.level = sinfo.signal; + wstats.qual.qual = sinfo.signal; + break; + } + default: + wstats.qual.updated |= IW_QUAL_LEVEL_INVALID; + wstats.qual.updated |= IW_QUAL_QUAL_INVALID; + } + + wstats.qual.updated |= IW_QUAL_NOISE_INVALID; + + return &wstats; +} +EXPORT_SYMBOL_GPL(cfg80211_wireless_stats); -- cgit v1.2.3 From 517357c685ccc4b5783cc7dbdae8824ada19a97f Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Thu, 2 Jul 2009 17:18:40 +0200 Subject: cfg80211: assimilate and export ieee80211_bss_get_ie This function from mac80211 seems generally useful, and I will need it in cfg80211 soon. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- include/net/cfg80211.h | 9 +++++++++ net/mac80211/mlme.c | 25 +++---------------------- net/wireless/util.c | 21 +++++++++++++++++++++ 3 files changed, 33 insertions(+), 22 deletions(-) (limited to 'include') diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index fe87819954a5..eb026541f928 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -604,6 +604,15 @@ struct cfg80211_bss { u8 priv[0] __attribute__((__aligned__(sizeof(void *)))); }; +/** + * ieee80211_bss_get_ie - find IE with given ID + * @bss: the bss to search + * @ie: the IE ID + * Returns %NULL if not found. + */ +const u8 *ieee80211_bss_get_ie(struct cfg80211_bss *bss, u8 ie); + + /** * struct cfg80211_crypto_settings - Crypto settings * @wpa_versions: indicates which, if any, WPA versions are enabled diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index 2d9b6663253c..5748cda659c2 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -46,26 +46,6 @@ static int ecw2cw(int ecw) return (1 << ecw) - 1; } -static u8 *ieee80211_bss_get_ie(struct ieee80211_bss *bss, u8 ie) -{ - u8 *end, *pos; - - pos = bss->cbss.information_elements; - if (pos == NULL) - return NULL; - end = pos + bss->cbss.len_information_elements; - - while (pos + 1 < end) { - if (pos + 2 + pos[1] > end) - break; - if (pos[0] == ie) - return pos; - pos += 2 + pos[1]; - } - - return NULL; -} - static int ieee80211_compatible_rates(struct ieee80211_bss *bss, struct ieee80211_supported_band *sband, u32 *rates) @@ -181,7 +161,8 @@ static void ieee80211_send_assoc(struct ieee80211_sub_if_data *sdata) struct ieee80211_local *local = sdata->local; struct sk_buff *skb; struct ieee80211_mgmt *mgmt; - u8 *pos, *ies, *ht_ie; + u8 *pos; + const u8 *ies, *ht_ie; int i, len, count, rates_len, supp_rates_len; u16 capab; struct ieee80211_bss *bss; @@ -345,7 +326,7 @@ static void ieee80211_send_assoc(struct ieee80211_sub_if_data *sdata) */ if (wmm && (ifmgd->flags & IEEE80211_STA_WMM_ENABLED) && sband->ht_cap.ht_supported && - (ht_ie = ieee80211_bss_get_ie(bss, WLAN_EID_HT_INFORMATION)) && + (ht_ie = ieee80211_bss_get_ie(&bss->cbss, WLAN_EID_HT_INFORMATION)) && ht_ie[1] >= sizeof(struct ieee80211_ht_info) && (!(ifmgd->flags & IEEE80211_STA_DISABLE_11N))) { struct ieee80211_ht_info *ht_info = diff --git a/net/wireless/util.c b/net/wireless/util.c index 25550692dda6..28f8f96801d4 100644 --- a/net/wireless/util.c +++ b/net/wireless/util.c @@ -502,3 +502,24 @@ unsigned int cfg80211_classify8021d(struct sk_buff *skb) return dscp >> 5; } EXPORT_SYMBOL(cfg80211_classify8021d); + +const u8 *ieee80211_bss_get_ie(struct cfg80211_bss *bss, u8 ie) +{ + u8 *end, *pos; + + pos = bss->information_elements; + if (pos == NULL) + return NULL; + end = pos + bss->len_information_elements; + + while (pos + 1 < end) { + if (pos + 2 + pos[1] > end) + break; + if (pos[0] == ie) + return pos; + pos += 2 + pos[1]; + } + + return NULL; +} +EXPORT_SYMBOL(ieee80211_bss_get_ie); -- cgit v1.2.3 From 19957bb399e2722719c0e20c9ae91cf8b6aaff04 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Thu, 2 Jul 2009 17:20:43 +0200 Subject: cfg80211: keep track of BSSes In order to avoid problems with BSS structs going away while they're in use, I've long wanted to make cfg80211 keep track of them. Without the SME, that wasn't doable but now that we have the SME we can do this too. It can keep track of up to four separate authentications and one association, regardless of whether it's controlled by the cfg80211 SME or the userspace SME. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- include/net/cfg80211.h | 86 +++--------- net/mac80211/cfg.c | 22 ++- net/mac80211/mlme.c | 6 +- net/wireless/core.c | 5 +- net/wireless/core.h | 41 +++++- net/wireless/ibss.c | 12 +- net/wireless/mlme.c | 357 ++++++++++++++++++++++++++++++++++++++++++++++-- net/wireless/nl80211.c | 144 +++++++++---------- net/wireless/scan.c | 31 +---- net/wireless/sme.c | 156 ++++++++++++--------- net/wireless/wext-sme.c | 4 +- 11 files changed, 589 insertions(+), 275 deletions(-) (limited to 'include') diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index eb026541f928..ca986cc91098 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -584,7 +584,6 @@ enum cfg80211_signal_type { * is no guarantee that these are well-formed!) * @len_information_elements: total length of the information elements * @signal: signal strength value (type depends on the wiphy's signal_type) - * @hold: BSS should not expire * @free_priv: function pointer to free private data * @priv: private area for driver use, has at least wiphy->bss_priv_size bytes */ @@ -642,33 +641,17 @@ struct cfg80211_crypto_settings { * * This structure provides information needed to complete IEEE 802.11 * authentication. - * NOTE: This structure will likely change when more code from mac80211 is - * moved into cfg80211 so that non-mac80211 drivers can benefit from it, too. - * Before using this in a driver that does not use mac80211, it would be better - * to check the status of that work and better yet, volunteer to work on it. - * - * @chan: The channel to use or %NULL if not specified (auto-select based on - * scan results) - * @peer_addr: The address of the peer STA (AP BSSID in infrastructure case); - * this field is required to be present; if the driver wants to help with - * BSS selection, it should use (yet to be added) MLME event to allow user - * space SME to be notified of roaming candidate, so that the SME can then - * use the authentication request with the recommended BSSID and whatever - * other data may be needed for authentication/association - * @ssid: SSID or %NULL if not yet available - * @ssid_len: Length of ssid in octets + * + * @bss: The BSS to authenticate with. * @auth_type: Authentication type (algorithm) * @ie: Extra IEs to add to Authentication frame or %NULL * @ie_len: Length of ie buffer in octets */ struct cfg80211_auth_request { - struct ieee80211_channel *chan; - u8 *peer_addr; - const u8 *ssid; - size_t ssid_len; - enum nl80211_auth_type auth_type; + struct cfg80211_bss *bss; const u8 *ie; size_t ie_len; + enum nl80211_auth_type auth_type; }; /** @@ -676,32 +659,18 @@ struct cfg80211_auth_request { * * This structure provides information needed to complete IEEE 802.11 * (re)association. - * NOTE: This structure will likely change when more code from mac80211 is - * moved into cfg80211 so that non-mac80211 drivers can benefit from it, too. - * Before using this in a driver that does not use mac80211, it would be better - * to check the status of that work and better yet, volunteer to work on it. - * - * @chan: The channel to use or %NULL if not specified (auto-select based on - * scan results) - * @peer_addr: The address of the peer STA (AP BSSID); this field is required - * to be present and the STA must be in State 2 (authenticated) with the - * peer STA - * @ssid: SSID - * @ssid_len: Length of ssid in octets + * @bss: The BSS to associate with. * @ie: Extra IEs to add to (Re)Association Request frame or %NULL * @ie_len: Length of ie buffer in octets * @use_mfp: Use management frame protection (IEEE 802.11w) in this association * @crypto: crypto settings */ struct cfg80211_assoc_request { - struct ieee80211_channel *chan; - u8 *peer_addr; - const u8 *ssid; - size_t ssid_len; + struct cfg80211_bss *bss; const u8 *ie; size_t ie_len; - bool use_mfp; struct cfg80211_crypto_settings crypto; + bool use_mfp; }; /** @@ -710,16 +679,16 @@ struct cfg80211_assoc_request { * This structure provides information needed to complete IEEE 802.11 * deauthentication. * - * @peer_addr: The address of the peer STA (AP BSSID); this field is required - * to be present and the STA must be authenticated with the peer STA + * @bss: the BSS to deauthenticate from * @ie: Extra IEs to add to Deauthentication frame or %NULL * @ie_len: Length of ie buffer in octets + * @reason_code: The reason code for the deauthentication */ struct cfg80211_deauth_request { - u8 *peer_addr; - u16 reason_code; + struct cfg80211_bss *bss; const u8 *ie; size_t ie_len; + u16 reason_code; }; /** @@ -728,16 +697,16 @@ struct cfg80211_deauth_request { * This structure provides information needed to complete IEEE 802.11 * disassocation. * - * @peer_addr: The address of the peer STA (AP BSSID); this field is required - * to be present and the STA must be associated with the peer STA + * @bss: the BSS to disassociate from * @ie: Extra IEs to add to Disassociation frame or %NULL * @ie_len: Length of ie buffer in octets + * @reason_code: The reason code for the disassociation */ struct cfg80211_disassoc_request { - u8 *peer_addr; - u16 reason_code; + struct cfg80211_bss *bss; const u8 *ie; size_t ie_len; + u16 reason_code; }; /** @@ -1252,6 +1221,9 @@ extern void wiphy_free(struct wiphy *wiphy); /* internal struct */ struct cfg80211_conn; +struct cfg80211_internal_bss; + +#define MAX_AUTH_BSSES 4 /** * struct wireless_dev - wireless per-netdev state @@ -1281,7 +1253,6 @@ struct wireless_dev { struct net_device *netdev; /* currently used for IBSS and SME - might be rearranged later */ - struct cfg80211_bss *current_bss; u8 ssid[IEEE80211_MAX_SSID_LEN]; u8 ssid_len; enum { @@ -1291,6 +1262,10 @@ struct wireless_dev { } sme_state; struct cfg80211_conn *conn; + struct cfg80211_internal_bss *authtry_bsses[MAX_AUTH_BSSES]; + struct cfg80211_internal_bss *auth_bsses[MAX_AUTH_BSSES]; + struct cfg80211_internal_bss *current_bss; /* associated / joined */ + #ifdef CONFIG_WIRELESS_EXT /* wext data */ struct { @@ -1812,23 +1787,6 @@ void cfg80211_send_deauth(struct net_device *dev, const u8 *buf, size_t len, gfp */ void cfg80211_send_disassoc(struct net_device *dev, const u8 *buf, size_t len, gfp_t gfp); -/** - * cfg80211_hold_bss - exclude bss from expiration - * @bss: bss which should not expire - * - * In a case when the BSS is not updated but it shouldn't expire this - * function can be used to mark the BSS to be excluded from expiration. - */ -void cfg80211_hold_bss(struct cfg80211_bss *bss); - -/** - * cfg80211_unhold_bss - remove expiration exception from the BSS - * @bss: bss which can expire again - * - * This function marks the BSS to be expirable again. - */ -void cfg80211_unhold_bss(struct cfg80211_bss *bss); - /** * cfg80211_michael_mic_failure - notification of Michael MIC failure (TKIP) * @dev: network device diff --git a/net/mac80211/cfg.c b/net/mac80211/cfg.c index 7606571d4581..0f29cd0580c9 100644 --- a/net/mac80211/cfg.c +++ b/net/mac80211/cfg.c @@ -1173,6 +1173,7 @@ static int ieee80211_auth(struct wiphy *wiphy, struct net_device *dev, struct cfg80211_auth_request *req) { struct ieee80211_sub_if_data *sdata; + const u8 *ssid; sdata = IEEE80211_DEV_TO_SUB_IF(dev); @@ -1193,15 +1194,16 @@ static int ieee80211_auth(struct wiphy *wiphy, struct net_device *dev, return -EOPNOTSUPP; } - memcpy(sdata->u.mgd.bssid, req->peer_addr, ETH_ALEN); + memcpy(sdata->u.mgd.bssid, req->bss->bssid, ETH_ALEN); - sdata->local->oper_channel = req->chan; + sdata->local->oper_channel = req->bss->channel; ieee80211_hw_config(sdata->local, 0); - if (!req->ssid) + ssid = ieee80211_bss_get_ie(req->bss, WLAN_EID_SSID); + if (!ssid) return -EINVAL; - memcpy(sdata->u.mgd.ssid, req->ssid, req->ssid_len); - sdata->u.mgd.ssid_len = req->ssid_len; + sdata->u.mgd.ssid_len = *(ssid + 1); + memcpy(sdata->u.mgd.ssid, ssid + 2, sdata->u.mgd.ssid_len); kfree(sdata->u.mgd.sme_auth_ie); sdata->u.mgd.sme_auth_ie = NULL; @@ -1227,7 +1229,7 @@ static int ieee80211_assoc(struct wiphy *wiphy, struct net_device *dev, sdata = IEEE80211_DEV_TO_SUB_IF(dev); - if (memcmp(sdata->u.mgd.bssid, req->peer_addr, ETH_ALEN) != 0 || + if (memcmp(sdata->u.mgd.bssid, req->bss->bssid, ETH_ALEN) != 0 || !(sdata->u.mgd.flags & IEEE80211_STA_AUTHENTICATED)) return -ENOLINK; /* not authenticated */ @@ -1239,15 +1241,9 @@ static int ieee80211_assoc(struct wiphy *wiphy, struct net_device *dev, req->crypto.ciphers_pairwise[i] == WLAN_CIPHER_SUITE_WEP104) sdata->u.mgd.flags |= IEEE80211_STA_DISABLE_11N; - sdata->local->oper_channel = req->chan; + sdata->local->oper_channel = req->bss->channel; ieee80211_hw_config(sdata->local, 0); - if (!req->ssid) - return -EINVAL; - - memcpy(sdata->u.mgd.ssid, req->ssid, req->ssid_len); - sdata->u.mgd.ssid_len = req->ssid_len; - ret = ieee80211_sta_set_extra_ie(sdata, req->ie, req->ie_len); if (ret && ret != -EALREADY) return ret; diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index 5748cda659c2..aa1829ae431d 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -876,8 +876,6 @@ static void ieee80211_set_associated(struct ieee80211_sub_if_data *sdata, bss_info_changed |= ieee80211_handle_bss_capability(sdata, bss->cbss.capability, bss->has_erp_value, bss->erp_value); - cfg80211_hold_bss(&bss->cbss); - ieee80211_rx_bss_put(local, bss); } @@ -1031,10 +1029,8 @@ static void ieee80211_set_disassoc(struct ieee80211_sub_if_data *sdata, conf->channel->center_freq, ifmgd->ssid, ifmgd->ssid_len); - if (bss) { - cfg80211_unhold_bss(&bss->cbss); + if (bss) ieee80211_rx_bss_put(local, bss); - } if (self_disconnected) { if (deauth) diff --git a/net/wireless/core.c b/net/wireless/core.c index 5f6a8322bcb3..7b66cf15349a 100644 --- a/net/wireless/core.c +++ b/net/wireless/core.c @@ -583,15 +583,12 @@ static int cfg80211_netdev_notifier_call(struct notifier_block * nb, #endif cfg80211_disconnect(rdev, dev, WLAN_REASON_DEAUTH_LEAVING, true); + cfg80211_mlme_down(rdev, dev); break; default: break; } break; - case NETDEV_DOWN: - kfree(wdev->conn); - wdev->conn = NULL; - break; case NETDEV_UP: #ifdef CONFIG_WIRELESS_EXT switch (wdev->iftype) { diff --git a/net/wireless/core.h b/net/wireless/core.h index 5209acb0ff7e..82918f5896a5 100644 --- a/net/wireless/core.h +++ b/net/wireless/core.h @@ -110,12 +110,30 @@ struct cfg80211_internal_bss { struct rb_node rbn; unsigned long ts; struct kref ref; - bool hold, ies_allocated; + atomic_t hold; + bool ies_allocated; /* must be last because of priv member */ struct cfg80211_bss pub; }; +static inline struct cfg80211_internal_bss *bss_from_pub(struct cfg80211_bss *pub) +{ + return container_of(pub, struct cfg80211_internal_bss, pub); +} + +static inline void cfg80211_hold_bss(struct cfg80211_internal_bss *bss) +{ + atomic_inc(&bss->hold); +} + +static inline void cfg80211_unhold_bss(struct cfg80211_internal_bss *bss) +{ + int r = atomic_dec_return(&bss->hold); + WARN_ON(r < 0); +} + + struct cfg80211_registered_device *cfg80211_drv_by_wiphy_idx(int wiphy_idx); int get_wiphy_idx(struct wiphy *wiphy); @@ -176,6 +194,26 @@ void cfg80211_clear_ibss(struct net_device *dev, bool nowext); int cfg80211_leave_ibss(struct cfg80211_registered_device *rdev, struct net_device *dev, bool nowext); +/* MLME */ +int cfg80211_mlme_auth(struct cfg80211_registered_device *rdev, + struct net_device *dev, struct ieee80211_channel *chan, + enum nl80211_auth_type auth_type, const u8 *bssid, + const u8 *ssid, int ssid_len, + const u8 *ie, int ie_len); +int cfg80211_mlme_assoc(struct cfg80211_registered_device *rdev, + struct net_device *dev, struct ieee80211_channel *chan, + const u8 *bssid, const u8 *ssid, int ssid_len, + const u8 *ie, int ie_len, bool use_mfp, + struct cfg80211_crypto_settings *crypt); +int cfg80211_mlme_deauth(struct cfg80211_registered_device *rdev, + struct net_device *dev, const u8 *bssid, + const u8 *ie, int ie_len, u16 reason); +int cfg80211_mlme_disassoc(struct cfg80211_registered_device *rdev, + struct net_device *dev, const u8 *bssid, + const u8 *ie, int ie_len, u16 reason); +void cfg80211_mlme_down(struct cfg80211_registered_device *rdev, + struct net_device *dev); + /* SME */ int cfg80211_connect(struct cfg80211_registered_device *rdev, struct net_device *dev, @@ -193,5 +231,6 @@ void __cfg80211_disconnected(struct net_device *dev, gfp_t gfp, u8 *ie, size_t ie_len, u16 reason, bool from_ap); void cfg80211_sme_scan_done(struct net_device *dev); void cfg80211_sme_rx_auth(struct net_device *dev, const u8 *buf, size_t len); +void cfg80211_sme_disassoc(struct net_device *dev, int idx); #endif /* __NET_WIRELESS_CORE_H */ diff --git a/net/wireless/ibss.c b/net/wireless/ibss.c index 34b11eae30c8..c92b542d54b0 100644 --- a/net/wireless/ibss.c +++ b/net/wireless/ibss.c @@ -33,11 +33,11 @@ void cfg80211_ibss_joined(struct net_device *dev, const u8 *bssid, gfp_t gfp) if (wdev->current_bss) { cfg80211_unhold_bss(wdev->current_bss); - cfg80211_put_bss(wdev->current_bss); + cfg80211_put_bss(&wdev->current_bss->pub); } - cfg80211_hold_bss(bss); - wdev->current_bss = bss; + cfg80211_hold_bss(bss_from_pub(bss)); + wdev->current_bss = bss_from_pub(bss); nl80211_send_ibss_bssid(wiphy_to_dev(wdev->wiphy), dev, bssid, gfp); #ifdef CONFIG_WIRELESS_EXT @@ -78,7 +78,7 @@ void cfg80211_clear_ibss(struct net_device *dev, bool nowext) if (wdev->current_bss) { cfg80211_unhold_bss(wdev->current_bss); - cfg80211_put_bss(wdev->current_bss); + cfg80211_put_bss(&wdev->current_bss->pub); } wdev->current_bss = NULL; @@ -212,7 +212,7 @@ int cfg80211_ibss_wext_giwfreq(struct net_device *dev, return -EINVAL; if (wdev->current_bss) - chan = wdev->current_bss->channel; + chan = wdev->current_bss->pub.channel; else if (wdev->wext.ibss.channel) chan = wdev->wext.ibss.channel; @@ -352,7 +352,7 @@ int cfg80211_ibss_wext_giwap(struct net_device *dev, ap_addr->sa_family = ARPHRD_ETHER; if (wdev->current_bss) - memcpy(ap_addr->sa_data, wdev->current_bss->bssid, ETH_ALEN); + memcpy(ap_addr->sa_data, wdev->current_bss->pub.bssid, ETH_ALEN); else memcpy(ap_addr->sa_data, wdev->wext.ibss.bssid, ETH_ALEN); return 0; diff --git a/net/wireless/mlme.c b/net/wireless/mlme.c index 3427fe73d3c3..1a92bf7597bf 100644 --- a/net/wireless/mlme.c +++ b/net/wireless/mlme.c @@ -14,8 +14,32 @@ void cfg80211_send_rx_auth(struct net_device *dev, const u8 *buf, size_t len, gfp_t gfp) { - struct wiphy *wiphy = dev->ieee80211_ptr->wiphy; + struct wireless_dev *wdev = dev->ieee80211_ptr; + struct wiphy *wiphy = wdev->wiphy; struct cfg80211_registered_device *rdev = wiphy_to_dev(wiphy); + struct ieee80211_mgmt *mgmt = (struct ieee80211_mgmt *)buf; + u8 *bssid = mgmt->bssid; + int i; + u16 status = le16_to_cpu(mgmt->u.auth.status_code); + bool done = false; + + for (i = 0; i < MAX_AUTH_BSSES; i++) { + if (wdev->authtry_bsses[i] && + memcmp(wdev->authtry_bsses[i]->pub.bssid, bssid, + ETH_ALEN) == 0) { + if (status == WLAN_STATUS_SUCCESS) { + wdev->auth_bsses[i] = wdev->authtry_bsses[i]; + } else { + cfg80211_unhold_bss(wdev->authtry_bsses[i]); + cfg80211_put_bss(&wdev->authtry_bsses[i]->pub); + } + wdev->authtry_bsses[i] = NULL; + done = true; + break; + } + } + + WARN_ON(!done); nl80211_send_rx_auth(rdev, dev, buf, len, gfp); cfg80211_sme_rx_auth(dev, buf, len); @@ -30,7 +54,8 @@ void cfg80211_send_rx_assoc(struct net_device *dev, const u8 *buf, size_t len, g struct cfg80211_registered_device *rdev = wiphy_to_dev(wiphy); struct ieee80211_mgmt *mgmt = (struct ieee80211_mgmt *)buf; u8 *ie = mgmt->u.assoc_resp.variable; - int ieoffs = offsetof(struct ieee80211_mgmt, u.assoc_resp.variable); + int i, ieoffs = offsetof(struct ieee80211_mgmt, u.assoc_resp.variable); + bool done; status_code = le16_to_cpu(mgmt->u.assoc_resp.status_code); @@ -38,6 +63,20 @@ void cfg80211_send_rx_assoc(struct net_device *dev, const u8 *buf, size_t len, g cfg80211_connect_result(dev, mgmt->bssid, NULL, 0, ie, len - ieoffs, status_code, gfp); + + if (status_code == WLAN_STATUS_SUCCESS) { + for (i = 0; wdev->current_bss && i < MAX_AUTH_BSSES; i++) { + if (wdev->auth_bsses[i] == wdev->current_bss) { + cfg80211_unhold_bss(wdev->auth_bsses[i]); + cfg80211_put_bss(&wdev->auth_bsses[i]->pub); + wdev->auth_bsses[i] = NULL; + done = true; + break; + } + } + + WARN_ON(!done); + } } EXPORT_SYMBOL(cfg80211_send_rx_assoc); @@ -47,9 +86,45 @@ void cfg80211_send_deauth(struct net_device *dev, const u8 *buf, size_t len, gfp struct wiphy *wiphy = wdev->wiphy; struct cfg80211_registered_device *rdev = wiphy_to_dev(wiphy); struct ieee80211_mgmt *mgmt = (struct ieee80211_mgmt *)buf; + const u8 *bssid = mgmt->bssid; + int i; + bool done = false; nl80211_send_deauth(rdev, dev, buf, len, gfp); + if (wdev->current_bss && + memcmp(wdev->current_bss->pub.bssid, bssid, ETH_ALEN) == 0) { + done = true; + cfg80211_unhold_bss(wdev->current_bss); + cfg80211_put_bss(&wdev->current_bss->pub); + wdev->current_bss = NULL; + } else for (i = 0; i < MAX_AUTH_BSSES; i++) { + if (wdev->auth_bsses[i] && + memcmp(wdev->auth_bsses[i]->pub.bssid, bssid, ETH_ALEN) == 0) { + cfg80211_unhold_bss(wdev->auth_bsses[i]); + cfg80211_put_bss(&wdev->auth_bsses[i]->pub); + wdev->auth_bsses[i] = NULL; + done = true; + break; + } + if (wdev->authtry_bsses[i] && + memcmp(wdev->authtry_bsses[i]->pub.bssid, bssid, ETH_ALEN) == 0) { + cfg80211_unhold_bss(wdev->authtry_bsses[i]); + cfg80211_put_bss(&wdev->authtry_bsses[i]->pub); + wdev->authtry_bsses[i] = NULL; + done = true; + break; + } + } +/* + * mac80211 currently triggers this warning, + * so disable for now (it's harmless, just + * means that we got a spurious event) + + WARN_ON(!done); + + */ + if (wdev->sme_state == CFG80211_SME_CONNECTED) { u16 reason_code; bool from_ap; @@ -59,8 +134,6 @@ void cfg80211_send_deauth(struct net_device *dev, const u8 *buf, size_t len, gfp from_ap = memcmp(mgmt->da, dev->dev_addr, ETH_ALEN) == 0; __cfg80211_disconnected(dev, gfp, NULL, 0, reason_code, from_ap); - - wdev->sme_state = CFG80211_SME_IDLE; } else if (wdev->sme_state == CFG80211_SME_CONNECTING) { cfg80211_connect_result(dev, mgmt->bssid, NULL, 0, NULL, 0, WLAN_STATUS_UNSPECIFIED_FAILURE, gfp); @@ -74,21 +147,38 @@ void cfg80211_send_disassoc(struct net_device *dev, const u8 *buf, size_t len, g struct wiphy *wiphy = wdev->wiphy; struct cfg80211_registered_device *rdev = wiphy_to_dev(wiphy); struct ieee80211_mgmt *mgmt = (struct ieee80211_mgmt *)buf; + const u8 *bssid = mgmt->bssid; + int i; + u16 reason_code; + bool from_ap; + bool done = false; nl80211_send_disassoc(rdev, dev, buf, len, gfp); - if (wdev->sme_state == CFG80211_SME_CONNECTED) { - u16 reason_code; - bool from_ap; + if (!wdev->sme_state == CFG80211_SME_CONNECTED) + return; - reason_code = le16_to_cpu(mgmt->u.disassoc.reason_code); + if (wdev->current_bss && + memcmp(wdev->current_bss, bssid, ETH_ALEN) == 0) { + for (i = 0; i < MAX_AUTH_BSSES; i++) { + if (wdev->authtry_bsses[i] || wdev->auth_bsses[i]) + continue; + wdev->auth_bsses[i] = wdev->current_bss; + wdev->current_bss = NULL; + done = true; + cfg80211_sme_disassoc(dev, i); + break; + } + WARN_ON(!done); + } else + WARN_ON(1); - from_ap = memcmp(mgmt->da, dev->dev_addr, ETH_ALEN) == 0; - __cfg80211_disconnected(dev, gfp, NULL, 0, - reason_code, from_ap); - wdev->sme_state = CFG80211_SME_IDLE; - } + reason_code = le16_to_cpu(mgmt->u.disassoc.reason_code); + + from_ap = memcmp(mgmt->da, dev->dev_addr, ETH_ALEN) == 0; + __cfg80211_disconnected(dev, gfp, NULL, 0, + reason_code, from_ap); } EXPORT_SYMBOL(cfg80211_send_disassoc); @@ -97,11 +187,27 @@ void cfg80211_send_auth_timeout(struct net_device *dev, const u8 *addr, gfp_t gf struct wireless_dev *wdev = dev->ieee80211_ptr; struct wiphy *wiphy = wdev->wiphy; struct cfg80211_registered_device *rdev = wiphy_to_dev(wiphy); + int i; + bool done = false; + nl80211_send_auth_timeout(rdev, dev, addr, gfp); if (wdev->sme_state == CFG80211_SME_CONNECTING) cfg80211_connect_result(dev, addr, NULL, 0, NULL, 0, WLAN_STATUS_UNSPECIFIED_FAILURE, gfp); - wdev->sme_state = CFG80211_SME_IDLE; + + for (i = 0; addr && i < MAX_AUTH_BSSES; i++) { + if (wdev->authtry_bsses[i] && + memcmp(wdev->authtry_bsses[i]->pub.bssid, + addr, ETH_ALEN) == 0) { + cfg80211_unhold_bss(wdev->authtry_bsses[i]); + cfg80211_put_bss(&wdev->authtry_bsses[i]->pub); + wdev->authtry_bsses[i] = NULL; + done = true; + break; + } + } + + WARN_ON(!done); } EXPORT_SYMBOL(cfg80211_send_auth_timeout); @@ -110,11 +216,27 @@ void cfg80211_send_assoc_timeout(struct net_device *dev, const u8 *addr, gfp_t g struct wireless_dev *wdev = dev->ieee80211_ptr; struct wiphy *wiphy = wdev->wiphy; struct cfg80211_registered_device *rdev = wiphy_to_dev(wiphy); + int i; + bool done = false; + nl80211_send_assoc_timeout(rdev, dev, addr, gfp); if (wdev->sme_state == CFG80211_SME_CONNECTING) cfg80211_connect_result(dev, addr, NULL, 0, NULL, 0, WLAN_STATUS_UNSPECIFIED_FAILURE, gfp); - wdev->sme_state = CFG80211_SME_IDLE; + + for (i = 0; addr && i < MAX_AUTH_BSSES; i++) { + if (wdev->auth_bsses[i] && + memcmp(wdev->auth_bsses[i]->pub.bssid, + addr, ETH_ALEN) == 0) { + cfg80211_unhold_bss(wdev->auth_bsses[i]); + cfg80211_put_bss(&wdev->auth_bsses[i]->pub); + wdev->auth_bsses[i] = NULL; + done = true; + break; + } + } + + WARN_ON(!done); } EXPORT_SYMBOL(cfg80211_send_assoc_timeout); @@ -143,3 +265,208 @@ void cfg80211_michael_mic_failure(struct net_device *dev, const u8 *addr, nl80211_michael_mic_failure(rdev, dev, addr, key_type, key_id, tsc, gfp); } EXPORT_SYMBOL(cfg80211_michael_mic_failure); + +/* some MLME handling for userspace SME */ +int cfg80211_mlme_auth(struct cfg80211_registered_device *rdev, + struct net_device *dev, struct ieee80211_channel *chan, + enum nl80211_auth_type auth_type, const u8 *bssid, + const u8 *ssid, int ssid_len, + const u8 *ie, int ie_len) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + struct cfg80211_auth_request req; + struct cfg80211_internal_bss *bss; + int i, err, slot = -1, nfree = 0; + + memset(&req, 0, sizeof(req)); + + req.ie = ie; + req.ie_len = ie_len; + req.auth_type = auth_type; + req.bss = cfg80211_get_bss(&rdev->wiphy, chan, bssid, ssid, ssid_len, + WLAN_CAPABILITY_ESS, WLAN_CAPABILITY_ESS); + if (!req.bss) + return -ENOENT; + + bss = bss_from_pub(req.bss); + + for (i = 0; i < MAX_AUTH_BSSES; i++) { + if (bss == wdev->auth_bsses[i]) { + err = -EALREADY; + goto out; + } + } + + for (i = 0; i < MAX_AUTH_BSSES; i++) { + if (!wdev->auth_bsses[i] && !wdev->authtry_bsses[i]) { + slot = i; + nfree++; + } + } + + /* we need one free slot for disassoc and one for this auth */ + if (nfree < 2) { + err = -ENOSPC; + goto out; + } + + wdev->authtry_bsses[slot] = bss; + cfg80211_hold_bss(bss); + + err = rdev->ops->auth(&rdev->wiphy, dev, &req); + if (err) { + wdev->authtry_bsses[slot] = NULL; + cfg80211_unhold_bss(bss); + } + + out: + if (err) + cfg80211_put_bss(req.bss); + return err; +} + +int cfg80211_mlme_assoc(struct cfg80211_registered_device *rdev, + struct net_device *dev, struct ieee80211_channel *chan, + const u8 *bssid, const u8 *ssid, int ssid_len, + const u8 *ie, int ie_len, bool use_mfp, + struct cfg80211_crypto_settings *crypt) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + struct cfg80211_assoc_request req; + struct cfg80211_internal_bss *bss; + int i, err, slot = -1; + + memset(&req, 0, sizeof(req)); + + if (wdev->current_bss) + return -EALREADY; + + req.ie = ie; + req.ie_len = ie_len; + memcpy(&req.crypto, crypt, sizeof(req.crypto)); + req.use_mfp = use_mfp; + req.bss = cfg80211_get_bss(&rdev->wiphy, chan, bssid, ssid, ssid_len, + WLAN_CAPABILITY_ESS, WLAN_CAPABILITY_ESS); + if (!req.bss) + return -ENOENT; + + bss = bss_from_pub(req.bss); + + for (i = 0; i < MAX_AUTH_BSSES; i++) { + if (bss == wdev->auth_bsses[i]) { + slot = i; + break; + } + } + + if (slot < 0) { + err = -ENOTCONN; + goto out; + } + + err = rdev->ops->assoc(&rdev->wiphy, dev, &req); + out: + /* still a reference in wdev->auth_bsses[slot] */ + cfg80211_put_bss(req.bss); + return err; +} + +int cfg80211_mlme_deauth(struct cfg80211_registered_device *rdev, + struct net_device *dev, const u8 *bssid, + const u8 *ie, int ie_len, u16 reason) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + struct cfg80211_deauth_request req; + int i; + + memset(&req, 0, sizeof(req)); + req.reason_code = reason; + req.ie = ie; + req.ie_len = ie_len; + if (wdev->current_bss && + memcmp(wdev->current_bss->pub.bssid, bssid, ETH_ALEN) == 0) { + req.bss = &wdev->current_bss->pub; + } else for (i = 0; i < MAX_AUTH_BSSES; i++) { + if (wdev->auth_bsses[i] && + memcmp(bssid, wdev->auth_bsses[i]->pub.bssid, ETH_ALEN) == 0) { + req.bss = &wdev->auth_bsses[i]->pub; + break; + } + if (wdev->authtry_bsses[i] && + memcmp(bssid, wdev->authtry_bsses[i]->pub.bssid, ETH_ALEN) == 0) { + req.bss = &wdev->authtry_bsses[i]->pub; + break; + } + } + + if (!req.bss) + return -ENOTCONN; + + return rdev->ops->deauth(&rdev->wiphy, dev, &req); +} + +int cfg80211_mlme_disassoc(struct cfg80211_registered_device *rdev, + struct net_device *dev, const u8 *bssid, + const u8 *ie, int ie_len, u16 reason) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + struct cfg80211_disassoc_request req; + + memset(&req, 0, sizeof(req)); + req.reason_code = reason; + req.ie = ie; + req.ie_len = ie_len; + if (memcmp(wdev->current_bss->pub.bssid, bssid, ETH_ALEN) == 0) + req.bss = &wdev->current_bss->pub; + else + return -ENOTCONN; + + return rdev->ops->disassoc(&rdev->wiphy, dev, &req); +} + +void cfg80211_mlme_down(struct cfg80211_registered_device *rdev, + struct net_device *dev) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + struct cfg80211_deauth_request req; + int i; + + if (!rdev->ops->deauth) + return; + + memset(&req, 0, sizeof(req)); + req.reason_code = WLAN_REASON_DEAUTH_LEAVING; + req.ie = NULL; + req.ie_len = 0; + + if (wdev->current_bss) { + req.bss = &wdev->current_bss->pub; + rdev->ops->deauth(&rdev->wiphy, dev, &req); + if (wdev->current_bss) { + cfg80211_unhold_bss(wdev->current_bss); + cfg80211_put_bss(&wdev->current_bss->pub); + wdev->current_bss = NULL; + } + } + + for (i = 0; i < MAX_AUTH_BSSES; i++) { + if (wdev->auth_bsses[i]) { + req.bss = &wdev->auth_bsses[i]->pub; + rdev->ops->deauth(&rdev->wiphy, dev, &req); + if (wdev->auth_bsses[i]) { + cfg80211_unhold_bss(wdev->auth_bsses[i]); + cfg80211_put_bss(&wdev->auth_bsses[i]->pub); + wdev->auth_bsses[i] = NULL; + } + } + if (wdev->authtry_bsses[i]) { + req.bss = &wdev->authtry_bsses[i]->pub; + rdev->ops->deauth(&rdev->wiphy, dev, &req); + if (wdev->authtry_bsses[i]) { + cfg80211_unhold_bss(wdev->authtry_bsses[i]); + cfg80211_put_bss(&wdev->authtry_bsses[i]->pub); + wdev->authtry_bsses[i] = NULL; + } + } + } +} diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index 0008144b354b..aa2b3f35cc48 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -3044,9 +3044,10 @@ static int nl80211_authenticate(struct sk_buff *skb, struct genl_info *info) { struct cfg80211_registered_device *drv; struct net_device *dev; - struct cfg80211_auth_request req; - struct wiphy *wiphy; - int err; + struct ieee80211_channel *chan; + const u8 *bssid, *ssid, *ie = NULL; + int err, ssid_len, ie_len = 0; + enum nl80211_auth_type auth_type; if (!is_valid_ie_attr(info->attrs[NL80211_ATTR_IE])) return -EINVAL; @@ -3057,6 +3058,12 @@ static int nl80211_authenticate(struct sk_buff *skb, struct genl_info *info) if (!info->attrs[NL80211_ATTR_AUTH_TYPE]) return -EINVAL; + if (!info->attrs[NL80211_ATTR_SSID]) + return -EINVAL; + + if (!info->attrs[NL80211_ATTR_WIPHY_FREQ]) + return -EINVAL; + rtnl_lock(); err = get_drv_dev_by_info_ifindex(info->attrs, &drv, &dev); @@ -3078,38 +3085,30 @@ static int nl80211_authenticate(struct sk_buff *skb, struct genl_info *info) goto out; } - wiphy = &drv->wiphy; - memset(&req, 0, sizeof(req)); - - req.peer_addr = nla_data(info->attrs[NL80211_ATTR_MAC]); - - if (info->attrs[NL80211_ATTR_WIPHY_FREQ]) { - req.chan = ieee80211_get_channel( - wiphy, - nla_get_u32(info->attrs[NL80211_ATTR_WIPHY_FREQ])); - if (!req.chan) { - err = -EINVAL; - goto out; - } + bssid = nla_data(info->attrs[NL80211_ATTR_MAC]); + chan = ieee80211_get_channel(&drv->wiphy, + nla_get_u32(info->attrs[NL80211_ATTR_WIPHY_FREQ])); + if (!chan || (chan->flags & IEEE80211_CHAN_DISABLED)) { + err = -EINVAL; + goto out; } - if (info->attrs[NL80211_ATTR_SSID]) { - req.ssid = nla_data(info->attrs[NL80211_ATTR_SSID]); - req.ssid_len = nla_len(info->attrs[NL80211_ATTR_SSID]); - } + ssid = nla_data(info->attrs[NL80211_ATTR_SSID]); + ssid_len = nla_len(info->attrs[NL80211_ATTR_SSID]); if (info->attrs[NL80211_ATTR_IE]) { - req.ie = nla_data(info->attrs[NL80211_ATTR_IE]); - req.ie_len = nla_len(info->attrs[NL80211_ATTR_IE]); + ie = nla_data(info->attrs[NL80211_ATTR_IE]); + ie_len = nla_len(info->attrs[NL80211_ATTR_IE]); } - req.auth_type = nla_get_u32(info->attrs[NL80211_ATTR_AUTH_TYPE]); - if (!nl80211_valid_auth_type(req.auth_type)) { + auth_type = nla_get_u32(info->attrs[NL80211_ATTR_AUTH_TYPE]); + if (!nl80211_valid_auth_type(auth_type)) { err = -EINVAL; goto out; } - err = drv->ops->auth(&drv->wiphy, dev, &req); + err = cfg80211_mlme_auth(drv, dev, chan, auth_type, bssid, + ssid, ssid_len, ie, ie_len); out: cfg80211_put_dev(drv); @@ -3183,26 +3182,29 @@ static int nl80211_crypto_settings(struct genl_info *info, static int nl80211_associate(struct sk_buff *skb, struct genl_info *info) { - struct cfg80211_registered_device *drv; + struct cfg80211_registered_device *rdev; struct net_device *dev; - struct cfg80211_assoc_request req; - struct wiphy *wiphy; - int err; + struct cfg80211_crypto_settings crypto; + struct ieee80211_channel *chan; + const u8 *bssid, *ssid, *ie = NULL; + int err, ssid_len, ie_len = 0; + bool use_mfp = false; if (!is_valid_ie_attr(info->attrs[NL80211_ATTR_IE])) return -EINVAL; if (!info->attrs[NL80211_ATTR_MAC] || - !info->attrs[NL80211_ATTR_SSID]) + !info->attrs[NL80211_ATTR_SSID] || + !info->attrs[NL80211_ATTR_WIPHY_FREQ]) return -EINVAL; rtnl_lock(); - err = get_drv_dev_by_info_ifindex(info->attrs, &drv, &dev); + err = get_drv_dev_by_info_ifindex(info->attrs, &rdev, &dev); if (err) goto unlock_rtnl; - if (!drv->ops->assoc) { + if (!rdev->ops->assoc) { err = -EOPNOTSUPP; goto out; } @@ -3217,46 +3219,42 @@ static int nl80211_associate(struct sk_buff *skb, struct genl_info *info) goto out; } - wiphy = &drv->wiphy; - memset(&req, 0, sizeof(req)); - - req.peer_addr = nla_data(info->attrs[NL80211_ATTR_MAC]); + bssid = nla_data(info->attrs[NL80211_ATTR_MAC]); - if (info->attrs[NL80211_ATTR_WIPHY_FREQ]) { - req.chan = ieee80211_get_channel( - wiphy, - nla_get_u32(info->attrs[NL80211_ATTR_WIPHY_FREQ])); - if (!req.chan) { - err = -EINVAL; - goto out; - } + chan = ieee80211_get_channel(&rdev->wiphy, + nla_get_u32(info->attrs[NL80211_ATTR_WIPHY_FREQ])); + if (!chan || (chan->flags & IEEE80211_CHAN_DISABLED)) { + err = -EINVAL; + goto out; } - req.ssid = nla_data(info->attrs[NL80211_ATTR_SSID]); - req.ssid_len = nla_len(info->attrs[NL80211_ATTR_SSID]); + ssid = nla_data(info->attrs[NL80211_ATTR_SSID]); + ssid_len = nla_len(info->attrs[NL80211_ATTR_SSID]); if (info->attrs[NL80211_ATTR_IE]) { - req.ie = nla_data(info->attrs[NL80211_ATTR_IE]); - req.ie_len = nla_len(info->attrs[NL80211_ATTR_IE]); + ie = nla_data(info->attrs[NL80211_ATTR_IE]); + ie_len = nla_len(info->attrs[NL80211_ATTR_IE]); } if (info->attrs[NL80211_ATTR_USE_MFP]) { enum nl80211_mfp use_mfp = nla_get_u32(info->attrs[NL80211_ATTR_USE_MFP]); if (use_mfp == NL80211_MFP_REQUIRED) - req.use_mfp = true; + use_mfp = true; else if (use_mfp != NL80211_MFP_NO) { err = -EINVAL; goto out; } } - err = nl80211_crypto_settings(info, &req.crypto); + err = nl80211_crypto_settings(info, &crypto); if (!err) - err = drv->ops->assoc(&drv->wiphy, dev, &req); + err = cfg80211_mlme_assoc(rdev, dev, chan, bssid, ssid, + ssid_len, ie, ie_len, use_mfp, + &crypto); out: - cfg80211_put_dev(drv); + cfg80211_put_dev(rdev); dev_put(dev); unlock_rtnl: rtnl_unlock(); @@ -3267,9 +3265,9 @@ static int nl80211_deauthenticate(struct sk_buff *skb, struct genl_info *info) { struct cfg80211_registered_device *drv; struct net_device *dev; - struct cfg80211_deauth_request req; - struct wiphy *wiphy; - int err; + const u8 *ie = NULL, *bssid; + int err, ie_len = 0; + u16 reason_code; if (!is_valid_ie_attr(info->attrs[NL80211_ATTR_IE])) return -EINVAL; @@ -3301,24 +3299,21 @@ static int nl80211_deauthenticate(struct sk_buff *skb, struct genl_info *info) goto out; } - wiphy = &drv->wiphy; - memset(&req, 0, sizeof(req)); - - req.peer_addr = nla_data(info->attrs[NL80211_ATTR_MAC]); + bssid = nla_data(info->attrs[NL80211_ATTR_MAC]); - req.reason_code = nla_get_u16(info->attrs[NL80211_ATTR_REASON_CODE]); - if (req.reason_code == 0) { + reason_code = nla_get_u16(info->attrs[NL80211_ATTR_REASON_CODE]); + if (reason_code == 0) { /* Reason Code 0 is reserved */ err = -EINVAL; goto out; } if (info->attrs[NL80211_ATTR_IE]) { - req.ie = nla_data(info->attrs[NL80211_ATTR_IE]); - req.ie_len = nla_len(info->attrs[NL80211_ATTR_IE]); + ie = nla_data(info->attrs[NL80211_ATTR_IE]); + ie_len = nla_len(info->attrs[NL80211_ATTR_IE]); } - err = drv->ops->deauth(&drv->wiphy, dev, &req); + err = cfg80211_mlme_deauth(drv, dev, bssid, ie, ie_len, reason_code); out: cfg80211_put_dev(drv); @@ -3332,9 +3327,9 @@ static int nl80211_disassociate(struct sk_buff *skb, struct genl_info *info) { struct cfg80211_registered_device *drv; struct net_device *dev; - struct cfg80211_disassoc_request req; - struct wiphy *wiphy; - int err; + const u8 *ie = NULL, *bssid; + int err, ie_len = 0; + u16 reason_code; if (!is_valid_ie_attr(info->attrs[NL80211_ATTR_IE])) return -EINVAL; @@ -3366,24 +3361,21 @@ static int nl80211_disassociate(struct sk_buff *skb, struct genl_info *info) goto out; } - wiphy = &drv->wiphy; - memset(&req, 0, sizeof(req)); - - req.peer_addr = nla_data(info->attrs[NL80211_ATTR_MAC]); + bssid = nla_data(info->attrs[NL80211_ATTR_MAC]); - req.reason_code = nla_get_u16(info->attrs[NL80211_ATTR_REASON_CODE]); - if (req.reason_code == 0) { + reason_code = nla_get_u16(info->attrs[NL80211_ATTR_REASON_CODE]); + if (reason_code == 0) { /* Reason Code 0 is reserved */ err = -EINVAL; goto out; } if (info->attrs[NL80211_ATTR_IE]) { - req.ie = nla_data(info->attrs[NL80211_ATTR_IE]); - req.ie_len = nla_len(info->attrs[NL80211_ATTR_IE]); + ie = nla_data(info->attrs[NL80211_ATTR_IE]); + ie_len = nla_len(info->attrs[NL80211_ATTR_IE]); } - err = drv->ops->disassoc(&drv->wiphy, dev, &req); + err = cfg80211_mlme_disassoc(drv, dev, bssid, ie, ie_len, reason_code); out: cfg80211_put_dev(drv); diff --git a/net/wireless/scan.c b/net/wireless/scan.c index 82b33e708488..925399462a79 100644 --- a/net/wireless/scan.c +++ b/net/wireless/scan.c @@ -70,6 +70,8 @@ static void bss_release(struct kref *ref) if (bss->ies_allocated) kfree(bss->pub.information_elements); + BUG_ON(atomic_read(&bss->hold)); + kfree(bss); } @@ -92,8 +94,9 @@ void cfg80211_bss_expire(struct cfg80211_registered_device *dev) bool expired = false; list_for_each_entry_safe(bss, tmp, &dev->bss_list, list) { - if (bss->hold || - !time_after(jiffies, bss->ts + IEEE80211_SCAN_RESULT_EXPIRE)) + if (atomic_read(&bss->hold)) + continue; + if (!time_after(jiffies, bss->ts + IEEE80211_SCAN_RESULT_EXPIRE)) continue; list_del(&bss->list); rb_erase(&bss->rbn, &dev->bss_tree); @@ -553,30 +556,6 @@ void cfg80211_unlink_bss(struct wiphy *wiphy, struct cfg80211_bss *pub) } EXPORT_SYMBOL(cfg80211_unlink_bss); -void cfg80211_hold_bss(struct cfg80211_bss *pub) -{ - struct cfg80211_internal_bss *bss; - - if (!pub) - return; - - bss = container_of(pub, struct cfg80211_internal_bss, pub); - bss->hold = true; -} -EXPORT_SYMBOL(cfg80211_hold_bss); - -void cfg80211_unhold_bss(struct cfg80211_bss *pub) -{ - struct cfg80211_internal_bss *bss; - - if (!pub) - return; - - bss = container_of(pub, struct cfg80211_internal_bss, pub); - bss->hold = false; -} -EXPORT_SYMBOL(cfg80211_unhold_bss); - #ifdef CONFIG_WIRELESS_EXT int cfg80211_wext_siwscan(struct net_device *dev, struct iw_request_info *info, diff --git a/net/wireless/sme.c b/net/wireless/sme.c index d4e0b4065cbc..412161f7b08e 100644 --- a/net/wireless/sme.c +++ b/net/wireless/sme.c @@ -103,44 +103,37 @@ static int cfg80211_conn_scan(struct wireless_dev *wdev) static int cfg80211_conn_do_work(struct wireless_dev *wdev) { struct cfg80211_registered_device *drv = wiphy_to_dev(wdev->wiphy); - union { - struct cfg80211_auth_request auth_req; - struct cfg80211_assoc_request assoc_req; - } u; - - memset(&u, 0, sizeof(u)); + struct cfg80211_connect_params *params; + int err; if (!wdev->conn) return 0; + params = &wdev->conn->params; + switch (wdev->conn->state) { case CFG80211_CONN_SCAN_AGAIN: return cfg80211_conn_scan(wdev); case CFG80211_CONN_AUTHENTICATE_NEXT: - u.auth_req.chan = wdev->conn->params.channel; - u.auth_req.peer_addr = wdev->conn->params.bssid; - u.auth_req.ssid = wdev->conn->params.ssid; - u.auth_req.ssid_len = wdev->conn->params.ssid_len; - u.auth_req.auth_type = wdev->conn->params.auth_type; - u.auth_req.ie = NULL; - u.auth_req.ie_len = 0; - wdev->conn->state = CFG80211_CONN_AUTHENTICATING; BUG_ON(!drv->ops->auth); - return drv->ops->auth(wdev->wiphy, wdev->netdev, &u.auth_req); + wdev->conn->state = CFG80211_CONN_AUTHENTICATING; + return cfg80211_mlme_auth(drv, wdev->netdev, + params->channel, params->auth_type, + params->bssid, + params->ssid, params->ssid_len, + NULL, 0); case CFG80211_CONN_ASSOCIATE_NEXT: - u.assoc_req.chan = wdev->conn->params.channel; - u.assoc_req.peer_addr = wdev->conn->params.bssid; - u.assoc_req.ssid = wdev->conn->params.ssid; - u.assoc_req.ssid_len = wdev->conn->params.ssid_len; - u.assoc_req.ie = wdev->conn->params.ie; - u.assoc_req.ie_len = wdev->conn->params.ie_len; - u.assoc_req.use_mfp = false; - memcpy(&u.assoc_req.crypto, &wdev->conn->params.crypto, - sizeof(u.assoc_req.crypto)); - wdev->conn->state = CFG80211_CONN_ASSOCIATING; BUG_ON(!drv->ops->assoc); - return drv->ops->assoc(wdev->wiphy, wdev->netdev, - &u.assoc_req); + wdev->conn->state = CFG80211_CONN_ASSOCIATING; + err = cfg80211_mlme_assoc(drv, wdev->netdev, + params->channel, params->bssid, + params->ssid, params->ssid_len, + params->ie, params->ie_len, + false, ¶ms->crypto); + if (err) + cfg80211_mlme_deauth(drv, wdev->netdev, params->bssid, + NULL, 0, WLAN_REASON_DEAUTH_LEAVING); + return err; default: return 0; } @@ -186,7 +179,6 @@ static bool cfg80211_get_conn_bss(struct wireless_dev *wdev) wdev->conn->params.ssid_len, WLAN_CAPABILITY_ESS | WLAN_CAPABILITY_PRIVACY, capa); - if (!bss) return false; @@ -264,9 +256,11 @@ void cfg80211_sme_rx_auth(struct net_device *dev, const u8 *buf, size_t len) } wdev->conn->state = CFG80211_CONN_AUTHENTICATE_NEXT; schedule_work(&rdev->conn_work); - } else if (status_code != WLAN_STATUS_SUCCESS) + } else if (status_code != WLAN_STATUS_SUCCESS) { wdev->sme_state = CFG80211_SME_IDLE; - else if (wdev->sme_state == CFG80211_SME_CONNECTING && + kfree(wdev->conn); + wdev->conn = NULL; + } else if (wdev->sme_state == CFG80211_SME_CONNECTING && wdev->conn->state == CFG80211_CONN_AUTHENTICATING) { wdev->conn->state = CFG80211_CONN_ASSOCIATE_NEXT; schedule_work(&rdev->conn_work); @@ -330,10 +324,13 @@ static void __cfg80211_connect_result(struct net_device *dev, const u8 *bssid, if (wdev->current_bss) { cfg80211_unhold_bss(wdev->current_bss); - cfg80211_put_bss(wdev->current_bss); + cfg80211_put_bss(&wdev->current_bss->pub); wdev->current_bss = NULL; } + if (wdev->conn) + wdev->conn->state = CFG80211_CONN_IDLE; + if (status == WLAN_STATUS_SUCCESS) { bss = cfg80211_get_bss(wdev->wiphy, NULL, bssid, wdev->ssid, wdev->ssid_len, @@ -343,16 +340,15 @@ static void __cfg80211_connect_result(struct net_device *dev, const u8 *bssid, if (WARN_ON(!bss)) return; - cfg80211_hold_bss(bss); - wdev->current_bss = bss; + cfg80211_hold_bss(bss_from_pub(bss)); + wdev->current_bss = bss_from_pub(bss); wdev->sme_state = CFG80211_SME_CONNECTED; } else { wdev->sme_state = CFG80211_SME_IDLE; + kfree(wdev->conn); + wdev->conn = NULL; } - - if (wdev->conn) - wdev->conn->state = CFG80211_CONN_IDLE; } void cfg80211_connect_result(struct net_device *dev, const u8 *bssid, @@ -387,7 +383,7 @@ void cfg80211_roamed(struct net_device *dev, const u8 *bssid, } cfg80211_unhold_bss(wdev->current_bss); - cfg80211_put_bss(wdev->current_bss); + cfg80211_put_bss(&wdev->current_bss->pub); wdev->current_bss = NULL; bss = cfg80211_get_bss(wdev->wiphy, NULL, bssid, @@ -397,8 +393,8 @@ void cfg80211_roamed(struct net_device *dev, const u8 *bssid, if (WARN_ON(!bss)) return; - cfg80211_hold_bss(bss); - wdev->current_bss = bss; + cfg80211_hold_bss(bss_from_pub(bss)); + wdev->current_bss = bss_from_pub(bss); nl80211_send_roamed(wiphy_to_dev(wdev->wiphy), dev, bssid, req_ie, req_ie_len, resp_ie, resp_ie_len, gfp); @@ -440,7 +436,7 @@ void __cfg80211_disconnected(struct net_device *dev, gfp_t gfp, u8 *ie, if (wdev->current_bss) { cfg80211_unhold_bss(wdev->current_bss); - cfg80211_put_bss(wdev->current_bss); + cfg80211_put_bss(&wdev->current_bss->pub); } wdev->current_bss = NULL; @@ -449,6 +445,8 @@ void __cfg80211_disconnected(struct net_device *dev, gfp_t gfp, u8 *ie, if (wdev->conn) { kfree(wdev->conn->ie); wdev->conn->ie = NULL; + kfree(wdev->conn); + wdev->conn = NULL; } nl80211_send_disconnected(wiphy_to_dev(wdev->wiphy), dev, @@ -482,12 +480,12 @@ int cfg80211_connect(struct cfg80211_registered_device *rdev, if (!rdev->ops->auth || !rdev->ops->assoc) return -EOPNOTSUPP; - if (!wdev->conn) { - wdev->conn = kzalloc(sizeof(*wdev->conn), GFP_KERNEL); - if (!wdev->conn) - return -ENOMEM; - } else - memset(wdev->conn, 0, sizeof(*wdev->conn)); + if (WARN_ON(wdev->conn)) + return -EINPROGRESS; + + wdev->conn = kzalloc(sizeof(*wdev->conn), GFP_KERNEL); + if (!wdev->conn) + return -ENOMEM; /* * Copy all parameters, and treat explicitly IEs, BSSID, SSID. @@ -502,8 +500,11 @@ int cfg80211_connect(struct cfg80211_registered_device *rdev, wdev->conn->ie = kmemdup(connect->ie, connect->ie_len, GFP_KERNEL); wdev->conn->params.ie = wdev->conn->ie; - if (!wdev->conn->ie) + if (!wdev->conn->ie) { + kfree(wdev->conn); + wdev->conn = NULL; return -ENOMEM; + } } if (connect->auth_type == NL80211_AUTHTYPE_AUTOMATIC) { @@ -543,8 +544,11 @@ int cfg80211_connect(struct cfg80211_registered_device *rdev, wdev->conn->state = CFG80211_CONN_SCAN_AGAIN; } } - if (err) + if (err) { + kfree(wdev->conn); + wdev->conn = NULL; wdev->sme_state = CFG80211_SME_IDLE; + } return err; } else { @@ -572,31 +576,27 @@ int cfg80211_disconnect(struct cfg80211_registered_device *rdev, return -EINVAL; if (!rdev->ops->disconnect) { - struct cfg80211_deauth_request deauth; - u8 bssid[ETH_ALEN]; + if (!rdev->ops->deauth) + return -EOPNOTSUPP; - /* internal bug. */ - if (WARN_ON(!wdev->conn)) - return -EINVAL; + /* was it connected by userspace SME? */ + if (!wdev->conn) { + cfg80211_mlme_down(rdev, dev); + return 0; + } if (wdev->sme_state == CFG80211_SME_CONNECTING && (wdev->conn->state == CFG80211_CONN_SCANNING || wdev->conn->state == CFG80211_CONN_SCAN_AGAIN)) { wdev->sme_state = CFG80211_SME_IDLE; + kfree(wdev->conn); + wdev->conn = NULL; return 0; } - if (!rdev->ops->deauth) - return -EOPNOTSUPP; - - memset(&deauth, 0, sizeof(deauth)); - /* wdev->conn->params.bssid must be set if > SCANNING */ - memcpy(bssid, wdev->conn->params.bssid, ETH_ALEN); - deauth.peer_addr = bssid; - deauth.reason_code = reason; - - err = rdev->ops->deauth(&rdev->wiphy, dev, &deauth); + err = cfg80211_mlme_deauth(rdev, dev, wdev->conn->params.bssid, + NULL, 0, reason); if (err) return err; } else { @@ -614,3 +614,33 @@ int cfg80211_disconnect(struct cfg80211_registered_device *rdev, return 0; } + +void cfg80211_sme_disassoc(struct net_device *dev, int idx) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + struct cfg80211_registered_device *rdev = wiphy_to_dev(wdev->wiphy); + u8 bssid[ETH_ALEN]; + + if (!wdev->conn) + return; + + if (wdev->conn->state == CFG80211_CONN_IDLE) + return; + + /* + * Ok, so the association was made by this SME -- we don't + * want it any more so deauthenticate too. + */ + + if (!wdev->auth_bsses[idx]) + return; + + memcpy(bssid, wdev->auth_bsses[idx]->pub.bssid, ETH_ALEN); + if (cfg80211_mlme_deauth(rdev, dev, bssid, + NULL, 0, WLAN_REASON_DEAUTH_LEAVING)) { + /* whatever -- assume gone anyway */ + cfg80211_unhold_bss(wdev->auth_bsses[idx]); + cfg80211_put_bss(&wdev->auth_bsses[idx]->pub); + wdev->auth_bsses[idx] = NULL; + } +} diff --git a/net/wireless/wext-sme.c b/net/wireless/wext-sme.c index 3b531d572b69..fe1987acb891 100644 --- a/net/wireless/wext-sme.c +++ b/net/wireless/wext-sme.c @@ -93,7 +93,7 @@ int cfg80211_mgd_wext_giwfreq(struct net_device *dev, return -EINVAL; if (wdev->current_bss) - chan = wdev->current_bss->channel; + chan = wdev->current_bss->pub.channel; else if (wdev->wext.connect.channel) chan = wdev->wext.connect.channel; @@ -244,7 +244,7 @@ int cfg80211_mgd_wext_giwap(struct net_device *dev, ap_addr->sa_family = ARPHRD_ETHER; if (wdev->current_bss) - memcpy(ap_addr->sa_data, wdev->current_bss->bssid, ETH_ALEN); + memcpy(ap_addr->sa_data, wdev->current_bss->pub.bssid, ETH_ALEN); else if (wdev->wext.connect.bssid) memcpy(ap_addr->sa_data, wdev->wext.connect.bssid, ETH_ALEN); else -- cgit v1.2.3 From 1be491fca12ff599c37ceaf7e9042ebee9f0068e Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Sun, 5 Jul 2009 14:51:06 +0200 Subject: rfkill: prep for rfkill API changes We've designed the /dev/rfkill API in a way that we can increase the event struct by adding members at the end, should it become necessary. To validate the events, userspace and the kernel need to have the proper event size to check for -- when reading from the other end they need to verify that it's at least version 1 of the event API, with the current struct size, so define a constant for that and make the code a little more 'future proof'. Not that I expect that we'll have to change the event size any time soon, but it's better to write the code in a way that lends itself to extending. Due to the current size of the event struct, the code is currently equivalent, but should the event struct ever need to be increased the new code might not need changing. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- include/linux/rfkill.h | 14 ++++++++++++++ net/rfkill/core.c | 10 ++++++++-- 2 files changed, 22 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/include/linux/rfkill.h b/include/linux/rfkill.h index 2ce29831feb6..f3d5812693d6 100644 --- a/include/linux/rfkill.h +++ b/include/linux/rfkill.h @@ -82,6 +82,20 @@ struct rfkill_event { __u8 soft, hard; } __packed; +/* + * We are planning to be backward and forward compatible with changes + * to the event struct, by adding new, optional, members at the end. + * When reading an event (whether the kernel from userspace or vice + * versa) we need to accept anything that's at least as large as the + * version 1 event size, but might be able to accept other sizes in + * the future. + * + * One exception is the kernel -- we already have two event sizes in + * that we've made the 'hard' member optional since our only option + * is to ignore it anyway. + */ +#define RFKILL_EVENT_SIZE_V1 8 + /* ioctl for turning off rfkill-input (if present) */ #define RFKILL_IOC_MAGIC 'R' #define RFKILL_IOC_NOINPUT 1 diff --git a/net/rfkill/core.c b/net/rfkill/core.c index 79693fe2001e..47497c97c8d9 100644 --- a/net/rfkill/core.c +++ b/net/rfkill/core.c @@ -1076,10 +1076,16 @@ static ssize_t rfkill_fop_write(struct file *file, const char __user *buf, struct rfkill_event ev; /* we don't need the 'hard' variable but accept it */ - if (count < sizeof(ev) - 1) + if (count < RFKILL_EVENT_SIZE_V1 - 1) return -EINVAL; - if (copy_from_user(&ev, buf, sizeof(ev) - 1)) + /* + * Copy as much data as we can accept into our 'ev' buffer, + * but tell userspace how much we've copied so it can determine + * our API version even in a write() call, if it cares. + */ + count = min(count, sizeof(ev)); + if (copy_from_user(&ev, buf, count)) return -EFAULT; if (ev.op != RFKILL_OP_CHANGE && ev.op != RFKILL_OP_CHANGE_ALL) -- cgit v1.2.3 From 3e5d7649a64e558e4146ddfad4dfcf13fc65dd47 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Tue, 7 Jul 2009 14:37:26 +0200 Subject: cfg80211: let SME control reassociation vs. association Since we don't really know that well in the kernel, let's let the SME control whether it wants to use reassociation or not, by allowing it to give the previous BSSID in the associate() parameters. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- include/linux/nl80211.h | 5 +++++ include/net/cfg80211.h | 3 ++- net/mac80211/cfg.c | 6 ++++++ net/mac80211/mlme.c | 7 ------- net/wireless/core.h | 3 ++- net/wireless/mlme.c | 4 +++- net/wireless/nl80211.c | 10 +++++++--- net/wireless/sme.c | 8 +++++++- 8 files changed, 32 insertions(+), 14 deletions(-) (limited to 'include') diff --git a/include/linux/nl80211.h b/include/linux/nl80211.h index b34c17f52f3e..e496a2daf7ef 100644 --- a/include/linux/nl80211.h +++ b/include/linux/nl80211.h @@ -564,6 +564,9 @@ enum nl80211_commands { * @NL80211_ATTR_RESP_IE: (Re)association response information elements as * sent by peer, for ROAM and successful CONNECT events. * + * @NL80211_ATTR_PREV_BSSID: previous BSSID, to be used by in ASSOCIATE + * commands to specify using a reassociate frame + * * @NL80211_ATTR_MAX: highest attribute number currently defined * @__NL80211_ATTR_AFTER_LAST: internal use */ @@ -687,6 +690,8 @@ enum nl80211_attrs { NL80211_ATTR_REQ_IE, NL80211_ATTR_RESP_IE, + NL80211_ATTR_PREV_BSSID, + /* add attributes here, update the policy in nl80211.c */ __NL80211_ATTR_AFTER_LAST, diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index ca986cc91098..71847d3c2640 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -664,10 +664,11 @@ struct cfg80211_auth_request { * @ie_len: Length of ie buffer in octets * @use_mfp: Use management frame protection (IEEE 802.11w) in this association * @crypto: crypto settings + * @prev_bssid: previous BSSID, if not %NULL use reassociate frame */ struct cfg80211_assoc_request { struct cfg80211_bss *bss; - const u8 *ie; + const u8 *ie, *prev_bssid; size_t ie_len; struct cfg80211_crypto_settings crypto; bool use_mfp; diff --git a/net/mac80211/cfg.c b/net/mac80211/cfg.c index 0f29cd0580c9..e6d8860f26f2 100644 --- a/net/mac80211/cfg.c +++ b/net/mac80211/cfg.c @@ -1256,6 +1256,12 @@ static int ieee80211_assoc(struct wiphy *wiphy, struct net_device *dev, sdata->u.mgd.flags &= ~IEEE80211_STA_MFP_ENABLED; } + if (req->prev_bssid) { + sdata->u.mgd.flags |= IEEE80211_STA_PREV_BSSID_SET; + memcpy(sdata->u.mgd.prev_bssid, req->prev_bssid, ETH_ALEN); + } else + sdata->u.mgd.flags &= ~IEEE80211_STA_PREV_BSSID_SET; + if (req->crypto.control_port) sdata->u.mgd.flags |= IEEE80211_STA_CONTROL_PORT; else diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index aa1829ae431d..24486455e505 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -879,9 +879,6 @@ static void ieee80211_set_associated(struct ieee80211_sub_if_data *sdata, ieee80211_rx_bss_put(local, bss); } - ifmgd->flags |= IEEE80211_STA_PREV_BSSID_SET; - memcpy(ifmgd->prev_bssid, sdata->u.mgd.bssid, ETH_ALEN); - ifmgd->last_probe = jiffies; ieee80211_led_assoc(local, 1); @@ -1470,10 +1467,6 @@ static void ieee80211_rx_mgmt_assoc_resp(struct ieee80211_sub_if_data *sdata, if (status_code != WLAN_STATUS_SUCCESS) { printk(KERN_DEBUG "%s: AP denied association (code=%d)\n", sdata->dev->name, status_code); - /* if this was a reassociation, ensure we try a "full" - * association next time. This works around some broken APs - * which do not correctly reject reassociation requests. */ - ifmgd->flags &= ~IEEE80211_STA_PREV_BSSID_SET; cfg80211_send_rx_assoc(sdata->dev, (u8 *) mgmt, len, GFP_KERNEL); /* Wait for SME to decide what to do next */ diff --git a/net/wireless/core.h b/net/wireless/core.h index 82918f5896a5..4554453c116a 100644 --- a/net/wireless/core.h +++ b/net/wireless/core.h @@ -202,7 +202,8 @@ int cfg80211_mlme_auth(struct cfg80211_registered_device *rdev, const u8 *ie, int ie_len); int cfg80211_mlme_assoc(struct cfg80211_registered_device *rdev, struct net_device *dev, struct ieee80211_channel *chan, - const u8 *bssid, const u8 *ssid, int ssid_len, + const u8 *bssid, const u8 *prev_bssid, + const u8 *ssid, int ssid_len, const u8 *ie, int ie_len, bool use_mfp, struct cfg80211_crypto_settings *crypt); int cfg80211_mlme_deauth(struct cfg80211_registered_device *rdev, diff --git a/net/wireless/mlme.c b/net/wireless/mlme.c index 020f33b38467..087d3377958f 100644 --- a/net/wireless/mlme.c +++ b/net/wireless/mlme.c @@ -335,7 +335,8 @@ int cfg80211_mlme_auth(struct cfg80211_registered_device *rdev, int cfg80211_mlme_assoc(struct cfg80211_registered_device *rdev, struct net_device *dev, struct ieee80211_channel *chan, - const u8 *bssid, const u8 *ssid, int ssid_len, + const u8 *bssid, const u8 *prev_bssid, + const u8 *ssid, int ssid_len, const u8 *ie, int ie_len, bool use_mfp, struct cfg80211_crypto_settings *crypt) { @@ -353,6 +354,7 @@ int cfg80211_mlme_assoc(struct cfg80211_registered_device *rdev, req.ie_len = ie_len; memcpy(&req.crypto, crypt, sizeof(req.crypto)); req.use_mfp = use_mfp; + req.prev_bssid = prev_bssid; req.bss = cfg80211_get_bss(&rdev->wiphy, chan, bssid, ssid, ssid_len, WLAN_CAPABILITY_ESS, WLAN_CAPABILITY_ESS); if (!req.bss) diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index 723512b48f2e..44c520c264fc 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -71,6 +71,7 @@ static struct nla_policy nl80211_policy[NL80211_ATTR_MAX+1] __read_mostly = { [NL80211_ATTR_IFNAME] = { .type = NLA_NUL_STRING, .len = IFNAMSIZ-1 }, [NL80211_ATTR_MAC] = { .type = NLA_BINARY, .len = ETH_ALEN }, + [NL80211_ATTR_PREV_BSSID] = { .type = NLA_BINARY, .len = ETH_ALEN }, [NL80211_ATTR_KEY_DATA] = { .type = NLA_BINARY, .len = WLAN_MAX_KEY_LEN }, @@ -3187,7 +3188,7 @@ static int nl80211_associate(struct sk_buff *skb, struct genl_info *info) struct net_device *dev; struct cfg80211_crypto_settings crypto; struct ieee80211_channel *chan; - const u8 *bssid, *ssid, *ie = NULL; + const u8 *bssid, *ssid, *ie = NULL, *prev_bssid = NULL; int err, ssid_len, ie_len = 0; bool use_mfp = false; @@ -3248,10 +3249,13 @@ static int nl80211_associate(struct sk_buff *skb, struct genl_info *info) } } + if (info->attrs[NL80211_ATTR_PREV_BSSID]) + prev_bssid = nla_data(info->attrs[NL80211_ATTR_PREV_BSSID]); + err = nl80211_crypto_settings(info, &crypto, 1); if (!err) - err = cfg80211_mlme_assoc(rdev, dev, chan, bssid, ssid, - ssid_len, ie, ie_len, use_mfp, + err = cfg80211_mlme_assoc(rdev, dev, chan, bssid, prev_bssid, + ssid, ssid_len, ie, ie_len, use_mfp, &crypto); out: diff --git a/net/wireless/sme.c b/net/wireless/sme.c index 412161f7b08e..066a19ef9d73 100644 --- a/net/wireless/sme.c +++ b/net/wireless/sme.c @@ -125,8 +125,14 @@ static int cfg80211_conn_do_work(struct wireless_dev *wdev) case CFG80211_CONN_ASSOCIATE_NEXT: BUG_ON(!drv->ops->assoc); wdev->conn->state = CFG80211_CONN_ASSOCIATING; + /* + * We could, later, implement roaming here and then actually + * set prev_bssid to non-NULL. But then we need to be aware + * that some APs don't like that -- so we'd need to retry + * the association. + */ err = cfg80211_mlme_assoc(drv, wdev->netdev, - params->channel, params->bssid, + params->channel, params->bssid, NULL, params->ssid, params->ssid_len, params->ie, params->ie_len, false, ¶ms->crypto); -- cgit v1.2.3 From c238c8ac63f2d33ea5e7c0b9e9e0ccd8ae9a34e4 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Tue, 7 Jul 2009 03:56:06 +0200 Subject: cfg80211: dont use union for wext Otherwise it becomes very hard to reset the structs correctly since wext can be configured while the interface is down. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- include/net/cfg80211.h | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index 71847d3c2640..fe49833242d7 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -1270,10 +1270,8 @@ struct wireless_dev { #ifdef CONFIG_WIRELESS_EXT /* wext data */ struct { - union { - struct cfg80211_ibss_params ibss; - struct cfg80211_connect_params connect; - }; + struct cfg80211_ibss_params ibss; + struct cfg80211_connect_params connect; u8 *ie; size_t ie_len; u8 bssid[ETH_ALEN]; -- cgit v1.2.3 From cb0b4beb93d14429bf0c50fc1ab8e26348dca880 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Tue, 7 Jul 2009 03:56:07 +0200 Subject: cfg80211: mlme API must be able to sleep After the mac80211 mlme cleanup, we can require that the MLME functions in cfg80211 can sleep. This will simplify future work in cfg80211 a lot. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- include/net/cfg80211.h | 30 ++++++++++++++--------------- net/mac80211/mlme.c | 26 +++++++++---------------- net/wireless/mlme.c | 51 ++++++++++++++++++++++++++++++++------------------ 3 files changed, 56 insertions(+), 51 deletions(-) (limited to 'include') diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index fe49833242d7..60c1f11da45f 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -1721,70 +1721,68 @@ void cfg80211_unlink_bss(struct wiphy *wiphy, struct cfg80211_bss *bss); * @dev: network device * @buf: authentication frame (header + body) * @len: length of the frame data - * @gfp: allocation flags * * This function is called whenever an authentication has been processed in * station mode. The driver is required to call either this function or * cfg80211_send_auth_timeout() to indicate the result of cfg80211_ops::auth() - * call. + * call. This function may sleep. */ -void cfg80211_send_rx_auth(struct net_device *dev, const u8 *buf, size_t len, gfp_t gfp); +void cfg80211_send_rx_auth(struct net_device *dev, const u8 *buf, size_t len); /** * cfg80211_send_auth_timeout - notification of timed out authentication * @dev: network device * @addr: The MAC address of the device with which the authentication timed out - * @gfp: allocation flags + * + * This function may sleep. */ -void cfg80211_send_auth_timeout(struct net_device *dev, const u8 *addr, gfp_t gfp); +void cfg80211_send_auth_timeout(struct net_device *dev, const u8 *addr); /** * cfg80211_send_rx_assoc - notification of processed association * @dev: network device * @buf: (re)association response frame (header + body) * @len: length of the frame data - * @gfp: allocation flags * * This function is called whenever a (re)association response has been * processed in station mode. The driver is required to call either this * function or cfg80211_send_assoc_timeout() to indicate the result of - * cfg80211_ops::assoc() call. + * cfg80211_ops::assoc() call. This function may sleep. */ -void cfg80211_send_rx_assoc(struct net_device *dev, const u8 *buf, size_t len, gfp_t gfp); +void cfg80211_send_rx_assoc(struct net_device *dev, const u8 *buf, size_t len); /** * cfg80211_send_assoc_timeout - notification of timed out association * @dev: network device * @addr: The MAC address of the device with which the association timed out - * @gfp: allocation flags + * + * This function may sleep. */ -void cfg80211_send_assoc_timeout(struct net_device *dev, const u8 *addr, gfp_t gfp); +void cfg80211_send_assoc_timeout(struct net_device *dev, const u8 *addr); /** * cfg80211_send_deauth - notification of processed deauthentication * @dev: network device * @buf: deauthentication frame (header + body) * @len: length of the frame data - * @gfp: allocation flags * * This function is called whenever deauthentication has been processed in * station mode. This includes both received deauthentication frames and - * locally generated ones. + * locally generated ones. This function may sleep. */ -void cfg80211_send_deauth(struct net_device *dev, const u8 *buf, size_t len, gfp_t gfp); +void cfg80211_send_deauth(struct net_device *dev, const u8 *buf, size_t len); /** * cfg80211_send_disassoc - notification of processed disassociation * @dev: network device * @buf: disassociation response frame (header + body) * @len: length of the frame data - * @gfp: allocation flags * * This function is called whenever disassociation has been processed in * station mode. This includes both received disassociation frames and locally - * generated ones. + * generated ones. This function may sleep. */ -void cfg80211_send_disassoc(struct net_device *dev, const u8 *buf, size_t len, gfp_t gfp); +void cfg80211_send_disassoc(struct net_device *dev, const u8 *buf, size_t len); /** * cfg80211_michael_mic_failure - notification of Michael MIC failure (TKIP) diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index 108e8c9c60fd..15dbb57ab55e 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -412,11 +412,9 @@ static void ieee80211_send_deauth_disassoc(struct ieee80211_sub_if_data *sdata, mgmt->u.deauth.reason_code = cpu_to_le16(reason); if (stype == IEEE80211_STYPE_DEAUTH) - cfg80211_send_deauth(sdata->dev, (u8 *) mgmt, skb->len, - GFP_KERNEL); + cfg80211_send_deauth(sdata->dev, (u8 *) mgmt, skb->len); else - cfg80211_send_disassoc(sdata->dev, (u8 *) mgmt, skb->len, - GFP_KERNEL); + cfg80211_send_disassoc(sdata->dev, (u8 *) mgmt, skb->len); ieee80211_tx_skb(sdata, skb, ifmgd->flags & IEEE80211_STA_MFP_ENABLED); } @@ -1839,12 +1837,10 @@ static void ieee80211_sta_rx_queued_mgmt(struct ieee80211_sub_if_data *sdata, /* no action */ break; case RX_MGMT_CFG80211_DEAUTH: - cfg80211_send_deauth(sdata->dev, (u8 *) mgmt, - skb->len, GFP_KERNEL); + cfg80211_send_deauth(sdata->dev, (u8 *)mgmt, skb->len); break; case RX_MGMT_CFG80211_DISASSOC: - cfg80211_send_disassoc(sdata->dev, (u8 *) mgmt, - skb->len, GFP_KERNEL); + cfg80211_send_disassoc(sdata->dev, (u8 *)mgmt, skb->len); break; default: WARN(1, "unexpected: %d", rma); @@ -1893,12 +1889,10 @@ static void ieee80211_sta_rx_queued_mgmt(struct ieee80211_sub_if_data *sdata, /* no action */ break; case RX_MGMT_CFG80211_AUTH: - cfg80211_send_rx_auth(sdata->dev, (u8 *) mgmt, skb->len, - GFP_KERNEL); + cfg80211_send_rx_auth(sdata->dev, (u8 *) mgmt, skb->len); break; case RX_MGMT_CFG80211_ASSOC: - cfg80211_send_rx_assoc(sdata->dev, (u8 *) mgmt, skb->len, - GFP_KERNEL); + cfg80211_send_rx_assoc(sdata->dev, (u8 *) mgmt, skb->len); break; default: WARN(1, "unexpected: %d", rma); @@ -2026,13 +2020,11 @@ static void ieee80211_sta_work(struct work_struct *work) switch (wk->tries) { case RX_MGMT_CFG80211_AUTH_TO: cfg80211_send_auth_timeout(sdata->dev, - wk->bss->cbss.bssid, - GFP_KERNEL); + wk->bss->cbss.bssid); break; case RX_MGMT_CFG80211_ASSOC_TO: - cfg80211_send_auth_timeout(sdata->dev, - wk->bss->cbss.bssid, - GFP_KERNEL); + cfg80211_send_assoc_timeout(sdata->dev, + wk->bss->cbss.bssid); break; default: WARN(1, "unexpected: %d", wk->tries); diff --git a/net/wireless/mlme.c b/net/wireless/mlme.c index 087d3377958f..f7dc7524e145 100644 --- a/net/wireless/mlme.c +++ b/net/wireless/mlme.c @@ -12,7 +12,7 @@ #include "core.h" #include "nl80211.h" -void cfg80211_send_rx_auth(struct net_device *dev, const u8 *buf, size_t len, gfp_t gfp) +void cfg80211_send_rx_auth(struct net_device *dev, const u8 *buf, size_t len) { struct wireless_dev *wdev = dev->ieee80211_ptr; struct wiphy *wiphy = wdev->wiphy; @@ -23,6 +23,8 @@ void cfg80211_send_rx_auth(struct net_device *dev, const u8 *buf, size_t len, gf u16 status = le16_to_cpu(mgmt->u.auth.status_code); bool done = false; + might_sleep(); + for (i = 0; i < MAX_AUTH_BSSES; i++) { if (wdev->authtry_bsses[i] && memcmp(wdev->authtry_bsses[i]->pub.bssid, bssid, @@ -41,12 +43,12 @@ void cfg80211_send_rx_auth(struct net_device *dev, const u8 *buf, size_t len, gf WARN_ON(!done); - nl80211_send_rx_auth(rdev, dev, buf, len, gfp); + nl80211_send_rx_auth(rdev, dev, buf, len, GFP_KERNEL); cfg80211_sme_rx_auth(dev, buf, len); } EXPORT_SYMBOL(cfg80211_send_rx_auth); -void cfg80211_send_rx_assoc(struct net_device *dev, const u8 *buf, size_t len, gfp_t gfp) +void cfg80211_send_rx_assoc(struct net_device *dev, const u8 *buf, size_t len) { u16 status_code; struct wireless_dev *wdev = dev->ieee80211_ptr; @@ -57,12 +59,14 @@ void cfg80211_send_rx_assoc(struct net_device *dev, const u8 *buf, size_t len, g int i, ieoffs = offsetof(struct ieee80211_mgmt, u.assoc_resp.variable); bool done; + might_sleep(); + status_code = le16_to_cpu(mgmt->u.assoc_resp.status_code); - nl80211_send_rx_assoc(rdev, dev, buf, len, gfp); + nl80211_send_rx_assoc(rdev, dev, buf, len, GFP_KERNEL); cfg80211_connect_result(dev, mgmt->bssid, NULL, 0, ie, len - ieoffs, - status_code, gfp); + status_code, GFP_KERNEL); if (status_code == WLAN_STATUS_SUCCESS) { for (i = 0; wdev->current_bss && i < MAX_AUTH_BSSES; i++) { @@ -80,7 +84,7 @@ void cfg80211_send_rx_assoc(struct net_device *dev, const u8 *buf, size_t len, g } EXPORT_SYMBOL(cfg80211_send_rx_assoc); -void cfg80211_send_deauth(struct net_device *dev, const u8 *buf, size_t len, gfp_t gfp) +void cfg80211_send_deauth(struct net_device *dev, const u8 *buf, size_t len) { struct wireless_dev *wdev = dev->ieee80211_ptr; struct wiphy *wiphy = wdev->wiphy; @@ -90,7 +94,9 @@ void cfg80211_send_deauth(struct net_device *dev, const u8 *buf, size_t len, gfp int i; bool done = false; - nl80211_send_deauth(rdev, dev, buf, len, gfp); + might_sleep(); + + nl80211_send_deauth(rdev, dev, buf, len, GFP_KERNEL); if (wdev->current_bss && memcmp(wdev->current_bss->pub.bssid, bssid, ETH_ALEN) == 0) { @@ -132,16 +138,17 @@ void cfg80211_send_deauth(struct net_device *dev, const u8 *buf, size_t len, gfp reason_code = le16_to_cpu(mgmt->u.deauth.reason_code); from_ap = memcmp(mgmt->da, dev->dev_addr, ETH_ALEN) == 0; - __cfg80211_disconnected(dev, gfp, NULL, 0, + __cfg80211_disconnected(dev, GFP_KERNEL, NULL, 0, reason_code, from_ap); } else if (wdev->sme_state == CFG80211_SME_CONNECTING) { cfg80211_connect_result(dev, mgmt->bssid, NULL, 0, NULL, 0, - WLAN_STATUS_UNSPECIFIED_FAILURE, gfp); + WLAN_STATUS_UNSPECIFIED_FAILURE, + GFP_KERNEL); } } EXPORT_SYMBOL(cfg80211_send_deauth); -void cfg80211_send_disassoc(struct net_device *dev, const u8 *buf, size_t len, gfp_t gfp) +void cfg80211_send_disassoc(struct net_device *dev, const u8 *buf, size_t len) { struct wireless_dev *wdev = dev->ieee80211_ptr; struct wiphy *wiphy = wdev->wiphy; @@ -153,7 +160,9 @@ void cfg80211_send_disassoc(struct net_device *dev, const u8 *buf, size_t len, g bool from_ap; bool done = false; - nl80211_send_disassoc(rdev, dev, buf, len, gfp); + might_sleep(); + + nl80211_send_disassoc(rdev, dev, buf, len, GFP_KERNEL); if (!wdev->sme_state == CFG80211_SME_CONNECTED) return; @@ -177,12 +186,12 @@ void cfg80211_send_disassoc(struct net_device *dev, const u8 *buf, size_t len, g reason_code = le16_to_cpu(mgmt->u.disassoc.reason_code); from_ap = memcmp(mgmt->da, dev->dev_addr, ETH_ALEN) == 0; - __cfg80211_disconnected(dev, gfp, NULL, 0, + __cfg80211_disconnected(dev, GFP_KERNEL, NULL, 0, reason_code, from_ap); } EXPORT_SYMBOL(cfg80211_send_disassoc); -void cfg80211_send_auth_timeout(struct net_device *dev, const u8 *addr, gfp_t gfp) +void cfg80211_send_auth_timeout(struct net_device *dev, const u8 *addr) { struct wireless_dev *wdev = dev->ieee80211_ptr; struct wiphy *wiphy = wdev->wiphy; @@ -190,10 +199,13 @@ void cfg80211_send_auth_timeout(struct net_device *dev, const u8 *addr, gfp_t gf int i; bool done = false; - nl80211_send_auth_timeout(rdev, dev, addr, gfp); + might_sleep(); + + nl80211_send_auth_timeout(rdev, dev, addr, GFP_KERNEL); if (wdev->sme_state == CFG80211_SME_CONNECTING) cfg80211_connect_result(dev, addr, NULL, 0, NULL, 0, - WLAN_STATUS_UNSPECIFIED_FAILURE, gfp); + WLAN_STATUS_UNSPECIFIED_FAILURE, + GFP_KERNEL); for (i = 0; addr && i < MAX_AUTH_BSSES; i++) { if (wdev->authtry_bsses[i] && @@ -211,7 +223,7 @@ void cfg80211_send_auth_timeout(struct net_device *dev, const u8 *addr, gfp_t gf } EXPORT_SYMBOL(cfg80211_send_auth_timeout); -void cfg80211_send_assoc_timeout(struct net_device *dev, const u8 *addr, gfp_t gfp) +void cfg80211_send_assoc_timeout(struct net_device *dev, const u8 *addr) { struct wireless_dev *wdev = dev->ieee80211_ptr; struct wiphy *wiphy = wdev->wiphy; @@ -219,10 +231,13 @@ void cfg80211_send_assoc_timeout(struct net_device *dev, const u8 *addr, gfp_t g int i; bool done = false; - nl80211_send_assoc_timeout(rdev, dev, addr, gfp); + might_sleep(); + + nl80211_send_assoc_timeout(rdev, dev, addr, GFP_KERNEL); if (wdev->sme_state == CFG80211_SME_CONNECTING) cfg80211_connect_result(dev, addr, NULL, 0, NULL, 0, - WLAN_STATUS_UNSPECIFIED_FAILURE, gfp); + WLAN_STATUS_UNSPECIFIED_FAILURE, + GFP_KERNEL); for (i = 0; addr && i < MAX_AUTH_BSSES; i++) { if (wdev->auth_bsses[i] && -- cgit v1.2.3 From 667503ddcb96f3b10211f997fe55907fa7509841 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Tue, 7 Jul 2009 03:56:11 +0200 Subject: cfg80211: fix locking Over time, a lot of locking issues have crept into the smarts of cfg80211, so e.g. scan completion can race against a new scan, IBSS join can race against leaving an IBSS, etc. Introduce a new per-interface lock that protects most of the per-interface data that we need to keep track of, and sprinkle assertions about that lock everywhere. Some things now need to be offloaded to work structs so that we don't require being able to sleep in functions the drivers call. The exception to that are the MLME callbacks (rx_auth etc.) that currently only mac80211 calls because it was easier to do that there instead of in cfg80211, and future drivers implementing those calls will, if they ever exist, probably need to use a similar scheme like mac80211 anyway... In order to be able to handle _deauth and _disassoc properly, introduce a cookie passed to it that will determine locking requirements. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- include/net/cfg80211.h | 24 ++++- net/mac80211/cfg.c | 12 ++- net/mac80211/ieee80211_i.h | 6 +- net/mac80211/mlme.c | 25 +++-- net/wireless/core.c | 92 +++++++++++++++-- net/wireless/core.h | 100 +++++++++++++++++- net/wireless/ibss.c | 133 +++++++++++++++++++----- net/wireless/mlme.c | 214 +++++++++++++++++++++++++++++--------- net/wireless/nl80211.c | 8 +- net/wireless/nl80211.h | 2 +- net/wireless/scan.c | 30 +++++- net/wireless/sme.c | 252 +++++++++++++++++++++++++++++++++++---------- net/wireless/wext-sme.c | 125 +++++++++++++++------- 13 files changed, 823 insertions(+), 200 deletions(-) (limited to 'include') diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index 60c1f11da45f..83c2c727d71e 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -555,6 +555,7 @@ struct cfg80211_scan_request { /* internal */ struct wiphy *wiphy; int ifidx; + bool aborted; }; /** @@ -998,9 +999,11 @@ struct cfg80211_ops { int (*assoc)(struct wiphy *wiphy, struct net_device *dev, struct cfg80211_assoc_request *req); int (*deauth)(struct wiphy *wiphy, struct net_device *dev, - struct cfg80211_deauth_request *req); + struct cfg80211_deauth_request *req, + void *cookie); int (*disassoc)(struct wiphy *wiphy, struct net_device *dev, - struct cfg80211_disassoc_request *req); + struct cfg80211_disassoc_request *req, + void *cookie); int (*connect)(struct wiphy *wiphy, struct net_device *dev, struct cfg80211_connect_params *sme); @@ -1249,10 +1252,12 @@ struct wireless_dev { struct wiphy *wiphy; enum nl80211_iftype iftype; - /* private to the generic wireless code */ + /* the remainder of this struct should be private to cfg80211 */ struct list_head list; struct net_device *netdev; + struct mutex mtx; + /* currently used for IBSS and SME - might be rearranged later */ u8 ssid[IEEE80211_MAX_SSID_LEN]; u8 ssid_len; @@ -1263,6 +1268,9 @@ struct wireless_dev { } sme_state; struct cfg80211_conn *conn; + struct list_head event_list; + spinlock_t event_lock; + struct cfg80211_internal_bss *authtry_bsses[MAX_AUTH_BSSES]; struct cfg80211_internal_bss *auth_bsses[MAX_AUTH_BSSES]; struct cfg80211_internal_bss *current_bss; /* associated / joined */ @@ -1765,24 +1773,30 @@ void cfg80211_send_assoc_timeout(struct net_device *dev, const u8 *addr); * @dev: network device * @buf: deauthentication frame (header + body) * @len: length of the frame data + * @cookie: cookie from ->deauth if called within that callback, + * %NULL otherwise * * This function is called whenever deauthentication has been processed in * station mode. This includes both received deauthentication frames and * locally generated ones. This function may sleep. */ -void cfg80211_send_deauth(struct net_device *dev, const u8 *buf, size_t len); +void cfg80211_send_deauth(struct net_device *dev, const u8 *buf, size_t len, + void *cookie); /** * cfg80211_send_disassoc - notification of processed disassociation * @dev: network device * @buf: disassociation response frame (header + body) * @len: length of the frame data + * @cookie: cookie from ->disassoc if called within that callback, + * %NULL otherwise * * This function is called whenever disassociation has been processed in * station mode. This includes both received disassociation frames and locally * generated ones. This function may sleep. */ -void cfg80211_send_disassoc(struct net_device *dev, const u8 *buf, size_t len); +void cfg80211_send_disassoc(struct net_device *dev, const u8 *buf, size_t len, + void *cookie); /** * cfg80211_michael_mic_failure - notification of Michael MIC failure (TKIP) diff --git a/net/mac80211/cfg.c b/net/mac80211/cfg.c index 7cfc14e4ca07..36f8f245fa4c 100644 --- a/net/mac80211/cfg.c +++ b/net/mac80211/cfg.c @@ -1182,15 +1182,19 @@ static int ieee80211_assoc(struct wiphy *wiphy, struct net_device *dev, } static int ieee80211_deauth(struct wiphy *wiphy, struct net_device *dev, - struct cfg80211_deauth_request *req) + struct cfg80211_deauth_request *req, + void *cookie) { - return ieee80211_mgd_deauth(IEEE80211_DEV_TO_SUB_IF(dev), req); + return ieee80211_mgd_deauth(IEEE80211_DEV_TO_SUB_IF(dev), + req, cookie); } static int ieee80211_disassoc(struct wiphy *wiphy, struct net_device *dev, - struct cfg80211_disassoc_request *req) + struct cfg80211_disassoc_request *req, + void *cookie) { - return ieee80211_mgd_disassoc(IEEE80211_DEV_TO_SUB_IF(dev), req); + return ieee80211_mgd_disassoc(IEEE80211_DEV_TO_SUB_IF(dev), + req, cookie); } static int ieee80211_join_ibss(struct wiphy *wiphy, struct net_device *dev, diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h index 2e92bbd9b2d0..327aabc07abe 100644 --- a/net/mac80211/ieee80211_i.h +++ b/net/mac80211/ieee80211_i.h @@ -918,9 +918,11 @@ int ieee80211_mgd_auth(struct ieee80211_sub_if_data *sdata, int ieee80211_mgd_assoc(struct ieee80211_sub_if_data *sdata, struct cfg80211_assoc_request *req); int ieee80211_mgd_deauth(struct ieee80211_sub_if_data *sdata, - struct cfg80211_deauth_request *req); + struct cfg80211_deauth_request *req, + void *cookie); int ieee80211_mgd_disassoc(struct ieee80211_sub_if_data *sdata, - struct cfg80211_disassoc_request *req); + struct cfg80211_disassoc_request *req, + void *cookie); ieee80211_rx_result ieee80211_sta_rx_mgmt(struct ieee80211_sub_if_data *sdata, struct sk_buff *skb); void ieee80211_send_pspoll(struct ieee80211_local *local, diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index 15dbb57ab55e..c9db9646025b 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -386,7 +386,8 @@ static void ieee80211_send_assoc(struct ieee80211_sub_if_data *sdata, static void ieee80211_send_deauth_disassoc(struct ieee80211_sub_if_data *sdata, - const u8 *bssid, u16 stype, u16 reason) + const u8 *bssid, u16 stype, u16 reason, + void *cookie) { struct ieee80211_local *local = sdata->local; struct ieee80211_if_managed *ifmgd = &sdata->u.mgd; @@ -412,9 +413,9 @@ static void ieee80211_send_deauth_disassoc(struct ieee80211_sub_if_data *sdata, mgmt->u.deauth.reason_code = cpu_to_le16(reason); if (stype == IEEE80211_STYPE_DEAUTH) - cfg80211_send_deauth(sdata->dev, (u8 *) mgmt, skb->len); + cfg80211_send_deauth(sdata->dev, (u8 *)mgmt, skb->len, cookie); else - cfg80211_send_disassoc(sdata->dev, (u8 *) mgmt, skb->len); + cfg80211_send_disassoc(sdata->dev, (u8 *)mgmt, skb->len, cookie); ieee80211_tx_skb(sdata, skb, ifmgd->flags & IEEE80211_STA_MFP_ENABLED); } @@ -1837,10 +1838,12 @@ static void ieee80211_sta_rx_queued_mgmt(struct ieee80211_sub_if_data *sdata, /* no action */ break; case RX_MGMT_CFG80211_DEAUTH: - cfg80211_send_deauth(sdata->dev, (u8 *)mgmt, skb->len); + cfg80211_send_deauth(sdata->dev, (u8 *)mgmt, skb->len, + NULL); break; case RX_MGMT_CFG80211_DISASSOC: - cfg80211_send_disassoc(sdata->dev, (u8 *)mgmt, skb->len); + cfg80211_send_disassoc(sdata->dev, (u8 *)mgmt, skb->len, + NULL); break; default: WARN(1, "unexpected: %d", rma); @@ -2273,7 +2276,8 @@ int ieee80211_mgd_assoc(struct ieee80211_sub_if_data *sdata, } int ieee80211_mgd_deauth(struct ieee80211_sub_if_data *sdata, - struct cfg80211_deauth_request *req) + struct cfg80211_deauth_request *req, + void *cookie) { struct ieee80211_if_managed *ifmgd = &sdata->u.mgd; struct ieee80211_mgd_work *wk; @@ -2305,13 +2309,15 @@ int ieee80211_mgd_deauth(struct ieee80211_sub_if_data *sdata, mutex_unlock(&ifmgd->mtx); ieee80211_send_deauth_disassoc(sdata, bssid, - IEEE80211_STYPE_DEAUTH, req->reason_code); + IEEE80211_STYPE_DEAUTH, req->reason_code, + cookie); return 0; } int ieee80211_mgd_disassoc(struct ieee80211_sub_if_data *sdata, - struct cfg80211_disassoc_request *req) + struct cfg80211_disassoc_request *req, + void *cookie) { struct ieee80211_if_managed *ifmgd = &sdata->u.mgd; @@ -2331,6 +2337,7 @@ int ieee80211_mgd_disassoc(struct ieee80211_sub_if_data *sdata, mutex_unlock(&ifmgd->mtx); ieee80211_send_deauth_disassoc(sdata, req->bss->bssid, - IEEE80211_STYPE_DISASSOC, req->reason_code); + IEEE80211_STYPE_DISASSOC, req->reason_code, + cookie); return 0; } diff --git a/net/wireless/core.c b/net/wireless/core.c index c6813beded06..9c73769440a8 100644 --- a/net/wireless/core.c +++ b/net/wireless/core.c @@ -257,6 +257,71 @@ static void cfg80211_rfkill_sync_work(struct work_struct *work) cfg80211_rfkill_set_block(drv, rfkill_blocked(drv->rfkill)); } +static void cfg80211_process_events(struct wireless_dev *wdev) +{ + struct cfg80211_event *ev; + unsigned long flags; + + spin_lock_irqsave(&wdev->event_lock, flags); + while (!list_empty(&wdev->event_list)) { + ev = list_first_entry(&wdev->event_list, + struct cfg80211_event, list); + list_del(&ev->list); + spin_unlock_irqrestore(&wdev->event_lock, flags); + + wdev_lock(wdev); + switch (ev->type) { + case EVENT_CONNECT_RESULT: + __cfg80211_connect_result( + wdev->netdev, ev->cr.bssid, + ev->cr.req_ie, ev->cr.req_ie_len, + ev->cr.resp_ie, ev->cr.resp_ie_len, + ev->cr.status, + ev->cr.status == WLAN_STATUS_SUCCESS); + break; + case EVENT_ROAMED: + __cfg80211_roamed(wdev, ev->rm.bssid, + ev->rm.req_ie, ev->rm.req_ie_len, + ev->rm.resp_ie, ev->rm.resp_ie_len); + break; + case EVENT_DISCONNECTED: + __cfg80211_disconnected(wdev->netdev, + ev->dc.ie, ev->dc.ie_len, + ev->dc.reason, true); + break; + case EVENT_IBSS_JOINED: + __cfg80211_ibss_joined(wdev->netdev, ev->ij.bssid); + break; + } + wdev_unlock(wdev); + + kfree(ev); + + spin_lock_irqsave(&wdev->event_lock, flags); + } + spin_unlock_irqrestore(&wdev->event_lock, flags); +} + +static void cfg80211_event_work(struct work_struct *work) +{ + struct cfg80211_registered_device *rdev; + struct wireless_dev *wdev; + + rdev = container_of(work, struct cfg80211_registered_device, + event_work); + + rtnl_lock(); + cfg80211_lock_rdev(rdev); + mutex_lock(&rdev->devlist_mtx); + + list_for_each_entry(wdev, &rdev->netdev_list, list) + cfg80211_process_events(wdev); + + mutex_unlock(&rdev->devlist_mtx); + cfg80211_unlock_rdev(rdev); + rtnl_unlock(); +} + /* exported functions */ struct wiphy *wiphy_new(const struct cfg80211_ops *ops, int sizeof_priv) @@ -299,6 +364,7 @@ struct wiphy *wiphy_new(const struct cfg80211_ops *ops, int sizeof_priv) INIT_LIST_HEAD(&drv->netdev_list); spin_lock_init(&drv->bss_lock); INIT_LIST_HEAD(&drv->bss_list); + INIT_WORK(&drv->scan_done_wk, __cfg80211_scan_done); device_initialize(&drv->wiphy.dev); drv->wiphy.dev.class = &ieee80211_class; @@ -316,6 +382,7 @@ struct wiphy *wiphy_new(const struct cfg80211_ops *ops, int sizeof_priv) INIT_WORK(&drv->rfkill_sync, cfg80211_rfkill_sync_work); INIT_WORK(&drv->conn_work, cfg80211_conn_work); + INIT_WORK(&drv->event_work, cfg80211_event_work); /* * Initialize wiphy parameters to IEEE 802.11 MIB default values. @@ -477,6 +544,9 @@ void wiphy_unregister(struct wiphy *wiphy) mutex_unlock(&drv->mtx); cancel_work_sync(&drv->conn_work); + cancel_work_sync(&drv->scan_done_wk); + kfree(drv->scan_req); + flush_work(&drv->event_work); cfg80211_debugfs_drv_del(drv); @@ -535,6 +605,9 @@ static int cfg80211_netdev_notifier_call(struct notifier_block * nb, switch (state) { case NETDEV_REGISTER: + mutex_init(&wdev->mtx); + INIT_LIST_HEAD(&wdev->event_list); + spin_lock_init(&wdev->event_lock); mutex_lock(&rdev->devlist_mtx); list_add(&wdev->list, &rdev->netdev_list); if (sysfs_create_link(&dev->dev.kobj, &rdev->wiphy.dev.kobj, @@ -566,15 +639,17 @@ static int cfg80211_netdev_notifier_call(struct notifier_block * nb, cfg80211_leave_ibss(rdev, dev, true); break; case NL80211_IFTYPE_STATION: + wdev_lock(wdev); #ifdef CONFIG_WIRELESS_EXT kfree(wdev->wext.ie); wdev->wext.ie = NULL; wdev->wext.ie_len = 0; wdev->wext.connect.auth_type = NL80211_AUTHTYPE_AUTOMATIC; #endif - cfg80211_disconnect(rdev, dev, - WLAN_REASON_DEAUTH_LEAVING, true); + __cfg80211_disconnect(rdev, dev, + WLAN_REASON_DEAUTH_LEAVING, true); cfg80211_mlme_down(rdev, dev); + wdev_unlock(wdev); break; default: break; @@ -582,20 +657,24 @@ static int cfg80211_netdev_notifier_call(struct notifier_block * nb, break; case NETDEV_UP: #ifdef CONFIG_WIRELESS_EXT + cfg80211_lock_rdev(rdev); + wdev_lock(wdev); switch (wdev->iftype) { case NL80211_IFTYPE_ADHOC: if (wdev->wext.ibss.ssid_len) - cfg80211_join_ibss(rdev, dev, - &wdev->wext.ibss); + __cfg80211_join_ibss(rdev, dev, + &wdev->wext.ibss); break; case NL80211_IFTYPE_STATION: if (wdev->wext.connect.ssid_len) - cfg80211_connect(rdev, dev, - &wdev->wext.connect); + __cfg80211_connect(rdev, dev, + &wdev->wext.connect); break; default: break; } + wdev_unlock(wdev); + cfg80211_unlock_rdev(rdev); #endif break; case NETDEV_UNREGISTER: @@ -605,6 +684,7 @@ static int cfg80211_netdev_notifier_call(struct notifier_block * nb, list_del_init(&wdev->list); } mutex_unlock(&rdev->devlist_mtx); + mutex_destroy(&wdev->mtx); break; case NETDEV_PRE_UP: if (rfkill_blocked(rdev->rfkill)) diff --git a/net/wireless/core.h b/net/wireless/core.h index 92da612b3f9e..5ccd642e183b 100644 --- a/net/wireless/core.h +++ b/net/wireless/core.h @@ -57,12 +57,14 @@ struct cfg80211_registered_device { u32 bss_generation; struct cfg80211_scan_request *scan_req; /* protected by RTNL */ unsigned long suspend_at; + struct work_struct scan_done_wk; #ifdef CONFIG_NL80211_TESTMODE struct genl_info *testmode_info; #endif struct work_struct conn_work; + struct work_struct event_work; #ifdef CONFIG_CFG80211_DEBUGFS /* Debugfs entries */ @@ -170,12 +172,73 @@ struct wiphy *wiphy_idx_to_wiphy(int wiphy_idx); extern struct cfg80211_registered_device * cfg80211_get_dev_from_ifindex(int ifindex); +static inline void cfg80211_lock_rdev(struct cfg80211_registered_device *drv) +{ + mutex_lock(&drv->mtx); +} + static inline void cfg80211_unlock_rdev(struct cfg80211_registered_device *drv) { BUG_ON(IS_ERR(drv) || !drv); mutex_unlock(&drv->mtx); } +static inline void wdev_lock(struct wireless_dev *wdev) + __acquires(wdev) +{ + mutex_lock(&wdev->mtx); + __acquire(wdev->mtx); +} + +static inline void wdev_unlock(struct wireless_dev *wdev) + __releases(wdev) +{ + __release(wdev->mtx); + mutex_unlock(&wdev->mtx); +} + +#define ASSERT_RDEV_LOCK(rdev) WARN_ON(!mutex_is_locked(&(rdev)->mtx)); +#define ASSERT_WDEV_LOCK(wdev) WARN_ON(!mutex_is_locked(&(wdev)->mtx)); + +enum cfg80211_event_type { + EVENT_CONNECT_RESULT, + EVENT_ROAMED, + EVENT_DISCONNECTED, + EVENT_IBSS_JOINED, +}; + +struct cfg80211_event { + struct list_head list; + enum cfg80211_event_type type; + + union { + struct { + u8 bssid[ETH_ALEN]; + const u8 *req_ie; + const u8 *resp_ie; + size_t req_ie_len; + size_t resp_ie_len; + u16 status; + } cr; + struct { + u8 bssid[ETH_ALEN]; + const u8 *req_ie; + const u8 *resp_ie; + size_t req_ie_len; + size_t resp_ie_len; + } rm; + struct { + const u8 *ie; + size_t ie_len; + u16 reason; + } dc; + struct { + u8 bssid[ETH_ALEN]; + } ij; + }; +}; + + /* free object */ extern void cfg80211_dev_free(struct cfg80211_registered_device *drv); @@ -191,25 +254,46 @@ void cfg80211_bss_age(struct cfg80211_registered_device *dev, unsigned long age_secs); /* IBSS */ +int __cfg80211_join_ibss(struct cfg80211_registered_device *rdev, + struct net_device *dev, + struct cfg80211_ibss_params *params); int cfg80211_join_ibss(struct cfg80211_registered_device *rdev, struct net_device *dev, struct cfg80211_ibss_params *params); void cfg80211_clear_ibss(struct net_device *dev, bool nowext); int cfg80211_leave_ibss(struct cfg80211_registered_device *rdev, struct net_device *dev, bool nowext); +void __cfg80211_ibss_joined(struct net_device *dev, const u8 *bssid); /* MLME */ +int __cfg80211_mlme_auth(struct cfg80211_registered_device *rdev, + struct net_device *dev, + struct ieee80211_channel *chan, + enum nl80211_auth_type auth_type, + const u8 *bssid, + const u8 *ssid, int ssid_len, + const u8 *ie, int ie_len); int cfg80211_mlme_auth(struct cfg80211_registered_device *rdev, struct net_device *dev, struct ieee80211_channel *chan, enum nl80211_auth_type auth_type, const u8 *bssid, const u8 *ssid, int ssid_len, const u8 *ie, int ie_len); +int __cfg80211_mlme_assoc(struct cfg80211_registered_device *rdev, + struct net_device *dev, + struct ieee80211_channel *chan, + const u8 *bssid, const u8 *prev_bssid, + const u8 *ssid, int ssid_len, + const u8 *ie, int ie_len, bool use_mfp, + struct cfg80211_crypto_settings *crypt); int cfg80211_mlme_assoc(struct cfg80211_registered_device *rdev, struct net_device *dev, struct ieee80211_channel *chan, const u8 *bssid, const u8 *prev_bssid, const u8 *ssid, int ssid_len, const u8 *ie, int ie_len, bool use_mfp, struct cfg80211_crypto_settings *crypt); +int __cfg80211_mlme_deauth(struct cfg80211_registered_device *rdev, + struct net_device *dev, const u8 *bssid, + const u8 *ie, int ie_len, u16 reason); int cfg80211_mlme_deauth(struct cfg80211_registered_device *rdev, struct net_device *dev, const u8 *bssid, const u8 *ie, int ie_len, u16 reason); @@ -218,24 +302,38 @@ int cfg80211_mlme_disassoc(struct cfg80211_registered_device *rdev, const u8 *ie, int ie_len, u16 reason); void cfg80211_mlme_down(struct cfg80211_registered_device *rdev, struct net_device *dev); +void __cfg80211_connect_result(struct net_device *dev, const u8 *bssid, + const u8 *req_ie, size_t req_ie_len, + const u8 *resp_ie, size_t resp_ie_len, + u16 status, bool wextev); /* SME */ +int __cfg80211_connect(struct cfg80211_registered_device *rdev, + struct net_device *dev, + struct cfg80211_connect_params *connect); int cfg80211_connect(struct cfg80211_registered_device *rdev, struct net_device *dev, struct cfg80211_connect_params *connect); +int __cfg80211_disconnect(struct cfg80211_registered_device *rdev, + struct net_device *dev, u16 reason, + bool wextev); int cfg80211_disconnect(struct cfg80211_registered_device *rdev, struct net_device *dev, u16 reason, bool wextev); +void __cfg80211_roamed(struct wireless_dev *wdev, const u8 *bssid, + const u8 *req_ie, size_t req_ie_len, + const u8 *resp_ie, size_t resp_ie_len); void cfg80211_conn_work(struct work_struct *work); /* internal helpers */ int cfg80211_validate_key_settings(struct key_params *params, int key_idx, const u8 *mac_addr); -void __cfg80211_disconnected(struct net_device *dev, gfp_t gfp, u8 *ie, +void __cfg80211_disconnected(struct net_device *dev, const u8 *ie, size_t ie_len, u16 reason, bool from_ap); void cfg80211_sme_scan_done(struct net_device *dev); void cfg80211_sme_rx_auth(struct net_device *dev, const u8 *buf, size_t len); void cfg80211_sme_disassoc(struct net_device *dev, int idx); +void __cfg80211_scan_done(struct work_struct *wk); #endif /* __NET_WIRELESS_CORE_H */ diff --git a/net/wireless/ibss.c b/net/wireless/ibss.c index a5330c5a5477..99ef9364b7e8 100644 --- a/net/wireless/ibss.c +++ b/net/wireless/ibss.c @@ -10,7 +10,7 @@ #include "nl80211.h" -void cfg80211_ibss_joined(struct net_device *dev, const u8 *bssid, gfp_t gfp) +void __cfg80211_ibss_joined(struct net_device *dev, const u8 *bssid) { struct wireless_dev *wdev = dev->ieee80211_ptr; struct cfg80211_bss *bss; @@ -39,22 +39,45 @@ void cfg80211_ibss_joined(struct net_device *dev, const u8 *bssid, gfp_t gfp) cfg80211_hold_bss(bss_from_pub(bss)); wdev->current_bss = bss_from_pub(bss); - nl80211_send_ibss_bssid(wiphy_to_dev(wdev->wiphy), dev, bssid, gfp); + nl80211_send_ibss_bssid(wiphy_to_dev(wdev->wiphy), dev, bssid, + GFP_KERNEL); #ifdef CONFIG_WIRELESS_EXT memset(&wrqu, 0, sizeof(wrqu)); memcpy(wrqu.ap_addr.sa_data, bssid, ETH_ALEN); wireless_send_event(dev, SIOCGIWAP, &wrqu, NULL); #endif } + +void cfg80211_ibss_joined(struct net_device *dev, const u8 *bssid, gfp_t gfp) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + struct cfg80211_registered_device *rdev = wiphy_to_dev(wdev->wiphy); + struct cfg80211_event *ev; + unsigned long flags; + + ev = kzalloc(sizeof(*ev), gfp); + if (!ev) + return; + + ev->type = EVENT_IBSS_JOINED; + memcpy(ev->cr.bssid, bssid, ETH_ALEN); + + spin_lock_irqsave(&wdev->event_lock, flags); + list_add_tail(&ev->list, &wdev->event_list); + spin_unlock_irqrestore(&wdev->event_lock, flags); + schedule_work(&rdev->event_work); +} EXPORT_SYMBOL(cfg80211_ibss_joined); -int cfg80211_join_ibss(struct cfg80211_registered_device *rdev, - struct net_device *dev, - struct cfg80211_ibss_params *params) +int __cfg80211_join_ibss(struct cfg80211_registered_device *rdev, + struct net_device *dev, + struct cfg80211_ibss_params *params) { struct wireless_dev *wdev = dev->ieee80211_ptr; int err; + ASSERT_WDEV_LOCK(wdev); + if (wdev->ssid_len) return -EALREADY; @@ -72,10 +95,26 @@ int cfg80211_join_ibss(struct cfg80211_registered_device *rdev, return 0; } -void cfg80211_clear_ibss(struct net_device *dev, bool nowext) +int cfg80211_join_ibss(struct cfg80211_registered_device *rdev, + struct net_device *dev, + struct cfg80211_ibss_params *params) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + int err; + + wdev_lock(wdev); + err = __cfg80211_join_ibss(rdev, dev, params); + wdev_unlock(wdev); + + return err; +} + +static void __cfg80211_clear_ibss(struct net_device *dev, bool nowext) { struct wireless_dev *wdev = dev->ieee80211_ptr; + ASSERT_WDEV_LOCK(wdev); + if (wdev->current_bss) { cfg80211_unhold_bss(wdev->current_bss); cfg80211_put_bss(&wdev->current_bss->pub); @@ -89,12 +128,23 @@ void cfg80211_clear_ibss(struct net_device *dev, bool nowext) #endif } -int cfg80211_leave_ibss(struct cfg80211_registered_device *rdev, - struct net_device *dev, bool nowext) +void cfg80211_clear_ibss(struct net_device *dev, bool nowext) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + + wdev_lock(wdev); + __cfg80211_clear_ibss(dev, nowext); + wdev_unlock(wdev); +} + +static int __cfg80211_leave_ibss(struct cfg80211_registered_device *rdev, + struct net_device *dev, bool nowext) { struct wireless_dev *wdev = dev->ieee80211_ptr; int err; + ASSERT_WDEV_LOCK(wdev); + if (!wdev->ssid_len) return -ENOLINK; @@ -103,11 +153,24 @@ int cfg80211_leave_ibss(struct cfg80211_registered_device *rdev, if (err) return err; - cfg80211_clear_ibss(dev, nowext); + __cfg80211_clear_ibss(dev, nowext); return 0; } +int cfg80211_leave_ibss(struct cfg80211_registered_device *rdev, + struct net_device *dev, bool nowext) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + int err; + + wdev_lock(wdev); + err = __cfg80211_leave_ibss(rdev, dev, nowext); + wdev_unlock(wdev); + + return err; +} + #ifdef CONFIG_WIRELESS_EXT static int cfg80211_ibss_wext_join(struct cfg80211_registered_device *rdev, struct wireless_dev *wdev) @@ -184,12 +247,15 @@ int cfg80211_ibss_wext_siwfreq(struct net_device *dev, if (wdev->wext.ibss.channel == chan) return 0; - if (wdev->ssid_len) { - err = cfg80211_leave_ibss(wiphy_to_dev(wdev->wiphy), - dev, true); - if (err) - return err; - } + wdev_lock(wdev); + err = 0; + if (wdev->ssid_len) + err = __cfg80211_leave_ibss(wiphy_to_dev(wdev->wiphy), + dev, true); + wdev_unlock(wdev); + + if (err) + return err; if (chan) { wdev->wext.ibss.channel = chan; @@ -215,10 +281,12 @@ int cfg80211_ibss_wext_giwfreq(struct net_device *dev, if (WARN_ON(wdev->iftype != NL80211_IFTYPE_ADHOC)) return -EINVAL; + wdev_lock(wdev); if (wdev->current_bss) chan = wdev->current_bss->pub.channel; else if (wdev->wext.ibss.channel) chan = wdev->wext.ibss.channel; + wdev_unlock(wdev); if (chan) { freq->m = chan->center_freq; @@ -247,12 +315,15 @@ int cfg80211_ibss_wext_siwessid(struct net_device *dev, if (!wiphy_to_dev(wdev->wiphy)->ops->join_ibss) return -EOPNOTSUPP; - if (wdev->ssid_len) { - err = cfg80211_leave_ibss(wiphy_to_dev(wdev->wiphy), - dev, true); - if (err) - return err; - } + wdev_lock(wdev); + err = 0; + if (wdev->ssid_len) + err = __cfg80211_leave_ibss(wiphy_to_dev(wdev->wiphy), + dev, true); + wdev_unlock(wdev); + + if (err) + return err; /* iwconfig uses nul termination in SSID.. */ if (len > 0 && ssid[len - 1] == '\0') @@ -279,6 +350,7 @@ int cfg80211_ibss_wext_giwessid(struct net_device *dev, data->flags = 0; + wdev_lock(wdev); if (wdev->ssid_len) { data->flags = 1; data->length = wdev->ssid_len; @@ -288,6 +360,7 @@ int cfg80211_ibss_wext_giwessid(struct net_device *dev, data->length = wdev->wext.ibss.ssid_len; memcpy(ssid, wdev->wext.ibss.ssid, data->length); } + wdev_unlock(wdev); return 0; } @@ -325,12 +398,15 @@ int cfg80211_ibss_wext_siwap(struct net_device *dev, compare_ether_addr(bssid, wdev->wext.ibss.bssid) == 0) return 0; - if (wdev->ssid_len) { - err = cfg80211_leave_ibss(wiphy_to_dev(wdev->wiphy), - dev, true); - if (err) - return err; - } + wdev_lock(wdev); + err = 0; + if (wdev->ssid_len) + err = __cfg80211_leave_ibss(wiphy_to_dev(wdev->wiphy), + dev, true); + wdev_unlock(wdev); + + if (err) + return err; if (bssid) { memcpy(wdev->wext.bssid, bssid, ETH_ALEN); @@ -355,10 +431,13 @@ int cfg80211_ibss_wext_giwap(struct net_device *dev, ap_addr->sa_family = ARPHRD_ETHER; + wdev_lock(wdev); if (wdev->current_bss) memcpy(ap_addr->sa_data, wdev->current_bss->pub.bssid, ETH_ALEN); else memcpy(ap_addr->sa_data, wdev->wext.ibss.bssid, ETH_ALEN); + wdev_unlock(wdev); + return 0; } /* temporary symbol - mark GPL - in the future the handler won't be */ diff --git a/net/wireless/mlme.c b/net/wireless/mlme.c index 960bf60e44e5..1b2ca1fea7a1 100644 --- a/net/wireless/mlme.c +++ b/net/wireless/mlme.c @@ -23,7 +23,7 @@ void cfg80211_send_rx_auth(struct net_device *dev, const u8 *buf, size_t len) u16 status = le16_to_cpu(mgmt->u.auth.status_code); bool done = false; - might_sleep(); + wdev_lock(wdev); for (i = 0; i < MAX_AUTH_BSSES; i++) { if (wdev->authtry_bsses[i] && @@ -45,6 +45,8 @@ void cfg80211_send_rx_auth(struct net_device *dev, const u8 *buf, size_t len) nl80211_send_rx_auth(rdev, dev, buf, len, GFP_KERNEL); cfg80211_sme_rx_auth(dev, buf, len); + + wdev_unlock(wdev); } EXPORT_SYMBOL(cfg80211_send_rx_auth); @@ -59,14 +61,15 @@ void cfg80211_send_rx_assoc(struct net_device *dev, const u8 *buf, size_t len) int i, ieoffs = offsetof(struct ieee80211_mgmt, u.assoc_resp.variable); bool done; - might_sleep(); + wdev_lock(wdev); status_code = le16_to_cpu(mgmt->u.assoc_resp.status_code); nl80211_send_rx_assoc(rdev, dev, buf, len, GFP_KERNEL); - cfg80211_connect_result(dev, mgmt->bssid, NULL, 0, ie, len - ieoffs, - status_code, GFP_KERNEL); + __cfg80211_connect_result(dev, mgmt->bssid, NULL, 0, ie, len - ieoffs, + status_code, + status_code == WLAN_STATUS_SUCCESS); if (status_code == WLAN_STATUS_SUCCESS) { for (i = 0; wdev->current_bss && i < MAX_AUTH_BSSES; i++) { @@ -81,10 +84,13 @@ void cfg80211_send_rx_assoc(struct net_device *dev, const u8 *buf, size_t len) WARN_ON(!done); } + + wdev_unlock(wdev); } EXPORT_SYMBOL(cfg80211_send_rx_assoc); -void cfg80211_send_deauth(struct net_device *dev, const u8 *buf, size_t len) +static void __cfg80211_send_deauth(struct net_device *dev, + const u8 *buf, size_t len) { struct wireless_dev *wdev = dev->ieee80211_ptr; struct wiphy *wiphy = wdev->wiphy; @@ -94,7 +100,7 @@ void cfg80211_send_deauth(struct net_device *dev, const u8 *buf, size_t len) int i; bool done = false; - might_sleep(); + ASSERT_WDEV_LOCK(wdev); nl80211_send_deauth(rdev, dev, buf, len, GFP_KERNEL); @@ -132,17 +138,35 @@ void cfg80211_send_deauth(struct net_device *dev, const u8 *buf, size_t len) reason_code = le16_to_cpu(mgmt->u.deauth.reason_code); from_ap = memcmp(mgmt->da, dev->dev_addr, ETH_ALEN) == 0; - __cfg80211_disconnected(dev, GFP_KERNEL, NULL, 0, - reason_code, from_ap); + __cfg80211_disconnected(dev, NULL, 0, reason_code, from_ap); } else if (wdev->sme_state == CFG80211_SME_CONNECTING) { - cfg80211_connect_result(dev, mgmt->bssid, NULL, 0, NULL, 0, - WLAN_STATUS_UNSPECIFIED_FAILURE, - GFP_KERNEL); + __cfg80211_connect_result(dev, mgmt->bssid, NULL, 0, NULL, 0, + WLAN_STATUS_UNSPECIFIED_FAILURE, + false); + } +} + + +void cfg80211_send_deauth(struct net_device *dev, const u8 *buf, size_t len, + void *cookie) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + + BUG_ON(cookie && wdev != cookie); + + if (cookie) { + /* called within callback */ + __cfg80211_send_deauth(dev, buf, len); + } else { + wdev_lock(wdev); + __cfg80211_send_deauth(dev, buf, len); + wdev_unlock(wdev); } } EXPORT_SYMBOL(cfg80211_send_deauth); -void cfg80211_send_disassoc(struct net_device *dev, const u8 *buf, size_t len) +static void __cfg80211_send_disassoc(struct net_device *dev, + const u8 *buf, size_t len) { struct wireless_dev *wdev = dev->ieee80211_ptr; struct wiphy *wiphy = wdev->wiphy; @@ -154,12 +178,12 @@ void cfg80211_send_disassoc(struct net_device *dev, const u8 *buf, size_t len) bool from_ap; bool done = false; - might_sleep(); + wdev_lock(wdev); nl80211_send_disassoc(rdev, dev, buf, len, GFP_KERNEL); if (!wdev->sme_state == CFG80211_SME_CONNECTED) - return; + goto out; if (wdev->current_bss && memcmp(wdev->current_bss, bssid, ETH_ALEN) == 0) { @@ -180,8 +204,26 @@ void cfg80211_send_disassoc(struct net_device *dev, const u8 *buf, size_t len) reason_code = le16_to_cpu(mgmt->u.disassoc.reason_code); from_ap = memcmp(mgmt->da, dev->dev_addr, ETH_ALEN) == 0; - __cfg80211_disconnected(dev, GFP_KERNEL, NULL, 0, - reason_code, from_ap); + __cfg80211_disconnected(dev, NULL, 0, reason_code, from_ap); + out: + wdev_unlock(wdev); +} + +void cfg80211_send_disassoc(struct net_device *dev, const u8 *buf, size_t len, + void *cookie) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + + BUG_ON(cookie && wdev != cookie); + + if (cookie) { + /* called within callback */ + __cfg80211_send_disassoc(dev, buf, len); + } else { + wdev_lock(wdev); + __cfg80211_send_disassoc(dev, buf, len); + wdev_unlock(wdev); + } } EXPORT_SYMBOL(cfg80211_send_disassoc); @@ -193,13 +235,13 @@ void cfg80211_send_auth_timeout(struct net_device *dev, const u8 *addr) int i; bool done = false; - might_sleep(); + wdev_lock(wdev); nl80211_send_auth_timeout(rdev, dev, addr, GFP_KERNEL); if (wdev->sme_state == CFG80211_SME_CONNECTING) - cfg80211_connect_result(dev, addr, NULL, 0, NULL, 0, - WLAN_STATUS_UNSPECIFIED_FAILURE, - GFP_KERNEL); + __cfg80211_connect_result(dev, addr, NULL, 0, NULL, 0, + WLAN_STATUS_UNSPECIFIED_FAILURE, + false); for (i = 0; addr && i < MAX_AUTH_BSSES; i++) { if (wdev->authtry_bsses[i] && @@ -214,6 +256,8 @@ void cfg80211_send_auth_timeout(struct net_device *dev, const u8 *addr) } WARN_ON(!done); + + wdev_unlock(wdev); } EXPORT_SYMBOL(cfg80211_send_auth_timeout); @@ -225,13 +269,13 @@ void cfg80211_send_assoc_timeout(struct net_device *dev, const u8 *addr) int i; bool done = false; - might_sleep(); + wdev_lock(wdev); nl80211_send_assoc_timeout(rdev, dev, addr, GFP_KERNEL); if (wdev->sme_state == CFG80211_SME_CONNECTING) - cfg80211_connect_result(dev, addr, NULL, 0, NULL, 0, - WLAN_STATUS_UNSPECIFIED_FAILURE, - GFP_KERNEL); + __cfg80211_connect_result(dev, addr, NULL, 0, NULL, 0, + WLAN_STATUS_UNSPECIFIED_FAILURE, + false); for (i = 0; addr && i < MAX_AUTH_BSSES; i++) { if (wdev->auth_bsses[i] && @@ -246,6 +290,8 @@ void cfg80211_send_assoc_timeout(struct net_device *dev, const u8 *addr) } WARN_ON(!done); + + wdev_unlock(wdev); } EXPORT_SYMBOL(cfg80211_send_assoc_timeout); @@ -276,17 +322,21 @@ void cfg80211_michael_mic_failure(struct net_device *dev, const u8 *addr, EXPORT_SYMBOL(cfg80211_michael_mic_failure); /* some MLME handling for userspace SME */ -int cfg80211_mlme_auth(struct cfg80211_registered_device *rdev, - struct net_device *dev, struct ieee80211_channel *chan, - enum nl80211_auth_type auth_type, const u8 *bssid, - const u8 *ssid, int ssid_len, - const u8 *ie, int ie_len) +int __cfg80211_mlme_auth(struct cfg80211_registered_device *rdev, + struct net_device *dev, + struct ieee80211_channel *chan, + enum nl80211_auth_type auth_type, + const u8 *bssid, + const u8 *ssid, int ssid_len, + const u8 *ie, int ie_len) { struct wireless_dev *wdev = dev->ieee80211_ptr; struct cfg80211_auth_request req; struct cfg80211_internal_bss *bss; int i, err, slot = -1, nfree = 0; + ASSERT_WDEV_LOCK(wdev); + if (wdev->current_bss && memcmp(bssid, wdev->current_bss->pub.bssid, ETH_ALEN) == 0) return -EALREADY; @@ -342,18 +392,37 @@ int cfg80211_mlme_auth(struct cfg80211_registered_device *rdev, return err; } -int cfg80211_mlme_assoc(struct cfg80211_registered_device *rdev, - struct net_device *dev, struct ieee80211_channel *chan, - const u8 *bssid, const u8 *prev_bssid, - const u8 *ssid, int ssid_len, - const u8 *ie, int ie_len, bool use_mfp, - struct cfg80211_crypto_settings *crypt) +int cfg80211_mlme_auth(struct cfg80211_registered_device *rdev, + struct net_device *dev, struct ieee80211_channel *chan, + enum nl80211_auth_type auth_type, const u8 *bssid, + const u8 *ssid, int ssid_len, + const u8 *ie, int ie_len) +{ + int err; + + wdev_lock(dev->ieee80211_ptr); + err = __cfg80211_mlme_auth(rdev, dev, chan, auth_type, bssid, + ssid, ssid_len, ie, ie_len); + wdev_unlock(dev->ieee80211_ptr); + + return err; +} + +int __cfg80211_mlme_assoc(struct cfg80211_registered_device *rdev, + struct net_device *dev, + struct ieee80211_channel *chan, + const u8 *bssid, const u8 *prev_bssid, + const u8 *ssid, int ssid_len, + const u8 *ie, int ie_len, bool use_mfp, + struct cfg80211_crypto_settings *crypt) { struct wireless_dev *wdev = dev->ieee80211_ptr; struct cfg80211_assoc_request req; struct cfg80211_internal_bss *bss; int i, err, slot = -1; + ASSERT_WDEV_LOCK(wdev); + memset(&req, 0, sizeof(req)); if (wdev->current_bss) @@ -390,14 +459,35 @@ int cfg80211_mlme_assoc(struct cfg80211_registered_device *rdev, return err; } -int cfg80211_mlme_deauth(struct cfg80211_registered_device *rdev, - struct net_device *dev, const u8 *bssid, - const u8 *ie, int ie_len, u16 reason) +int cfg80211_mlme_assoc(struct cfg80211_registered_device *rdev, + struct net_device *dev, + struct ieee80211_channel *chan, + const u8 *bssid, const u8 *prev_bssid, + const u8 *ssid, int ssid_len, + const u8 *ie, int ie_len, bool use_mfp, + struct cfg80211_crypto_settings *crypt) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + int err; + + wdev_lock(wdev); + err = __cfg80211_mlme_assoc(rdev, dev, chan, bssid, prev_bssid, + ssid, ssid_len, ie, ie_len, use_mfp, crypt); + wdev_unlock(wdev); + + return err; +} + +int __cfg80211_mlme_deauth(struct cfg80211_registered_device *rdev, + struct net_device *dev, const u8 *bssid, + const u8 *ie, int ie_len, u16 reason) { struct wireless_dev *wdev = dev->ieee80211_ptr; struct cfg80211_deauth_request req; int i; + ASSERT_WDEV_LOCK(wdev); + memset(&req, 0, sizeof(req)); req.reason_code = reason; req.ie = ie; @@ -421,16 +511,32 @@ int cfg80211_mlme_deauth(struct cfg80211_registered_device *rdev, if (!req.bss) return -ENOTCONN; - return rdev->ops->deauth(&rdev->wiphy, dev, &req); + return rdev->ops->deauth(&rdev->wiphy, dev, &req, wdev); } -int cfg80211_mlme_disassoc(struct cfg80211_registered_device *rdev, - struct net_device *dev, const u8 *bssid, - const u8 *ie, int ie_len, u16 reason) +int cfg80211_mlme_deauth(struct cfg80211_registered_device *rdev, + struct net_device *dev, const u8 *bssid, + const u8 *ie, int ie_len, u16 reason) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + int err; + + wdev_lock(wdev); + err = __cfg80211_mlme_deauth(rdev, dev, bssid, ie, ie_len, reason); + wdev_unlock(wdev); + + return err; +} + +static int __cfg80211_mlme_disassoc(struct cfg80211_registered_device *rdev, + struct net_device *dev, const u8 *bssid, + const u8 *ie, int ie_len, u16 reason) { struct wireless_dev *wdev = dev->ieee80211_ptr; struct cfg80211_disassoc_request req; + ASSERT_WDEV_LOCK(wdev); + memset(&req, 0, sizeof(req)); req.reason_code = reason; req.ie = ie; @@ -440,7 +546,21 @@ int cfg80211_mlme_disassoc(struct cfg80211_registered_device *rdev, else return -ENOTCONN; - return rdev->ops->disassoc(&rdev->wiphy, dev, &req); + return rdev->ops->disassoc(&rdev->wiphy, dev, &req, wdev); +} + +int cfg80211_mlme_disassoc(struct cfg80211_registered_device *rdev, + struct net_device *dev, const u8 *bssid, + const u8 *ie, int ie_len, u16 reason) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + int err; + + wdev_lock(wdev); + err = __cfg80211_mlme_disassoc(rdev, dev, bssid, ie, ie_len, reason); + wdev_unlock(wdev); + + return err; } void cfg80211_mlme_down(struct cfg80211_registered_device *rdev, @@ -450,6 +570,8 @@ void cfg80211_mlme_down(struct cfg80211_registered_device *rdev, struct cfg80211_deauth_request req; int i; + ASSERT_WDEV_LOCK(wdev); + if (!rdev->ops->deauth) return; @@ -460,7 +582,7 @@ void cfg80211_mlme_down(struct cfg80211_registered_device *rdev, if (wdev->current_bss) { req.bss = &wdev->current_bss->pub; - rdev->ops->deauth(&rdev->wiphy, dev, &req); + rdev->ops->deauth(&rdev->wiphy, dev, &req, wdev); if (wdev->current_bss) { cfg80211_unhold_bss(wdev->current_bss); cfg80211_put_bss(&wdev->current_bss->pub); @@ -471,7 +593,7 @@ void cfg80211_mlme_down(struct cfg80211_registered_device *rdev, for (i = 0; i < MAX_AUTH_BSSES; i++) { if (wdev->auth_bsses[i]) { req.bss = &wdev->auth_bsses[i]->pub; - rdev->ops->deauth(&rdev->wiphy, dev, &req); + rdev->ops->deauth(&rdev->wiphy, dev, &req, wdev); if (wdev->auth_bsses[i]) { cfg80211_unhold_bss(wdev->auth_bsses[i]); cfg80211_put_bss(&wdev->auth_bsses[i]->pub); @@ -480,7 +602,7 @@ void cfg80211_mlme_down(struct cfg80211_registered_device *rdev, } if (wdev->authtry_bsses[i]) { req.bss = &wdev->authtry_bsses[i]->pub; - rdev->ops->deauth(&rdev->wiphy, dev, &req); + rdev->ops->deauth(&rdev->wiphy, dev, &req, wdev); if (wdev->authtry_bsses[i]) { cfg80211_unhold_bss(wdev->authtry_bsses[i]); cfg80211_put_bss(&wdev->authtry_bsses[i]->pub); diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index 4976eac888a3..cf4ac786b20a 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -4029,6 +4029,8 @@ static int nl80211_add_scan_req(struct sk_buff *msg, struct nlattr *nest; int i; + ASSERT_RDEV_LOCK(rdev); + if (WARN_ON(!req)) return 0; @@ -4391,12 +4393,12 @@ void nl80211_send_roamed(struct cfg80211_registered_device *rdev, void nl80211_send_disconnected(struct cfg80211_registered_device *rdev, struct net_device *netdev, u16 reason, - u8 *ie, size_t ie_len, bool from_ap, gfp_t gfp) + const u8 *ie, size_t ie_len, bool from_ap) { struct sk_buff *msg; void *hdr; - msg = nlmsg_new(NLMSG_GOODSIZE, gfp); + msg = nlmsg_new(NLMSG_GOODSIZE, GFP_KERNEL); if (!msg) return; @@ -4420,7 +4422,7 @@ void nl80211_send_disconnected(struct cfg80211_registered_device *rdev, return; } - genlmsg_multicast(msg, 0, nl80211_mlme_mcgrp.id, gfp); + genlmsg_multicast(msg, 0, nl80211_mlme_mcgrp.id, GFP_KERNEL); return; nla_put_failure: diff --git a/net/wireless/nl80211.h b/net/wireless/nl80211.h index cf3708b48c29..44cc2a76a1b0 100644 --- a/net/wireless/nl80211.h +++ b/net/wireless/nl80211.h @@ -42,7 +42,7 @@ void nl80211_send_roamed(struct cfg80211_registered_device *rdev, const u8 *resp_ie, size_t resp_ie_len, gfp_t gfp); void nl80211_send_disconnected(struct cfg80211_registered_device *rdev, struct net_device *netdev, u16 reason, - u8 *ie, size_t ie_len, bool from_ap, gfp_t gfp); + const u8 *ie, size_t ie_len, bool from_ap); void nl80211_michael_mic_failure(struct cfg80211_registered_device *rdev, diff --git a/net/wireless/scan.c b/net/wireless/scan.c index 1625faf1de57..4f552c3f29a3 100644 --- a/net/wireless/scan.c +++ b/net/wireless/scan.c @@ -17,13 +17,21 @@ #define IEEE80211_SCAN_RESULT_EXPIRE (10 * HZ) -void cfg80211_scan_done(struct cfg80211_scan_request *request, bool aborted) +void __cfg80211_scan_done(struct work_struct *wk) { + struct cfg80211_registered_device *rdev; + struct cfg80211_scan_request *request; struct net_device *dev; #ifdef CONFIG_WIRELESS_EXT union iwreq_data wrqu; #endif + rdev = container_of(wk, struct cfg80211_registered_device, + scan_done_wk); + + mutex_lock(&rdev->mtx); + request = rdev->scan_req; + dev = dev_get_by_index(&init_net, request->ifidx); if (!dev) goto out; @@ -35,7 +43,7 @@ void cfg80211_scan_done(struct cfg80211_scan_request *request, bool aborted) */ cfg80211_sme_scan_done(dev); - if (aborted) + if (request->aborted) nl80211_send_scan_aborted(wiphy_to_dev(request->wiphy), dev); else nl80211_send_scan_done(wiphy_to_dev(request->wiphy), dev); @@ -43,7 +51,7 @@ void cfg80211_scan_done(struct cfg80211_scan_request *request, bool aborted) wiphy_to_dev(request->wiphy)->scan_req = NULL; #ifdef CONFIG_WIRELESS_EXT - if (!aborted) { + if (!request->aborted) { memset(&wrqu, 0, sizeof(wrqu)); wireless_send_event(dev, SIOCGIWSCAN, &wrqu, NULL); @@ -53,8 +61,24 @@ void cfg80211_scan_done(struct cfg80211_scan_request *request, bool aborted) dev_put(dev); out: + cfg80211_unlock_rdev(rdev); kfree(request); } + +void cfg80211_scan_done(struct cfg80211_scan_request *request, bool aborted) +{ + struct net_device *dev = dev_get_by_index(&init_net, request->ifidx); + if (WARN_ON(!dev)) { + kfree(request); + return; + } + + WARN_ON(request != wiphy_to_dev(request->wiphy)->scan_req); + + request->aborted = aborted; + schedule_work(&wiphy_to_dev(request->wiphy)->scan_done_wk); + dev_put(dev); +} EXPORT_SYMBOL(cfg80211_scan_done); static void bss_release(struct kref *ref) diff --git a/net/wireless/sme.c b/net/wireless/sme.c index 066a19ef9d73..472e2412c781 100644 --- a/net/wireless/sme.c +++ b/net/wireless/sme.c @@ -38,6 +38,8 @@ static int cfg80211_conn_scan(struct wireless_dev *wdev) int n_channels, err; ASSERT_RTNL(); + ASSERT_RDEV_LOCK(drv); + ASSERT_WDEV_LOCK(wdev); if (drv->scan_req) return -EBUSY; @@ -106,6 +108,8 @@ static int cfg80211_conn_do_work(struct wireless_dev *wdev) struct cfg80211_connect_params *params; int err; + ASSERT_WDEV_LOCK(wdev); + if (!wdev->conn) return 0; @@ -117,11 +121,11 @@ static int cfg80211_conn_do_work(struct wireless_dev *wdev) case CFG80211_CONN_AUTHENTICATE_NEXT: BUG_ON(!drv->ops->auth); wdev->conn->state = CFG80211_CONN_AUTHENTICATING; - return cfg80211_mlme_auth(drv, wdev->netdev, - params->channel, params->auth_type, - params->bssid, - params->ssid, params->ssid_len, - NULL, 0); + return __cfg80211_mlme_auth(drv, wdev->netdev, + params->channel, params->auth_type, + params->bssid, + params->ssid, params->ssid_len, + NULL, 0); case CFG80211_CONN_ASSOCIATE_NEXT: BUG_ON(!drv->ops->assoc); wdev->conn->state = CFG80211_CONN_ASSOCIATING; @@ -131,14 +135,16 @@ static int cfg80211_conn_do_work(struct wireless_dev *wdev) * that some APs don't like that -- so we'd need to retry * the association. */ - err = cfg80211_mlme_assoc(drv, wdev->netdev, - params->channel, params->bssid, NULL, - params->ssid, params->ssid_len, - params->ie, params->ie_len, - false, ¶ms->crypto); + err = __cfg80211_mlme_assoc(drv, wdev->netdev, + params->channel, params->bssid, + NULL, + params->ssid, params->ssid_len, + params->ie, params->ie_len, + false, ¶ms->crypto); if (err) - cfg80211_mlme_deauth(drv, wdev->netdev, params->bssid, - NULL, 0, WLAN_REASON_DEAUTH_LEAVING); + __cfg80211_mlme_deauth(drv, wdev->netdev, params->bssid, + NULL, 0, + WLAN_REASON_DEAUTH_LEAVING); return err; default: return 0; @@ -152,22 +158,31 @@ void cfg80211_conn_work(struct work_struct *work) struct wireless_dev *wdev; rtnl_lock(); + cfg80211_lock_rdev(drv); mutex_lock(&drv->devlist_mtx); list_for_each_entry(wdev, &drv->netdev_list, list) { - if (!netif_running(wdev->netdev)) + wdev_lock(wdev); + if (!netif_running(wdev->netdev)) { + wdev_unlock(wdev); continue; - if (wdev->sme_state != CFG80211_SME_CONNECTING) + } + if (wdev->sme_state != CFG80211_SME_CONNECTING) { + wdev_unlock(wdev); continue; + } if (cfg80211_conn_do_work(wdev)) - cfg80211_connect_result(wdev->netdev, - wdev->conn->params.bssid, - NULL, 0, NULL, 0, - WLAN_STATUS_UNSPECIFIED_FAILURE, - GFP_ATOMIC); + __cfg80211_connect_result( + wdev->netdev, + wdev->conn->params.bssid, + NULL, 0, NULL, 0, + WLAN_STATUS_UNSPECIFIED_FAILURE, + false); + wdev_unlock(wdev); } mutex_unlock(&drv->devlist_mtx); + cfg80211_unlock_rdev(drv); rtnl_unlock(); } @@ -177,6 +192,8 @@ static bool cfg80211_get_conn_bss(struct wireless_dev *wdev) struct cfg80211_bss *bss; u16 capa = WLAN_CAPABILITY_ESS; + ASSERT_WDEV_LOCK(wdev); + if (wdev->conn->params.privacy) capa |= WLAN_CAPABILITY_PRIVACY; @@ -198,11 +215,13 @@ static bool cfg80211_get_conn_bss(struct wireless_dev *wdev) return true; } -void cfg80211_sme_scan_done(struct net_device *dev) +static void __cfg80211_sme_scan_done(struct net_device *dev) { struct wireless_dev *wdev = dev->ieee80211_ptr; struct cfg80211_registered_device *drv = wiphy_to_dev(wdev->wiphy); + ASSERT_WDEV_LOCK(wdev); + if (wdev->sme_state != CFG80211_SME_CONNECTING) return; @@ -218,15 +237,26 @@ void cfg80211_sme_scan_done(struct net_device *dev) if (wdev->conn->state == CFG80211_CONN_SCAN_AGAIN) schedule_work(&drv->conn_work); else - cfg80211_connect_result(dev, wdev->conn->params.bssid, - NULL, 0, NULL, 0, - WLAN_STATUS_UNSPECIFIED_FAILURE, - GFP_ATOMIC); - return; + __cfg80211_connect_result( + wdev->netdev, + wdev->conn->params.bssid, + NULL, 0, NULL, 0, + WLAN_STATUS_UNSPECIFIED_FAILURE, + false); } } -void cfg80211_sme_rx_auth(struct net_device *dev, const u8 *buf, size_t len) +void cfg80211_sme_scan_done(struct net_device *dev) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + + wdev_lock(wdev); + __cfg80211_sme_scan_done(dev); + wdev_unlock(wdev); +} + +void cfg80211_sme_rx_auth(struct net_device *dev, + const u8 *buf, size_t len) { struct wireless_dev *wdev = dev->ieee80211_ptr; struct wiphy *wiphy = wdev->wiphy; @@ -234,6 +264,8 @@ void cfg80211_sme_rx_auth(struct net_device *dev, const u8 *buf, size_t len) struct ieee80211_mgmt *mgmt = (struct ieee80211_mgmt *)buf; u16 status_code = le16_to_cpu(mgmt->u.auth.status_code); + ASSERT_WDEV_LOCK(wdev); + /* should only RX auth frames when connecting */ if (wdev->sme_state != CFG80211_SME_CONNECTING) return; @@ -273,10 +305,10 @@ void cfg80211_sme_rx_auth(struct net_device *dev, const u8 *buf, size_t len) } } -static void __cfg80211_connect_result(struct net_device *dev, const u8 *bssid, - const u8 *req_ie, size_t req_ie_len, - const u8 *resp_ie, size_t resp_ie_len, - u16 status, bool wextev, gfp_t gfp) +void __cfg80211_connect_result(struct net_device *dev, const u8 *bssid, + const u8 *req_ie, size_t req_ie_len, + const u8 *resp_ie, size_t resp_ie_len, + u16 status, bool wextev) { struct wireless_dev *wdev = dev->ieee80211_ptr; struct cfg80211_bss *bss; @@ -284,18 +316,20 @@ static void __cfg80211_connect_result(struct net_device *dev, const u8 *bssid, union iwreq_data wrqu; #endif + ASSERT_WDEV_LOCK(wdev); + if (WARN_ON(wdev->iftype != NL80211_IFTYPE_STATION)) return; if (wdev->sme_state == CFG80211_SME_CONNECTED) nl80211_send_roamed(wiphy_to_dev(wdev->wiphy), dev, bssid, req_ie, req_ie_len, - resp_ie, resp_ie_len, gfp); + resp_ie, resp_ie_len, GFP_KERNEL); else nl80211_send_connect_result(wiphy_to_dev(wdev->wiphy), dev, bssid, req_ie, req_ie_len, resp_ie, resp_ie_len, - status, gfp); + status, GFP_KERNEL); #ifdef CONFIG_WIRELESS_EXT if (wextev) { @@ -362,21 +396,43 @@ void cfg80211_connect_result(struct net_device *dev, const u8 *bssid, const u8 *resp_ie, size_t resp_ie_len, u16 status, gfp_t gfp) { - bool wextev = status == WLAN_STATUS_SUCCESS; - __cfg80211_connect_result(dev, bssid, req_ie, req_ie_len, resp_ie, resp_ie_len, status, wextev, gfp); + struct wireless_dev *wdev = dev->ieee80211_ptr; + struct cfg80211_registered_device *rdev = wiphy_to_dev(wdev->wiphy); + struct cfg80211_event *ev; + unsigned long flags; + + ev = kzalloc(sizeof(*ev) + req_ie_len + resp_ie_len, gfp); + if (!ev) + return; + + ev->type = EVENT_CONNECT_RESULT; + memcpy(ev->cr.bssid, bssid, ETH_ALEN); + ev->cr.req_ie = ((u8 *)ev) + sizeof(*ev); + ev->cr.req_ie_len = req_ie_len; + memcpy((void *)ev->cr.req_ie, req_ie, req_ie_len); + ev->cr.resp_ie = ((u8 *)ev) + sizeof(*ev) + req_ie_len; + ev->cr.resp_ie_len = resp_ie_len; + memcpy((void *)ev->cr.resp_ie, resp_ie, resp_ie_len); + ev->cr.status = status; + + spin_lock_irqsave(&wdev->event_lock, flags); + list_add_tail(&ev->list, &wdev->event_list); + spin_unlock_irqrestore(&wdev->event_lock, flags); + schedule_work(&rdev->event_work); } EXPORT_SYMBOL(cfg80211_connect_result); -void cfg80211_roamed(struct net_device *dev, const u8 *bssid, - const u8 *req_ie, size_t req_ie_len, - const u8 *resp_ie, size_t resp_ie_len, gfp_t gfp) +void __cfg80211_roamed(struct wireless_dev *wdev, const u8 *bssid, + const u8 *req_ie, size_t req_ie_len, + const u8 *resp_ie, size_t resp_ie_len) { - struct wireless_dev *wdev = dev->ieee80211_ptr; struct cfg80211_bss *bss; #ifdef CONFIG_WIRELESS_EXT union iwreq_data wrqu; #endif + ASSERT_WDEV_LOCK(wdev); + if (WARN_ON(wdev->iftype != NL80211_IFTYPE_STATION)) return; @@ -402,31 +458,62 @@ void cfg80211_roamed(struct net_device *dev, const u8 *bssid, cfg80211_hold_bss(bss_from_pub(bss)); wdev->current_bss = bss_from_pub(bss); - nl80211_send_roamed(wiphy_to_dev(wdev->wiphy), dev, bssid, - req_ie, req_ie_len, resp_ie, resp_ie_len, gfp); + nl80211_send_roamed(wiphy_to_dev(wdev->wiphy), wdev->netdev, bssid, + req_ie, req_ie_len, resp_ie, resp_ie_len, + GFP_KERNEL); #ifdef CONFIG_WIRELESS_EXT if (req_ie) { memset(&wrqu, 0, sizeof(wrqu)); wrqu.data.length = req_ie_len; - wireless_send_event(dev, IWEVASSOCRESPIE, &wrqu, req_ie); + wireless_send_event(wdev->netdev, IWEVASSOCRESPIE, + &wrqu, req_ie); } if (resp_ie) { memset(&wrqu, 0, sizeof(wrqu)); wrqu.data.length = resp_ie_len; - wireless_send_event(dev, IWEVASSOCRESPIE, &wrqu, resp_ie); + wireless_send_event(wdev->netdev, IWEVASSOCRESPIE, + &wrqu, resp_ie); } memset(&wrqu, 0, sizeof(wrqu)); wrqu.ap_addr.sa_family = ARPHRD_ETHER; memcpy(wrqu.ap_addr.sa_data, bssid, ETH_ALEN); - wireless_send_event(dev, SIOCGIWAP, &wrqu, NULL); + wireless_send_event(wdev->netdev, SIOCGIWAP, &wrqu, NULL); #endif } + +void cfg80211_roamed(struct net_device *dev, const u8 *bssid, + const u8 *req_ie, size_t req_ie_len, + const u8 *resp_ie, size_t resp_ie_len, gfp_t gfp) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + struct cfg80211_registered_device *rdev = wiphy_to_dev(wdev->wiphy); + struct cfg80211_event *ev; + unsigned long flags; + + ev = kzalloc(sizeof(*ev) + req_ie_len + resp_ie_len, gfp); + if (!ev) + return; + + ev->type = EVENT_ROAMED; + memcpy(ev->rm.bssid, bssid, ETH_ALEN); + ev->rm.req_ie = ((u8 *)ev) + sizeof(*ev); + ev->rm.req_ie_len = req_ie_len; + memcpy((void *)ev->rm.req_ie, req_ie, req_ie_len); + ev->rm.resp_ie = ((u8 *)ev) + sizeof(*ev) + req_ie_len; + ev->rm.resp_ie_len = resp_ie_len; + memcpy((void *)ev->rm.resp_ie, resp_ie, resp_ie_len); + + spin_lock_irqsave(&wdev->event_lock, flags); + list_add_tail(&ev->list, &wdev->event_list); + spin_unlock_irqrestore(&wdev->event_lock, flags); + schedule_work(&rdev->event_work); +} EXPORT_SYMBOL(cfg80211_roamed); -void __cfg80211_disconnected(struct net_device *dev, gfp_t gfp, u8 *ie, +void __cfg80211_disconnected(struct net_device *dev, const u8 *ie, size_t ie_len, u16 reason, bool from_ap) { struct wireless_dev *wdev = dev->ieee80211_ptr; @@ -434,6 +521,8 @@ void __cfg80211_disconnected(struct net_device *dev, gfp_t gfp, u8 *ie, union iwreq_data wrqu; #endif + ASSERT_WDEV_LOCK(wdev); + if (WARN_ON(wdev->iftype != NL80211_IFTYPE_STATION)) return; @@ -456,7 +545,7 @@ void __cfg80211_disconnected(struct net_device *dev, gfp_t gfp, u8 *ie, } nl80211_send_disconnected(wiphy_to_dev(wdev->wiphy), dev, - reason, ie, ie_len, from_ap, gfp); + reason, ie, ie_len, from_ap); #ifdef CONFIG_WIRELESS_EXT memset(&wrqu, 0, sizeof(wrqu)); @@ -468,16 +557,36 @@ void __cfg80211_disconnected(struct net_device *dev, gfp_t gfp, u8 *ie, void cfg80211_disconnected(struct net_device *dev, u16 reason, u8 *ie, size_t ie_len, gfp_t gfp) { - __cfg80211_disconnected(dev, gfp, ie, ie_len, reason, true); + struct wireless_dev *wdev = dev->ieee80211_ptr; + struct cfg80211_registered_device *rdev = wiphy_to_dev(wdev->wiphy); + struct cfg80211_event *ev; + unsigned long flags; + + ev = kzalloc(sizeof(*ev) + ie_len, gfp); + if (!ev) + return; + + ev->type = EVENT_DISCONNECTED; + ev->dc.ie = ((u8 *)ev) + sizeof(*ev); + ev->dc.ie_len = ie_len; + memcpy((void *)ev->dc.ie, ie, ie_len); + ev->dc.reason = reason; + + spin_lock_irqsave(&wdev->event_lock, flags); + list_add_tail(&ev->list, &wdev->event_list); + spin_unlock_irqrestore(&wdev->event_lock, flags); + schedule_work(&rdev->event_work); } EXPORT_SYMBOL(cfg80211_disconnected); -int cfg80211_connect(struct cfg80211_registered_device *rdev, - struct net_device *dev, - struct cfg80211_connect_params *connect) +int __cfg80211_connect(struct cfg80211_registered_device *rdev, + struct net_device *dev, + struct cfg80211_connect_params *connect) { - int err; struct wireless_dev *wdev = dev->ieee80211_ptr; + int err; + + ASSERT_WDEV_LOCK(wdev); if (wdev->sme_state != CFG80211_SME_IDLE) return -EALREADY; @@ -572,12 +681,27 @@ int cfg80211_connect(struct cfg80211_registered_device *rdev, } } -int cfg80211_disconnect(struct cfg80211_registered_device *rdev, - struct net_device *dev, u16 reason, bool wextev) +int cfg80211_connect(struct cfg80211_registered_device *rdev, + struct net_device *dev, + struct cfg80211_connect_params *connect) +{ + int err; + + wdev_lock(dev->ieee80211_ptr); + err = __cfg80211_connect(rdev, dev, connect); + wdev_unlock(dev->ieee80211_ptr); + + return err; +} + +int __cfg80211_disconnect(struct cfg80211_registered_device *rdev, + struct net_device *dev, u16 reason, bool wextev) { struct wireless_dev *wdev = dev->ieee80211_ptr; int err; + ASSERT_WDEV_LOCK(wdev); + if (wdev->sme_state == CFG80211_SME_IDLE) return -EINVAL; @@ -601,8 +725,9 @@ int cfg80211_disconnect(struct cfg80211_registered_device *rdev, } /* wdev->conn->params.bssid must be set if > SCANNING */ - err = cfg80211_mlme_deauth(rdev, dev, wdev->conn->params.bssid, - NULL, 0, reason); + err = __cfg80211_mlme_deauth(rdev, dev, + wdev->conn->params.bssid, + NULL, 0, reason); if (err) return err; } else { @@ -612,21 +737,36 @@ int cfg80211_disconnect(struct cfg80211_registered_device *rdev, } if (wdev->sme_state == CFG80211_SME_CONNECTED) - __cfg80211_disconnected(dev, GFP_KERNEL, NULL, 0, 0, false); + __cfg80211_disconnected(dev, NULL, 0, 0, false); else if (wdev->sme_state == CFG80211_SME_CONNECTING) __cfg80211_connect_result(dev, NULL, NULL, 0, NULL, 0, WLAN_STATUS_UNSPECIFIED_FAILURE, - wextev, GFP_KERNEL); + wextev); return 0; } +int cfg80211_disconnect(struct cfg80211_registered_device *rdev, + struct net_device *dev, + u16 reason, bool wextev) +{ + int err; + + wdev_lock(dev->ieee80211_ptr); + err = __cfg80211_disconnect(rdev, dev, reason, wextev); + wdev_unlock(dev->ieee80211_ptr); + + return err; +} + void cfg80211_sme_disassoc(struct net_device *dev, int idx) { struct wireless_dev *wdev = dev->ieee80211_ptr; struct cfg80211_registered_device *rdev = wiphy_to_dev(wdev->wiphy); u8 bssid[ETH_ALEN]; + ASSERT_WDEV_LOCK(wdev); + if (!wdev->conn) return; diff --git a/net/wireless/wext-sme.c b/net/wireless/wext-sme.c index fe1987acb891..6f75aaa7f795 100644 --- a/net/wireless/wext-sme.c +++ b/net/wireless/wext-sme.c @@ -15,6 +15,9 @@ static int cfg80211_mgd_wext_connect(struct cfg80211_registered_device *rdev, { int err; + ASSERT_RDEV_LOCK(rdev); + ASSERT_WDEV_LOCK(wdev); + if (!netif_running(wdev->netdev)) return 0; @@ -24,8 +27,8 @@ static int cfg80211_mgd_wext_connect(struct cfg80211_registered_device *rdev, err = 0; if (wdev->wext.connect.ssid_len != 0) - err = cfg80211_connect(rdev, wdev->netdev, - &wdev->wext.connect); + err = __cfg80211_connect(rdev, wdev->netdev, + &wdev->wext.connect); return err; } @@ -50,33 +53,43 @@ int cfg80211_mgd_wext_siwfreq(struct net_device *dev, if (chan && (chan->flags & IEEE80211_CHAN_DISABLED)) return -EINVAL; - if (wdev->wext.connect.channel == chan) - return 0; + cfg80211_lock_rdev(rdev); + wdev_lock(wdev); + + if (wdev->wext.connect.channel == chan) { + err = 0; + goto out; + } if (wdev->sme_state != CFG80211_SME_IDLE) { bool event = true; /* if SSID set, we'll try right again, avoid event */ if (wdev->wext.connect.ssid_len) event = false; - err = cfg80211_disconnect(wiphy_to_dev(wdev->wiphy), - dev, WLAN_REASON_DEAUTH_LEAVING, - event); + err = __cfg80211_disconnect(wiphy_to_dev(wdev->wiphy), + dev, WLAN_REASON_DEAUTH_LEAVING, + event); if (err) - return err; + goto out; } + wdev->wext.connect.channel = chan; /* SSID is not set, we just want to switch channel */ if (wdev->wext.connect.ssid_len && chan) { - if (!rdev->ops->set_channel) - return -EOPNOTSUPP; - - return rdev->ops->set_channel(wdev->wiphy, chan, - NL80211_CHAN_NO_HT); + err = -EOPNOTSUPP; + if (rdev->ops->set_channel) + err = rdev->ops->set_channel(wdev->wiphy, chan, + NL80211_CHAN_NO_HT); + goto out; } - return cfg80211_mgd_wext_connect(wiphy_to_dev(wdev->wiphy), wdev); + err = cfg80211_mgd_wext_connect(wiphy_to_dev(wdev->wiphy), wdev); + out: + wdev_unlock(wdev); + cfg80211_unlock_rdev(rdev); + return err; } /* temporary symbol - mark GPL - in the future the handler won't be */ EXPORT_SYMBOL_GPL(cfg80211_mgd_wext_siwfreq); @@ -92,10 +105,12 @@ int cfg80211_mgd_wext_giwfreq(struct net_device *dev, if (WARN_ON(wdev->iftype != NL80211_IFTYPE_STATION)) return -EINVAL; + wdev_lock(wdev); if (wdev->current_bss) chan = wdev->current_bss->pub.channel; else if (wdev->wext.connect.channel) chan = wdev->wext.connect.channel; + wdev_unlock(wdev); if (chan) { freq->m = chan->center_freq; @@ -128,21 +143,26 @@ int cfg80211_mgd_wext_siwessid(struct net_device *dev, if (len > 0 && ssid[len - 1] == '\0') len--; + cfg80211_lock_rdev(wiphy_to_dev(wdev->wiphy)); + wdev_lock(wdev); + + err = 0; + if (wdev->wext.connect.ssid && len && len == wdev->wext.connect.ssid_len && memcmp(wdev->wext.connect.ssid, ssid, len)) - return 0; + goto out; if (wdev->sme_state != CFG80211_SME_IDLE) { bool event = true; /* if SSID set now, we'll try to connect, avoid event */ if (len) event = false; - err = cfg80211_disconnect(wiphy_to_dev(wdev->wiphy), - dev, WLAN_REASON_DEAUTH_LEAVING, - event); + err = __cfg80211_disconnect(wiphy_to_dev(wdev->wiphy), + dev, WLAN_REASON_DEAUTH_LEAVING, + event); if (err) - return err; + goto out; } wdev->wext.connect.ssid = wdev->wext.ssid; @@ -151,7 +171,11 @@ int cfg80211_mgd_wext_siwessid(struct net_device *dev, wdev->wext.connect.crypto.control_port = false; - return cfg80211_mgd_wext_connect(wiphy_to_dev(wdev->wiphy), wdev); + err = cfg80211_mgd_wext_connect(wiphy_to_dev(wdev->wiphy), wdev); + out: + wdev_unlock(wdev); + cfg80211_unlock_rdev(wiphy_to_dev(wdev->wiphy)); + return err; } /* temporary symbol - mark GPL - in the future the handler won't be */ EXPORT_SYMBOL_GPL(cfg80211_mgd_wext_siwessid); @@ -168,6 +192,7 @@ int cfg80211_mgd_wext_giwessid(struct net_device *dev, data->flags = 0; + wdev_lock(wdev); if (wdev->ssid_len) { data->flags = 1; data->length = wdev->ssid_len; @@ -178,6 +203,7 @@ int cfg80211_mgd_wext_giwessid(struct net_device *dev, memcpy(ssid, wdev->wext.connect.ssid, data->length); } else data->flags = 0; + wdev_unlock(wdev); return 0; } @@ -203,21 +229,25 @@ int cfg80211_mgd_wext_siwap(struct net_device *dev, if (is_zero_ether_addr(bssid) || is_broadcast_ether_addr(bssid)) bssid = NULL; + cfg80211_lock_rdev(wiphy_to_dev(wdev->wiphy)); + wdev_lock(wdev); + + err = 0; /* both automatic */ if (!bssid && !wdev->wext.connect.bssid) - return 0; + goto out; /* fixed already - and no change */ if (wdev->wext.connect.bssid && bssid && compare_ether_addr(bssid, wdev->wext.connect.bssid) == 0) - return 0; + goto out; if (wdev->sme_state != CFG80211_SME_IDLE) { - err = cfg80211_disconnect(wiphy_to_dev(wdev->wiphy), - dev, WLAN_REASON_DEAUTH_LEAVING, - false); + err = __cfg80211_disconnect(wiphy_to_dev(wdev->wiphy), + dev, WLAN_REASON_DEAUTH_LEAVING, + false); if (err) - return err; + goto out; } if (bssid) { @@ -226,7 +256,11 @@ int cfg80211_mgd_wext_siwap(struct net_device *dev, } else wdev->wext.connect.bssid = NULL; - return cfg80211_mgd_wext_connect(wiphy_to_dev(wdev->wiphy), wdev); + err = cfg80211_mgd_wext_connect(wiphy_to_dev(wdev->wiphy), wdev); + out: + wdev_unlock(wdev); + cfg80211_unlock_rdev(wiphy_to_dev(wdev->wiphy)); + return err; } /* temporary symbol - mark GPL - in the future the handler won't be */ EXPORT_SYMBOL_GPL(cfg80211_mgd_wext_siwap); @@ -243,12 +277,14 @@ int cfg80211_mgd_wext_giwap(struct net_device *dev, ap_addr->sa_family = ARPHRD_ETHER; + wdev_lock(wdev); if (wdev->current_bss) memcpy(ap_addr->sa_data, wdev->current_bss->pub.bssid, ETH_ALEN); else if (wdev->wext.connect.bssid) memcpy(ap_addr->sa_data, wdev->wext.connect.bssid, ETH_ALEN); else memset(ap_addr->sa_data, 0, ETH_ALEN); + wdev_unlock(wdev); return 0; } @@ -270,15 +306,20 @@ int cfg80211_wext_siwgenie(struct net_device *dev, if (!ie_len) ie = NULL; + wdev_lock(wdev); + /* no change */ + err = 0; if (wdev->wext.ie_len == ie_len && memcmp(wdev->wext.ie, ie, ie_len) == 0) - return 0; + goto out; if (ie_len) { ie = kmemdup(extra, ie_len, GFP_KERNEL); - if (!ie) - return -ENOMEM; + if (!ie) { + err = -ENOMEM; + goto out; + } } else ie = NULL; @@ -287,14 +328,17 @@ int cfg80211_wext_siwgenie(struct net_device *dev, wdev->wext.ie_len = ie_len; if (wdev->sme_state != CFG80211_SME_IDLE) { - err = cfg80211_disconnect(rdev, dev, - WLAN_REASON_DEAUTH_LEAVING, false); + err = __cfg80211_disconnect(rdev, dev, + WLAN_REASON_DEAUTH_LEAVING, false); if (err) - return err; + goto out; } /* userspace better not think we'll reconnect */ - return 0; + err = 0; + out: + wdev_unlock(wdev); + return err; } EXPORT_SYMBOL_GPL(cfg80211_wext_siwgenie); @@ -305,6 +349,7 @@ int cfg80211_wext_siwmlme(struct net_device *dev, struct wireless_dev *wdev = dev->ieee80211_ptr; struct iw_mlme *mlme = (struct iw_mlme *)extra; struct cfg80211_registered_device *rdev; + int err; if (!wdev) return -EOPNOTSUPP; @@ -317,13 +362,19 @@ int cfg80211_wext_siwmlme(struct net_device *dev, if (mlme->addr.sa_family != ARPHRD_ETHER) return -EINVAL; + wdev_lock(wdev); switch (mlme->cmd) { case IW_MLME_DEAUTH: case IW_MLME_DISASSOC: - return cfg80211_disconnect(rdev, dev, mlme->reason_code, - true); + err = __cfg80211_disconnect(rdev, dev, mlme->reason_code, + true); + break; default: - return -EOPNOTSUPP; + err = -EOPNOTSUPP; + break; } + wdev_unlock(wdev); + + return err; } EXPORT_SYMBOL_GPL(cfg80211_wext_siwmlme); -- cgit v1.2.3 From c99e6efe1ba04561e7d93a81f0be07e37427e835 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Fri, 10 Jul 2009 14:57:56 +0200 Subject: sched: INIT_PREEMPT_COUNT Pull the initial preempt_count value into a single definition site. Maintainers for: alpha, ia64 and m68k, please have a look, your arch code is funny. The header magic is a bit odd, but similar to the KERNEL_DS one, CPP waits with expanding these macros until the INIT_THREAD_INFO macro itself is expanded, which is in arch/*/kernel/init_task.c where we've already included sched.h so we're good. Cc: tony.luck@intel.com Cc: rth@twiddle.net Cc: geert@linux-m68k.org Signed-off-by: Peter Zijlstra Acked-by: Matt Mackall Signed-off-by: Linus Torvalds --- arch/alpha/include/asm/thread_info.h | 1 + arch/arm/include/asm/thread_info.h | 2 +- arch/avr32/include/asm/thread_info.h | 2 +- arch/blackfin/include/asm/thread_info.h | 2 +- arch/cris/include/asm/thread_info.h | 4 +--- arch/frv/include/asm/thread_info.h | 4 +--- arch/h8300/include/asm/thread_info.h | 2 +- arch/ia64/include/asm/thread_info.h | 2 +- arch/m32r/include/asm/thread_info.h | 4 +--- arch/m68k/include/asm/thread_info_mm.h | 1 + arch/m68k/include/asm/thread_info_no.h | 1 + arch/microblaze/include/asm/thread_info.h | 4 +--- arch/mips/include/asm/thread_info.h | 4 +--- arch/mn10300/include/asm/thread_info.h | 4 +--- arch/parisc/include/asm/thread_info.h | 2 +- arch/powerpc/include/asm/thread_info.h | 4 +--- arch/s390/include/asm/thread_info.h | 2 +- arch/sh/include/asm/thread_info.h | 2 +- arch/sparc/include/asm/thread_info_32.h | 4 +--- arch/sparc/include/asm/thread_info_64.h | 4 +--- arch/um/include/asm/thread_info.h | 2 +- arch/x86/include/asm/thread_info.h | 2 +- arch/xtensa/include/asm/thread_info.h | 4 +--- include/linux/sched.h | 6 ++++++ 24 files changed, 29 insertions(+), 40 deletions(-) (limited to 'include') diff --git a/arch/alpha/include/asm/thread_info.h b/arch/alpha/include/asm/thread_info.h index d069526bd767..60c83abfde70 100644 --- a/arch/alpha/include/asm/thread_info.h +++ b/arch/alpha/include/asm/thread_info.h @@ -37,6 +37,7 @@ struct thread_info { .task = &tsk, \ .exec_domain = &default_exec_domain, \ .addr_limit = KERNEL_DS, \ + .preempt_count = INIT_PREEMPT_COUNT, \ .restart_block = { \ .fn = do_no_restart_syscall, \ }, \ diff --git a/arch/arm/include/asm/thread_info.h b/arch/arm/include/asm/thread_info.h index 4f8848260ee2..73394e50cbca 100644 --- a/arch/arm/include/asm/thread_info.h +++ b/arch/arm/include/asm/thread_info.h @@ -73,7 +73,7 @@ struct thread_info { .task = &tsk, \ .exec_domain = &default_exec_domain, \ .flags = 0, \ - .preempt_count = 1, \ + .preempt_count = INIT_PREEMPT_COUNT, \ .addr_limit = KERNEL_DS, \ .cpu_domain = domain_val(DOMAIN_USER, DOMAIN_MANAGER) | \ domain_val(DOMAIN_KERNEL, DOMAIN_MANAGER) | \ diff --git a/arch/avr32/include/asm/thread_info.h b/arch/avr32/include/asm/thread_info.h index 4442f8d2d423..fc42de5ca209 100644 --- a/arch/avr32/include/asm/thread_info.h +++ b/arch/avr32/include/asm/thread_info.h @@ -40,7 +40,7 @@ struct thread_info { .exec_domain = &default_exec_domain, \ .flags = 0, \ .cpu = 0, \ - .preempt_count = 1, \ + .preempt_count = INIT_PREEMPT_COUNT, \ .restart_block = { \ .fn = do_no_restart_syscall \ } \ diff --git a/arch/blackfin/include/asm/thread_info.h b/arch/blackfin/include/asm/thread_info.h index 2920087516f2..2bbfdd950afc 100644 --- a/arch/blackfin/include/asm/thread_info.h +++ b/arch/blackfin/include/asm/thread_info.h @@ -77,7 +77,7 @@ struct thread_info { .exec_domain = &default_exec_domain, \ .flags = 0, \ .cpu = 0, \ - .preempt_count = 1, \ + .preempt_count = INIT_PREEMPT_COUNT, \ .restart_block = { \ .fn = do_no_restart_syscall, \ }, \ diff --git a/arch/cris/include/asm/thread_info.h b/arch/cris/include/asm/thread_info.h index bc5b2935ca53..c3aade36c330 100644 --- a/arch/cris/include/asm/thread_info.h +++ b/arch/cris/include/asm/thread_info.h @@ -50,8 +50,6 @@ struct thread_info { /* * macros/functions for gaining access to the thread information structure - * - * preempt_count needs to be 1 initially, until the scheduler is functional. */ #ifndef __ASSEMBLY__ #define INIT_THREAD_INFO(tsk) \ @@ -60,7 +58,7 @@ struct thread_info { .exec_domain = &default_exec_domain, \ .flags = 0, \ .cpu = 0, \ - .preempt_count = 1, \ + .preempt_count = INIT_PREEMPT_COUNT, \ .addr_limit = KERNEL_DS, \ .restart_block = { \ .fn = do_no_restart_syscall, \ diff --git a/arch/frv/include/asm/thread_info.h b/arch/frv/include/asm/thread_info.h index e8a5ed7be021..e608e056bb53 100644 --- a/arch/frv/include/asm/thread_info.h +++ b/arch/frv/include/asm/thread_info.h @@ -56,8 +56,6 @@ struct thread_info { /* * macros/functions for gaining access to the thread information structure - * - * preempt_count needs to be 1 initially, until the scheduler is functional. */ #ifndef __ASSEMBLY__ @@ -67,7 +65,7 @@ struct thread_info { .exec_domain = &default_exec_domain, \ .flags = 0, \ .cpu = 0, \ - .preempt_count = 1, \ + .preempt_count = INIT_PREEMPT_COUNT, \ .addr_limit = KERNEL_DS, \ .restart_block = { \ .fn = do_no_restart_syscall, \ diff --git a/arch/h8300/include/asm/thread_info.h b/arch/h8300/include/asm/thread_info.h index 700014d2155f..8bbc8b0ee45d 100644 --- a/arch/h8300/include/asm/thread_info.h +++ b/arch/h8300/include/asm/thread_info.h @@ -36,7 +36,7 @@ struct thread_info { .exec_domain = &default_exec_domain, \ .flags = 0, \ .cpu = 0, \ - .preempt_count = 1, \ + .preempt_count = INIT_PREEMPT_COUNT, \ .restart_block = { \ .fn = do_no_restart_syscall, \ }, \ diff --git a/arch/ia64/include/asm/thread_info.h b/arch/ia64/include/asm/thread_info.h index ae6922626bf4..8ce2e388e37c 100644 --- a/arch/ia64/include/asm/thread_info.h +++ b/arch/ia64/include/asm/thread_info.h @@ -48,7 +48,7 @@ struct thread_info { .flags = 0, \ .cpu = 0, \ .addr_limit = KERNEL_DS, \ - .preempt_count = 0, \ + .preempt_count = INIT_PREEMPT_COUNT, \ .restart_block = { \ .fn = do_no_restart_syscall, \ }, \ diff --git a/arch/m32r/include/asm/thread_info.h b/arch/m32r/include/asm/thread_info.h index 8589d462df27..07bb5bd00e2a 100644 --- a/arch/m32r/include/asm/thread_info.h +++ b/arch/m32r/include/asm/thread_info.h @@ -57,8 +57,6 @@ struct thread_info { /* * macros/functions for gaining access to the thread information structure - * - * preempt_count needs to be 1 initially, until the scheduler is functional. */ #ifndef __ASSEMBLY__ @@ -68,7 +66,7 @@ struct thread_info { .exec_domain = &default_exec_domain, \ .flags = 0, \ .cpu = 0, \ - .preempt_count = 1, \ + .preempt_count = INIT_PREEMPT_COUNT, \ .addr_limit = KERNEL_DS, \ .restart_block = { \ .fn = do_no_restart_syscall, \ diff --git a/arch/m68k/include/asm/thread_info_mm.h b/arch/m68k/include/asm/thread_info_mm.h index af0fda46e94b..6ea5c33b3c56 100644 --- a/arch/m68k/include/asm/thread_info_mm.h +++ b/arch/m68k/include/asm/thread_info_mm.h @@ -19,6 +19,7 @@ struct thread_info { { \ .task = &tsk, \ .exec_domain = &default_exec_domain, \ + .preempt_count = INIT_PREEMPT_COUNT, \ .restart_block = { \ .fn = do_no_restart_syscall, \ }, \ diff --git a/arch/m68k/include/asm/thread_info_no.h b/arch/m68k/include/asm/thread_info_no.h index 82529f424ea3..c2bde5e24b0b 100644 --- a/arch/m68k/include/asm/thread_info_no.h +++ b/arch/m68k/include/asm/thread_info_no.h @@ -49,6 +49,7 @@ struct thread_info { .exec_domain = &default_exec_domain, \ .flags = 0, \ .cpu = 0, \ + .preempt_count = INIT_PREEMPT_COUNT, \ .restart_block = { \ .fn = do_no_restart_syscall, \ }, \ diff --git a/arch/microblaze/include/asm/thread_info.h b/arch/microblaze/include/asm/thread_info.h index 7fac44498445..6e92885d381a 100644 --- a/arch/microblaze/include/asm/thread_info.h +++ b/arch/microblaze/include/asm/thread_info.h @@ -75,8 +75,6 @@ struct thread_info { /* * macros/functions for gaining access to the thread information structure - * - * preempt_count needs to be 1 initially, until the scheduler is functional. */ #define INIT_THREAD_INFO(tsk) \ { \ @@ -84,7 +82,7 @@ struct thread_info { .exec_domain = &default_exec_domain, \ .flags = 0, \ .cpu = 0, \ - .preempt_count = 1, \ + .preempt_count = INIT_PREEMPT_COUNT, \ .addr_limit = KERNEL_DS, \ .restart_block = { \ .fn = do_no_restart_syscall, \ diff --git a/arch/mips/include/asm/thread_info.h b/arch/mips/include/asm/thread_info.h index 143a48136a4b..f9df720d2e40 100644 --- a/arch/mips/include/asm/thread_info.h +++ b/arch/mips/include/asm/thread_info.h @@ -39,8 +39,6 @@ struct thread_info { /* * macros/functions for gaining access to the thread information structure - * - * preempt_count needs to be 1 initially, until the scheduler is functional. */ #define INIT_THREAD_INFO(tsk) \ { \ @@ -48,7 +46,7 @@ struct thread_info { .exec_domain = &default_exec_domain, \ .flags = _TIF_FIXADE, \ .cpu = 0, \ - .preempt_count = 1, \ + .preempt_count = INIT_PREEMPT_COUNT, \ .addr_limit = KERNEL_DS, \ .restart_block = { \ .fn = do_no_restart_syscall, \ diff --git a/arch/mn10300/include/asm/thread_info.h b/arch/mn10300/include/asm/thread_info.h index 78a3881f3c12..58d64f8b2cc3 100644 --- a/arch/mn10300/include/asm/thread_info.h +++ b/arch/mn10300/include/asm/thread_info.h @@ -65,8 +65,6 @@ struct thread_info { /* * macros/functions for gaining access to the thread information structure - * - * preempt_count needs to be 1 initially, until the scheduler is functional. */ #ifndef __ASSEMBLY__ @@ -76,7 +74,7 @@ struct thread_info { .exec_domain = &default_exec_domain, \ .flags = 0, \ .cpu = 0, \ - .preempt_count = 1, \ + .preempt_count = INIT_PREEMPT_COUNT, \ .addr_limit = KERNEL_DS, \ .restart_block = { \ .fn = do_no_restart_syscall, \ diff --git a/arch/parisc/include/asm/thread_info.h b/arch/parisc/include/asm/thread_info.h index 0407959da489..4ce0edfbe969 100644 --- a/arch/parisc/include/asm/thread_info.h +++ b/arch/parisc/include/asm/thread_info.h @@ -23,7 +23,7 @@ struct thread_info { .flags = 0, \ .cpu = 0, \ .addr_limit = KERNEL_DS, \ - .preempt_count = 1, \ + .preempt_count = INIT_PREEMPT_COUNT, \ .restart_block = { \ .fn = do_no_restart_syscall \ } \ diff --git a/arch/powerpc/include/asm/thread_info.h b/arch/powerpc/include/asm/thread_info.h index 9aba5a38a7c4..c8b329255678 100644 --- a/arch/powerpc/include/asm/thread_info.h +++ b/arch/powerpc/include/asm/thread_info.h @@ -46,15 +46,13 @@ struct thread_info { /* * macros/functions for gaining access to the thread information structure - * - * preempt_count needs to be 1 initially, until the scheduler is functional. */ #define INIT_THREAD_INFO(tsk) \ { \ .task = &tsk, \ .exec_domain = &default_exec_domain, \ .cpu = 0, \ - .preempt_count = 1, \ + .preempt_count = INIT_PREEMPT_COUNT, \ .restart_block = { \ .fn = do_no_restart_syscall, \ }, \ diff --git a/arch/s390/include/asm/thread_info.h b/arch/s390/include/asm/thread_info.h index 925bcc649035..ba1cab9fc1f9 100644 --- a/arch/s390/include/asm/thread_info.h +++ b/arch/s390/include/asm/thread_info.h @@ -61,7 +61,7 @@ struct thread_info { .exec_domain = &default_exec_domain, \ .flags = 0, \ .cpu = 0, \ - .preempt_count = 1, \ + .preempt_count = INIT_PREEMPT_COUNT, \ .restart_block = { \ .fn = do_no_restart_syscall, \ }, \ diff --git a/arch/sh/include/asm/thread_info.h b/arch/sh/include/asm/thread_info.h index f09ac4806294..d570ac2e5cb9 100644 --- a/arch/sh/include/asm/thread_info.h +++ b/arch/sh/include/asm/thread_info.h @@ -51,7 +51,7 @@ struct thread_info { .exec_domain = &default_exec_domain, \ .flags = 0, \ .cpu = 0, \ - .preempt_count = 1, \ + .preempt_count = INIT_PREEMPT_COUNT, \ .addr_limit = KERNEL_DS, \ .restart_block = { \ .fn = do_no_restart_syscall, \ diff --git a/arch/sparc/include/asm/thread_info_32.h b/arch/sparc/include/asm/thread_info_32.h index 0f7b0e5fb1c7..844d73a0340c 100644 --- a/arch/sparc/include/asm/thread_info_32.h +++ b/arch/sparc/include/asm/thread_info_32.h @@ -54,8 +54,6 @@ struct thread_info { /* * macros/functions for gaining access to the thread information structure - * - * preempt_count needs to be 1 initially, until the scheduler is functional. */ #define INIT_THREAD_INFO(tsk) \ { \ @@ -64,7 +62,7 @@ struct thread_info { .exec_domain = &default_exec_domain, \ .flags = 0, \ .cpu = 0, \ - .preempt_count = 1, \ + .preempt_count = INIT_PREEMPT_COUNT, \ .restart_block = { \ .fn = do_no_restart_syscall, \ }, \ diff --git a/arch/sparc/include/asm/thread_info_64.h b/arch/sparc/include/asm/thread_info_64.h index 65865726b283..1b45a7bbe407 100644 --- a/arch/sparc/include/asm/thread_info_64.h +++ b/arch/sparc/include/asm/thread_info_64.h @@ -125,8 +125,6 @@ struct thread_info { /* * macros/functions for gaining access to the thread information structure - * - * preempt_count needs to be 1 initially, until the scheduler is functional. */ #ifndef __ASSEMBLY__ @@ -135,7 +133,7 @@ struct thread_info { .task = &tsk, \ .flags = ((unsigned long)ASI_P) << TI_FLAG_CURRENT_DS_SHIFT, \ .exec_domain = &default_exec_domain, \ - .preempt_count = 1, \ + .preempt_count = INIT_PREEMPT_COUNT, \ .restart_block = { \ .fn = do_no_restart_syscall, \ }, \ diff --git a/arch/um/include/asm/thread_info.h b/arch/um/include/asm/thread_info.h index 62274ab9471f..fd911f855367 100644 --- a/arch/um/include/asm/thread_info.h +++ b/arch/um/include/asm/thread_info.h @@ -32,7 +32,7 @@ struct thread_info { .exec_domain = &default_exec_domain, \ .flags = 0, \ .cpu = 0, \ - .preempt_count = 1, \ + .preempt_count = INIT_PREEMPT_COUNT, \ .addr_limit = KERNEL_DS, \ .restart_block = { \ .fn = do_no_restart_syscall, \ diff --git a/arch/x86/include/asm/thread_info.h b/arch/x86/include/asm/thread_info.h index b0783520988b..fad7d40b75f8 100644 --- a/arch/x86/include/asm/thread_info.h +++ b/arch/x86/include/asm/thread_info.h @@ -49,7 +49,7 @@ struct thread_info { .exec_domain = &default_exec_domain, \ .flags = 0, \ .cpu = 0, \ - .preempt_count = 1, \ + .preempt_count = INIT_PREEMPT_COUNT, \ .addr_limit = KERNEL_DS, \ .restart_block = { \ .fn = do_no_restart_syscall, \ diff --git a/arch/xtensa/include/asm/thread_info.h b/arch/xtensa/include/asm/thread_info.h index 0f4fe1faf9ba..13165641cc51 100644 --- a/arch/xtensa/include/asm/thread_info.h +++ b/arch/xtensa/include/asm/thread_info.h @@ -80,8 +80,6 @@ struct thread_info { /* * macros/functions for gaining access to the thread information structure - * - * preempt_count needs to be 1 initially, until the scheduler is functional. */ #ifndef __ASSEMBLY__ @@ -92,7 +90,7 @@ struct thread_info { .exec_domain = &default_exec_domain, \ .flags = 0, \ .cpu = 0, \ - .preempt_count = 1, \ + .preempt_count = INIT_PREEMPT_COUNT, \ .addr_limit = KERNEL_DS, \ .restart_block = { \ .fn = do_no_restart_syscall, \ diff --git a/include/linux/sched.h b/include/linux/sched.h index 0085d758d645..2a99f1c15cf8 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -498,6 +498,12 @@ struct task_cputime { .sum_exec_runtime = 0, \ } +/* + * Disable preemption until the scheduler is running. + * Reset by start_kernel()->sched_init()->init_idle(). + */ +#define INIT_PREEMPT_COUNT (1) + /** * struct thread_group_cputimer - thread group interval timer counts * @cputime: thread group interval timers. -- cgit v1.2.3 From d86ee4809d0329d4aa0d0f2c76c2295a16862799 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Fri, 10 Jul 2009 14:57:57 +0200 Subject: sched: optimize cond_resched() Optimize cond_resched() by removing one conditional. Currently cond_resched() checks system_state == SYSTEM_RUNNING in order to avoid scheduling before the scheduler is running. We can however, as per suggestion of Matt, use PREEMPT_ACTIVE to accomplish that very same. Suggested-by: Matt Mackall Signed-off-by: Peter Zijlstra Acked-by: Matt Mackall Signed-off-by: Linus Torvalds --- include/linux/sched.h | 5 ++++- kernel/sched.c | 14 +++++++++----- 2 files changed, 13 insertions(+), 6 deletions(-) (limited to 'include') diff --git a/include/linux/sched.h b/include/linux/sched.h index 2a99f1c15cf8..16a982e389fb 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -501,8 +501,11 @@ struct task_cputime { /* * Disable preemption until the scheduler is running. * Reset by start_kernel()->sched_init()->init_idle(). + * + * We include PREEMPT_ACTIVE to avoid cond_resched() from working + * before the scheduler is active -- see should_resched(). */ -#define INIT_PREEMPT_COUNT (1) +#define INIT_PREEMPT_COUNT (1 + PREEMPT_ACTIVE) /** * struct thread_group_cputimer - thread group interval timer counts diff --git a/kernel/sched.c b/kernel/sched.c index 7c9098d186e6..01f55ada3598 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -6541,6 +6541,11 @@ SYSCALL_DEFINE0(sched_yield) return 0; } +static inline int should_resched(void) +{ + return need_resched() && !(preempt_count() & PREEMPT_ACTIVE); +} + static void __cond_resched(void) { #ifdef CONFIG_DEBUG_SPINLOCK_SLEEP @@ -6560,8 +6565,7 @@ static void __cond_resched(void) int __sched _cond_resched(void) { - if (need_resched() && !(preempt_count() & PREEMPT_ACTIVE) && - system_state == SYSTEM_RUNNING) { + if (should_resched()) { __cond_resched(); return 1; } @@ -6579,12 +6583,12 @@ EXPORT_SYMBOL(_cond_resched); */ int cond_resched_lock(spinlock_t *lock) { - int resched = need_resched() && system_state == SYSTEM_RUNNING; + int resched = should_resched(); int ret = 0; if (spin_needbreak(lock) || resched) { spin_unlock(lock); - if (resched && need_resched()) + if (resched) __cond_resched(); else cpu_relax(); @@ -6599,7 +6603,7 @@ int __sched cond_resched_softirq(void) { BUG_ON(!in_softirq()); - if (need_resched() && system_state == SYSTEM_RUNNING) { + if (should_resched()) { local_bh_enable(); __cond_resched(); local_bh_disable(); -- cgit v1.2.3 From 5a421ce3c062a87db0a9e7f2a0a7ee0a5b869aab Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Fri, 10 Jul 2009 12:37:40 +0300 Subject: nfsd41: gather and report statistics also for v4.1 ops Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- include/linux/nfs4.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/nfs4.h b/include/linux/nfs4.h index bd2eba530667..aff924a24abb 100644 --- a/include/linux/nfs4.h +++ b/include/linux/nfs4.h @@ -234,7 +234,7 @@ enum nfs_opnum4 { Needs to be updated if more operations are defined in future.*/ #define FIRST_NFS4_OP OP_ACCESS -#define LAST_NFS4_OP OP_RELEASE_LOCKOWNER +#define LAST_NFS4_OP OP_RECLAIM_COMPLETE enum nfsstat4 { NFS4_OK = 0, -- cgit v1.2.3 From 24a15a62dcb1968bf4ffdae55c88fa934d971180 Mon Sep 17 00:00:00 2001 From: Alan Cox Date: Thu, 9 Jul 2009 13:36:22 +0100 Subject: tty: Fix USB kref leak The sysrq code acquired a kref leak. Fix it by passing the tty separately from the caller (thus effectively using the callers kref which all the callers hold anyway) Signed-off-by: Alan Cox Signed-off-by: Linus Torvalds --- drivers/usb/serial/ftdi_sio.c | 2 +- drivers/usb/serial/generic.c | 7 ++++--- drivers/usb/serial/pl2303.c | 2 +- include/linux/usb/serial.h | 3 ++- 4 files changed, 8 insertions(+), 6 deletions(-) (limited to 'include') diff --git a/drivers/usb/serial/ftdi_sio.c b/drivers/usb/serial/ftdi_sio.c index 3dc3768ca71c..5f08702f672f 100644 --- a/drivers/usb/serial/ftdi_sio.c +++ b/drivers/usb/serial/ftdi_sio.c @@ -2121,7 +2121,7 @@ static void ftdi_process_read(struct work_struct *work) /* Note that the error flag is duplicated for every character received since we don't know which character it applied to */ - if (!usb_serial_handle_sysrq_char(port, + if (!usb_serial_handle_sysrq_char(tty, port, data[packet_offset + i])) tty_insert_flip_char(tty, data[packet_offset + i], diff --git a/drivers/usb/serial/generic.c b/drivers/usb/serial/generic.c index 3d8dc5671bea..ce57f6a32bdf 100644 --- a/drivers/usb/serial/generic.c +++ b/drivers/usb/serial/generic.c @@ -432,7 +432,7 @@ static void flush_and_resubmit_read_urb(struct usb_serial_port *port) else { /* Push data to tty */ for (i = 0; i < urb->actual_length; i++, ch++) { - if (!usb_serial_handle_sysrq_char(port, *ch)) + if (!usb_serial_handle_sysrq_char(tty, port, *ch)) tty_insert_flip_char(tty, *ch, TTY_NORMAL); } } @@ -534,11 +534,12 @@ void usb_serial_generic_unthrottle(struct tty_struct *tty) } } -int usb_serial_handle_sysrq_char(struct usb_serial_port *port, unsigned int ch) +int usb_serial_handle_sysrq_char(struct tty_struct *tty, + struct usb_serial_port *port, unsigned int ch) { if (port->sysrq && port->console) { if (ch && time_before(jiffies, port->sysrq)) { - handle_sysrq(ch, tty_port_tty_get(&port->port)); + handle_sysrq(ch, tty); port->sysrq = 0; return 1; } diff --git a/drivers/usb/serial/pl2303.c b/drivers/usb/serial/pl2303.c index ec6c132a25b5..8835802067f7 100644 --- a/drivers/usb/serial/pl2303.c +++ b/drivers/usb/serial/pl2303.c @@ -1038,7 +1038,7 @@ static void pl2303_read_bulk_callback(struct urb *urb) if (line_status & UART_OVERRUN_ERROR) tty_insert_flip_char(tty, 0, TTY_OVERRUN); for (i = 0; i < urb->actual_length; ++i) - if (!usb_serial_handle_sysrq_char(port, data[i])) + if (!usb_serial_handle_sysrq_char(tty, port, data[i])) tty_insert_flip_char(tty, data[i], tty_flag); tty_flip_buffer_push(tty); } diff --git a/include/linux/usb/serial.h b/include/linux/usb/serial.h index 44801d26a37a..0ec50ba62139 100644 --- a/include/linux/usb/serial.h +++ b/include/linux/usb/serial.h @@ -317,7 +317,8 @@ extern int usb_serial_generic_register(int debug); extern void usb_serial_generic_deregister(void); extern void usb_serial_generic_resubmit_read_urb(struct usb_serial_port *port, gfp_t mem_flags); -extern int usb_serial_handle_sysrq_char(struct usb_serial_port *port, +extern int usb_serial_handle_sysrq_char(struct tty_struct *tty, + struct usb_serial_port *port, unsigned int ch); extern int usb_serial_handle_break(struct usb_serial_port *port); -- cgit v1.2.3 From 99d27e1c59e34869605de625b033c52163f5bfa7 Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Thu, 9 Jul 2009 20:30:57 +0800 Subject: crypto: shash - Export/import hash state only This patch replaces the full descriptor export with an export of the partial hash state. This allows the use of a consistent export format across all implementations of a given algorithm. This is useful because a number of cases require the use of the partial hash state, e.g., PadLock can use the SHA1 hash state to get around the fact that it can only hash contiguous data chunks. Signed-off-by: Herbert Xu --- crypto/shash.c | 25 ++++++++++++++----------- include/crypto/hash.h | 18 ++++++++++++++---- 2 files changed, 28 insertions(+), 15 deletions(-) (limited to 'include') diff --git a/crypto/shash.c b/crypto/shash.c index 23e05a1e9038..14a3b707a31f 100644 --- a/crypto/shash.c +++ b/crypto/shash.c @@ -172,19 +172,15 @@ int crypto_shash_digest(struct shash_desc *desc, const u8 *data, } EXPORT_SYMBOL_GPL(crypto_shash_digest); -int crypto_shash_import(struct shash_desc *desc, const u8 *in) +static int shash_no_export(struct shash_desc *desc, void *out) { - struct crypto_shash *tfm = desc->tfm; - struct shash_alg *alg = crypto_shash_alg(tfm); - - memcpy(shash_desc_ctx(desc), in, crypto_shash_descsize(tfm)); - - if (alg->reinit) - return alg->reinit(desc); + return -ENOSYS; +} - return 0; +static int shash_no_import(struct shash_desc *desc, const void *in) +{ + return -ENOSYS; } -EXPORT_SYMBOL_GPL(crypto_shash_import); static int shash_async_setkey(struct crypto_ahash *tfm, const u8 *key, unsigned int keylen) @@ -484,12 +480,19 @@ static int shash_prepare_alg(struct shash_alg *alg) struct crypto_alg *base = &alg->base; if (alg->digestsize > PAGE_SIZE / 8 || - alg->descsize > PAGE_SIZE / 8) + alg->descsize > PAGE_SIZE / 8 || + alg->statesize > PAGE_SIZE / 8) return -EINVAL; base->cra_type = &crypto_shash_type; base->cra_flags &= ~CRYPTO_ALG_TYPE_MASK; base->cra_flags |= CRYPTO_ALG_TYPE_SHASH; + + if (!alg->import) + alg->import = shash_no_import; + if (!alg->export) + alg->export = shash_no_export; + return 0; } diff --git a/include/crypto/hash.h b/include/crypto/hash.h index d56bb71617c3..3c4cce6a425c 100644 --- a/include/crypto/hash.h +++ b/include/crypto/hash.h @@ -24,7 +24,6 @@ struct shash_desc { struct shash_alg { int (*init)(struct shash_desc *desc); - int (*reinit)(struct shash_desc *desc); int (*update)(struct shash_desc *desc, const u8 *data, unsigned int len); int (*final)(struct shash_desc *desc, u8 *out); @@ -32,11 +31,14 @@ struct shash_alg { unsigned int len, u8 *out); int (*digest)(struct shash_desc *desc, const u8 *data, unsigned int len, u8 *out); + int (*export)(struct shash_desc *desc, void *out); + int (*import)(struct shash_desc *desc, const void *in); int (*setkey)(struct crypto_shash *tfm, const u8 *key, unsigned int keylen); unsigned int descsize; unsigned int digestsize; + unsigned int statesize; struct crypto_alg base; }; @@ -251,6 +253,11 @@ static inline unsigned int crypto_shash_digestsize(struct crypto_shash *tfm) return crypto_shash_alg(tfm)->digestsize; } +static inline unsigned int crypto_shash_statesize(struct crypto_shash *tfm) +{ + return crypto_shash_alg(tfm)->statesize; +} + static inline u32 crypto_shash_get_flags(struct crypto_shash *tfm) { return crypto_tfm_get_flags(crypto_shash_tfm(tfm)); @@ -281,12 +288,15 @@ int crypto_shash_setkey(struct crypto_shash *tfm, const u8 *key, int crypto_shash_digest(struct shash_desc *desc, const u8 *data, unsigned int len, u8 *out); -static inline void crypto_shash_export(struct shash_desc *desc, u8 *out) +static inline int crypto_shash_export(struct shash_desc *desc, void *out) { - memcpy(out, shash_desc_ctx(desc), crypto_shash_descsize(desc->tfm)); + return crypto_shash_alg(desc->tfm)->export(desc, out); } -int crypto_shash_import(struct shash_desc *desc, const u8 *in); +static inline int crypto_shash_import(struct shash_desc *desc, const void *in) +{ + return crypto_shash_alg(desc->tfm)->import(desc, in); +} static inline int crypto_shash_init(struct shash_desc *desc) { -- cgit v1.2.3 From e2a7ce4e185a94462698cc0e5192495ee3d22a2f Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Thu, 9 Jul 2009 21:27:13 +0800 Subject: crypto: sha1_generic - Add export/import support This patch adds export/import support to sha1_generic. The exported type is defined by struct sha1_state, which is basically the entire descriptor state of sha1_generic. Signed-off-by: Herbert Xu --- crypto/sha1_generic.c | 41 +++++++++++++++++++++++++---------------- include/crypto/sha.h | 8 ++++++++ 2 files changed, 33 insertions(+), 16 deletions(-) (limited to 'include') diff --git a/crypto/sha1_generic.c b/crypto/sha1_generic.c index 9efef20454cb..0416091bf45a 100644 --- a/crypto/sha1_generic.c +++ b/crypto/sha1_generic.c @@ -25,31 +25,21 @@ #include #include -struct sha1_ctx { - u64 count; - u32 state[5]; - u8 buffer[64]; -}; - static int sha1_init(struct shash_desc *desc) { - struct sha1_ctx *sctx = shash_desc_ctx(desc); + struct sha1_state *sctx = shash_desc_ctx(desc); - static const struct sha1_ctx initstate = { - 0, - { SHA1_H0, SHA1_H1, SHA1_H2, SHA1_H3, SHA1_H4 }, - { 0, } + *sctx = (struct sha1_state){ + .state = { SHA1_H0, SHA1_H1, SHA1_H2, SHA1_H3, SHA1_H4 }, }; - *sctx = initstate; - return 0; } static int sha1_update(struct shash_desc *desc, const u8 *data, unsigned int len) { - struct sha1_ctx *sctx = shash_desc_ctx(desc); + struct sha1_state *sctx = shash_desc_ctx(desc); unsigned int partial, done; const u8 *src; @@ -85,7 +75,7 @@ static int sha1_update(struct shash_desc *desc, const u8 *data, /* Add padding and return the message digest. */ static int sha1_final(struct shash_desc *desc, u8 *out) { - struct sha1_ctx *sctx = shash_desc_ctx(desc); + struct sha1_state *sctx = shash_desc_ctx(desc); __be32 *dst = (__be32 *)out; u32 i, index, padlen; __be64 bits; @@ -111,12 +101,31 @@ static int sha1_final(struct shash_desc *desc, u8 *out) return 0; } +static int sha1_export(struct shash_desc *desc, void *out) +{ + struct sha1_state *sctx = shash_desc_ctx(desc); + + memcpy(out, sctx, sizeof(*sctx)); + return 0; +} + +static int sha1_import(struct shash_desc *desc, const void *in) +{ + struct sha1_state *sctx = shash_desc_ctx(desc); + + memcpy(sctx, in, sizeof(*sctx)); + return 0; +} + static struct shash_alg alg = { .digestsize = SHA1_DIGEST_SIZE, .init = sha1_init, .update = sha1_update, .final = sha1_final, - .descsize = sizeof(struct sha1_ctx), + .export = sha1_export, + .import = sha1_import, + .descsize = sizeof(struct sha1_state), + .statesize = sizeof(struct sha1_state), .base = { .cra_name = "sha1", .cra_driver_name= "sha1-generic", diff --git a/include/crypto/sha.h b/include/crypto/sha.h index c0ccc2b1a2d8..922a248bd04d 100644 --- a/include/crypto/sha.h +++ b/include/crypto/sha.h @@ -5,6 +5,8 @@ #ifndef _CRYPTO_SHA_H #define _CRYPTO_SHA_H +#include + #define SHA1_DIGEST_SIZE 20 #define SHA1_BLOCK_SIZE 64 @@ -62,4 +64,10 @@ #define SHA512_H6 0x1f83d9abfb41bd6bULL #define SHA512_H7 0x5be0cd19137e2179ULL +struct sha1_state { + u64 count; + u32 state[SHA1_DIGEST_SIZE / 4]; + u8 buffer[SHA1_BLOCK_SIZE]; +}; + #endif -- cgit v1.2.3 From 9b2fda7b94a769af13c24582739e50664b0e27a8 Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Fri, 10 Jul 2009 13:00:27 +0800 Subject: crypto: sha256_generic - Add export/import support This patch adds export/import support to sha256_generic. The exported type is defined by struct sha256_state, which is basically the entire descriptor state of sha256_generic. Signed-off-by: Herbert Xu --- crypto/sha256_generic.c | 37 +++++++++++++++++++++++++------------ include/crypto/sha.h | 6 ++++++ 2 files changed, 31 insertions(+), 12 deletions(-) (limited to 'include') diff --git a/crypto/sha256_generic.c b/crypto/sha256_generic.c index e58c71bdf333..c48459ebf05b 100644 --- a/crypto/sha256_generic.c +++ b/crypto/sha256_generic.c @@ -25,12 +25,6 @@ #include #include -struct sha256_ctx { - u64 count; - u32 state[8]; - u8 buf[128]; -}; - static inline u32 Ch(u32 x, u32 y, u32 z) { return z ^ (x & (y ^ z)); @@ -222,7 +216,7 @@ static void sha256_transform(u32 *state, const u8 *input) static int sha224_init(struct shash_desc *desc) { - struct sha256_ctx *sctx = shash_desc_ctx(desc); + struct sha256_state *sctx = shash_desc_ctx(desc); sctx->state[0] = SHA224_H0; sctx->state[1] = SHA224_H1; sctx->state[2] = SHA224_H2; @@ -238,7 +232,7 @@ static int sha224_init(struct shash_desc *desc) static int sha256_init(struct shash_desc *desc) { - struct sha256_ctx *sctx = shash_desc_ctx(desc); + struct sha256_state *sctx = shash_desc_ctx(desc); sctx->state[0] = SHA256_H0; sctx->state[1] = SHA256_H1; sctx->state[2] = SHA256_H2; @@ -255,7 +249,7 @@ static int sha256_init(struct shash_desc *desc) static int sha256_update(struct shash_desc *desc, const u8 *data, unsigned int len) { - struct sha256_ctx *sctx = shash_desc_ctx(desc); + struct sha256_state *sctx = shash_desc_ctx(desc); unsigned int partial, done; const u8 *src; @@ -286,7 +280,7 @@ static int sha256_update(struct shash_desc *desc, const u8 *data, static int sha256_final(struct shash_desc *desc, u8 *out) { - struct sha256_ctx *sctx = shash_desc_ctx(desc); + struct sha256_state *sctx = shash_desc_ctx(desc); __be32 *dst = (__be32 *)out; __be64 bits; unsigned int index, pad_len; @@ -326,12 +320,31 @@ static int sha224_final(struct shash_desc *desc, u8 *hash) return 0; } +static int sha256_export(struct shash_desc *desc, void *out) +{ + struct sha256_state *sctx = shash_desc_ctx(desc); + + memcpy(out, sctx, sizeof(*sctx)); + return 0; +} + +static int sha256_import(struct shash_desc *desc, const void *in) +{ + struct sha256_state *sctx = shash_desc_ctx(desc); + + memcpy(sctx, in, sizeof(*sctx)); + return 0; +} + static struct shash_alg sha256 = { .digestsize = SHA256_DIGEST_SIZE, .init = sha256_init, .update = sha256_update, .final = sha256_final, - .descsize = sizeof(struct sha256_ctx), + .export = sha256_export, + .import = sha256_import, + .descsize = sizeof(struct sha256_state), + .statesize = sizeof(struct sha256_state), .base = { .cra_name = "sha256", .cra_driver_name= "sha256-generic", @@ -346,7 +359,7 @@ static struct shash_alg sha224 = { .init = sha224_init, .update = sha256_update, .final = sha224_final, - .descsize = sizeof(struct sha256_ctx), + .descsize = sizeof(struct sha256_state), .base = { .cra_name = "sha224", .cra_driver_name= "sha224-generic", diff --git a/include/crypto/sha.h b/include/crypto/sha.h index 922a248bd04d..88ef5eb9514d 100644 --- a/include/crypto/sha.h +++ b/include/crypto/sha.h @@ -70,4 +70,10 @@ struct sha1_state { u8 buffer[SHA1_BLOCK_SIZE]; }; +struct sha256_state { + u64 count; + u32 state[SHA256_DIGEST_SIZE / 4]; + u8 buf[SHA256_BLOCK_SIZE]; +}; + #endif -- cgit v1.2.3 From 373c0a7ed3ea3b34efedb7c83ffb521adff7c894 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Sat, 11 Jul 2009 10:06:54 -0400 Subject: Fix compile error due to congestion_wait() changes Move the definition of BLK_RW_ASYNC/BLK_RW_SYNC into linux/backing-dev.h so that it is available to all callers of set/clear_bdi_congested(). This replaces commit 097041e576ee3a50d92dd643ee8ca65bf6a62e21 ("fuse: Fix build error"), which will be reverted. Signed-off-by: Trond Myklebust Acked-by: Larry Finger Cc: Jens Axboe Cc: Miklos Szeredi Signed-off-by: Linus Torvalds --- include/linux/backing-dev.h | 5 +++++ include/linux/blkdev.h | 5 ----- 2 files changed, 5 insertions(+), 5 deletions(-) (limited to 'include') diff --git a/include/linux/backing-dev.h b/include/linux/backing-dev.h index 3a52a63c1351..1d52425a6118 100644 --- a/include/linux/backing-dev.h +++ b/include/linux/backing-dev.h @@ -229,6 +229,11 @@ static inline int bdi_rw_congested(struct backing_dev_info *bdi) (1 << BDI_async_congested)); } +enum { + BLK_RW_ASYNC = 0, + BLK_RW_SYNC = 1, +}; + void clear_bdi_congested(struct backing_dev_info *bdi, int sync); void set_bdi_congested(struct backing_dev_info *bdi, int sync); long congestion_wait(int sync, long timeout); diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index 0146e0fecf1a..e7cb5dbf6c26 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -70,11 +70,6 @@ enum rq_cmd_type_bits { REQ_TYPE_ATA_PC, }; -enum { - BLK_RW_ASYNC = 0, - BLK_RW_SYNC = 1, -}; - /* * For request of type REQ_TYPE_LINUX_BLOCK, rq->cmd[0] is the opcode being * sent down (similar to how REQ_TYPE_BLOCK_PC means that ->cmd[] holds a -- cgit v1.2.3 From aef73cfcb913eae3d0deeb60eb385f75039db40b Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Sat, 11 Jul 2009 22:22:14 +0800 Subject: crypto: async - Use kzfree for requests This patch changes the kfree call to kzfree for async requests. As the request may contain sensitive data it needs to be zeroed before it can be reallocated by others. Signed-off-by: Herbert Xu --- include/crypto/hash.h | 2 +- include/linux/crypto.h | 4 ++-- 2 files changed, 3 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/include/crypto/hash.h b/include/crypto/hash.h index 3c4cce6a425c..f74214a4b01b 100644 --- a/include/crypto/hash.h +++ b/include/crypto/hash.h @@ -186,7 +186,7 @@ static inline struct ahash_request *ahash_request_alloc( static inline void ahash_request_free(struct ahash_request *req) { - kfree(req); + kzfree(req); } static inline struct ahash_request *ahash_request_cast( diff --git a/include/linux/crypto.h b/include/linux/crypto.h index ec29fa268b94..274f9c7da90c 100644 --- a/include/linux/crypto.h +++ b/include/linux/crypto.h @@ -770,7 +770,7 @@ static inline struct ablkcipher_request *ablkcipher_request_alloc( static inline void ablkcipher_request_free(struct ablkcipher_request *req) { - kfree(req); + kzfree(req); } static inline void ablkcipher_request_set_callback( @@ -901,7 +901,7 @@ static inline struct aead_request *aead_request_alloc(struct crypto_aead *tfm, static inline void aead_request_free(struct aead_request *req) { - kfree(req); + kzfree(req); } static inline void aead_request_set_callback(struct aead_request *req, -- cgit v1.2.3 From f9fabcb58a6d26d6efde842d1703ac7cfa9427b6 Mon Sep 17 00:00:00 2001 From: Julien Tinnes Date: Fri, 26 Jun 2009 20:27:40 +0200 Subject: personality: fix PER_CLEAR_ON_SETID We have found that the current PER_CLEAR_ON_SETID mask on Linux doesn't include neither ADDR_COMPAT_LAYOUT, nor MMAP_PAGE_ZERO. The current mask is READ_IMPLIES_EXEC|ADDR_NO_RANDOMIZE. We believe it is important to add MMAP_PAGE_ZERO, because by using this personality it is possible to have the first page mapped inside a process running as setuid root. This could be used in those scenarios: - Exploiting a NULL pointer dereference issue in a setuid root binary - Bypassing the mmap_min_addr restrictions of the Linux kernel: by running a setuid binary that would drop privileges before giving us control back (for instance by loading a user-supplied library), we could get the first page mapped in a process we control. By further using mremap and mprotect on this mapping, we can then completely bypass the mmap_min_addr restrictions. Less importantly, we believe ADDR_COMPAT_LAYOUT should also be added since on x86 32bits it will in practice disable most of the address space layout randomization (only the stack will remain randomized). Signed-off-by: Julien Tinnes Signed-off-by: Tavis Ormandy Cc: stable@kernel.org Acked-by: Christoph Hellwig Acked-by: Kees Cook Acked-by: Eugene Teo [ Shortened lines and fixed whitespace as per Christophs' suggestion ] Signed-off-by: Linus Torvalds --- include/linux/personality.h | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/personality.h b/include/linux/personality.h index a84e9ff9b27e..126120819a0d 100644 --- a/include/linux/personality.h +++ b/include/linux/personality.h @@ -40,7 +40,10 @@ enum { * Security-relevant compatibility flags that must be * cleared upon setuid or setgid exec: */ -#define PER_CLEAR_ON_SETID (READ_IMPLIES_EXEC|ADDR_NO_RANDOMIZE) +#define PER_CLEAR_ON_SETID (READ_IMPLIES_EXEC | \ + ADDR_NO_RANDOMIZE | \ + ADDR_COMPAT_LAYOUT | \ + MMAP_PAGE_ZERO) /* * Personality types. -- cgit v1.2.3 From 405f55712dfe464b3240d7816cc4fe4174831be2 Mon Sep 17 00:00:00 2001 From: Alexey Dobriyan Date: Sat, 11 Jul 2009 22:08:37 +0400 Subject: headers: smp_lock.h redux * Remove smp_lock.h from files which don't need it (including some headers!) * Add smp_lock.h to files which do need it * Make smp_lock.h include conditional in hardirq.h It's needed only for one kernel_locked() usage which is under CONFIG_PREEMPT This will make hardirq.h inclusion cheaper for every PREEMPT=n config (which includes allmodconfig/allyesconfig, BTW) Signed-off-by: Alexey Dobriyan Signed-off-by: Linus Torvalds --- arch/alpha/kernel/ptrace.c | 1 - arch/blackfin/kernel/ptrace.c | 1 - arch/blackfin/kernel/sys_bfin.c | 1 - arch/cris/kernel/sys_cris.c | 1 - arch/ia64/kernel/ptrace.c | 1 - arch/m32r/kernel/ptrace.c | 1 - arch/microblaze/kernel/ptrace.c | 1 - arch/microblaze/kernel/signal.c | 1 - arch/microblaze/kernel/sys_microblaze.c | 1 - arch/mips/kernel/ptrace32.c | 1 - arch/mips/mm/hugetlbpage.c | 1 - arch/mn10300/kernel/ptrace.c | 1 - arch/mn10300/kernel/signal.c | 1 - arch/mn10300/kernel/sys_mn10300.c | 1 - arch/mn10300/kernel/traps.c | 1 - arch/mn10300/mm/fault.c | 1 - arch/mn10300/mm/misalignment.c | 1 - arch/powerpc/kernel/ptrace32.c | 1 - arch/s390/kernel/dis.c | 1 - arch/s390/kernel/ptrace.c | 1 - arch/s390/mm/fault.c | 1 - arch/sh/mm/tlb-sh3.c | 1 - arch/sparc/kernel/ptrace_32.c | 1 - arch/sparc/kernel/ptrace_64.c | 1 - arch/sparc/kernel/time_64.c | 1 - arch/sparc/kernel/traps_32.c | 1 - drivers/block/DAC960.c | 1 + drivers/block/cciss.c | 1 + drivers/block/loop.c | 1 - drivers/bluetooth/hci_vhci.c | 1 - drivers/char/amiserial.c | 1 + drivers/char/cyclades.c | 1 + drivers/char/epca.c | 1 + drivers/char/isicom.c | 1 + drivers/char/istallion.c | 1 + drivers/char/moxa.c | 1 + drivers/char/mxser.c | 1 + drivers/char/n_hdlc.c | 1 + drivers/char/n_r3964.c | 1 + drivers/char/pty.c | 1 + drivers/char/rio/rio_linux.c | 1 + drivers/char/riscom8.c | 1 + drivers/char/rocket.c | 1 + drivers/char/serial167.c | 1 + drivers/char/specialix.c | 1 + drivers/char/sx.c | 1 + drivers/char/synclink.c | 1 + drivers/char/synclink_gt.c | 1 + drivers/char/synclinkmp.c | 1 + drivers/char/tpm/tpm.c | 1 - drivers/char/tty_ioctl.c | 1 - drivers/char/tty_ldisc.c | 1 - drivers/char/vt.c | 1 + drivers/char/vt_ioctl.c | 1 + drivers/gpio/vr41xx_giu.c | 1 - drivers/hid/usbhid/hid-core.c | 1 - drivers/isdn/hisax/hfc_usb.c | 1 - drivers/isdn/i4l/isdn_tty.c | 1 + drivers/isdn/mISDN/stack.c | 1 + drivers/media/dvb/bt8xx/dst_ca.c | 1 + drivers/media/dvb/dvb-core/dvbdev.h | 1 - drivers/media/dvb/ttpci/av7110.c | 1 - drivers/media/radio/radio-mr800.c | 1 + drivers/media/radio/radio-si470x.c | 1 + drivers/media/video/bt8xx/bttv-driver.c | 1 + drivers/media/video/cx23885/cx23885-417.c | 1 + drivers/media/video/cx23885/cx23885-video.c | 1 + drivers/media/video/cx88/cx88-blackbird.c | 1 + drivers/media/video/cx88/cx88-video.c | 1 + drivers/media/video/dabusb.c | 1 + drivers/media/video/pwc/pwc-if.c | 1 + drivers/media/video/pwc/pwc.h | 1 - drivers/media/video/s2255drv.c | 1 + drivers/media/video/saa5246a.c | 1 - drivers/media/video/saa5249.c | 1 - drivers/media/video/saa7134/saa7134-empress.c | 1 + drivers/media/video/se401.c | 1 + drivers/media/video/stk-webcam.c | 1 + drivers/media/video/stradis.c | 1 + drivers/media/video/stv680.c | 1 + drivers/media/video/usbvideo/vicam.c | 1 + drivers/media/video/usbvision/usbvision-video.c | 1 + drivers/media/video/v4l2-dev.c | 1 - drivers/media/video/zoran/zoran_driver.c | 1 + drivers/misc/sgi-gru/grufile.c | 1 - drivers/misc/sgi-gru/grukservices.c | 1 - drivers/net/irda/irtty-sir.c | 1 - drivers/pci/hotplug/cpci_hotplug_core.c | 1 - drivers/pci/hotplug/cpqphp_ctrl.c | 1 - drivers/pci/hotplug/cpqphp_sysfs.c | 1 + drivers/pci/hotplug/pciehp_ctrl.c | 1 - drivers/pci/syscall.c | 1 - drivers/s390/block/dasd_ioctl.c | 1 + drivers/scsi/qla2xxx/qla_mid.c | 1 - drivers/telephony/ixj.c | 1 + drivers/telephony/phonedev.c | 1 - drivers/usb/class/cdc-wdm.c | 1 - drivers/usb/gadget/amd5536udc.c | 1 - drivers/usb/gadget/langwell_udc.c | 1 - drivers/usb/gadget/s3c2410_udc.c | 1 - drivers/usb/host/r8a66597-hcd.c | 1 - drivers/usb/misc/iowarrior.c | 1 + drivers/usb/misc/rio500.c | 1 + drivers/usb/misc/usblcd.c | 1 + drivers/usb/musb/cppi_dma.h | 1 - drivers/usb/musb/musb_core.h | 1 - drivers/usb/serial/ftdi_sio.c | 1 + drivers/usb/serial/mos7840.c | 1 + drivers/usb/serial/usb-serial.c | 1 + drivers/video/fbmem.c | 1 - fs/adfs/super.c | 1 + fs/afs/super.c | 1 + fs/autofs4/dev-ioctl.c | 1 - fs/bfs/dir.c | 1 - fs/bfs/file.c | 1 - fs/btrfs/compression.c | 1 - fs/btrfs/file.c | 1 - fs/btrfs/inode.c | 1 - fs/btrfs/ioctl.c | 1 - fs/btrfs/super.c | 1 - fs/char_dev.c | 1 - fs/compat.c | 1 - fs/compat_ioctl.c | 1 + fs/exofs/super.c | 1 + fs/ext2/ioctl.c | 1 - fs/ext4/ioctl.c | 1 - fs/fat/dir.c | 1 - fs/fat/namei_msdos.c | 1 - fs/fat/namei_vfat.c | 1 - fs/fcntl.c | 1 - fs/freevxfs/vxfs_super.c | 1 + fs/hfs/super.c | 1 + fs/hfsplus/super.c | 1 + fs/hpfs/dir.c | 1 + fs/hpfs/file.c | 1 + fs/hpfs/hpfs_fn.h | 1 - fs/hpfs/inode.c | 1 + fs/hpfs/namei.c | 1 + fs/jffs2/super.c | 1 + fs/lockd/clntproc.c | 1 + fs/lockd/svc4proc.c | 1 + fs/lockd/svcproc.c | 1 + fs/nfs/delegation.c | 1 + fs/nfs/dir.c | 1 - fs/nfs/file.c | 1 - fs/nfs/inode.c | 1 - fs/nfs/nfs4proc.c | 1 - fs/nfs/read.c | 1 - fs/nfsd/nfsctl.c | 1 - fs/nfsd/nfssvc.c | 1 - fs/nilfs2/dir.c | 1 - fs/ocfs2/ioctl.c | 1 - fs/reiserfs/xattr.c | 1 - fs/squashfs/super.c | 1 + fs/ubifs/ioctl.c | 1 - fs/xfs/linux-2.6/xfs_file.c | 1 - include/linux/crash_dump.h | 1 - include/linux/hardirq.h | 2 ++ include/linux/quotaops.h | 1 - include/linux/sunrpc/xdr.h | 1 - kernel/power/user.c | 1 - kernel/trace/blktrace.c | 1 + kernel/trace/trace.c | 1 + net/appletalk/ddp.c | 1 + net/ipx/af_ipx.c | 1 + net/irda/af_irda.c | 1 + net/irda/irnet/irnet.h | 1 - net/irda/irnet/irnet_ppp.c | 1 + net/sunrpc/clnt.c | 1 - net/sunrpc/sched.c | 1 - net/sunrpc/svc_xprt.c | 1 + net/wanrouter/wanmain.c | 1 + net/x25/af_x25.c | 1 + 173 files changed, 81 insertions(+), 93 deletions(-) (limited to 'include') diff --git a/arch/alpha/kernel/ptrace.c b/arch/alpha/kernel/ptrace.c index 1e9ad52c460e..e072041d19f8 100644 --- a/arch/alpha/kernel/ptrace.c +++ b/arch/alpha/kernel/ptrace.c @@ -8,7 +8,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/blackfin/kernel/ptrace.c b/arch/blackfin/kernel/ptrace.c index d76618db50df..6a387eec6b65 100644 --- a/arch/blackfin/kernel/ptrace.c +++ b/arch/blackfin/kernel/ptrace.c @@ -31,7 +31,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/blackfin/kernel/sys_bfin.c b/arch/blackfin/kernel/sys_bfin.c index a8f1329c15a4..3da60fb13ce4 100644 --- a/arch/blackfin/kernel/sys_bfin.c +++ b/arch/blackfin/kernel/sys_bfin.c @@ -29,7 +29,6 @@ * 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA */ -#include #include #include #include diff --git a/arch/cris/kernel/sys_cris.c b/arch/cris/kernel/sys_cris.c index a79fbd87021b..2ad962c7e88e 100644 --- a/arch/cris/kernel/sys_cris.c +++ b/arch/cris/kernel/sys_cris.c @@ -15,7 +15,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/ia64/kernel/ptrace.c b/arch/ia64/kernel/ptrace.c index 92c9689b7d97..9daa87fdb018 100644 --- a/arch/ia64/kernel/ptrace.c +++ b/arch/ia64/kernel/ptrace.c @@ -15,7 +15,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/m32r/kernel/ptrace.c b/arch/m32r/kernel/ptrace.c index bf0abe9e1f73..98b8feb12ed8 100644 --- a/arch/m32r/kernel/ptrace.c +++ b/arch/m32r/kernel/ptrace.c @@ -19,7 +19,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/microblaze/kernel/ptrace.c b/arch/microblaze/kernel/ptrace.c index b86aa623e36d..53ff39af6a5c 100644 --- a/arch/microblaze/kernel/ptrace.c +++ b/arch/microblaze/kernel/ptrace.c @@ -27,7 +27,6 @@ #include #include #include -#include #include #include diff --git a/arch/microblaze/kernel/signal.c b/arch/microblaze/kernel/signal.c index 493819c25fba..1c80e4fc40ce 100644 --- a/arch/microblaze/kernel/signal.c +++ b/arch/microblaze/kernel/signal.c @@ -21,7 +21,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/microblaze/kernel/sys_microblaze.c b/arch/microblaze/kernel/sys_microblaze.c index 8c9ebac5da10..e000bce09b2b 100644 --- a/arch/microblaze/kernel/sys_microblaze.c +++ b/arch/microblaze/kernel/sys_microblaze.c @@ -15,7 +15,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/mips/kernel/ptrace32.c b/arch/mips/kernel/ptrace32.c index c4f9ac17474a..32644b4a0714 100644 --- a/arch/mips/kernel/ptrace32.c +++ b/arch/mips/kernel/ptrace32.c @@ -22,7 +22,6 @@ #include #include #include -#include #include #include diff --git a/arch/mips/mm/hugetlbpage.c b/arch/mips/mm/hugetlbpage.c index 471c09aa1614..8c2834f5919d 100644 --- a/arch/mips/mm/hugetlbpage.c +++ b/arch/mips/mm/hugetlbpage.c @@ -16,7 +16,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/mn10300/kernel/ptrace.c b/arch/mn10300/kernel/ptrace.c index e143339ad28e..cf847dabc1bd 100644 --- a/arch/mn10300/kernel/ptrace.c +++ b/arch/mn10300/kernel/ptrace.c @@ -13,7 +13,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/mn10300/kernel/signal.c b/arch/mn10300/kernel/signal.c index 9f7572a0f578..feb2f2e810db 100644 --- a/arch/mn10300/kernel/signal.c +++ b/arch/mn10300/kernel/signal.c @@ -12,7 +12,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/mn10300/kernel/sys_mn10300.c b/arch/mn10300/kernel/sys_mn10300.c index bca5a84dc72c..29d196b83d25 100644 --- a/arch/mn10300/kernel/sys_mn10300.c +++ b/arch/mn10300/kernel/sys_mn10300.c @@ -13,7 +13,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/mn10300/kernel/traps.c b/arch/mn10300/kernel/traps.c index 0dfdc5001124..91365adba4f5 100644 --- a/arch/mn10300/kernel/traps.c +++ b/arch/mn10300/kernel/traps.c @@ -17,7 +17,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/mn10300/mm/fault.c b/arch/mn10300/mm/fault.c index a62e1e138bc1..53bb17d0f068 100644 --- a/arch/mn10300/mm/fault.c +++ b/arch/mn10300/mm/fault.c @@ -20,7 +20,6 @@ #include #include #include -#include #include #include #include /* For unblank_screen() */ diff --git a/arch/mn10300/mm/misalignment.c b/arch/mn10300/mm/misalignment.c index 94c4a4358065..30016251f658 100644 --- a/arch/mn10300/mm/misalignment.c +++ b/arch/mn10300/mm/misalignment.c @@ -17,7 +17,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/powerpc/kernel/ptrace32.c b/arch/powerpc/kernel/ptrace32.c index 297632cba047..8a6daf4129f6 100644 --- a/arch/powerpc/kernel/ptrace32.c +++ b/arch/powerpc/kernel/ptrace32.c @@ -21,7 +21,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/s390/kernel/dis.c b/arch/s390/kernel/dis.c index d2f270c995d9..db943a7ec513 100644 --- a/arch/s390/kernel/dis.c +++ b/arch/s390/kernel/dis.c @@ -15,7 +15,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/s390/kernel/ptrace.c b/arch/s390/kernel/ptrace.c index 490b39934d65..43acd73105b7 100644 --- a/arch/s390/kernel/ptrace.c +++ b/arch/s390/kernel/ptrace.c @@ -26,7 +26,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/s390/mm/fault.c b/arch/s390/mm/fault.c index 74eb26bf1970..e5e119fe03b2 100644 --- a/arch/s390/mm/fault.c +++ b/arch/s390/mm/fault.c @@ -22,7 +22,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/sh/mm/tlb-sh3.c b/arch/sh/mm/tlb-sh3.c index 7fbfd5a11ffa..17cb7c3adf22 100644 --- a/arch/sh/mm/tlb-sh3.c +++ b/arch/sh/mm/tlb-sh3.c @@ -18,7 +18,6 @@ #include #include #include -#include #include #include diff --git a/arch/sparc/kernel/ptrace_32.c b/arch/sparc/kernel/ptrace_32.c index 8ce6285a06d5..7e3dfd9bb97e 100644 --- a/arch/sparc/kernel/ptrace_32.c +++ b/arch/sparc/kernel/ptrace_32.c @@ -16,7 +16,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/sparc/kernel/ptrace_64.c b/arch/sparc/kernel/ptrace_64.c index a941c610e7ce..4ae91dc2feb9 100644 --- a/arch/sparc/kernel/ptrace_64.c +++ b/arch/sparc/kernel/ptrace_64.c @@ -17,7 +17,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/sparc/kernel/time_64.c b/arch/sparc/kernel/time_64.c index 5c12e79b4bdf..da1218e8ee87 100644 --- a/arch/sparc/kernel/time_64.c +++ b/arch/sparc/kernel/time_64.c @@ -11,7 +11,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/sparc/kernel/traps_32.c b/arch/sparc/kernel/traps_32.c index 358283341b47..c0490c7bbde0 100644 --- a/arch/sparc/kernel/traps_32.c +++ b/arch/sparc/kernel/traps_32.c @@ -13,7 +13,6 @@ #include #include #include -#include #include #include diff --git a/drivers/block/DAC960.c b/drivers/block/DAC960.c index 668dc234b8e2..1e6b7c14f697 100644 --- a/drivers/block/DAC960.c +++ b/drivers/block/DAC960.c @@ -36,6 +36,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/block/cciss.c b/drivers/block/cciss.c index 65a0655e7fc8..a52cc7fe45ea 100644 --- a/drivers/block/cciss.c +++ b/drivers/block/cciss.c @@ -26,6 +26,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/block/loop.c b/drivers/block/loop.c index 801f4ab83302..5757188cd1fb 100644 --- a/drivers/block/loop.c +++ b/drivers/block/loop.c @@ -61,7 +61,6 @@ #include #include #include -#include #include #include #include diff --git a/drivers/bluetooth/hci_vhci.c b/drivers/bluetooth/hci_vhci.c index 1df9dda2e377..d5cde6d86f89 100644 --- a/drivers/bluetooth/hci_vhci.c +++ b/drivers/bluetooth/hci_vhci.c @@ -28,7 +28,6 @@ #include #include #include -#include #include #include #include diff --git a/drivers/char/amiserial.c b/drivers/char/amiserial.c index 72429b6b2fa8..6c32fbf07164 100644 --- a/drivers/char/amiserial.c +++ b/drivers/char/amiserial.c @@ -81,6 +81,7 @@ static char *serial_version = "4.30"; #include #include #include +#include #include #include diff --git a/drivers/char/cyclades.c b/drivers/char/cyclades.c index f3366d3f06cf..2dafc2da0648 100644 --- a/drivers/char/cyclades.c +++ b/drivers/char/cyclades.c @@ -633,6 +633,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/char/epca.c b/drivers/char/epca.c index abef1f7d84fe..ff647ca1c489 100644 --- a/drivers/char/epca.c +++ b/drivers/char/epca.c @@ -36,6 +36,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/char/isicom.c b/drivers/char/isicom.c index 621d1184673c..4f1f4cd670da 100644 --- a/drivers/char/isicom.c +++ b/drivers/char/isicom.c @@ -122,6 +122,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/char/istallion.c b/drivers/char/istallion.c index 0c999f5bb3db..ab2f3349c5c4 100644 --- a/drivers/char/istallion.c +++ b/drivers/char/istallion.c @@ -20,6 +20,7 @@ #include #include +#include #include #include #include diff --git a/drivers/char/moxa.c b/drivers/char/moxa.c index 65b6ff2442c6..dd0083bbb64a 100644 --- a/drivers/char/moxa.c +++ b/drivers/char/moxa.c @@ -34,6 +34,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/char/mxser.c b/drivers/char/mxser.c index 52d953eb30c3..dbf8d52f31d0 100644 --- a/drivers/char/mxser.c +++ b/drivers/char/mxser.c @@ -23,6 +23,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/char/n_hdlc.c b/drivers/char/n_hdlc.c index 1c43c8cdee25..c68118efad84 100644 --- a/drivers/char/n_hdlc.c +++ b/drivers/char/n_hdlc.c @@ -97,6 +97,7 @@ #include #include #include +#include #include /* used in new tty drivers */ #include /* used in new tty drivers */ #include diff --git a/drivers/char/n_r3964.c b/drivers/char/n_r3964.c index 2e99158ebb8a..6934025a1ac1 100644 --- a/drivers/char/n_r3964.c +++ b/drivers/char/n_r3964.c @@ -58,6 +58,7 @@ #include #include #include +#include #include #include #include /* used in new tty drivers */ diff --git a/drivers/char/pty.c b/drivers/char/pty.c index 9d1b4f548f67..6e6942c45f5b 100644 --- a/drivers/char/pty.c +++ b/drivers/char/pty.c @@ -22,6 +22,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/char/rio/rio_linux.c b/drivers/char/rio/rio_linux.c index ce81da5b2da9..d58c2eb07f07 100644 --- a/drivers/char/rio/rio_linux.c +++ b/drivers/char/rio/rio_linux.c @@ -44,6 +44,7 @@ #include #include #include +#include #include #include diff --git a/drivers/char/riscom8.c b/drivers/char/riscom8.c index 217660451237..171711acf5cd 100644 --- a/drivers/char/riscom8.c +++ b/drivers/char/riscom8.c @@ -47,6 +47,7 @@ #include #include #include +#include #include #include diff --git a/drivers/char/rocket.c b/drivers/char/rocket.c index 63d5b628477a..0e29a23ec4c5 100644 --- a/drivers/char/rocket.c +++ b/drivers/char/rocket.c @@ -73,6 +73,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/char/serial167.c b/drivers/char/serial167.c index f1f24f0ee26f..51e7a46787be 100644 --- a/drivers/char/serial167.c +++ b/drivers/char/serial167.c @@ -52,6 +52,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/char/specialix.c b/drivers/char/specialix.c index e72be4190a44..bfe4cdb2febb 100644 --- a/drivers/char/specialix.c +++ b/drivers/char/specialix.c @@ -87,6 +87,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/char/sx.c b/drivers/char/sx.c index 518f2a25d91e..a81ec4fcf6ff 100644 --- a/drivers/char/sx.c +++ b/drivers/char/sx.c @@ -216,6 +216,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/char/synclink.c b/drivers/char/synclink.c index afded3a2379c..813552f14884 100644 --- a/drivers/char/synclink.c +++ b/drivers/char/synclink.c @@ -81,6 +81,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/char/synclink_gt.c b/drivers/char/synclink_gt.c index a2e67e6df3a1..91f20a92fddf 100644 --- a/drivers/char/synclink_gt.c +++ b/drivers/char/synclink_gt.c @@ -62,6 +62,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/char/synclinkmp.c b/drivers/char/synclinkmp.c index 6f727e3c53ad..8d4a2a8a0a70 100644 --- a/drivers/char/synclinkmp.c +++ b/drivers/char/synclinkmp.c @@ -52,6 +52,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/char/tpm/tpm.c b/drivers/char/tpm/tpm.c index ccdd828adcef..b0603b2e5684 100644 --- a/drivers/char/tpm/tpm.c +++ b/drivers/char/tpm/tpm.c @@ -26,7 +26,6 @@ #include #include #include -#include #include "tpm.h" diff --git a/drivers/char/tty_ioctl.c b/drivers/char/tty_ioctl.c index b24f6c6a1ea3..ad6ba4ed2808 100644 --- a/drivers/char/tty_ioctl.c +++ b/drivers/char/tty_ioctl.c @@ -21,7 +21,6 @@ #include #include #include -#include #include #include diff --git a/drivers/char/tty_ldisc.c b/drivers/char/tty_ldisc.c index 913aa8d3f1c5..0ef0dc97ba20 100644 --- a/drivers/char/tty_ldisc.c +++ b/drivers/char/tty_ldisc.c @@ -21,7 +21,6 @@ #include #include #include -#include #include #include #include diff --git a/drivers/char/vt.c b/drivers/char/vt.c index d9113b4c76e3..7947bd1b4cf7 100644 --- a/drivers/char/vt.c +++ b/drivers/char/vt.c @@ -89,6 +89,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/char/vt_ioctl.c b/drivers/char/vt_ioctl.c index 7539bed0f7e0..95189f288f8c 100644 --- a/drivers/char/vt_ioctl.c +++ b/drivers/char/vt_ioctl.c @@ -25,6 +25,7 @@ #include #include #include +#include #include #include diff --git a/drivers/gpio/vr41xx_giu.c b/drivers/gpio/vr41xx_giu.c index b70e06133e78..b16c9a8c03f5 100644 --- a/drivers/gpio/vr41xx_giu.c +++ b/drivers/gpio/vr41xx_giu.c @@ -29,7 +29,6 @@ #include #include #include -#include #include #include diff --git a/drivers/hid/usbhid/hid-core.c b/drivers/hid/usbhid/hid-core.c index 76c4bbe9dccb..3c1fcb7640ab 100644 --- a/drivers/hid/usbhid/hid-core.c +++ b/drivers/hid/usbhid/hid-core.c @@ -22,7 +22,6 @@ #include #include #include -#include #include #include #include diff --git a/drivers/isdn/hisax/hfc_usb.c b/drivers/isdn/hisax/hfc_usb.c index 8df889b0c1a9..9de54202c90c 100644 --- a/drivers/isdn/hisax/hfc_usb.c +++ b/drivers/isdn/hisax/hfc_usb.c @@ -37,7 +37,6 @@ #include #include #include -#include #include #include #include "hisax.h" diff --git a/drivers/isdn/i4l/isdn_tty.c b/drivers/isdn/i4l/isdn_tty.c index b4d4522e5071..2881a66c1aa9 100644 --- a/drivers/isdn/i4l/isdn_tty.c +++ b/drivers/isdn/i4l/isdn_tty.c @@ -13,6 +13,7 @@ #include #include +#include #include "isdn_common.h" #include "isdn_tty.h" #ifdef CONFIG_ISDN_AUDIO diff --git a/drivers/isdn/mISDN/stack.c b/drivers/isdn/mISDN/stack.c index e2f45019ebf0..3e1532a180ff 100644 --- a/drivers/isdn/mISDN/stack.c +++ b/drivers/isdn/mISDN/stack.c @@ -17,6 +17,7 @@ #include #include +#include #include "core.h" static u_int *debug; diff --git a/drivers/media/dvb/bt8xx/dst_ca.c b/drivers/media/dvb/bt8xx/dst_ca.c index 4601b059b2b2..0e246eaad05a 100644 --- a/drivers/media/dvb/bt8xx/dst_ca.c +++ b/drivers/media/dvb/bt8xx/dst_ca.c @@ -21,6 +21,7 @@ #include #include #include +#include #include #include #include "dvbdev.h" diff --git a/drivers/media/dvb/dvb-core/dvbdev.h b/drivers/media/dvb/dvb-core/dvbdev.h index 79927305e84d..487919bea7ae 100644 --- a/drivers/media/dvb/dvb-core/dvbdev.h +++ b/drivers/media/dvb/dvb-core/dvbdev.h @@ -27,7 +27,6 @@ #include #include #include -#include #define DVB_MAJOR 212 diff --git a/drivers/media/dvb/ttpci/av7110.c b/drivers/media/dvb/ttpci/av7110.c index d1d959ed37b7..8d65c652ba50 100644 --- a/drivers/media/dvb/ttpci/av7110.c +++ b/drivers/media/dvb/ttpci/av7110.c @@ -36,7 +36,6 @@ #include #include #include -#include #include #include diff --git a/drivers/media/radio/radio-mr800.c b/drivers/media/radio/radio-mr800.c index 837467f93805..575bf9d89419 100644 --- a/drivers/media/radio/radio-mr800.c +++ b/drivers/media/radio/radio-mr800.c @@ -58,6 +58,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/media/radio/radio-si470x.c b/drivers/media/radio/radio-si470x.c index 46d216329611..e85f318b4d2b 100644 --- a/drivers/media/radio/radio-si470x.c +++ b/drivers/media/radio/radio-si470x.c @@ -127,6 +127,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/media/video/bt8xx/bttv-driver.c b/drivers/media/video/bt8xx/bttv-driver.c index 5eb1464af670..d147d29bb0d3 100644 --- a/drivers/media/video/bt8xx/bttv-driver.c +++ b/drivers/media/video/bt8xx/bttv-driver.c @@ -41,6 +41,7 @@ #include #include #include +#include #include #include #include "bttvp.h" diff --git a/drivers/media/video/cx23885/cx23885-417.c b/drivers/media/video/cx23885/cx23885-417.c index 2943bfd32a94..428f0c45e6b7 100644 --- a/drivers/media/video/cx23885/cx23885-417.c +++ b/drivers/media/video/cx23885/cx23885-417.c @@ -31,6 +31,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/media/video/cx23885/cx23885-video.c b/drivers/media/video/cx23885/cx23885-video.c index 70836af3ab48..5d6093336300 100644 --- a/drivers/media/video/cx23885/cx23885-video.c +++ b/drivers/media/video/cx23885/cx23885-video.c @@ -26,6 +26,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/media/video/cx88/cx88-blackbird.c b/drivers/media/video/cx88/cx88-blackbird.c index 44eacfb0d0d6..356d6896da3f 100644 --- a/drivers/media/video/cx88/cx88-blackbird.c +++ b/drivers/media/video/cx88/cx88-blackbird.c @@ -32,6 +32,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/media/video/cx88/cx88-video.c b/drivers/media/video/cx88/cx88-video.c index b12770848c00..2bb54c3ef5cd 100644 --- a/drivers/media/video/cx88/cx88-video.c +++ b/drivers/media/video/cx88/cx88-video.c @@ -31,6 +31,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/media/video/dabusb.c b/drivers/media/video/dabusb.c index ec2f45dde164..0664d111085f 100644 --- a/drivers/media/video/dabusb.c +++ b/drivers/media/video/dabusb.c @@ -32,6 +32,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/media/video/pwc/pwc-if.c b/drivers/media/video/pwc/pwc-if.c index db25c3034c11..8d17cf613306 100644 --- a/drivers/media/video/pwc/pwc-if.c +++ b/drivers/media/video/pwc/pwc-if.c @@ -62,6 +62,7 @@ #include #include #include +#include #ifdef CONFIG_USB_PWC_INPUT_EVDEV #include #endif diff --git a/drivers/media/video/pwc/pwc.h b/drivers/media/video/pwc/pwc.h index 0be6f814f539..0b658dee05a4 100644 --- a/drivers/media/video/pwc/pwc.h +++ b/drivers/media/video/pwc/pwc.h @@ -29,7 +29,6 @@ #include #include #include -#include #include #include #include diff --git a/drivers/media/video/s2255drv.c b/drivers/media/video/s2255drv.c index 6be845ccc7d7..9e3262c0ba37 100644 --- a/drivers/media/video/s2255drv.c +++ b/drivers/media/video/s2255drv.c @@ -48,6 +48,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/media/video/saa5246a.c b/drivers/media/video/saa5246a.c index 155804b061e9..b624a4c01fdc 100644 --- a/drivers/media/video/saa5246a.c +++ b/drivers/media/video/saa5246a.c @@ -43,7 +43,6 @@ #include #include #include -#include #include #include #include diff --git a/drivers/media/video/saa5249.c b/drivers/media/video/saa5249.c index 271d6e931b75..12835fb82c95 100644 --- a/drivers/media/video/saa5249.c +++ b/drivers/media/video/saa5249.c @@ -46,7 +46,6 @@ #include #include #include -#include #include #include #include diff --git a/drivers/media/video/saa7134/saa7134-empress.c b/drivers/media/video/saa7134/saa7134-empress.c index add1757f8930..296788c3bf0e 100644 --- a/drivers/media/video/saa7134/saa7134-empress.c +++ b/drivers/media/video/saa7134/saa7134-empress.c @@ -22,6 +22,7 @@ #include #include #include +#include #include #include "saa7134-reg.h" diff --git a/drivers/media/video/se401.c b/drivers/media/video/se401.c index c8f05297d0f0..85ffc2cba039 100644 --- a/drivers/media/video/se401.c +++ b/drivers/media/video/se401.c @@ -31,6 +31,7 @@ static const char version[] = "0.24"; #include #include #include +#include #include #include #include "se401.h" diff --git a/drivers/media/video/stk-webcam.c b/drivers/media/video/stk-webcam.c index 2e5937047278..4d6785e63455 100644 --- a/drivers/media/video/stk-webcam.c +++ b/drivers/media/video/stk-webcam.c @@ -27,6 +27,7 @@ #include #include #include +#include #include #include diff --git a/drivers/media/video/stradis.c b/drivers/media/video/stradis.c index 0eb313082c97..eaada39c76fd 100644 --- a/drivers/media/video/stradis.c +++ b/drivers/media/video/stradis.c @@ -26,6 +26,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/media/video/stv680.c b/drivers/media/video/stv680.c index 75f286f7a2e9..8b4e7dafce7b 100644 --- a/drivers/media/video/stv680.c +++ b/drivers/media/video/stv680.c @@ -62,6 +62,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/media/video/usbvideo/vicam.c b/drivers/media/video/usbvideo/vicam.c index 8d73979596f9..45fce39ec9ad 100644 --- a/drivers/media/video/usbvideo/vicam.c +++ b/drivers/media/video/usbvideo/vicam.c @@ -43,6 +43,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/media/video/usbvision/usbvision-video.c b/drivers/media/video/usbvision/usbvision-video.c index 90b58914f984..90d9b5c0e9a7 100644 --- a/drivers/media/video/usbvision/usbvision-video.c +++ b/drivers/media/video/usbvision/usbvision-video.c @@ -50,6 +50,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/media/video/v4l2-dev.c b/drivers/media/video/v4l2-dev.c index 31eac66411d7..a7f1b69a7dab 100644 --- a/drivers/media/video/v4l2-dev.c +++ b/drivers/media/video/v4l2-dev.c @@ -25,7 +25,6 @@ #include #include #include -#include #include #include diff --git a/drivers/media/video/zoran/zoran_driver.c b/drivers/media/video/zoran/zoran_driver.c index 3d7df32a3d87..bcdefb1bcb3d 100644 --- a/drivers/media/video/zoran/zoran_driver.c +++ b/drivers/media/video/zoran/zoran_driver.c @@ -49,6 +49,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/misc/sgi-gru/grufile.c b/drivers/misc/sgi-gru/grufile.c index fa2d93a9fb8d..aed609832bc2 100644 --- a/drivers/misc/sgi-gru/grufile.c +++ b/drivers/misc/sgi-gru/grufile.c @@ -29,7 +29,6 @@ #include #include #include -#include #include #include #include diff --git a/drivers/misc/sgi-gru/grukservices.c b/drivers/misc/sgi-gru/grukservices.c index eedbf9c32760..79689b10f937 100644 --- a/drivers/misc/sgi-gru/grukservices.c +++ b/drivers/misc/sgi-gru/grukservices.c @@ -24,7 +24,6 @@ #include #include #include -#include #include #include #include diff --git a/drivers/net/irda/irtty-sir.c b/drivers/net/irda/irtty-sir.c index d53aa9582137..20f9bc626688 100644 --- a/drivers/net/irda/irtty-sir.c +++ b/drivers/net/irda/irtty-sir.c @@ -31,7 +31,6 @@ #include #include #include -#include #include #include diff --git a/drivers/pci/hotplug/cpci_hotplug_core.c b/drivers/pci/hotplug/cpci_hotplug_core.c index a5b9f6ae507b..d703e73fffa7 100644 --- a/drivers/pci/hotplug/cpci_hotplug_core.c +++ b/drivers/pci/hotplug/cpci_hotplug_core.c @@ -32,7 +32,6 @@ #include #include #include -#include #include #include #include diff --git a/drivers/pci/hotplug/cpqphp_ctrl.c b/drivers/pci/hotplug/cpqphp_ctrl.c index 2fa47af992a8..0ff689afa757 100644 --- a/drivers/pci/hotplug/cpqphp_ctrl.c +++ b/drivers/pci/hotplug/cpqphp_ctrl.c @@ -34,7 +34,6 @@ #include #include #include -#include #include #include #include diff --git a/drivers/pci/hotplug/cpqphp_sysfs.c b/drivers/pci/hotplug/cpqphp_sysfs.c index 8450f4a6568a..e6089bdb6e5b 100644 --- a/drivers/pci/hotplug/cpqphp_sysfs.c +++ b/drivers/pci/hotplug/cpqphp_sysfs.c @@ -33,6 +33,7 @@ #include #include #include +#include #include #include "cpqphp.h" diff --git a/drivers/pci/hotplug/pciehp_ctrl.c b/drivers/pci/hotplug/pciehp_ctrl.c index ff4034502d24..8aab8edf123e 100644 --- a/drivers/pci/hotplug/pciehp_ctrl.c +++ b/drivers/pci/hotplug/pciehp_ctrl.c @@ -30,7 +30,6 @@ #include #include #include -#include #include #include #include "../pci.h" diff --git a/drivers/pci/syscall.c b/drivers/pci/syscall.c index ec22284eed30..e1c1ec540893 100644 --- a/drivers/pci/syscall.c +++ b/drivers/pci/syscall.c @@ -9,7 +9,6 @@ #include #include -#include #include #include #include "pci.h" diff --git a/drivers/s390/block/dasd_ioctl.c b/drivers/s390/block/dasd_ioctl.c index 4ce3f72ee1c1..df918ef27965 100644 --- a/drivers/s390/block/dasd_ioctl.c +++ b/drivers/s390/block/dasd_ioctl.c @@ -16,6 +16,7 @@ #include #include #include +#include #include #include diff --git a/drivers/scsi/qla2xxx/qla_mid.c b/drivers/scsi/qla2xxx/qla_mid.c index 650bcef08f2a..cd78c501803a 100644 --- a/drivers/scsi/qla2xxx/qla_mid.c +++ b/drivers/scsi/qla2xxx/qla_mid.c @@ -9,7 +9,6 @@ #include #include -#include #include #include diff --git a/drivers/telephony/ixj.c b/drivers/telephony/ixj.c index a913efc69669..40de151f2789 100644 --- a/drivers/telephony/ixj.c +++ b/drivers/telephony/ixj.c @@ -257,6 +257,7 @@ #include /* everything... */ #include /* error codes */ #include +#include #include #include #include diff --git a/drivers/telephony/phonedev.c b/drivers/telephony/phonedev.c index b52cc830c0b4..f3873f650bb4 100644 --- a/drivers/telephony/phonedev.c +++ b/drivers/telephony/phonedev.c @@ -23,7 +23,6 @@ #include #include #include -#include #include #include diff --git a/drivers/usb/class/cdc-wdm.c b/drivers/usb/class/cdc-wdm.c index 0fe434505ac4..ba589d4ca8bc 100644 --- a/drivers/usb/class/cdc-wdm.c +++ b/drivers/usb/class/cdc-wdm.c @@ -15,7 +15,6 @@ #include #include #include -#include #include #include #include diff --git a/drivers/usb/gadget/amd5536udc.c b/drivers/usb/gadget/amd5536udc.c index 826f3adde5d8..77352ccc245e 100644 --- a/drivers/usb/gadget/amd5536udc.c +++ b/drivers/usb/gadget/amd5536udc.c @@ -48,7 +48,6 @@ #include #include #include -#include #include #include #include diff --git a/drivers/usb/gadget/langwell_udc.c b/drivers/usb/gadget/langwell_udc.c index 6829d5961359..a3913519fd58 100644 --- a/drivers/usb/gadget/langwell_udc.c +++ b/drivers/usb/gadget/langwell_udc.c @@ -34,7 +34,6 @@ #include #include #include -#include #include #include #include diff --git a/drivers/usb/gadget/s3c2410_udc.c b/drivers/usb/gadget/s3c2410_udc.c index 9a2b8920532d..a9b452fe6221 100644 --- a/drivers/usb/gadget/s3c2410_udc.c +++ b/drivers/usb/gadget/s3c2410_udc.c @@ -28,7 +28,6 @@ #include #include #include -#include #include #include #include diff --git a/drivers/usb/host/r8a66597-hcd.c b/drivers/usb/host/r8a66597-hcd.c index 56976cc0352a..e18f74946e68 100644 --- a/drivers/usb/host/r8a66597-hcd.c +++ b/drivers/usb/host/r8a66597-hcd.c @@ -26,7 +26,6 @@ #include #include #include -#include #include #include #include diff --git a/drivers/usb/misc/iowarrior.c b/drivers/usb/misc/iowarrior.c index 3c5fe5cee05a..90e1a8dedfa9 100644 --- a/drivers/usb/misc/iowarrior.c +++ b/drivers/usb/misc/iowarrior.c @@ -18,6 +18,7 @@ #include #include #include +#include #include #include diff --git a/drivers/usb/misc/rio500.c b/drivers/usb/misc/rio500.c index deb95bb49fd1..d645f3899fe1 100644 --- a/drivers/usb/misc/rio500.c +++ b/drivers/usb/misc/rio500.c @@ -32,6 +32,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/usb/misc/usblcd.c b/drivers/usb/misc/usblcd.c index e0ff9ccd866b..29092b8e59ce 100644 --- a/drivers/usb/misc/usblcd.c +++ b/drivers/usb/misc/usblcd.c @@ -16,6 +16,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/usb/musb/cppi_dma.h b/drivers/usb/musb/cppi_dma.h index 8a39de3e6e47..59bf949e589b 100644 --- a/drivers/usb/musb/cppi_dma.h +++ b/drivers/usb/musb/cppi_dma.h @@ -5,7 +5,6 @@ #include #include -#include #include #include diff --git a/drivers/usb/musb/musb_core.h b/drivers/usb/musb/musb_core.h index f3772ca3b2cf..381d648a36b8 100644 --- a/drivers/usb/musb/musb_core.h +++ b/drivers/usb/musb/musb_core.h @@ -38,7 +38,6 @@ #include #include #include -#include #include #include #include diff --git a/drivers/usb/serial/ftdi_sio.c b/drivers/usb/serial/ftdi_sio.c index 5f08702f672f..5a8ae274d258 100644 --- a/drivers/usb/serial/ftdi_sio.c +++ b/drivers/usb/serial/ftdi_sio.c @@ -33,6 +33,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/usb/serial/mos7840.c b/drivers/usb/serial/mos7840.c index c40f95c1951c..c31940a307f8 100644 --- a/drivers/usb/serial/mos7840.c +++ b/drivers/usb/serial/mos7840.c @@ -26,6 +26,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/usb/serial/usb-serial.c b/drivers/usb/serial/usb-serial.c index a84216464ca0..0c39b55aeef4 100644 --- a/drivers/usb/serial/usb-serial.c +++ b/drivers/usb/serial/usb-serial.c @@ -21,6 +21,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/video/fbmem.c b/drivers/video/fbmem.c index 53ea05645ff8..a85c818be945 100644 --- a/drivers/video/fbmem.c +++ b/drivers/video/fbmem.c @@ -16,7 +16,6 @@ #include #include #include -#include #include #include #include diff --git a/fs/adfs/super.c b/fs/adfs/super.c index aad92f0a1048..6910a98bd73c 100644 --- a/fs/adfs/super.c +++ b/fs/adfs/super.c @@ -13,6 +13,7 @@ #include #include #include +#include #include #include "adfs.h" #include "dir_f.h" diff --git a/fs/afs/super.c b/fs/afs/super.c index ad0514d0115f..e1ea1c240b6a 100644 --- a/fs/afs/super.c +++ b/fs/afs/super.c @@ -18,6 +18,7 @@ #include #include #include +#include #include #include #include diff --git a/fs/autofs4/dev-ioctl.c b/fs/autofs4/dev-ioctl.c index f3da2eb51f56..00bf8fcb245f 100644 --- a/fs/autofs4/dev-ioctl.c +++ b/fs/autofs4/dev-ioctl.c @@ -19,7 +19,6 @@ #include #include #include -#include #include #include #include diff --git a/fs/bfs/dir.c b/fs/bfs/dir.c index 54bd07d44e68..1e41aadb1068 100644 --- a/fs/bfs/dir.c +++ b/fs/bfs/dir.c @@ -8,7 +8,6 @@ #include #include #include -#include #include #include #include "bfs.h" diff --git a/fs/bfs/file.c b/fs/bfs/file.c index 6a021265f018..88b9a3ff44e4 100644 --- a/fs/bfs/file.c +++ b/fs/bfs/file.c @@ -11,7 +11,6 @@ #include #include -#include #include "bfs.h" #undef DEBUG diff --git a/fs/btrfs/compression.c b/fs/btrfs/compression.c index de1e2fd32080..9d8ba4d54a37 100644 --- a/fs/btrfs/compression.c +++ b/fs/btrfs/compression.c @@ -26,7 +26,6 @@ #include #include #include -#include #include #include #include diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index 7c3cd248d8d6..4b833972273a 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -22,7 +22,6 @@ #include #include #include -#include #include #include #include diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 7ffa3d34ea19..791eab19e330 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -26,7 +26,6 @@ #include #include #include -#include #include #include #include diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index 9f4db848db10..bd88f25889f7 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -27,7 +27,6 @@ #include #include #include -#include #include #include #include diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index 9f179d4832d5..6d6d06cb6dfc 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -26,7 +26,6 @@ #include #include #include -#include #include #include #include diff --git a/fs/char_dev.c b/fs/char_dev.c index b7c9d5187a75..a173551e19d7 100644 --- a/fs/char_dev.c +++ b/fs/char_dev.c @@ -13,7 +13,6 @@ #include #include #include -#include #include #include diff --git a/fs/compat.c b/fs/compat.c index fbadb947727b..94502dab972a 100644 --- a/fs/compat.c +++ b/fs/compat.c @@ -32,7 +32,6 @@ #include #include #include -#include #include #include #include diff --git a/fs/compat_ioctl.c b/fs/compat_ioctl.c index 626c7483b4de..f28f070a60fc 100644 --- a/fs/compat_ioctl.c +++ b/fs/compat_ioctl.c @@ -19,6 +19,7 @@ #include #include #include +#include #include #include #include diff --git a/fs/exofs/super.c b/fs/exofs/super.c index a343b4ea62f6..5ab10c3bbebe 100644 --- a/fs/exofs/super.c +++ b/fs/exofs/super.c @@ -31,6 +31,7 @@ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA */ +#include #include #include #include diff --git a/fs/ext2/ioctl.c b/fs/ext2/ioctl.c index 7cb4badef927..e7431309bdca 100644 --- a/fs/ext2/ioctl.c +++ b/fs/ext2/ioctl.c @@ -13,7 +13,6 @@ #include #include #include -#include #include #include diff --git a/fs/ext4/ioctl.c b/fs/ext4/ioctl.c index bb415408fdb6..24a6abb2aef5 100644 --- a/fs/ext4/ioctl.c +++ b/fs/ext4/ioctl.c @@ -12,7 +12,6 @@ #include #include #include -#include #include #include #include diff --git a/fs/fat/dir.c b/fs/fat/dir.c index 38ff75a0fe22..530b4ca01510 100644 --- a/fs/fat/dir.c +++ b/fs/fat/dir.c @@ -16,7 +16,6 @@ #include #include #include -#include #include #include #include diff --git a/fs/fat/namei_msdos.c b/fs/fat/namei_msdos.c index 82f88733b681..bbc94ae4fd77 100644 --- a/fs/fat/namei_msdos.c +++ b/fs/fat/namei_msdos.c @@ -9,7 +9,6 @@ #include #include #include -#include #include "fat.h" /* Characters that are undesirable in an MS-DOS file name */ diff --git a/fs/fat/namei_vfat.c b/fs/fat/namei_vfat.c index 73471b7ecc8c..cb6e83557112 100644 --- a/fs/fat/namei_vfat.c +++ b/fs/fat/namei_vfat.c @@ -19,7 +19,6 @@ #include #include #include -#include #include #include #include "fat.h" diff --git a/fs/fcntl.c b/fs/fcntl.c index a040b764f8e3..ae413086db97 100644 --- a/fs/fcntl.c +++ b/fs/fcntl.c @@ -19,7 +19,6 @@ #include #include #include -#include #include #include diff --git a/fs/freevxfs/vxfs_super.c b/fs/freevxfs/vxfs_super.c index cdbd1654e4cd..1e8af939b3e4 100644 --- a/fs/freevxfs/vxfs_super.c +++ b/fs/freevxfs/vxfs_super.c @@ -38,6 +38,7 @@ #include #include #include +#include #include #include #include diff --git a/fs/hfs/super.c b/fs/hfs/super.c index 6f833dc8e910..f7fcbe49da72 100644 --- a/fs/hfs/super.c +++ b/fs/hfs/super.c @@ -19,6 +19,7 @@ #include #include #include +#include #include #include "hfs_fs.h" diff --git a/fs/hfsplus/super.c b/fs/hfsplus/super.c index 9fc3af0c0dab..c0759fe0855b 100644 --- a/fs/hfsplus/super.c +++ b/fs/hfsplus/super.c @@ -12,6 +12,7 @@ #include #include #include +#include #include #include diff --git a/fs/hpfs/dir.c b/fs/hpfs/dir.c index 6916c41d7017..8865c94f55f6 100644 --- a/fs/hpfs/dir.c +++ b/fs/hpfs/dir.c @@ -6,6 +6,7 @@ * directory VFS functions */ +#include #include "hpfs_fn.h" static int hpfs_dir_release(struct inode *inode, struct file *filp) diff --git a/fs/hpfs/file.c b/fs/hpfs/file.c index 64ab52259204..3efabff00367 100644 --- a/fs/hpfs/file.c +++ b/fs/hpfs/file.c @@ -6,6 +6,7 @@ * file VFS functions */ +#include #include "hpfs_fn.h" #define BLOCKS(size) (((size) + 511) >> 9) diff --git a/fs/hpfs/hpfs_fn.h b/fs/hpfs/hpfs_fn.h index c2ea31bae313..701ca54c0867 100644 --- a/fs/hpfs/hpfs_fn.h +++ b/fs/hpfs/hpfs_fn.h @@ -13,7 +13,6 @@ #include #include #include -#include #include "hpfs.h" diff --git a/fs/hpfs/inode.c b/fs/hpfs/inode.c index 39a1bfbea312..fe703ae46bc7 100644 --- a/fs/hpfs/inode.c +++ b/fs/hpfs/inode.c @@ -6,6 +6,7 @@ * inode VFS functions */ +#include #include "hpfs_fn.h" void hpfs_init_inode(struct inode *i) diff --git a/fs/hpfs/namei.c b/fs/hpfs/namei.c index b649232dde97..82b9c4ba9ed0 100644 --- a/fs/hpfs/namei.c +++ b/fs/hpfs/namei.c @@ -6,6 +6,7 @@ * adding & removing files & directories */ #include +#include #include "hpfs_fn.h" static int hpfs_mkdir(struct inode *dir, struct dentry *dentry, int mode) diff --git a/fs/jffs2/super.c b/fs/jffs2/super.c index 07a22caf2687..0035c021395a 100644 --- a/fs/jffs2/super.c +++ b/fs/jffs2/super.c @@ -12,6 +12,7 @@ #include #include #include +#include #include #include #include diff --git a/fs/lockd/clntproc.c b/fs/lockd/clntproc.c index f2fdcbce143e..4336adba952a 100644 --- a/fs/lockd/clntproc.c +++ b/fs/lockd/clntproc.c @@ -7,6 +7,7 @@ */ #include +#include #include #include #include diff --git a/fs/lockd/svc4proc.c b/fs/lockd/svc4proc.c index 1725037374c5..bd173a6ca3b1 100644 --- a/fs/lockd/svc4proc.c +++ b/fs/lockd/svc4proc.c @@ -10,6 +10,7 @@ #include #include #include +#include #include #include #include diff --git a/fs/lockd/svcproc.c b/fs/lockd/svcproc.c index 3688e55901fc..e1d28ddd2169 100644 --- a/fs/lockd/svcproc.c +++ b/fs/lockd/svcproc.c @@ -10,6 +10,7 @@ #include #include #include +#include #include #include #include diff --git a/fs/nfs/delegation.c b/fs/nfs/delegation.c index af05b918cb5b..6dd48a4405b4 100644 --- a/fs/nfs/delegation.c +++ b/fs/nfs/delegation.c @@ -10,6 +10,7 @@ #include #include #include +#include #include #include diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c index 89f98e9a024b..38d42c29fb92 100644 --- a/fs/nfs/dir.c +++ b/fs/nfs/dir.c @@ -29,7 +29,6 @@ #include #include #include -#include #include #include #include diff --git a/fs/nfs/file.c b/fs/nfs/file.c index 0055b813ec2c..05062329b678 100644 --- a/fs/nfs/file.c +++ b/fs/nfs/file.c @@ -26,7 +26,6 @@ #include #include #include -#include #include #include diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c index 64f87194d390..bd7938eda6a8 100644 --- a/fs/nfs/inode.c +++ b/fs/nfs/inode.c @@ -30,7 +30,6 @@ #include #include #include -#include #include #include #include diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 92ce43517814..ff0c080db59b 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -45,7 +45,6 @@ #include #include #include -#include #include #include #include diff --git a/fs/nfs/read.c b/fs/nfs/read.c index 96c4ebfa46f4..73ea5e8d66ce 100644 --- a/fs/nfs/read.c +++ b/fs/nfs/read.c @@ -18,7 +18,6 @@ #include #include #include -#include #include diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index 1250fb978ac1..6d0847562d87 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -25,7 +25,6 @@ #include #include #include -#include #include #include diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c index d4c9884cd54b..492c79b7800b 100644 --- a/fs/nfsd/nfssvc.c +++ b/fs/nfsd/nfssvc.c @@ -18,7 +18,6 @@ #include #include #include -#include #include #include #include diff --git a/fs/nilfs2/dir.c b/fs/nilfs2/dir.c index 54100acc1102..1a4fa04cf071 100644 --- a/fs/nilfs2/dir.c +++ b/fs/nilfs2/dir.c @@ -43,7 +43,6 @@ */ #include -#include #include "nilfs.h" #include "page.h" diff --git a/fs/ocfs2/ioctl.c b/fs/ocfs2/ioctl.c index 9fcd36dcc9a0..467b413bec21 100644 --- a/fs/ocfs2/ioctl.c +++ b/fs/ocfs2/ioctl.c @@ -7,7 +7,6 @@ #include #include -#include #define MLOG_MASK_PREFIX ML_INODE #include diff --git a/fs/reiserfs/xattr.c b/fs/reiserfs/xattr.c index f3d47d856848..6925b835a43b 100644 --- a/fs/reiserfs/xattr.c +++ b/fs/reiserfs/xattr.c @@ -46,7 +46,6 @@ #include #include #include -#include #include #include diff --git a/fs/squashfs/super.c b/fs/squashfs/super.c index 3b52770f46ff..cb5fc57e370b 100644 --- a/fs/squashfs/super.c +++ b/fs/squashfs/super.c @@ -30,6 +30,7 @@ #include #include #include +#include #include #include #include diff --git a/fs/ubifs/ioctl.c b/fs/ubifs/ioctl.c index 6db7a6be6c97..8aacd64957a2 100644 --- a/fs/ubifs/ioctl.c +++ b/fs/ubifs/ioctl.c @@ -25,7 +25,6 @@ /* This file implements EXT2-compatible extended attribute ioctl() calls */ #include -#include #include #include "ubifs.h" diff --git a/fs/xfs/linux-2.6/xfs_file.c b/fs/xfs/linux-2.6/xfs_file.c index f4e255441574..0542fd507649 100644 --- a/fs/xfs/linux-2.6/xfs_file.c +++ b/fs/xfs/linux-2.6/xfs_file.c @@ -41,7 +41,6 @@ #include "xfs_ioctl.h" #include -#include static struct vm_operations_struct xfs_file_vm_ops; diff --git a/include/linux/crash_dump.h b/include/linux/crash_dump.h index 2dac064d8359..0026f267da20 100644 --- a/include/linux/crash_dump.h +++ b/include/linux/crash_dump.h @@ -3,7 +3,6 @@ #ifdef CONFIG_CRASH_DUMP #include -#include #include #include diff --git a/include/linux/hardirq.h b/include/linux/hardirq.h index 45257475623c..8246c697863d 100644 --- a/include/linux/hardirq.h +++ b/include/linux/hardirq.h @@ -2,7 +2,9 @@ #define LINUX_HARDIRQ_H #include +#ifdef CONFIG_PREEMPT #include +#endif #include #include #include diff --git a/include/linux/quotaops.h b/include/linux/quotaops.h index 7bc457593684..26361c4c037a 100644 --- a/include/linux/quotaops.h +++ b/include/linux/quotaops.h @@ -7,7 +7,6 @@ #ifndef _LINUX_QUOTAOPS_ #define _LINUX_QUOTAOPS_ -#include #include static inline struct quota_info *sb_dqopt(struct super_block *sb) diff --git a/include/linux/sunrpc/xdr.h b/include/linux/sunrpc/xdr.h index d8910b68e1bd..b99c625fddfe 100644 --- a/include/linux/sunrpc/xdr.h +++ b/include/linux/sunrpc/xdr.h @@ -12,7 +12,6 @@ #include #include #include -#include /* * Buffer adjustment diff --git a/kernel/power/user.c b/kernel/power/user.c index ed97375daae9..bf0014d6a5f0 100644 --- a/kernel/power/user.c +++ b/kernel/power/user.c @@ -23,7 +23,6 @@ #include #include #include -#include #include #include diff --git a/kernel/trace/blktrace.c b/kernel/trace/blktrace.c index 39af8af6fc30..1090b0aed9ba 100644 --- a/kernel/trace/blktrace.c +++ b/kernel/trace/blktrace.c @@ -22,6 +22,7 @@ #include #include #include +#include #include #include diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 3aa0a0dfdfa8..8bc8d8afea6a 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -17,6 +17,7 @@ #include #include #include +#include #include #include #include diff --git a/net/appletalk/ddp.c b/net/appletalk/ddp.c index 590b83963622..bfbe13786bb4 100644 --- a/net/appletalk/ddp.c +++ b/net/appletalk/ddp.c @@ -54,6 +54,7 @@ #include #include #include +#include #include /* For TIOCOUTQ/INQ */ #include #include diff --git a/net/ipx/af_ipx.c b/net/ipx/af_ipx.c index 417b0e309495..f1118d92a191 100644 --- a/net/ipx/af_ipx.c +++ b/net/ipx/af_ipx.c @@ -41,6 +41,7 @@ #include #include #include +#include #include #include #include diff --git a/net/irda/af_irda.c b/net/irda/af_irda.c index cb762c8723ea..80cf29aae096 100644 --- a/net/irda/af_irda.c +++ b/net/irda/af_irda.c @@ -45,6 +45,7 @@ #include #include #include +#include #include #include #include diff --git a/net/irda/irnet/irnet.h b/net/irda/irnet/irnet.h index bccf4d0059f0..b001c361ad30 100644 --- a/net/irda/irnet/irnet.h +++ b/net/irda/irnet/irnet.h @@ -241,7 +241,6 @@ #include #include -#include #include #include #include diff --git a/net/irda/irnet/irnet_ppp.c b/net/irda/irnet/irnet_ppp.c index 6d8ae03c14f5..68cbcb19cbd8 100644 --- a/net/irda/irnet/irnet_ppp.c +++ b/net/irda/irnet/irnet_ppp.c @@ -13,6 +13,7 @@ * 2) as a control channel (write commands, read events) */ +#include #include "irnet_ppp.h" /* Private header */ /* Please put other headers in irnet.h - Thanks */ diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index 5bc2f45bddf0..ebfcf9b89909 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -28,7 +28,6 @@ #include #include #include -#include #include #include #include diff --git a/net/sunrpc/sched.c b/net/sunrpc/sched.c index 1102ce1251f7..8f459abe97cf 100644 --- a/net/sunrpc/sched.c +++ b/net/sunrpc/sched.c @@ -16,7 +16,6 @@ #include #include #include -#include #include #include diff --git a/net/sunrpc/svc_xprt.c b/net/sunrpc/svc_xprt.c index 6f33d33cc064..27d44332f017 100644 --- a/net/sunrpc/svc_xprt.c +++ b/net/sunrpc/svc_xprt.c @@ -5,6 +5,7 @@ */ #include +#include #include #include #include diff --git a/net/wanrouter/wanmain.c b/net/wanrouter/wanmain.c index 466e2d22d256..258daa80ad92 100644 --- a/net/wanrouter/wanmain.c +++ b/net/wanrouter/wanmain.c @@ -48,6 +48,7 @@ #include #include /* support for loadable modules */ #include /* kmalloc(), kfree() */ +#include #include #include /* inline mem*, str* functions */ diff --git a/net/x25/af_x25.c b/net/x25/af_x25.c index 21cdc872004e..5e6c072c64d3 100644 --- a/net/x25/af_x25.c +++ b/net/x25/af_x25.c @@ -40,6 +40,7 @@ #include #include #include +#include #include #include #include -- cgit v1.2.3 From 4ead0a2b6b3aa0b527871ec978c3ef12aafb6bfb Mon Sep 17 00:00:00 2001 From: Kay Sievers Date: Thu, 2 Jul 2009 23:25:44 +0200 Subject: Driver Core: remove BUS_ID_SIZE The name size limit is gone from the driver-core, this is the removal of the last left-over. Signed-off-by: Kay Sievers Signed-off-by: Greg Kroah-Hartman --- include/linux/device.h | 2 -- 1 file changed, 2 deletions(-) (limited to 'include') diff --git a/include/linux/device.h b/include/linux/device.h index ed4e39f2c423..aebb81036db2 100644 --- a/include/linux/device.h +++ b/include/linux/device.h @@ -25,8 +25,6 @@ #include #include -#define BUS_ID_SIZE 20 - struct device; struct device_private; struct device_driver; -- cgit v1.2.3 From 11a28d373ed2539a110d56419457e2e7db221ac7 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Fri, 10 Jul 2009 09:51:33 +0000 Subject: net: make namespace iteration possible under RCU All we need to take care of is using proper RCU list add/del primitives and inserting a synchronize_rcu() at one place to make sure the exit notifiers are run after everybody has stopped iterating the list. Signed-off-by: Johannes Berg Signed-off-by: David S. Miller --- include/net/net_namespace.h | 3 +++ net/core/net_namespace.c | 14 +++++++++++--- 2 files changed, 14 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/include/net/net_namespace.h b/include/net/net_namespace.h index ded434b032a4..b34a6ee73754 100644 --- a/include/net/net_namespace.h +++ b/include/net/net_namespace.h @@ -208,6 +208,9 @@ static inline struct net *read_pnet(struct net * const *pnet) #define for_each_net(VAR) \ list_for_each_entry(VAR, &net_namespace_list, list) +#define for_each_net_rcu(VAR) \ + list_for_each_entry_rcu(VAR, &net_namespace_list, list) + #ifdef CONFIG_NET_NS #define __net_init #define __net_exit diff --git a/net/core/net_namespace.c b/net/core/net_namespace.c index b7292a2719dc..5cd0b22e649d 100644 --- a/net/core/net_namespace.c +++ b/net/core/net_namespace.c @@ -6,6 +6,7 @@ #include #include #include +#include #include #include @@ -127,7 +128,7 @@ static struct net *net_create(void) rv = setup_net(net); if (rv == 0) { rtnl_lock(); - list_add_tail(&net->list, &net_namespace_list); + list_add_tail_rcu(&net->list, &net_namespace_list); rtnl_unlock(); } mutex_unlock(&net_mutex); @@ -156,9 +157,16 @@ static void cleanup_net(struct work_struct *work) /* Don't let anyone else find us. */ rtnl_lock(); - list_del(&net->list); + list_del_rcu(&net->list); rtnl_unlock(); + /* + * Another CPU might be rcu-iterating the list, wait for it. + * This needs to be before calling the exit() notifiers, so + * the rcu_barrier() below isn't sufficient alone. + */ + synchronize_rcu(); + /* Run all of the network namespace exit methods */ list_for_each_entry_reverse(ops, &pernet_list, list) { if (ops->exit) @@ -219,7 +227,7 @@ static int __init net_ns_init(void) panic("Could not setup the initial network namespace"); rtnl_lock(); - list_add_tail(&init_net.list, &net_namespace_list); + list_add_tail_rcu(&init_net.list, &net_namespace_list); rtnl_unlock(); mutex_unlock(&net_mutex); -- cgit v1.2.3 From 134e63756d5f3d0f7604dfcca847b09d1b14fd66 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Fri, 10 Jul 2009 09:51:34 +0000 Subject: genetlink: make netns aware This makes generic netlink network namespace aware. No generic netlink families except for the controller family are made namespace aware, they need to be checked one by one and then set the family->netnsok member to true. A new function genlmsg_multicast_netns() is introduced to allow sending a multicast message in a given namespace, for example when it applies to an object that lives in that namespace, a new function genlmsg_multicast_allns() to send a message to all network namespaces (for objects that do not have an associated netns). The function genlmsg_multicast() is changed to multicast the message in just init_net, which is currently correct for all generic netlink families since they only work in init_net right now. Some will later want to work in all net namespaces because they do not care about the netns at all -- those will have to be converted to use one of the new functions genlmsg_multicast_allns() or genlmsg_multicast_netns() whenever they are made netns aware in some way. After this patch families can easily decide whether or not they should be available in all net namespaces. Many genl families us it for objects not related to networking and should therefore be available in all namespaces, but that will have to be done on a per family basis. Note that this doesn't touch on the checkpoint/restart problem where network namespaces could be used, genl families and multicast groups are numbered globally and I see no easy way of changing that, especially since it must be possible to multicast to all network namespaces for those families that do not care about netns. Signed-off-by: Johannes Berg Signed-off-by: David S. Miller --- fs/dlm/netlink.c | 2 +- include/net/genetlink.h | 66 +++++++++++++-- include/net/net_namespace.h | 2 + kernel/taskstats.c | 10 +-- net/irda/irnetlink.c | 2 +- net/netfilter/ipvs/ip_vs_ctl.c | 2 +- net/netlink/genetlink.c | 186 ++++++++++++++++++++++++++++++++--------- net/tipc/netlink.c | 2 +- net/wireless/nl80211.c | 14 ++-- 9 files changed, 223 insertions(+), 63 deletions(-) (limited to 'include') diff --git a/fs/dlm/netlink.c b/fs/dlm/netlink.c index ccc9d62c462d..55ea369f43a9 100644 --- a/fs/dlm/netlink.c +++ b/fs/dlm/netlink.c @@ -63,7 +63,7 @@ static int send_data(struct sk_buff *skb) return rv; } - return genlmsg_unicast(skb, listener_nlpid); + return genlmsg_unicast(&init_net, skb, listener_nlpid); } static int user_cmd(struct sk_buff *skb, struct genl_info *info) diff --git a/include/net/genetlink.h b/include/net/genetlink.h index 1b0e3ee4ddd8..2a1c06874c42 100644 --- a/include/net/genetlink.h +++ b/include/net/genetlink.h @@ -3,6 +3,7 @@ #include #include +#include /** * struct genl_multicast_group - generic netlink multicast group @@ -27,6 +28,8 @@ struct genl_multicast_group * @name: name of family * @version: protocol version * @maxattr: maximum number of attributes supported + * @netnsok: set to true if the family can handle network + * namespaces and should be presented in all of them * @attrbuf: buffer to store parsed attributes * @ops_list: list of all assigned operations * @family_list: family list @@ -39,6 +42,7 @@ struct genl_family char name[GENL_NAMSIZ]; unsigned int version; unsigned int maxattr; + bool netnsok; struct nlattr ** attrbuf; /* private */ struct list_head ops_list; /* private */ struct list_head family_list; /* private */ @@ -62,8 +66,32 @@ struct genl_info struct genlmsghdr * genlhdr; void * userhdr; struct nlattr ** attrs; +#ifdef CONFIG_NET_NS + struct net * _net; +#endif }; +#ifdef CONFIG_NET_NS +static inline struct net *genl_info_net(struct genl_info *info) +{ + return info->_net; +} + +static inline void genl_info_net_set(struct genl_info *info, struct net *net) +{ + info->_net = net; +} +#else +static inline struct net *genl_info_net(struct genl_info *info) +{ + return &init_net; +} + +static inline void genl_info_net_set(struct genl_info *info, struct net *net) +{ +} +#endif + /** * struct genl_ops - generic netlink operations * @cmd: command identifier @@ -98,8 +126,6 @@ extern int genl_register_mc_group(struct genl_family *family, extern void genl_unregister_mc_group(struct genl_family *family, struct genl_multicast_group *grp); -extern struct sock *genl_sock; - /** * genlmsg_put - Add generic netlink header to netlink message * @skb: socket buffer holding the message @@ -170,7 +196,21 @@ static inline void genlmsg_cancel(struct sk_buff *skb, void *hdr) } /** - * genlmsg_multicast - multicast a netlink message + * genlmsg_multicast_netns - multicast a netlink message to a specific netns + * @net: the net namespace + * @skb: netlink message as socket buffer + * @pid: own netlink pid to avoid sending to yourself + * @group: multicast group id + * @flags: allocation flags + */ +static inline int genlmsg_multicast_netns(struct net *net, struct sk_buff *skb, + u32 pid, unsigned int group, gfp_t flags) +{ + return nlmsg_multicast(net->genl_sock, skb, pid, group, flags); +} + +/** + * genlmsg_multicast - multicast a netlink message to the default netns * @skb: netlink message as socket buffer * @pid: own netlink pid to avoid sending to yourself * @group: multicast group id @@ -179,17 +219,29 @@ static inline void genlmsg_cancel(struct sk_buff *skb, void *hdr) static inline int genlmsg_multicast(struct sk_buff *skb, u32 pid, unsigned int group, gfp_t flags) { - return nlmsg_multicast(genl_sock, skb, pid, group, flags); + return genlmsg_multicast_netns(&init_net, skb, pid, group, flags); } +/** + * genlmsg_multicast_allns - multicast a netlink message to all net namespaces + * @skb: netlink message as socket buffer + * @pid: own netlink pid to avoid sending to yourself + * @group: multicast group id + * @flags: allocation flags + * + * This function must hold the RTNL or rcu_read_lock(). + */ +int genlmsg_multicast_allns(struct sk_buff *skb, u32 pid, + unsigned int group, gfp_t flags); + /** * genlmsg_unicast - unicast a netlink message * @skb: netlink message as socket buffer * @pid: netlink pid of the destination socket */ -static inline int genlmsg_unicast(struct sk_buff *skb, u32 pid) +static inline int genlmsg_unicast(struct net *net, struct sk_buff *skb, u32 pid) { - return nlmsg_unicast(genl_sock, skb, pid); + return nlmsg_unicast(net->genl_sock, skb, pid); } /** @@ -199,7 +251,7 @@ static inline int genlmsg_unicast(struct sk_buff *skb, u32 pid) */ static inline int genlmsg_reply(struct sk_buff *skb, struct genl_info *info) { - return genlmsg_unicast(skb, info->snd_pid); + return genlmsg_unicast(genl_info_net(info), skb, info->snd_pid); } /** diff --git a/include/net/net_namespace.h b/include/net/net_namespace.h index b34a6ee73754..3173d12d946b 100644 --- a/include/net/net_namespace.h +++ b/include/net/net_namespace.h @@ -26,6 +26,7 @@ struct net_device; struct sock; struct ctl_table_header; struct net_generic; +struct sock; struct net { atomic_t count; /* To decided when the network @@ -57,6 +58,7 @@ struct net { spinlock_t rules_mod_lock; struct sock *rtnl; /* rtnetlink socket */ + struct sock *genl_sock; struct netns_core core; struct netns_mib mib; diff --git a/kernel/taskstats.c b/kernel/taskstats.c index 888adbcca30c..ea8384d3caa7 100644 --- a/kernel/taskstats.c +++ b/kernel/taskstats.c @@ -108,7 +108,7 @@ static int prepare_reply(struct genl_info *info, u8 cmd, struct sk_buff **skbp, /* * Send taskstats data in @skb to listener with nl_pid @pid */ -static int send_reply(struct sk_buff *skb, pid_t pid) +static int send_reply(struct sk_buff *skb, struct genl_info *info) { struct genlmsghdr *genlhdr = nlmsg_data(nlmsg_hdr(skb)); void *reply = genlmsg_data(genlhdr); @@ -120,7 +120,7 @@ static int send_reply(struct sk_buff *skb, pid_t pid) return rc; } - return genlmsg_unicast(skb, pid); + return genlmsg_reply(skb, info); } /* @@ -150,7 +150,7 @@ static void send_cpu_listeners(struct sk_buff *skb, if (!skb_next) break; } - rc = genlmsg_unicast(skb_cur, s->pid); + rc = genlmsg_unicast(&init_net, skb_cur, s->pid); if (rc == -ECONNREFUSED) { s->valid = 0; delcount++; @@ -418,7 +418,7 @@ static int cgroupstats_user_cmd(struct sk_buff *skb, struct genl_info *info) goto err; } - rc = send_reply(rep_skb, info->snd_pid); + rc = send_reply(rep_skb, info); err: fput_light(file, fput_needed); @@ -487,7 +487,7 @@ free_return_rc: } else goto err; - return send_reply(rep_skb, info->snd_pid); + return send_reply(rep_skb, info); err: nlmsg_free(rep_skb); return rc; diff --git a/net/irda/irnetlink.c b/net/irda/irnetlink.c index 8dd7ed7e7c1f..476b307bd801 100644 --- a/net/irda/irnetlink.c +++ b/net/irda/irnetlink.c @@ -115,7 +115,7 @@ static int irda_nl_get_mode(struct sk_buff *skb, struct genl_info *info) genlmsg_end(msg, hdr); - return genlmsg_unicast(msg, info->snd_pid); + return genlmsg_reply(msg, info); err_out: nlmsg_free(msg); diff --git a/net/netfilter/ipvs/ip_vs_ctl.c b/net/netfilter/ipvs/ip_vs_ctl.c index 7c1333c67ff3..2d24d81474ce 100644 --- a/net/netfilter/ipvs/ip_vs_ctl.c +++ b/net/netfilter/ipvs/ip_vs_ctl.c @@ -3231,7 +3231,7 @@ static int ip_vs_genl_get_cmd(struct sk_buff *skb, struct genl_info *info) } genlmsg_end(msg, reply); - ret = genlmsg_unicast(msg, info->snd_pid); + ret = genlmsg_reply(msg, info); goto out; nla_put_failure: diff --git a/net/netlink/genetlink.c b/net/netlink/genetlink.c index eed4c6a8afc0..575c64341508 100644 --- a/net/netlink/genetlink.c +++ b/net/netlink/genetlink.c @@ -18,8 +18,6 @@ #include #include -struct sock *genl_sock = NULL; - static DEFINE_MUTEX(genl_mutex); /* serialization of message processing */ static inline void genl_lock(void) @@ -175,10 +173,31 @@ int genl_register_mc_group(struct genl_family *family, mc_groups_longs++; } - err = netlink_change_ngroups(genl_sock, - mc_groups_longs * BITS_PER_LONG); - if (err) - goto out; + if (family->netnsok) { + struct net *net; + + rcu_read_lock(); + for_each_net_rcu(net) { + err = netlink_change_ngroups(net->genl_sock, + mc_groups_longs * BITS_PER_LONG); + if (err) { + /* + * No need to roll back, can only fail if + * memory allocation fails and then the + * number of _possible_ groups has been + * increased on some sockets which is ok. + */ + rcu_read_unlock(); + goto out; + } + } + rcu_read_unlock(); + } else { + err = netlink_change_ngroups(init_net.genl_sock, + mc_groups_longs * BITS_PER_LONG); + if (err) + goto out; + } grp->id = id; set_bit(id, mc_groups); @@ -195,8 +214,14 @@ EXPORT_SYMBOL(genl_register_mc_group); static void __genl_unregister_mc_group(struct genl_family *family, struct genl_multicast_group *grp) { + struct net *net; BUG_ON(grp->family != family); - netlink_clear_multicast_users(genl_sock, grp->id); + + rcu_read_lock(); + for_each_net_rcu(net) + netlink_clear_multicast_users(net->genl_sock, grp->id); + rcu_read_unlock(); + clear_bit(grp->id, mc_groups); list_del(&grp->list); genl_ctrl_event(CTRL_CMD_DELMCAST_GRP, grp); @@ -467,6 +492,7 @@ static int genl_rcv_msg(struct sk_buff *skb, struct nlmsghdr *nlh) { struct genl_ops *ops; struct genl_family *family; + struct net *net = sock_net(skb->sk); struct genl_info info; struct genlmsghdr *hdr = nlmsg_data(nlh); int hdrlen, err; @@ -475,6 +501,10 @@ static int genl_rcv_msg(struct sk_buff *skb, struct nlmsghdr *nlh) if (family == NULL) return -ENOENT; + /* this family doesn't exist in this netns */ + if (!family->netnsok && !net_eq(net, &init_net)) + return -ENOENT; + hdrlen = GENL_HDRLEN + family->hdrsize; if (nlh->nlmsg_len < nlmsg_msg_size(hdrlen)) return -EINVAL; @@ -492,7 +522,7 @@ static int genl_rcv_msg(struct sk_buff *skb, struct nlmsghdr *nlh) return -EOPNOTSUPP; genl_unlock(); - err = netlink_dump_start(genl_sock, skb, nlh, + err = netlink_dump_start(net->genl_sock, skb, nlh, ops->dumpit, ops->done); genl_lock(); return err; @@ -514,6 +544,7 @@ static int genl_rcv_msg(struct sk_buff *skb, struct nlmsghdr *nlh) info.genlhdr = nlmsg_data(nlh); info.userhdr = nlmsg_data(nlh) + GENL_HDRLEN; info.attrs = family->attrbuf; + genl_info_net_set(&info, net); return ops->doit(skb, &info); } @@ -534,6 +565,7 @@ static struct genl_family genl_ctrl = { .name = "nlctrl", .version = 0x2, .maxattr = CTRL_ATTR_MAX, + .netnsok = true, }; static int ctrl_fill_info(struct genl_family *family, u32 pid, u32 seq, @@ -650,6 +682,7 @@ static int ctrl_dumpfamily(struct sk_buff *skb, struct netlink_callback *cb) int i, n = 0; struct genl_family *rt; + struct net *net = sock_net(skb->sk); int chains_to_skip = cb->args[0]; int fams_to_skip = cb->args[1]; @@ -658,6 +691,8 @@ static int ctrl_dumpfamily(struct sk_buff *skb, struct netlink_callback *cb) continue; n = 0; list_for_each_entry(rt, genl_family_chain(i), family_list) { + if (!rt->netnsok && !net_eq(net, &init_net)) + continue; if (++n < fams_to_skip) continue; if (ctrl_fill_info(rt, NETLINK_CB(cb->skb).pid, @@ -729,6 +764,7 @@ static int ctrl_getfamily(struct sk_buff *skb, struct genl_info *info) if (info->attrs[CTRL_ATTR_FAMILY_ID]) { u16 id = nla_get_u16(info->attrs[CTRL_ATTR_FAMILY_ID]); res = genl_family_find_byid(id); + err = -ENOENT; } if (info->attrs[CTRL_ATTR_FAMILY_NAME]) { @@ -736,49 +772,61 @@ static int ctrl_getfamily(struct sk_buff *skb, struct genl_info *info) name = nla_data(info->attrs[CTRL_ATTR_FAMILY_NAME]); res = genl_family_find_byname(name); + err = -ENOENT; } - if (res == NULL) { - err = -ENOENT; - goto errout; + if (res == NULL) + return err; + + if (!res->netnsok && !net_eq(genl_info_net(info), &init_net)) { + /* family doesn't exist here */ + return -ENOENT; } msg = ctrl_build_family_msg(res, info->snd_pid, info->snd_seq, CTRL_CMD_NEWFAMILY); - if (IS_ERR(msg)) { - err = PTR_ERR(msg); - goto errout; - } + if (IS_ERR(msg)) + return PTR_ERR(msg); - err = genlmsg_reply(msg, info); -errout: - return err; + return genlmsg_reply(msg, info); } static int genl_ctrl_event(int event, void *data) { struct sk_buff *msg; + struct genl_family *family; + struct genl_multicast_group *grp; - if (genl_sock == NULL) + /* genl is still initialising */ + if (!init_net.genl_sock) return 0; switch (event) { case CTRL_CMD_NEWFAMILY: case CTRL_CMD_DELFAMILY: - msg = ctrl_build_family_msg(data, 0, 0, event); - if (IS_ERR(msg)) - return PTR_ERR(msg); - - genlmsg_multicast(msg, 0, GENL_ID_CTRL, GFP_KERNEL); + family = data; + msg = ctrl_build_family_msg(family, 0, 0, event); break; case CTRL_CMD_NEWMCAST_GRP: case CTRL_CMD_DELMCAST_GRP: + grp = data; + family = grp->family; msg = ctrl_build_mcgrp_msg(data, 0, 0, event); - if (IS_ERR(msg)) - return PTR_ERR(msg); - - genlmsg_multicast(msg, 0, GENL_ID_CTRL, GFP_KERNEL); break; + default: + return -EINVAL; + } + + if (IS_ERR(msg)) + return PTR_ERR(msg); + + if (!family->netnsok) { + genlmsg_multicast_netns(&init_net, msg, 0, + GENL_ID_CTRL, GFP_KERNEL); + } else { + rcu_read_lock(); + genlmsg_multicast_allns(msg, 0, GENL_ID_CTRL, GFP_ATOMIC); + rcu_read_unlock(); } return 0; @@ -795,6 +843,33 @@ static struct genl_multicast_group notify_grp = { .name = "notify", }; +static int __net_init genl_pernet_init(struct net *net) +{ + /* we'll bump the group number right afterwards */ + net->genl_sock = netlink_kernel_create(net, NETLINK_GENERIC, 0, + genl_rcv, &genl_mutex, + THIS_MODULE); + + if (!net->genl_sock && net_eq(net, &init_net)) + panic("GENL: Cannot initialize generic netlink\n"); + + if (!net->genl_sock) + return -ENOMEM; + + return 0; +} + +static void __net_exit genl_pernet_exit(struct net *net) +{ + netlink_kernel_release(net->genl_sock); + net->genl_sock = NULL; +} + +static struct pernet_operations genl_pernet_ops = { + .init = genl_pernet_init, + .exit = genl_pernet_exit, +}; + static int __init genl_init(void) { int i, err; @@ -804,36 +879,67 @@ static int __init genl_init(void) err = genl_register_family(&genl_ctrl); if (err < 0) - goto errout; + goto problem; err = genl_register_ops(&genl_ctrl, &genl_ctrl_ops); if (err < 0) - goto errout_register; + goto problem; netlink_set_nonroot(NETLINK_GENERIC, NL_NONROOT_RECV); - /* we'll bump the group number right afterwards */ - genl_sock = netlink_kernel_create(&init_net, NETLINK_GENERIC, 0, - genl_rcv, &genl_mutex, THIS_MODULE); - if (genl_sock == NULL) - panic("GENL: Cannot initialize generic netlink\n"); + err = register_pernet_subsys(&genl_pernet_ops); + if (err) + goto problem; err = genl_register_mc_group(&genl_ctrl, ¬ify_grp); if (err < 0) - goto errout_register; + goto problem; return 0; -errout_register: - genl_unregister_family(&genl_ctrl); -errout: +problem: panic("GENL: Cannot register controller: %d\n", err); } subsys_initcall(genl_init); -EXPORT_SYMBOL(genl_sock); EXPORT_SYMBOL(genl_register_ops); EXPORT_SYMBOL(genl_unregister_ops); EXPORT_SYMBOL(genl_register_family); EXPORT_SYMBOL(genl_unregister_family); + +static int genlmsg_mcast(struct sk_buff *skb, u32 pid, unsigned long group, + gfp_t flags) +{ + struct sk_buff *tmp; + struct net *net, *prev = NULL; + int err; + + for_each_net_rcu(net) { + if (prev) { + tmp = skb_clone(skb, flags); + if (!tmp) { + err = -ENOMEM; + goto error; + } + err = nlmsg_multicast(prev->genl_sock, tmp, + pid, group, flags); + if (err) + goto error; + } + + prev = net; + } + + return nlmsg_multicast(prev->genl_sock, skb, pid, group, flags); + error: + kfree_skb(skb); + return err; +} + +int genlmsg_multicast_allns(struct sk_buff *skb, u32 pid, unsigned int group, + gfp_t flags) +{ + return genlmsg_mcast(skb, pid, group, flags); +} +EXPORT_SYMBOL(genlmsg_multicast_allns); diff --git a/net/tipc/netlink.c b/net/tipc/netlink.c index 3c57005e44d1..7bda8e3d1398 100644 --- a/net/tipc/netlink.c +++ b/net/tipc/netlink.c @@ -62,7 +62,7 @@ static int handle_cmd(struct sk_buff *skb, struct genl_info *info) rep_nlh = nlmsg_hdr(rep_buf); memcpy(rep_nlh, req_nlh, hdr_space); rep_nlh->nlmsg_len = rep_buf->len; - genlmsg_unicast(rep_buf, NETLINK_CB(skb).pid); + genlmsg_unicast(&init_net, rep_buf, NETLINK_CB(skb).pid); } return 0; diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index 9deb12f73c44..2a04beba4369 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -413,7 +413,7 @@ static int nl80211_get_wiphy(struct sk_buff *skb, struct genl_info *info) cfg80211_unlock_rdev(dev); - return genlmsg_unicast(msg, info->snd_pid); + return genlmsg_reply(msg, info); out_free: nlmsg_free(msg); @@ -739,7 +739,7 @@ static int nl80211_get_interface(struct sk_buff *skb, struct genl_info *info) dev_put(netdev); cfg80211_unlock_rdev(dev); - return genlmsg_unicast(msg, info->snd_pid); + return genlmsg_reply(msg, info); out_free: nlmsg_free(msg); @@ -1030,7 +1030,7 @@ static int nl80211_get_key(struct sk_buff *skb, struct genl_info *info) goto nla_put_failure; genlmsg_end(msg, hdr); - err = genlmsg_unicast(msg, info->snd_pid); + err = genlmsg_reply(msg, info); goto out; nla_put_failure: @@ -1618,7 +1618,7 @@ static int nl80211_get_station(struct sk_buff *skb, struct genl_info *info) dev, mac_addr, &sinfo) < 0) goto out_free; - err = genlmsg_unicast(msg, info->snd_pid); + err = genlmsg_reply(msg, info); goto out; out_free: @@ -2087,7 +2087,7 @@ static int nl80211_get_mpath(struct sk_buff *skb, struct genl_info *info) dev, dst, next_hop, &pinfo) < 0) goto out_free; - err = genlmsg_unicast(msg, info->snd_pid); + err = genlmsg_reply(msg, info); goto out; out_free: @@ -2436,7 +2436,7 @@ static int nl80211_get_mesh_params(struct sk_buff *skb, cur_params.dot11MeshHWMPnetDiameterTraversalTime); nla_nest_end(msg, pinfoattr); genlmsg_end(msg, hdr); - err = genlmsg_unicast(msg, info->snd_pid); + err = genlmsg_reply(msg, info); goto out; nla_put_failure: @@ -2624,7 +2624,7 @@ static int nl80211_get_reg(struct sk_buff *skb, struct genl_info *info) nla_nest_end(msg, nl_reg_rules); genlmsg_end(msg, hdr); - err = genlmsg_unicast(msg, info->snd_pid); + err = genlmsg_reply(msg, info); goto out; nla_put_failure: -- cgit v1.2.3 From 30ffee8480c13fbcf8ab6c28e31f79dfff683117 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Fri, 10 Jul 2009 09:51:35 +0000 Subject: net: move and export get_net_ns_by_pid The function get_net_ns_by_pid(), to get a network namespace from a pid_t, will be required in cfg80211 as well. Therefore, let's move it to net_namespace.c and export it. We can't make it a static inline in the !NETNS case because it needs to verify that the given pid even exists (and return -ESRCH). Signed-off-by: Johannes Berg Signed-off-by: David S. Miller --- include/net/net_namespace.h | 2 ++ net/core/net_namespace.c | 21 +++++++++++++++++++++ net/core/rtnetlink.c | 21 +-------------------- 3 files changed, 24 insertions(+), 20 deletions(-) (limited to 'include') diff --git a/include/net/net_namespace.h b/include/net/net_namespace.h index 3173d12d946b..3882db1e263a 100644 --- a/include/net/net_namespace.h +++ b/include/net/net_namespace.h @@ -108,6 +108,8 @@ static inline struct net *copy_net_ns(unsigned long flags, struct net *net_ns) extern struct list_head net_namespace_list; +extern struct net *get_net_ns_by_pid(pid_t pid); + #ifdef CONFIG_NET_NS extern void __put_net(struct net *net); diff --git a/net/core/net_namespace.c b/net/core/net_namespace.c index 5cd0b22e649d..ddd2cd2b1775 100644 --- a/net/core/net_namespace.c +++ b/net/core/net_namespace.c @@ -7,6 +7,7 @@ #include #include #include +#include #include #include @@ -201,6 +202,26 @@ struct net *copy_net_ns(unsigned long flags, struct net *old_net) } #endif +struct net *get_net_ns_by_pid(pid_t pid) +{ + struct task_struct *tsk; + struct net *net; + + /* Lookup the network namespace */ + net = ERR_PTR(-ESRCH); + rcu_read_lock(); + tsk = find_task_by_vpid(pid); + if (tsk) { + struct nsproxy *nsproxy; + nsproxy = task_nsproxy(tsk); + if (nsproxy) + net = get_net(nsproxy->net_ns); + } + rcu_read_unlock(); + return net; +} +EXPORT_SYMBOL_GPL(get_net_ns_by_pid); + static int __init net_ns_init(void) { struct net_generic *ng; diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c index d78030f88bd0..b44775f9f2bf 100644 --- a/net/core/rtnetlink.c +++ b/net/core/rtnetlink.c @@ -35,7 +35,6 @@ #include #include #include -#include #include #include @@ -52,6 +51,7 @@ #include #include #include +#include struct rtnl_link { @@ -725,25 +725,6 @@ static const struct nla_policy ifla_info_policy[IFLA_INFO_MAX+1] = { [IFLA_INFO_DATA] = { .type = NLA_NESTED }, }; -static struct net *get_net_ns_by_pid(pid_t pid) -{ - struct task_struct *tsk; - struct net *net; - - /* Lookup the network namespace */ - net = ERR_PTR(-ESRCH); - rcu_read_lock(); - tsk = find_task_by_vpid(pid); - if (tsk) { - struct nsproxy *nsproxy; - nsproxy = task_nsproxy(tsk); - if (nsproxy) - net = get_net(nsproxy->net_ns); - } - rcu_read_unlock(); - return net; -} - static int validate_linkmsg(struct net_device *dev, struct nlattr *tb[]) { if (dev) { -- cgit v1.2.3 From d7ca4cc01fd154f2da30ae6dae160fa5800af758 Mon Sep 17 00:00:00 2001 From: Sridhar Samudrala Date: Thu, 9 Jul 2009 08:09:47 +0000 Subject: udpv4: Handle large incoming UDP/IPv4 packets and support software UFO. - validate and forward GSO UDP/IPv4 packets from untrusted sources. - do software UFO if the outgoing device doesn't support UFO. Signed-off-by: Sridhar Samudrala Acked-by: Herbert Xu Signed-off-by: David S. Miller --- include/net/udp.h | 3 +++ net/ipv4/af_inet.c | 12 ++++++++++- net/ipv4/udp.c | 61 ++++++++++++++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 75 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/net/udp.h b/include/net/udp.h index 90e6ce56be65..5fb029f817a3 100644 --- a/include/net/udp.h +++ b/include/net/udp.h @@ -207,4 +207,7 @@ extern void udp4_proc_exit(void); #endif extern void udp_init(void); + +extern int udp4_ufo_send_check(struct sk_buff *skb); +extern struct sk_buff *udp4_ufo_fragment(struct sk_buff *skb, int features); #endif /* _UDP_H */ diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c index 566ea6c4321d..197d024b2536 100644 --- a/net/ipv4/af_inet.c +++ b/net/ipv4/af_inet.c @@ -1187,6 +1187,7 @@ static struct sk_buff *inet_gso_segment(struct sk_buff *skb, int features) int proto; int ihl; int id; + unsigned int offset = 0; if (!(features & NETIF_F_V4_CSUM)) features &= ~NETIF_F_SG; @@ -1229,7 +1230,14 @@ static struct sk_buff *inet_gso_segment(struct sk_buff *skb, int features) skb = segs; do { iph = ip_hdr(skb); - iph->id = htons(id++); + if (proto == IPPROTO_UDP) { + iph->id = htons(id); + iph->frag_off = htons(offset >> 3); + if (skb->next != NULL) + iph->frag_off |= htons(IP_MF); + offset += (skb->len - skb->mac_len - iph->ihl * 4); + } else + iph->id = htons(id++); iph->tot_len = htons(skb->len - skb->mac_len); iph->check = 0; iph->check = ip_fast_csum(skb_network_header(skb), iph->ihl); @@ -1425,6 +1433,8 @@ static struct net_protocol tcp_protocol = { static struct net_protocol udp_protocol = { .handler = udp_rcv, .err_handler = udp_err, + .gso_send_check = udp4_ufo_send_check, + .gso_segment = udp4_ufo_fragment, .no_policy = 1, .netns_ok = 1, }; diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c index 80e3812837ad..7bc2d082a49e 100644 --- a/net/ipv4/udp.c +++ b/net/ipv4/udp.c @@ -1816,6 +1816,67 @@ void __init udp_init(void) sysctl_udp_wmem_min = SK_MEM_QUANTUM; } +int udp4_ufo_send_check(struct sk_buff *skb) +{ + const struct iphdr *iph; + struct udphdr *uh; + + if (!pskb_may_pull(skb, sizeof(*uh))) + return -EINVAL; + + iph = ip_hdr(skb); + uh = udp_hdr(skb); + + uh->check = ~csum_tcpudp_magic(iph->saddr, iph->daddr, skb->len, + IPPROTO_UDP, 0); + skb->csum_start = skb_transport_header(skb) - skb->head; + skb->csum_offset = offsetof(struct udphdr, check); + skb->ip_summed = CHECKSUM_PARTIAL; + return 0; +} + +struct sk_buff *udp4_ufo_fragment(struct sk_buff *skb, int features) +{ + struct sk_buff *segs = ERR_PTR(-EINVAL); + unsigned int mss; + int offset; + __wsum csum; + + mss = skb_shinfo(skb)->gso_size; + if (unlikely(skb->len <= mss)) + goto out; + + if (skb_gso_ok(skb, features | NETIF_F_GSO_ROBUST)) { + /* Packet is from an untrusted source, reset gso_segs. */ + int type = skb_shinfo(skb)->gso_type; + + if (unlikely(type & ~(SKB_GSO_UDP | SKB_GSO_DODGY) || + !(type & (SKB_GSO_UDP)))) + goto out; + + skb_shinfo(skb)->gso_segs = DIV_ROUND_UP(skb->len, mss); + + segs = NULL; + goto out; + } + + /* Do software UFO. Complete and fill in the UDP checksum as HW cannot + * do checksum of UDP packets sent as multiple IP fragments. + */ + offset = skb->csum_start - skb_headroom(skb); + csum = skb_checksum(skb, offset, skb->len- offset, 0); + offset += skb->csum_offset; + *(__sum16 *)(skb->data + offset) = csum_fold(csum); + skb->ip_summed = CHECKSUM_NONE; + + /* Fragment the skb. IP headers of the fragments are updated in + * inet_gso_segment() + */ + segs = skb_segment(skb, features); +out: + return segs; +} + EXPORT_SYMBOL(udp_disconnect); EXPORT_SYMBOL(udp_ioctl); EXPORT_SYMBOL(udp_prot); -- cgit v1.2.3 From 7ea2f2c5a66e4e9a8d96296ac47ad895c467ee1d Mon Sep 17 00:00:00 2001 From: Sridhar Samudrala Date: Thu, 9 Jul 2009 08:10:01 +0000 Subject: udpv6: Remove unused skb argument of ipv6_select_ident() - move ipv6_select_ident() inline function to ipv6.h and remove the unused skb argument Signed-off-by: Sridhar Samudrala Acked-by: Herbert Xu Signed-off-by: David S. Miller --- include/net/ipv6.h | 12 ++++++++++++ net/ipv6/ip6_output.c | 18 +++--------------- 2 files changed, 15 insertions(+), 15 deletions(-) (limited to 'include') diff --git a/include/net/ipv6.h b/include/net/ipv6.h index f27fd83d67d8..ad9a51130254 100644 --- a/include/net/ipv6.h +++ b/include/net/ipv6.h @@ -441,6 +441,18 @@ static inline int ipv6_addr_diff(const struct in6_addr *a1, const struct in6_add return __ipv6_addr_diff(a1, a2, sizeof(struct in6_addr)); } +static __inline__ void ipv6_select_ident(struct frag_hdr *fhdr) +{ + static u32 ipv6_fragmentation_id = 1; + static DEFINE_SPINLOCK(ip6_id_lock); + + spin_lock_bh(&ip6_id_lock); + fhdr->identification = htonl(ipv6_fragmentation_id); + if (++ipv6_fragmentation_id == 0) + ipv6_fragmentation_id = 1; + spin_unlock_bh(&ip6_id_lock); +} + /* * Prototypes exported by ipv6 */ diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c index 1c6f0fc43690..dd1a980b8ac9 100644 --- a/net/ipv6/ip6_output.c +++ b/net/ipv6/ip6_output.c @@ -57,18 +57,6 @@ static int ip6_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *)); -static __inline__ void ipv6_select_ident(struct sk_buff *skb, struct frag_hdr *fhdr) -{ - static u32 ipv6_fragmentation_id = 1; - static DEFINE_SPINLOCK(ip6_id_lock); - - spin_lock_bh(&ip6_id_lock); - fhdr->identification = htonl(ipv6_fragmentation_id); - if (++ipv6_fragmentation_id == 0) - ipv6_fragmentation_id = 1; - spin_unlock_bh(&ip6_id_lock); -} - int __ip6_local_out(struct sk_buff *skb) { int len; @@ -706,7 +694,7 @@ static int ip6_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *)) skb_reset_network_header(skb); memcpy(skb_network_header(skb), tmp_hdr, hlen); - ipv6_select_ident(skb, fh); + ipv6_select_ident(fh); fh->nexthdr = nexthdr; fh->reserved = 0; fh->frag_off = htons(IP6_MF); @@ -844,7 +832,7 @@ slow_path: fh->nexthdr = nexthdr; fh->reserved = 0; if (!frag_id) { - ipv6_select_ident(skb, fh); + ipv6_select_ident(fh); frag_id = fh->identification; } else fh->identification = frag_id; @@ -1093,7 +1081,7 @@ static inline int ip6_ufo_append_data(struct sock *sk, skb_shinfo(skb)->gso_size = (mtu - fragheaderlen - sizeof(struct frag_hdr)) & ~7; skb_shinfo(skb)->gso_type = SKB_GSO_UDP; - ipv6_select_ident(skb, &fhdr); + ipv6_select_ident(&fhdr); skb_shinfo(skb)->ip6_frag_id = fhdr.identification; __skb_queue_tail(&sk->sk_write_queue, skb); -- cgit v1.2.3 From b3a633c8527ef155b1a4e22e8f5abc58f7af54c9 Mon Sep 17 00:00:00 2001 From: Julien Tinnes Date: Fri, 10 Jul 2009 10:46:30 -0700 Subject: personality handling: fix PER_CLEAR_ON_SETID for security reasons We have found that the current PER_CLEAR_ON_SETID mask on Linux doesn't include neither ADDR_COMPAT_LAYOUT, nor MMAP_PAGE_ZERO. The current mask is READ_IMPLIES_EXEC|ADDR_NO_RANDOMIZE. We believe it is important to add MMAP_PAGE_ZERO, because by using this personality it is possible to have the first page mapped inside a process running as setuid root. This could be used in those scenarios: - Exploiting a NULL pointer dereference issue in a setuid root binary - Bypassing the mmap_min_addr restrictions of the Linux kernel: by running a setuid binary that would drop privileges before giving us control back (for instance by loading a user-supplied library), we could get the first page mapped in a process we control. By further using mremap and mprotect on this mapping, we can then completely bypass the mmap_min_addr restrictions. Less importantly, we believe ADDR_COMPAT_LAYOUT should also be added since on x86 32bits it will in practice disable most of the address space layout randomization (only the stack will remain randomized). Signed-off-by: Julien Tinnes Signed-off-by: Tavis Ormandy Acked-by: Christoph Hellwig Acked-by: Kees Cook Signed-off-by: James Morris --- include/linux/personality.h | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/personality.h b/include/linux/personality.h index a84e9ff9b27e..b7f578dac544 100644 --- a/include/linux/personality.h +++ b/include/linux/personality.h @@ -40,7 +40,11 @@ enum { * Security-relevant compatibility flags that must be * cleared upon setuid or setgid exec: */ -#define PER_CLEAR_ON_SETID (READ_IMPLIES_EXEC|ADDR_NO_RANDOMIZE) +#define PER_CLEAR_ON_SETID \ + (READ_IMPLIES_EXEC | \ + ADDR_NO_RANDOMIZE | \ + ADDR_COMPAT_LAYOUT | \ + MMAP_PAGE_ZERO) /* * Personality types. -- cgit v1.2.3 From 1a74826fa1cd6c2e382f927403b4440675f0f55a Mon Sep 17 00:00:00 2001 From: Greg Kroah-Hartman Date: Tue, 23 Jun 2009 15:58:48 -0700 Subject: Revert "USB: Add Intel Langwell USB OTG Transceiver Drive" This reverts commit 453f77558810ffa669ed5a510a7173ec49def396. The driver should not have been accepted as the MSRT code is not in the main kernel yet, which this depends on. Cc: Alan Cox Cc: Hao Wu Signed-off-by: Greg Kroah-Hartman --- drivers/usb/otg/Kconfig | 14 - drivers/usb/otg/Makefile | 1 - drivers/usb/otg/langwell_otg.c | 1915 -------------------------------------- include/linux/usb/langwell_otg.h | 177 ---- 4 files changed, 2107 deletions(-) delete mode 100644 drivers/usb/otg/langwell_otg.c delete mode 100644 include/linux/usb/langwell_otg.h (limited to 'include') diff --git a/drivers/usb/otg/Kconfig b/drivers/usb/otg/Kconfig index 69feeec1628c..aa884d072f0b 100644 --- a/drivers/usb/otg/Kconfig +++ b/drivers/usb/otg/Kconfig @@ -59,18 +59,4 @@ config NOP_USB_XCEIV built-in with usb ip or which are autonomous and doesn't require any phy programming such as ISP1x04 etc. -config USB_LANGWELL_OTG - tristate "Intel Langwell USB OTG dual-role support" - depends on USB && MRST - select USB_OTG - select USB_OTG_UTILS - help - Say Y here if you want to build Intel Langwell USB OTG - transciever driver in kernel. This driver implements role - switch between EHCI host driver and Langwell USB OTG - client driver. - - To compile this driver as a module, choose M here: the - module will be called langwell_otg. - endif # USB || OTG diff --git a/drivers/usb/otg/Makefile b/drivers/usb/otg/Makefile index 6d1abdd3c0ac..208167856529 100644 --- a/drivers/usb/otg/Makefile +++ b/drivers/usb/otg/Makefile @@ -9,7 +9,6 @@ obj-$(CONFIG_USB_OTG_UTILS) += otg.o obj-$(CONFIG_USB_GPIO_VBUS) += gpio_vbus.o obj-$(CONFIG_ISP1301_OMAP) += isp1301_omap.o obj-$(CONFIG_TWL4030_USB) += twl4030-usb.o -obj-$(CONFIG_USB_LANGWELL_OTG) += langwell_otg.o obj-$(CONFIG_NOP_USB_XCEIV) += nop-usb-xceiv.o ccflags-$(CONFIG_USB_DEBUG) += -DDEBUG diff --git a/drivers/usb/otg/langwell_otg.c b/drivers/usb/otg/langwell_otg.c deleted file mode 100644 index 6f628d0e9f39..000000000000 --- a/drivers/usb/otg/langwell_otg.c +++ /dev/null @@ -1,1915 +0,0 @@ -/* - * Intel Langwell USB OTG transceiver driver - * Copyright (C) 2008 - 2009, Intel Corporation. - * - * This program is free software; you can redistribute it and/or modify it - * under the terms and conditions of the GNU General Public License, - * version 2, as published by the Free Software Foundation. - * - * This program is distributed in the hope it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for - * more details. - * - * You should have received a copy of the GNU General Public License along with - * this program; if not, write to the Free Software Foundation, Inc., - * 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA. - * - */ -/* This driver helps to switch Langwell OTG controller function between host - * and peripheral. It works with EHCI driver and Langwell client controller - * driver together. - */ -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include "../core/hcd.h" - -#include - -#define DRIVER_DESC "Intel Langwell USB OTG transceiver driver" -#define DRIVER_VERSION "3.0.0.32L.0002" - -MODULE_DESCRIPTION(DRIVER_DESC); -MODULE_AUTHOR("Henry Yuan , Hao Wu "); -MODULE_VERSION(DRIVER_VERSION); -MODULE_LICENSE("GPL"); - -static const char driver_name[] = "langwell_otg"; - -static int langwell_otg_probe(struct pci_dev *pdev, - const struct pci_device_id *id); -static void langwell_otg_remove(struct pci_dev *pdev); -static int langwell_otg_suspend(struct pci_dev *pdev, pm_message_t message); -static int langwell_otg_resume(struct pci_dev *pdev); - -static int langwell_otg_set_host(struct otg_transceiver *otg, - struct usb_bus *host); -static int langwell_otg_set_peripheral(struct otg_transceiver *otg, - struct usb_gadget *gadget); -static int langwell_otg_start_srp(struct otg_transceiver *otg); - -static const struct pci_device_id pci_ids[] = {{ - .class = ((PCI_CLASS_SERIAL_USB << 8) | 0xfe), - .class_mask = ~0, - .vendor = 0x8086, - .device = 0x0811, - .subvendor = PCI_ANY_ID, - .subdevice = PCI_ANY_ID, -}, { /* end: all zeroes */ } -}; - -static struct pci_driver otg_pci_driver = { - .name = (char *) driver_name, - .id_table = pci_ids, - - .probe = langwell_otg_probe, - .remove = langwell_otg_remove, - - .suspend = langwell_otg_suspend, - .resume = langwell_otg_resume, -}; - -static const char *state_string(enum usb_otg_state state) -{ - switch (state) { - case OTG_STATE_A_IDLE: - return "a_idle"; - case OTG_STATE_A_WAIT_VRISE: - return "a_wait_vrise"; - case OTG_STATE_A_WAIT_BCON: - return "a_wait_bcon"; - case OTG_STATE_A_HOST: - return "a_host"; - case OTG_STATE_A_SUSPEND: - return "a_suspend"; - case OTG_STATE_A_PERIPHERAL: - return "a_peripheral"; - case OTG_STATE_A_WAIT_VFALL: - return "a_wait_vfall"; - case OTG_STATE_A_VBUS_ERR: - return "a_vbus_err"; - case OTG_STATE_B_IDLE: - return "b_idle"; - case OTG_STATE_B_SRP_INIT: - return "b_srp_init"; - case OTG_STATE_B_PERIPHERAL: - return "b_peripheral"; - case OTG_STATE_B_WAIT_ACON: - return "b_wait_acon"; - case OTG_STATE_B_HOST: - return "b_host"; - default: - return "UNDEFINED"; - } -} - -/* HSM timers */ -static inline struct langwell_otg_timer *otg_timer_initializer -(void (*function)(unsigned long), unsigned long expires, unsigned long data) -{ - struct langwell_otg_timer *timer; - timer = kmalloc(sizeof(struct langwell_otg_timer), GFP_KERNEL); - timer->function = function; - timer->expires = expires; - timer->data = data; - return timer; -} - -static struct langwell_otg_timer *a_wait_vrise_tmr, *a_wait_bcon_tmr, - *a_aidl_bdis_tmr, *b_ase0_brst_tmr, *b_se0_srp_tmr, *b_srp_res_tmr, - *b_bus_suspend_tmr; - -static struct list_head active_timers; - -static struct langwell_otg *the_transceiver; - -/* host/client notify transceiver when event affects HNP state */ -void langwell_update_transceiver() -{ - otg_dbg("transceiver driver is notified\n"); - queue_work(the_transceiver->qwork, &the_transceiver->work); -} -EXPORT_SYMBOL(langwell_update_transceiver); - -static int langwell_otg_set_host(struct otg_transceiver *otg, - struct usb_bus *host) -{ - otg->host = host; - - return 0; -} - -static int langwell_otg_set_peripheral(struct otg_transceiver *otg, - struct usb_gadget *gadget) -{ - otg->gadget = gadget; - - return 0; -} - -static int langwell_otg_set_power(struct otg_transceiver *otg, - unsigned mA) -{ - return 0; -} - -/* A-device drives vbus, controlled through PMIC CHRGCNTL register*/ -static void langwell_otg_drv_vbus(int on) -{ - struct ipc_pmic_reg_data pmic_data = {0}; - struct ipc_pmic_reg_data battery_data; - - /* Check if battery is attached or not */ - battery_data.pmic_reg_data[0].register_address = 0xd2; - battery_data.ioc = 0; - battery_data.num_entries = 1; - if (ipc_pmic_register_read(&battery_data)) { - otg_dbg("Failed to read PMIC register 0xd2.\n"); - return; - } - - if ((battery_data.pmic_reg_data[0].value & 0x20) == 0) { - otg_dbg("no battery attached\n"); - return; - } - - /* Workaround for battery attachment issue */ - if (battery_data.pmic_reg_data[0].value == 0x34) { - otg_dbg("battery \n"); - return; - } - - otg_dbg("battery attached\n"); - - pmic_data.ioc = 0; - pmic_data.pmic_reg_data[0].register_address = 0xD4; - pmic_data.num_entries = 1; - if (on) - pmic_data.pmic_reg_data[0].value = 0x20; - else - pmic_data.pmic_reg_data[0].value = 0xc0; - - if (ipc_pmic_register_write(&pmic_data, TRUE)) - otg_dbg("Failed to write PMIC.\n"); - -} - -/* charge vbus or discharge vbus through a resistor to ground */ -static void langwell_otg_chrg_vbus(int on) -{ - - u32 val; - - val = readl(the_transceiver->regs + CI_OTGSC); - - if (on) - writel((val & ~OTGSC_INTSTS_MASK) | OTGSC_VC, - the_transceiver->regs + CI_OTGSC); - else - writel((val & ~OTGSC_INTSTS_MASK) | OTGSC_VD, - the_transceiver->regs + CI_OTGSC); - -} - -/* Start SRP */ -static int langwell_otg_start_srp(struct otg_transceiver *otg) -{ - u32 val; - - otg_dbg("Start SRP ->\n"); - - val = readl(the_transceiver->regs + CI_OTGSC); - - writel((val & ~OTGSC_INTSTS_MASK) | OTGSC_HADP, - the_transceiver->regs + CI_OTGSC); - - /* Check if the data plus is finished or not */ - msleep(8); - val = readl(the_transceiver->regs + CI_OTGSC); - if (val & (OTGSC_HADP | OTGSC_DP)) - otg_dbg("DataLine SRP Error\n"); - - /* FIXME: VBus SRP */ - - return 0; -} - - -/* stop SOF via bus_suspend */ -static void langwell_otg_loc_sof(int on) -{ - struct usb_hcd *hcd; - int err; - - otg_dbg("loc_sof -> %d\n", on); - - hcd = bus_to_hcd(the_transceiver->otg.host); - if (on) - err = hcd->driver->bus_resume(hcd); - else - err = hcd->driver->bus_suspend(hcd); - - if (err) - otg_dbg("Failed to resume/suspend bus - %d\n", err); -} - -static void langwell_otg_phy_low_power(int on) -{ - u32 val; - - otg_dbg("phy low power mode-> %d\n", on); - - val = readl(the_transceiver->regs + CI_HOSTPC1); - if (on) - writel(val | HOSTPC1_PHCD, the_transceiver->regs + CI_HOSTPC1); - else - writel(val & ~HOSTPC1_PHCD, the_transceiver->regs + CI_HOSTPC1); -} - -/* Enable/Disable OTG interrupt */ -static void langwell_otg_intr(int on) -{ - u32 val; - - otg_dbg("interrupt -> %d\n", on); - - val = readl(the_transceiver->regs + CI_OTGSC); - if (on) { - val = val | (OTGSC_INTEN_MASK | OTGSC_IDPU); - writel(val, the_transceiver->regs + CI_OTGSC); - } else { - val = val & ~(OTGSC_INTEN_MASK | OTGSC_IDPU); - writel(val, the_transceiver->regs + CI_OTGSC); - } -} - -/* set HAAR: Hardware Assist Auto-Reset */ -static void langwell_otg_HAAR(int on) -{ - u32 val; - - otg_dbg("HAAR -> %d\n", on); - - val = readl(the_transceiver->regs + CI_OTGSC); - if (on) - writel((val & ~OTGSC_INTSTS_MASK) | OTGSC_HAAR, - the_transceiver->regs + CI_OTGSC); - else - writel((val & ~OTGSC_INTSTS_MASK) & ~OTGSC_HAAR, - the_transceiver->regs + CI_OTGSC); -} - -/* set HABA: Hardware Assist B-Disconnect to A-Connect */ -static void langwell_otg_HABA(int on) -{ - u32 val; - - otg_dbg("HABA -> %d\n", on); - - val = readl(the_transceiver->regs + CI_OTGSC); - if (on) - writel((val & ~OTGSC_INTSTS_MASK) | OTGSC_HABA, - the_transceiver->regs + CI_OTGSC); - else - writel((val & ~OTGSC_INTSTS_MASK) & ~OTGSC_HABA, - the_transceiver->regs + CI_OTGSC); -} - -static int langwell_otg_check_se0_srp(int on) -{ - u32 val; - - int delay_time = TB_SE0_SRP * 10; /* step is 100us */ - - otg_dbg("check_se0_srp -> \n"); - - do { - udelay(100); - if (!delay_time--) - break; - val = readl(the_transceiver->regs + CI_PORTSC1); - val &= PORTSC_LS; - } while (!val); - - otg_dbg("check_se0_srp <- \n"); - return val; -} - -/* The timeout callback function to set time out bit */ -static void set_tmout(unsigned long indicator) -{ - *(int *)indicator = 1; -} - -void langwell_otg_nsf_msg(unsigned long indicator) -{ - switch (indicator) { - case 2: - case 4: - case 6: - case 7: - printk(KERN_ERR "OTG:NSF-%lu - deivce not responding\n", - indicator); - break; - case 3: - printk(KERN_ERR "OTG:NSF-%lu - deivce not supported\n", - indicator); - break; - default: - printk(KERN_ERR "Do not have this kind of NSF\n"); - break; - } -} - -/* Initialize timers */ -static void langwell_otg_init_timers(struct otg_hsm *hsm) -{ - /* HSM used timers */ - a_wait_vrise_tmr = otg_timer_initializer(&set_tmout, TA_WAIT_VRISE, - (unsigned long)&hsm->a_wait_vrise_tmout); - a_wait_bcon_tmr = otg_timer_initializer(&set_tmout, TA_WAIT_BCON, - (unsigned long)&hsm->a_wait_bcon_tmout); - a_aidl_bdis_tmr = otg_timer_initializer(&set_tmout, TA_AIDL_BDIS, - (unsigned long)&hsm->a_aidl_bdis_tmout); - b_ase0_brst_tmr = otg_timer_initializer(&set_tmout, TB_ASE0_BRST, - (unsigned long)&hsm->b_ase0_brst_tmout); - b_se0_srp_tmr = otg_timer_initializer(&set_tmout, TB_SE0_SRP, - (unsigned long)&hsm->b_se0_srp); - b_srp_res_tmr = otg_timer_initializer(&set_tmout, TB_SRP_RES, - (unsigned long)&hsm->b_srp_res_tmout); - b_bus_suspend_tmr = otg_timer_initializer(&set_tmout, TB_BUS_SUSPEND, - (unsigned long)&hsm->b_bus_suspend_tmout); -} - -/* Free timers */ -static void langwell_otg_free_timers(void) -{ - kfree(a_wait_vrise_tmr); - kfree(a_wait_bcon_tmr); - kfree(a_aidl_bdis_tmr); - kfree(b_ase0_brst_tmr); - kfree(b_se0_srp_tmr); - kfree(b_srp_res_tmr); - kfree(b_bus_suspend_tmr); -} - -/* Add timer to timer list */ -static void langwell_otg_add_timer(void *gtimer) -{ - struct langwell_otg_timer *timer = (struct langwell_otg_timer *)gtimer; - struct langwell_otg_timer *tmp_timer; - u32 val32; - - /* Check if the timer is already in the active list, - * if so update timer count - */ - list_for_each_entry(tmp_timer, &active_timers, list) - if (tmp_timer == timer) { - timer->count = timer->expires; - return; - } - timer->count = timer->expires; - - if (list_empty(&active_timers)) { - val32 = readl(the_transceiver->regs + CI_OTGSC); - writel(val32 | OTGSC_1MSE, the_transceiver->regs + CI_OTGSC); - } - - list_add_tail(&timer->list, &active_timers); -} - -/* Remove timer from the timer list; clear timeout status */ -static void langwell_otg_del_timer(void *gtimer) -{ - struct langwell_otg_timer *timer = (struct langwell_otg_timer *)gtimer; - struct langwell_otg_timer *tmp_timer, *del_tmp; - u32 val32; - - list_for_each_entry_safe(tmp_timer, del_tmp, &active_timers, list) - if (tmp_timer == timer) - list_del(&timer->list); - - if (list_empty(&active_timers)) { - val32 = readl(the_transceiver->regs + CI_OTGSC); - writel(val32 & ~OTGSC_1MSE, the_transceiver->regs + CI_OTGSC); - } -} - -/* Reduce timer count by 1, and find timeout conditions.*/ -static int langwell_otg_tick_timer(u32 *int_sts) -{ - struct langwell_otg_timer *tmp_timer, *del_tmp; - int expired = 0; - - list_for_each_entry_safe(tmp_timer, del_tmp, &active_timers, list) { - tmp_timer->count--; - /* check if timer expires */ - if (!tmp_timer->count) { - list_del(&tmp_timer->list); - tmp_timer->function(tmp_timer->data); - expired = 1; - } - } - - if (list_empty(&active_timers)) { - otg_dbg("tick timer: disable 1ms int\n"); - *int_sts = *int_sts & ~OTGSC_1MSE; - } - return expired; -} - -static void reset_otg(void) -{ - u32 val; - int delay_time = 1000; - - otg_dbg("reseting OTG controller ...\n"); - val = readl(the_transceiver->regs + CI_USBCMD); - writel(val | USBCMD_RST, the_transceiver->regs + CI_USBCMD); - do { - udelay(100); - if (!delay_time--) - otg_dbg("reset timeout\n"); - val = readl(the_transceiver->regs + CI_USBCMD); - val &= USBCMD_RST; - } while (val != 0); - otg_dbg("reset done.\n"); -} - -static void set_host_mode(void) -{ - u32 val; - - reset_otg(); - val = readl(the_transceiver->regs + CI_USBMODE); - val = (val & (~USBMODE_CM)) | USBMODE_HOST; - writel(val, the_transceiver->regs + CI_USBMODE); -} - -static void set_client_mode(void) -{ - u32 val; - - reset_otg(); - val = readl(the_transceiver->regs + CI_USBMODE); - val = (val & (~USBMODE_CM)) | USBMODE_DEVICE; - writel(val, the_transceiver->regs + CI_USBMODE); -} - -static void init_hsm(void) -{ - struct langwell_otg *langwell = the_transceiver; - u32 val32; - - /* read OTGSC after reset */ - val32 = readl(langwell->regs + CI_OTGSC); - otg_dbg("%s: OTGSC init value = 0x%x\n", __func__, val32); - - /* set init state */ - if (val32 & OTGSC_ID) { - langwell->hsm.id = 1; - langwell->otg.default_a = 0; - set_client_mode(); - langwell->otg.state = OTG_STATE_B_IDLE; - langwell_otg_drv_vbus(0); - } else { - langwell->hsm.id = 0; - langwell->otg.default_a = 1; - set_host_mode(); - langwell->otg.state = OTG_STATE_A_IDLE; - } - - /* set session indicator */ - if (val32 & OTGSC_BSE) - langwell->hsm.b_sess_end = 1; - if (val32 & OTGSC_BSV) - langwell->hsm.b_sess_vld = 1; - if (val32 & OTGSC_ASV) - langwell->hsm.a_sess_vld = 1; - if (val32 & OTGSC_AVV) - langwell->hsm.a_vbus_vld = 1; - - /* defautly power the bus */ - langwell->hsm.a_bus_req = 1; - langwell->hsm.a_bus_drop = 0; - /* defautly don't request bus as B device */ - langwell->hsm.b_bus_req = 0; - /* no system error */ - langwell->hsm.a_clr_err = 0; -} - -static irqreturn_t otg_dummy_irq(int irq, void *_dev) -{ - void __iomem *reg_base = _dev; - u32 val; - u32 int_mask = 0; - - val = readl(reg_base + CI_USBMODE); - if ((val & USBMODE_CM) != USBMODE_DEVICE) - return IRQ_NONE; - - val = readl(reg_base + CI_USBSTS); - int_mask = val & INTR_DUMMY_MASK; - - if (int_mask == 0) - return IRQ_NONE; - - /* clear hsm.b_conn here since host driver can't detect it - * otg_dummy_irq called means B-disconnect happened. - */ - if (the_transceiver->hsm.b_conn) { - the_transceiver->hsm.b_conn = 0; - if (spin_trylock(&the_transceiver->wq_lock)) { - queue_work(the_transceiver->qwork, - &the_transceiver->work); - spin_unlock(&the_transceiver->wq_lock); - } - } - /* Clear interrupts */ - writel(int_mask, reg_base + CI_USBSTS); - return IRQ_HANDLED; -} - -static irqreturn_t otg_irq(int irq, void *_dev) -{ - struct langwell_otg *langwell = _dev; - u32 int_sts, int_en; - u32 int_mask = 0; - int flag = 0; - - int_sts = readl(langwell->regs + CI_OTGSC); - int_en = (int_sts & OTGSC_INTEN_MASK) >> 8; - int_mask = int_sts & int_en; - if (int_mask == 0) - return IRQ_NONE; - - if (int_mask & OTGSC_IDIS) { - otg_dbg("%s: id change int\n", __func__); - langwell->hsm.id = (int_sts & OTGSC_ID) ? 1 : 0; - flag = 1; - } - if (int_mask & OTGSC_DPIS) { - otg_dbg("%s: data pulse int\n", __func__); - langwell->hsm.a_srp_det = (int_sts & OTGSC_DPS) ? 1 : 0; - flag = 1; - } - if (int_mask & OTGSC_BSEIS) { - otg_dbg("%s: b session end int\n", __func__); - langwell->hsm.b_sess_end = (int_sts & OTGSC_BSE) ? 1 : 0; - flag = 1; - } - if (int_mask & OTGSC_BSVIS) { - otg_dbg("%s: b session valid int\n", __func__); - langwell->hsm.b_sess_vld = (int_sts & OTGSC_BSV) ? 1 : 0; - flag = 1; - } - if (int_mask & OTGSC_ASVIS) { - otg_dbg("%s: a session valid int\n", __func__); - langwell->hsm.a_sess_vld = (int_sts & OTGSC_ASV) ? 1 : 0; - flag = 1; - } - if (int_mask & OTGSC_AVVIS) { - otg_dbg("%s: a vbus valid int\n", __func__); - langwell->hsm.a_vbus_vld = (int_sts & OTGSC_AVV) ? 1 : 0; - flag = 1; - } - - if (int_mask & OTGSC_1MSS) { - /* need to schedule otg_work if any timer is expired */ - if (langwell_otg_tick_timer(&int_sts)) - flag = 1; - } - - writel((int_sts & ~OTGSC_INTSTS_MASK) | int_mask, - langwell->regs + CI_OTGSC); - if (flag) - queue_work(langwell->qwork, &langwell->work); - - return IRQ_HANDLED; -} - -static void langwell_otg_work(struct work_struct *work) -{ - struct langwell_otg *langwell = container_of(work, - struct langwell_otg, work); - int retval; - - otg_dbg("%s: old state = %s\n", __func__, - state_string(langwell->otg.state)); - - switch (langwell->otg.state) { - case OTG_STATE_UNDEFINED: - case OTG_STATE_B_IDLE: - if (!langwell->hsm.id) { - langwell_otg_del_timer(b_srp_res_tmr); - langwell->otg.default_a = 1; - langwell->hsm.a_srp_det = 0; - - langwell_otg_chrg_vbus(0); - langwell_otg_drv_vbus(0); - - set_host_mode(); - langwell->otg.state = OTG_STATE_A_IDLE; - queue_work(langwell->qwork, &langwell->work); - } else if (langwell->hsm.b_srp_res_tmout) { - langwell->hsm.b_srp_res_tmout = 0; - langwell->hsm.b_bus_req = 0; - langwell_otg_nsf_msg(6); - } else if (langwell->hsm.b_sess_vld) { - langwell_otg_del_timer(b_srp_res_tmr); - langwell->hsm.b_sess_end = 0; - langwell->hsm.a_bus_suspend = 0; - - langwell_otg_chrg_vbus(0); - if (langwell->client_ops) { - langwell->client_ops->resume(langwell->pdev); - langwell->otg.state = OTG_STATE_B_PERIPHERAL; - } else - otg_dbg("client driver not loaded.\n"); - - } else if (langwell->hsm.b_bus_req && - (langwell->hsm.b_sess_end)) { - /* workaround for b_se0_srp detection */ - retval = langwell_otg_check_se0_srp(0); - if (retval) { - langwell->hsm.b_bus_req = 0; - otg_dbg("LS is not SE0, try again later\n"); - } else { - /* Start SRP */ - langwell_otg_start_srp(&langwell->otg); - langwell_otg_add_timer(b_srp_res_tmr); - } - } - break; - case OTG_STATE_B_SRP_INIT: - if (!langwell->hsm.id) { - langwell->otg.default_a = 1; - langwell->hsm.a_srp_det = 0; - - langwell_otg_drv_vbus(0); - langwell_otg_chrg_vbus(0); - - langwell->otg.state = OTG_STATE_A_IDLE; - queue_work(langwell->qwork, &langwell->work); - } else if (langwell->hsm.b_sess_vld) { - langwell_otg_chrg_vbus(0); - if (langwell->client_ops) { - langwell->client_ops->resume(langwell->pdev); - langwell->otg.state = OTG_STATE_B_PERIPHERAL; - } else - otg_dbg("client driver not loaded.\n"); - } - break; - case OTG_STATE_B_PERIPHERAL: - if (!langwell->hsm.id) { - langwell->otg.default_a = 1; - langwell->hsm.a_srp_det = 0; - - langwell_otg_drv_vbus(0); - langwell_otg_chrg_vbus(0); - set_host_mode(); - - if (langwell->client_ops) { - langwell->client_ops->suspend(langwell->pdev, - PMSG_FREEZE); - } else - otg_dbg("client driver has been removed.\n"); - - langwell->otg.state = OTG_STATE_A_IDLE; - queue_work(langwell->qwork, &langwell->work); - } else if (!langwell->hsm.b_sess_vld) { - langwell->hsm.b_hnp_enable = 0; - - if (langwell->client_ops) { - langwell->client_ops->suspend(langwell->pdev, - PMSG_FREEZE); - } else - otg_dbg("client driver has been removed.\n"); - - langwell->otg.state = OTG_STATE_B_IDLE; - } else if (langwell->hsm.b_bus_req && langwell->hsm.b_hnp_enable - && langwell->hsm.a_bus_suspend) { - - if (langwell->client_ops) { - langwell->client_ops->suspend(langwell->pdev, - PMSG_FREEZE); - } else - otg_dbg("client driver has been removed.\n"); - - langwell_otg_HAAR(1); - langwell->hsm.a_conn = 0; - - if (langwell->host_ops) { - langwell->host_ops->probe(langwell->pdev, - langwell->host_ops->id_table); - langwell->otg.state = OTG_STATE_B_WAIT_ACON; - } else - otg_dbg("host driver not loaded.\n"); - - langwell->hsm.a_bus_resume = 0; - langwell->hsm.b_ase0_brst_tmout = 0; - langwell_otg_add_timer(b_ase0_brst_tmr); - } - break; - - case OTG_STATE_B_WAIT_ACON: - if (!langwell->hsm.id) { - langwell_otg_del_timer(b_ase0_brst_tmr); - langwell->otg.default_a = 1; - langwell->hsm.a_srp_det = 0; - - langwell_otg_drv_vbus(0); - langwell_otg_chrg_vbus(0); - set_host_mode(); - - langwell_otg_HAAR(0); - if (langwell->host_ops) - langwell->host_ops->remove(langwell->pdev); - else - otg_dbg("host driver has been removed.\n"); - langwell->otg.state = OTG_STATE_A_IDLE; - queue_work(langwell->qwork, &langwell->work); - } else if (!langwell->hsm.b_sess_vld) { - langwell_otg_del_timer(b_ase0_brst_tmr); - langwell->hsm.b_hnp_enable = 0; - langwell->hsm.b_bus_req = 0; - langwell_otg_chrg_vbus(0); - langwell_otg_HAAR(0); - - if (langwell->host_ops) - langwell->host_ops->remove(langwell->pdev); - else - otg_dbg("host driver has been removed.\n"); - langwell->otg.state = OTG_STATE_B_IDLE; - } else if (langwell->hsm.a_conn) { - langwell_otg_del_timer(b_ase0_brst_tmr); - langwell_otg_HAAR(0); - langwell->otg.state = OTG_STATE_B_HOST; - queue_work(langwell->qwork, &langwell->work); - } else if (langwell->hsm.a_bus_resume || - langwell->hsm.b_ase0_brst_tmout) { - langwell_otg_del_timer(b_ase0_brst_tmr); - langwell_otg_HAAR(0); - langwell_otg_nsf_msg(7); - - if (langwell->host_ops) - langwell->host_ops->remove(langwell->pdev); - else - otg_dbg("host driver has been removed.\n"); - - langwell->hsm.a_bus_suspend = 0; - langwell->hsm.b_bus_req = 0; - - if (langwell->client_ops) - langwell->client_ops->resume(langwell->pdev); - else - otg_dbg("client driver not loaded.\n"); - - langwell->otg.state = OTG_STATE_B_PERIPHERAL; - } - break; - - case OTG_STATE_B_HOST: - if (!langwell->hsm.id) { - langwell->otg.default_a = 1; - langwell->hsm.a_srp_det = 0; - - langwell_otg_drv_vbus(0); - langwell_otg_chrg_vbus(0); - set_host_mode(); - if (langwell->host_ops) - langwell->host_ops->remove(langwell->pdev); - else - otg_dbg("host driver has been removed.\n"); - langwell->otg.state = OTG_STATE_A_IDLE; - queue_work(langwell->qwork, &langwell->work); - } else if (!langwell->hsm.b_sess_vld) { - langwell->hsm.b_hnp_enable = 0; - langwell->hsm.b_bus_req = 0; - langwell_otg_chrg_vbus(0); - if (langwell->host_ops) - langwell->host_ops->remove(langwell->pdev); - else - otg_dbg("host driver has been removed.\n"); - langwell->otg.state = OTG_STATE_B_IDLE; - } else if ((!langwell->hsm.b_bus_req) || - (!langwell->hsm.a_conn)) { - langwell->hsm.b_bus_req = 0; - langwell_otg_loc_sof(0); - if (langwell->host_ops) - langwell->host_ops->remove(langwell->pdev); - else - otg_dbg("host driver has been removed.\n"); - - langwell->hsm.a_bus_suspend = 0; - - if (langwell->client_ops) - langwell->client_ops->resume(langwell->pdev); - else - otg_dbg("client driver not loaded.\n"); - - langwell->otg.state = OTG_STATE_B_PERIPHERAL; - } - break; - - case OTG_STATE_A_IDLE: - langwell->otg.default_a = 1; - if (langwell->hsm.id) { - langwell->otg.default_a = 0; - langwell->hsm.b_bus_req = 0; - langwell_otg_drv_vbus(0); - langwell_otg_chrg_vbus(0); - - langwell->otg.state = OTG_STATE_B_IDLE; - queue_work(langwell->qwork, &langwell->work); - } else if (langwell->hsm.a_sess_vld) { - langwell_otg_drv_vbus(1); - langwell->hsm.a_srp_det = 1; - langwell->hsm.a_wait_vrise_tmout = 0; - langwell_otg_add_timer(a_wait_vrise_tmr); - langwell->otg.state = OTG_STATE_A_WAIT_VRISE; - queue_work(langwell->qwork, &langwell->work); - } else if (!langwell->hsm.a_bus_drop && - (langwell->hsm.a_srp_det || langwell->hsm.a_bus_req)) { - langwell_otg_drv_vbus(1); - langwell->hsm.a_wait_vrise_tmout = 0; - langwell_otg_add_timer(a_wait_vrise_tmr); - langwell->otg.state = OTG_STATE_A_WAIT_VRISE; - queue_work(langwell->qwork, &langwell->work); - } - break; - case OTG_STATE_A_WAIT_VRISE: - if (langwell->hsm.id) { - langwell_otg_del_timer(a_wait_vrise_tmr); - langwell->hsm.b_bus_req = 0; - langwell->otg.default_a = 0; - langwell_otg_drv_vbus(0); - langwell->otg.state = OTG_STATE_B_IDLE; - } else if (langwell->hsm.a_vbus_vld) { - langwell_otg_del_timer(a_wait_vrise_tmr); - if (langwell->host_ops) - langwell->host_ops->probe(langwell->pdev, - langwell->host_ops->id_table); - else - otg_dbg("host driver not loaded.\n"); - langwell->hsm.b_conn = 0; - langwell->hsm.a_set_b_hnp_en = 0; - langwell->hsm.a_wait_bcon_tmout = 0; - langwell_otg_add_timer(a_wait_bcon_tmr); - langwell->otg.state = OTG_STATE_A_WAIT_BCON; - } else if (langwell->hsm.a_wait_vrise_tmout) { - if (langwell->hsm.a_vbus_vld) { - if (langwell->host_ops) - langwell->host_ops->probe( - langwell->pdev, - langwell->host_ops->id_table); - else - otg_dbg("host driver not loaded.\n"); - langwell->hsm.b_conn = 0; - langwell->hsm.a_set_b_hnp_en = 0; - langwell->hsm.a_wait_bcon_tmout = 0; - langwell_otg_add_timer(a_wait_bcon_tmr); - langwell->otg.state = OTG_STATE_A_WAIT_BCON; - } else { - langwell_otg_drv_vbus(0); - langwell->otg.state = OTG_STATE_A_VBUS_ERR; - } - } - break; - case OTG_STATE_A_WAIT_BCON: - if (langwell->hsm.id) { - langwell_otg_del_timer(a_wait_bcon_tmr); - - langwell->otg.default_a = 0; - langwell->hsm.b_bus_req = 0; - if (langwell->host_ops) - langwell->host_ops->remove(langwell->pdev); - else - otg_dbg("host driver has been removed.\n"); - langwell_otg_drv_vbus(0); - langwell->otg.state = OTG_STATE_B_IDLE; - queue_work(langwell->qwork, &langwell->work); - } else if (!langwell->hsm.a_vbus_vld) { - langwell_otg_del_timer(a_wait_bcon_tmr); - - if (langwell->host_ops) - langwell->host_ops->remove(langwell->pdev); - else - otg_dbg("host driver has been removed.\n"); - langwell_otg_drv_vbus(0); - langwell->otg.state = OTG_STATE_A_VBUS_ERR; - } else if (langwell->hsm.a_bus_drop || - (langwell->hsm.a_wait_bcon_tmout && - !langwell->hsm.a_bus_req)) { - langwell_otg_del_timer(a_wait_bcon_tmr); - - if (langwell->host_ops) - langwell->host_ops->remove(langwell->pdev); - else - otg_dbg("host driver has been removed.\n"); - langwell_otg_drv_vbus(0); - langwell->otg.state = OTG_STATE_A_WAIT_VFALL; - } else if (langwell->hsm.b_conn) { - langwell_otg_del_timer(a_wait_bcon_tmr); - - langwell->hsm.a_suspend_req = 0; - langwell->otg.state = OTG_STATE_A_HOST; - if (!langwell->hsm.a_bus_req && - langwell->hsm.a_set_b_hnp_en) { - /* It is not safe enough to do a fast - * transistion from A_WAIT_BCON to - * A_SUSPEND */ - msleep(10000); - if (langwell->hsm.a_bus_req) - break; - - if (request_irq(langwell->pdev->irq, - otg_dummy_irq, IRQF_SHARED, - driver_name, langwell->regs) != 0) { - otg_dbg("request interrupt %d fail\n", - langwell->pdev->irq); - } - - langwell_otg_HABA(1); - langwell->hsm.b_bus_resume = 0; - langwell->hsm.a_aidl_bdis_tmout = 0; - langwell_otg_add_timer(a_aidl_bdis_tmr); - - langwell_otg_loc_sof(0); - langwell->otg.state = OTG_STATE_A_SUSPEND; - } else if (!langwell->hsm.a_bus_req && - !langwell->hsm.a_set_b_hnp_en) { - struct pci_dev *pdev = langwell->pdev; - if (langwell->host_ops) - langwell->host_ops->remove(pdev); - else - otg_dbg("host driver removed.\n"); - langwell_otg_drv_vbus(0); - langwell->otg.state = OTG_STATE_A_WAIT_VFALL; - } - } - break; - case OTG_STATE_A_HOST: - if (langwell->hsm.id) { - langwell->otg.default_a = 0; - langwell->hsm.b_bus_req = 0; - if (langwell->host_ops) - langwell->host_ops->remove(langwell->pdev); - else - otg_dbg("host driver has been removed.\n"); - langwell_otg_drv_vbus(0); - langwell->otg.state = OTG_STATE_B_IDLE; - queue_work(langwell->qwork, &langwell->work); - } else if (langwell->hsm.a_bus_drop || - (!langwell->hsm.a_set_b_hnp_en && !langwell->hsm.a_bus_req)) { - if (langwell->host_ops) - langwell->host_ops->remove(langwell->pdev); - else - otg_dbg("host driver has been removed.\n"); - langwell_otg_drv_vbus(0); - langwell->otg.state = OTG_STATE_A_WAIT_VFALL; - } else if (!langwell->hsm.a_vbus_vld) { - if (langwell->host_ops) - langwell->host_ops->remove(langwell->pdev); - else - otg_dbg("host driver has been removed.\n"); - langwell_otg_drv_vbus(0); - langwell->otg.state = OTG_STATE_A_VBUS_ERR; - } else if (langwell->hsm.a_set_b_hnp_en - && !langwell->hsm.a_bus_req) { - /* Set HABA to enable hardware assistance to signal - * A-connect after receiver B-disconnect. Hardware - * will then set client mode and enable URE, SLE and - * PCE after the assistance. otg_dummy_irq is used to - * clean these ints when client driver is not resumed. - */ - if (request_irq(langwell->pdev->irq, - otg_dummy_irq, IRQF_SHARED, driver_name, - langwell->regs) != 0) { - otg_dbg("request interrupt %d failed\n", - langwell->pdev->irq); - } - - /* set HABA */ - langwell_otg_HABA(1); - langwell->hsm.b_bus_resume = 0; - langwell->hsm.a_aidl_bdis_tmout = 0; - langwell_otg_add_timer(a_aidl_bdis_tmr); - langwell_otg_loc_sof(0); - langwell->otg.state = OTG_STATE_A_SUSPEND; - } else if (!langwell->hsm.b_conn || !langwell->hsm.a_bus_req) { - langwell->hsm.a_wait_bcon_tmout = 0; - langwell->hsm.a_set_b_hnp_en = 0; - langwell_otg_add_timer(a_wait_bcon_tmr); - langwell->otg.state = OTG_STATE_A_WAIT_BCON; - } - break; - case OTG_STATE_A_SUSPEND: - if (langwell->hsm.id) { - langwell_otg_del_timer(a_aidl_bdis_tmr); - langwell_otg_HABA(0); - free_irq(langwell->pdev->irq, langwell->regs); - langwell->otg.default_a = 0; - langwell->hsm.b_bus_req = 0; - if (langwell->host_ops) - langwell->host_ops->remove(langwell->pdev); - else - otg_dbg("host driver has been removed.\n"); - langwell_otg_drv_vbus(0); - langwell->otg.state = OTG_STATE_B_IDLE; - queue_work(langwell->qwork, &langwell->work); - } else if (langwell->hsm.a_bus_req || - langwell->hsm.b_bus_resume) { - langwell_otg_del_timer(a_aidl_bdis_tmr); - langwell_otg_HABA(0); - free_irq(langwell->pdev->irq, langwell->regs); - langwell->hsm.a_suspend_req = 0; - langwell_otg_loc_sof(1); - langwell->otg.state = OTG_STATE_A_HOST; - } else if (langwell->hsm.a_aidl_bdis_tmout || - langwell->hsm.a_bus_drop) { - langwell_otg_del_timer(a_aidl_bdis_tmr); - langwell_otg_HABA(0); - free_irq(langwell->pdev->irq, langwell->regs); - if (langwell->host_ops) - langwell->host_ops->remove(langwell->pdev); - else - otg_dbg("host driver has been removed.\n"); - langwell_otg_drv_vbus(0); - langwell->otg.state = OTG_STATE_A_WAIT_VFALL; - } else if (!langwell->hsm.b_conn && - langwell->hsm.a_set_b_hnp_en) { - langwell_otg_del_timer(a_aidl_bdis_tmr); - langwell_otg_HABA(0); - free_irq(langwell->pdev->irq, langwell->regs); - - if (langwell->host_ops) - langwell->host_ops->remove(langwell->pdev); - else - otg_dbg("host driver has been removed.\n"); - - langwell->hsm.b_bus_suspend = 0; - langwell->hsm.b_bus_suspend_vld = 0; - langwell->hsm.b_bus_suspend_tmout = 0; - - /* msleep(200); */ - if (langwell->client_ops) - langwell->client_ops->resume(langwell->pdev); - else - otg_dbg("client driver not loaded.\n"); - - langwell_otg_add_timer(b_bus_suspend_tmr); - langwell->otg.state = OTG_STATE_A_PERIPHERAL; - break; - } else if (!langwell->hsm.a_vbus_vld) { - langwell_otg_del_timer(a_aidl_bdis_tmr); - langwell_otg_HABA(0); - free_irq(langwell->pdev->irq, langwell->regs); - if (langwell->host_ops) - langwell->host_ops->remove(langwell->pdev); - else - otg_dbg("host driver has been removed.\n"); - langwell_otg_drv_vbus(0); - langwell->otg.state = OTG_STATE_A_VBUS_ERR; - } - break; - case OTG_STATE_A_PERIPHERAL: - if (langwell->hsm.id) { - langwell_otg_del_timer(b_bus_suspend_tmr); - langwell->otg.default_a = 0; - langwell->hsm.b_bus_req = 0; - if (langwell->client_ops) - langwell->client_ops->suspend(langwell->pdev, - PMSG_FREEZE); - else - otg_dbg("client driver has been removed.\n"); - langwell_otg_drv_vbus(0); - langwell->otg.state = OTG_STATE_B_IDLE; - queue_work(langwell->qwork, &langwell->work); - } else if (!langwell->hsm.a_vbus_vld) { - langwell_otg_del_timer(b_bus_suspend_tmr); - if (langwell->client_ops) - langwell->client_ops->suspend(langwell->pdev, - PMSG_FREEZE); - else - otg_dbg("client driver has been removed.\n"); - langwell_otg_drv_vbus(0); - langwell->otg.state = OTG_STATE_A_VBUS_ERR; - } else if (langwell->hsm.a_bus_drop) { - langwell_otg_del_timer(b_bus_suspend_tmr); - if (langwell->client_ops) - langwell->client_ops->suspend(langwell->pdev, - PMSG_FREEZE); - else - otg_dbg("client driver has been removed.\n"); - langwell_otg_drv_vbus(0); - langwell->otg.state = OTG_STATE_A_WAIT_VFALL; - } else if (langwell->hsm.b_bus_suspend) { - langwell_otg_del_timer(b_bus_suspend_tmr); - if (langwell->client_ops) - langwell->client_ops->suspend(langwell->pdev, - PMSG_FREEZE); - else - otg_dbg("client driver has been removed.\n"); - - if (langwell->host_ops) - langwell->host_ops->probe(langwell->pdev, - langwell->host_ops->id_table); - else - otg_dbg("host driver not loaded.\n"); - langwell->hsm.a_set_b_hnp_en = 0; - langwell->hsm.a_wait_bcon_tmout = 0; - langwell_otg_add_timer(a_wait_bcon_tmr); - langwell->otg.state = OTG_STATE_A_WAIT_BCON; - } else if (langwell->hsm.b_bus_suspend_tmout) { - u32 val; - val = readl(langwell->regs + CI_PORTSC1); - if (!(val & PORTSC_SUSP)) - break; - if (langwell->client_ops) - langwell->client_ops->suspend(langwell->pdev, - PMSG_FREEZE); - else - otg_dbg("client driver has been removed.\n"); - if (langwell->host_ops) - langwell->host_ops->probe(langwell->pdev, - langwell->host_ops->id_table); - else - otg_dbg("host driver not loaded.\n"); - langwell->hsm.a_set_b_hnp_en = 0; - langwell->hsm.a_wait_bcon_tmout = 0; - langwell_otg_add_timer(a_wait_bcon_tmr); - langwell->otg.state = OTG_STATE_A_WAIT_BCON; - } - break; - case OTG_STATE_A_VBUS_ERR: - if (langwell->hsm.id) { - langwell->otg.default_a = 0; - langwell->hsm.a_clr_err = 0; - langwell->hsm.a_srp_det = 0; - langwell->otg.state = OTG_STATE_B_IDLE; - queue_work(langwell->qwork, &langwell->work); - } else if (langwell->hsm.a_clr_err) { - langwell->hsm.a_clr_err = 0; - langwell->hsm.a_srp_det = 0; - reset_otg(); - init_hsm(); - if (langwell->otg.state == OTG_STATE_A_IDLE) - queue_work(langwell->qwork, &langwell->work); - } - break; - case OTG_STATE_A_WAIT_VFALL: - if (langwell->hsm.id) { - langwell->otg.default_a = 0; - langwell->otg.state = OTG_STATE_B_IDLE; - queue_work(langwell->qwork, &langwell->work); - } else if (langwell->hsm.a_bus_req) { - langwell_otg_drv_vbus(1); - langwell->hsm.a_wait_vrise_tmout = 0; - langwell_otg_add_timer(a_wait_vrise_tmr); - langwell->otg.state = OTG_STATE_A_WAIT_VRISE; - } else if (!langwell->hsm.a_sess_vld) { - langwell->hsm.a_srp_det = 0; - langwell_otg_drv_vbus(0); - set_host_mode(); - langwell->otg.state = OTG_STATE_A_IDLE; - } - break; - default: - ; - } - - otg_dbg("%s: new state = %s\n", __func__, - state_string(langwell->otg.state)); -} - - static ssize_t -show_registers(struct device *_dev, struct device_attribute *attr, char *buf) -{ - struct langwell_otg *langwell; - char *next; - unsigned size; - unsigned t; - - langwell = the_transceiver; - next = buf; - size = PAGE_SIZE; - - t = scnprintf(next, size, - "\n" - "USBCMD = 0x%08x \n" - "USBSTS = 0x%08x \n" - "USBINTR = 0x%08x \n" - "ASYNCLISTADDR = 0x%08x \n" - "PORTSC1 = 0x%08x \n" - "HOSTPC1 = 0x%08x \n" - "OTGSC = 0x%08x \n" - "USBMODE = 0x%08x \n", - readl(langwell->regs + 0x30), - readl(langwell->regs + 0x34), - readl(langwell->regs + 0x38), - readl(langwell->regs + 0x48), - readl(langwell->regs + 0x74), - readl(langwell->regs + 0xb4), - readl(langwell->regs + 0xf4), - readl(langwell->regs + 0xf8) - ); - size -= t; - next += t; - - return PAGE_SIZE - size; -} -static DEVICE_ATTR(registers, S_IRUGO, show_registers, NULL); - -static ssize_t -show_hsm(struct device *_dev, struct device_attribute *attr, char *buf) -{ - struct langwell_otg *langwell; - char *next; - unsigned size; - unsigned t; - - langwell = the_transceiver; - next = buf; - size = PAGE_SIZE; - - t = scnprintf(next, size, - "\n" - "current state = %s\n" - "a_bus_resume = \t%d\n" - "a_bus_suspend = \t%d\n" - "a_conn = \t%d\n" - "a_sess_vld = \t%d\n" - "a_srp_det = \t%d\n" - "a_vbus_vld = \t%d\n" - "b_bus_resume = \t%d\n" - "b_bus_suspend = \t%d\n" - "b_conn = \t%d\n" - "b_se0_srp = \t%d\n" - "b_sess_end = \t%d\n" - "b_sess_vld = \t%d\n" - "id = \t%d\n" - "a_set_b_hnp_en = \t%d\n" - "b_srp_done = \t%d\n" - "b_hnp_enable = \t%d\n" - "a_wait_vrise_tmout = \t%d\n" - "a_wait_bcon_tmout = \t%d\n" - "a_aidl_bdis_tmout = \t%d\n" - "b_ase0_brst_tmout = \t%d\n" - "a_bus_drop = \t%d\n" - "a_bus_req = \t%d\n" - "a_clr_err = \t%d\n" - "a_suspend_req = \t%d\n" - "b_bus_req = \t%d\n" - "b_bus_suspend_tmout = \t%d\n" - "b_bus_suspend_vld = \t%d\n", - state_string(langwell->otg.state), - langwell->hsm.a_bus_resume, - langwell->hsm.a_bus_suspend, - langwell->hsm.a_conn, - langwell->hsm.a_sess_vld, - langwell->hsm.a_srp_det, - langwell->hsm.a_vbus_vld, - langwell->hsm.b_bus_resume, - langwell->hsm.b_bus_suspend, - langwell->hsm.b_conn, - langwell->hsm.b_se0_srp, - langwell->hsm.b_sess_end, - langwell->hsm.b_sess_vld, - langwell->hsm.id, - langwell->hsm.a_set_b_hnp_en, - langwell->hsm.b_srp_done, - langwell->hsm.b_hnp_enable, - langwell->hsm.a_wait_vrise_tmout, - langwell->hsm.a_wait_bcon_tmout, - langwell->hsm.a_aidl_bdis_tmout, - langwell->hsm.b_ase0_brst_tmout, - langwell->hsm.a_bus_drop, - langwell->hsm.a_bus_req, - langwell->hsm.a_clr_err, - langwell->hsm.a_suspend_req, - langwell->hsm.b_bus_req, - langwell->hsm.b_bus_suspend_tmout, - langwell->hsm.b_bus_suspend_vld - ); - size -= t; - next += t; - - return PAGE_SIZE - size; -} -static DEVICE_ATTR(hsm, S_IRUGO, show_hsm, NULL); - -static ssize_t -get_a_bus_req(struct device *dev, struct device_attribute *attr, char *buf) -{ - struct langwell_otg *langwell; - char *next; - unsigned size; - unsigned t; - - langwell = the_transceiver; - next = buf; - size = PAGE_SIZE; - - t = scnprintf(next, size, "%d", langwell->hsm.a_bus_req); - size -= t; - next += t; - - return PAGE_SIZE - size; -} - -static ssize_t -set_a_bus_req(struct device *dev, struct device_attribute *attr, - const char *buf, size_t count) -{ - struct langwell_otg *langwell; - langwell = the_transceiver; - if (!langwell->otg.default_a) - return -1; - if (count > 2) - return -1; - - if (buf[0] == '0') { - langwell->hsm.a_bus_req = 0; - otg_dbg("a_bus_req = 0\n"); - } else if (buf[0] == '1') { - /* If a_bus_drop is TRUE, a_bus_req can't be set */ - if (langwell->hsm.a_bus_drop) - return -1; - langwell->hsm.a_bus_req = 1; - otg_dbg("a_bus_req = 1\n"); - } - if (spin_trylock(&langwell->wq_lock)) { - queue_work(langwell->qwork, &langwell->work); - spin_unlock(&langwell->wq_lock); - } - return count; -} -static DEVICE_ATTR(a_bus_req, S_IRUGO | S_IWUGO, get_a_bus_req, set_a_bus_req); - -static ssize_t -get_a_bus_drop(struct device *dev, struct device_attribute *attr, char *buf) -{ - struct langwell_otg *langwell; - char *next; - unsigned size; - unsigned t; - - langwell = the_transceiver; - next = buf; - size = PAGE_SIZE; - - t = scnprintf(next, size, "%d", langwell->hsm.a_bus_drop); - size -= t; - next += t; - - return PAGE_SIZE - size; -} - -static ssize_t -set_a_bus_drop(struct device *dev, struct device_attribute *attr, - const char *buf, size_t count) -{ - struct langwell_otg *langwell; - langwell = the_transceiver; - if (!langwell->otg.default_a) - return -1; - if (count > 2) - return -1; - - if (buf[0] == '0') { - langwell->hsm.a_bus_drop = 0; - otg_dbg("a_bus_drop = 0\n"); - } else if (buf[0] == '1') { - langwell->hsm.a_bus_drop = 1; - langwell->hsm.a_bus_req = 0; - otg_dbg("a_bus_drop = 1, then a_bus_req = 0\n"); - } - if (spin_trylock(&langwell->wq_lock)) { - queue_work(langwell->qwork, &langwell->work); - spin_unlock(&langwell->wq_lock); - } - return count; -} -static DEVICE_ATTR(a_bus_drop, S_IRUGO | S_IWUGO, - get_a_bus_drop, set_a_bus_drop); - -static ssize_t -get_b_bus_req(struct device *dev, struct device_attribute *attr, char *buf) -{ - struct langwell_otg *langwell; - char *next; - unsigned size; - unsigned t; - - langwell = the_transceiver; - next = buf; - size = PAGE_SIZE; - - t = scnprintf(next, size, "%d", langwell->hsm.b_bus_req); - size -= t; - next += t; - - return PAGE_SIZE - size; -} - -static ssize_t -set_b_bus_req(struct device *dev, struct device_attribute *attr, - const char *buf, size_t count) -{ - struct langwell_otg *langwell; - langwell = the_transceiver; - - if (langwell->otg.default_a) - return -1; - - if (count > 2) - return -1; - - if (buf[0] == '0') { - langwell->hsm.b_bus_req = 0; - otg_dbg("b_bus_req = 0\n"); - } else if (buf[0] == '1') { - langwell->hsm.b_bus_req = 1; - otg_dbg("b_bus_req = 1\n"); - } - if (spin_trylock(&langwell->wq_lock)) { - queue_work(langwell->qwork, &langwell->work); - spin_unlock(&langwell->wq_lock); - } - return count; -} -static DEVICE_ATTR(b_bus_req, S_IRUGO | S_IWUGO, get_b_bus_req, set_b_bus_req); - -static ssize_t -set_a_clr_err(struct device *dev, struct device_attribute *attr, - const char *buf, size_t count) -{ - struct langwell_otg *langwell; - langwell = the_transceiver; - - if (!langwell->otg.default_a) - return -1; - if (count > 2) - return -1; - - if (buf[0] == '1') { - langwell->hsm.a_clr_err = 1; - otg_dbg("a_clr_err = 1\n"); - } - if (spin_trylock(&langwell->wq_lock)) { - queue_work(langwell->qwork, &langwell->work); - spin_unlock(&langwell->wq_lock); - } - return count; -} -static DEVICE_ATTR(a_clr_err, S_IWUGO, NULL, set_a_clr_err); - -static struct attribute *inputs_attrs[] = { - &dev_attr_a_bus_req.attr, - &dev_attr_a_bus_drop.attr, - &dev_attr_b_bus_req.attr, - &dev_attr_a_clr_err.attr, - NULL, -}; - -static struct attribute_group debug_dev_attr_group = { - .name = "inputs", - .attrs = inputs_attrs, -}; - -int langwell_register_host(struct pci_driver *host_driver) -{ - int ret = 0; - - the_transceiver->host_ops = host_driver; - queue_work(the_transceiver->qwork, &the_transceiver->work); - otg_dbg("host controller driver is registered\n"); - - return ret; -} -EXPORT_SYMBOL(langwell_register_host); - -void langwell_unregister_host(struct pci_driver *host_driver) -{ - if (the_transceiver->host_ops) - the_transceiver->host_ops->remove(the_transceiver->pdev); - the_transceiver->host_ops = NULL; - the_transceiver->hsm.a_bus_drop = 1; - queue_work(the_transceiver->qwork, &the_transceiver->work); - otg_dbg("host controller driver is unregistered\n"); -} -EXPORT_SYMBOL(langwell_unregister_host); - -int langwell_register_peripheral(struct pci_driver *client_driver) -{ - int ret = 0; - - if (client_driver) - ret = client_driver->probe(the_transceiver->pdev, - client_driver->id_table); - if (!ret) { - the_transceiver->client_ops = client_driver; - queue_work(the_transceiver->qwork, &the_transceiver->work); - otg_dbg("client controller driver is registered\n"); - } - - return ret; -} -EXPORT_SYMBOL(langwell_register_peripheral); - -void langwell_unregister_peripheral(struct pci_driver *client_driver) -{ - if (the_transceiver->client_ops) - the_transceiver->client_ops->remove(the_transceiver->pdev); - the_transceiver->client_ops = NULL; - the_transceiver->hsm.b_bus_req = 0; - queue_work(the_transceiver->qwork, &the_transceiver->work); - otg_dbg("client controller driver is unregistered\n"); -} -EXPORT_SYMBOL(langwell_unregister_peripheral); - -static int langwell_otg_probe(struct pci_dev *pdev, - const struct pci_device_id *id) -{ - unsigned long resource, len; - void __iomem *base = NULL; - int retval; - u32 val32; - struct langwell_otg *langwell; - char qname[] = "langwell_otg_queue"; - - retval = 0; - otg_dbg("\notg controller is detected.\n"); - if (pci_enable_device(pdev) < 0) { - retval = -ENODEV; - goto done; - } - - langwell = kzalloc(sizeof *langwell, GFP_KERNEL); - if (langwell == NULL) { - retval = -ENOMEM; - goto done; - } - the_transceiver = langwell; - - /* control register: BAR 0 */ - resource = pci_resource_start(pdev, 0); - len = pci_resource_len(pdev, 0); - if (!request_mem_region(resource, len, driver_name)) { - retval = -EBUSY; - goto err; - } - langwell->region = 1; - - base = ioremap_nocache(resource, len); - if (base == NULL) { - retval = -EFAULT; - goto err; - } - langwell->regs = base; - - if (!pdev->irq) { - otg_dbg("No IRQ.\n"); - retval = -ENODEV; - goto err; - } - - langwell->qwork = create_workqueue(qname); - if (!langwell->qwork) { - otg_dbg("cannot create workqueue %s\n", qname); - retval = -ENOMEM; - goto err; - } - INIT_WORK(&langwell->work, langwell_otg_work); - - /* OTG common part */ - langwell->pdev = pdev; - langwell->otg.dev = &pdev->dev; - langwell->otg.label = driver_name; - langwell->otg.set_host = langwell_otg_set_host; - langwell->otg.set_peripheral = langwell_otg_set_peripheral; - langwell->otg.set_power = langwell_otg_set_power; - langwell->otg.start_srp = langwell_otg_start_srp; - langwell->otg.state = OTG_STATE_UNDEFINED; - if (otg_set_transceiver(&langwell->otg)) { - otg_dbg("can't set transceiver\n"); - retval = -EBUSY; - goto err; - } - - reset_otg(); - init_hsm(); - - spin_lock_init(&langwell->lock); - spin_lock_init(&langwell->wq_lock); - INIT_LIST_HEAD(&active_timers); - langwell_otg_init_timers(&langwell->hsm); - - if (request_irq(pdev->irq, otg_irq, IRQF_SHARED, - driver_name, langwell) != 0) { - otg_dbg("request interrupt %d failed\n", pdev->irq); - retval = -EBUSY; - goto err; - } - - /* enable OTGSC int */ - val32 = OTGSC_DPIE | OTGSC_BSEIE | OTGSC_BSVIE | - OTGSC_ASVIE | OTGSC_AVVIE | OTGSC_IDIE | OTGSC_IDPU; - writel(val32, langwell->regs + CI_OTGSC); - - retval = device_create_file(&pdev->dev, &dev_attr_registers); - if (retval < 0) { - otg_dbg("Can't register sysfs attribute: %d\n", retval); - goto err; - } - - retval = device_create_file(&pdev->dev, &dev_attr_hsm); - if (retval < 0) { - otg_dbg("Can't hsm sysfs attribute: %d\n", retval); - goto err; - } - - retval = sysfs_create_group(&pdev->dev.kobj, &debug_dev_attr_group); - if (retval < 0) { - otg_dbg("Can't register sysfs attr group: %d\n", retval); - goto err; - } - - if (langwell->otg.state == OTG_STATE_A_IDLE) - queue_work(langwell->qwork, &langwell->work); - - return 0; - -err: - if (the_transceiver) - langwell_otg_remove(pdev); -done: - return retval; -} - -static void langwell_otg_remove(struct pci_dev *pdev) -{ - struct langwell_otg *langwell; - - langwell = the_transceiver; - - if (langwell->qwork) { - flush_workqueue(langwell->qwork); - destroy_workqueue(langwell->qwork); - } - langwell_otg_free_timers(); - - /* disable OTGSC interrupt as OTGSC doesn't change in reset */ - writel(0, langwell->regs + CI_OTGSC); - - if (pdev->irq) - free_irq(pdev->irq, langwell); - if (langwell->regs) - iounmap(langwell->regs); - if (langwell->region) - release_mem_region(pci_resource_start(pdev, 0), - pci_resource_len(pdev, 0)); - - otg_set_transceiver(NULL); - pci_disable_device(pdev); - sysfs_remove_group(&pdev->dev.kobj, &debug_dev_attr_group); - device_remove_file(&pdev->dev, &dev_attr_hsm); - device_remove_file(&pdev->dev, &dev_attr_registers); - kfree(langwell); - langwell = NULL; -} - -static void transceiver_suspend(struct pci_dev *pdev) -{ - pci_save_state(pdev); - pci_set_power_state(pdev, PCI_D3hot); - langwell_otg_phy_low_power(1); -} - -static int langwell_otg_suspend(struct pci_dev *pdev, pm_message_t message) -{ - int ret = 0; - struct langwell_otg *langwell; - - langwell = the_transceiver; - - /* Disbale OTG interrupts */ - langwell_otg_intr(0); - - if (pdev->irq) - free_irq(pdev->irq, langwell); - - /* Prevent more otg_work */ - flush_workqueue(langwell->qwork); - spin_lock(&langwell->wq_lock); - - /* start actions */ - switch (langwell->otg.state) { - case OTG_STATE_A_IDLE: - case OTG_STATE_B_IDLE: - case OTG_STATE_A_WAIT_VFALL: - case OTG_STATE_A_VBUS_ERR: - transceiver_suspend(pdev); - break; - case OTG_STATE_A_WAIT_VRISE: - langwell_otg_del_timer(a_wait_vrise_tmr); - langwell->hsm.a_srp_det = 0; - langwell_otg_drv_vbus(0); - langwell->otg.state = OTG_STATE_A_IDLE; - transceiver_suspend(pdev); - break; - case OTG_STATE_A_WAIT_BCON: - langwell_otg_del_timer(a_wait_bcon_tmr); - if (langwell->host_ops) - ret = langwell->host_ops->suspend(pdev, message); - langwell_otg_drv_vbus(0); - break; - case OTG_STATE_A_HOST: - if (langwell->host_ops) - ret = langwell->host_ops->suspend(pdev, message); - langwell_otg_drv_vbus(0); - langwell_otg_phy_low_power(1); - break; - case OTG_STATE_A_SUSPEND: - langwell_otg_del_timer(a_aidl_bdis_tmr); - langwell_otg_HABA(0); - if (langwell->host_ops) - langwell->host_ops->remove(pdev); - else - otg_dbg("host driver has been removed.\n"); - langwell_otg_drv_vbus(0); - transceiver_suspend(pdev); - langwell->otg.state = OTG_STATE_A_WAIT_VFALL; - break; - case OTG_STATE_A_PERIPHERAL: - if (langwell->client_ops) - ret = langwell->client_ops->suspend(pdev, message); - else - otg_dbg("client driver has been removed.\n"); - langwell_otg_drv_vbus(0); - transceiver_suspend(pdev); - langwell->otg.state = OTG_STATE_A_WAIT_VFALL; - break; - case OTG_STATE_B_HOST: - if (langwell->host_ops) - langwell->host_ops->remove(pdev); - else - otg_dbg("host driver has been removed.\n"); - langwell->hsm.b_bus_req = 0; - transceiver_suspend(pdev); - langwell->otg.state = OTG_STATE_B_IDLE; - break; - case OTG_STATE_B_PERIPHERAL: - if (langwell->client_ops) - ret = langwell->client_ops->suspend(pdev, message); - else - otg_dbg("client driver has been removed.\n"); - break; - case OTG_STATE_B_WAIT_ACON: - langwell_otg_del_timer(b_ase0_brst_tmr); - langwell_otg_HAAR(0); - if (langwell->host_ops) - langwell->host_ops->remove(pdev); - else - otg_dbg("host driver has been removed.\n"); - langwell->hsm.b_bus_req = 0; - langwell->otg.state = OTG_STATE_B_IDLE; - transceiver_suspend(pdev); - break; - default: - otg_dbg("error state before suspend\n "); - break; - } - spin_unlock(&langwell->wq_lock); - - return ret; -} - -static void transceiver_resume(struct pci_dev *pdev) -{ - pci_restore_state(pdev); - pci_set_power_state(pdev, PCI_D0); - langwell_otg_phy_low_power(0); -} - -static int langwell_otg_resume(struct pci_dev *pdev) -{ - int ret = 0; - struct langwell_otg *langwell; - - langwell = the_transceiver; - - spin_lock(&langwell->wq_lock); - - switch (langwell->otg.state) { - case OTG_STATE_A_IDLE: - case OTG_STATE_B_IDLE: - case OTG_STATE_A_WAIT_VFALL: - case OTG_STATE_A_VBUS_ERR: - transceiver_resume(pdev); - break; - case OTG_STATE_A_WAIT_BCON: - langwell_otg_add_timer(a_wait_bcon_tmr); - langwell_otg_drv_vbus(1); - if (langwell->host_ops) - ret = langwell->host_ops->resume(pdev); - break; - case OTG_STATE_A_HOST: - langwell_otg_drv_vbus(1); - langwell_otg_phy_low_power(0); - if (langwell->host_ops) - ret = langwell->host_ops->resume(pdev); - break; - case OTG_STATE_B_PERIPHERAL: - if (langwell->client_ops) - ret = langwell->client_ops->resume(pdev); - else - otg_dbg("client driver not loaded.\n"); - break; - default: - otg_dbg("error state before suspend\n "); - break; - } - - if (request_irq(pdev->irq, otg_irq, IRQF_SHARED, - driver_name, the_transceiver) != 0) { - otg_dbg("request interrupt %d failed\n", pdev->irq); - ret = -EBUSY; - } - - /* enable OTG interrupts */ - langwell_otg_intr(1); - - spin_unlock(&langwell->wq_lock); - - queue_work(langwell->qwork, &langwell->work); - - - return ret; -} - -static int __init langwell_otg_init(void) -{ - return pci_register_driver(&otg_pci_driver); -} -module_init(langwell_otg_init); - -static void __exit langwell_otg_cleanup(void) -{ - pci_unregister_driver(&otg_pci_driver); -} -module_exit(langwell_otg_cleanup); diff --git a/include/linux/usb/langwell_otg.h b/include/linux/usb/langwell_otg.h deleted file mode 100644 index e115ae6df1da..000000000000 --- a/include/linux/usb/langwell_otg.h +++ /dev/null @@ -1,177 +0,0 @@ -/* - * Intel Langwell USB OTG transceiver driver - * Copyright (C) 2008, Intel Corporation. - * - * This program is free software; you can redistribute it and/or modify it - * under the terms and conditions of the GNU General Public License, - * version 2, as published by the Free Software Foundation. - * - * This program is distributed in the hope it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for - * more details. - * - * You should have received a copy of the GNU General Public License along with - * this program; if not, write to the Free Software Foundation, Inc., - * 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA. - * - */ - -#ifndef __LANGWELL_OTG_H__ -#define __LANGWELL_OTG_H__ - -/* notify transceiver driver about OTG events */ -extern void langwell_update_transceiver(void); -/* HCD register bus driver */ -extern int langwell_register_host(struct pci_driver *host_driver); -/* HCD unregister bus driver */ -extern void langwell_unregister_host(struct pci_driver *host_driver); -/* DCD register bus driver */ -extern int langwell_register_peripheral(struct pci_driver *client_driver); -/* DCD unregister bus driver */ -extern void langwell_unregister_peripheral(struct pci_driver *client_driver); -/* No silent failure, output warning message */ -extern void langwell_otg_nsf_msg(unsigned long message); - -#define CI_USBCMD 0x30 -# define USBCMD_RST BIT(1) -# define USBCMD_RS BIT(0) -#define CI_USBSTS 0x34 -# define USBSTS_SLI BIT(8) -# define USBSTS_URI BIT(6) -# define USBSTS_PCI BIT(2) -#define CI_PORTSC1 0x74 -# define PORTSC_PP BIT(12) -# define PORTSC_LS (BIT(11) | BIT(10)) -# define PORTSC_SUSP BIT(7) -# define PORTSC_CCS BIT(0) -#define CI_HOSTPC1 0xb4 -# define HOSTPC1_PHCD BIT(22) -#define CI_OTGSC 0xf4 -# define OTGSC_DPIE BIT(30) -# define OTGSC_1MSE BIT(29) -# define OTGSC_BSEIE BIT(28) -# define OTGSC_BSVIE BIT(27) -# define OTGSC_ASVIE BIT(26) -# define OTGSC_AVVIE BIT(25) -# define OTGSC_IDIE BIT(24) -# define OTGSC_DPIS BIT(22) -# define OTGSC_1MSS BIT(21) -# define OTGSC_BSEIS BIT(20) -# define OTGSC_BSVIS BIT(19) -# define OTGSC_ASVIS BIT(18) -# define OTGSC_AVVIS BIT(17) -# define OTGSC_IDIS BIT(16) -# define OTGSC_DPS BIT(14) -# define OTGSC_1MST BIT(13) -# define OTGSC_BSE BIT(12) -# define OTGSC_BSV BIT(11) -# define OTGSC_ASV BIT(10) -# define OTGSC_AVV BIT(9) -# define OTGSC_ID BIT(8) -# define OTGSC_HABA BIT(7) -# define OTGSC_HADP BIT(6) -# define OTGSC_IDPU BIT(5) -# define OTGSC_DP BIT(4) -# define OTGSC_OT BIT(3) -# define OTGSC_HAAR BIT(2) -# define OTGSC_VC BIT(1) -# define OTGSC_VD BIT(0) -# define OTGSC_INTEN_MASK (0x7f << 24) -# define OTGSC_INTSTS_MASK (0x7f << 16) -#define CI_USBMODE 0xf8 -# define USBMODE_CM (BIT(1) | BIT(0)) -# define USBMODE_IDLE 0 -# define USBMODE_DEVICE 0x2 -# define USBMODE_HOST 0x3 - -#define INTR_DUMMY_MASK (USBSTS_SLI | USBSTS_URI | USBSTS_PCI) - -struct otg_hsm { - /* Input */ - int a_bus_resume; - int a_bus_suspend; - int a_conn; - int a_sess_vld; - int a_srp_det; - int a_vbus_vld; - int b_bus_resume; - int b_bus_suspend; - int b_conn; - int b_se0_srp; - int b_sess_end; - int b_sess_vld; - int id; - - /* Internal variables */ - int a_set_b_hnp_en; - int b_srp_done; - int b_hnp_enable; - - /* Timeout indicator for timers */ - int a_wait_vrise_tmout; - int a_wait_bcon_tmout; - int a_aidl_bdis_tmout; - int b_ase0_brst_tmout; - int b_bus_suspend_tmout; - int b_srp_res_tmout; - - /* Informative variables */ - int a_bus_drop; - int a_bus_req; - int a_clr_err; - int a_suspend_req; - int b_bus_req; - - /* Output */ - int drv_vbus; - int loc_conn; - int loc_sof; - - /* Others */ - int b_bus_suspend_vld; -}; - -#define TA_WAIT_VRISE 100 -#define TA_WAIT_BCON 30000 -#define TA_AIDL_BDIS 15000 -#define TB_ASE0_BRST 5000 -#define TB_SE0_SRP 2 -#define TB_SRP_RES 100 -#define TB_BUS_SUSPEND 500 - -struct langwell_otg_timer { - unsigned long expires; /* Number of count increase to timeout */ - unsigned long count; /* Tick counter */ - void (*function)(unsigned long); /* Timeout function */ - unsigned long data; /* Data passed to function */ - struct list_head list; -}; - -struct langwell_otg { - struct otg_transceiver otg; - struct otg_hsm hsm; - void __iomem *regs; - unsigned region; - struct pci_driver *host_ops; - struct pci_driver *client_ops; - struct pci_dev *pdev; - struct work_struct work; - struct workqueue_struct *qwork; - spinlock_t lock; - spinlock_t wq_lock; -}; - -static inline struct langwell_otg *otg_to_langwell(struct otg_transceiver *otg) -{ - return container_of(otg, struct langwell_otg, otg); -} - -#ifdef DEBUG -#define otg_dbg(fmt, args...) \ - printk(KERN_DEBUG fmt , ## args) -#else -#define otg_dbg(fmt, args...) \ - do { } while (0) -#endif /* DEBUG */ -#endif /* __LANGWELL_OTG_H__ */ -- cgit v1.2.3 From e376bbbb6a82cf119c93bde66937f66c72cba27b Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Thu, 18 Jun 2009 10:39:11 -0700 Subject: USB: usb.h: fix kernel-doc notation Fix usb.h kernel-doc warnings: Warning(include/linux/usb.h:918): Excess struct/union/enum/typedef member 'nodename' description in 'usb_device_driver' Warning(include/linux/usb.h:939): No description found for parameter 'nodename' Warning(include/linux/usb.h:1219): No description found for parameter 'sg' Warning(include/linux/usb.h:1219): No description found for parameter 'num_sgs' Signed-off-by: Randy Dunlap Signed-off-by: Greg Kroah-Hartman --- include/linux/usb.h | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/include/linux/usb.h b/include/linux/usb.h index 84929e914034..b1e3c2fbfe11 100644 --- a/include/linux/usb.h +++ b/include/linux/usb.h @@ -888,8 +888,6 @@ struct usb_driver { * struct usb_device_driver - identifies USB device driver to usbcore * @name: The driver name should be unique among USB drivers, * and should normally be the same as the module name. - * @nodename: Callback to provide a naming hint for a possible - * device node to create. * @probe: Called to see if the driver is willing to manage a particular * device. If it is, probe returns zero and uses dev_set_drvdata() * to associate driver-specific data with the device. If unwilling @@ -924,6 +922,8 @@ extern struct bus_type usb_bus_type; /** * struct usb_class_driver - identifies a USB driver that wants to use the USB major number * @name: the usb class device name for this driver. Will show up in sysfs. + * @nodename: Callback to provide a naming hint for a possible + * device node to create. * @fops: pointer to the struct file_operations of this driver. * @minor_base: the start of the minor range for this driver. * @@ -1046,6 +1046,8 @@ typedef void (*usb_complete_t)(struct urb *); * the device driver is saying that it provided this DMA address, * which the host controller driver should use in preference to the * transfer_buffer. + * @sg: scatter gather buffer list + * @num_sgs: number of entries in the sg list * @transfer_buffer_length: How big is transfer_buffer. The transfer may * be broken up into chunks according to the current maximum packet * size for the endpoint, which is a function of the configuration -- cgit v1.2.3 From 440c1ce178d6a6743e02d136a55b2de3f6d62a20 Mon Sep 17 00:00:00 2001 From: Huang Weiyi Date: Fri, 10 Jul 2009 15:33:49 +0000 Subject: dropmon: remove duplicated #include Remove duplicated #include('s) in include/linux/net_dropmon.h Signed-off-by: Huang Weiyi Signed-off-by: David S. Miller --- include/linux/net_dropmon.h | 1 - 1 file changed, 1 deletion(-) (limited to 'include') diff --git a/include/linux/net_dropmon.h b/include/linux/net_dropmon.h index 3ceb0cc1bc78..2a739462caeb 100644 --- a/include/linux/net_dropmon.h +++ b/include/linux/net_dropmon.h @@ -3,7 +3,6 @@ #include #include -#include struct net_dm_drop_point { __u8 pc[8]; -- cgit v1.2.3 From d0b6e04a4cd8360e3c9c419f7c30a3081a0c142a Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Mon, 13 Jul 2009 10:33:21 +0800 Subject: tracing/events: Move TRACE_SYSTEM outside of include guard If TRACE_INCLDUE_FILE is defined, will be included and compiled, otherwise it will be So TRACE_SYSTEM should be defined outside of #if proctection, just like TRACE_INCLUDE_FILE. Imaging this scenario: #include -> TRACE_SYSTEM == foo ... #include -> TRACE_SYSTEM == bar ... #define CREATE_TRACE_POINTS #include -> TRACE_SYSTEM == bar !!! and then bar.h will be included and compiled. Signed-off-by: Li Zefan Cc: Steven Rostedt Cc: Frederic Weisbecker LKML-Reference: <4A5A9CF1.2010007@cn.fujitsu.com> Signed-off-by: Ingo Molnar --- fs/gfs2/trace_gfs2.h | 8 +++---- include/trace/events/block.h | 6 ++--- include/trace/events/ext4.h | 6 ++--- include/trace/events/irq.h | 6 ++--- include/trace/events/jbd2.h | 6 ++--- include/trace/events/kmem.h | 6 ++--- include/trace/events/lockdep.h | 6 ++--- include/trace/events/sched.h | 6 ++--- include/trace/events/skb.h | 6 ++--- include/trace/events/workqueue.h | 6 ++--- samples/trace_events/trace-events-sample.h | 37 ++++++++++++++++-------------- 11 files changed, 51 insertions(+), 48 deletions(-) (limited to 'include') diff --git a/fs/gfs2/trace_gfs2.h b/fs/gfs2/trace_gfs2.h index 98d6ef1c1dc0..148d55c14171 100644 --- a/fs/gfs2/trace_gfs2.h +++ b/fs/gfs2/trace_gfs2.h @@ -1,12 +1,11 @@ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM gfs2 + #if !defined(_TRACE_GFS2_H) || defined(TRACE_HEADER_MULTI_READ) #define _TRACE_GFS2_H #include -#undef TRACE_SYSTEM -#define TRACE_SYSTEM gfs2 -#define TRACE_INCLUDE_FILE trace_gfs2 - #include #include #include @@ -403,5 +402,6 @@ TRACE_EVENT(gfs2_block_alloc, /* This part must be outside protection */ #undef TRACE_INCLUDE_PATH #define TRACE_INCLUDE_PATH . +#define TRACE_INCLUDE_FILE trace_gfs2 #include diff --git a/include/trace/events/block.h b/include/trace/events/block.h index d6b05f42dd44..9a74b468a229 100644 --- a/include/trace/events/block.h +++ b/include/trace/events/block.h @@ -1,3 +1,6 @@ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM block + #if !defined(_TRACE_BLOCK_H) || defined(TRACE_HEADER_MULTI_READ) #define _TRACE_BLOCK_H @@ -5,9 +8,6 @@ #include #include -#undef TRACE_SYSTEM -#define TRACE_SYSTEM block - TRACE_EVENT(block_rq_abort, TP_PROTO(struct request_queue *q, struct request *rq), diff --git a/include/trace/events/ext4.h b/include/trace/events/ext4.h index acf4cc9cd36d..8eb82cc8d149 100644 --- a/include/trace/events/ext4.h +++ b/include/trace/events/ext4.h @@ -1,9 +1,9 @@ -#if !defined(_TRACE_EXT4_H) || defined(TRACE_HEADER_MULTI_READ) -#define _TRACE_EXT4_H - #undef TRACE_SYSTEM #define TRACE_SYSTEM ext4 +#if !defined(_TRACE_EXT4_H) || defined(TRACE_HEADER_MULTI_READ) +#define _TRACE_EXT4_H + #include #include "../../../fs/ext4/ext4.h" #include "../../../fs/ext4/mballoc.h" diff --git a/include/trace/events/irq.h b/include/trace/events/irq.h index b0c7ede55eb1..1cb0c3aa11e6 100644 --- a/include/trace/events/irq.h +++ b/include/trace/events/irq.h @@ -1,12 +1,12 @@ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM irq + #if !defined(_TRACE_IRQ_H) || defined(TRACE_HEADER_MULTI_READ) #define _TRACE_IRQ_H #include #include -#undef TRACE_SYSTEM -#define TRACE_SYSTEM irq - #define softirq_name(sirq) { sirq##_SOFTIRQ, #sirq } #define show_softirq_name(val) \ __print_symbolic(val, \ diff --git a/include/trace/events/jbd2.h b/include/trace/events/jbd2.h index 845b0b4b48fd..10813fa0c8d0 100644 --- a/include/trace/events/jbd2.h +++ b/include/trace/events/jbd2.h @@ -1,12 +1,12 @@ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM jbd2 + #if !defined(_TRACE_JBD2_H) || defined(TRACE_HEADER_MULTI_READ) #define _TRACE_JBD2_H #include #include -#undef TRACE_SYSTEM -#define TRACE_SYSTEM jbd2 - TRACE_EVENT(jbd2_checkpoint, TP_PROTO(journal_t *journal, int result), diff --git a/include/trace/events/kmem.h b/include/trace/events/kmem.h index 9baba50d6512..1493c541f9c4 100644 --- a/include/trace/events/kmem.h +++ b/include/trace/events/kmem.h @@ -1,12 +1,12 @@ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM kmem + #if !defined(_TRACE_KMEM_H) || defined(TRACE_HEADER_MULTI_READ) #define _TRACE_KMEM_H #include #include -#undef TRACE_SYSTEM -#define TRACE_SYSTEM kmem - /* * The order of these masks is important. Matching masks will be seen * first and the left over flags will end up showing by themselves. diff --git a/include/trace/events/lockdep.h b/include/trace/events/lockdep.h index 0e956c9dfd7e..bcf1d209a00d 100644 --- a/include/trace/events/lockdep.h +++ b/include/trace/events/lockdep.h @@ -1,12 +1,12 @@ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM lockdep + #if !defined(_TRACE_LOCKDEP_H) || defined(TRACE_HEADER_MULTI_READ) #define _TRACE_LOCKDEP_H #include #include -#undef TRACE_SYSTEM -#define TRACE_SYSTEM lockdep - #ifdef CONFIG_LOCKDEP TRACE_EVENT(lock_acquire, diff --git a/include/trace/events/sched.h b/include/trace/events/sched.h index 24ab5bcff7b2..8949bb7eb082 100644 --- a/include/trace/events/sched.h +++ b/include/trace/events/sched.h @@ -1,12 +1,12 @@ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM sched + #if !defined(_TRACE_SCHED_H) || defined(TRACE_HEADER_MULTI_READ) #define _TRACE_SCHED_H #include #include -#undef TRACE_SYSTEM -#define TRACE_SYSTEM sched - /* * Tracepoint for calling kthread_stop, performed to end a kthread: */ diff --git a/include/trace/events/skb.h b/include/trace/events/skb.h index 1e8fabb57c06..e499863b9669 100644 --- a/include/trace/events/skb.h +++ b/include/trace/events/skb.h @@ -1,12 +1,12 @@ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM skb + #if !defined(_TRACE_SKB_H) || defined(TRACE_HEADER_MULTI_READ) #define _TRACE_SKB_H #include #include -#undef TRACE_SYSTEM -#define TRACE_SYSTEM skb - /* * Tracepoint for free an sk_buff: */ diff --git a/include/trace/events/workqueue.h b/include/trace/events/workqueue.h index 035f1bff288e..fcfd9a1e4b96 100644 --- a/include/trace/events/workqueue.h +++ b/include/trace/events/workqueue.h @@ -1,3 +1,6 @@ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM workqueue + #if !defined(_TRACE_WORKQUEUE_H) || defined(TRACE_HEADER_MULTI_READ) #define _TRACE_WORKQUEUE_H @@ -5,9 +8,6 @@ #include #include -#undef TRACE_SYSTEM -#define TRACE_SYSTEM workqueue - TRACE_EVENT(workqueue_insertion, TP_PROTO(struct task_struct *wq_thread, struct work_struct *work), diff --git a/samples/trace_events/trace-events-sample.h b/samples/trace_events/trace-events-sample.h index 9977a756fb32..f24ae370e514 100644 --- a/samples/trace_events/trace-events-sample.h +++ b/samples/trace_events/trace-events-sample.h @@ -1,20 +1,3 @@ -/* - * Notice that this file is not protected like a normal header. - * We also must allow for rereading of this file. The - * - * || defined(TRACE_HEADER_MULTI_READ) - * - * serves this purpose. - */ -#if !defined(_TRACE_EVENT_SAMPLE_H) || defined(TRACE_HEADER_MULTI_READ) -#define _TRACE_EVENT_SAMPLE_H - -/* - * All trace headers should include tracepoint.h, until we finally - * make it into a standard header. - */ -#include - /* * If TRACE_SYSTEM is defined, that will be the directory created * in the ftrace directory under /debugfs/tracing/events/ @@ -34,10 +17,30 @@ * #define TRACE_INCLUDE_FILE trace-events-sample * * As we do an the bottom of this file. + * + * Notice that TRACE_SYSTEM should be defined outside of #if + * protection, just like TRACE_INCLUDE_FILE. */ #undef TRACE_SYSTEM #define TRACE_SYSTEM sample +/* + * Notice that this file is not protected like a normal header. + * We also must allow for rereading of this file. The + * + * || defined(TRACE_HEADER_MULTI_READ) + * + * serves this purpose. + */ +#if !defined(_TRACE_EVENT_SAMPLE_H) || defined(TRACE_HEADER_MULTI_READ) +#define _TRACE_EVENT_SAMPLE_H + +/* + * All trace headers should include tracepoint.h, until we finally + * make it into a standard header. + */ +#include + /* * The TRACE_EVENT macro is broken up into 5 parts. * -- cgit v1.2.3 From 113adefc73c291f93f875fe515a46d8f76252fff Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Tue, 14 Jul 2009 12:50:12 +0800 Subject: crypto: shash - Make descsize a run-time attribute This patch changes descsize to a run-time attribute so that implementations can change it in their init functions. Signed-off-by: Herbert Xu --- crypto/shash.c | 39 ++++++++++++++++++++++++++++----------- include/crypto/hash.h | 3 ++- include/crypto/internal/hash.h | 2 +- 3 files changed, 31 insertions(+), 13 deletions(-) (limited to 'include') diff --git a/crypto/shash.c b/crypto/shash.c index 131e14d2b572..d8e0f8f8d672 100644 --- a/crypto/shash.c +++ b/crypto/shash.c @@ -300,14 +300,16 @@ static int crypto_init_shash_ops_async(struct crypto_tfm *tfm) static int shash_compat_setkey(struct crypto_hash *tfm, const u8 *key, unsigned int keylen) { - struct shash_desc *desc = crypto_hash_ctx(tfm); + struct shash_desc **descp = crypto_hash_ctx(tfm); + struct shash_desc *desc = *descp; return crypto_shash_setkey(desc->tfm, key, keylen); } static int shash_compat_init(struct hash_desc *hdesc) { - struct shash_desc *desc = crypto_hash_ctx(hdesc->tfm); + struct shash_desc **descp = crypto_hash_ctx(hdesc->tfm); + struct shash_desc *desc = *descp; desc->flags = hdesc->flags; @@ -317,7 +319,8 @@ static int shash_compat_init(struct hash_desc *hdesc) static int shash_compat_update(struct hash_desc *hdesc, struct scatterlist *sg, unsigned int len) { - struct shash_desc *desc = crypto_hash_ctx(hdesc->tfm); + struct shash_desc **descp = crypto_hash_ctx(hdesc->tfm); + struct shash_desc *desc = *descp; struct crypto_hash_walk walk; int nbytes; @@ -330,7 +333,9 @@ static int shash_compat_update(struct hash_desc *hdesc, struct scatterlist *sg, static int shash_compat_final(struct hash_desc *hdesc, u8 *out) { - return crypto_shash_final(crypto_hash_ctx(hdesc->tfm), out); + struct shash_desc **descp = crypto_hash_ctx(hdesc->tfm); + + return crypto_shash_final(*descp, out); } static int shash_compat_digest(struct hash_desc *hdesc, struct scatterlist *sg, @@ -340,7 +345,8 @@ static int shash_compat_digest(struct hash_desc *hdesc, struct scatterlist *sg, int err; if (nbytes < min(sg->length, ((unsigned int)(PAGE_SIZE)) - offset)) { - struct shash_desc *desc = crypto_hash_ctx(hdesc->tfm); + struct shash_desc **descp = crypto_hash_ctx(hdesc->tfm); + struct shash_desc *desc = *descp; void *data; desc->flags = hdesc->flags; @@ -368,9 +374,11 @@ out: static void crypto_exit_shash_ops_compat(struct crypto_tfm *tfm) { - struct shash_desc *desc= crypto_tfm_ctx(tfm); + struct shash_desc **descp = crypto_tfm_ctx(tfm); + struct shash_desc *desc = *descp; crypto_free_shash(desc->tfm); + kzfree(desc); } static int crypto_init_shash_ops_compat(struct crypto_tfm *tfm) @@ -378,8 +386,9 @@ static int crypto_init_shash_ops_compat(struct crypto_tfm *tfm) struct hash_tfm *crt = &tfm->crt_hash; struct crypto_alg *calg = tfm->__crt_alg; struct shash_alg *alg = __crypto_shash_alg(calg); - struct shash_desc *desc = crypto_tfm_ctx(tfm); + struct shash_desc **descp = crypto_tfm_ctx(tfm); struct crypto_shash *shash; + struct shash_desc *desc; if (!crypto_mod_get(calg)) return -EAGAIN; @@ -390,6 +399,14 @@ static int crypto_init_shash_ops_compat(struct crypto_tfm *tfm) return PTR_ERR(shash); } + desc = kmalloc(sizeof(*desc) + crypto_shash_descsize(shash), + GFP_KERNEL); + if (!desc) { + crypto_free_shash(shash); + return -ENOMEM; + } + + *descp = desc; desc->tfm = shash; tfm->exit = crypto_exit_shash_ops_compat; @@ -419,11 +436,9 @@ static int crypto_init_shash_ops(struct crypto_tfm *tfm, u32 type, u32 mask) static unsigned int crypto_shash_ctxsize(struct crypto_alg *alg, u32 type, u32 mask) { - struct shash_alg *salg = __crypto_shash_alg(alg); - switch (mask & CRYPTO_ALG_TYPE_MASK) { case CRYPTO_ALG_TYPE_HASH_MASK: - return sizeof(struct shash_desc) + salg->descsize; + return sizeof(struct shash_desc *); case CRYPTO_ALG_TYPE_AHASH_MASK: return sizeof(struct crypto_shash *); } @@ -434,6 +449,9 @@ static unsigned int crypto_shash_ctxsize(struct crypto_alg *alg, u32 type, static int crypto_shash_init_tfm(struct crypto_tfm *tfm, const struct crypto_type *frontend) { + struct crypto_shash *hash = __crypto_shash_cast(tfm); + + hash->descsize = crypto_shash_alg(hash)->descsize; return 0; } @@ -452,7 +470,6 @@ static void crypto_shash_show(struct seq_file *m, struct crypto_alg *alg) seq_printf(m, "type : shash\n"); seq_printf(m, "blocksize : %u\n", alg->cra_blocksize); seq_printf(m, "digestsize : %u\n", salg->digestsize); - seq_printf(m, "descsize : %u\n", salg->descsize); } static const struct crypto_type crypto_shash_type = { diff --git a/include/crypto/hash.h b/include/crypto/hash.h index f74214a4b01b..fcc02d978231 100644 --- a/include/crypto/hash.h +++ b/include/crypto/hash.h @@ -48,6 +48,7 @@ struct crypto_ahash { }; struct crypto_shash { + unsigned int descsize; struct crypto_tfm base; }; @@ -275,7 +276,7 @@ static inline void crypto_shash_clear_flags(struct crypto_shash *tfm, u32 flags) static inline unsigned int crypto_shash_descsize(struct crypto_shash *tfm) { - return crypto_shash_alg(tfm)->descsize; + return tfm->descsize; } static inline void *shash_desc_ctx(struct shash_desc *desc) diff --git a/include/crypto/internal/hash.h b/include/crypto/internal/hash.h index 069b93e1a8ec..2d1b10f23c91 100644 --- a/include/crypto/internal/hash.h +++ b/include/crypto/internal/hash.h @@ -135,7 +135,7 @@ static inline void *crypto_shash_ctx_aligned(struct crypto_shash *tfm) static inline struct crypto_shash *__crypto_shash_cast(struct crypto_tfm *tfm) { - return (struct crypto_shash *)tfm; + return container_of(tfm, struct crypto_shash, base); } #endif /* _CRYPTO_INTERNAL_HASH_H */ -- cgit v1.2.3 From 7eddf95ec5440d60f10963f453e27f82f394044e Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Sun, 12 Jul 2009 21:25:20 +0800 Subject: crypto: shash - Export async functions This patch exports the async functions so that they can be reused by cryptd when it switches over to using shash. Signed-off-by: Herbert Xu --- crypto/shash.c | 42 ++++++++++++++++++++++-------------------- include/crypto/internal/hash.h | 3 +++ 2 files changed, 25 insertions(+), 20 deletions(-) (limited to 'include') diff --git a/crypto/shash.c b/crypto/shash.c index d8e0f8f8d672..3b1b06b3dcf3 100644 --- a/crypto/shash.c +++ b/crypto/shash.c @@ -202,9 +202,8 @@ static int shash_async_init(struct ahash_request *req) return crypto_shash_init(desc); } -static int shash_async_update(struct ahash_request *req) +int shash_ahash_update(struct ahash_request *req, struct shash_desc *desc) { - struct shash_desc *desc = ahash_request_ctx(req); struct crypto_hash_walk walk; int nbytes; @@ -214,13 +213,19 @@ static int shash_async_update(struct ahash_request *req) return nbytes; } +EXPORT_SYMBOL_GPL(shash_ahash_update); + +static int shash_async_update(struct ahash_request *req) +{ + return shash_ahash_update(req, ahash_request_ctx(req)); +} static int shash_async_final(struct ahash_request *req) { return crypto_shash_final(ahash_request_ctx(req), req->result); } -static int shash_async_digest(struct ahash_request *req) +int shash_ahash_digest(struct ahash_request *req, struct shash_desc *desc) { struct scatterlist *sg = req->src; unsigned int offset = sg->offset; @@ -228,34 +233,31 @@ static int shash_async_digest(struct ahash_request *req) int err; if (nbytes < min(sg->length, ((unsigned int)(PAGE_SIZE)) - offset)) { - struct crypto_shash **ctx = - crypto_ahash_ctx(crypto_ahash_reqtfm(req)); - struct shash_desc *desc = ahash_request_ctx(req); void *data; - desc->tfm = *ctx; - desc->flags = req->base.flags; - data = crypto_kmap(sg_page(sg), 0); err = crypto_shash_digest(desc, data + offset, nbytes, req->result); crypto_kunmap(data, 0); crypto_yield(desc->flags); - goto out; - } + } else + err = crypto_shash_init(desc) ?: + shash_ahash_update(req, desc) ?: + crypto_shash_final(desc, req->result); - err = shash_async_init(req); - if (err) - goto out; + return err; +} +EXPORT_SYMBOL_GPL(shash_ahash_digest); - err = shash_async_update(req); - if (err) - goto out; +static int shash_async_digest(struct ahash_request *req) +{ + struct crypto_shash **ctx = crypto_ahash_ctx(crypto_ahash_reqtfm(req)); + struct shash_desc *desc = ahash_request_ctx(req); - err = shash_async_final(req); + desc->tfm = *ctx; + desc->flags = req->base.flags; -out: - return err; + return shash_ahash_digest(req, desc); } static void crypto_exit_shash_ops_async(struct crypto_tfm *tfm) diff --git a/include/crypto/internal/hash.h b/include/crypto/internal/hash.h index 2d1b10f23c91..6e283495374e 100644 --- a/include/crypto/internal/hash.h +++ b/include/crypto/internal/hash.h @@ -63,6 +63,9 @@ int crypto_init_shash_spawn(struct crypto_shash_spawn *spawn, struct shash_alg *shash_attr_alg(struct rtattr *rta, u32 type, u32 mask); +int shash_ahash_update(struct ahash_request *req, struct shash_desc *desc); +int shash_ahash_digest(struct ahash_request *req, struct shash_desc *desc); + static inline void *crypto_ahash_ctx(struct crypto_ahash *tfm) { return crypto_tfm_ctx(&tfm->base); -- cgit v1.2.3 From fc00127fb67b2a7d2b66f0f4096a5367b581f045 Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Sun, 12 Jul 2009 23:05:48 +0800 Subject: crypto: ahash - Add crypto_ahash_set_reqsize This patch adds the helper crypto_ahash_set_reqsize so that implementations do not directly access the crypto_ahash structure. Signed-off-by: Herbert Xu --- include/crypto/internal/hash.h | 6 ++++++ 1 file changed, 6 insertions(+) (limited to 'include') diff --git a/include/crypto/internal/hash.h b/include/crypto/internal/hash.h index 6e283495374e..5e45818f3351 100644 --- a/include/crypto/internal/hash.h +++ b/include/crypto/internal/hash.h @@ -77,6 +77,12 @@ static inline struct ahash_alg *crypto_ahash_alg( return &crypto_ahash_tfm(tfm)->__crt_alg->cra_ahash; } +static inline void crypto_ahash_set_reqsize(struct crypto_ahash *tfm, + unsigned int reqsize) +{ + crypto_ahash_crt(tfm)->reqsize = reqsize; +} + static inline int ahash_enqueue_request(struct crypto_queue *queue, struct ahash_request *request) { -- cgit v1.2.3 From 2ca33da1dea3ba53d1425226a6bac073c5e8568c Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Mon, 13 Jul 2009 20:46:25 +0800 Subject: crypto: api - Remove frontend argument from extsize/init_tfm As the extsize and init_tfm functions belong to the frontend the frontend argument is superfluous. Signed-off-by: Herbert Xu --- crypto/api.c | 4 ++-- crypto/pcompress.c | 6 ++---- crypto/shash.c | 6 ++---- include/crypto/algapi.h | 6 ++---- 4 files changed, 8 insertions(+), 14 deletions(-) (limited to 'include') diff --git a/crypto/api.c b/crypto/api.c index c8ac18f7ac1e..798526d90538 100644 --- a/crypto/api.c +++ b/crypto/api.c @@ -457,7 +457,7 @@ void *crypto_create_tfm(struct crypto_alg *alg, int err = -ENOMEM; tfmsize = frontend->tfmsize; - total = tfmsize + sizeof(*tfm) + frontend->extsize(alg, frontend); + total = tfmsize + sizeof(*tfm) + frontend->extsize(alg); mem = kzalloc(total, GFP_KERNEL); if (mem == NULL) @@ -466,7 +466,7 @@ void *crypto_create_tfm(struct crypto_alg *alg, tfm = (struct crypto_tfm *)(mem + tfmsize); tfm->__crt_alg = alg; - err = frontend->init_tfm(tfm, frontend); + err = frontend->init_tfm(tfm); if (err) goto out_free_tfm; diff --git a/crypto/pcompress.c b/crypto/pcompress.c index bcadc03726b7..f7c4a7d7412e 100644 --- a/crypto/pcompress.c +++ b/crypto/pcompress.c @@ -36,14 +36,12 @@ static int crypto_pcomp_init(struct crypto_tfm *tfm, u32 type, u32 mask) return 0; } -static unsigned int crypto_pcomp_extsize(struct crypto_alg *alg, - const struct crypto_type *frontend) +static unsigned int crypto_pcomp_extsize(struct crypto_alg *alg) { return alg->cra_ctxsize; } -static int crypto_pcomp_init_tfm(struct crypto_tfm *tfm, - const struct crypto_type *frontend) +static int crypto_pcomp_init_tfm(struct crypto_tfm *tfm) { return 0; } diff --git a/crypto/shash.c b/crypto/shash.c index 3b1b06b3dcf3..7063e1421504 100644 --- a/crypto/shash.c +++ b/crypto/shash.c @@ -448,8 +448,7 @@ static unsigned int crypto_shash_ctxsize(struct crypto_alg *alg, u32 type, return 0; } -static int crypto_shash_init_tfm(struct crypto_tfm *tfm, - const struct crypto_type *frontend) +static int crypto_shash_init_tfm(struct crypto_tfm *tfm) { struct crypto_shash *hash = __crypto_shash_cast(tfm); @@ -457,8 +456,7 @@ static int crypto_shash_init_tfm(struct crypto_tfm *tfm, return 0; } -static unsigned int crypto_shash_extsize(struct crypto_alg *alg, - const struct crypto_type *frontend) +static unsigned int crypto_shash_extsize(struct crypto_alg *alg) { return alg->cra_ctxsize; } diff --git a/include/crypto/algapi.h b/include/crypto/algapi.h index 1d15a926041e..7635fde7b1a2 100644 --- a/include/crypto/algapi.h +++ b/include/crypto/algapi.h @@ -22,11 +22,9 @@ struct seq_file; struct crypto_type { unsigned int (*ctxsize)(struct crypto_alg *alg, u32 type, u32 mask); - unsigned int (*extsize)(struct crypto_alg *alg, - const struct crypto_type *frontend); + unsigned int (*extsize)(struct crypto_alg *alg); int (*init)(struct crypto_tfm *tfm, u32 type, u32 mask); - int (*init_tfm)(struct crypto_tfm *tfm, - const struct crypto_type *frontend); + int (*init_tfm)(struct crypto_tfm *tfm); void (*show)(struct seq_file *m, struct crypto_alg *alg); struct crypto_alg *(*lookup)(const char *name, u32 type, u32 mask); -- cgit v1.2.3 From 88056ec346ccf41f63dbc7080b24b5fd19d1358d Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Tue, 14 Jul 2009 12:28:26 +0800 Subject: crypto: ahash - Convert to new style algorithms This patch converts crypto_ahash to the new style. The old ahash algorithm type is retained until the existing ahash implementations are also converted. All ahash users will automatically get the new crypto_ahash type. Signed-off-by: Herbert Xu --- crypto/ahash.c | 82 +++++++++++++++++++++---------- crypto/shash.c | 8 +-- include/crypto/hash.h | 109 +++++++++++++++++++++++++++++------------ include/crypto/internal/hash.h | 11 +++-- include/linux/crypto.h | 29 ++--------- 5 files changed, 148 insertions(+), 91 deletions(-) (limited to 'include') diff --git a/crypto/ahash.c b/crypto/ahash.c index f3476374f764..838519386215 100644 --- a/crypto/ahash.c +++ b/crypto/ahash.c @@ -24,6 +24,12 @@ #include "internal.h" +static inline struct ahash_alg *crypto_ahash_alg(struct crypto_ahash *hash) +{ + return container_of(crypto_hash_alg_common(hash), struct ahash_alg, + halg); +} + static int hash_walk_next(struct crypto_hash_walk *walk) { unsigned int alignmask = walk->alignmask; @@ -169,30 +175,11 @@ static int ahash_nosetkey(struct crypto_ahash *tfm, const u8 *key, return -ENOSYS; } -int crypto_ahash_import(struct ahash_request *req, const u8 *in) -{ - struct crypto_ahash *tfm = crypto_ahash_reqtfm(req); - struct ahash_alg *alg = crypto_ahash_alg(tfm); - - memcpy(ahash_request_ctx(req), in, crypto_ahash_reqsize(tfm)); - - if (alg->reinit) - alg->reinit(req); - - return 0; -} -EXPORT_SYMBOL_GPL(crypto_ahash_import); - -static unsigned int crypto_ahash_ctxsize(struct crypto_alg *alg, u32 type, - u32 mask) -{ - return alg->cra_ctxsize; -} - static int crypto_init_ahash_ops(struct crypto_tfm *tfm, u32 type, u32 mask) { - struct ahash_alg *alg = &tfm->__crt_alg->cra_ahash; - struct ahash_tfm *crt = &tfm->crt_ahash; + struct old_ahash_alg *alg = &tfm->__crt_alg->cra_ahash; + struct crypto_ahash *crt = __crypto_ahash_cast(tfm); + struct ahash_alg *nalg = crypto_ahash_alg(crt); if (alg->digestsize > PAGE_SIZE / 8) return -EINVAL; @@ -204,9 +191,42 @@ static int crypto_init_ahash_ops(struct crypto_tfm *tfm, u32 type, u32 mask) crt->setkey = alg->setkey ? ahash_setkey : ahash_nosetkey; crt->digestsize = alg->digestsize; + nalg->setkey = alg->setkey; + nalg->halg.digestsize = alg->digestsize; + return 0; } +static int crypto_ahash_init_tfm(struct crypto_tfm *tfm) +{ + struct crypto_ahash *hash = __crypto_ahash_cast(tfm); + struct ahash_alg *alg = crypto_ahash_alg(hash); + struct old_ahash_alg *oalg = crypto_old_ahash_alg(hash); + + if (tfm->__crt_alg->cra_type != &crypto_ahash_type) + return crypto_init_shash_ops_async(tfm); + + if (oalg->init) + return crypto_init_ahash_ops(tfm, 0, 0); + + hash->init = alg->init; + hash->update = alg->update; + hash->final = alg->final; + hash->digest = alg->digest; + hash->setkey = alg->setkey ? ahash_setkey : ahash_nosetkey; + hash->digestsize = alg->halg.digestsize; + + return 0; +} + +static unsigned int crypto_ahash_extsize(struct crypto_alg *alg) +{ + if (alg->cra_type == &crypto_ahash_type) + return alg->cra_ctxsize; + + return sizeof(struct crypto_shash *); +} + static void crypto_ahash_show(struct seq_file *m, struct crypto_alg *alg) __attribute__ ((unused)); static void crypto_ahash_show(struct seq_file *m, struct crypto_alg *alg) @@ -215,17 +235,29 @@ static void crypto_ahash_show(struct seq_file *m, struct crypto_alg *alg) seq_printf(m, "async : %s\n", alg->cra_flags & CRYPTO_ALG_ASYNC ? "yes" : "no"); seq_printf(m, "blocksize : %u\n", alg->cra_blocksize); - seq_printf(m, "digestsize : %u\n", alg->cra_ahash.digestsize); + seq_printf(m, "digestsize : %u\n", + __crypto_hash_alg_common(alg)->digestsize); } const struct crypto_type crypto_ahash_type = { - .ctxsize = crypto_ahash_ctxsize, - .init = crypto_init_ahash_ops, + .extsize = crypto_ahash_extsize, + .init_tfm = crypto_ahash_init_tfm, #ifdef CONFIG_PROC_FS .show = crypto_ahash_show, #endif + .maskclear = ~CRYPTO_ALG_TYPE_MASK, + .maskset = CRYPTO_ALG_TYPE_AHASH_MASK, + .type = CRYPTO_ALG_TYPE_AHASH, + .tfmsize = offsetof(struct crypto_ahash, base), }; EXPORT_SYMBOL_GPL(crypto_ahash_type); +struct crypto_ahash *crypto_alloc_ahash(const char *alg_name, u32 type, + u32 mask) +{ + return crypto_alloc_tfm(alg_name, &crypto_ahash_type, type, mask); +} +EXPORT_SYMBOL_GPL(crypto_alloc_ahash); + MODULE_LICENSE("GPL"); MODULE_DESCRIPTION("Asynchronous cryptographic hash type"); diff --git a/crypto/shash.c b/crypto/shash.c index 7063e1421504..615a5f4d9b7a 100644 --- a/crypto/shash.c +++ b/crypto/shash.c @@ -267,11 +267,11 @@ static void crypto_exit_shash_ops_async(struct crypto_tfm *tfm) crypto_free_shash(*ctx); } -static int crypto_init_shash_ops_async(struct crypto_tfm *tfm) +int crypto_init_shash_ops_async(struct crypto_tfm *tfm) { struct crypto_alg *calg = tfm->__crt_alg; struct shash_alg *alg = __crypto_shash_alg(calg); - struct ahash_tfm *crt = &tfm->crt_ahash; + struct crypto_ahash *crt = __crypto_ahash_cast(tfm); struct crypto_shash **ctx = crypto_tfm_ctx(tfm); struct crypto_shash *shash; @@ -428,8 +428,6 @@ static int crypto_init_shash_ops(struct crypto_tfm *tfm, u32 type, u32 mask) switch (mask & CRYPTO_ALG_TYPE_MASK) { case CRYPTO_ALG_TYPE_HASH_MASK: return crypto_init_shash_ops_compat(tfm); - case CRYPTO_ALG_TYPE_AHASH_MASK: - return crypto_init_shash_ops_async(tfm); } return -EINVAL; @@ -441,8 +439,6 @@ static unsigned int crypto_shash_ctxsize(struct crypto_alg *alg, u32 type, switch (mask & CRYPTO_ALG_TYPE_MASK) { case CRYPTO_ALG_TYPE_HASH_MASK: return sizeof(struct shash_desc *); - case CRYPTO_ALG_TYPE_AHASH_MASK: - return sizeof(struct crypto_shash *); } return 0; diff --git a/include/crypto/hash.h b/include/crypto/hash.h index fcc02d978231..262861d8f0cb 100644 --- a/include/crypto/hash.h +++ b/include/crypto/hash.h @@ -15,6 +15,39 @@ #include +struct crypto_ahash; + +struct hash_alg_common { + unsigned int digestsize; + unsigned int statesize; + + struct crypto_alg base; +}; + +struct ahash_request { + struct crypto_async_request base; + + unsigned int nbytes; + struct scatterlist *src; + u8 *result; + + void *__ctx[] CRYPTO_MINALIGN_ATTR; +}; + +struct ahash_alg { + int (*init)(struct ahash_request *req); + int (*update)(struct ahash_request *req); + int (*final)(struct ahash_request *req); + int (*finup)(struct ahash_request *req); + int (*digest)(struct ahash_request *req); + int (*export)(struct ahash_request *req, void *out); + int (*import)(struct ahash_request *req, const void *in); + int (*setkey)(struct crypto_ahash *tfm, const u8 *key, + unsigned int keylen); + + struct hash_alg_common halg; +}; + struct shash_desc { struct crypto_shash *tfm; u32 flags; @@ -37,6 +70,8 @@ struct shash_alg { unsigned int keylen); unsigned int descsize; + + /* These fields must match hash_alg_common. */ unsigned int digestsize; unsigned int statesize; @@ -44,6 +79,18 @@ struct shash_alg { }; struct crypto_ahash { + int (*init)(struct ahash_request *req); + int (*update)(struct ahash_request *req); + int (*final)(struct ahash_request *req); + int (*finup)(struct ahash_request *req); + int (*digest)(struct ahash_request *req); + int (*export)(struct ahash_request *req, void *out); + int (*import)(struct ahash_request *req, const void *in); + int (*setkey)(struct crypto_ahash *tfm, const u8 *key, + unsigned int keylen); + + unsigned int digestsize; + unsigned int reqsize; struct crypto_tfm base; }; @@ -54,19 +101,11 @@ struct crypto_shash { static inline struct crypto_ahash *__crypto_ahash_cast(struct crypto_tfm *tfm) { - return (struct crypto_ahash *)tfm; + return container_of(tfm, struct crypto_ahash, base); } -static inline struct crypto_ahash *crypto_alloc_ahash(const char *alg_name, - u32 type, u32 mask) -{ - type &= ~CRYPTO_ALG_TYPE_MASK; - mask &= ~CRYPTO_ALG_TYPE_MASK; - type |= CRYPTO_ALG_TYPE_AHASH; - mask |= CRYPTO_ALG_TYPE_AHASH_MASK; - - return __crypto_ahash_cast(crypto_alloc_base(alg_name, type, mask)); -} +struct crypto_ahash *crypto_alloc_ahash(const char *alg_name, u32 type, + u32 mask); static inline struct crypto_tfm *crypto_ahash_tfm(struct crypto_ahash *tfm) { @@ -75,7 +114,7 @@ static inline struct crypto_tfm *crypto_ahash_tfm(struct crypto_ahash *tfm) static inline void crypto_free_ahash(struct crypto_ahash *tfm) { - crypto_free_tfm(crypto_ahash_tfm(tfm)); + crypto_destroy_tfm(tfm, crypto_ahash_tfm(tfm)); } static inline unsigned int crypto_ahash_alignmask( @@ -84,14 +123,26 @@ static inline unsigned int crypto_ahash_alignmask( return crypto_tfm_alg_alignmask(crypto_ahash_tfm(tfm)); } -static inline struct ahash_tfm *crypto_ahash_crt(struct crypto_ahash *tfm) +static inline struct hash_alg_common *__crypto_hash_alg_common( + struct crypto_alg *alg) { - return &crypto_ahash_tfm(tfm)->crt_ahash; + return container_of(alg, struct hash_alg_common, base); +} + +static inline struct hash_alg_common *crypto_hash_alg_common( + struct crypto_ahash *tfm) +{ + return __crypto_hash_alg_common(crypto_ahash_tfm(tfm)->__crt_alg); } static inline unsigned int crypto_ahash_digestsize(struct crypto_ahash *tfm) { - return crypto_ahash_crt(tfm)->digestsize; + return tfm->digestsize; +} + +static inline unsigned int crypto_ahash_statesize(struct crypto_ahash *tfm) +{ + return crypto_hash_alg_common(tfm)->statesize; } static inline u32 crypto_ahash_get_flags(struct crypto_ahash *tfm) @@ -117,7 +168,7 @@ static inline struct crypto_ahash *crypto_ahash_reqtfm( static inline unsigned int crypto_ahash_reqsize(struct crypto_ahash *tfm) { - return crypto_ahash_crt(tfm)->reqsize; + return tfm->reqsize; } static inline void *ahash_request_ctx(struct ahash_request *req) @@ -128,41 +179,37 @@ static inline void *ahash_request_ctx(struct ahash_request *req) static inline int crypto_ahash_setkey(struct crypto_ahash *tfm, const u8 *key, unsigned int keylen) { - struct ahash_tfm *crt = crypto_ahash_crt(tfm); - - return crt->setkey(tfm, key, keylen); + return tfm->setkey(tfm, key, keylen); } static inline int crypto_ahash_digest(struct ahash_request *req) { - struct ahash_tfm *crt = crypto_ahash_crt(crypto_ahash_reqtfm(req)); - return crt->digest(req); + return crypto_ahash_reqtfm(req)->digest(req); } -static inline void crypto_ahash_export(struct ahash_request *req, u8 *out) +static inline int crypto_ahash_export(struct ahash_request *req, void *out) { - memcpy(out, ahash_request_ctx(req), - crypto_ahash_reqsize(crypto_ahash_reqtfm(req))); + return crypto_ahash_reqtfm(req)->export(req, out); } -int crypto_ahash_import(struct ahash_request *req, const u8 *in); +static inline int crypto_ahash_import(struct ahash_request *req, const void *in) +{ + return crypto_ahash_reqtfm(req)->import(req, in); +} static inline int crypto_ahash_init(struct ahash_request *req) { - struct ahash_tfm *crt = crypto_ahash_crt(crypto_ahash_reqtfm(req)); - return crt->init(req); + return crypto_ahash_reqtfm(req)->init(req); } static inline int crypto_ahash_update(struct ahash_request *req) { - struct ahash_tfm *crt = crypto_ahash_crt(crypto_ahash_reqtfm(req)); - return crt->update(req); + return crypto_ahash_reqtfm(req)->update(req); } static inline int crypto_ahash_final(struct ahash_request *req) { - struct ahash_tfm *crt = crypto_ahash_crt(crypto_ahash_reqtfm(req)); - return crt->final(req); + return crypto_ahash_reqtfm(req)->final(req); } static inline void ahash_request_set_tfm(struct ahash_request *req, diff --git a/include/crypto/internal/hash.h b/include/crypto/internal/hash.h index 5e45818f3351..08bdffafefad 100644 --- a/include/crypto/internal/hash.h +++ b/include/crypto/internal/hash.h @@ -51,6 +51,9 @@ int crypto_hash_walk_first_compat(struct hash_desc *hdesc, struct crypto_hash_walk *walk, struct scatterlist *sg, unsigned int len); +int crypto_register_ahash(struct ahash_alg *alg); +int crypto_unregister_ahash(struct ahash_alg *alg); + int crypto_register_shash(struct shash_alg *alg); int crypto_unregister_shash(struct shash_alg *alg); int shash_register_instance(struct crypto_template *tmpl, @@ -66,12 +69,14 @@ struct shash_alg *shash_attr_alg(struct rtattr *rta, u32 type, u32 mask); int shash_ahash_update(struct ahash_request *req, struct shash_desc *desc); int shash_ahash_digest(struct ahash_request *req, struct shash_desc *desc); +int crypto_init_shash_ops_async(struct crypto_tfm *tfm); + static inline void *crypto_ahash_ctx(struct crypto_ahash *tfm) { - return crypto_tfm_ctx(&tfm->base); + return crypto_tfm_ctx(crypto_ahash_tfm(tfm)); } -static inline struct ahash_alg *crypto_ahash_alg( +static inline struct old_ahash_alg *crypto_old_ahash_alg( struct crypto_ahash *tfm) { return &crypto_ahash_tfm(tfm)->__crt_alg->cra_ahash; @@ -80,7 +85,7 @@ static inline struct ahash_alg *crypto_ahash_alg( static inline void crypto_ahash_set_reqsize(struct crypto_ahash *tfm, unsigned int reqsize) { - crypto_ahash_crt(tfm)->reqsize = reqsize; + tfm->reqsize = reqsize; } static inline int ahash_enqueue_request(struct crypto_queue *queue, diff --git a/include/linux/crypto.h b/include/linux/crypto.h index 274f9c7da90c..9e7e9b62a3dc 100644 --- a/include/linux/crypto.h +++ b/include/linux/crypto.h @@ -120,6 +120,7 @@ struct crypto_rng; struct crypto_tfm; struct crypto_type; struct aead_givcrypt_request; +struct ahash_request; struct skcipher_givcrypt_request; typedef void (*crypto_completion_t)(struct crypto_async_request *req, int err); @@ -146,16 +147,6 @@ struct ablkcipher_request { void *__ctx[] CRYPTO_MINALIGN_ATTR; }; -struct ahash_request { - struct crypto_async_request base; - - unsigned int nbytes; - struct scatterlist *src; - u8 *result; - - void *__ctx[] CRYPTO_MINALIGN_ATTR; -}; - /** * struct aead_request - AEAD request * @base: Common attributes for async crypto requests @@ -220,7 +211,7 @@ struct ablkcipher_alg { unsigned int ivsize; }; -struct ahash_alg { +struct old_ahash_alg { int (*init)(struct ahash_request *req); int (*reinit)(struct ahash_request *req); int (*update)(struct ahash_request *req); @@ -346,7 +337,7 @@ struct crypto_alg { struct cipher_alg cipher; struct digest_alg digest; struct hash_alg hash; - struct ahash_alg ahash; + struct old_ahash_alg ahash; struct compress_alg compress; struct rng_alg rng; } cra_u; @@ -433,18 +424,6 @@ struct hash_tfm { unsigned int digestsize; }; -struct ahash_tfm { - int (*init)(struct ahash_request *req); - int (*update)(struct ahash_request *req); - int (*final)(struct ahash_request *req); - int (*digest)(struct ahash_request *req); - int (*setkey)(struct crypto_ahash *tfm, const u8 *key, - unsigned int keylen); - - unsigned int digestsize; - unsigned int reqsize; -}; - struct compress_tfm { int (*cot_compress)(struct crypto_tfm *tfm, const u8 *src, unsigned int slen, @@ -465,7 +444,6 @@ struct rng_tfm { #define crt_blkcipher crt_u.blkcipher #define crt_cipher crt_u.cipher #define crt_hash crt_u.hash -#define crt_ahash crt_u.ahash #define crt_compress crt_u.compress #define crt_rng crt_u.rng @@ -479,7 +457,6 @@ struct crypto_tfm { struct blkcipher_tfm blkcipher; struct cipher_tfm cipher; struct hash_tfm hash; - struct ahash_tfm ahash; struct compress_tfm compress; struct rng_tfm rng; } crt_u; -- cgit v1.2.3 From 01c2dece4316dadc0f9fad1ad0b56d493980e492 Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Tue, 14 Jul 2009 14:06:06 +0800 Subject: crypto: ahash - Add instance/spawn support This patch adds support for creating ahash instances and using ahash as spawns. Signed-off-by: Herbert Xu --- crypto/ahash.c | 72 ++++++++++++++++++++++++++++++++++++++++++ include/crypto/internal/hash.h | 51 ++++++++++++++++++++++++++++++ 2 files changed, 123 insertions(+) (limited to 'include') diff --git a/crypto/ahash.c b/crypto/ahash.c index 838519386215..7f599d26086a 100644 --- a/crypto/ahash.c +++ b/crypto/ahash.c @@ -259,5 +259,77 @@ struct crypto_ahash *crypto_alloc_ahash(const char *alg_name, u32 type, } EXPORT_SYMBOL_GPL(crypto_alloc_ahash); +static int ahash_prepare_alg(struct ahash_alg *alg) +{ + struct crypto_alg *base = &alg->halg.base; + + if (alg->halg.digestsize > PAGE_SIZE / 8 || + alg->halg.statesize > PAGE_SIZE / 8) + return -EINVAL; + + base->cra_type = &crypto_ahash_type; + base->cra_flags &= ~CRYPTO_ALG_TYPE_MASK; + base->cra_flags |= CRYPTO_ALG_TYPE_AHASH; + + return 0; +} + +int crypto_register_ahash(struct ahash_alg *alg) +{ + struct crypto_alg *base = &alg->halg.base; + int err; + + err = ahash_prepare_alg(alg); + if (err) + return err; + + return crypto_register_alg(base); +} +EXPORT_SYMBOL_GPL(crypto_register_ahash); + +int crypto_unregister_ahash(struct ahash_alg *alg) +{ + return crypto_unregister_alg(&alg->halg.base); +} +EXPORT_SYMBOL_GPL(crypto_unregister_ahash); + +int ahash_register_instance(struct crypto_template *tmpl, + struct ahash_instance *inst) +{ + int err; + + err = ahash_prepare_alg(&inst->alg); + if (err) + return err; + + return crypto_register_instance(tmpl, ahash_crypto_instance(inst)); +} +EXPORT_SYMBOL_GPL(ahash_register_instance); + +void ahash_free_instance(struct crypto_instance *inst) +{ + crypto_drop_spawn(crypto_instance_ctx(inst)); + kfree(ahash_instance(inst)); +} +EXPORT_SYMBOL_GPL(ahash_free_instance); + +int crypto_init_ahash_spawn(struct crypto_ahash_spawn *spawn, + struct hash_alg_common *alg, + struct crypto_instance *inst) +{ + return crypto_init_spawn2(&spawn->base, &alg->base, inst, + &crypto_ahash_type); +} +EXPORT_SYMBOL_GPL(crypto_init_ahash_spawn); + +struct hash_alg_common *ahash_attr_alg(struct rtattr *rta, u32 type, u32 mask) +{ + struct crypto_alg *alg; + + alg = crypto_attr_alg2(rta, &crypto_ahash_type, type, mask); + return IS_ERR(alg) ? ERR_CAST(alg) : __crypto_hash_alg_common(alg); +} +EXPORT_SYMBOL_GPL(ahash_attr_alg); + MODULE_LICENSE("GPL"); MODULE_DESCRIPTION("Asynchronous cryptographic hash type"); diff --git a/include/crypto/internal/hash.h b/include/crypto/internal/hash.h index 08bdffafefad..34eda233ee67 100644 --- a/include/crypto/internal/hash.h +++ b/include/crypto/internal/hash.h @@ -34,10 +34,18 @@ struct crypto_hash_walk { unsigned int flags; }; +struct ahash_instance { + struct ahash_alg alg; +}; + struct shash_instance { struct shash_alg alg; }; +struct crypto_ahash_spawn { + struct crypto_spawn base; +}; + struct crypto_shash_spawn { struct crypto_spawn base; }; @@ -53,6 +61,15 @@ int crypto_hash_walk_first_compat(struct hash_desc *hdesc, int crypto_register_ahash(struct ahash_alg *alg); int crypto_unregister_ahash(struct ahash_alg *alg); +int ahash_register_instance(struct crypto_template *tmpl, + struct ahash_instance *inst); +void ahash_free_instance(struct crypto_instance *inst); + +int crypto_init_ahash_spawn(struct crypto_ahash_spawn *spawn, + struct hash_alg_common *alg, + struct crypto_instance *inst); + +struct hash_alg_common *ahash_attr_alg(struct rtattr *rta, u32 type, u32 mask); int crypto_register_shash(struct shash_alg *alg); int crypto_unregister_shash(struct shash_alg *alg); @@ -88,6 +105,40 @@ static inline void crypto_ahash_set_reqsize(struct crypto_ahash *tfm, tfm->reqsize = reqsize; } +static inline struct crypto_instance *ahash_crypto_instance( + struct ahash_instance *inst) +{ + return container_of(&inst->alg.halg.base, struct crypto_instance, alg); +} + +static inline struct ahash_instance *ahash_instance( + struct crypto_instance *inst) +{ + return container_of(&inst->alg, struct ahash_instance, alg.halg.base); +} + +static inline void *ahash_instance_ctx(struct ahash_instance *inst) +{ + return crypto_instance_ctx(ahash_crypto_instance(inst)); +} + +static inline unsigned int ahash_instance_headroom(void) +{ + return sizeof(struct ahash_alg) - sizeof(struct crypto_alg); +} + +static inline struct ahash_instance *ahash_alloc_instance( + const char *name, struct crypto_alg *alg) +{ + return crypto_alloc_instance2(name, alg, ahash_instance_headroom()); +} + +static inline struct crypto_ahash *crypto_spawn_ahash( + struct crypto_ahash_spawn *spawn) +{ + return crypto_spawn_tfm2(&spawn->base); +} + static inline int ahash_enqueue_request(struct crypto_queue *queue, struct ahash_request *request) { -- cgit v1.2.3 From 52861c7cd711fac97b37ae0f4842a9ad26ecae72 Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Tue, 14 Jul 2009 18:30:24 +0800 Subject: crypto: hash - Add helpers to free spawns This patch adds the helpers crypto_drop_ahash and crypto_drop_shash so that these spawns can be dropped without ugly casts. Signed-off-by: Herbert Xu --- include/crypto/internal/hash.h | 10 ++++++++++ 1 file changed, 10 insertions(+) (limited to 'include') diff --git a/include/crypto/internal/hash.h b/include/crypto/internal/hash.h index 34eda233ee67..9d30316e9f10 100644 --- a/include/crypto/internal/hash.h +++ b/include/crypto/internal/hash.h @@ -69,6 +69,11 @@ int crypto_init_ahash_spawn(struct crypto_ahash_spawn *spawn, struct hash_alg_common *alg, struct crypto_instance *inst); +static inline void crypto_drop_ahash(struct crypto_ahash_spawn *spawn) +{ + crypto_drop_spawn(&spawn->base); +} + struct hash_alg_common *ahash_attr_alg(struct rtattr *rta, u32 type, u32 mask); int crypto_register_shash(struct shash_alg *alg); @@ -81,6 +86,11 @@ int crypto_init_shash_spawn(struct crypto_shash_spawn *spawn, struct shash_alg *alg, struct crypto_instance *inst); +static inline void crypto_drop_shash(struct crypto_shash_spawn *spawn) +{ + crypto_drop_spawn(&spawn->base); +} + struct shash_alg *shash_attr_alg(struct rtattr *rta, u32 type, u32 mask); int shash_ahash_update(struct ahash_request *req, struct shash_desc *desc); -- cgit v1.2.3 From 9cd899a32f611eb6328014f1d9e0ba31977812d9 Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Tue, 14 Jul 2009 18:45:45 +0800 Subject: crypto: cryptd - Switch to template create API This patch changes cryptd to use the template->create function instead of alloc in anticipation for the switch to new style ahash algorithms. Signed-off-by: Herbert Xu --- crypto/cryptd.c | 53 +++++++++++++++++++++++++++---------------------- crypto/internal.h | 3 --- include/crypto/algapi.h | 3 +++ 3 files changed, 32 insertions(+), 27 deletions(-) (limited to 'include') diff --git a/crypto/cryptd.c b/crypto/cryptd.c index 6e6722edd0d9..ad58f513ba8b 100644 --- a/crypto/cryptd.c +++ b/crypto/cryptd.c @@ -287,8 +287,9 @@ out_free_inst: goto out; } -static struct crypto_instance *cryptd_alloc_blkcipher( - struct rtattr **tb, struct cryptd_queue *queue) +static int cryptd_create_blkcipher(struct crypto_template *tmpl, + struct rtattr **tb, + struct cryptd_queue *queue) { struct cryptd_instance_ctx *ctx; struct crypto_instance *inst; @@ -298,7 +299,7 @@ static struct crypto_instance *cryptd_alloc_blkcipher( alg = crypto_get_attr_alg(tb, CRYPTO_ALG_TYPE_BLKCIPHER, CRYPTO_ALG_TYPE_MASK); if (IS_ERR(alg)) - return ERR_CAST(alg); + return PTR_ERR(alg); inst = cryptd_alloc_instance(alg, sizeof(*ctx)); if (IS_ERR(inst)) @@ -330,14 +331,16 @@ static struct crypto_instance *cryptd_alloc_blkcipher( inst->alg.cra_ablkcipher.encrypt = cryptd_blkcipher_encrypt_enqueue; inst->alg.cra_ablkcipher.decrypt = cryptd_blkcipher_decrypt_enqueue; + err = crypto_register_instance(tmpl, inst); + if (err) { + crypto_drop_spawn(&ctx->spawn); +out_free_inst: + kfree(inst); + } + out_put_alg: crypto_mod_put(alg); - return inst; - -out_free_inst: - kfree(inst); - inst = ERR_PTR(err); - goto out_put_alg; + return err; } static int cryptd_hash_init_tfm(struct crypto_tfm *tfm) @@ -502,8 +505,8 @@ static int cryptd_hash_digest_enqueue(struct ahash_request *req) return cryptd_hash_enqueue(req, cryptd_hash_digest); } -static struct crypto_instance *cryptd_alloc_hash( - struct rtattr **tb, struct cryptd_queue *queue) +static int cryptd_create_hash(struct crypto_template *tmpl, struct rtattr **tb, + struct cryptd_queue *queue) { struct hashd_instance_ctx *ctx; struct crypto_instance *inst; @@ -513,7 +516,7 @@ static struct crypto_instance *cryptd_alloc_hash( salg = shash_attr_alg(tb[1], 0, 0); if (IS_ERR(salg)) - return ERR_CAST(salg); + return PTR_ERR(salg); alg = &salg->base; inst = cryptd_alloc_instance(alg, sizeof(*ctx)); @@ -542,34 +545,36 @@ static struct crypto_instance *cryptd_alloc_hash( inst->alg.cra_ahash.setkey = cryptd_hash_setkey; inst->alg.cra_ahash.digest = cryptd_hash_digest_enqueue; + err = crypto_register_instance(tmpl, inst); + if (err) { + crypto_drop_shash(&ctx->spawn); +out_free_inst: + kfree(inst); + } + out_put_alg: crypto_mod_put(alg); - return inst; - -out_free_inst: - kfree(inst); - inst = ERR_PTR(err); - goto out_put_alg; + return err; } static struct cryptd_queue queue; -static struct crypto_instance *cryptd_alloc(struct rtattr **tb) +static int cryptd_create(struct crypto_template *tmpl, struct rtattr **tb) { struct crypto_attr_type *algt; algt = crypto_get_attr_type(tb); if (IS_ERR(algt)) - return ERR_CAST(algt); + return PTR_ERR(algt); switch (algt->type & algt->mask & CRYPTO_ALG_TYPE_MASK) { case CRYPTO_ALG_TYPE_BLKCIPHER: - return cryptd_alloc_blkcipher(tb, &queue); + return cryptd_create_blkcipher(tmpl, tb, &queue); case CRYPTO_ALG_TYPE_DIGEST: - return cryptd_alloc_hash(tb, &queue); + return cryptd_create_hash(tmpl, tb, &queue); } - return ERR_PTR(-EINVAL); + return -EINVAL; } static void cryptd_free(struct crypto_instance *inst) @@ -582,7 +587,7 @@ static void cryptd_free(struct crypto_instance *inst) static struct crypto_template cryptd_tmpl = { .name = "cryptd", - .alloc = cryptd_alloc, + .create = cryptd_create, .free = cryptd_free, .module = THIS_MODULE, }; diff --git a/crypto/internal.h b/crypto/internal.h index 030e22921c30..2d226362e594 100644 --- a/crypto/internal.h +++ b/crypto/internal.h @@ -97,9 +97,6 @@ struct crypto_alg *crypto_find_alg(const char *alg_name, void *crypto_alloc_tfm(const char *alg_name, const struct crypto_type *frontend, u32 type, u32 mask); -int crypto_register_instance(struct crypto_template *tmpl, - struct crypto_instance *inst); - int crypto_register_notifier(struct notifier_block *nb); int crypto_unregister_notifier(struct notifier_block *nb); int crypto_probing_notify(unsigned long val, void *v); diff --git a/include/crypto/algapi.h b/include/crypto/algapi.h index 7635fde7b1a2..9de6c38f4069 100644 --- a/include/crypto/algapi.h +++ b/include/crypto/algapi.h @@ -114,6 +114,9 @@ int crypto_register_template(struct crypto_template *tmpl); void crypto_unregister_template(struct crypto_template *tmpl); struct crypto_template *crypto_lookup_template(const char *name); +int crypto_register_instance(struct crypto_template *tmpl, + struct crypto_instance *inst); + int crypto_init_spawn(struct crypto_spawn *spawn, struct crypto_alg *alg, struct crypto_instance *inst, u32 mask); int crypto_init_spawn2(struct crypto_spawn *spawn, struct crypto_alg *alg, -- cgit v1.2.3 From 4dc10c0142ce0af8c20ec44dc6928ae63ad4f73a Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Tue, 14 Jul 2009 20:21:46 +0800 Subject: crypto: crypto4xx - Switch to new style ahash This patch changes crypto4xx to use the new style ahash type. In particular, we now use ahash_alg to define ahash algorithms instead of crypto_alg. This is achieved by introducing a union that encapsulates the new type and the existing crypto_alg structure. They're told apart through a u32 field containing the type value. Signed-off-by: Herbert Xu --- drivers/crypto/amcc/crypto4xx_core.c | 85 +++++++++++++++++++++--------------- drivers/crypto/amcc/crypto4xx_core.h | 25 +++++++++-- include/crypto/internal/hash.h | 6 +++ 3 files changed, 77 insertions(+), 39 deletions(-) (limited to 'include') diff --git a/drivers/crypto/amcc/crypto4xx_core.c b/drivers/crypto/amcc/crypto4xx_core.c index cb7be4283aa0..857e35efedac 100644 --- a/drivers/crypto/amcc/crypto4xx_core.c +++ b/drivers/crypto/amcc/crypto4xx_core.c @@ -31,8 +31,6 @@ #include #include #include -#include -#include #include #include #include "crypto4xx_reg_def.h" @@ -998,11 +996,15 @@ static int crypto4xx_alg_init(struct crypto_tfm *tfm) ctx->sa_out_dma_addr = 0; ctx->sa_len = 0; - if (alg->cra_type == &crypto_ablkcipher_type) + switch (alg->cra_flags & CRYPTO_ALG_TYPE_MASK) { + default: tfm->crt_ablkcipher.reqsize = sizeof(struct crypto4xx_ctx); - else if (alg->cra_type == &crypto_ahash_type) + break; + case CRYPTO_ALG_TYPE_AHASH: crypto_ahash_set_reqsize(__crypto_ahash_cast(tfm), sizeof(struct crypto4xx_ctx)); + break; + } return 0; } @@ -1016,7 +1018,8 @@ static void crypto4xx_alg_exit(struct crypto_tfm *tfm) } int crypto4xx_register_alg(struct crypto4xx_device *sec_dev, - struct crypto_alg *crypto_alg, int array_size) + struct crypto4xx_alg_common *crypto_alg, + int array_size) { struct crypto4xx_alg *alg; int i; @@ -1028,13 +1031,18 @@ int crypto4xx_register_alg(struct crypto4xx_device *sec_dev, return -ENOMEM; alg->alg = crypto_alg[i]; - INIT_LIST_HEAD(&alg->alg.cra_list); - if (alg->alg.cra_init == NULL) - alg->alg.cra_init = crypto4xx_alg_init; - if (alg->alg.cra_exit == NULL) - alg->alg.cra_exit = crypto4xx_alg_exit; alg->dev = sec_dev; - rc = crypto_register_alg(&alg->alg); + + switch (alg->alg.type) { + case CRYPTO_ALG_TYPE_AHASH: + rc = crypto_register_ahash(&alg->alg.u.hash); + break; + + default: + rc = crypto_register_alg(&alg->alg.u.cipher); + break; + } + if (rc) { list_del(&alg->entry); kfree(alg); @@ -1052,7 +1060,14 @@ static void crypto4xx_unregister_alg(struct crypto4xx_device *sec_dev) list_for_each_entry_safe(alg, tmp, &sec_dev->alg_list, entry) { list_del(&alg->entry); - crypto_unregister_alg(&alg->alg); + switch (alg->alg.type) { + case CRYPTO_ALG_TYPE_AHASH: + crypto_unregister_ahash(&alg->alg.u.hash); + break; + + default: + crypto_unregister_alg(&alg->alg.u.cipher); + } kfree(alg); } } @@ -1105,17 +1120,18 @@ static irqreturn_t crypto4xx_ce_interrupt_handler(int irq, void *data) /** * Supported Crypto Algorithms */ -struct crypto_alg crypto4xx_alg[] = { +struct crypto4xx_alg_common crypto4xx_alg[] = { /* Crypto AES modes */ - { + { .type = CRYPTO_ALG_TYPE_ABLKCIPHER, .u.cipher = { .cra_name = "cbc(aes)", .cra_driver_name = "cbc-aes-ppc4xx", .cra_priority = CRYPTO4XX_CRYPTO_PRIORITY, .cra_flags = CRYPTO_ALG_TYPE_ABLKCIPHER | CRYPTO_ALG_ASYNC, .cra_blocksize = AES_BLOCK_SIZE, .cra_ctxsize = sizeof(struct crypto4xx_ctx), - .cra_alignmask = 0, .cra_type = &crypto_ablkcipher_type, + .cra_init = crypto4xx_alg_init, + .cra_exit = crypto4xx_alg_exit, .cra_module = THIS_MODULE, .cra_u = { .ablkcipher = { @@ -1127,29 +1143,26 @@ struct crypto_alg crypto4xx_alg[] = { .decrypt = crypto4xx_decrypt, } } - }, + }}, /* Hash SHA1 */ - { - .cra_name = "sha1", - .cra_driver_name = "sha1-ppc4xx", - .cra_priority = CRYPTO4XX_CRYPTO_PRIORITY, - .cra_flags = CRYPTO_ALG_TYPE_AHASH | CRYPTO_ALG_ASYNC, - .cra_blocksize = SHA1_BLOCK_SIZE, - .cra_ctxsize = sizeof(struct crypto4xx_ctx), - .cra_alignmask = 0, - .cra_type = &crypto_ahash_type, - .cra_init = crypto4xx_sha1_alg_init, - .cra_module = THIS_MODULE, - .cra_u = { - .ahash = { - .digestsize = SHA1_DIGEST_SIZE, - .init = crypto4xx_hash_init, - .update = crypto4xx_hash_update, - .final = crypto4xx_hash_final, - .digest = crypto4xx_hash_digest, - } + { .type = CRYPTO_ALG_TYPE_AHASH, .u.hash = { + .init = crypto4xx_hash_init, + .update = crypto4xx_hash_update, + .final = crypto4xx_hash_final, + .digest = crypto4xx_hash_digest, + .halg.digestsize = SHA1_DIGEST_SIZE, + .halg.base = { + .cra_name = "sha1", + .cra_driver_name = "sha1-ppc4xx", + .cra_priority = CRYPTO4XX_CRYPTO_PRIORITY, + .cra_flags = CRYPTO_ALG_ASYNC, + .cra_blocksize = SHA1_BLOCK_SIZE, + .cra_ctxsize = sizeof(struct crypto4xx_ctx), + .cra_init = crypto4xx_sha1_alg_init, + .cra_exit = crypto4xx_alg_exit, + .cra_module = THIS_MODULE, } - }, + }}, }; /** diff --git a/drivers/crypto/amcc/crypto4xx_core.h b/drivers/crypto/amcc/crypto4xx_core.h index 1ef103449364..da9cbe3b9fc3 100644 --- a/drivers/crypto/amcc/crypto4xx_core.h +++ b/drivers/crypto/amcc/crypto4xx_core.h @@ -22,6 +22,8 @@ #ifndef __CRYPTO4XX_CORE_H__ #define __CRYPTO4XX_CORE_H__ +#include + #define PPC460SX_SDR0_SRST 0x201 #define PPC405EX_SDR0_SRST 0x200 #define PPC460EX_SDR0_SRST 0x201 @@ -138,14 +140,31 @@ struct crypto4xx_req_ctx { u16 sa_len; }; +struct crypto4xx_alg_common { + u32 type; + union { + struct crypto_alg cipher; + struct ahash_alg hash; + } u; +}; + struct crypto4xx_alg { struct list_head entry; - struct crypto_alg alg; + struct crypto4xx_alg_common alg; struct crypto4xx_device *dev; }; -#define crypto_alg_to_crypto4xx_alg(x) \ - container_of(x, struct crypto4xx_alg, alg) +static inline struct crypto4xx_alg *crypto_alg_to_crypto4xx_alg( + struct crypto_alg *x) +{ + switch (x->cra_flags & CRYPTO_ALG_TYPE_MASK) { + case CRYPTO_ALG_TYPE_AHASH: + return container_of(__crypto_ahash_alg(x), + struct crypto4xx_alg, alg.u.hash); + } + + return container_of(x, struct crypto4xx_alg, alg.u.cipher); +} extern int crypto4xx_alloc_sa(struct crypto4xx_ctx *ctx, u32 size); extern void crypto4xx_free_sa(struct crypto4xx_ctx *ctx); diff --git a/include/crypto/internal/hash.h b/include/crypto/internal/hash.h index 9d30316e9f10..e3a82514d61e 100644 --- a/include/crypto/internal/hash.h +++ b/include/crypto/internal/hash.h @@ -103,6 +103,12 @@ static inline void *crypto_ahash_ctx(struct crypto_ahash *tfm) return crypto_tfm_ctx(crypto_ahash_tfm(tfm)); } +static inline struct ahash_alg *__crypto_ahash_alg(struct crypto_alg *alg) +{ + return container_of(__crypto_hash_alg_common(alg), struct ahash_alg, + halg); +} + static inline struct old_ahash_alg *crypto_old_ahash_alg( struct crypto_ahash *tfm) { -- cgit v1.2.3 From 500b3e3c3dc8e4845b77ae81e5b7b085ab183ce6 Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Tue, 14 Jul 2009 20:29:57 +0800 Subject: crypto: ahash - Remove old_ahash_alg Now that all ahash implementations have been converted to the new ahash type, we can remove old_ahash_alg and its associated support. Signed-off-by: Herbert Xu --- crypto/ahash.c | 27 --------------------------- crypto/shash.c | 2 -- include/crypto/hash.h | 3 +-- include/crypto/internal/hash.h | 6 ------ include/linux/crypto.h | 16 ---------------- 5 files changed, 1 insertion(+), 53 deletions(-) (limited to 'include') diff --git a/crypto/ahash.c b/crypto/ahash.c index 7f599d26086a..cc824ef25830 100644 --- a/crypto/ahash.c +++ b/crypto/ahash.c @@ -175,46 +175,19 @@ static int ahash_nosetkey(struct crypto_ahash *tfm, const u8 *key, return -ENOSYS; } -static int crypto_init_ahash_ops(struct crypto_tfm *tfm, u32 type, u32 mask) -{ - struct old_ahash_alg *alg = &tfm->__crt_alg->cra_ahash; - struct crypto_ahash *crt = __crypto_ahash_cast(tfm); - struct ahash_alg *nalg = crypto_ahash_alg(crt); - - if (alg->digestsize > PAGE_SIZE / 8) - return -EINVAL; - - crt->init = alg->init; - crt->update = alg->update; - crt->final = alg->final; - crt->digest = alg->digest; - crt->setkey = alg->setkey ? ahash_setkey : ahash_nosetkey; - crt->digestsize = alg->digestsize; - - nalg->setkey = alg->setkey; - nalg->halg.digestsize = alg->digestsize; - - return 0; -} - static int crypto_ahash_init_tfm(struct crypto_tfm *tfm) { struct crypto_ahash *hash = __crypto_ahash_cast(tfm); struct ahash_alg *alg = crypto_ahash_alg(hash); - struct old_ahash_alg *oalg = crypto_old_ahash_alg(hash); if (tfm->__crt_alg->cra_type != &crypto_ahash_type) return crypto_init_shash_ops_async(tfm); - if (oalg->init) - return crypto_init_ahash_ops(tfm, 0, 0); - hash->init = alg->init; hash->update = alg->update; hash->final = alg->final; hash->digest = alg->digest; hash->setkey = alg->setkey ? ahash_setkey : ahash_nosetkey; - hash->digestsize = alg->halg.digestsize; return 0; } diff --git a/crypto/shash.c b/crypto/shash.c index 615a5f4d9b7a..fd92c03b38fc 100644 --- a/crypto/shash.c +++ b/crypto/shash.c @@ -270,7 +270,6 @@ static void crypto_exit_shash_ops_async(struct crypto_tfm *tfm) int crypto_init_shash_ops_async(struct crypto_tfm *tfm) { struct crypto_alg *calg = tfm->__crt_alg; - struct shash_alg *alg = __crypto_shash_alg(calg); struct crypto_ahash *crt = __crypto_ahash_cast(tfm); struct crypto_shash **ctx = crypto_tfm_ctx(tfm); struct crypto_shash *shash; @@ -293,7 +292,6 @@ int crypto_init_shash_ops_async(struct crypto_tfm *tfm) crt->digest = shash_async_digest; crt->setkey = shash_async_setkey; - crt->digestsize = alg->digestsize; crt->reqsize = sizeof(struct shash_desc) + crypto_shash_descsize(shash); return 0; diff --git a/include/crypto/hash.h b/include/crypto/hash.h index 262861d8f0cb..45c2bddfdf32 100644 --- a/include/crypto/hash.h +++ b/include/crypto/hash.h @@ -89,7 +89,6 @@ struct crypto_ahash { int (*setkey)(struct crypto_ahash *tfm, const u8 *key, unsigned int keylen); - unsigned int digestsize; unsigned int reqsize; struct crypto_tfm base; }; @@ -137,7 +136,7 @@ static inline struct hash_alg_common *crypto_hash_alg_common( static inline unsigned int crypto_ahash_digestsize(struct crypto_ahash *tfm) { - return tfm->digestsize; + return crypto_hash_alg_common(tfm)->digestsize; } static inline unsigned int crypto_ahash_statesize(struct crypto_ahash *tfm) diff --git a/include/crypto/internal/hash.h b/include/crypto/internal/hash.h index e3a82514d61e..179dd8fdfa14 100644 --- a/include/crypto/internal/hash.h +++ b/include/crypto/internal/hash.h @@ -109,12 +109,6 @@ static inline struct ahash_alg *__crypto_ahash_alg(struct crypto_alg *alg) halg); } -static inline struct old_ahash_alg *crypto_old_ahash_alg( - struct crypto_ahash *tfm) -{ - return &crypto_ahash_tfm(tfm)->__crt_alg->cra_ahash; -} - static inline void crypto_ahash_set_reqsize(struct crypto_ahash *tfm, unsigned int reqsize) { diff --git a/include/linux/crypto.h b/include/linux/crypto.h index 9e7e9b62a3dc..fd929889e8dc 100644 --- a/include/linux/crypto.h +++ b/include/linux/crypto.h @@ -115,12 +115,10 @@ struct crypto_async_request; struct crypto_aead; struct crypto_blkcipher; struct crypto_hash; -struct crypto_ahash; struct crypto_rng; struct crypto_tfm; struct crypto_type; struct aead_givcrypt_request; -struct ahash_request; struct skcipher_givcrypt_request; typedef void (*crypto_completion_t)(struct crypto_async_request *req, int err); @@ -211,18 +209,6 @@ struct ablkcipher_alg { unsigned int ivsize; }; -struct old_ahash_alg { - int (*init)(struct ahash_request *req); - int (*reinit)(struct ahash_request *req); - int (*update)(struct ahash_request *req); - int (*final)(struct ahash_request *req); - int (*digest)(struct ahash_request *req); - int (*setkey)(struct crypto_ahash *tfm, const u8 *key, - unsigned int keylen); - - unsigned int digestsize; -}; - struct aead_alg { int (*setkey)(struct crypto_aead *tfm, const u8 *key, unsigned int keylen); @@ -309,7 +295,6 @@ struct rng_alg { #define cra_cipher cra_u.cipher #define cra_digest cra_u.digest #define cra_hash cra_u.hash -#define cra_ahash cra_u.ahash #define cra_compress cra_u.compress #define cra_rng cra_u.rng @@ -337,7 +322,6 @@ struct crypto_alg { struct cipher_alg cipher; struct digest_alg digest; struct hash_alg hash; - struct old_ahash_alg ahash; struct compress_alg compress; struct rng_alg rng; } cra_u; -- cgit v1.2.3 From 8660c1240ec6016522b882c88751cb4ce40bf0e8 Mon Sep 17 00:00:00 2001 From: Tobias Klauser Date: Mon, 13 Jul 2009 22:48:16 +0000 Subject: skbuff.h: Fix comment for NET_IP_ALIGN Use the correct function call for skb_reserve in the comment for NET_IP_ALIGN. Signed-off-by: Tobias Klauser Signed-off-by: David S. Miller --- include/linux/skbuff.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index b47b3f039d14..f2c69a2cca17 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -1342,12 +1342,12 @@ static inline int skb_network_offset(const struct sk_buff *skb) * shifting the start of the packet by 2 bytes. Drivers should do this * with: * - * skb_reserve(NET_IP_ALIGN); + * skb_reserve(skb, NET_IP_ALIGN); * * The downside to this alignment of the IP header is that the DMA is now * unaligned. On some architectures the cost of an unaligned DMA is high * and this cost outweighs the gains made by aligning the IP header. - * + * * Since this trade off varies between architectures, we allow NET_IP_ALIGN * to be overridden. */ -- cgit v1.2.3 From 4bd9b0f4afc76cf972578c702e1bc1b6f2d10ba5 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 24 Jun 2009 15:37:45 -0400 Subject: nfsd41: use globals for DRC limits The version 4.1 DRC memory limit and tracking variables are server wide and session specific. Replace struct svc_serv fields with globals. Stop using the svc_serv sv_lock. Add a spinlock to serialize access to the DRC limit management variables which change on session creation and deletion (usage counter) or (future) administrative action to adjust the total DRC memory limit. Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy --- fs/nfsd/nfs4state.c | 10 +++++----- fs/nfsd/nfssvc.c | 19 +++++++++++++++---- include/linux/nfsd/nfsd.h | 3 +++ include/linux/sunrpc/svc.h | 2 -- 4 files changed, 23 insertions(+), 11 deletions(-) (limited to 'include') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 980a216a48c8..2e6a44e3d2fe 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -430,11 +430,11 @@ static int set_forechannel_maxreqs(struct nfsd4_channel_attrs *fchan) else if (fchan->maxreqs > NFSD_MAX_SLOTS_PER_SESSION) fchan->maxreqs = NFSD_MAX_SLOTS_PER_SESSION; - spin_lock(&nfsd_serv->sv_lock); - if (np + nfsd_serv->sv_drc_pages_used > nfsd_serv->sv_drc_max_pages) - np = nfsd_serv->sv_drc_max_pages - nfsd_serv->sv_drc_pages_used; - nfsd_serv->sv_drc_pages_used += np; - spin_unlock(&nfsd_serv->sv_lock); + spin_lock(&nfsd_drc_lock); + if (np + nfsd_drc_pages_used > nfsd_drc_max_pages) + np = nfsd_drc_max_pages - nfsd_drc_pages_used; + nfsd_drc_pages_used += np; + spin_unlock(&nfsd_drc_lock); if (np <= 0) { status = nfserr_resource; diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c index d4c9884cd54b..78d8fcd883fb 100644 --- a/fs/nfsd/nfssvc.c +++ b/fs/nfsd/nfssvc.c @@ -67,6 +67,16 @@ struct timeval nfssvc_boot; DEFINE_MUTEX(nfsd_mutex); struct svc_serv *nfsd_serv; +/* + * nfsd_drc_lock protects nfsd_drc_max_pages and nfsd_drc_pages_used. + * nfsd_drc_max_pages limits the total amount of memory available for + * version 4.1 DRC caches. + * nfsd_drc_pages_used tracks the current version 4.1 DRC memory usage. + */ +spinlock_t nfsd_drc_lock; +unsigned int nfsd_drc_max_pages; +unsigned int nfsd_drc_pages_used; + #if defined(CONFIG_NFSD_V2_ACL) || defined(CONFIG_NFSD_V3_ACL) static struct svc_stat nfsd_acl_svcstats; static struct svc_version * nfsd_acl_version[] = { @@ -238,11 +248,12 @@ static void set_max_drc(void) { /* The percent of nr_free_buffer_pages used by the V4.1 server DRC */ #define NFSD_DRC_SIZE_SHIFT 7 - nfsd_serv->sv_drc_max_pages = nr_free_buffer_pages() + nfsd_drc_max_pages = nr_free_buffer_pages() >> NFSD_DRC_SIZE_SHIFT; - nfsd_serv->sv_drc_pages_used = 0; - dprintk("%s svc_drc_max_pages %u\n", __func__, - nfsd_serv->sv_drc_max_pages); + nfsd_drc_pages_used = 0; + spin_lock_init(&nfsd_drc_lock); + dprintk("%s nfsd_drc_max_pages %u\n", __func__, + nfsd_drc_max_pages); } int nfsd_create_serv(void) diff --git a/include/linux/nfsd/nfsd.h b/include/linux/nfsd/nfsd.h index 2b49d676d0c9..2571f856908f 100644 --- a/include/linux/nfsd/nfsd.h +++ b/include/linux/nfsd/nfsd.h @@ -56,6 +56,9 @@ extern struct svc_version nfsd_version2, nfsd_version3, extern u32 nfsd_supported_minorversion; extern struct mutex nfsd_mutex; extern struct svc_serv *nfsd_serv; +extern spinlock_t nfsd_drc_lock; +extern unsigned int nfsd_drc_max_pages; +extern unsigned int nfsd_drc_pages_used; extern struct seq_operations nfs_exports_op; diff --git a/include/linux/sunrpc/svc.h b/include/linux/sunrpc/svc.h index ea8009695c69..52e8cb0a7569 100644 --- a/include/linux/sunrpc/svc.h +++ b/include/linux/sunrpc/svc.h @@ -94,8 +94,6 @@ struct svc_serv { struct module * sv_module; /* optional module to count when * adding threads */ svc_thread_fn sv_function; /* main function for threads */ - unsigned int sv_drc_max_pages; /* Total pages for DRC */ - unsigned int sv_drc_pages_used;/* DRC pages used */ #if defined(CONFIG_NFS_V4_1) struct list_head sv_cb_list; /* queue for callback requests * that arrive over the same -- cgit v1.2.3 From d0cb43b35d64877b2944bd37719708be5d7bbf99 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Thu, 9 Jul 2009 09:27:50 +0900 Subject: libata: implement and use HORKAGE_NOSETXFER, take#2 PIONEER DVD-RW DVRTD08 times out SETXFER if no media is present. The device is SATA and simply skipping SETXFER works around the problem. Implement ATA_HORKAGE_NOSETXFER and apply it to the device. Reported by Moritz Rigler in the following thread. http://thread.gmane.org/gmane.linux.ide/36790 and by Lars in bko#9540. Updated to whine and ignore NOSETXFER if PATA component is detected as suggested by Alan Cox. Signed-off-by: Tejun Heo Reported-by: Moritz Rigler Reported-by: Lars Cc: Alan Cox Signed-off-by: Jeff Garzik --- drivers/ata/libata-core.c | 20 ++++++++++++++++++-- include/linux/libata.h | 1 + 2 files changed, 19 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c index 045a486a09ea..2c6aedaef718 100644 --- a/drivers/ata/libata-core.c +++ b/drivers/ata/libata-core.c @@ -3392,17 +3392,27 @@ int ata_down_xfermask_limit(struct ata_device *dev, unsigned int sel) static int ata_dev_set_mode(struct ata_device *dev) { + struct ata_port *ap = dev->link->ap; struct ata_eh_context *ehc = &dev->link->eh_context; + const bool nosetxfer = dev->horkage & ATA_HORKAGE_NOSETXFER; const char *dev_err_whine = ""; int ign_dev_err = 0; - unsigned int err_mask; + unsigned int err_mask = 0; int rc; dev->flags &= ~ATA_DFLAG_PIO; if (dev->xfer_shift == ATA_SHIFT_PIO) dev->flags |= ATA_DFLAG_PIO; - err_mask = ata_dev_set_xfermode(dev); + if (nosetxfer && ap->flags & ATA_FLAG_SATA && ata_id_is_sata(dev->id)) + dev_err_whine = " (SET_XFERMODE skipped)"; + else { + if (nosetxfer) + ata_dev_printk(dev, KERN_WARNING, + "NOSETXFER but PATA detected - can't " + "skip SETXFER, might malfunction\n"); + err_mask = ata_dev_set_xfermode(dev); + } if (err_mask & ~AC_ERR_DEV) goto fail; @@ -4297,6 +4307,12 @@ static const struct ata_blacklist_entry ata_device_blacklist [] = { /* Devices which aren't very happy with higher link speeds */ { "WD My Book", NULL, ATA_HORKAGE_1_5_GBPS, }, + /* + * Devices which choke on SETXFER. Applies only if both the + * device and controller are SATA. + */ + { "PIONEER DVD-RW DVRTD08", "1.00", ATA_HORKAGE_NOSETXFER }, + /* End Marker */ { } }; diff --git a/include/linux/libata.h b/include/linux/libata.h index 3d501db36a26..79b6d7fd4ac2 100644 --- a/include/linux/libata.h +++ b/include/linux/libata.h @@ -385,6 +385,7 @@ enum { not multiple of 16 bytes */ ATA_HORKAGE_FIRMWARE_WARN = (1 << 12), /* firmware update warning */ ATA_HORKAGE_1_5_GBPS = (1 << 13), /* force 1.5 Gbps */ + ATA_HORKAGE_NOSETXFER = (1 << 14), /* skip SETXFER, SATA only */ /* DMA mask for user DMA control: User visible values; DO NOT renumber */ -- cgit v1.2.3 From 66f6ce5e52f2f209d5bf1f06167cec888f4f4c13 Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Wed, 15 Jul 2009 12:40:40 +0800 Subject: crypto: ahash - Add unaligned handling and default operations This patch exports the finup operation where available and adds a default finup operation for ahash. The operations final, finup and digest also will now deal with unaligned result pointers by copying it. Finally export/import operations are will now be exported too. Signed-off-by: Herbert Xu --- crypto/ahash.c | 204 ++++++++++++++++++++++++++++++++++++++++- crypto/shash.c | 52 ++++++++++- include/crypto/hash.h | 23 ++--- include/crypto/internal/hash.h | 6 ++ 4 files changed, 263 insertions(+), 22 deletions(-) (limited to 'include') diff --git a/crypto/ahash.c b/crypto/ahash.c index a196055b73d3..ac0798d2824e 100644 --- a/crypto/ahash.c +++ b/crypto/ahash.c @@ -24,6 +24,13 @@ #include "internal.h" +struct ahash_request_priv { + crypto_completion_t complete; + void *data; + u8 *result; + void *ubuf[] CRYPTO_MINALIGN_ATTR; +}; + static inline struct ahash_alg *crypto_ahash_alg(struct crypto_ahash *hash) { return container_of(crypto_hash_alg_common(hash), struct ahash_alg, @@ -156,7 +163,7 @@ static int ahash_setkey_unaligned(struct crypto_ahash *tfm, const u8 *key, return ret; } -static int ahash_setkey(struct crypto_ahash *tfm, const u8 *key, +int crypto_ahash_setkey(struct crypto_ahash *tfm, const u8 *key, unsigned int keylen) { struct ahash_alg *ahash = crypto_ahash_alg(tfm); @@ -167,6 +174,7 @@ static int ahash_setkey(struct crypto_ahash *tfm, const u8 *key, return ahash->setkey(tfm, key, keylen); } +EXPORT_SYMBOL_GPL(crypto_ahash_setkey); static int ahash_nosetkey(struct crypto_ahash *tfm, const u8 *key, unsigned int keylen) @@ -174,19 +182,209 @@ static int ahash_nosetkey(struct crypto_ahash *tfm, const u8 *key, return -ENOSYS; } +static inline unsigned int ahash_align_buffer_size(unsigned len, + unsigned long mask) +{ + return len + (mask & ~(crypto_tfm_ctx_alignment() - 1)); +} + +static void ahash_op_unaligned_finish(struct ahash_request *req, int err) +{ + struct ahash_request_priv *priv = req->priv; + + if (err == -EINPROGRESS) + return; + + if (!err) + memcpy(priv->result, req->result, + crypto_ahash_digestsize(crypto_ahash_reqtfm(req))); + + kzfree(priv); +} + +static void ahash_op_unaligned_done(struct crypto_async_request *req, int err) +{ + struct ahash_request *areq = req->data; + struct ahash_request_priv *priv = areq->priv; + crypto_completion_t complete = priv->complete; + void *data = priv->data; + + ahash_op_unaligned_finish(areq, err); + + complete(data, err); +} + +static int ahash_op_unaligned(struct ahash_request *req, + int (*op)(struct ahash_request *)) +{ + struct crypto_ahash *tfm = crypto_ahash_reqtfm(req); + unsigned long alignmask = crypto_ahash_alignmask(tfm); + unsigned int ds = crypto_ahash_digestsize(tfm); + struct ahash_request_priv *priv; + int err; + + priv = kmalloc(sizeof(*priv) + ahash_align_buffer_size(ds, alignmask), + (req->base.flags & CRYPTO_TFM_REQ_MAY_SLEEP) ? + GFP_ATOMIC : GFP_ATOMIC); + if (!priv) + return -ENOMEM; + + priv->result = req->result; + priv->complete = req->base.complete; + priv->data = req->base.data; + + req->result = PTR_ALIGN((u8 *)priv->ubuf, alignmask + 1); + req->base.complete = ahash_op_unaligned_done; + req->base.data = req; + req->priv = priv; + + err = op(req); + ahash_op_unaligned_finish(req, err); + + return err; +} + +static int crypto_ahash_op(struct ahash_request *req, + int (*op)(struct ahash_request *)) +{ + struct crypto_ahash *tfm = crypto_ahash_reqtfm(req); + unsigned long alignmask = crypto_ahash_alignmask(tfm); + + if ((unsigned long)req->result & alignmask) + return ahash_op_unaligned(req, op); + + return op(req); +} + +int crypto_ahash_final(struct ahash_request *req) +{ + return crypto_ahash_op(req, crypto_ahash_reqtfm(req)->final); +} +EXPORT_SYMBOL_GPL(crypto_ahash_final); + +int crypto_ahash_finup(struct ahash_request *req) +{ + return crypto_ahash_op(req, crypto_ahash_reqtfm(req)->finup); +} +EXPORT_SYMBOL_GPL(crypto_ahash_finup); + +int crypto_ahash_digest(struct ahash_request *req) +{ + return crypto_ahash_op(req, crypto_ahash_reqtfm(req)->digest); +} +EXPORT_SYMBOL_GPL(crypto_ahash_digest); + +static void ahash_def_finup_finish2(struct ahash_request *req, int err) +{ + struct ahash_request_priv *priv = req->priv; + + if (err == -EINPROGRESS) + return; + + if (!err) + memcpy(priv->result, req->result, + crypto_ahash_digestsize(crypto_ahash_reqtfm(req))); + + kzfree(priv); +} + +static void ahash_def_finup_done2(struct crypto_async_request *req, int err) +{ + struct ahash_request *areq = req->data; + struct ahash_request_priv *priv = areq->priv; + crypto_completion_t complete = priv->complete; + void *data = priv->data; + + ahash_def_finup_finish2(areq, err); + + complete(data, err); +} + +static int ahash_def_finup_finish1(struct ahash_request *req, int err) +{ + if (err) + goto out; + + req->base.complete = ahash_def_finup_done2; + req->base.flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; + err = crypto_ahash_reqtfm(req)->final(req); + +out: + ahash_def_finup_finish2(req, err); + return err; +} + +static void ahash_def_finup_done1(struct crypto_async_request *req, int err) +{ + struct ahash_request *areq = req->data; + struct ahash_request_priv *priv = areq->priv; + crypto_completion_t complete = priv->complete; + void *data = priv->data; + + err = ahash_def_finup_finish1(areq, err); + + complete(data, err); +} + +static int ahash_def_finup(struct ahash_request *req) +{ + struct crypto_ahash *tfm = crypto_ahash_reqtfm(req); + unsigned long alignmask = crypto_ahash_alignmask(tfm); + unsigned int ds = crypto_ahash_digestsize(tfm); + struct ahash_request_priv *priv; + + priv = kmalloc(sizeof(*priv) + ahash_align_buffer_size(ds, alignmask), + (req->base.flags & CRYPTO_TFM_REQ_MAY_SLEEP) ? + GFP_ATOMIC : GFP_ATOMIC); + if (!priv) + return -ENOMEM; + + priv->result = req->result; + priv->complete = req->base.complete; + priv->data = req->base.data; + + req->result = PTR_ALIGN((u8 *)priv->ubuf, alignmask + 1); + req->base.complete = ahash_def_finup_done1; + req->base.data = req; + req->priv = priv; + + return ahash_def_finup_finish1(req, tfm->update(req)); +} + +static int ahash_no_export(struct ahash_request *req, void *out) +{ + return -ENOSYS; +} + +static int ahash_no_import(struct ahash_request *req, const void *in) +{ + return -ENOSYS; +} + static int crypto_ahash_init_tfm(struct crypto_tfm *tfm) { struct crypto_ahash *hash = __crypto_ahash_cast(tfm); struct ahash_alg *alg = crypto_ahash_alg(hash); + hash->setkey = ahash_nosetkey; + hash->export = ahash_no_export; + hash->import = ahash_no_import; + if (tfm->__crt_alg->cra_type != &crypto_ahash_type) return crypto_init_shash_ops_async(tfm); hash->init = alg->init; hash->update = alg->update; - hash->final = alg->final; + hash->final = alg->final; + hash->finup = alg->finup ?: ahash_def_finup; hash->digest = alg->digest; - hash->setkey = alg->setkey ? ahash_setkey : ahash_nosetkey; + + if (alg->setkey) + hash->setkey = alg->setkey; + if (alg->export) + hash->export = alg->export; + if (alg->import) + hash->import = alg->import; return 0; } diff --git a/crypto/shash.c b/crypto/shash.c index 171c8f052f89..834d9d24cdae 100644 --- a/crypto/shash.c +++ b/crypto/shash.c @@ -235,6 +235,33 @@ static int shash_async_final(struct ahash_request *req) return crypto_shash_final(ahash_request_ctx(req), req->result); } +int shash_ahash_finup(struct ahash_request *req, struct shash_desc *desc) +{ + struct crypto_hash_walk walk; + int nbytes; + + for (nbytes = crypto_hash_walk_first(req, &walk); nbytes > 0; + nbytes = crypto_hash_walk_done(&walk, nbytes)) + nbytes = crypto_hash_walk_last(&walk) ? + crypto_shash_finup(desc, walk.data, nbytes, + req->result) : + crypto_shash_update(desc, walk.data, nbytes); + + return nbytes; +} +EXPORT_SYMBOL_GPL(shash_ahash_finup); + +static int shash_async_finup(struct ahash_request *req) +{ + struct crypto_shash **ctx = crypto_ahash_ctx(crypto_ahash_reqtfm(req)); + struct shash_desc *desc = ahash_request_ctx(req); + + desc->tfm = *ctx; + desc->flags = req->base.flags; + + return shash_ahash_finup(req, desc); +} + int shash_ahash_digest(struct ahash_request *req, struct shash_desc *desc) { struct scatterlist *sg = req->src; @@ -252,8 +279,7 @@ int shash_ahash_digest(struct ahash_request *req, struct shash_desc *desc) crypto_yield(desc->flags); } else err = crypto_shash_init(desc) ?: - shash_ahash_update(req, desc) ?: - crypto_shash_final(desc, req->result); + shash_ahash_finup(req, desc); return err; } @@ -270,6 +296,16 @@ static int shash_async_digest(struct ahash_request *req) return shash_ahash_digest(req, desc); } +static int shash_async_export(struct ahash_request *req, void *out) +{ + return crypto_shash_export(ahash_request_ctx(req), out); +} + +static int shash_async_import(struct ahash_request *req, const void *in) +{ + return crypto_shash_import(ahash_request_ctx(req), in); +} + static void crypto_exit_shash_ops_async(struct crypto_tfm *tfm) { struct crypto_shash **ctx = crypto_tfm_ctx(tfm); @@ -280,6 +316,7 @@ static void crypto_exit_shash_ops_async(struct crypto_tfm *tfm) int crypto_init_shash_ops_async(struct crypto_tfm *tfm) { struct crypto_alg *calg = tfm->__crt_alg; + struct shash_alg *alg = __crypto_shash_alg(calg); struct crypto_ahash *crt = __crypto_ahash_cast(tfm); struct crypto_shash **ctx = crypto_tfm_ctx(tfm); struct crypto_shash *shash; @@ -298,9 +335,16 @@ int crypto_init_shash_ops_async(struct crypto_tfm *tfm) crt->init = shash_async_init; crt->update = shash_async_update; - crt->final = shash_async_final; + crt->final = shash_async_final; + crt->finup = shash_async_finup; crt->digest = shash_async_digest; - crt->setkey = shash_async_setkey; + + if (alg->setkey) + crt->setkey = shash_async_setkey; + if (alg->export) + crt->export = shash_async_export; + if (alg->setkey) + crt->import = shash_async_import; crt->reqsize = sizeof(struct shash_desc) + crypto_shash_descsize(shash); diff --git a/include/crypto/hash.h b/include/crypto/hash.h index 45c2bddfdf32..3e89ce16b59c 100644 --- a/include/crypto/hash.h +++ b/include/crypto/hash.h @@ -31,6 +31,9 @@ struct ahash_request { struct scatterlist *src; u8 *result; + /* This field may only be used by the ahash API code. */ + void *priv; + void *__ctx[] CRYPTO_MINALIGN_ATTR; }; @@ -175,16 +178,11 @@ static inline void *ahash_request_ctx(struct ahash_request *req) return req->__ctx; } -static inline int crypto_ahash_setkey(struct crypto_ahash *tfm, - const u8 *key, unsigned int keylen) -{ - return tfm->setkey(tfm, key, keylen); -} - -static inline int crypto_ahash_digest(struct ahash_request *req) -{ - return crypto_ahash_reqtfm(req)->digest(req); -} +int crypto_ahash_setkey(struct crypto_ahash *tfm, const u8 *key, + unsigned int keylen); +int crypto_ahash_finup(struct ahash_request *req); +int crypto_ahash_final(struct ahash_request *req); +int crypto_ahash_digest(struct ahash_request *req); static inline int crypto_ahash_export(struct ahash_request *req, void *out) { @@ -206,11 +204,6 @@ static inline int crypto_ahash_update(struct ahash_request *req) return crypto_ahash_reqtfm(req)->update(req); } -static inline int crypto_ahash_final(struct ahash_request *req) -{ - return crypto_ahash_reqtfm(req)->final(req); -} - static inline void ahash_request_set_tfm(struct ahash_request *req, struct crypto_ahash *tfm) { diff --git a/include/crypto/internal/hash.h b/include/crypto/internal/hash.h index 179dd8fdfa14..5bfad8c80595 100644 --- a/include/crypto/internal/hash.h +++ b/include/crypto/internal/hash.h @@ -59,6 +59,11 @@ int crypto_hash_walk_first_compat(struct hash_desc *hdesc, struct crypto_hash_walk *walk, struct scatterlist *sg, unsigned int len); +static inline int crypto_hash_walk_last(struct crypto_hash_walk *walk) +{ + return !(walk->entrylen | walk->total); +} + int crypto_register_ahash(struct ahash_alg *alg); int crypto_unregister_ahash(struct ahash_alg *alg); int ahash_register_instance(struct crypto_template *tmpl, @@ -94,6 +99,7 @@ static inline void crypto_drop_shash(struct crypto_shash_spawn *spawn) struct shash_alg *shash_attr_alg(struct rtattr *rta, u32 type, u32 mask); int shash_ahash_update(struct ahash_request *req, struct shash_desc *desc); +int shash_ahash_finup(struct ahash_request *req, struct shash_desc *desc); int shash_ahash_digest(struct ahash_request *req, struct shash_desc *desc); int crypto_init_shash_ops_async(struct crypto_tfm *tfm); -- cgit v1.2.3 From 069a9dce384e211784ce6fdfaf1f13921327480d Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Wed, 1 Jul 2009 13:03:52 -0400 Subject: drm/radeon: add some missing pci ids Also, fix ordering for a couple others Signed-off-by: Alex Deucher Signed-off-by: Dave Airlie --- include/drm/drm_pciids.h | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/include/drm/drm_pciids.h b/include/drm/drm_pciids.h index 45c18672b093..7174818c2c13 100644 --- a/include/drm/drm_pciids.h +++ b/include/drm/drm_pciids.h @@ -43,6 +43,7 @@ {0x1002, 0x4A4F, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_R420|RADEON_NEW_MEMMAP}, \ {0x1002, 0x4A50, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_R420|RADEON_NEW_MEMMAP}, \ {0x1002, 0x4A54, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_R420|RADEON_NEW_MEMMAP}, \ + {0x1002, 0x4B48, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_R420|RADEON_NEW_MEMMAP}, \ {0x1002, 0x4B49, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_R420|RADEON_NEW_MEMMAP}, \ {0x1002, 0x4B4A, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_R420|RADEON_NEW_MEMMAP}, \ {0x1002, 0x4B4B, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_R420|RADEON_NEW_MEMMAP}, \ @@ -262,6 +263,7 @@ {0x1002, 0x9440, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV770|RADEON_NEW_MEMMAP}, \ {0x1002, 0x9441, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV770|RADEON_NEW_MEMMAP}, \ {0x1002, 0x9442, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV770|RADEON_NEW_MEMMAP}, \ + {0x1002, 0x9443, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV770|RADEON_NEW_MEMMAP}, \ {0x1002, 0x9444, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV770|RADEON_NEW_MEMMAP}, \ {0x1002, 0x9446, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV770|RADEON_NEW_MEMMAP}, \ {0x1002, 0x944A, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV770|RADEON_IS_MOBILITY|RADEON_NEW_MEMMAP}, \ @@ -346,12 +348,12 @@ {0x1002, 0x9599, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV635|RADEON_NEW_MEMMAP}, \ {0x1002, 0x959B, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV635|RADEON_IS_MOBILITY|RADEON_NEW_MEMMAP}, \ {0x1002, 0x95C0, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV620|RADEON_NEW_MEMMAP}, \ + {0x1002, 0x95C2, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV620|RADEON_IS_MOBILITY|RADEON_NEW_MEMMAP}, \ + {0x1002, 0x95C4, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV620|RADEON_IS_MOBILITY|RADEON_NEW_MEMMAP}, \ {0x1002, 0x95C5, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV620|RADEON_NEW_MEMMAP}, \ {0x1002, 0x95C6, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV620|RADEON_NEW_MEMMAP}, \ {0x1002, 0x95C7, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV620|RADEON_NEW_MEMMAP}, \ {0x1002, 0x95C9, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV620|RADEON_NEW_MEMMAP}, \ - {0x1002, 0x95C2, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV620|RADEON_IS_MOBILITY|RADEON_NEW_MEMMAP}, \ - {0x1002, 0x95C4, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV620|RADEON_IS_MOBILITY|RADEON_NEW_MEMMAP}, \ {0x1002, 0x95CC, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV620|RADEON_NEW_MEMMAP}, \ {0x1002, 0x95CD, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV620|RADEON_NEW_MEMMAP}, \ {0x1002, 0x95CE, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV620|RADEON_NEW_MEMMAP}, \ -- cgit v1.2.3 From d782c3f95c9263dc0b98e7115f75f1e18b9600b3 Mon Sep 17 00:00:00 2001 From: Zhao Yakui Date: Mon, 22 Jun 2009 13:17:08 +0800 Subject: drm/mode: add the CVT algorithm in kernel space Add the CVT algorithm in kernel space. And this function can be called to generate the required modeline. I copied it from the file of xserver/hw/xfree86/modes/xf86cvt.c. What I have done is to translate it by using integer calculation. This is to avoid the float-point calculation in kernel space. [airlied:- cleaned up some bits] Signed-off-by: Zhao Yakui Signed-off-by: Dave Airlie --- drivers/gpu/drm/drm_modes.c | 219 ++++++++++++++++++++++++++++++++++++++++++++ include/drm/drm_crtc.h | 3 + 2 files changed, 222 insertions(+) (limited to 'include') diff --git a/drivers/gpu/drm/drm_modes.c b/drivers/gpu/drm/drm_modes.c index 54f492a488a9..0dbc7e4f8643 100644 --- a/drivers/gpu/drm/drm_modes.c +++ b/drivers/gpu/drm/drm_modes.c @@ -8,6 +8,7 @@ * Copyright © 2007 Dave Airlie * Copyright © 2007-2008 Intel Corporation * Jesse Barnes + * Copyright 2005-2006 Luc Verhaegen * * Permission is hereby granted, free of charge, to any person obtaining a * copy of this software and associated documentation files (the "Software"), @@ -61,6 +62,224 @@ void drm_mode_debug_printmodeline(struct drm_display_mode *mode) } EXPORT_SYMBOL(drm_mode_debug_printmodeline); +/** + * drm_cvt_mode -create a modeline based on CVT algorithm + * @dev: DRM device + * @hdisplay: hdisplay size + * @vdisplay: vdisplay size + * @vrefresh : vrefresh rate + * @reduced : Whether the GTF calculation is simplified + * @interlaced:Whether the interlace is supported + * + * LOCKING: + * none. + * + * return the modeline based on CVT algorithm + * + * This function is called to generate the modeline based on CVT algorithm + * according to the hdisplay, vdisplay, vrefresh. + * It is based from the VESA(TM) Coordinated Video Timing Generator by + * Graham Loveridge April 9, 2003 available at + * http://www.vesa.org/public/CVT/CVTd6r1.xls + * + * And it is copied from xf86CVTmode in xserver/hw/xfree86/modes/xf86cvt.c. + * What I have done is to translate it by using integer calculation. + */ +#define HV_FACTOR 1000 +struct drm_display_mode *drm_cvt_mode(struct drm_device *dev, int hdisplay, + int vdisplay, int vrefresh, + bool reduced, bool interlaced) +{ + /* 1) top/bottom margin size (% of height) - default: 1.8, */ +#define CVT_MARGIN_PERCENTAGE 18 + /* 2) character cell horizontal granularity (pixels) - default 8 */ +#define CVT_H_GRANULARITY 8 + /* 3) Minimum vertical porch (lines) - default 3 */ +#define CVT_MIN_V_PORCH 3 + /* 4) Minimum number of vertical back porch lines - default 6 */ +#define CVT_MIN_V_BPORCH 6 + /* Pixel Clock step (kHz) */ +#define CVT_CLOCK_STEP 250 + struct drm_display_mode *drm_mode; + bool margins = false; + unsigned int vfieldrate, hperiod; + int hdisplay_rnd, hmargin, vdisplay_rnd, vmargin, vsync; + int interlace; + + /* allocate the drm_display_mode structure. If failure, we will + * return directly + */ + drm_mode = drm_mode_create(dev); + if (!drm_mode) + return NULL; + + /* the CVT default refresh rate is 60Hz */ + if (!vrefresh) + vrefresh = 60; + + /* the required field fresh rate */ + if (interlaced) + vfieldrate = vrefresh * 2; + else + vfieldrate = vrefresh; + + /* horizontal pixels */ + hdisplay_rnd = hdisplay - (hdisplay % CVT_H_GRANULARITY); + + /* determine the left&right borders */ + hmargin = 0; + if (margins) { + hmargin = hdisplay_rnd * CVT_MARGIN_PERCENTAGE / 1000; + hmargin -= hmargin % CVT_H_GRANULARITY; + } + /* find the total active pixels */ + drm_mode->hdisplay = hdisplay_rnd + 2 * hmargin; + + /* find the number of lines per field */ + if (interlaced) + vdisplay_rnd = vdisplay / 2; + else + vdisplay_rnd = vdisplay; + + /* find the top & bottom borders */ + vmargin = 0; + if (margins) + vmargin = vdisplay_rnd * CVT_MARGIN_PERCENTAGE / 1000; + + drm_mode->vdisplay = vdisplay_rnd + 2 * vmargin; + + /* Interlaced */ + if (interlaced) + interlace = 1; + else + interlace = 0; + + /* Determine VSync Width from aspect ratio */ + if (!(vdisplay % 3) && ((vdisplay * 4 / 3) == hdisplay)) + vsync = 4; + else if (!(vdisplay % 9) && ((vdisplay * 16 / 9) == hdisplay)) + vsync = 5; + else if (!(vdisplay % 10) && ((vdisplay * 16 / 10) == hdisplay)) + vsync = 6; + else if (!(vdisplay % 4) && ((vdisplay * 5 / 4) == hdisplay)) + vsync = 7; + else if (!(vdisplay % 9) && ((vdisplay * 15 / 9) == hdisplay)) + vsync = 7; + else /* custom */ + vsync = 10; + + if (!reduced) { + /* simplify the GTF calculation */ + /* 4) Minimum time of vertical sync + back porch interval (µs) + * default 550.0 + */ + int tmp1, tmp2; +#define CVT_MIN_VSYNC_BP 550 + /* 3) Nominal HSync width (% of line period) - default 8 */ +#define CVT_HSYNC_PERCENTAGE 8 + unsigned int hblank_percentage; + int vsyncandback_porch, vback_porch, hblank; + + /* estimated the horizontal period */ + tmp1 = HV_FACTOR * 1000000 - + CVT_MIN_VSYNC_BP * HV_FACTOR * vfieldrate; + tmp2 = (vdisplay_rnd + 2 * vmargin + CVT_MIN_V_PORCH) * 2 + + interlace; + hperiod = tmp1 * 2 / (tmp2 * vfieldrate); + + tmp1 = CVT_MIN_VSYNC_BP * HV_FACTOR / hperiod + 1; + /* 9. Find number of lines in sync + backporch */ + if (tmp1 < (vsync + CVT_MIN_V_PORCH)) + vsyncandback_porch = vsync + CVT_MIN_V_PORCH; + else + vsyncandback_porch = tmp1; + /* 10. Find number of lines in back porch */ + vback_porch = vsyncandback_porch - vsync; + drm_mode->vtotal = vdisplay_rnd + 2 * vmargin + + vsyncandback_porch + CVT_MIN_V_PORCH; + /* 5) Definition of Horizontal blanking time limitation */ + /* Gradient (%/kHz) - default 600 */ +#define CVT_M_FACTOR 600 + /* Offset (%) - default 40 */ +#define CVT_C_FACTOR 40 + /* Blanking time scaling factor - default 128 */ +#define CVT_K_FACTOR 128 + /* Scaling factor weighting - default 20 */ +#define CVT_J_FACTOR 20 +#define CVT_M_PRIME (CVT_M_FACTOR * CVT_K_FACTOR / 256) +#define CVT_C_PRIME ((CVT_C_FACTOR - CVT_J_FACTOR) * CVT_K_FACTOR / 256 + \ + CVT_J_FACTOR) + /* 12. Find ideal blanking duty cycle from formula */ + hblank_percentage = CVT_C_PRIME * HV_FACTOR - CVT_M_PRIME * + hperiod / 1000; + /* 13. Blanking time */ + if (hblank_percentage < 20 * HV_FACTOR) + hblank_percentage = 20 * HV_FACTOR; + hblank = drm_mode->hdisplay * hblank_percentage / + (100 * HV_FACTOR - hblank_percentage); + hblank -= hblank % (2 * CVT_H_GRANULARITY); + /* 14. find the total pixes per line */ + drm_mode->htotal = drm_mode->hdisplay + hblank; + drm_mode->hsync_end = drm_mode->hdisplay + hblank / 2; + drm_mode->hsync_start = drm_mode->hsync_end - + (drm_mode->htotal * CVT_HSYNC_PERCENTAGE) / 100; + drm_mode->hsync_start += CVT_H_GRANULARITY - + drm_mode->hsync_start % CVT_H_GRANULARITY; + /* fill the Vsync values */ + drm_mode->vsync_start = drm_mode->vdisplay + CVT_MIN_V_PORCH; + drm_mode->vsync_end = drm_mode->vsync_start + vsync; + } else { + /* Reduced blanking */ + /* Minimum vertical blanking interval time (µs)- default 460 */ +#define CVT_RB_MIN_VBLANK 460 + /* Fixed number of clocks for horizontal sync */ +#define CVT_RB_H_SYNC 32 + /* Fixed number of clocks for horizontal blanking */ +#define CVT_RB_H_BLANK 160 + /* Fixed number of lines for vertical front porch - default 3*/ +#define CVT_RB_VFPORCH 3 + int vbilines; + int tmp1, tmp2; + /* 8. Estimate Horizontal period. */ + tmp1 = HV_FACTOR * 1000000 - + CVT_RB_MIN_VBLANK * HV_FACTOR * vfieldrate; + tmp2 = vdisplay_rnd + 2 * vmargin; + hperiod = tmp1 / (tmp2 * vfieldrate); + /* 9. Find number of lines in vertical blanking */ + vbilines = CVT_RB_MIN_VBLANK * HV_FACTOR / hperiod + 1; + /* 10. Check if vertical blanking is sufficient */ + if (vbilines < (CVT_RB_VFPORCH + vsync + CVT_MIN_V_BPORCH)) + vbilines = CVT_RB_VFPORCH + vsync + CVT_MIN_V_BPORCH; + /* 11. Find total number of lines in vertical field */ + drm_mode->vtotal = vdisplay_rnd + 2 * vmargin + vbilines; + /* 12. Find total number of pixels in a line */ + drm_mode->htotal = drm_mode->hdisplay + CVT_RB_H_BLANK; + /* Fill in HSync values */ + drm_mode->hsync_end = drm_mode->hdisplay + CVT_RB_H_BLANK / 2; + drm_mode->hsync_start = drm_mode->hsync_end = CVT_RB_H_SYNC; + } + /* 15/13. Find pixel clock frequency (kHz for xf86) */ + drm_mode->clock = drm_mode->htotal * HV_FACTOR * 1000 / hperiod; + drm_mode->clock -= drm_mode->clock % CVT_CLOCK_STEP; + /* 18/16. Find actual vertical frame frequency */ + /* ignore - just set the mode flag for interlaced */ + if (interlaced) + drm_mode->vtotal *= 2; + /* Fill the mode line name */ + drm_mode_set_name(drm_mode); + if (reduced) + drm_mode->flags |= (DRM_MODE_FLAG_PHSYNC | + DRM_MODE_FLAG_NVSYNC); + else + drm_mode->flags |= (DRM_MODE_FLAG_PVSYNC | + DRM_MODE_FLAG_NHSYNC); + if (interlaced) + drm_mode->flags |= DRM_MODE_FLAG_INTERLACE; + + return drm_mode; +} +EXPORT_SYMBOL(drm_cvt_mode); + /** * drm_mode_set_name - set the name on a mode * @mode: name will be set in this mode diff --git a/include/drm/drm_crtc.h b/include/drm/drm_crtc.h index 7300fb866767..820bc0977e5e 100644 --- a/include/drm/drm_crtc.h +++ b/include/drm/drm_crtc.h @@ -736,4 +736,7 @@ extern int drm_mode_gamma_get_ioctl(struct drm_device *dev, extern int drm_mode_gamma_set_ioctl(struct drm_device *dev, void *data, struct drm_file *file_priv); extern bool drm_detect_hdmi_monitor(struct edid *edid); +extern struct drm_display_mode *drm_cvt_mode(struct drm_device *dev, + int hdisplay, int vdisplay, int vrefresh, + bool reduced, bool interlaced); #endif /* __DRM_CRTC_H__ */ -- cgit v1.2.3 From 26bbdadad356ec02d33657858d91675f3e9aca94 Mon Sep 17 00:00:00 2001 From: Zhao Yakui Date: Mon, 22 Jun 2009 13:17:09 +0800 Subject: drm/mode: add the GTF algorithm in kernel space Add the GTF algorithm in kernel space. And this function can be called to generate the required modeline. I copied it from the file of xserver/hw/xfree86/modes/xf86gtf.c. What I have done is to translate it by using integer calculation. This is to avoid the float-point calculation in kernel space. At the same tie I also refer to the function of fb_get_mode in drivers/video/fbmon.c Signed-off-by: Zhao Yakui Signed-off-by: Dave Airlie --- drivers/gpu/drm/drm_modes.c | 197 ++++++++++++++++++++++++++++++++++++++++++++ include/drm/drm_crtc.h | 3 + 2 files changed, 200 insertions(+) (limited to 'include') diff --git a/drivers/gpu/drm/drm_modes.c b/drivers/gpu/drm/drm_modes.c index 0dbc7e4f8643..fd489d76fbbc 100644 --- a/drivers/gpu/drm/drm_modes.c +++ b/drivers/gpu/drm/drm_modes.c @@ -9,6 +9,7 @@ * Copyright © 2007-2008 Intel Corporation * Jesse Barnes * Copyright 2005-2006 Luc Verhaegen + * Copyright (c) 2001, Andy Ritger aritger@nvidia.com * * Permission is hereby granted, free of charge, to any person obtaining a * copy of this software and associated documentation files (the "Software"), @@ -280,6 +281,202 @@ struct drm_display_mode *drm_cvt_mode(struct drm_device *dev, int hdisplay, } EXPORT_SYMBOL(drm_cvt_mode); +/** + * drm_gtf_mode - create the modeline based on GTF algorithm + * + * @dev :drm device + * @hdisplay :hdisplay size + * @vdisplay :vdisplay size + * @vrefresh :vrefresh rate. + * @interlaced :whether the interlace is supported + * @margins :whether the margin is supported + * + * LOCKING. + * none. + * + * return the modeline based on GTF algorithm + * + * This function is to create the modeline based on the GTF algorithm. + * Generalized Timing Formula is derived from: + * GTF Spreadsheet by Andy Morrish (1/5/97) + * available at http://www.vesa.org + * + * And it is copied from the file of xserver/hw/xfree86/modes/xf86gtf.c. + * What I have done is to translate it by using integer calculation. + * I also refer to the function of fb_get_mode in the file of + * drivers/video/fbmon.c + */ +struct drm_display_mode *drm_gtf_mode(struct drm_device *dev, int hdisplay, + int vdisplay, int vrefresh, + bool interlaced, int margins) +{ + /* 1) top/bottom margin size (% of height) - default: 1.8, */ +#define GTF_MARGIN_PERCENTAGE 18 + /* 2) character cell horizontal granularity (pixels) - default 8 */ +#define GTF_CELL_GRAN 8 + /* 3) Minimum vertical porch (lines) - default 3 */ +#define GTF_MIN_V_PORCH 1 + /* width of vsync in lines */ +#define V_SYNC_RQD 3 + /* width of hsync as % of total line */ +#define H_SYNC_PERCENT 8 + /* min time of vsync + back porch (microsec) */ +#define MIN_VSYNC_PLUS_BP 550 + /* blanking formula gradient */ +#define GTF_M 600 + /* blanking formula offset */ +#define GTF_C 40 + /* blanking formula scaling factor */ +#define GTF_K 128 + /* blanking formula scaling factor */ +#define GTF_J 20 + /* C' and M' are part of the Blanking Duty Cycle computation */ +#define GTF_C_PRIME (((GTF_C - GTF_J) * GTF_K / 256) + GTF_J) +#define GTF_M_PRIME (GTF_K * GTF_M / 256) + struct drm_display_mode *drm_mode; + unsigned int hdisplay_rnd, vdisplay_rnd, vfieldrate_rqd; + int top_margin, bottom_margin; + int interlace; + unsigned int hfreq_est; + int vsync_plus_bp, vback_porch; + unsigned int vtotal_lines, vfieldrate_est, hperiod; + unsigned int vfield_rate, vframe_rate; + int left_margin, right_margin; + unsigned int total_active_pixels, ideal_duty_cycle; + unsigned int hblank, total_pixels, pixel_freq; + int hsync, hfront_porch, vodd_front_porch_lines; + unsigned int tmp1, tmp2; + + drm_mode = drm_mode_create(dev); + if (!drm_mode) + return NULL; + + /* 1. In order to give correct results, the number of horizontal + * pixels requested is first processed to ensure that it is divisible + * by the character size, by rounding it to the nearest character + * cell boundary: + */ + hdisplay_rnd = (hdisplay + GTF_CELL_GRAN / 2) / GTF_CELL_GRAN; + hdisplay_rnd = hdisplay_rnd * GTF_CELL_GRAN; + + /* 2. If interlace is requested, the number of vertical lines assumed + * by the calculation must be halved, as the computation calculates + * the number of vertical lines per field. + */ + if (interlaced) + vdisplay_rnd = vdisplay / 2; + else + vdisplay_rnd = vdisplay; + + /* 3. Find the frame rate required: */ + if (interlaced) + vfieldrate_rqd = vrefresh * 2; + else + vfieldrate_rqd = vrefresh; + + /* 4. Find number of lines in Top margin: */ + top_margin = 0; + if (margins) + top_margin = (vdisplay_rnd * GTF_MARGIN_PERCENTAGE + 500) / + 1000; + /* 5. Find number of lines in bottom margin: */ + bottom_margin = top_margin; + + /* 6. If interlace is required, then set variable interlace: */ + if (interlaced) + interlace = 1; + else + interlace = 0; + + /* 7. Estimate the Horizontal frequency */ + { + tmp1 = (1000000 - MIN_VSYNC_PLUS_BP * vfieldrate_rqd) / 500; + tmp2 = (vdisplay_rnd + 2 * top_margin + GTF_MIN_V_PORCH) * + 2 + interlace; + hfreq_est = (tmp2 * 1000 * vfieldrate_rqd) / tmp1; + } + + /* 8. Find the number of lines in V sync + back porch */ + /* [V SYNC+BP] = RINT(([MIN VSYNC+BP] * hfreq_est / 1000000)) */ + vsync_plus_bp = MIN_VSYNC_PLUS_BP * hfreq_est / 1000; + vsync_plus_bp = (vsync_plus_bp + 500) / 1000; + /* 9. Find the number of lines in V back porch alone: */ + vback_porch = vsync_plus_bp - V_SYNC_RQD; + /* 10. Find the total number of lines in Vertical field period: */ + vtotal_lines = vdisplay_rnd + top_margin + bottom_margin + + vsync_plus_bp + GTF_MIN_V_PORCH; + /* 11. Estimate the Vertical field frequency: */ + vfieldrate_est = hfreq_est / vtotal_lines; + /* 12. Find the actual horizontal period: */ + hperiod = 1000000 / (vfieldrate_rqd * vtotal_lines); + + /* 13. Find the actual Vertical field frequency: */ + vfield_rate = hfreq_est / vtotal_lines; + /* 14. Find the Vertical frame frequency: */ + if (interlaced) + vframe_rate = vfield_rate / 2; + else + vframe_rate = vfield_rate; + /* 15. Find number of pixels in left margin: */ + if (margins) + left_margin = (hdisplay_rnd * GTF_MARGIN_PERCENTAGE + 500) / + 1000; + else + left_margin = 0; + + /* 16.Find number of pixels in right margin: */ + right_margin = left_margin; + /* 17.Find total number of active pixels in image and left and right */ + total_active_pixels = hdisplay_rnd + left_margin + right_margin; + /* 18.Find the ideal blanking duty cycle from blanking duty cycle */ + ideal_duty_cycle = GTF_C_PRIME * 1000 - + (GTF_M_PRIME * 1000000 / hfreq_est); + /* 19.Find the number of pixels in the blanking time to the nearest + * double character cell: */ + hblank = total_active_pixels * ideal_duty_cycle / + (100000 - ideal_duty_cycle); + hblank = (hblank + GTF_CELL_GRAN) / (2 * GTF_CELL_GRAN); + hblank = hblank * 2 * GTF_CELL_GRAN; + /* 20.Find total number of pixels: */ + total_pixels = total_active_pixels + hblank; + /* 21.Find pixel clock frequency: */ + pixel_freq = total_pixels * hfreq_est / 1000; + /* Stage 1 computations are now complete; I should really pass + * the results to another function and do the Stage 2 computations, + * but I only need a few more values so I'll just append the + * computations here for now */ + /* 17. Find the number of pixels in the horizontal sync period: */ + hsync = H_SYNC_PERCENT * total_pixels / 100; + hsync = (hsync + GTF_CELL_GRAN / 2) / GTF_CELL_GRAN; + hsync = hsync * GTF_CELL_GRAN; + /* 18. Find the number of pixels in horizontal front porch period */ + hfront_porch = hblank / 2 - hsync; + /* 36. Find the number of lines in the odd front porch period: */ + vodd_front_porch_lines = GTF_MIN_V_PORCH ; + + /* finally, pack the results in the mode struct */ + drm_mode->hdisplay = hdisplay_rnd; + drm_mode->hsync_start = hdisplay_rnd + hfront_porch; + drm_mode->hsync_end = drm_mode->hsync_start + hsync; + drm_mode->htotal = total_pixels; + drm_mode->vdisplay = vdisplay_rnd; + drm_mode->vsync_start = vdisplay_rnd + vodd_front_porch_lines; + drm_mode->vsync_end = drm_mode->vsync_start + V_SYNC_RQD; + drm_mode->vtotal = vtotal_lines; + + drm_mode->clock = pixel_freq; + + drm_mode_set_name(drm_mode); + drm_mode->flags = DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_PVSYNC; + + if (interlaced) { + drm_mode->vtotal *= 2; + drm_mode->flags |= DRM_MODE_FLAG_INTERLACE; + } + + return drm_mode; +} +EXPORT_SYMBOL(drm_gtf_mode); /** * drm_mode_set_name - set the name on a mode * @mode: name will be set in this mode diff --git a/include/drm/drm_crtc.h b/include/drm/drm_crtc.h index 820bc0977e5e..125994d8ac0b 100644 --- a/include/drm/drm_crtc.h +++ b/include/drm/drm_crtc.h @@ -739,4 +739,7 @@ extern bool drm_detect_hdmi_monitor(struct edid *edid); extern struct drm_display_mode *drm_cvt_mode(struct drm_device *dev, int hdisplay, int vdisplay, int vrefresh, bool reduced, bool interlaced); +extern struct drm_display_mode *drm_gtf_mode(struct drm_device *dev, + int hdisplay, int vdisplay, int vrefresh, + bool interlaced, int margins); #endif /* __DRM_CRTC_H__ */ -- cgit v1.2.3 From d1724078d6a01177c1db4ea0b75fda1ca8a73d57 Mon Sep 17 00:00:00 2001 From: Thomas Hellstrom Date: Wed, 24 Jun 2009 19:57:35 +0200 Subject: ttm: Make messages more readable. Signed-off-by: Thomas Hellstrom Signed-off-by: Dave Airlie --- include/drm/ttm/ttm_module.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/drm/ttm/ttm_module.h b/include/drm/ttm/ttm_module.h index 889a4c7958ae..d1d433834e4f 100644 --- a/include/drm/ttm/ttm_module.h +++ b/include/drm/ttm/ttm_module.h @@ -33,7 +33,7 @@ #include -#define TTM_PFX "[TTM]" +#define TTM_PFX "[TTM] " enum ttm_global_types { TTM_GLOBAL_TTM_MEM = 0, -- cgit v1.2.3 From ad49f501867cba87e1e45e5ebae0b12435d68bf1 Mon Sep 17 00:00:00 2001 From: Dave Airlie Date: Fri, 10 Jul 2009 22:36:26 +1000 Subject: drm/ttm/radeon: add dma32 support. This add support for using dma32 memory on gpus that really need it. Currently IGPs are left without DMA32 but we might need to change that unless we can fix rs690. Signed-off-by: Dave Airlie --- drivers/gpu/drm/radeon/radeon.h | 1 + drivers/gpu/drm/radeon/radeon_device.c | 17 +++++++++++++++-- drivers/gpu/drm/radeon/radeon_ttm.c | 3 ++- drivers/gpu/drm/ttm/ttm_bo.c | 7 ++++++- drivers/gpu/drm/ttm/ttm_tt.c | 9 +++++++-- include/drm/ttm/ttm_bo_driver.h | 5 ++++- 6 files changed, 35 insertions(+), 7 deletions(-) (limited to 'include') diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h index e7662ba9abfb..3060ce14071e 100644 --- a/drivers/gpu/drm/radeon/radeon.h +++ b/drivers/gpu/drm/radeon/radeon.h @@ -624,6 +624,7 @@ struct radeon_device { bool gpu_lockup; bool shutdown; bool suspend; + bool need_dma32; }; int radeon_device_init(struct radeon_device *rdev, diff --git a/drivers/gpu/drm/radeon/radeon_device.c b/drivers/gpu/drm/radeon/radeon_device.c index 332911267ebe..27a5ac969953 100644 --- a/drivers/gpu/drm/radeon/radeon_device.c +++ b/drivers/gpu/drm/radeon/radeon_device.c @@ -450,6 +450,7 @@ int radeon_device_init(struct radeon_device *rdev, uint32_t flags) { int r, ret; + int dma_bits; DRM_INFO("radeon: Initializing kernel modesetting.\n"); rdev->shutdown = false; @@ -492,8 +493,20 @@ int radeon_device_init(struct radeon_device *rdev, return r; } - /* Report DMA addressing limitation */ - r = pci_set_dma_mask(rdev->pdev, DMA_BIT_MASK(32)); + /* set DMA mask + need_dma32 flags. + * PCIE - can handle 40-bits. + * IGP - can handle 40-bits (in theory) + * AGP - generally dma32 is safest + * PCI - only dma32 + */ + rdev->need_dma32 = false; + if (rdev->flags & RADEON_IS_AGP) + rdev->need_dma32 = true; + if (rdev->flags & RADEON_IS_PCI) + rdev->need_dma32 = true; + + dma_bits = rdev->need_dma32 ? 32 : 40; + r = pci_set_dma_mask(rdev->pdev, DMA_BIT_MASK(dma_bits)); if (r) { printk(KERN_WARNING "radeon: No suitable DMA available.\n"); } diff --git a/drivers/gpu/drm/radeon/radeon_ttm.c b/drivers/gpu/drm/radeon/radeon_ttm.c index 1227a97f5169..4ca9aa9203d0 100644 --- a/drivers/gpu/drm/radeon/radeon_ttm.c +++ b/drivers/gpu/drm/radeon/radeon_ttm.c @@ -442,7 +442,8 @@ int radeon_ttm_init(struct radeon_device *rdev) /* No others user of address space so set it to 0 */ r = ttm_bo_device_init(&rdev->mman.bdev, rdev->mman.mem_global_ref.object, - &radeon_bo_driver, DRM_FILE_PAGE_OFFSET); + &radeon_bo_driver, DRM_FILE_PAGE_OFFSET, + rdev->need_dma32); if (r) { DRM_ERROR("failed initializing buffer object driver(%d).\n", r); return r; diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c index a753598a5e35..e55e7972c897 100644 --- a/drivers/gpu/drm/ttm/ttm_bo.c +++ b/drivers/gpu/drm/ttm/ttm_bo.c @@ -224,6 +224,9 @@ static int ttm_bo_add_ttm(struct ttm_buffer_object *bo, bool zero_alloc) TTM_ASSERT_LOCKED(&bo->mutex); bo->ttm = NULL; + if (bdev->need_dma32) + page_flags |= TTM_PAGE_FLAG_DMA32; + switch (bo->type) { case ttm_bo_type_device: if (zero_alloc) @@ -1332,7 +1335,8 @@ EXPORT_SYMBOL(ttm_bo_device_release); int ttm_bo_device_init(struct ttm_bo_device *bdev, struct ttm_mem_global *mem_glob, - struct ttm_bo_driver *driver, uint64_t file_page_offset) + struct ttm_bo_driver *driver, uint64_t file_page_offset, + bool need_dma32) { int ret = -EINVAL; @@ -1369,6 +1373,7 @@ int ttm_bo_device_init(struct ttm_bo_device *bdev, INIT_LIST_HEAD(&bdev->ddestroy); INIT_LIST_HEAD(&bdev->swap_lru); bdev->dev_mapping = NULL; + bdev->need_dma32 = need_dma32; ttm_mem_init_shrink(&bdev->shrink, ttm_bo_swapout); ret = ttm_mem_register_shrink(mem_glob, &bdev->shrink); if (unlikely(ret != 0)) { diff --git a/drivers/gpu/drm/ttm/ttm_tt.c b/drivers/gpu/drm/ttm/ttm_tt.c index 75dc8bd24592..81ab81f030a3 100644 --- a/drivers/gpu/drm/ttm/ttm_tt.c +++ b/drivers/gpu/drm/ttm/ttm_tt.c @@ -131,10 +131,15 @@ static void ttm_tt_free_page_directory(struct ttm_tt *ttm) static struct page *ttm_tt_alloc_page(unsigned page_flags) { + gfp_t gfp_flags = GFP_HIGHUSER; + if (page_flags & TTM_PAGE_FLAG_ZERO_ALLOC) - return alloc_page(GFP_HIGHUSER | __GFP_ZERO); + gfp_flags |= __GFP_ZERO; + + if (page_flags & TTM_PAGE_FLAG_DMA32) + gfp_flags |= __GFP_DMA32; - return alloc_page(GFP_HIGHUSER); + return alloc_page(gfp_flags); } static void ttm_tt_free_user_pages(struct ttm_tt *ttm) diff --git a/include/drm/ttm/ttm_bo_driver.h b/include/drm/ttm/ttm_bo_driver.h index 62ed733c52a2..ea83dd23a4d7 100644 --- a/include/drm/ttm/ttm_bo_driver.h +++ b/include/drm/ttm/ttm_bo_driver.h @@ -121,6 +121,7 @@ struct ttm_backend { #define TTM_PAGE_FLAG_SWAPPED (1 << 4) #define TTM_PAGE_FLAG_PERSISTANT_SWAP (1 << 5) #define TTM_PAGE_FLAG_ZERO_ALLOC (1 << 6) +#define TTM_PAGE_FLAG_DMA32 (1 << 7) enum ttm_caching_state { tt_uncached, @@ -429,6 +430,8 @@ struct ttm_bo_device { */ struct delayed_work wq; + + bool need_dma32; }; /** @@ -648,7 +651,7 @@ extern int ttm_bo_device_release(struct ttm_bo_device *bdev); extern int ttm_bo_device_init(struct ttm_bo_device *bdev, struct ttm_mem_global *mem_glob, struct ttm_bo_driver *driver, - uint64_t file_page_offset); + uint64_t file_page_offset, bool need_dma32); /** * ttm_bo_reserve: -- cgit v1.2.3 From 3d39cecc4841e8d4c4abdb401d10180f5faaded0 Mon Sep 17 00:00:00 2001 From: David Woodhouse Date: Wed, 8 Jul 2009 15:23:30 +0100 Subject: intel-iommu: Remove superfluous iova_alloc_lock from IOVA code We only ever obtain this lock immediately before the iova_rbtree_lock, and release it immediately after the iova_rbtree_lock. So ditch it and just use iova_rbtree_lock. [v2: Remove the lockdep bits this time too] Signed-off-by: David Woodhouse --- drivers/pci/intel-iommu.c | 3 --- drivers/pci/iova.c | 16 ++++------------ include/linux/iova.h | 1 - 3 files changed, 4 insertions(+), 16 deletions(-) (limited to 'include') diff --git a/drivers/pci/intel-iommu.c b/drivers/pci/intel-iommu.c index c5f7c73cbb55..d6a857397ec3 100644 --- a/drivers/pci/intel-iommu.c +++ b/drivers/pci/intel-iommu.c @@ -1309,7 +1309,6 @@ static void iommu_detach_domain(struct dmar_domain *domain, } static struct iova_domain reserved_iova_list; -static struct lock_class_key reserved_alloc_key; static struct lock_class_key reserved_rbtree_key; static void dmar_init_reserved_ranges(void) @@ -1320,8 +1319,6 @@ static void dmar_init_reserved_ranges(void) init_iova_domain(&reserved_iova_list, DMA_32BIT_PFN); - lockdep_set_class(&reserved_iova_list.iova_alloc_lock, - &reserved_alloc_key); lockdep_set_class(&reserved_iova_list.iova_rbtree_lock, &reserved_rbtree_key); diff --git a/drivers/pci/iova.c b/drivers/pci/iova.c index 46dd440e2315..7914951ef29a 100644 --- a/drivers/pci/iova.c +++ b/drivers/pci/iova.c @@ -22,7 +22,6 @@ void init_iova_domain(struct iova_domain *iovad, unsigned long pfn_32bit) { - spin_lock_init(&iovad->iova_alloc_lock); spin_lock_init(&iovad->iova_rbtree_lock); iovad->rbroot = RB_ROOT; iovad->cached32_node = NULL; @@ -205,7 +204,6 @@ alloc_iova(struct iova_domain *iovad, unsigned long size, unsigned long limit_pfn, bool size_aligned) { - unsigned long flags; struct iova *new_iova; int ret; @@ -219,11 +217,9 @@ alloc_iova(struct iova_domain *iovad, unsigned long size, if (size_aligned) size = __roundup_pow_of_two(size); - spin_lock_irqsave(&iovad->iova_alloc_lock, flags); ret = __alloc_and_insert_iova_range(iovad, size, limit_pfn, new_iova, size_aligned); - spin_unlock_irqrestore(&iovad->iova_alloc_lock, flags); if (ret) { free_iova_mem(new_iova); return NULL; @@ -381,8 +377,7 @@ reserve_iova(struct iova_domain *iovad, struct iova *iova; unsigned int overlap = 0; - spin_lock_irqsave(&iovad->iova_alloc_lock, flags); - spin_lock(&iovad->iova_rbtree_lock); + spin_lock_irqsave(&iovad->iova_rbtree_lock, flags); for (node = rb_first(&iovad->rbroot); node; node = rb_next(node)) { if (__is_range_overlap(node, pfn_lo, pfn_hi)) { iova = container_of(node, struct iova, node); @@ -402,8 +397,7 @@ reserve_iova(struct iova_domain *iovad, iova = __insert_new_range(iovad, pfn_lo, pfn_hi); finish: - spin_unlock(&iovad->iova_rbtree_lock); - spin_unlock_irqrestore(&iovad->iova_alloc_lock, flags); + spin_unlock_irqrestore(&iovad->iova_rbtree_lock, flags); return iova; } @@ -420,8 +414,7 @@ copy_reserved_iova(struct iova_domain *from, struct iova_domain *to) unsigned long flags; struct rb_node *node; - spin_lock_irqsave(&from->iova_alloc_lock, flags); - spin_lock(&from->iova_rbtree_lock); + spin_lock_irqsave(&from->iova_rbtree_lock, flags); for (node = rb_first(&from->rbroot); node; node = rb_next(node)) { struct iova *iova = container_of(node, struct iova, node); struct iova *new_iova; @@ -430,6 +423,5 @@ copy_reserved_iova(struct iova_domain *from, struct iova_domain *to) printk(KERN_ERR "Reserve iova range %lx@%lx failed\n", iova->pfn_lo, iova->pfn_lo); } - spin_unlock(&from->iova_rbtree_lock); - spin_unlock_irqrestore(&from->iova_alloc_lock, flags); + spin_unlock_irqrestore(&from->iova_rbtree_lock, flags); } diff --git a/include/linux/iova.h b/include/linux/iova.h index 228f6c94b69c..76a0759e88ec 100644 --- a/include/linux/iova.h +++ b/include/linux/iova.h @@ -28,7 +28,6 @@ struct iova { /* holds all the iova translations for a domain */ struct iova_domain { - spinlock_t iova_alloc_lock;/* Lock to protect iova allocation */ spinlock_t iova_rbtree_lock; /* Lock to protect update of rbtree */ struct rb_root rbroot; /* iova domain rbtree root */ struct rb_node *cached32_node; /* Save last alloced node */ -- cgit v1.2.3 From fa64966473830219fe74952029ddb0e981a87749 Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Wed, 15 Jul 2009 21:16:05 +0800 Subject: crypto: shash - Fix digest size offset When an shash algorithm is exported as ahash, ahash will access its digest size through hash_alg_common. That's why the shash layout needs to match hash_alg_common. This wasn't the case because the alignment weren't identical. This patch fixes the problem. Signed-off-by: Herbert Xu --- include/crypto/hash.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/crypto/hash.h b/include/crypto/hash.h index 3e89ce16b59c..26cb1eb16f4c 100644 --- a/include/crypto/hash.h +++ b/include/crypto/hash.h @@ -75,7 +75,8 @@ struct shash_alg { unsigned int descsize; /* These fields must match hash_alg_common. */ - unsigned int digestsize; + unsigned int digestsize + __attribute__ ((aligned(__alignof__(struct hash_alg_common)))); unsigned int statesize; struct crypto_alg base; -- cgit v1.2.3 From a76761b621bcd8336065c4fe3a74f046858bc34c Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Wed, 15 Jul 2009 23:35:14 +0900 Subject: percpu: add dummy pcpu_lpage_remapped() for !CONFIG_SMP !CONFIG_SMP was missing pcpu_lpage_remapped() definition causing build failure. Add dummy implementation. This was discovered by linux-next testing. Signed-off-by: Tejun Heo Cc: Randy Dunlap Cc: Kamalesh Babulal Cc: Stephen Rothwell --- include/linux/percpu.h | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'include') diff --git a/include/linux/percpu.h b/include/linux/percpu.h index 8ce91af4aa19..e134c8229631 100644 --- a/include/linux/percpu.h +++ b/include/linux/percpu.h @@ -184,6 +184,11 @@ static inline void free_percpu(void *p) static inline void __init setup_per_cpu_areas(void) { } +static inline void *pcpu_lpage_remapped(void *kaddr) +{ + return NULL; +} + #endif /* CONFIG_SMP */ #define alloc_percpu(type) (type *)__alloc_percpu(sizeof(type), \ -- cgit v1.2.3 From b333b3d22822cf9b295990866798e9239c9dee72 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Wed, 24 Jun 2009 01:34:48 +0000 Subject: wireless extensions: make netns aware This makes wireless extensions netns aware. The tasklet sending the events is converted to a work struct so that we can rtnl_lock() in it. Signed-off-by: Johannes Berg Signed-off-by: David S. Miller --- include/net/net_namespace.h | 3 +++ net/wireless/wext.c | 61 +++++++++++++++++++++------------------------ 2 files changed, 32 insertions(+), 32 deletions(-) (limited to 'include') diff --git a/include/net/net_namespace.h b/include/net/net_namespace.h index 3882db1e263a..5c5136fceea8 100644 --- a/include/net/net_namespace.h +++ b/include/net/net_namespace.h @@ -79,6 +79,9 @@ struct net { #endif #ifdef CONFIG_XFRM struct netns_xfrm xfrm; +#endif +#ifdef CONFIG_WIRELESS_EXT + struct sk_buff_head wext_nlevents; #endif struct net_generic *gen; }; diff --git a/net/wireless/wext.c b/net/wireless/wext.c index 425f7d58b961..db8351a5a87d 100644 --- a/net/wireless/wext.c +++ b/net/wireless/wext.c @@ -1257,48 +1257,48 @@ int compat_wext_handle_ioctl(struct net *net, unsigned int cmd, } #endif -/************************* EVENT PROCESSING *************************/ -/* - * Process events generated by the wireless layer or the driver. - * Most often, the event will be propagated through rtnetlink - */ +static int __net_init wext_pernet_init(struct net *net) +{ + skb_queue_head_init(&net->wext_nlevents); + return 0; +} -/* ---------------------------------------------------------------- */ -/* - * Locking... - * ---------- - * - * Thanks to Herbert Xu for fixing - * the locking issue in here and implementing this code ! - * - * The issue : wireless_send_event() is often called in interrupt context, - * while the Netlink layer can never be called in interrupt context. - * The fully formed RtNetlink events are queued, and then a tasklet is run - * to feed those to Netlink. - * The skb_queue is interrupt safe, and its lock is not held while calling - * Netlink, so there is no possibility of dealock. - * Jean II - */ +static void __net_exit wext_pernet_exit(struct net *net) +{ + skb_queue_purge(&net->wext_nlevents); +} -static struct sk_buff_head wireless_nlevent_queue; +static struct pernet_operations wext_pernet_ops = { + .init = wext_pernet_init, + .exit = wext_pernet_exit, +}; static int __init wireless_nlevent_init(void) { - skb_queue_head_init(&wireless_nlevent_queue); + return register_pernet_subsys(&wext_pernet_ops); return 0; } subsys_initcall(wireless_nlevent_init); -static void wireless_nlevent_process(unsigned long data) +/* Process events generated by the wireless layer or the driver. */ +static void wireless_nlevent_process(struct work_struct *work) { struct sk_buff *skb; + struct net *net; + + rtnl_lock(); - while ((skb = skb_dequeue(&wireless_nlevent_queue))) - rtnl_notify(skb, &init_net, 0, RTNLGRP_LINK, NULL, GFP_ATOMIC); + for_each_net(net) { + while ((skb = skb_dequeue(&net->wext_nlevents))) + rtnl_notify(skb, net, 0, RTNLGRP_LINK, NULL, + GFP_KERNEL); + } + + rtnl_unlock(); } -static DECLARE_TASKLET(wireless_nlevent_tasklet, wireless_nlevent_process, 0); +static DECLARE_WORK(wireless_nlevent_work, wireless_nlevent_process); /* ---------------------------------------------------------------- */ /* @@ -1348,9 +1348,6 @@ static void rtmsg_iwinfo(struct net_device *dev, char *event, int event_len) struct sk_buff *skb; int err; - if (!net_eq(dev_net(dev), &init_net)) - return; - skb = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_ATOMIC); if (!skb) return; @@ -1363,8 +1360,8 @@ static void rtmsg_iwinfo(struct net_device *dev, char *event, int event_len) } NETLINK_CB(skb).dst_group = RTNLGRP_LINK; - skb_queue_tail(&wireless_nlevent_queue, skb); - tasklet_schedule(&wireless_nlevent_tasklet); + skb_queue_tail(&dev_net(dev)->wext_nlevents, skb); + schedule_work(&wireless_nlevent_work); } /* ---------------------------------------------------------------- */ -- cgit v1.2.3 From 1dacc76d0014a034b8aca14237c127d7c19d7726 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Wed, 1 Jul 2009 11:26:02 +0000 Subject: net/compat/wext: send different messages to compat tasks Wireless extensions have the unfortunate problem that events are multicast netlink messages, and are not independent of pointer size. Thus, currently 32-bit tasks on 64-bit platforms cannot properly receive events and fail with all kinds of strange problems, for instance wpa_supplicant never notices disassociations, due to the way the 64-bit event looks (to a 32-bit process), the fact that the address is all zeroes is lost, it thinks instead it is 00:00:00:00:01:00. The same problem existed with the ioctls, until David Miller fixed those some time ago in an heroic effort. A different problem caused by this is that we cannot send the ASSOCREQIE/ASSOCRESPIE events because sending them causes a 32-bit wpa_supplicant on a 64-bit system to overwrite its internal information, which is worse than it not getting the information at all -- so we currently resort to sending a custom string event that it then parses. This, however, has a severe size limitation we are frequently hitting with modern access points; this limitation would can be lifted after this patch by sending the correct binary, not custom, event. A similar problem apparently happens for some other netlink users on x86_64 with 32-bit tasks due to the alignment for 64-bit quantities. In order to fix these problems, I have implemented a way to send compat messages to tasks. When sending an event, we send the non-compat event data together with a compat event data in skb_shinfo(main_skb)->frag_list. Then, when the event is read from the socket, the netlink code makes sure to pass out only the skb that is compatible with the task. This approach was suggested by David Miller, my original approach required always sending two skbs but that had various small problems. To determine whether compat is needed or not, I have used the MSG_CMSG_COMPAT flag, and adjusted the call path for recv and recvfrom to include it, even if those calls do not have a cmsg parameter. I have not solved one small part of the problem, and I don't think it is necessary to: if a 32-bit application uses read() rather than any form of recvmsg() it will still get the wrong (64-bit) event. However, neither do applications actually do this, nor would it be a regression. Signed-off-by: Johannes Berg Signed-off-by: David S. Miller --- arch/mips/kernel/scall64-n32.S | 2 +- arch/mips/kernel/scall64-o32.S | 4 +-- arch/sparc/kernel/sys32.S | 2 +- include/linux/wireless.h | 8 +++++ net/Kconfig | 20 +++++++++++ net/compat.c | 17 +++++++-- net/netlink/af_netlink.c | 36 ++++++++++++++++++- net/wireless/wext.c | 78 ++++++++++++++++++++++++++++++++++++++++++ 8 files changed, 160 insertions(+), 7 deletions(-) (limited to 'include') diff --git a/arch/mips/kernel/scall64-n32.S b/arch/mips/kernel/scall64-n32.S index 15874f9812cc..7c4a94f43706 100644 --- a/arch/mips/kernel/scall64-n32.S +++ b/arch/mips/kernel/scall64-n32.S @@ -164,7 +164,7 @@ EXPORT(sysn32_call_table) PTR sys_connect PTR sys_accept PTR sys_sendto - PTR sys_recvfrom + PTR compat_sys_recvfrom PTR compat_sys_sendmsg /* 6045 */ PTR compat_sys_recvmsg PTR sys_shutdown diff --git a/arch/mips/kernel/scall64-o32.S b/arch/mips/kernel/scall64-o32.S index 781e0f1e9533..821fc978673d 100644 --- a/arch/mips/kernel/scall64-o32.S +++ b/arch/mips/kernel/scall64-o32.S @@ -378,8 +378,8 @@ sys_call_table: PTR sys_getsockname PTR sys_getsockopt PTR sys_listen - PTR sys_recv /* 4175 */ - PTR sys_recvfrom + PTR compat_sys_recv /* 4175 */ + PTR compat_sys_recvfrom PTR compat_sys_recvmsg PTR sys_send PTR compat_sys_sendmsg diff --git a/arch/sparc/kernel/sys32.S b/arch/sparc/kernel/sys32.S index f061c4dda9ef..3762f6c78944 100644 --- a/arch/sparc/kernel/sys32.S +++ b/arch/sparc/kernel/sys32.S @@ -121,7 +121,7 @@ SIGN2(sys32_syslog, sys_syslog, %o0, %o2) SIGN1(sys32_umask, sys_umask, %o0) SIGN3(sys32_tgkill, sys_tgkill, %o0, %o1, %o2) SIGN1(sys32_sendto, sys_sendto, %o0) -SIGN1(sys32_recvfrom, sys_recvfrom, %o0) +SIGN1(sys32_recvfrom, compat_sys_recvfrom, %o0) SIGN3(sys32_socket, sys_socket, %o0, %o1, %o2) SIGN2(sys32_connect, sys_connect, %o0, %o2) SIGN2(sys32_bind, sys_bind, %o0, %o2) diff --git a/include/linux/wireless.h b/include/linux/wireless.h index cb24204851f7..5b4c6c772a9b 100644 --- a/include/linux/wireless.h +++ b/include/linux/wireless.h @@ -1132,6 +1132,14 @@ struct __compat_iw_event { }; #define IW_EV_COMPAT_LCP_LEN offsetof(struct __compat_iw_event, pointer) #define IW_EV_COMPAT_POINT_OFF offsetof(struct compat_iw_point, length) + +/* Size of the various events for compat */ +#define IW_EV_COMPAT_CHAR_LEN (IW_EV_COMPAT_LCP_LEN + IFNAMSIZ) +#define IW_EV_COMPAT_UINT_LEN (IW_EV_COMPAT_LCP_LEN + sizeof(__u32)) +#define IW_EV_COMPAT_FREQ_LEN (IW_EV_COMPAT_LCP_LEN + sizeof(struct iw_freq)) +#define IW_EV_COMPAT_PARAM_LEN (IW_EV_COMPAT_LCP_LEN + sizeof(struct iw_param)) +#define IW_EV_COMPAT_ADDR_LEN (IW_EV_COMPAT_LCP_LEN + sizeof(struct sockaddr)) +#define IW_EV_COMPAT_QUAL_LEN (IW_EV_COMPAT_LCP_LEN + sizeof(struct iw_quality)) #define IW_EV_COMPAT_POINT_LEN \ (IW_EV_COMPAT_LCP_LEN + sizeof(struct compat_iw_point) - \ IW_EV_COMPAT_POINT_OFF) diff --git a/net/Kconfig b/net/Kconfig index 7051b9710675..041c35edb763 100644 --- a/net/Kconfig +++ b/net/Kconfig @@ -23,6 +23,26 @@ menuconfig NET if NET +config WANT_COMPAT_NETLINK_MESSAGES + bool + help + This option can be selected by other options that need compat + netlink messages. + +config COMPAT_NETLINK_MESSAGES + def_bool y + depends on COMPAT + depends on WIRELESS_EXT || WANT_COMPAT_NETLINK_MESSAGES + help + This option makes it possible to send different netlink messages + to tasks depending on whether the task is a compat task or not. To + achieve this, you need to set skb_shinfo(skb)->frag_list to the + compat skb before sending the skb, the netlink code will sort out + which message to actually pass to the task. + + Newly written code should NEVER need this option but do + compat-independent messages instead! + menu "Networking options" source "net/packet/Kconfig" diff --git a/net/compat.c b/net/compat.c index 8d739053afe4..12728b17a226 100644 --- a/net/compat.c +++ b/net/compat.c @@ -743,6 +743,18 @@ asmlinkage long compat_sys_recvmsg(int fd, struct compat_msghdr __user *msg, uns return sys_recvmsg(fd, (struct msghdr __user *)msg, flags | MSG_CMSG_COMPAT); } +asmlinkage long compat_sys_recv(int fd, void __user *buf, size_t len, unsigned flags) +{ + return sys_recv(fd, buf, len, flags | MSG_CMSG_COMPAT); +} + +asmlinkage long compat_sys_recvfrom(int fd, void __user *buf, size_t len, + unsigned flags, struct sockaddr __user *addr, + int __user *addrlen) +{ + return sys_recvfrom(fd, buf, len, flags | MSG_CMSG_COMPAT, addr, addrlen); +} + asmlinkage long compat_sys_socketcall(int call, u32 __user *args) { int ret; @@ -788,10 +800,11 @@ asmlinkage long compat_sys_socketcall(int call, u32 __user *args) ret = sys_sendto(a0, compat_ptr(a1), a[2], a[3], compat_ptr(a[4]), a[5]); break; case SYS_RECV: - ret = sys_recv(a0, compat_ptr(a1), a[2], a[3]); + ret = compat_sys_recv(a0, compat_ptr(a1), a[2], a[3]); break; case SYS_RECVFROM: - ret = sys_recvfrom(a0, compat_ptr(a1), a[2], a[3], compat_ptr(a[4]), compat_ptr(a[5])); + ret = compat_sys_recvfrom(a0, compat_ptr(a1), a[2], a[3], + compat_ptr(a[4]), compat_ptr(a[5])); break; case SYS_SHUTDOWN: ret = sys_shutdown(a0,a1); diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c index d46da6cb92e4..da3163d15ef0 100644 --- a/net/netlink/af_netlink.c +++ b/net/netlink/af_netlink.c @@ -1361,7 +1361,7 @@ static int netlink_recvmsg(struct kiocb *kiocb, struct socket *sock, struct netlink_sock *nlk = nlk_sk(sk); int noblock = flags&MSG_DONTWAIT; size_t copied; - struct sk_buff *skb; + struct sk_buff *skb, *frag __maybe_unused = NULL; int err; if (flags&MSG_OOB) @@ -1373,6 +1373,35 @@ static int netlink_recvmsg(struct kiocb *kiocb, struct socket *sock, if (skb == NULL) goto out; +#ifdef CONFIG_COMPAT_NETLINK_MESSAGES + if (unlikely(skb_shinfo(skb)->frag_list)) { + bool need_compat = !!(flags & MSG_CMSG_COMPAT); + + /* + * If this skb has a frag_list, then here that means that + * we will have to use the frag_list skb for compat tasks + * and the regular skb for non-compat tasks. + * + * The skb might (and likely will) be cloned, so we can't + * just reset frag_list and go on with things -- we need to + * keep that. For the compat case that's easy -- simply get + * a reference to the compat skb and free the regular one + * including the frag. For the non-compat case, we need to + * avoid sending the frag to the user -- so assign NULL but + * restore it below before freeing the skb. + */ + if (need_compat) { + struct sk_buff *compskb = skb_shinfo(skb)->frag_list; + skb_get(compskb); + kfree_skb(skb); + skb = compskb; + } else { + frag = skb_shinfo(skb)->frag_list; + skb_shinfo(skb)->frag_list = NULL; + } + } +#endif + msg->msg_namelen = 0; copied = skb->len; @@ -1403,6 +1432,11 @@ static int netlink_recvmsg(struct kiocb *kiocb, struct socket *sock, siocb->scm->creds = *NETLINK_CREDS(skb); if (flags & MSG_TRUNC) copied = skb->len; + +#ifdef CONFIG_COMPAT_NETLINK_MESSAGES + skb_shinfo(skb)->frag_list = frag; +#endif + skb_free_datagram(sk, skb); if (nlk->cb && atomic_read(&sk->sk_rmem_alloc) <= sk->sk_rcvbuf / 2) diff --git a/net/wireless/wext.c b/net/wireless/wext.c index ee35e6422f1f..3fe3c2c0ce11 100644 --- a/net/wireless/wext.c +++ b/net/wireless/wext.c @@ -417,6 +417,21 @@ static const int event_type_size[] = { IW_EV_QUAL_LEN, /* IW_HEADER_TYPE_QUAL */ }; +#ifdef CONFIG_COMPAT +static const int compat_event_type_size[] = { + IW_EV_COMPAT_LCP_LEN, /* IW_HEADER_TYPE_NULL */ + 0, + IW_EV_COMPAT_CHAR_LEN, /* IW_HEADER_TYPE_CHAR */ + 0, + IW_EV_COMPAT_UINT_LEN, /* IW_HEADER_TYPE_UINT */ + IW_EV_COMPAT_FREQ_LEN, /* IW_HEADER_TYPE_FREQ */ + IW_EV_COMPAT_ADDR_LEN, /* IW_HEADER_TYPE_ADDR */ + 0, + IW_EV_COMPAT_POINT_LEN, /* Without variable payload */ + IW_EV_COMPAT_PARAM_LEN, /* IW_HEADER_TYPE_PARAM */ + IW_EV_COMPAT_QUAL_LEN, /* IW_HEADER_TYPE_QUAL */ +}; +#endif /************************ COMMON SUBROUTINES ************************/ /* @@ -1348,6 +1363,22 @@ void wireless_send_event(struct net_device * dev, struct sk_buff *skb; struct nlmsghdr *nlh; struct nlattr *nla; +#ifdef CONFIG_COMPAT + struct __compat_iw_event *compat_event; + struct compat_iw_point compat_wrqu; + struct sk_buff *compskb; +#endif + + /* + * Nothing in the kernel sends scan events with data, be safe. + * This is necessary because we cannot fix up scan event data + * for compat, due to being contained in 'extra', but normally + * applications are required to retrieve the scan data anyway + * and no data is included in the event, this codifies that + * practice. + */ + if (WARN_ON(cmd == SIOCGIWSCAN && extra)) + extra = NULL; /* Get the description of the Event */ if (cmd <= SIOCIWLAST) { @@ -1446,7 +1477,54 @@ void wireless_send_event(struct net_device * dev, memcpy(((char *) event) + hdr_len, extra, extra_len); nlmsg_end(skb, nlh); +#ifdef CONFIG_COMPAT + hdr_len = compat_event_type_size[descr->header_type]; + event_len = hdr_len + extra_len; + compskb = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_ATOMIC); + if (!compskb) { + kfree_skb(skb); + return; + } + + /* Send via the RtNetlink event channel */ + nlh = rtnetlink_ifinfo_prep(dev, compskb); + if (WARN_ON(!nlh)) { + kfree_skb(skb); + kfree_skb(compskb); + return; + } + + /* Add the wireless events in the netlink packet */ + nla = nla_reserve(compskb, IFLA_WIRELESS, event_len); + if (!nla) { + kfree_skb(skb); + kfree_skb(compskb); + return; + } + compat_event = nla_data(nla); + + compat_event->len = event_len; + compat_event->cmd = cmd; + if (descr->header_type == IW_HEADER_TYPE_POINT) { + compat_wrqu.length = wrqu->data.length; + compat_wrqu.flags = wrqu->data.flags; + memcpy(&compat_event->pointer, + ((char *) &compat_wrqu) + IW_EV_COMPAT_POINT_OFF, + hdr_len - IW_EV_COMPAT_LCP_LEN); + if (extra_len) + memcpy(((char *) compat_event) + hdr_len, + extra, extra_len); + } else { + /* extra_len must be zero, so no if (extra) needed */ + memcpy(&compat_event->pointer, wrqu, + hdr_len - IW_EV_COMPAT_LCP_LEN); + } + + nlmsg_end(compskb, nlh); + + skb_shinfo(skb)->frag_list = compskb; +#endif skb_queue_tail(&dev_net(dev)->wext_nlevents, skb); schedule_work(&wireless_nlevent_work); } -- cgit v1.2.3 From 43237b5490e8f2f4679decd660064ff35ce490cc Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Wed, 20 May 2009 18:41:58 +0200 Subject: ext3: Get rid of extenddisksize parameter of ext3_get_blocks_handle() Get rid of extenddisksize parameter of ext3_get_blocks_handle(). This seems to be a relict from some old days and setting disksize in this function does not make much sence. Currently it was set only by ext3_getblk(). Since the parameter has some effect only if create == 1, it is easy to check that the three callers which end up calling ext3_getblk() with create == 1 (ext3_append, ext3_quota_write, ext3_mkdir) do the right thing and set disksize themselves. Signed-off-by: Jan Kara --- fs/ext3/dir.c | 3 +-- fs/ext3/inode.c | 13 +++---------- include/linux/ext3_fs.h | 2 +- 3 files changed, 5 insertions(+), 13 deletions(-) (limited to 'include') diff --git a/fs/ext3/dir.c b/fs/ext3/dir.c index 3d724a95882f..373fa90c796a 100644 --- a/fs/ext3/dir.c +++ b/fs/ext3/dir.c @@ -130,8 +130,7 @@ static int ext3_readdir(struct file * filp, struct buffer_head *bh = NULL; map_bh.b_state = 0; - err = ext3_get_blocks_handle(NULL, inode, blk, 1, - &map_bh, 0, 0); + err = ext3_get_blocks_handle(NULL, inode, blk, 1, &map_bh, 0); if (err > 0) { pgoff_t index = map_bh.b_blocknr >> (PAGE_CACHE_SHIFT - inode->i_blkbits); diff --git a/fs/ext3/inode.c b/fs/ext3/inode.c index 4d7da6f6184b..b49908a167ae 100644 --- a/fs/ext3/inode.c +++ b/fs/ext3/inode.c @@ -788,7 +788,7 @@ err_out: int ext3_get_blocks_handle(handle_t *handle, struct inode *inode, sector_t iblock, unsigned long maxblocks, struct buffer_head *bh_result, - int create, int extend_disksize) + int create) { int err = -EIO; int offsets[4]; @@ -911,13 +911,6 @@ int ext3_get_blocks_handle(handle_t *handle, struct inode *inode, if (!err) err = ext3_splice_branch(handle, inode, iblock, partial, indirect_blks, count); - /* - * i_disksize growing is protected by truncate_mutex. Don't forget to - * protect it if you're about to implement concurrent - * ext3_get_block() -bzzz - */ - if (!err && extend_disksize && inode->i_size > ei->i_disksize) - ei->i_disksize = inode->i_size; mutex_unlock(&ei->truncate_mutex); if (err) goto cleanup; @@ -972,7 +965,7 @@ static int ext3_get_block(struct inode *inode, sector_t iblock, } ret = ext3_get_blocks_handle(handle, inode, iblock, - max_blocks, bh_result, create, 0); + max_blocks, bh_result, create); if (ret > 0) { bh_result->b_size = (ret << inode->i_blkbits); ret = 0; @@ -1005,7 +998,7 @@ struct buffer_head *ext3_getblk(handle_t *handle, struct inode *inode, dummy.b_blocknr = -1000; buffer_trace_init(&dummy.b_history); err = ext3_get_blocks_handle(handle, inode, block, 1, - &dummy, create, 1); + &dummy, create); /* * ext3_get_blocks_handle() returns number of blocks * mapped. 0 in case of a HOLE. diff --git a/include/linux/ext3_fs.h b/include/linux/ext3_fs.h index 634a5e5aba3e..7499b3667798 100644 --- a/include/linux/ext3_fs.h +++ b/include/linux/ext3_fs.h @@ -874,7 +874,7 @@ struct buffer_head * ext3_getblk (handle_t *, struct inode *, long, int, int *); struct buffer_head * ext3_bread (handle_t *, struct inode *, int, int, int *); int ext3_get_blocks_handle(handle_t *handle, struct inode *inode, sector_t iblock, unsigned long maxblocks, struct buffer_head *bh_result, - int create, int extend_disksize); + int create); extern struct inode *ext3_iget(struct super_block *, unsigned long); extern int ext3_write_inode (struct inode *, int); -- cgit v1.2.3 From 5c9228f0cfb09a098a8a380116b42ae099e967b6 Mon Sep 17 00:00:00 2001 From: Johannes Weiner Date: Thu, 16 Jul 2009 16:06:09 +0100 Subject: vt: drop bootmem/slab memory distinction Bootmem is not used for the vt screen buffer anymore as slab is now available at the time the console is initialized. Get rid of the now superfluous distinction between slab and bootmem, it's always slab. This also fixes a kmalloc leak which Catalin described thusly: Commit a5f4f52e ("vt: use kzalloc() instead of the bootmem allocator") replaced the alloc_bootmem() with kzalloc() but didn't set vc_kmalloced to 1 and the memory block is later leaked. The corresponding kmemleak trace: unreferenced object 0xdf828000 (size 8192): comm "swapper", pid 0, jiffies 4294937296 backtrace: [] __save_stack_trace+0x17/0x1c [] log_early+0x55/0x84 [] kmemleak_alloc+0x33/0x3c [] __kmalloc+0xd7/0xe4 [] con_init+0xbf/0x1b8 [] console_init+0x11/0x20 [] start_kernel+0x137/0x1e4 Signed-off-by: Johannes Weiner Reviewed-by: Pekka Enberg Tested-by: Catalin Marinas Signed-off-by: Andrew Morton Signed-off-by: Alan Cox Signed-off-by: Linus Torvalds --- drivers/char/vt.c | 12 +++--------- include/linux/console_struct.h | 1 - 2 files changed, 3 insertions(+), 10 deletions(-) (limited to 'include') diff --git a/drivers/char/vt.c b/drivers/char/vt.c index 7947bd1b4cf7..404f4c1ee431 100644 --- a/drivers/char/vt.c +++ b/drivers/char/vt.c @@ -770,14 +770,12 @@ int vc_allocate(unsigned int currcons) /* return 0 on success */ visual_init(vc, currcons, 1); if (!*vc->vc_uni_pagedir_loc) con_set_default_unimap(vc); - if (!vc->vc_kmalloced) - vc->vc_screenbuf = kmalloc(vc->vc_screenbuf_size, GFP_KERNEL); + vc->vc_screenbuf = kmalloc(vc->vc_screenbuf_size, GFP_KERNEL); if (!vc->vc_screenbuf) { kfree(vc); vc_cons[currcons].d = NULL; return -ENOMEM; } - vc->vc_kmalloced = 1; vc_init(vc, vc->vc_rows, vc->vc_cols, 1); vcs_make_sysfs(currcons); atomic_notifier_call_chain(&vt_notifier_list, VT_ALLOCATE, ¶m); @@ -913,10 +911,8 @@ static int vc_do_resize(struct tty_struct *tty, struct vc_data *vc, if (new_scr_end > new_origin) scr_memsetw((void *)new_origin, vc->vc_video_erase_char, new_scr_end - new_origin); - if (vc->vc_kmalloced) - kfree(vc->vc_screenbuf); + kfree(vc->vc_screenbuf); vc->vc_screenbuf = newscreen; - vc->vc_kmalloced = 1; vc->vc_screenbuf_size = new_screen_size; set_origin(vc); @@ -995,8 +991,7 @@ void vc_deallocate(unsigned int currcons) vc->vc_sw->con_deinit(vc); put_pid(vc->vt_pid); module_put(vc->vc_sw->owner); - if (vc->vc_kmalloced) - kfree(vc->vc_screenbuf); + kfree(vc->vc_screenbuf); if (currcons >= MIN_NR_CONSOLES) kfree(vc); vc_cons[currcons].d = NULL; @@ -2881,7 +2876,6 @@ static int __init con_init(void) INIT_WORK(&vc_cons[currcons].SAK_work, vc_SAK); visual_init(vc, currcons, 1); vc->vc_screenbuf = kzalloc(vc->vc_screenbuf_size, GFP_NOWAIT); - vc->vc_kmalloced = 0; vc_init(vc, vc->vc_rows, vc->vc_cols, currcons || !vc->vc_sw->con_save_screen); } diff --git a/include/linux/console_struct.h b/include/linux/console_struct.h index d71f7c0f931b..38fe59dc89ae 100644 --- a/include/linux/console_struct.h +++ b/include/linux/console_struct.h @@ -89,7 +89,6 @@ struct vc_data { unsigned int vc_need_wrap : 1; unsigned int vc_can_do_color : 1; unsigned int vc_report_mouse : 2; - unsigned int vc_kmalloced : 1; unsigned char vc_utf : 1; /* Unicode UTF-8 encoding */ unsigned char vc_utf_count; int vc_utf_char; -- cgit v1.2.3 From 5bb459bb45d1ad3c177485dcf0af01580aa31125 Mon Sep 17 00:00:00 2001 From: Oleg Nesterov Date: Fri, 10 Jul 2009 03:48:23 +0200 Subject: kernel: rename is_single_threaded(task) to current_is_single_threaded(void) - is_single_threaded(task) is not safe unless task == current, we can't use task->signal or task->mm. - it doesn't make sense unless task == current, the task can fork right after the check. Rename it to current_is_single_threaded() and kill the argument. Signed-off-by: Oleg Nesterov Acked-by: David Howells Signed-off-by: James Morris --- include/linux/sched.h | 2 +- lib/is_single_threaded.c | 3 ++- security/keys/process_keys.c | 2 +- security/selinux/hooks.c | 2 +- 4 files changed, 5 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/include/linux/sched.h b/include/linux/sched.h index 16a982e389fb..0839a2c9b952 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -2075,7 +2075,7 @@ static inline unsigned long wait_task_inactive(struct task_struct *p, #define for_each_process(p) \ for (p = &init_task ; (p = next_task(p)) != &init_task ; ) -extern bool is_single_threaded(struct task_struct *); +extern bool current_is_single_threaded(void); /* * Careful: do_each_thread/while_each_thread is a double loop so diff --git a/lib/is_single_threaded.c b/lib/is_single_threaded.c index 2762516e0a5e..434010980bdf 100644 --- a/lib/is_single_threaded.c +++ b/lib/is_single_threaded.c @@ -15,8 +15,9 @@ /* * Returns true if the task does not share ->mm with another thread/process. */ -bool is_single_threaded(struct task_struct *task) +bool current_is_single_threaded(void) { + struct task_struct *task = current; struct mm_struct *mm = task->mm; struct task_struct *p, *t; bool ret; diff --git a/security/keys/process_keys.c b/security/keys/process_keys.c index 276d27882ce8..ed929af466d3 100644 --- a/security/keys/process_keys.c +++ b/security/keys/process_keys.c @@ -702,7 +702,7 @@ long join_session_keyring(const char *name) /* only permit this if there's a single thread in the thread group - * this avoids us having to adjust the creds on all threads and risking * ENOMEM */ - if (!is_single_threaded(current)) + if (!current_is_single_threaded()) return -EMLINK; new = prepare_creds(); diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c index 2081055f6783..e65677da36bd 100644 --- a/security/selinux/hooks.c +++ b/security/selinux/hooks.c @@ -5187,7 +5187,7 @@ static int selinux_setprocattr(struct task_struct *p, /* Only allow single threaded processes to change context */ error = -EPERM; - if (!is_single_threaded(p)) { + if (!current_is_single_threaded()) { error = security_bounded_transition(tsec->sid, sid); if (error) goto abort_change; -- cgit v1.2.3 From 4dc6dc7162c08b9965163c9ab3f9375d4adff2c7 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 15 Jul 2009 23:13:10 +0000 Subject: net: sock_copy() fixes Commit e912b1142be8f1e2c71c71001dc992c6e5eb2ec1 (net: sk_prot_alloc() should not blindly overwrite memory) took care of not zeroing whole new socket at allocation time. sock_copy() is another spot where we should be very careful. We should not set refcnt to a non null value, until we are sure other fields are correctly setup, or a lockless reader could catch this socket by mistake, while not fully (re)initialized. This patch puts sk_node & sk_refcnt to the very beginning of struct sock to ease sock_copy() & sk_prot_alloc() job. We add appropriate smp_wmb() before sk_refcnt initializations to match our RCU requirements (changes to sock keys should be committed to memory before sk_refcnt setting) Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- include/net/sock.h | 32 +++++++++++++++++++------------- net/core/sock.c | 20 ++++++++++++++++++-- 2 files changed, 37 insertions(+), 15 deletions(-) (limited to 'include') diff --git a/include/net/sock.h b/include/net/sock.h index 2c0da9239b95..950409dcec3d 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -104,15 +104,15 @@ struct net; /** * struct sock_common - minimal network layer representation of sockets + * @skc_node: main hash linkage for various protocol lookup tables + * @skc_nulls_node: main hash linkage for UDP/UDP-Lite protocol + * @skc_refcnt: reference count + * @skc_hash: hash value used with various protocol lookup tables * @skc_family: network address family * @skc_state: Connection state * @skc_reuse: %SO_REUSEADDR setting * @skc_bound_dev_if: bound device index if != 0 - * @skc_node: main hash linkage for various protocol lookup tables - * @skc_nulls_node: main hash linkage for UDP/UDP-Lite protocol * @skc_bind_node: bind hash linkage for various protocol lookup tables - * @skc_refcnt: reference count - * @skc_hash: hash value used with various protocol lookup tables * @skc_prot: protocol handlers inside a network family * @skc_net: reference to the network namespace of this socket * @@ -120,17 +120,21 @@ struct net; * for struct sock and struct inet_timewait_sock. */ struct sock_common { - unsigned short skc_family; - volatile unsigned char skc_state; - unsigned char skc_reuse; - int skc_bound_dev_if; + /* + * first fields are not copied in sock_copy() + */ union { struct hlist_node skc_node; struct hlist_nulls_node skc_nulls_node; }; - struct hlist_node skc_bind_node; atomic_t skc_refcnt; + unsigned int skc_hash; + unsigned short skc_family; + volatile unsigned char skc_state; + unsigned char skc_reuse; + int skc_bound_dev_if; + struct hlist_node skc_bind_node; struct proto *skc_prot; #ifdef CONFIG_NET_NS struct net *skc_net; @@ -208,15 +212,17 @@ struct sock { * don't add nothing before this first member (__sk_common) --acme */ struct sock_common __sk_common; +#define sk_node __sk_common.skc_node +#define sk_nulls_node __sk_common.skc_nulls_node +#define sk_refcnt __sk_common.skc_refcnt + +#define sk_copy_start __sk_common.skc_hash +#define sk_hash __sk_common.skc_hash #define sk_family __sk_common.skc_family #define sk_state __sk_common.skc_state #define sk_reuse __sk_common.skc_reuse #define sk_bound_dev_if __sk_common.skc_bound_dev_if -#define sk_node __sk_common.skc_node -#define sk_nulls_node __sk_common.skc_nulls_node #define sk_bind_node __sk_common.skc_bind_node -#define sk_refcnt __sk_common.skc_refcnt -#define sk_hash __sk_common.skc_hash #define sk_prot __sk_common.skc_prot #define sk_net __sk_common.skc_net kmemcheck_bitfield_begin(flags); diff --git a/net/core/sock.c b/net/core/sock.c index ba5d2116aea1..d9eec153d531 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -919,13 +919,19 @@ static inline void sock_lock_init(struct sock *sk) af_family_keys + sk->sk_family); } +/* + * Copy all fields from osk to nsk but nsk->sk_refcnt must not change yet, + * even temporarly, because of RCU lookups. sk_node should also be left as is. + */ static void sock_copy(struct sock *nsk, const struct sock *osk) { #ifdef CONFIG_SECURITY_NETWORK void *sptr = nsk->sk_security; #endif - - memcpy(nsk, osk, osk->sk_prot->obj_size); + BUILD_BUG_ON(offsetof(struct sock, sk_copy_start) != + sizeof(osk->sk_node) + sizeof(osk->sk_refcnt)); + memcpy(&nsk->sk_copy_start, &osk->sk_copy_start, + osk->sk_prot->obj_size - offsetof(struct sock, sk_copy_start)); #ifdef CONFIG_SECURITY_NETWORK nsk->sk_security = sptr; security_sk_clone(osk, nsk); @@ -1140,6 +1146,11 @@ struct sock *sk_clone(const struct sock *sk, const gfp_t priority) newsk->sk_err = 0; newsk->sk_priority = 0; + /* + * Before updating sk_refcnt, we must commit prior changes to memory + * (Documentation/RCU/rculist_nulls.txt for details) + */ + smp_wmb(); atomic_set(&newsk->sk_refcnt, 2); /* @@ -1855,6 +1866,11 @@ void sock_init_data(struct socket *sock, struct sock *sk) sk->sk_stamp = ktime_set(-1L, 0); + /* + * Before updating sk_refcnt, we must commit prior changes to memory + * (Documentation/RCU/rculist_nulls.txt for details) + */ + smp_wmb(); atomic_set(&sk->sk_refcnt, 1); atomic_set(&sk->sk_wmem_alloc, 1); atomic_set(&sk->sk_drops, 0); -- cgit v1.2.3 From 5780888bcac316508eb5f4dd23bbea8b5057647c Mon Sep 17 00:00:00 2001 From: Matias Zabaljauregui Date: Thu, 18 Jun 2009 11:44:06 -0300 Subject: lguest: fix journey fix: "make Guest" was complaining about duplicated G:032 Signed-off-by: Matias Zabaljauregui Signed-off-by: Rusty Russell --- arch/x86/include/asm/lguest_hcall.h | 2 +- arch/x86/lguest/boot.c | 2 +- include/linux/lguest.h | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/arch/x86/include/asm/lguest_hcall.h b/arch/x86/include/asm/lguest_hcall.h index d31c4a684078..33600a66755f 100644 --- a/arch/x86/include/asm/lguest_hcall.h +++ b/arch/x86/include/asm/lguest_hcall.h @@ -30,7 +30,7 @@ #include #include -/*G:031 But first, how does our Guest contact the Host to ask for privileged +/*G:030 But first, how does our Guest contact the Host to ask for privileged * operations? There are two ways: the direct way is to make a "hypercall", * to make requests of the Host Itself. * diff --git a/arch/x86/lguest/boot.c b/arch/x86/lguest/boot.c index 7bc65f0f62c4..0188fd37b6c0 100644 --- a/arch/x86/lguest/boot.c +++ b/arch/x86/lguest/boot.c @@ -1079,7 +1079,7 @@ static unsigned lguest_patch(u8 type, u16 clobber, void *ibuf, return insn_len; } -/*G:030 Once we get to lguest_init(), we know we're a Guest. The various +/*G:029 Once we get to lguest_init(), we know we're a Guest. The various * pv_ops structures in the kernel provide points for (almost) every routine we * have to override to avoid privileged instructions. */ __init void lguest_init(void) diff --git a/include/linux/lguest.h b/include/linux/lguest.h index 7bc1440fc473..dbf2479e808e 100644 --- a/include/linux/lguest.h +++ b/include/linux/lguest.h @@ -11,7 +11,7 @@ #define LG_CLOCK_MIN_DELTA 100UL #define LG_CLOCK_MAX_DELTA ULONG_MAX -/*G:032 The second method of communicating with the Host is to via "struct +/*G:031 The second method of communicating with the Host is to via "struct * lguest_data". Once the Guest's initialization hypercall tells the Host where * this is, the Guest and Host both publish information in it. :*/ struct lguest_data -- cgit v1.2.3 From e79f07e2925b10f09a2621650c16f3d6ea778747 Mon Sep 17 00:00:00 2001 From: Alex Williamson Date: Tue, 7 Jul 2009 08:47:10 -0600 Subject: virtio_net: Sync header with qemu Qemu added support for a few extra RX modes that Linux doesn't currently make use of. Sync the headers to maintain consistency. Signed-off-by: Alex Williamson Signed-off-by: Rusty Russell --- include/linux/virtio_net.h | 14 ++++++++++---- 1 file changed, 10 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/include/linux/virtio_net.h b/include/linux/virtio_net.h index cec79adbe3ea..9c543d6ac535 100644 --- a/include/linux/virtio_net.h +++ b/include/linux/virtio_net.h @@ -27,6 +27,7 @@ #define VIRTIO_NET_F_CTRL_VQ 17 /* Control channel available */ #define VIRTIO_NET_F_CTRL_RX 18 /* Control channel RX mode support */ #define VIRTIO_NET_F_CTRL_VLAN 19 /* Control channel VLAN filtering */ +#define VIRTIO_NET_F_CTRL_RX_EXTRA 20 /* Extra RX mode control support */ #define VIRTIO_NET_S_LINK_UP 1 /* Link is up */ @@ -81,14 +82,19 @@ typedef __u8 virtio_net_ctrl_ack; #define VIRTIO_NET_ERR 1 /* - * Control the RX mode, ie. promisucous and allmulti. PROMISC and - * ALLMULTI commands require an "out" sg entry containing a 1 byte - * state value, zero = disable, non-zero = enable. These commands - * are supported with the VIRTIO_NET_F_CTRL_RX feature. + * Control the RX mode, ie. promisucous, allmulti, etc... + * All commands require an "out" sg entry containing a 1 byte + * state value, zero = disable, non-zero = enable. Commands + * 0 and 1 are supported with the VIRTIO_NET_F_CTRL_RX feature. + * Commands 2-5 are added with VIRTIO_NET_F_CTRL_RX_EXTRA. */ #define VIRTIO_NET_CTRL_RX 0 #define VIRTIO_NET_CTRL_RX_PROMISC 0 #define VIRTIO_NET_CTRL_RX_ALLMULTI 1 + #define VIRTIO_NET_CTRL_RX_ALLUNI 2 + #define VIRTIO_NET_CTRL_RX_NOMULTI 3 + #define VIRTIO_NET_CTRL_RX_NOUNI 4 + #define VIRTIO_NET_CTRL_RX_NOBCAST 5 /* * Control the MAC filter table. -- cgit v1.2.3 From e36aa25a533962b08402530e8443ac804a454e27 Mon Sep 17 00:00:00 2001 From: Sridhar Samudrala Date: Tue, 14 Jul 2009 14:21:04 +0000 Subject: tun: Allow tap device to send/receive UFO packets. - Allow setting UFO on tap device and handle UFO packets. Signed-off-by: Sridhar Samudrala --------------------------------------------------------- Signed-off-by: David S. Miller --- drivers/net/tun.c | 13 ++++++++++++- include/linux/if_tun.h | 1 + 2 files changed, 13 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/drivers/net/tun.c b/drivers/net/tun.c index dfc1054e4cbd..a998b6a9c245 100644 --- a/drivers/net/tun.c +++ b/drivers/net/tun.c @@ -641,6 +641,9 @@ static __inline__ ssize_t tun_get_user(struct tun_struct *tun, case VIRTIO_NET_HDR_GSO_TCPV6: skb_shinfo(skb)->gso_type = SKB_GSO_TCPV6; break; + case VIRTIO_NET_HDR_GSO_UDP: + skb_shinfo(skb)->gso_type = SKB_GSO_UDP; + break; default: tun->dev->stats.rx_frame_errors++; kfree_skb(skb); @@ -726,6 +729,8 @@ static __inline__ ssize_t tun_put_user(struct tun_struct *tun, gso.gso_type = VIRTIO_NET_HDR_GSO_TCPV4; else if (sinfo->gso_type & SKB_GSO_TCPV6) gso.gso_type = VIRTIO_NET_HDR_GSO_TCPV6; + else if (sinfo->gso_type & SKB_GSO_UDP) + gso.gso_type = VIRTIO_NET_HDR_GSO_UDP; else BUG(); if (sinfo->gso_type & SKB_GSO_TCP_ECN) @@ -1073,7 +1078,8 @@ static int set_offload(struct net_device *dev, unsigned long arg) old_features = dev->features; /* Unset features, set them as we chew on the arg. */ features = (old_features & ~(NETIF_F_HW_CSUM|NETIF_F_SG|NETIF_F_FRAGLIST - |NETIF_F_TSO_ECN|NETIF_F_TSO|NETIF_F_TSO6)); + |NETIF_F_TSO_ECN|NETIF_F_TSO|NETIF_F_TSO6 + |NETIF_F_UFO)); if (arg & TUN_F_CSUM) { features |= NETIF_F_HW_CSUM|NETIF_F_SG|NETIF_F_FRAGLIST; @@ -1090,6 +1096,11 @@ static int set_offload(struct net_device *dev, unsigned long arg) features |= NETIF_F_TSO6; arg &= ~(TUN_F_TSO4|TUN_F_TSO6); } + + if (arg & TUN_F_UFO) { + features |= NETIF_F_UFO; + arg &= ~TUN_F_UFO; + } } /* This gives the user a way to test for new features in future by diff --git a/include/linux/if_tun.h b/include/linux/if_tun.h index 915ba5789f0e..3f5fd523b49d 100644 --- a/include/linux/if_tun.h +++ b/include/linux/if_tun.h @@ -62,6 +62,7 @@ #define TUN_F_TSO4 0x02 /* I can handle TSO for IPv4 packets */ #define TUN_F_TSO6 0x04 /* I can handle TSO for IPv6 packets */ #define TUN_F_TSO_ECN 0x08 /* I can handle TSO with ECN bits. */ +#define TUN_F_UFO 0x10 /* I can handle UFO packets */ /* Protocol info prepended to the packets (when IFF_NO_PI is not set) */ #define TUN_PKT_STRIP 0x0001 -- cgit v1.2.3 From 0741241c6b80bfd58417e95de984d60c9e9ef2a0 Mon Sep 17 00:00:00 2001 From: Mike Frysinger Date: Fri, 17 Jul 2009 10:13:21 -0700 Subject: connector: make callback argument type explicit The connector documentation states that the argument to the callback function is always a pointer to a struct cn_msg, but rather than encode it in the API itself, it uses a void pointer everywhere. This doesn't make much sense to encode the pointer in documentation as it prevents proper C type checking from occurring and can easily allow people to use the wrong pointer type. So convert the argument type to an explicit struct cn_msg pointer. Signed-off-by: Mike Frysinger Signed-off-by: David S. Miller --- Documentation/connector/cn_test.c | 4 +--- drivers/connector/cn_proc.c | 3 +-- drivers/connector/cn_queue.c | 7 +++++-- drivers/connector/connector.c | 6 +++--- drivers/staging/dst/dcore.c | 3 +-- drivers/video/uvesafb.c | 3 +-- drivers/w1/w1_netlink.c | 3 +-- include/linux/connector.h | 6 +++--- 8 files changed, 16 insertions(+), 19 deletions(-) (limited to 'include') diff --git a/Documentation/connector/cn_test.c b/Documentation/connector/cn_test.c index f688eba87704..50d5ce4899c8 100644 --- a/Documentation/connector/cn_test.c +++ b/Documentation/connector/cn_test.c @@ -32,10 +32,8 @@ static char cn_test_name[] = "cn_test"; static struct sock *nls; static struct timer_list cn_test_timer; -void cn_test_callback(void *data) +void cn_test_callback(struct cn_msg *msg) { - struct cn_msg *msg = (struct cn_msg *)data; - printk("%s: %lu: idx=%x, val=%x, seq=%u, ack=%u, len=%d: %s.\n", __func__, jiffies, msg->id.idx, msg->id.val, msg->seq, msg->ack, msg->len, (char *)msg->data); diff --git a/drivers/connector/cn_proc.c b/drivers/connector/cn_proc.c index c5afc98e2675..85e5dc0431fe 100644 --- a/drivers/connector/cn_proc.c +++ b/drivers/connector/cn_proc.c @@ -202,9 +202,8 @@ static void cn_proc_ack(int err, int rcvd_seq, int rcvd_ack) * cn_proc_mcast_ctl * @data: message sent from userspace via the connector */ -static void cn_proc_mcast_ctl(void *data) +static void cn_proc_mcast_ctl(struct cn_msg *msg) { - struct cn_msg *msg = data; enum proc_cn_mcast_op *mc_op = NULL; int err = 0; diff --git a/drivers/connector/cn_queue.c b/drivers/connector/cn_queue.c index c769ef269fb5..d478aefcd3be 100644 --- a/drivers/connector/cn_queue.c +++ b/drivers/connector/cn_queue.c @@ -87,7 +87,9 @@ void cn_queue_wrapper(struct work_struct *work) kfree(d->free); } -static struct cn_callback_entry *cn_queue_alloc_callback_entry(char *name, struct cb_id *id, void (*callback)(void *)) +static struct cn_callback_entry * +cn_queue_alloc_callback_entry(char *name, struct cb_id *id, + void (*callback)(struct cn_msg *)) { struct cn_callback_entry *cbq; @@ -120,7 +122,8 @@ int cn_cb_equal(struct cb_id *i1, struct cb_id *i2) return ((i1->idx == i2->idx) && (i1->val == i2->val)); } -int cn_queue_add_callback(struct cn_queue_dev *dev, char *name, struct cb_id *id, void (*callback)(void *)) +int cn_queue_add_callback(struct cn_queue_dev *dev, char *name, struct cb_id *id, + void (*callback)(struct cn_msg *)) { struct cn_callback_entry *cbq, *__cbq; int found = 0; diff --git a/drivers/connector/connector.c b/drivers/connector/connector.c index fd336c5a9057..3f45669f5d76 100644 --- a/drivers/connector/connector.c +++ b/drivers/connector/connector.c @@ -269,7 +269,8 @@ static void cn_notify(struct cb_id *id, u32 notify_event) * * May sleep. */ -int cn_add_callback(struct cb_id *id, char *name, void (*callback)(void *)) +int cn_add_callback(struct cb_id *id, char *name, + void (*callback)(struct cn_msg *)) { int err; struct cn_dev *dev = &cdev; @@ -351,9 +352,8 @@ static int cn_ctl_msg_equals(struct cn_ctl_msg *m1, struct cn_ctl_msg *m2) * * Used for notification of a request's processing. */ -static void cn_callback(void *data) +static void cn_callback(struct cn_msg *msg) { - struct cn_msg *msg = data; struct cn_ctl_msg *ctl; struct cn_ctl_entry *ent; u32 size; diff --git a/drivers/staging/dst/dcore.c b/drivers/staging/dst/dcore.c index fad25b753042..84724187ec3e 100644 --- a/drivers/staging/dst/dcore.c +++ b/drivers/staging/dst/dcore.c @@ -846,10 +846,9 @@ static dst_command_func dst_commands[] = { /* * Configuration parser. */ -static void cn_dst_callback(void *data) +static void cn_dst_callback(struct cn_msg *msg) { struct dst_ctl *ctl; - struct cn_msg *msg = data; int err; struct dst_ctl_ack ack; struct dst_node *n = NULL, *tmp; diff --git a/drivers/video/uvesafb.c b/drivers/video/uvesafb.c index ca5b4643a401..e98baf6916b8 100644 --- a/drivers/video/uvesafb.c +++ b/drivers/video/uvesafb.c @@ -67,9 +67,8 @@ static DEFINE_MUTEX(uvfb_lock); * find the kernel part of the task struct, copy the registers and * the buffer contents and then complete the task. */ -static void uvesafb_cn_callback(void *data) +static void uvesafb_cn_callback(struct cn_msg *msg) { - struct cn_msg *msg = data; struct uvesafb_task *utask; struct uvesafb_ktask *task; diff --git a/drivers/w1/w1_netlink.c b/drivers/w1/w1_netlink.c index fdf72851c574..52ccb3d3a963 100644 --- a/drivers/w1/w1_netlink.c +++ b/drivers/w1/w1_netlink.c @@ -306,9 +306,8 @@ static int w1_netlink_send_error(struct cn_msg *rcmsg, struct w1_netlink_msg *rm return error; } -static void w1_cn_callback(void *data) +static void w1_cn_callback(struct cn_msg *msg) { - struct cn_msg *msg = data; struct w1_netlink_msg *m = (struct w1_netlink_msg *)(msg + 1); struct w1_netlink_cmd *cmd; struct w1_slave *sl; diff --git a/include/linux/connector.h b/include/linux/connector.h index b68d27850d51..47ebf416f512 100644 --- a/include/linux/connector.h +++ b/include/linux/connector.h @@ -136,7 +136,7 @@ struct cn_callback_data { void *ddata; void *callback_priv; - void (*callback) (void *); + void (*callback) (struct cn_msg *); void *free; }; @@ -167,11 +167,11 @@ struct cn_dev { struct cn_queue_dev *cbdev; }; -int cn_add_callback(struct cb_id *, char *, void (*callback) (void *)); +int cn_add_callback(struct cb_id *, char *, void (*callback) (struct cn_msg *)); void cn_del_callback(struct cb_id *); int cn_netlink_send(struct cn_msg *, u32, gfp_t); -int cn_queue_add_callback(struct cn_queue_dev *dev, char *name, struct cb_id *id, void (*callback)(void *)); +int cn_queue_add_callback(struct cn_queue_dev *dev, char *name, struct cb_id *id, void (*callback)(struct cn_msg *)); void cn_queue_del_callback(struct cn_queue_dev *dev, struct cb_id *id); int queue_cn_work(struct cn_callback_entry *cbq, struct work_struct *work); -- cgit v1.2.3 From 04e448d9a386640a79a4aa71251aa1cdd314f662 Mon Sep 17 00:00:00 2001 From: Tim Abbott Date: Sun, 12 Jul 2009 18:23:33 -0400 Subject: vmlinux.lds.h: restructure BSS linker script macros The BSS section macros in vmlinux.lds.h currently place the .sbss input section outside the bounds of [__bss_start, __bss_end]. On all architectures except for microblaze that handle both .sbss and __bss_start/__bss_end, this is wrong: the .sbss input section is within the range [__bss_start, __bss_end]. Relatedly, the example code at the top of the file actually has __bss_start/__bss_end defined twice; I believe the right fix here is to define them in the BSS_SECTION macro but not in the BSS macro. Another problem with the current macros is that several architectures have an ALIGN(4) or some other small number just before __bss_stop in their linker scripts. The BSS_SECTION macro currently hardcodes this to 4; while it should really be an argument. It also ignores its sbss_align argument; fix that. mn10300 is the only user at present of any of the macros touched by this patch. It looks like mn10300 actually was incorrectly converted to use the new BSS() macro (the alignment of 4 prior to conversion was a __bss_stop alignment, but the argument to the BSS macro is a start alignment). So fix this as well. I'd like acks from Sam and David on this one. Also CCing Paul, since he has a patch from me which will need to be updated to use BSS_SECTION(0, PAGE_SIZE, 4) once this gets merged. Signed-off-by: Tim Abbott Cc: Paul Mundt Cc: David Howells Signed-off-by: Sam Ravnborg --- arch/mn10300/kernel/vmlinux.lds.S | 2 +- include/asm-generic/vmlinux.lds.h | 19 +++++++++---------- 2 files changed, 10 insertions(+), 11 deletions(-) (limited to 'include') diff --git a/arch/mn10300/kernel/vmlinux.lds.S b/arch/mn10300/kernel/vmlinux.lds.S index c96ba3da95ac..f4aa07934654 100644 --- a/arch/mn10300/kernel/vmlinux.lds.S +++ b/arch/mn10300/kernel/vmlinux.lds.S @@ -107,7 +107,7 @@ SECTIONS __init_end = .; /* freed after init ends here */ - BSS(4) + BSS_SECTION(0, PAGE_SIZE, 4) _end = . ; diff --git a/include/asm-generic/vmlinux.lds.h b/include/asm-generic/vmlinux.lds.h index a553f1041cf1..6ad76bf5fb40 100644 --- a/include/asm-generic/vmlinux.lds.h +++ b/include/asm-generic/vmlinux.lds.h @@ -30,9 +30,7 @@ * EXCEPTION_TABLE(...) * NOTES * - * __bss_start = .; - * BSS_SECTION(0, 0) - * __bss_stop = .; + * BSS_SECTION(0, 0, 0) * _end = .; * * /DISCARD/ : { @@ -489,7 +487,8 @@ * bss (Block Started by Symbol) - uninitialized data * zeroed during startup */ -#define SBSS \ +#define SBSS(sbss_align) \ + . = ALIGN(sbss_align); \ .sbss : AT(ADDR(.sbss) - LOAD_OFFSET) { \ *(.sbss) \ *(.scommon) \ @@ -498,12 +497,10 @@ #define BSS(bss_align) \ . = ALIGN(bss_align); \ .bss : AT(ADDR(.bss) - LOAD_OFFSET) { \ - VMLINUX_SYMBOL(__bss_start) = .; \ *(.bss.page_aligned) \ *(.dynbss) \ *(.bss) \ *(COMMON) \ - VMLINUX_SYMBOL(__bss_stop) = .; \ } /* @@ -735,8 +732,10 @@ INIT_RAM_FS \ } -#define BSS_SECTION(sbss_align, bss_align) \ - SBSS \ +#define BSS_SECTION(sbss_align, bss_align, stop_align) \ + . = ALIGN(sbss_align); \ + VMLINUX_SYMBOL(__bss_start) = .; \ + SBSS(sbss_align) \ BSS(bss_align) \ - . = ALIGN(4); - + . = ALIGN(stop_align); \ + VMLINUX_SYMBOL(__bss_stop) = .; -- cgit v1.2.3 From 6301cb95c119ebf324bb96ee226fa9ddffad80a7 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Fri, 17 Jul 2009 14:15:47 +0200 Subject: sched: fix nr_uninterruptible accounting of frozen tasks really commit e3c8ca8336 (sched: do not count frozen tasks toward load) broke the nr_uninterruptible accounting on freeze/thaw. On freeze the task is excluded from accounting with a check for (task->flags & PF_FROZEN), but that flag is cleared before the task is thawed. So while we prevent that the task with state TASK_UNINTERRUPTIBLE is accounted to nr_uninterruptible on freeze we decrement nr_uninterruptible on thaw. Use a separate flag which is handled by the freezing task itself. Set it before calling the scheduler with TASK_UNINTERRUPTIBLE state and clear it after we return from frozen state. Cc: Signed-off-by: Thomas Gleixner --- include/linux/sched.h | 3 ++- kernel/freezer.c | 7 +++++++ 2 files changed, 9 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/sched.h b/include/linux/sched.h index 16a982e389fb..3ab08e4bb6b8 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -209,7 +209,7 @@ extern unsigned long long time_sync_thresh; ((task->state & (__TASK_STOPPED | __TASK_TRACED)) != 0) #define task_contributes_to_load(task) \ ((task->state & TASK_UNINTERRUPTIBLE) != 0 && \ - (task->flags & PF_FROZEN) == 0) + (task->flags & PF_FREEZING) == 0) #define __set_task_state(tsk, state_value) \ do { (tsk)->state = (state_value); } while (0) @@ -1680,6 +1680,7 @@ extern cputime_t task_gtime(struct task_struct *p); #define PF_MEMALLOC 0x00000800 /* Allocating memory */ #define PF_FLUSHER 0x00001000 /* responsible for disk writeback */ #define PF_USED_MATH 0x00002000 /* if unset the fpu must be initialized before use */ +#define PF_FREEZING 0x00004000 /* freeze in progress. do not account to load */ #define PF_NOFREEZE 0x00008000 /* this thread should not be frozen */ #define PF_FROZEN 0x00010000 /* frozen for system suspend */ #define PF_FSTRANS 0x00020000 /* inside a filesystem transaction */ diff --git a/kernel/freezer.c b/kernel/freezer.c index 2f4936cf7083..bd1d42b17cb2 100644 --- a/kernel/freezer.c +++ b/kernel/freezer.c @@ -44,12 +44,19 @@ void refrigerator(void) recalc_sigpending(); /* We sent fake signal, clean it up */ spin_unlock_irq(¤t->sighand->siglock); + /* prevent accounting of that task to load */ + current->flags |= PF_FREEZING; + for (;;) { set_current_state(TASK_UNINTERRUPTIBLE); if (!frozen(current)) break; schedule(); } + + /* Remove the accounting blocker */ + current->flags &= ~PF_FREEZING; + pr_debug("%s left refrigerator\n", current->comm); __set_current_state(save); } -- cgit v1.2.3 From e09758fae8ccde97e026c704319eaa18d488dc86 Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Thu, 16 Jul 2009 15:44:29 +0200 Subject: sched: Cover the CONFIG_DEBUG_SPINLOCK_SLEEP off-case for __might_sleep() Cover the off case for __might_sleep(), so that we avoid #ifdefs in files that make use of it. Especially, this prepares for the __might_sleep() pull up on cond_resched(). Signed-off-by: Frederic Weisbecker Signed-off-by: Peter Zijlstra LKML-Reference: <1247725694-6082-3-git-send-email-fweisbec@gmail.com> Signed-off-by: Ingo Molnar --- include/linux/kernel.h | 1 + kernel/sched.c | 3 +-- 2 files changed, 2 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/include/linux/kernel.h b/include/linux/kernel.h index d6320a3e8def..b804f694ae6a 100644 --- a/include/linux/kernel.h +++ b/include/linux/kernel.h @@ -139,6 +139,7 @@ extern int _cond_resched(void); # define might_sleep() \ do { __might_sleep(__FILE__, __LINE__); might_resched(); } while (0) #else + static inline void __might_sleep(char *file, int line) { } # define might_sleep() do { might_resched(); } while (0) #endif diff --git a/kernel/sched.c b/kernel/sched.c index 4d39e96044b9..370a6c31c5e1 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -6610,9 +6610,8 @@ static inline int should_resched(void) static void __cond_resched(void) { -#ifdef CONFIG_DEBUG_SPINLOCK_SLEEP __might_sleep(__FILE__, __LINE__); -#endif + add_preempt_count(PREEMPT_ACTIVE); schedule(); sub_preempt_count(PREEMPT_ACTIVE); -- cgit v1.2.3 From e4aafea2d4bde8b44d6500c4ee7195bbfc51269e Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Thu, 16 Jul 2009 15:44:29 +0200 Subject: sched: Add a preempt count base offset to __might_sleep() Add a preempt count base offset to compare against the current preempt level count. It prepares to pull up the might_sleep check from cond_resched() to cond_resched_lock() and cond_resched_bh(). For these two helpers, we need to respectively ensure that once we'll unlock the given spinlock / reenable local softirqs, we will reach a sleepable state. Signed-off-by: Frederic Weisbecker [ Move and rename preempt_count_equals() ] Signed-off-by: Peter Zijlstra LKML-Reference: <1247725694-6082-4-git-send-email-fweisbec@gmail.com> Signed-off-by: Ingo Molnar --- include/linux/kernel.h | 6 +++--- kernel/sched.c | 15 +++++++++++---- 2 files changed, 14 insertions(+), 7 deletions(-) (limited to 'include') diff --git a/include/linux/kernel.h b/include/linux/kernel.h index b804f694ae6a..2b5b1e0899a8 100644 --- a/include/linux/kernel.h +++ b/include/linux/kernel.h @@ -125,7 +125,7 @@ extern int _cond_resched(void); #endif #ifdef CONFIG_DEBUG_SPINLOCK_SLEEP - void __might_sleep(char *file, int line); + void __might_sleep(char *file, int line, int preempt_offset); /** * might_sleep - annotation for functions that can sleep * @@ -137,9 +137,9 @@ extern int _cond_resched(void); * supposed to. */ # define might_sleep() \ - do { __might_sleep(__FILE__, __LINE__); might_resched(); } while (0) + do { __might_sleep(__FILE__, __LINE__, 0); might_resched(); } while (0) #else - static inline void __might_sleep(char *file, int line) { } + static inline void __might_sleep(char *file, int line, int preempt_offset) { } # define might_sleep() do { might_resched(); } while (0) #endif diff --git a/kernel/sched.c b/kernel/sched.c index 370a6c31c5e1..3ff4d004bd95 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -6610,7 +6610,7 @@ static inline int should_resched(void) static void __cond_resched(void) { - __might_sleep(__FILE__, __LINE__); + __might_sleep(__FILE__, __LINE__, 0); add_preempt_count(PREEMPT_ACTIVE); schedule(); @@ -9429,13 +9429,20 @@ void __init sched_init(void) } #ifdef CONFIG_DEBUG_SPINLOCK_SLEEP -void __might_sleep(char *file, int line) +static inline int preempt_count_equals(int preempt_offset) +{ + int nested = preempt_count() & ~PREEMPT_ACTIVE; + + return (nested == PREEMPT_INATOMIC_BASE + preempt_offset); +} + +void __might_sleep(char *file, int line, int preempt_offset) { #ifdef in_atomic static unsigned long prev_jiffy; /* ratelimiting */ - if ((!in_atomic() && !irqs_disabled()) || - system_state != SYSTEM_RUNNING || oops_in_progress) + if ((preempt_count_equals(preempt_offset) && !irqs_disabled()) || + system_state != SYSTEM_RUNNING || oops_in_progress) return; if (time_before(jiffies, prev_jiffy + HZ) && prev_jiffy) return; -- cgit v1.2.3 From 6f80bd985fe242c2e6a8b6209ed20b0495d3d63b Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Thu, 16 Jul 2009 15:44:29 +0200 Subject: sched: Remove the CONFIG_PREEMPT_BKL case definition of cond_resched() CONFIG_PREEMPT_BKL doesn't exist anymore. So remove this config-on case definition of cond_resched(). Reported-by: Peter Zijlstra Reported-by: Ingo Molnar Signed-off-by: Frederic Weisbecker Signed-off-by: Peter Zijlstra LKML-Reference: <1247725694-6082-5-git-send-email-fweisbec@gmail.com> Signed-off-by: Ingo Molnar --- include/linux/sched.h | 9 ++------- 1 file changed, 2 insertions(+), 7 deletions(-) (limited to 'include') diff --git a/include/linux/sched.h b/include/linux/sched.h index 9bada20e2b23..e2bdf18e05c4 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -2285,17 +2285,12 @@ static inline int need_resched(void) * cond_resched_softirq() will enable bhs before scheduling. */ extern int _cond_resched(void); -#ifdef CONFIG_PREEMPT_BKL -static inline int cond_resched(void) -{ - return 0; -} -#else + static inline int cond_resched(void) { return _cond_resched(); } -#endif + extern int cond_resched_lock(spinlock_t * lock); extern int cond_resched_softirq(void); static inline int cond_resched_bkl(void) -- cgit v1.2.3 From 613afbf83298efaead05ebcac23d2285609d7160 Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Thu, 16 Jul 2009 15:44:29 +0200 Subject: sched: Pull up the might_sleep() check into cond_resched() might_sleep() is called late-ish in cond_resched(), after the need_resched()/preempt enabled/system running tests are checked. It's better to check the sleeps while atomic earlier and not depend on some environment datas that reduce the chances to detect a problem. Also define cond_resched_*() helpers as macros, so that the FILE/LINE reported in the sleeping while atomic warning displays the real origin and not sched.h Changes in v2: - Call __might_sleep() directly instead of might_sleep() which may call cond_resched() - Turn cond_resched() into a macro so that the file:line couple reported refers to the caller of cond_resched() and not __cond_resched() itself. Changes in v3: - Also propagate this __might_sleep() pull up to cond_resched_lock() and cond_resched_softirq() Signed-off-by: Frederic Weisbecker Signed-off-by: Peter Zijlstra LKML-Reference: <1247725694-6082-6-git-send-email-fweisbec@gmail.com> Signed-off-by: Ingo Molnar --- fs/dcache.c | 1 + include/linux/sched.h | 29 +++++++++++++++++++---------- kernel/sched.c | 12 +++++------- 3 files changed, 25 insertions(+), 17 deletions(-) (limited to 'include') diff --git a/fs/dcache.c b/fs/dcache.c index 9e5cd3c3a6ba..a100fa35a48f 100644 --- a/fs/dcache.c +++ b/fs/dcache.c @@ -32,6 +32,7 @@ #include #include #include +#include #include "internal.h" int sysctl_vfs_cache_pressure __read_mostly = 100; diff --git a/include/linux/sched.h b/include/linux/sched.h index e2bdf18e05c4..c41d424db887 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -2286,17 +2286,26 @@ static inline int need_resched(void) */ extern int _cond_resched(void); -static inline int cond_resched(void) -{ - return _cond_resched(); -} +#define cond_resched() ({ \ + __might_sleep(__FILE__, __LINE__, 0); \ + _cond_resched(); \ +}) -extern int cond_resched_lock(spinlock_t * lock); -extern int cond_resched_softirq(void); -static inline int cond_resched_bkl(void) -{ - return _cond_resched(); -} +extern int __cond_resched_lock(spinlock_t *lock); + +#define cond_resched_lock(lock) ({ \ + __might_sleep(__FILE__, __LINE__, PREEMPT_OFFSET); \ + __cond_resched_lock(lock); \ +}) + +extern int __cond_resched_softirq(void); + +#define cond_resched_softirq() ({ \ + __might_sleep(__FILE__, __LINE__, SOFTIRQ_OFFSET); \ + __cond_resched_softirq(); \ +}) + +#define cond_resched_bkl() cond_resched() /* * Does a critical section need to be broken due to another diff --git a/kernel/sched.c b/kernel/sched.c index 3ff4d004bd95..1f7919add8ae 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -6610,8 +6610,6 @@ static inline int should_resched(void) static void __cond_resched(void) { - __might_sleep(__FILE__, __LINE__, 0); - add_preempt_count(PREEMPT_ACTIVE); schedule(); sub_preempt_count(PREEMPT_ACTIVE); @@ -6628,14 +6626,14 @@ int __sched _cond_resched(void) EXPORT_SYMBOL(_cond_resched); /* - * cond_resched_lock() - if a reschedule is pending, drop the given lock, + * __cond_resched_lock() - if a reschedule is pending, drop the given lock, * call schedule, and on return reacquire the lock. * * This works OK both with and without CONFIG_PREEMPT. We do strange low-level * operations here to prevent schedule() from being called twice (once via * spin_unlock(), once by hand). */ -int cond_resched_lock(spinlock_t *lock) +int __cond_resched_lock(spinlock_t *lock) { int resched = should_resched(); int ret = 0; @@ -6651,9 +6649,9 @@ int cond_resched_lock(spinlock_t *lock) } return ret; } -EXPORT_SYMBOL(cond_resched_lock); +EXPORT_SYMBOL(__cond_resched_lock); -int __sched cond_resched_softirq(void) +int __sched __cond_resched_softirq(void) { BUG_ON(!in_softirq()); @@ -6665,7 +6663,7 @@ int __sched cond_resched_softirq(void) } return 0; } -EXPORT_SYMBOL(cond_resched_softirq); +EXPORT_SYMBOL(__cond_resched_softirq); /** * yield - yield the current processor to other threads. -- cgit v1.2.3 From def01bc53d03881acfc393bd10a5c7575187e008 Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Thu, 16 Jul 2009 15:44:29 +0200 Subject: sched: Convert the only user of cond_resched_bkl to use cond_resched() fs/locks.c:flock_lock_file() is the only user of cond_resched_bkl() This helper doesn't do anything more than cond_resched(). The latter naming is enough to explain that we are rescheduling if needed. The bkl suffix suggests another semantics but it's actually a synonym of cond_resched(). Reported-by: Peter Zijlstra Signed-off-by: Frederic Weisbecker Signed-off-by: Peter Zijlstra LKML-Reference: <1247725694-6082-7-git-send-email-fweisbec@gmail.com> Signed-off-by: Ingo Molnar --- fs/locks.c | 2 +- include/linux/sched.h | 2 -- 2 files changed, 1 insertion(+), 3 deletions(-) (limited to 'include') diff --git a/fs/locks.c b/fs/locks.c index b6440f52178f..2eb81975c99c 100644 --- a/fs/locks.c +++ b/fs/locks.c @@ -768,7 +768,7 @@ static int flock_lock_file(struct file *filp, struct file_lock *request) * give it the opportunity to lock the file. */ if (found) - cond_resched_bkl(); + cond_resched(); find_conflict: for_each_lock(inode, before) { diff --git a/include/linux/sched.h b/include/linux/sched.h index c41d424db887..cbbfca69aa4a 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -2305,8 +2305,6 @@ extern int __cond_resched_softirq(void); __cond_resched_softirq(); \ }) -#define cond_resched_bkl() cond_resched() - /* * Does a critical section need to be broken due to another * task waiting?: (technically does not depend on CONFIG_PREEMPT, -- cgit v1.2.3 From 719a72b7c75bb239ca6184190ab994b71a31c6dc Mon Sep 17 00:00:00 2001 From: Magnus Damm Date: Fri, 17 Jul 2009 14:59:55 +0000 Subject: usb: r8a66597-hcd platform data on_chip support Convert the r8a66597-hcd driver to use the on_chip flag from platform data to enable on chip behaviour instead of relying on CONFIG_SUPERH_ON_CHIP_R8A66597 ugliness. This makes the code cleaner and also allows us to support both external and internal r8a66597 with the same kernel. It also makes the Kconfig part more future proof since we with this patch can add support for new processors with on-chip r8a66597 without modifying the Kconfig. Signed-off-by: Magnus Damm Signed-off-by: Paul Mundt --- arch/sh/boards/mach-se/7724/setup.c | 1 + arch/sh/kernel/cpu/sh4a/setup-sh7366.c | 2 +- arch/sh/kernel/cpu/sh4a/setup-sh7723.c | 2 +- drivers/usb/host/Kconfig | 7 -- drivers/usb/host/r8a66597-hcd.c | 187 +++++++++++++++++++-------------- drivers/usb/host/r8a66597.h | 76 ++++++-------- include/linux/usb/r8a66597.h | 3 + 7 files changed, 147 insertions(+), 131 deletions(-) (limited to 'include') diff --git a/arch/sh/boards/mach-se/7724/setup.c b/arch/sh/boards/mach-se/7724/setup.c index 8fed45a2fb85..4fb7e48e2843 100644 --- a/arch/sh/boards/mach-se/7724/setup.c +++ b/arch/sh/boards/mach-se/7724/setup.c @@ -304,6 +304,7 @@ static struct platform_device sh_eth_device = { }; static struct r8a66597_platdata sh7724_usb0_host_data = { + .on_chip = 1, }; static struct resource sh7724_usb0_host_resources[] = { diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7366.c b/arch/sh/kernel/cpu/sh4a/setup-sh7366.c index c18f7d09281b..f6d208813564 100644 --- a/arch/sh/kernel/cpu/sh4a/setup-sh7366.c +++ b/arch/sh/kernel/cpu/sh4a/setup-sh7366.c @@ -40,7 +40,7 @@ static struct platform_device iic_device = { }; static struct r8a66597_platdata r8a66597_data = { - /* This set zero to all members */ + .on_chip = 1, }; static struct resource usb_host_resources[] = { diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7723.c b/arch/sh/kernel/cpu/sh4a/setup-sh7723.c index e1bb80b2a27b..28516499a2c4 100644 --- a/arch/sh/kernel/cpu/sh4a/setup-sh7723.c +++ b/arch/sh/kernel/cpu/sh4a/setup-sh7723.c @@ -398,7 +398,7 @@ static struct platform_device rtc_device = { }; static struct r8a66597_platdata r8a66597_data = { - /* This set zero to all members */ + .on_chip = 1, }; static struct resource sh7723_usb_host_resources[] = { diff --git a/drivers/usb/host/Kconfig b/drivers/usb/host/Kconfig index 1a920c70b5a1..f21ca7d27a43 100644 --- a/drivers/usb/host/Kconfig +++ b/drivers/usb/host/Kconfig @@ -336,13 +336,6 @@ config USB_R8A66597_HCD To compile this driver as a module, choose M here: the module will be called r8a66597-hcd. -config SUPERH_ON_CHIP_R8A66597 - boolean "Enable SuperH on-chip R8A66597 USB" - depends on USB_R8A66597_HCD && (CPU_SUBTYPE_SH7366 || CPU_SUBTYPE_SH7723 || CPU_SUBTYPE_SH7724) - help - This driver enables support for the on-chip R8A66597 in the - SH7366, SH7723 and SH7724 processors. - config USB_WHCI_HCD tristate "Wireless USB Host Controller Interface (WHCI) driver (EXPERIMENTAL)" depends on EXPERIMENTAL diff --git a/drivers/usb/host/r8a66597-hcd.c b/drivers/usb/host/r8a66597-hcd.c index 09895a97c10b..82dce3e0d4d7 100644 --- a/drivers/usb/host/r8a66597-hcd.c +++ b/drivers/usb/host/r8a66597-hcd.c @@ -91,43 +91,43 @@ static int r8a66597_clock_enable(struct r8a66597 *r8a66597) u16 tmp; int i = 0; -#if defined(CONFIG_SUPERH_ON_CHIP_R8A66597) -#if defined(CONFIG_HAVE_CLK) - clk_enable(r8a66597->clk); + if (r8a66597->pdata->on_chip) { +#ifdef CONFIG_HAVE_CLK + clk_enable(r8a66597->clk); #endif - do { - r8a66597_write(r8a66597, SCKE, SYSCFG0); - tmp = r8a66597_read(r8a66597, SYSCFG0); - if (i++ > 1000) { - printk(KERN_ERR "r8a66597: register access fail.\n"); - return -ENXIO; - } - } while ((tmp & SCKE) != SCKE); - r8a66597_write(r8a66597, 0x04, 0x02); -#else - do { - r8a66597_write(r8a66597, USBE, SYSCFG0); - tmp = r8a66597_read(r8a66597, SYSCFG0); - if (i++ > 1000) { - printk(KERN_ERR "r8a66597: register access fail.\n"); - return -ENXIO; - } - } while ((tmp & USBE) != USBE); - r8a66597_bclr(r8a66597, USBE, SYSCFG0); - r8a66597_mdfy(r8a66597, get_xtal_from_pdata(r8a66597->pdata), XTAL, - SYSCFG0); + do { + r8a66597_write(r8a66597, SCKE, SYSCFG0); + tmp = r8a66597_read(r8a66597, SYSCFG0); + if (i++ > 1000) { + printk(KERN_ERR "r8a66597: reg access fail.\n"); + return -ENXIO; + } + } while ((tmp & SCKE) != SCKE); + r8a66597_write(r8a66597, 0x04, 0x02); + } else { + do { + r8a66597_write(r8a66597, USBE, SYSCFG0); + tmp = r8a66597_read(r8a66597, SYSCFG0); + if (i++ > 1000) { + printk(KERN_ERR "r8a66597: reg access fail.\n"); + return -ENXIO; + } + } while ((tmp & USBE) != USBE); + r8a66597_bclr(r8a66597, USBE, SYSCFG0); + r8a66597_mdfy(r8a66597, get_xtal_from_pdata(r8a66597->pdata), + XTAL, SYSCFG0); - i = 0; - r8a66597_bset(r8a66597, XCKE, SYSCFG0); - do { - msleep(1); - tmp = r8a66597_read(r8a66597, SYSCFG0); - if (i++ > 500) { - printk(KERN_ERR "r8a66597: register access fail.\n"); - return -ENXIO; - } - } while ((tmp & SCKE) != SCKE); -#endif /* #if defined(CONFIG_SUPERH_ON_CHIP_R8A66597) */ + i = 0; + r8a66597_bset(r8a66597, XCKE, SYSCFG0); + do { + msleep(1); + tmp = r8a66597_read(r8a66597, SYSCFG0); + if (i++ > 500) { + printk(KERN_ERR "r8a66597: reg access fail.\n"); + return -ENXIO; + } + } while ((tmp & SCKE) != SCKE); + } return 0; } @@ -136,15 +136,16 @@ static void r8a66597_clock_disable(struct r8a66597 *r8a66597) { r8a66597_bclr(r8a66597, SCKE, SYSCFG0); udelay(1); -#if defined(CONFIG_SUPERH_ON_CHIP_R8A66597) -#if defined(CONFIG_HAVE_CLK) - clk_disable(r8a66597->clk); -#endif -#else - r8a66597_bclr(r8a66597, PLLC, SYSCFG0); - r8a66597_bclr(r8a66597, XCKE, SYSCFG0); - r8a66597_bclr(r8a66597, USBE, SYSCFG0); + + if (r8a66597->pdata->on_chip) { +#ifdef CONFIG_HAVE_CLK + clk_disable(r8a66597->clk); #endif + } else { + r8a66597_bclr(r8a66597, PLLC, SYSCFG0); + r8a66597_bclr(r8a66597, XCKE, SYSCFG0); + r8a66597_bclr(r8a66597, USBE, SYSCFG0); + } } static void r8a66597_enable_port(struct r8a66597 *r8a66597, int port) @@ -205,7 +206,7 @@ static int enable_controller(struct r8a66597 *r8a66597) r8a66597_bset(r8a66597, SIGNE | SACKE, INTENB1); - for (port = 0; port < R8A66597_MAX_ROOT_HUB; port++) + for (port = 0; port < r8a66597->max_root_hub; port++) r8a66597_enable_port(r8a66597, port); return 0; @@ -218,7 +219,7 @@ static void disable_controller(struct r8a66597 *r8a66597) r8a66597_write(r8a66597, 0, INTENB0); r8a66597_write(r8a66597, 0, INTSTS0); - for (port = 0; port < R8A66597_MAX_ROOT_HUB; port++) + for (port = 0; port < r8a66597->max_root_hub; port++) r8a66597_disable_port(r8a66597, port); r8a66597_clock_disable(r8a66597); @@ -249,11 +250,12 @@ static int is_hub_limit(char *devpath) return ((strlen(devpath) >= 4) ? 1 : 0); } -static void get_port_number(char *devpath, u16 *root_port, u16 *hub_port) +static void get_port_number(struct r8a66597 *r8a66597, + char *devpath, u16 *root_port, u16 *hub_port) { if (root_port) { *root_port = (devpath[0] & 0x0F) - 1; - if (*root_port >= R8A66597_MAX_ROOT_HUB) + if (*root_port >= r8a66597->max_root_hub) printk(KERN_ERR "r8a66597: Illegal root port number.\n"); } if (hub_port) @@ -355,7 +357,8 @@ static int make_r8a66597_device(struct r8a66597 *r8a66597, INIT_LIST_HEAD(&dev->device_list); list_add_tail(&dev->device_list, &r8a66597->child_device); - get_port_number(urb->dev->devpath, &dev->root_port, &dev->hub_port); + get_port_number(r8a66597, urb->dev->devpath, + &dev->root_port, &dev->hub_port); if (!is_child_device(urb->dev->devpath)) r8a66597->root_hub[dev->root_port].dev = dev; @@ -420,7 +423,7 @@ static void free_usb_address(struct r8a66597 *r8a66597, list_del(&dev->device_list); kfree(dev); - for (port = 0; port < R8A66597_MAX_ROOT_HUB; port++) { + for (port = 0; port < r8a66597->max_root_hub; port++) { if (r8a66597->root_hub[port].dev == dev) { r8a66597->root_hub[port].dev = NULL; break; @@ -495,10 +498,20 @@ static void r8a66597_pipe_toggle(struct r8a66597 *r8a66597, r8a66597_bset(r8a66597, SQCLR, pipe->pipectr); } +static inline unsigned short mbw_value(struct r8a66597 *r8a66597) +{ + if (r8a66597->pdata->on_chip) + return MBW_32; + else + return MBW_16; +} + /* this function must be called with interrupt disabled */ static inline void cfifo_change(struct r8a66597 *r8a66597, u16 pipenum) { - r8a66597_mdfy(r8a66597, MBW | pipenum, MBW | CURPIPE, CFIFOSEL); + unsigned short mbw = mbw_value(r8a66597); + + r8a66597_mdfy(r8a66597, mbw | pipenum, mbw | CURPIPE, CFIFOSEL); r8a66597_reg_wait(r8a66597, CFIFOSEL, CURPIPE, pipenum); } @@ -506,11 +519,13 @@ static inline void cfifo_change(struct r8a66597 *r8a66597, u16 pipenum) static inline void fifo_change_from_pipe(struct r8a66597 *r8a66597, struct r8a66597_pipe *pipe) { + unsigned short mbw = mbw_value(r8a66597); + cfifo_change(r8a66597, 0); - r8a66597_mdfy(r8a66597, MBW | 0, MBW | CURPIPE, D0FIFOSEL); - r8a66597_mdfy(r8a66597, MBW | 0, MBW | CURPIPE, D1FIFOSEL); + r8a66597_mdfy(r8a66597, mbw | 0, mbw | CURPIPE, D0FIFOSEL); + r8a66597_mdfy(r8a66597, mbw | 0, mbw | CURPIPE, D1FIFOSEL); - r8a66597_mdfy(r8a66597, MBW | pipe->info.pipenum, MBW | CURPIPE, + r8a66597_mdfy(r8a66597, mbw | pipe->info.pipenum, mbw | CURPIPE, pipe->fifosel); r8a66597_reg_wait(r8a66597, pipe->fifosel, CURPIPE, pipe->info.pipenum); } @@ -742,9 +757,13 @@ static void enable_r8a66597_pipe_dma(struct r8a66597 *r8a66597, struct r8a66597_pipe *pipe, struct urb *urb) { -#if !defined(CONFIG_SUPERH_ON_CHIP_R8A66597) int i; struct r8a66597_pipe_info *info = &pipe->info; + unsigned short mbw = mbw_value(r8a66597); + + /* pipe dma is only for external controlles */ + if (r8a66597->pdata->on_chip) + return; if ((pipe->info.pipenum != 0) && (info->type != R8A66597_INT)) { for (i = 0; i < R8A66597_MAX_DMA_CHANNEL; i++) { @@ -763,8 +782,8 @@ static void enable_r8a66597_pipe_dma(struct r8a66597 *r8a66597, set_pipe_reg_addr(pipe, i); cfifo_change(r8a66597, 0); - r8a66597_mdfy(r8a66597, MBW | pipe->info.pipenum, - MBW | CURPIPE, pipe->fifosel); + r8a66597_mdfy(r8a66597, mbw | pipe->info.pipenum, + mbw | CURPIPE, pipe->fifosel); r8a66597_reg_wait(r8a66597, pipe->fifosel, CURPIPE, pipe->info.pipenum); @@ -772,7 +791,6 @@ static void enable_r8a66597_pipe_dma(struct r8a66597 *r8a66597, break; } } -#endif /* #if defined(CONFIG_SUPERH_ON_CHIP_R8A66597) */ } /* this function must be called with interrupt disabled */ @@ -1769,7 +1787,7 @@ static void r8a66597_timer(unsigned long _r8a66597) spin_lock_irqsave(&r8a66597->lock, flags); - for (port = 0; port < R8A66597_MAX_ROOT_HUB; port++) + for (port = 0; port < r8a66597->max_root_hub; port++) r8a66597_root_hub_control(r8a66597, port); spin_unlock_irqrestore(&r8a66597->lock, flags); @@ -1807,7 +1825,7 @@ static void set_address_zero(struct r8a66597 *r8a66597, struct urb *urb) u16 root_port, hub_port; if (usb_address == 0) { - get_port_number(urb->dev->devpath, + get_port_number(r8a66597, urb->dev->devpath, &root_port, &hub_port); set_devadd_reg(r8a66597, 0, get_r8a66597_usb_speed(urb->dev->speed), @@ -2082,7 +2100,7 @@ static int r8a66597_hub_status_data(struct usb_hcd *hcd, char *buf) *buf = 0; /* initialize (no change) */ - for (i = 0; i < R8A66597_MAX_ROOT_HUB; i++) { + for (i = 0; i < r8a66597->max_root_hub; i++) { if (r8a66597->root_hub[i].port & 0xffff0000) *buf |= 1 << (i + 1); } @@ -2097,11 +2115,11 @@ static void r8a66597_hub_descriptor(struct r8a66597 *r8a66597, { desc->bDescriptorType = 0x29; desc->bHubContrCurrent = 0; - desc->bNbrPorts = R8A66597_MAX_ROOT_HUB; + desc->bNbrPorts = r8a66597->max_root_hub; desc->bDescLength = 9; desc->bPwrOn2PwrGood = 0; desc->wHubCharacteristics = cpu_to_le16(0x0011); - desc->bitmap[0] = ((1 << R8A66597_MAX_ROOT_HUB) - 1) << 1; + desc->bitmap[0] = ((1 << r8a66597->max_root_hub) - 1) << 1; desc->bitmap[1] = ~0; } @@ -2129,7 +2147,7 @@ static int r8a66597_hub_control(struct usb_hcd *hcd, u16 typeReq, u16 wValue, } break; case ClearPortFeature: - if (wIndex > R8A66597_MAX_ROOT_HUB) + if (wIndex > r8a66597->max_root_hub) goto error; if (wLength != 0) goto error; @@ -2162,12 +2180,12 @@ static int r8a66597_hub_control(struct usb_hcd *hcd, u16 typeReq, u16 wValue, *buf = 0x00; break; case GetPortStatus: - if (wIndex > R8A66597_MAX_ROOT_HUB) + if (wIndex > r8a66597->max_root_hub) goto error; *(__le32 *)buf = cpu_to_le32(rh->port); break; case SetPortFeature: - if (wIndex > R8A66597_MAX_ROOT_HUB) + if (wIndex > r8a66597->max_root_hub) goto error; if (wLength != 0) goto error; @@ -2216,7 +2234,7 @@ static int r8a66597_bus_suspend(struct usb_hcd *hcd) dbg("%s", __func__); - for (port = 0; port < R8A66597_MAX_ROOT_HUB; port++) { + for (port = 0; port < r8a66597->max_root_hub; port++) { struct r8a66597_root_hub *rh = &r8a66597->root_hub[port]; unsigned long dvstctr_reg = get_dvstctr_reg(port); @@ -2247,7 +2265,7 @@ static int r8a66597_bus_resume(struct usb_hcd *hcd) dbg("%s", __func__); - for (port = 0; port < R8A66597_MAX_ROOT_HUB; port++) { + for (port = 0; port < r8a66597->max_root_hub; port++) { struct r8a66597_root_hub *rh = &r8a66597->root_hub[port]; unsigned long dvstctr_reg = get_dvstctr_reg(port); @@ -2314,7 +2332,7 @@ static int r8a66597_suspend(struct device *dev) disable_controller(r8a66597); - for (port = 0; port < R8A66597_MAX_ROOT_HUB; port++) { + for (port = 0; port < r8a66597->max_root_hub; port++) { struct r8a66597_root_hub *rh = &r8a66597->root_hub[port]; rh->port = 0x00000000; @@ -2354,8 +2372,9 @@ static int __init_or_module r8a66597_remove(struct platform_device *pdev) del_timer_sync(&r8a66597->rh_timer); usb_remove_hcd(hcd); iounmap((void *)r8a66597->reg); -#if defined(CONFIG_SUPERH_ON_CHIP_R8A66597) && defined(CONFIG_HAVE_CLK) - clk_put(r8a66597->clk); +#ifdef CONFIG_HAVE_CLK + if (r8a66597->pdata->on_chip) + clk_put(r8a66597->clk); #endif usb_put_hcd(hcd); return 0; @@ -2363,7 +2382,7 @@ static int __init_or_module r8a66597_remove(struct platform_device *pdev) static int __devinit r8a66597_probe(struct platform_device *pdev) { -#if defined(CONFIG_SUPERH_ON_CHIP_R8A66597) && defined(CONFIG_HAVE_CLK) +#ifdef CONFIG_HAVE_CLK char clk_name[8]; #endif struct resource *res = NULL, *ires; @@ -2425,15 +2444,20 @@ static int __devinit r8a66597_probe(struct platform_device *pdev) r8a66597->pdata = pdev->dev.platform_data; r8a66597->irq_sense_low = irq_trigger == IRQF_TRIGGER_LOW; -#if defined(CONFIG_SUPERH_ON_CHIP_R8A66597) && defined(CONFIG_HAVE_CLK) - snprintf(clk_name, sizeof(clk_name), "usb%d", pdev->id); - r8a66597->clk = clk_get(&pdev->dev, clk_name); - if (IS_ERR(r8a66597->clk)) { - dev_err(&pdev->dev, "cannot get clock \"%s\"\n", clk_name); - ret = PTR_ERR(r8a66597->clk); - goto clean_up2; - } + if (r8a66597->pdata->on_chip) { +#ifdef CONFIG_HAVE_CLK + snprintf(clk_name, sizeof(clk_name), "usb%d", pdev->id); + r8a66597->clk = clk_get(&pdev->dev, clk_name); + if (IS_ERR(r8a66597->clk)) { + dev_err(&pdev->dev, "cannot get clock \"%s\"\n", + clk_name); + ret = PTR_ERR(r8a66597->clk); + goto clean_up2; + } #endif + r8a66597->max_root_hub = 1; + } else + r8a66597->max_root_hub = 2; spin_lock_init(&r8a66597->lock); init_timer(&r8a66597->rh_timer); @@ -2463,8 +2487,9 @@ static int __devinit r8a66597_probe(struct platform_device *pdev) return 0; clean_up3: -#if defined(CONFIG_SUPERH_ON_CHIP_R8A66597) && defined(CONFIG_HAVE_CLK) - clk_put(r8a66597->clk); +#ifdef CONFIG_HAVE_CLK + if (r8a66597->pdata->on_chip) + clk_put(r8a66597->clk); clean_up2: #endif usb_put_hcd(hcd); diff --git a/drivers/usb/host/r8a66597.h b/drivers/usb/host/r8a66597.h index d72680b433f9..eecbd917bc81 100644 --- a/drivers/usb/host/r8a66597.h +++ b/drivers/usb/host/r8a66597.h @@ -26,7 +26,7 @@ #ifndef __R8A66597_H__ #define __R8A66597_H__ -#if defined(CONFIG_SUPERH_ON_CHIP_R8A66597) && defined(CONFIG_HAVE_CLK) +#ifdef CONFIG_HAVE_CLK #include #endif @@ -193,13 +193,9 @@ #define REW 0x4000 /* b14: Buffer rewind */ #define DCLRM 0x2000 /* b13: DMA buffer clear mode */ #define DREQE 0x1000 /* b12: DREQ output enable */ -#if defined(CONFIG_SUPERH_ON_CHIP_R8A66597) -#define MBW 0x0800 -#else -#define MBW 0x0400 /* b10: Maximum bit width for FIFO access */ -#endif #define MBW_8 0x0000 /* 8bit */ #define MBW_16 0x0400 /* 16bit */ +#define MBW_32 0x0800 /* 32bit */ #define BIGEND 0x0100 /* b8: Big endian mode */ #define BYTE_LITTLE 0x0000 /* little dendian */ #define BYTE_BIG 0x0100 /* big endifan */ @@ -405,11 +401,7 @@ #define R8A66597_MAX_NUM_PIPE 10 #define R8A66597_BUF_BSIZE 8 #define R8A66597_MAX_DEVICE 10 -#if defined(CONFIG_SUPERH_ON_CHIP_R8A66597) -#define R8A66597_MAX_ROOT_HUB 1 -#else #define R8A66597_MAX_ROOT_HUB 2 -#endif #define R8A66597_MAX_SAMPLING 5 #define R8A66597_RH_POLL_TIME 10 #define R8A66597_MAX_DMA_CHANNEL 2 @@ -487,7 +479,7 @@ struct r8a66597_root_hub { struct r8a66597 { spinlock_t lock; unsigned long reg; -#if defined(CONFIG_SUPERH_ON_CHIP_R8A66597) && defined(CONFIG_HAVE_CLK) +#ifdef CONFIG_HAVE_CLK struct clk *clk; #endif struct r8a66597_platdata *pdata; @@ -504,6 +496,7 @@ struct r8a66597 { unsigned short interval_map; unsigned char pipe_cnt[R8A66597_MAX_NUM_PIPE]; unsigned char dma_map; + unsigned int max_root_hub; struct list_head child_device; unsigned long child_connect_map[4]; @@ -550,21 +543,22 @@ static inline void r8a66597_read_fifo(struct r8a66597 *r8a66597, unsigned long offset, u16 *buf, int len) { -#if defined(CONFIG_SUPERH_ON_CHIP_R8A66597) unsigned long fifoaddr = r8a66597->reg + offset; unsigned long count; - count = len / 4; - insl(fifoaddr, buf, count); + if (r8a66597->pdata->on_chip) { + count = len / 4; + insl(fifoaddr, buf, count); - if (len & 0x00000003) { - unsigned long tmp = inl(fifoaddr); - memcpy((unsigned char *)buf + count * 4, &tmp, len & 0x03); + if (len & 0x00000003) { + unsigned long tmp = inl(fifoaddr); + memcpy((unsigned char *)buf + count * 4, &tmp, + len & 0x03); + } + } else { + len = (len + 1) / 2; + insw(fifoaddr, buf, len); } -#else - len = (len + 1) / 2; - insw(r8a66597->reg + offset, buf, len); -#endif } static inline void r8a66597_write(struct r8a66597 *r8a66597, u16 val, @@ -578,33 +572,33 @@ static inline void r8a66597_write_fifo(struct r8a66597 *r8a66597, int len) { unsigned long fifoaddr = r8a66597->reg + offset; -#if defined(CONFIG_SUPERH_ON_CHIP_R8A66597) unsigned long count; unsigned char *pb; int i; - count = len / 4; - outsl(fifoaddr, buf, count); + if (r8a66597->pdata->on_chip) { + count = len / 4; + outsl(fifoaddr, buf, count); + + if (len & 0x00000003) { + pb = (unsigned char *)buf + count * 4; + for (i = 0; i < (len & 0x00000003); i++) { + if (r8a66597_read(r8a66597, CFIFOSEL) & BIGEND) + outb(pb[i], fifoaddr + i); + else + outb(pb[i], fifoaddr + 3 - i); + } + } + } else { + int odd = len & 0x0001; - if (len & 0x00000003) { - pb = (unsigned char *)buf + count * 4; - for (i = 0; i < (len & 0x00000003); i++) { - if (r8a66597_read(r8a66597, CFIFOSEL) & BIGEND) - outb(pb[i], fifoaddr + i); - else - outb(pb[i], fifoaddr + 3 - i); + len = len / 2; + outsw(fifoaddr, buf, len); + if (unlikely(odd)) { + buf = &buf[len]; + outb((unsigned char)*buf, fifoaddr); } } -#else - int odd = len & 0x0001; - - len = len / 2; - outsw(fifoaddr, buf, len); - if (unlikely(odd)) { - buf = &buf[len]; - outb((unsigned char)*buf, fifoaddr); - } -#endif } static inline void r8a66597_mdfy(struct r8a66597 *r8a66597, diff --git a/include/linux/usb/r8a66597.h b/include/linux/usb/r8a66597.h index e9f0384fa20c..460ee3f6a2c6 100644 --- a/include/linux/usb/r8a66597.h +++ b/include/linux/usb/r8a66597.h @@ -31,6 +31,9 @@ struct r8a66597_platdata { /* This ops can controll port power instead of DVSTCTR register. */ void (*port_power)(int port, int power); + /* set one = on chip controller, set zero = external controller */ + unsigned on_chip:1; + /* (external controller only) set R8A66597_PLATDATA_XTAL_nnMHZ */ unsigned xtal:2; -- cgit v1.2.3 From 1c388ad054fb1ead3dc354b1719570b99e464135 Mon Sep 17 00:00:00 2001 From: Paul Menage Date: Fri, 17 Jul 2009 16:16:18 -0700 Subject: include/linux/cred.h: work around gcc-4.2.4 warning in get_cred() With gcc 4.2.4 (building UML) I get the warning include/linux/cred.h: In function 'get_cred': include/linux/cred.h:189: warning: passing argument 1 of 'get_new_cred' discards qualifiers from pointer target type Inserting an additional local variable appears to keep the compiler happy, although it's not clear to me why this should be needed. Signed-off-by: Paul Menage Signed-off-by: Andrew Morton Acked-by: David Howells Signed-off-by: James Morris --- include/linux/cred.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/cred.h b/include/linux/cred.h index 4fa999696310..b3c76e815d66 100644 --- a/include/linux/cred.h +++ b/include/linux/cred.h @@ -186,7 +186,8 @@ static inline struct cred *get_new_cred(struct cred *cred) */ static inline const struct cred *get_cred(const struct cred *cred) { - return get_new_cred((struct cred *) cred); + struct cred *nonconst_cred = (struct cred *) cred; + return get_new_cred(nonconst_cred); } /** -- cgit v1.2.3 From 4d4036e0e7299c6cbb2d2421b4b30b7a409ce61a Mon Sep 17 00:00:00 2001 From: Jason Yeh Date: Wed, 8 Jul 2009 13:49:38 +0200 Subject: oprofile: Implement performance counter multiplexing The number of hardware counters is limited. The multiplexing feature enables OProfile to gather more events than counters are provided by the hardware. This is realized by switching between events at an user specified time interval. A new file (/dev/oprofile/time_slice) is added for the user to specify the timer interval in ms. If the number of events to profile is higher than the number of hardware counters available, the patch will schedule a work queue that switches the event counter and re-writes the different sets of values into it. The switching mechanism needs to be implemented for each architecture to support multiplexing. This patch only implements AMD CPU support, but multiplexing can be easily extended for other models and architectures. There are follow-on patches that rework parts of this patch. Signed-off-by: Jason Yeh Signed-off-by: Robert Richter --- arch/Kconfig | 12 +++ arch/x86/oprofile/nmi_int.c | 162 ++++++++++++++++++++++++++++++++++++-- arch/x86/oprofile/op_counter.h | 2 +- arch/x86/oprofile/op_model_amd.c | 110 ++++++++++++++++++++++---- arch/x86/oprofile/op_model_p4.c | 4 + arch/x86/oprofile/op_model_ppro.c | 2 + arch/x86/oprofile/op_x86_model.h | 7 ++ drivers/oprofile/oprof.c | 78 ++++++++++++++++++ drivers/oprofile/oprof.h | 2 + drivers/oprofile/oprofile_files.c | 43 ++++++++++ drivers/oprofile/oprofile_stats.c | 10 +++ include/linux/oprofile.h | 3 + 12 files changed, 415 insertions(+), 20 deletions(-) (limited to 'include') diff --git a/arch/Kconfig b/arch/Kconfig index 99193b160232..beea3ccebb5e 100644 --- a/arch/Kconfig +++ b/arch/Kconfig @@ -30,6 +30,18 @@ config OPROFILE_IBS If unsure, say N. +config OPROFILE_EVENT_MULTIPLEX + bool "OProfile multiplexing support (EXPERIMENTAL)" + default n + depends on OPROFILE && X86 + help + The number of hardware counters is limited. The multiplexing + feature enables OProfile to gather more events than counters + are provided by the hardware. This is realized by switching + between events at an user specified time interval. + + If unsure, say N. + config HAVE_OPROFILE bool diff --git a/arch/x86/oprofile/nmi_int.c b/arch/x86/oprofile/nmi_int.c index fca8dc94531e..e54f6a0b35ac 100644 --- a/arch/x86/oprofile/nmi_int.c +++ b/arch/x86/oprofile/nmi_int.c @@ -1,11 +1,14 @@ /** * @file nmi_int.c * - * @remark Copyright 2002-2008 OProfile authors + * @remark Copyright 2002-2009 OProfile authors * @remark Read the file COPYING * * @author John Levon * @author Robert Richter + * @author Barry Kasindorf + * @author Jason Yeh + * @author Suravee Suthikulpanit */ #include @@ -24,6 +27,12 @@ #include "op_counter.h" #include "op_x86_model.h" + +#ifdef CONFIG_OPROFILE_EVENT_MULTIPLEX +DEFINE_PER_CPU(int, switch_index); +#endif + + static struct op_x86_model_spec const *model; static DEFINE_PER_CPU(struct op_msrs, cpu_msrs); static DEFINE_PER_CPU(unsigned long, saved_lvtpc); @@ -31,6 +40,13 @@ static DEFINE_PER_CPU(unsigned long, saved_lvtpc); /* 0 == registered but off, 1 == registered and on */ static int nmi_enabled = 0; + +#ifdef CONFIG_OPROFILE_EVENT_MULTIPLEX +extern atomic_t multiplex_counter; +#endif + +struct op_counter_config counter_config[OP_MAX_COUNTER]; + /* common functions */ u64 op_x86_get_ctrl(struct op_x86_model_spec const *model, @@ -95,6 +111,11 @@ static void free_msrs(void) per_cpu(cpu_msrs, i).counters = NULL; kfree(per_cpu(cpu_msrs, i).controls); per_cpu(cpu_msrs, i).controls = NULL; + +#ifdef CONFIG_OPROFILE_EVENT_MULTIPLEX + kfree(per_cpu(cpu_msrs, i).multiplex); + per_cpu(cpu_msrs, i).multiplex = NULL; +#endif } } @@ -103,6 +124,9 @@ static int allocate_msrs(void) int success = 1; size_t controls_size = sizeof(struct op_msr) * model->num_controls; size_t counters_size = sizeof(struct op_msr) * model->num_counters; +#ifdef CONFIG_OPROFILE_EVENT_MULTIPLEX + size_t multiplex_size = sizeof(struct op_msr) * model->num_virt_counters; +#endif int i; for_each_possible_cpu(i) { @@ -118,6 +142,14 @@ static int allocate_msrs(void) success = 0; break; } +#ifdef CONFIG_OPROFILE_EVENT_MULTIPLEX + per_cpu(cpu_msrs, i).multiplex = + kmalloc(multiplex_size, GFP_KERNEL); + if (!per_cpu(cpu_msrs, i).multiplex) { + success = 0; + break; + } +#endif } if (!success) @@ -126,6 +158,25 @@ static int allocate_msrs(void) return success; } +#ifdef CONFIG_OPROFILE_EVENT_MULTIPLEX + +static void nmi_setup_cpu_mux(struct op_msrs const * const msrs) +{ + int i; + struct op_msr *multiplex = msrs->multiplex; + + for (i = 0; i < model->num_virt_counters; ++i) { + if (counter_config[i].enabled) { + multiplex[i].saved = -(u64)counter_config[i].count; + } else { + multiplex[i].addr = 0; + multiplex[i].saved = 0; + } + } +} + +#endif + static void nmi_cpu_setup(void *dummy) { int cpu = smp_processor_id(); @@ -133,6 +184,9 @@ static void nmi_cpu_setup(void *dummy) nmi_cpu_save_registers(msrs); spin_lock(&oprofilefs_lock); model->setup_ctrs(model, msrs); +#ifdef CONFIG_OPROFILE_EVENT_MULTIPLEX + nmi_setup_cpu_mux(msrs); +#endif spin_unlock(&oprofilefs_lock); per_cpu(saved_lvtpc, cpu) = apic_read(APIC_LVTPC); apic_write(APIC_LVTPC, APIC_DM_NMI); @@ -173,14 +227,52 @@ static int nmi_setup(void) memcpy(per_cpu(cpu_msrs, cpu).controls, per_cpu(cpu_msrs, 0).controls, sizeof(struct op_msr) * model->num_controls); +#ifdef CONFIG_OPROFILE_EVENT_MULTIPLEX + memcpy(per_cpu(cpu_msrs, cpu).multiplex, + per_cpu(cpu_msrs, 0).multiplex, + sizeof(struct op_msr) * model->num_virt_counters); +#endif } - } on_each_cpu(nmi_cpu_setup, NULL, 1); nmi_enabled = 1; return 0; } +#ifdef CONFIG_OPROFILE_EVENT_MULTIPLEX + +static void nmi_cpu_save_mpx_registers(struct op_msrs *msrs) +{ + unsigned int si = __get_cpu_var(switch_index); + struct op_msr *multiplex = msrs->multiplex; + unsigned int i; + + for (i = 0; i < model->num_counters; ++i) { + int offset = i + si; + if (multiplex[offset].addr) { + rdmsrl(multiplex[offset].addr, + multiplex[offset].saved); + } + } +} + +static void nmi_cpu_restore_mpx_registers(struct op_msrs *msrs) +{ + unsigned int si = __get_cpu_var(switch_index); + struct op_msr *multiplex = msrs->multiplex; + unsigned int i; + + for (i = 0; i < model->num_counters; ++i) { + int offset = i + si; + if (multiplex[offset].addr) { + wrmsrl(multiplex[offset].addr, + multiplex[offset].saved); + } + } +} + +#endif + static void nmi_cpu_restore_registers(struct op_msrs *msrs) { struct op_msr *counters = msrs->counters; @@ -214,6 +306,9 @@ static void nmi_cpu_shutdown(void *dummy) apic_write(APIC_LVTPC, per_cpu(saved_lvtpc, cpu)); apic_write(APIC_LVTERR, v); nmi_cpu_restore_registers(msrs); +#ifdef CONFIG_OPROFILE_EVENT_MULTIPLEX + __get_cpu_var(switch_index) = 0; +#endif } static void nmi_shutdown(void) @@ -252,16 +347,15 @@ static void nmi_stop(void) on_each_cpu(nmi_cpu_stop, NULL, 1); } -struct op_counter_config counter_config[OP_MAX_COUNTER]; - static int nmi_create_files(struct super_block *sb, struct dentry *root) { unsigned int i; - for (i = 0; i < model->num_counters; ++i) { + for (i = 0; i < model->num_virt_counters; ++i) { struct dentry *dir; char buf[4]; +#ifndef CONFIG_OPROFILE_EVENT_MULTIPLEX /* quick little hack to _not_ expose a counter if it is not * available for use. This should protect userspace app. * NOTE: assumes 1:1 mapping here (that counters are organized @@ -269,6 +363,7 @@ static int nmi_create_files(struct super_block *sb, struct dentry *root) */ if (unlikely(!avail_to_resrv_perfctr_nmi_bit(i))) continue; +#endif /* CONFIG_OPROFILE_EVENT_MULTIPLEX */ snprintf(buf, sizeof(buf), "%d", i); dir = oprofilefs_mkdir(sb, root, buf); @@ -283,6 +378,57 @@ static int nmi_create_files(struct super_block *sb, struct dentry *root) return 0; } +#ifdef CONFIG_OPROFILE_EVENT_MULTIPLEX + +static void nmi_cpu_switch(void *dummy) +{ + int cpu = smp_processor_id(); + int si = per_cpu(switch_index, cpu); + struct op_msrs *msrs = &per_cpu(cpu_msrs, cpu); + + nmi_cpu_stop(NULL); + nmi_cpu_save_mpx_registers(msrs); + + /* move to next set */ + si += model->num_counters; + if ((si > model->num_virt_counters) || (counter_config[si].count == 0)) + per_cpu(switch_index, cpu) = 0; + else + per_cpu(switch_index, cpu) = si; + + model->switch_ctrl(model, msrs); + nmi_cpu_restore_mpx_registers(msrs); + + nmi_cpu_start(NULL); +} + + +/* + * Quick check to see if multiplexing is necessary. + * The check should be sufficient since counters are used + * in ordre. + */ +static int nmi_multiplex_on(void) +{ + return counter_config[model->num_counters].count ? 0 : -EINVAL; +} + +static int nmi_switch_event(void) +{ + if (!model->switch_ctrl) + return -ENOSYS; /* not implemented */ + if (nmi_multiplex_on() < 0) + return -EINVAL; /* not necessary */ + + on_each_cpu(nmi_cpu_switch, NULL, 1); + + atomic_inc(&multiplex_counter); + + return 0; +} + +#endif + #ifdef CONFIG_SMP static int oprofile_cpu_notifier(struct notifier_block *b, unsigned long action, void *data) @@ -516,12 +662,18 @@ int __init op_nmi_init(struct oprofile_operations *ops) register_cpu_notifier(&oprofile_cpu_nb); #endif /* default values, can be overwritten by model */ +#ifdef CONFIG_OPROFILE_EVENT_MULTIPLEX + __raw_get_cpu_var(switch_index) = 0; +#endif ops->create_files = nmi_create_files; ops->setup = nmi_setup; ops->shutdown = nmi_shutdown; ops->start = nmi_start; ops->stop = nmi_stop; ops->cpu_type = cpu_type; +#ifdef CONFIG_OPROFILE_EVENT_MULTIPLEX + ops->switch_events = nmi_switch_event; +#endif if (model->init) ret = model->init(ops); diff --git a/arch/x86/oprofile/op_counter.h b/arch/x86/oprofile/op_counter.h index 91b6a116165e..e28398df0df2 100644 --- a/arch/x86/oprofile/op_counter.h +++ b/arch/x86/oprofile/op_counter.h @@ -10,7 +10,7 @@ #ifndef OP_COUNTER_H #define OP_COUNTER_H -#define OP_MAX_COUNTER 8 +#define OP_MAX_COUNTER 32 /* Per-perfctr configuration as set via * oprofilefs. diff --git a/arch/x86/oprofile/op_model_amd.c b/arch/x86/oprofile/op_model_amd.c index f676f8825a3f..fdbed3a0c877 100644 --- a/arch/x86/oprofile/op_model_amd.c +++ b/arch/x86/oprofile/op_model_amd.c @@ -9,12 +9,15 @@ * @author Philippe Elie * @author Graydon Hoare * @author Robert Richter - * @author Barry Kasindorf + * @author Barry Kasindorf + * @author Jason Yeh + * @author Suravee Suthikulpanit */ #include #include #include +#include #include #include @@ -25,12 +28,23 @@ #define NUM_COUNTERS 4 #define NUM_CONTROLS 4 +#ifdef CONFIG_OPROFILE_EVENT_MULTIPLEX +#define NUM_VIRT_COUNTERS 32 +#define NUM_VIRT_CONTROLS 32 +#else +#define NUM_VIRT_COUNTERS NUM_COUNTERS +#define NUM_VIRT_CONTROLS NUM_CONTROLS +#endif + #define OP_EVENT_MASK 0x0FFF #define OP_CTR_OVERFLOW (1ULL<<31) #define MSR_AMD_EVENTSEL_RESERVED ((0xFFFFFCF0ULL<<32)|(1ULL<<21)) -static unsigned long reset_value[NUM_COUNTERS]; +static unsigned long reset_value[NUM_VIRT_COUNTERS]; +#ifdef CONFIG_OPROFILE_EVENT_MULTIPLEX +DECLARE_PER_CPU(int, switch_index); +#endif #ifdef CONFIG_OPROFILE_IBS @@ -82,6 +96,16 @@ static void op_amd_fill_in_addresses(struct op_msrs * const msrs) else msrs->controls[i].addr = 0; } + +#ifdef CONFIG_OPROFILE_EVENT_MULTIPLEX + for (i = 0; i < NUM_VIRT_COUNTERS; i++) { + int hw_counter = i % NUM_CONTROLS; + if (reserve_perfctr_nmi(MSR_K7_PERFCTR0 + i)) + msrs->multiplex[i].addr = MSR_K7_PERFCTR0 + hw_counter; + else + msrs->multiplex[i].addr = 0; + } +#endif } static void op_amd_setup_ctrs(struct op_x86_model_spec const *model, @@ -90,6 +114,15 @@ static void op_amd_setup_ctrs(struct op_x86_model_spec const *model, u64 val; int i; + /* setup reset_value */ + for (i = 0; i < NUM_VIRT_COUNTERS; ++i) { + if (counter_config[i].enabled) { + reset_value[i] = counter_config[i].count; + } else { + reset_value[i] = 0; + } + } + /* clear all counters */ for (i = 0; i < NUM_CONTROLS; ++i) { if (unlikely(!msrs->controls[i].addr)) @@ -108,20 +141,49 @@ static void op_amd_setup_ctrs(struct op_x86_model_spec const *model, /* enable active counters */ for (i = 0; i < NUM_COUNTERS; ++i) { - if (counter_config[i].enabled && msrs->counters[i].addr) { - reset_value[i] = counter_config[i].count; - wrmsrl(msrs->counters[i].addr, - -(u64)counter_config[i].count); +#ifdef CONFIG_OPROFILE_EVENT_MULTIPLEX + int offset = i + __get_cpu_var(switch_index); +#else + int offset = i; +#endif + if (counter_config[offset].enabled && msrs->counters[i].addr) { + /* setup counter registers */ + wrmsrl(msrs->counters[i].addr, -(u64)reset_value[offset]); + + /* setup control registers */ rdmsrl(msrs->controls[i].addr, val); val &= model->reserved; - val |= op_x86_get_ctrl(model, &counter_config[i]); + val |= op_x86_get_ctrl(model, &counter_config[offset]); + wrmsrl(msrs->controls[i].addr, val); + } + } +} + + +#ifdef CONFIG_OPROFILE_EVENT_MULTIPLEX + +static void op_amd_switch_ctrl(struct op_x86_model_spec const *model, + struct op_msrs const * const msrs) +{ + u64 val; + int i; + + /* enable active counters */ + for (i = 0; i < NUM_COUNTERS; ++i) { + int offset = i + __get_cpu_var(switch_index); + if (counter_config[offset].enabled) { + /* setup control registers */ + rdmsrl(msrs->controls[i].addr, val); + val &= model->reserved; + val |= op_x86_get_ctrl(model, &counter_config[offset]); wrmsrl(msrs->controls[i].addr, val); - } else { - reset_value[i] = 0; } } } +#endif + + #ifdef CONFIG_OPROFILE_IBS static inline int @@ -230,14 +292,19 @@ static int op_amd_check_ctrs(struct pt_regs * const regs, int i; for (i = 0; i < NUM_COUNTERS; ++i) { - if (!reset_value[i]) +#ifdef CONFIG_OPROFILE_EVENT_MULTIPLEX + int offset = i + __get_cpu_var(switch_index); +#else + int offset = i; +#endif + if (!reset_value[offset]) continue; rdmsrl(msrs->counters[i].addr, val); /* bit is clear if overflowed: */ if (val & OP_CTR_OVERFLOW) continue; - oprofile_add_sample(regs, i); - wrmsrl(msrs->counters[i].addr, -(u64)reset_value[i]); + oprofile_add_sample(regs, offset); + wrmsrl(msrs->counters[i].addr, -(u64)reset_value[offset]); } op_amd_handle_ibs(regs, msrs); @@ -250,8 +317,14 @@ static void op_amd_start(struct op_msrs const * const msrs) { u64 val; int i; + for (i = 0; i < NUM_COUNTERS; ++i) { - if (reset_value[i]) { +#ifdef CONFIG_OPROFILE_EVENT_MULTIPLEX + int offset = i + __get_cpu_var(switch_index); +#else + int offset = i; +#endif + if (reset_value[offset]) { rdmsrl(msrs->controls[i].addr, val); val |= ARCH_PERFMON_EVENTSEL0_ENABLE; wrmsrl(msrs->controls[i].addr, val); @@ -271,7 +344,11 @@ static void op_amd_stop(struct op_msrs const * const msrs) * pm callback */ for (i = 0; i < NUM_COUNTERS; ++i) { +#ifdef CONFIG_OPROFILE_EVENT_MULTIPLEX + if (!reset_value[i + per_cpu(switch_index, smp_processor_id())]) +#else if (!reset_value[i]) +#endif continue; rdmsrl(msrs->controls[i].addr, val); val &= ~ARCH_PERFMON_EVENTSEL0_ENABLE; @@ -289,7 +366,7 @@ static void op_amd_shutdown(struct op_msrs const * const msrs) if (msrs->counters[i].addr) release_perfctr_nmi(MSR_K7_PERFCTR0 + i); } - for (i = 0; i < NUM_CONTROLS; ++i) { + for (i = 0; i < NUM_COUNTERS; ++i) { if (msrs->controls[i].addr) release_evntsel_nmi(MSR_K7_EVNTSEL0 + i); } @@ -463,6 +540,8 @@ static void op_amd_exit(void) {} struct op_x86_model_spec const op_amd_spec = { .num_counters = NUM_COUNTERS, .num_controls = NUM_CONTROLS, + .num_virt_counters = NUM_VIRT_COUNTERS, + .num_virt_controls = NUM_VIRT_CONTROLS, .reserved = MSR_AMD_EVENTSEL_RESERVED, .event_mask = OP_EVENT_MASK, .init = op_amd_init, @@ -473,4 +552,7 @@ struct op_x86_model_spec const op_amd_spec = { .start = &op_amd_start, .stop = &op_amd_stop, .shutdown = &op_amd_shutdown, +#ifdef CONFIG_OPROFILE_EVENT_MULTIPLEX + .switch_ctrl = &op_amd_switch_ctrl, +#endif }; diff --git a/arch/x86/oprofile/op_model_p4.c b/arch/x86/oprofile/op_model_p4.c index 5921b7fc724b..65b9237cde8b 100644 --- a/arch/x86/oprofile/op_model_p4.c +++ b/arch/x86/oprofile/op_model_p4.c @@ -698,6 +698,8 @@ static void p4_shutdown(struct op_msrs const * const msrs) struct op_x86_model_spec const op_p4_ht2_spec = { .num_counters = NUM_COUNTERS_HT2, .num_controls = NUM_CONTROLS_HT2, + .num_virt_counters = NUM_COUNTERS_HT2, + .num_virt_controls = NUM_CONTROLS_HT2, .fill_in_addresses = &p4_fill_in_addresses, .setup_ctrs = &p4_setup_ctrs, .check_ctrs = &p4_check_ctrs, @@ -710,6 +712,8 @@ struct op_x86_model_spec const op_p4_ht2_spec = { struct op_x86_model_spec const op_p4_spec = { .num_counters = NUM_COUNTERS_NON_HT, .num_controls = NUM_CONTROLS_NON_HT, + .num_virt_counters = NUM_COUNTERS_NON_HT, + .num_virt_controls = NUM_CONTROLS_NON_HT, .fill_in_addresses = &p4_fill_in_addresses, .setup_ctrs = &p4_setup_ctrs, .check_ctrs = &p4_check_ctrs, diff --git a/arch/x86/oprofile/op_model_ppro.c b/arch/x86/oprofile/op_model_ppro.c index 570d717c3308..098cbca5c0b0 100644 --- a/arch/x86/oprofile/op_model_ppro.c +++ b/arch/x86/oprofile/op_model_ppro.c @@ -206,6 +206,8 @@ static void ppro_shutdown(struct op_msrs const * const msrs) struct op_x86_model_spec const op_ppro_spec = { .num_counters = 2, .num_controls = 2, + .num_virt_counters = 2, + .num_virt_controls = 2, .reserved = MSR_PPRO_EVENTSEL_RESERVED, .fill_in_addresses = &ppro_fill_in_addresses, .setup_ctrs = &ppro_setup_ctrs, diff --git a/arch/x86/oprofile/op_x86_model.h b/arch/x86/oprofile/op_x86_model.h index 505489873b9d..0d07d23cb062 100644 --- a/arch/x86/oprofile/op_x86_model.h +++ b/arch/x86/oprofile/op_x86_model.h @@ -23,6 +23,7 @@ struct op_msr { struct op_msrs { struct op_msr *counters; struct op_msr *controls; + struct op_msr *multiplex; }; struct pt_regs; @@ -35,6 +36,8 @@ struct oprofile_operations; struct op_x86_model_spec { unsigned int num_counters; unsigned int num_controls; + unsigned int num_virt_counters; + unsigned int num_virt_controls; u64 reserved; u16 event_mask; int (*init)(struct oprofile_operations *ops); @@ -47,6 +50,10 @@ struct op_x86_model_spec { void (*start)(struct op_msrs const * const msrs); void (*stop)(struct op_msrs const * const msrs); void (*shutdown)(struct op_msrs const * const msrs); +#ifdef CONFIG_OPROFILE_EVENT_MULTIPLEX + void (*switch_ctrl)(struct op_x86_model_spec const *model, + struct op_msrs const * const msrs); +#endif }; struct op_counter_config; diff --git a/drivers/oprofile/oprof.c b/drivers/oprofile/oprof.c index 3cffce90f82a..7bc64af7cf99 100644 --- a/drivers/oprofile/oprof.c +++ b/drivers/oprofile/oprof.c @@ -12,6 +12,8 @@ #include #include #include +#include +#include #include #include "oprof.h" @@ -27,6 +29,15 @@ unsigned long oprofile_backtrace_depth; static unsigned long is_setup; static DEFINE_MUTEX(start_mutex); +#ifdef CONFIG_OPROFILE_EVENT_MULTIPLEX + +static void switch_worker(struct work_struct *work); +static DECLARE_DELAYED_WORK(switch_work, switch_worker); +unsigned long timeout_jiffies; +#define MULTIPLEXING_TIMER_DEFAULT 1 + +#endif + /* timer 0 - use performance monitoring hardware if available 1 - use the timer int mechanism regardless @@ -87,6 +98,20 @@ out: return err; } +#ifdef CONFIG_OPROFILE_EVENT_MULTIPLEX + +static void start_switch_worker(void) +{ + schedule_delayed_work(&switch_work, timeout_jiffies); +} + +static void switch_worker(struct work_struct *work) +{ + if (!oprofile_ops.switch_events()) + start_switch_worker(); +} + +#endif /* Actually start profiling (echo 1>/dev/oprofile/enable) */ int oprofile_start(void) @@ -108,6 +133,11 @@ int oprofile_start(void) if ((err = oprofile_ops.start())) goto out; +#ifdef CONFIG_OPROFILE_EVENT_MULTIPLEX + if (oprofile_ops.switch_events) + start_switch_worker(); +#endif + oprofile_started = 1; out: mutex_unlock(&start_mutex); @@ -123,6 +153,11 @@ void oprofile_stop(void) goto out; oprofile_ops.stop(); oprofile_started = 0; + +#ifdef CONFIG_OPROFILE_EVENT_MULTIPLEX + cancel_delayed_work_sync(&switch_work); +#endif + /* wake up the daemon to read what remains */ wake_up_buffer_waiter(); out: @@ -155,6 +190,36 @@ post_sync: mutex_unlock(&start_mutex); } +#ifdef CONFIG_OPROFILE_EVENT_MULTIPLEX + +/* User inputs in ms, converts to jiffies */ +int oprofile_set_timeout(unsigned long val_msec) +{ + int err = 0; + + mutex_lock(&start_mutex); + + if (oprofile_started) { + err = -EBUSY; + goto out; + } + + if (!oprofile_ops.switch_events) { + err = -EINVAL; + goto out; + } + + timeout_jiffies = msecs_to_jiffies(val_msec); + if (timeout_jiffies == MAX_JIFFY_OFFSET) + timeout_jiffies = msecs_to_jiffies(MULTIPLEXING_TIMER_DEFAULT); + +out: + mutex_unlock(&start_mutex); + return err; + +} + +#endif int oprofile_set_backtrace(unsigned long val) { @@ -179,10 +244,23 @@ out: return err; } +#ifdef CONFIG_OPROFILE_EVENT_MULTIPLEX + +static void __init oprofile_multiplexing_init(void) +{ + timeout_jiffies = msecs_to_jiffies(MULTIPLEXING_TIMER_DEFAULT); +} + +#endif + static int __init oprofile_init(void) { int err; +#ifdef CONFIG_OPROFILE_EVENT_MULTIPLEX + oprofile_multiplexing_init(); +#endif + err = oprofile_arch_init(&oprofile_ops); if (err < 0 || timer) { diff --git a/drivers/oprofile/oprof.h b/drivers/oprofile/oprof.h index c288d3c24b50..ee38abcc74f3 100644 --- a/drivers/oprofile/oprof.h +++ b/drivers/oprofile/oprof.h @@ -27,6 +27,7 @@ extern unsigned long oprofile_buffer_watershed; extern struct oprofile_operations oprofile_ops; extern unsigned long oprofile_started; extern unsigned long oprofile_backtrace_depth; +extern unsigned long timeout_jiffies; struct super_block; struct dentry; @@ -35,5 +36,6 @@ void oprofile_create_files(struct super_block *sb, struct dentry *root); void oprofile_timer_init(struct oprofile_operations *ops); int oprofile_set_backtrace(unsigned long depth); +int oprofile_set_timeout(unsigned long time); #endif /* OPROF_H */ diff --git a/drivers/oprofile/oprofile_files.c b/drivers/oprofile/oprofile_files.c index 5d36ffc30dd5..468ec3e4f856 100644 --- a/drivers/oprofile/oprofile_files.c +++ b/drivers/oprofile/oprofile_files.c @@ -9,6 +9,7 @@ #include #include +#include #include "event_buffer.h" #include "oprofile_stats.h" @@ -22,6 +23,45 @@ unsigned long oprofile_buffer_size; unsigned long oprofile_cpu_buffer_size; unsigned long oprofile_buffer_watershed; +#ifdef CONFIG_OPROFILE_EVENT_MULTIPLEX + +static ssize_t timeout_read(struct file *file, char __user *buf, + size_t count, loff_t *offset) +{ + return oprofilefs_ulong_to_user(jiffies_to_msecs(timeout_jiffies), + buf, count, offset); +} + + +static ssize_t timeout_write(struct file *file, char const __user *buf, + size_t count, loff_t *offset) +{ + unsigned long val; + int retval; + + if (*offset) + return -EINVAL; + + retval = oprofilefs_ulong_from_user(&val, buf, count); + if (retval) + return retval; + + retval = oprofile_set_timeout(val); + + if (retval) + return retval; + return count; +} + + +static const struct file_operations timeout_fops = { + .read = timeout_read, + .write = timeout_write, +}; + +#endif + + static ssize_t depth_read(struct file *file, char __user *buf, size_t count, loff_t *offset) { return oprofilefs_ulong_to_user(oprofile_backtrace_depth, buf, count, @@ -139,6 +179,9 @@ void oprofile_create_files(struct super_block *sb, struct dentry *root) oprofilefs_create_file(sb, root, "cpu_type", &cpu_type_fops); oprofilefs_create_file(sb, root, "backtrace_depth", &depth_fops); oprofilefs_create_file(sb, root, "pointer_size", &pointer_size_fops); +#ifdef CONFIG_OPROFILE_EVENT_MULTIPLEX + oprofilefs_create_file(sb, root, "time_slice", &timeout_fops); +#endif oprofile_create_stats_files(sb, root); if (oprofile_ops.create_files) oprofile_ops.create_files(sb, root); diff --git a/drivers/oprofile/oprofile_stats.c b/drivers/oprofile/oprofile_stats.c index 3c2270a8300c..77a57a6792f6 100644 --- a/drivers/oprofile/oprofile_stats.c +++ b/drivers/oprofile/oprofile_stats.c @@ -16,6 +16,9 @@ #include "cpu_buffer.h" struct oprofile_stat_struct oprofile_stats; +#ifdef CONFIG_OPROFILE_EVENT_MULTIPLEX +atomic_t multiplex_counter; +#endif void oprofile_reset_stats(void) { @@ -34,6 +37,9 @@ void oprofile_reset_stats(void) atomic_set(&oprofile_stats.sample_lost_no_mapping, 0); atomic_set(&oprofile_stats.event_lost_overflow, 0); atomic_set(&oprofile_stats.bt_lost_no_mapping, 0); +#ifdef CONFIG_OPROFILE_EVENT_MULTIPLEX + atomic_set(&multiplex_counter, 0); +#endif } @@ -76,4 +82,8 @@ void oprofile_create_stats_files(struct super_block *sb, struct dentry *root) &oprofile_stats.event_lost_overflow); oprofilefs_create_ro_atomic(sb, dir, "bt_lost_no_mapping", &oprofile_stats.bt_lost_no_mapping); +#ifdef CONFIG_OPROFILE_EVENT_MULTIPLEX + oprofilefs_create_ro_atomic(sb, dir, "multiplex_counter", + &multiplex_counter); +#endif } diff --git a/include/linux/oprofile.h b/include/linux/oprofile.h index d68d2ed94f15..5171639ecf0f 100644 --- a/include/linux/oprofile.h +++ b/include/linux/oprofile.h @@ -67,6 +67,9 @@ struct oprofile_operations { /* Initiate a stack backtrace. Optional. */ void (*backtrace)(struct pt_regs * const regs, unsigned int depth); + + /* Multiplex between different events. Optional. */ + int (*switch_events)(void); /* CPU identification string. */ char * cpu_type; }; -- cgit v1.2.3 From e3afe7b75ed8f809c1473ea9b39267487c187ccb Mon Sep 17 00:00:00 2001 From: John Dykstra Date: Thu, 16 Jul 2009 05:04:51 +0000 Subject: tcp: Fix MD5 signature checking on IPv4 mapped sockets Fix MD5 signature checking so that an IPv4 active open to an IPv6 socket can succeed. In particular, use the correct address family's signature generation function for the SYN/ACK. Reported-by: Stephen Hemminger Signed-off-by: John Dykstra Signed-off-by: David S. Miller --- include/net/tcp.h | 5 +++++ net/ipv4/tcp_ipv4.c | 1 + net/ipv4/tcp_output.c | 2 +- net/ipv6/tcp_ipv6.c | 1 + 4 files changed, 8 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/net/tcp.h b/include/net/tcp.h index 19f4150f4d4d..88af84306471 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -1425,6 +1425,11 @@ struct tcp_request_sock_ops { #ifdef CONFIG_TCP_MD5SIG struct tcp_md5sig_key *(*md5_lookup) (struct sock *sk, struct request_sock *req); + int (*calc_md5_hash) (char *location, + struct tcp_md5sig_key *md5, + struct sock *sk, + struct request_sock *req, + struct sk_buff *skb); #endif }; diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index 5a1ca2698c88..7c107eb876c8 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -1160,6 +1160,7 @@ struct request_sock_ops tcp_request_sock_ops __read_mostly = { #ifdef CONFIG_TCP_MD5SIG static struct tcp_request_sock_ops tcp_request_sock_ipv4_ops = { .md5_lookup = tcp_v4_reqsk_md5_lookup, + .calc_md5_hash = tcp_v4_md5_hash_skb, }; #endif diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c index 5bdf08d312d9..bd62712848fa 100644 --- a/net/ipv4/tcp_output.c +++ b/net/ipv4/tcp_output.c @@ -2261,7 +2261,7 @@ struct sk_buff *tcp_make_synack(struct sock *sk, struct dst_entry *dst, #ifdef CONFIG_TCP_MD5SIG /* Okay, we have all we need - do the md5 hash if needed */ if (md5) { - tp->af_specific->calc_md5_hash(md5_hash_location, + tcp_rsk(req)->af_specific->calc_md5_hash(md5_hash_location, md5, NULL, req, skb); } #endif diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c index 58810c65b635..ae3d65753562 100644 --- a/net/ipv6/tcp_ipv6.c +++ b/net/ipv6/tcp_ipv6.c @@ -896,6 +896,7 @@ struct request_sock_ops tcp6_request_sock_ops __read_mostly = { #ifdef CONFIG_TCP_MD5SIG static struct tcp_request_sock_ops tcp_request_sock_ipv6_ops = { .md5_lookup = tcp_v6_reqsk_md5_lookup, + .calc_md5_hash = tcp_v6_md5_hash_skb, }; #endif -- cgit v1.2.3 From 68fd60a8c8bca6af51805c45f286f0f2572ac977 Mon Sep 17 00:00:00 2001 From: Lai Jiangshan Date: Thu, 16 Jul 2009 10:53:34 +0800 Subject: tracing/events: add missing type info of dynamic arrays The format file doesn't contain enough information for __dynamic_array/__string. The type name is missing. Before: # cat format: name: irq_handler_entry ... field:__data_loc name; offset:16; size:2; After: # cat format: name: irq_handler_entry ... field:__data_loc char[] name; offset:16; size:2; Signed-off-by: Lai Jiangshan LKML-Reference: <4A5E962E.9020900@cn.fujitsu.com> Signed-off-by: Li Zefan Acked-by: Frederic Weisbecker Signed-off-by: Steven Rostedt --- include/trace/ftrace.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/include/trace/ftrace.h b/include/trace/ftrace.h index 1867553c61e5..cc78943e0038 100644 --- a/include/trace/ftrace.h +++ b/include/trace/ftrace.h @@ -120,7 +120,7 @@ #undef __dynamic_array #define __dynamic_array(type, item, len) \ - ret = trace_seq_printf(s, "\tfield:__data_loc " #item ";\t" \ + ret = trace_seq_printf(s, "\tfield:__data_loc " #type "[] " #item ";\t"\ "offset:%u;\tsize:%u;\n", \ (unsigned int)offsetof(typeof(field), \ __data_loc_##item), \ @@ -279,7 +279,7 @@ ftrace_raw_output_##call(struct trace_iterator *iter, int flags) \ #undef __dynamic_array #define __dynamic_array(type, item, len) \ - ret = trace_define_field(event_call, "__data_loc" "[" #type "]", #item,\ + ret = trace_define_field(event_call, "__data_loc " #type "[]", #item, \ offsetof(typeof(field), __data_loc_##item), \ sizeof(field.__data_loc_##item), 0); -- cgit v1.2.3 From 4edf547b4d0f886acf5aa5a0c8f8edbaff280830 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Wed, 15 Jul 2009 06:16:34 +0000 Subject: net: explain netns notifiers a little better Eric explained this to me -- and afterwards the comment made sense, but not before. Add the the critical point about interfaces having to be gone from the netns before subsys notifiers are called. Signed-off-by: Johannes Berg Signed-off-by: David S. Miller --- include/net/net_namespace.h | 16 +++++++++------- 1 file changed, 9 insertions(+), 7 deletions(-) (limited to 'include') diff --git a/include/net/net_namespace.h b/include/net/net_namespace.h index 5c5136fceea8..a1202841aadd 100644 --- a/include/net/net_namespace.h +++ b/include/net/net_namespace.h @@ -239,13 +239,15 @@ struct pernet_operations { * needs per network namespace operations use device pernet operations, * otherwise use pernet subsys operations. * - * This is critically important. Most of the network code cleanup - * runs with the assumption that dev_remove_pack has been called so no - * new packets will arrive during and after the cleanup functions have - * been called. dev_remove_pack is not per namespace so instead the - * guarantee of no more packets arriving in a network namespace is - * provided by ensuring that all network devices and all sockets have - * left the network namespace before the cleanup methods are called. + * Network interfaces need to be removed from a dying netns _before_ + * subsys notifiers can be called, as most of the network code cleanup + * (which is done from subsys notifiers) runs with the assumption that + * dev_remove_pack has been called so no new packets will arrive during + * and after the cleanup functions have been called. dev_remove_pack + * is not per namespace so instead the guarantee of no more packets + * arriving in a network namespace is provided by ensuring that all + * network devices and all sockets have left the network namespace + * before the cleanup methods are called. * * For the longest time the ipv4 icmp code was registered as a pernet * device which caused kernel oops, and panics during network -- cgit v1.2.3 From 7d536cb3fb9993bdcd5a2fbaa6b0670ded4e101c Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Thu, 16 Jul 2009 10:54:02 +0800 Subject: tracing/events: record the size of dynamic arrays When a dynamic array is defined, we add __data_loc_foo in trace_entry to record the offset of the array, but the size of the array is not recorded, which causes 2 problems: - the event filter just compares the first 2 chars of the strings. - parsers can't parse dynamic arrays. So we encode the size of each dynamic array in the higher 16 bits of __data_loc_foo, while the offset is in lower 16 bits. Signed-off-by: Li Zefan LKML-Reference: <4A5E964A.9000403@cn.fujitsu.com> Acked-by: Frederic Weisbecker Signed-off-by: Steven Rostedt --- include/trace/ftrace.h | 14 ++++++++------ kernel/trace/trace_events_filter.c | 6 ++++-- 2 files changed, 12 insertions(+), 8 deletions(-) (limited to 'include') diff --git a/include/trace/ftrace.h b/include/trace/ftrace.h index cc78943e0038..3cbb96ef34f4 100644 --- a/include/trace/ftrace.h +++ b/include/trace/ftrace.h @@ -25,7 +25,7 @@ #define __array(type, item, len) type item[len]; #undef __dynamic_array -#define __dynamic_array(type, item, len) unsigned short __data_loc_##item; +#define __dynamic_array(type, item, len) u32 __data_loc_##item; #undef __string #define __string(item, src) __dynamic_array(char, item, -1) @@ -51,13 +51,14 @@ * Include the following: * * struct ftrace_data_offsets_ { - * int ; - * int ; + * u32 ; + * u32 ; * [...] * }; * - * The __dynamic_array() macro will create each int , this is + * The __dynamic_array() macro will create each u32 , this is * to keep the offset of each array from the beginning of the event. + * The size of an array is also encoded, in the higher 16 bits of . */ #undef __field @@ -67,7 +68,7 @@ #define __array(type, item, len) #undef __dynamic_array -#define __dynamic_array(type, item, len) int item; +#define __dynamic_array(type, item, len) u32 item; #undef __string #define __string(item, src) __dynamic_array(char, item, -1) @@ -207,7 +208,7 @@ ftrace_format_##call(struct trace_seq *s) \ #undef __get_dynamic_array #define __get_dynamic_array(field) \ - ((void *)__entry + __entry->__data_loc_##field) + ((void *)__entry + (__entry->__data_loc_##field & 0xffff)) #undef __get_str #define __get_str(field) (char *)__get_dynamic_array(field) @@ -325,6 +326,7 @@ ftrace_define_fields_##call(void) \ #define __dynamic_array(type, item, len) \ __data_offsets->item = __data_size + \ offsetof(typeof(*entry), __data); \ + __data_offsets->item |= (len * sizeof(type)) << 16; \ __data_size += (len) * sizeof(type); #undef __string diff --git a/kernel/trace/trace_events_filter.c b/kernel/trace/trace_events_filter.c index b9aae72d13db..1c80ef702b83 100644 --- a/kernel/trace/trace_events_filter.c +++ b/kernel/trace/trace_events_filter.c @@ -176,11 +176,13 @@ static int filter_pred_string(struct filter_pred *pred, void *event, static int filter_pred_strloc(struct filter_pred *pred, void *event, int val1, int val2) { - unsigned short str_loc = *(unsigned short *)(event + pred->offset); + u32 str_item = *(u32 *)(event + pred->offset); + int str_loc = str_item & 0xffff; + int str_len = str_item >> 16; char *addr = (char *)(event + str_loc); int cmp, match; - cmp = strncmp(addr, pred->str_val, pred->str_len); + cmp = strncmp(addr, pred->str_val, str_len); match = (!cmp) ^ pred->not; -- cgit v1.2.3 From 1f9963cbb0280e0cd554161e00f1a0eeddbf1ae1 Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Mon, 20 Jul 2009 10:20:53 +0800 Subject: tracing/filters: improve subsystem filter Currently a subsystem filter should be applicable to all events under the subsystem, and if it failed, all the event filters will be cleared. Those behaviors make subsys filter much less useful: # echo 'vec == 1' > irq/softirq_entry/filter # echo 'irq == 5' > irq/filter bash: echo: write error: Invalid argument # cat irq/softirq_entry/filter none I'd expect it set the filter for irq_handler_entry/exit, and not touch softirq_entry/exit. The basic idea is, try to see if the filter can be applied to which events, and then just apply to the those events: # echo 'vec == 1' > softirq_entry/filter # echo 'irq == 5' > filter # cat irq_handler_entry/filter irq == 5 # cat softirq_entry/filter vec == 1 Changelog for v2: - do some cleanups to address Frederic's comments. Inspired-by: Steven Rostedt Signed-off-by: Li Zefan Acked-by: Frederic Weisbecker LKML-Reference: <4A63D485.7030703@cn.fujitsu.com> Signed-off-by: Steven Rostedt --- include/linux/ftrace_event.h | 4 +- kernel/trace/trace.h | 3 +- kernel/trace/trace_events_filter.c | 124 ++++++++++++++++++++++++------------- 3 files changed, 87 insertions(+), 44 deletions(-) (limited to 'include') diff --git a/include/linux/ftrace_event.h b/include/linux/ftrace_event.h index 5c093ffc655b..26d3673d5143 100644 --- a/include/linux/ftrace_event.h +++ b/include/linux/ftrace_event.h @@ -101,6 +101,8 @@ void trace_current_buffer_discard_commit(struct ring_buffer_event *event); void tracing_record_cmdline(struct task_struct *tsk); +struct event_filter; + struct ftrace_event_call { struct list_head list; char *name; @@ -116,7 +118,7 @@ struct ftrace_event_call { int (*define_fields)(void); struct list_head fields; int filter_active; - void *filter; + struct event_filter *filter; void *mod; #ifdef CONFIG_EVENT_PROFILE diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h index 94305c7bc11c..758b0dbed552 100644 --- a/kernel/trace/trace.h +++ b/kernel/trace/trace.h @@ -750,13 +750,14 @@ struct event_filter { int n_preds; struct filter_pred **preds; char *filter_string; + bool no_reset; }; struct event_subsystem { struct list_head list; const char *name; struct dentry *entry; - void *filter; + struct event_filter *filter; int nr_events; }; diff --git a/kernel/trace/trace_events_filter.c b/kernel/trace/trace_events_filter.c index 1c80ef702b83..27c2dbea3710 100644 --- a/kernel/trace/trace_events_filter.c +++ b/kernel/trace/trace_events_filter.c @@ -420,7 +420,14 @@ oom: } EXPORT_SYMBOL_GPL(init_preds); -static void filter_free_subsystem_preds(struct event_subsystem *system) +enum { + FILTER_DISABLE_ALL, + FILTER_INIT_NO_RESET, + FILTER_SKIP_NO_RESET, +}; + +static void filter_free_subsystem_preds(struct event_subsystem *system, + int flag) { struct ftrace_event_call *call; @@ -428,6 +435,14 @@ static void filter_free_subsystem_preds(struct event_subsystem *system) if (!call->define_fields) continue; + if (flag == FILTER_INIT_NO_RESET) { + call->filter->no_reset = false; + continue; + } + + if (flag == FILTER_SKIP_NO_RESET && call->filter->no_reset) + continue; + if (!strcmp(call->system, system->name)) { filter_disable_preds(call); remove_filter_string(call->filter); @@ -529,7 +544,8 @@ static filter_pred_fn_t select_comparison_fn(int op, int field_size, static int filter_add_pred(struct filter_parse_state *ps, struct ftrace_event_call *call, - struct filter_pred *pred) + struct filter_pred *pred, + bool dry_run) { struct ftrace_event_field *field; filter_pred_fn_t fn; @@ -541,10 +557,12 @@ static int filter_add_pred(struct filter_parse_state *ps, if (pred->op == OP_AND) { pred->pop_n = 2; - return filter_add_pred_fn(ps, call, pred, filter_pred_and); + fn = filter_pred_and; + goto add_pred_fn; } else if (pred->op == OP_OR) { pred->pop_n = 2; - return filter_add_pred_fn(ps, call, pred, filter_pred_or); + fn = filter_pred_or; + goto add_pred_fn; } field = find_event_field(call, pred->field_name); @@ -567,9 +585,6 @@ static int filter_add_pred(struct filter_parse_state *ps, else fn = filter_pred_strloc; pred->str_len = field->size; - if (pred->op == OP_NE) - pred->not = 1; - return filter_add_pred_fn(ps, call, pred, fn); } else { if (field->is_signed) ret = strict_strtoll(pred->str_val, 0, &val); @@ -580,27 +595,33 @@ static int filter_add_pred(struct filter_parse_state *ps, return -EINVAL; } pred->val = val; - } - fn = select_comparison_fn(pred->op, field->size, field->is_signed); - if (!fn) { - parse_error(ps, FILT_ERR_INVALID_OP, 0); - return -EINVAL; + fn = select_comparison_fn(pred->op, field->size, + field->is_signed); + if (!fn) { + parse_error(ps, FILT_ERR_INVALID_OP, 0); + return -EINVAL; + } } if (pred->op == OP_NE) pred->not = 1; - return filter_add_pred_fn(ps, call, pred, fn); +add_pred_fn: + if (!dry_run) + return filter_add_pred_fn(ps, call, pred, fn); + return 0; } static int filter_add_subsystem_pred(struct filter_parse_state *ps, struct event_subsystem *system, struct filter_pred *pred, - char *filter_string) + char *filter_string, + bool dry_run) { struct ftrace_event_call *call; int err = 0; + bool fail = true; list_for_each_entry(call, &ftrace_events, list) { @@ -610,16 +631,24 @@ static int filter_add_subsystem_pred(struct filter_parse_state *ps, if (strcmp(call->system, system->name)) continue; - err = filter_add_pred(ps, call, pred); - if (err) { - filter_free_subsystem_preds(system); - parse_error(ps, FILT_ERR_BAD_SUBSYS_FILTER, 0); - goto out; - } - replace_filter_string(call->filter, filter_string); + if (call->filter->no_reset) + continue; + + err = filter_add_pred(ps, call, pred, dry_run); + if (err) + call->filter->no_reset = true; + else + fail = false; + + if (!dry_run) + replace_filter_string(call->filter, filter_string); } -out: - return err; + + if (fail) { + parse_error(ps, FILT_ERR_BAD_SUBSYS_FILTER, 0); + return err; + } + return 0; } static void parse_init(struct filter_parse_state *ps, @@ -978,12 +1007,14 @@ static int check_preds(struct filter_parse_state *ps) static int replace_preds(struct event_subsystem *system, struct ftrace_event_call *call, struct filter_parse_state *ps, - char *filter_string) + char *filter_string, + bool dry_run) { char *operand1 = NULL, *operand2 = NULL; struct filter_pred *pred; struct postfix_elt *elt; int err; + int n_preds = 0; err = check_preds(ps); if (err) @@ -1002,19 +1033,14 @@ static int replace_preds(struct event_subsystem *system, continue; } + if (n_preds++ == MAX_FILTER_PRED) { + parse_error(ps, FILT_ERR_TOO_MANY_PREDS, 0); + return -ENOSPC; + } + if (elt->op == OP_AND || elt->op == OP_OR) { pred = create_logical_pred(elt->op); - if (call) - err = filter_add_pred(ps, call, pred); - else - err = filter_add_subsystem_pred(ps, system, - pred, filter_string); - filter_free_pred(pred); - if (err) - return err; - - operand1 = operand2 = NULL; - continue; + goto add_pred; } if (!operand1 || !operand2) { @@ -1023,11 +1049,12 @@ static int replace_preds(struct event_subsystem *system, } pred = create_pred(elt->op, operand1, operand2); +add_pred: if (call) - err = filter_add_pred(ps, call, pred); + err = filter_add_pred(ps, call, pred, false); else err = filter_add_subsystem_pred(ps, system, pred, - filter_string); + filter_string, dry_run); filter_free_pred(pred); if (err) return err; @@ -1068,7 +1095,7 @@ int apply_event_filter(struct ftrace_event_call *call, char *filter_string) goto out; } - err = replace_preds(NULL, call, ps, filter_string); + err = replace_preds(NULL, call, ps, filter_string, false); if (err) append_filter_err(ps, call->filter); @@ -1092,7 +1119,7 @@ int apply_subsystem_event_filter(struct event_subsystem *system, mutex_lock(&event_mutex); if (!strcmp(strstrip(filter_string), "0")) { - filter_free_subsystem_preds(system); + filter_free_subsystem_preds(system, FILTER_DISABLE_ALL); remove_filter_string(system->filter); mutex_unlock(&event_mutex); return 0; @@ -1103,7 +1130,6 @@ int apply_subsystem_event_filter(struct event_subsystem *system, if (!ps) goto out_unlock; - filter_free_subsystem_preds(system); replace_filter_string(system->filter, filter_string); parse_init(ps, filter_ops, filter_string); @@ -1113,9 +1139,23 @@ int apply_subsystem_event_filter(struct event_subsystem *system, goto out; } - err = replace_preds(system, NULL, ps, filter_string); - if (err) + filter_free_subsystem_preds(system, FILTER_INIT_NO_RESET); + + /* try to see the filter can be applied to which events */ + err = replace_preds(system, NULL, ps, filter_string, true); + if (err) { + append_filter_err(ps, system->filter); + goto out; + } + + filter_free_subsystem_preds(system, FILTER_SKIP_NO_RESET); + + /* really apply the filter to the events */ + err = replace_preds(system, NULL, ps, filter_string, false); + if (err) { append_filter_err(ps, system->filter); + filter_free_subsystem_preds(system, 2); + } out: filter_opstack_clear(ps); -- cgit v1.2.3 From 99fde513f57db2c8e1b202ade4be7d47033ff09b Mon Sep 17 00:00:00 2001 From: Marek Vasut Date: Mon, 20 Jul 2009 22:28:50 -0700 Subject: Input: wm97xx - add possibility to control the GPIO_STATUS shift This patch allows tweaking the behaviour of GPIO_STATUS register shift quirk that's in wm97xx-core. The problem with GPIO_STATUS register being shifted by one doesn't appear on all hardware it seems and causes problems with accelerated touchscreen drivers on Palm hardware. Therefore an accelerated touchscreen driver can select if the shift is/isn't happening on the hardware. Signed-off-by: Marek Vasut Acked-by: Mark Brown Signed-off-by: Dmitry Torokhov --- drivers/input/touchscreen/mainstone-wm97xx.c | 3 +++ drivers/input/touchscreen/wm97xx-core.c | 6 ++++-- include/linux/wm97xx.h | 7 +++++++ 3 files changed, 14 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/drivers/input/touchscreen/mainstone-wm97xx.c b/drivers/input/touchscreen/mainstone-wm97xx.c index c797bc04ee83..8fc3b08deb3b 100644 --- a/drivers/input/touchscreen/mainstone-wm97xx.c +++ b/drivers/input/touchscreen/mainstone-wm97xx.c @@ -198,6 +198,9 @@ static int wm97xx_acc_startup(struct wm97xx *wm) if (machine_is_palmt5() || machine_is_palmtx() || machine_is_palmld()) { pen_int = 1; irq = 27; + /* There is some obscure mutant of WM9712 interbred with WM9713 + * used on Palm HW */ + wm->variant = WM97xx_WM1613; } else if (machine_is_mainstone() && pen_int) irq = 4; diff --git a/drivers/input/touchscreen/wm97xx-core.c b/drivers/input/touchscreen/wm97xx-core.c index 2957d48e0045..252eb11fe9db 100644 --- a/drivers/input/touchscreen/wm97xx-core.c +++ b/drivers/input/touchscreen/wm97xx-core.c @@ -204,7 +204,7 @@ void wm97xx_set_gpio(struct wm97xx *wm, u32 gpio, else reg &= ~gpio; - if (wm->id == WM9712_ID2) + if (wm->id == WM9712_ID2 && wm->variant != WM97xx_WM1613) wm97xx_reg_write(wm, AC97_GPIO_STATUS, reg << 1); else wm97xx_reg_write(wm, AC97_GPIO_STATUS, reg); @@ -307,7 +307,7 @@ static void wm97xx_pen_irq_worker(struct work_struct *work) WM97XX_GPIO_13); } - if (wm->id == WM9712_ID2) + if (wm->id == WM9712_ID2 && wm->variant != WM97xx_WM1613) wm97xx_reg_write(wm, AC97_GPIO_STATUS, (status & ~WM97XX_GPIO_13) << 1); else @@ -582,6 +582,8 @@ static int wm97xx_probe(struct device *dev) wm->id = wm97xx_reg_read(wm, AC97_VENDOR_ID2); + wm->variant = WM97xx_GENERIC; + dev_info(wm->dev, "detected a wm97%02x codec\n", wm->id & 0xff); switch (wm->id & 0xff) { diff --git a/include/linux/wm97xx.h b/include/linux/wm97xx.h index 6f69968eab24..0c9878123d5f 100644 --- a/include/linux/wm97xx.h +++ b/include/linux/wm97xx.h @@ -15,6 +15,12 @@ #include /* Input device layer */ #include +/* + * WM97xx variants + */ +#define WM97xx_GENERIC 0x0000 +#define WM97xx_WM1613 0x1613 + /* * WM97xx AC97 Touchscreen registers */ @@ -283,6 +289,7 @@ struct wm97xx { unsigned pen_is_down:1; /* Pen is down */ unsigned aux_waiting:1; /* aux measurement waiting */ unsigned pen_probably_down:1; /* used in polling mode */ + u16 variant; /* WM97xx chip variant */ u16 suspend_mode; /* PRP in suspend mode */ }; -- cgit v1.2.3 From 591d2fb02ea80472d846c0b8507007806bdd69cc Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Tue, 21 Jul 2009 11:09:39 +0200 Subject: genirq: Delegate irq affinity setting to the irq thread irq_set_thread_affinity() calls set_cpus_allowed_ptr() which might sleep, but irq_set_thread_affinity() is called with desc->lock held and can be called from hard interrupt context as well. The code has another bug as it does not hold a ref on the task struct as required by set_cpus_allowed_ptr(). Just set the IRQTF_AFFINITY bit in action->thread_flags. The next time the thread runs it migrates itself. Solves all of the above problems nicely. Add kerneldoc to irq_set_thread_affinity() while at it. Signed-off-by: Thomas Gleixner LKML-Reference: --- include/linux/interrupt.h | 2 ++ kernel/irq/internals.h | 3 +-- kernel/irq/manage.c | 50 +++++++++++++++++++++++++++++++++++++++++------ kernel/irq/migration.c | 2 +- 4 files changed, 48 insertions(+), 9 deletions(-) (limited to 'include') diff --git a/include/linux/interrupt.h b/include/linux/interrupt.h index 2721f07e9354..88b056ac5629 100644 --- a/include/linux/interrupt.h +++ b/include/linux/interrupt.h @@ -64,11 +64,13 @@ * IRQTF_RUNTHREAD - signals that the interrupt handler thread should run * IRQTF_DIED - handler thread died * IRQTF_WARNED - warning "IRQ_WAKE_THREAD w/o thread_fn" has been printed + * IRQTF_AFFINITY - irq thread is requested to adjust affinity */ enum { IRQTF_RUNTHREAD, IRQTF_DIED, IRQTF_WARNED, + IRQTF_AFFINITY, }; typedef irqreturn_t (*irq_handler_t)(int, void *); diff --git a/kernel/irq/internals.h b/kernel/irq/internals.h index 73468253143b..e70ed5592eb9 100644 --- a/kernel/irq/internals.h +++ b/kernel/irq/internals.h @@ -42,8 +42,7 @@ static inline void unregister_handler_proc(unsigned int irq, extern int irq_select_affinity_usr(unsigned int irq); -extern void -irq_set_thread_affinity(struct irq_desc *desc, const struct cpumask *cpumask); +extern void irq_set_thread_affinity(struct irq_desc *desc); /* * Debugging printout: diff --git a/kernel/irq/manage.c b/kernel/irq/manage.c index 50da67672901..f0de36f13a44 100644 --- a/kernel/irq/manage.c +++ b/kernel/irq/manage.c @@ -80,14 +80,22 @@ int irq_can_set_affinity(unsigned int irq) return 1; } -void -irq_set_thread_affinity(struct irq_desc *desc, const struct cpumask *cpumask) +/** + * irq_set_thread_affinity - Notify irq threads to adjust affinity + * @desc: irq descriptor which has affitnity changed + * + * We just set IRQTF_AFFINITY and delegate the affinity setting + * to the interrupt thread itself. We can not call + * set_cpus_allowed_ptr() here as we hold desc->lock and this + * code can be called from hard interrupt context. + */ +void irq_set_thread_affinity(struct irq_desc *desc) { struct irqaction *action = desc->action; while (action) { if (action->thread) - set_cpus_allowed_ptr(action->thread, cpumask); + set_bit(IRQTF_AFFINITY, &action->thread_flags); action = action->next; } } @@ -112,7 +120,7 @@ int irq_set_affinity(unsigned int irq, const struct cpumask *cpumask) if (desc->status & IRQ_MOVE_PCNTXT) { if (!desc->chip->set_affinity(irq, cpumask)) { cpumask_copy(desc->affinity, cpumask); - irq_set_thread_affinity(desc, cpumask); + irq_set_thread_affinity(desc); } } else { @@ -122,7 +130,7 @@ int irq_set_affinity(unsigned int irq, const struct cpumask *cpumask) #else if (!desc->chip->set_affinity(irq, cpumask)) { cpumask_copy(desc->affinity, cpumask); - irq_set_thread_affinity(desc, cpumask); + irq_set_thread_affinity(desc); } #endif desc->status |= IRQ_AFFINITY_SET; @@ -176,7 +184,7 @@ int irq_select_affinity_usr(unsigned int irq) spin_lock_irqsave(&desc->lock, flags); ret = setup_affinity(irq, desc); if (!ret) - irq_set_thread_affinity(desc, desc->affinity); + irq_set_thread_affinity(desc); spin_unlock_irqrestore(&desc->lock, flags); return ret; @@ -443,6 +451,34 @@ static int irq_wait_for_interrupt(struct irqaction *action) return -1; } +/* + * Check whether we need to change the affinity of the interrupt thread. + */ +static void +irq_thread_check_affinity(struct irq_desc *desc, struct irqaction *action) +{ + cpumask_var_t mask; + + if (!test_and_clear_bit(IRQTF_AFFINITY, &action->thread_flags)) + return; + + /* + * In case we are out of memory we set IRQTF_AFFINITY again and + * try again next time + */ + if (!alloc_cpumask_var(&mask, GFP_KERNEL)) { + set_bit(IRQTF_AFFINITY, &action->thread_flags); + return; + } + + spin_lock_irq(&desc->lock); + cpumask_copy(mask, desc->affinity); + spin_unlock_irq(&desc->lock); + + set_cpus_allowed_ptr(current, mask); + free_cpumask_var(mask); +} + /* * Interrupt handler thread */ @@ -458,6 +494,8 @@ static int irq_thread(void *data) while (!irq_wait_for_interrupt(action)) { + irq_thread_check_affinity(desc, action); + atomic_inc(&desc->threads_active); spin_lock_irq(&desc->lock); diff --git a/kernel/irq/migration.c b/kernel/irq/migration.c index cfe767ca1545..fcb6c96f2627 100644 --- a/kernel/irq/migration.c +++ b/kernel/irq/migration.c @@ -45,7 +45,7 @@ void move_masked_irq(int irq) < nr_cpu_ids)) if (!desc->chip->set_affinity(irq, desc->pending_mask)) { cpumask_copy(desc->affinity, desc->pending_mask); - irq_set_thread_affinity(desc, desc->pending_mask); + irq_set_thread_affinity(desc); } cpumask_clear(desc->pending_mask); -- cgit v1.2.3 From e56f0975360369347725c49654ecfe3792710429 Mon Sep 17 00:00:00 2001 From: Alan Jenkins Date: Sat, 18 Jul 2009 19:20:20 +0100 Subject: rfkill: remove too-strict __must_check Some drivers don't need the return value of rfkill_set_hw_state(), so it should not be marked as __must_check. Signed-off-by: Alan Jenkins Acked-by: Johannes Berg Signed-off-by: John W. Linville --- include/linux/rfkill.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/rfkill.h b/include/linux/rfkill.h index 2ce29831feb6..278777fa8a3a 100644 --- a/include/linux/rfkill.h +++ b/include/linux/rfkill.h @@ -224,7 +224,7 @@ void rfkill_destroy(struct rfkill *rfkill); * should be blocked) so that drivers need not keep track of the soft * block state -- which they might not be able to. */ -bool __must_check rfkill_set_hw_state(struct rfkill *rfkill, bool blocked); +bool rfkill_set_hw_state(struct rfkill *rfkill, bool blocked); /** * rfkill_set_sw_state - Set the internal rfkill software block state -- cgit v1.2.3 From f44aebcc566d1d6275f7191867b9633dc11de2ee Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Wed, 15 Jul 2009 15:49:52 -0400 Subject: inotify: use GFP_NOFS under potential memory pressure inotify can have a watchs removed under filesystem reclaim. ================================= [ INFO: inconsistent lock state ] 2.6.31-rc2 #16 --------------------------------- inconsistent {IN-RECLAIM_FS-W} -> {RECLAIM_FS-ON-W} usage. khubd/217 [HC0[0]:SC0[0]:HE1:SE1] takes: (iprune_mutex){+.+.?.}, at: [] invalidate_inodes+0x20/0xe3 {IN-RECLAIM_FS-W} state was registered at: [] __lock_acquire+0x2c9/0xac4 [] lock_acquire+0x9f/0xc2 [] __mutex_lock_common+0x2d/0x323 [] mutex_lock_nested+0x2e/0x36 [] shrink_icache_memory+0x38/0x1b2 [] shrink_slab+0xe2/0x13c [] kswapd+0x3d1/0x55d [] kthread+0x66/0x6b [] kernel_thread_helper+0x7/0x10 [] 0xffffffff Two things are needed to fix this. First we need a method to tell fsnotify_create_event() to use GFP_NOFS and second we need to stop using one global IN_IGNORED event and allocate them one at a time. This solves current issues with multiple IN_IGNORED on a queue having tail drop problems and simplifies the allocations since we don't have to worry about two tasks opperating on the IGNORED event concurrently. Signed-off-by: Eric Paris --- fs/notify/fsnotify.c | 4 +++- fs/notify/inotify/inotify_user.c | 18 ++++++++++++------ fs/notify/notification.c | 9 +++++---- include/linux/fsnotify_backend.h | 2 +- 4 files changed, 21 insertions(+), 12 deletions(-) (limited to 'include') diff --git a/fs/notify/fsnotify.c b/fs/notify/fsnotify.c index ec2f7bd76818..037e878e03fc 100644 --- a/fs/notify/fsnotify.c +++ b/fs/notify/fsnotify.c @@ -159,7 +159,9 @@ void fsnotify(struct inode *to_tell, __u32 mask, void *data, int data_is, const if (!group->ops->should_send_event(group, to_tell, mask)) continue; if (!event) { - event = fsnotify_create_event(to_tell, mask, data, data_is, file_name, cookie); + event = fsnotify_create_event(to_tell, mask, data, + data_is, file_name, cookie, + GFP_KERNEL); /* shit, we OOM'd and now we can't tell, maybe * someday someone else will want to do something * here */ diff --git a/fs/notify/inotify/inotify_user.c b/fs/notify/inotify/inotify_user.c index 726118a5845b..f30d9bbc2e1b 100644 --- a/fs/notify/inotify/inotify_user.c +++ b/fs/notify/inotify/inotify_user.c @@ -57,7 +57,6 @@ int inotify_max_user_watches __read_mostly; static struct kmem_cache *inotify_inode_mark_cachep __read_mostly; struct kmem_cache *event_priv_cachep __read_mostly; -static struct fsnotify_event *inotify_ignored_event; /* * When inotify registers a new group it increments this and uses that @@ -384,12 +383,19 @@ void inotify_ignored_and_remove_idr(struct fsnotify_mark_entry *entry, struct fsnotify_group *group) { struct inotify_inode_mark_entry *ientry; + struct fsnotify_event *ignored_event; struct inotify_event_private_data *event_priv; struct fsnotify_event_private_data *fsn_event_priv; + ignored_event = fsnotify_create_event(NULL, FS_IN_IGNORED, NULL, + FSNOTIFY_EVENT_NONE, NULL, 0, + GFP_NOFS); + if (!ignored_event) + return; + ientry = container_of(entry, struct inotify_inode_mark_entry, fsn_entry); - event_priv = kmem_cache_alloc(event_priv_cachep, GFP_KERNEL); + event_priv = kmem_cache_alloc(event_priv_cachep, GFP_NOFS); if (unlikely(!event_priv)) goto skip_send_ignore; @@ -398,7 +404,7 @@ void inotify_ignored_and_remove_idr(struct fsnotify_mark_entry *entry, fsn_event_priv->group = group; event_priv->wd = ientry->wd; - fsnotify_add_notify_event(group, inotify_ignored_event, fsn_event_priv); + fsnotify_add_notify_event(group, ignored_event, fsn_event_priv); /* did the private data get added? */ if (list_empty(&fsn_event_priv->event_list)) @@ -406,6 +412,9 @@ void inotify_ignored_and_remove_idr(struct fsnotify_mark_entry *entry, skip_send_ignore: + /* matches the reference taken when the event was created */ + fsnotify_put_event(ignored_event); + /* remove this entry from the idr */ inotify_remove_from_idr(group, ientry); @@ -748,9 +757,6 @@ static int __init inotify_user_setup(void) inotify_inode_mark_cachep = KMEM_CACHE(inotify_inode_mark_entry, SLAB_PANIC); event_priv_cachep = KMEM_CACHE(inotify_event_private_data, SLAB_PANIC); - inotify_ignored_event = fsnotify_create_event(NULL, FS_IN_IGNORED, NULL, FSNOTIFY_EVENT_NONE, NULL, 0); - if (!inotify_ignored_event) - panic("unable to allocate the inotify ignored event\n"); inotify_max_queued_events = 16384; inotify_max_user_instances = 128; diff --git a/fs/notify/notification.c b/fs/notify/notification.c index 2b20feaf263a..521368574e97 100644 --- a/fs/notify/notification.c +++ b/fs/notify/notification.c @@ -153,7 +153,7 @@ static bool event_compare(struct fsnotify_event *old, struct fsnotify_event *new return true; break; case (FSNOTIFY_EVENT_NONE): - return true; + return false; }; } return false; @@ -345,18 +345,19 @@ static void initialize_event(struct fsnotify_event *event) * @name the filename, if available */ struct fsnotify_event *fsnotify_create_event(struct inode *to_tell, __u32 mask, void *data, - int data_type, const char *name, u32 cookie) + int data_type, const char *name, u32 cookie, + gfp_t gfp) { struct fsnotify_event *event; - event = kmem_cache_alloc(fsnotify_event_cachep, GFP_KERNEL); + event = kmem_cache_alloc(fsnotify_event_cachep, gfp); if (!event) return NULL; initialize_event(event); if (name) { - event->file_name = kstrdup(name, GFP_KERNEL); + event->file_name = kstrdup(name, gfp); if (!event->file_name) { kmem_cache_free(fsnotify_event_cachep, event); return NULL; diff --git a/include/linux/fsnotify_backend.h b/include/linux/fsnotify_backend.h index 6c3de999fb34..4d6f47b51189 100644 --- a/include/linux/fsnotify_backend.h +++ b/include/linux/fsnotify_backend.h @@ -352,7 +352,7 @@ extern void fsnotify_unmount_inodes(struct list_head *list); /* put here because inotify does some weird stuff when destroying watches */ extern struct fsnotify_event *fsnotify_create_event(struct inode *to_tell, __u32 mask, void *data, int data_is, const char *name, - u32 cookie); + u32 cookie, gfp_t gfp); #else -- cgit v1.2.3 From d7aacaddcac3971e33cf52d7e610c06696cb347f Mon Sep 17 00:00:00 2001 From: Magnus Damm Date: Wed, 8 Jul 2009 13:21:31 +0200 Subject: Driver Core: Add platform device arch data V3 Allow architecture specific data in struct platform_device V3. With this patch struct pdev_archdata is added to struct platform_device, similar to struct dev_archdata in found in struct device. Useful for architecture code that needs to keep extra data associated with each platform device. Struct pdev_archdata is different from dev.platform_data, the convention is that dev.platform_data points to driver-specific data. It may or may not be required by the driver. The format of this depends on driver but is the same across architectures. The structure pdev_archdata is a place for architecture specific data. This data is handled by architecture specific code (for example runtime PM), and since it is architecture specific it should _never_ be touched by device driver code. Exactly like struct dev_archdata but for platform devices. [rjw: This change is for power management mostly and that's why it goes through the suspend tree.] Signed-off-by: Magnus Damm Acked-by: Kevin Hilman Acked-by: Greg Kroah-Hartman Signed-off-by: Rafael J. Wysocki --- arch/arm/include/asm/device.h | 3 +++ arch/ia64/include/asm/device.h | 3 +++ arch/microblaze/include/asm/device.h | 3 +++ arch/powerpc/include/asm/device.h | 3 +++ arch/sparc/include/asm/device.h | 3 +++ arch/x86/include/asm/device.h | 3 +++ include/asm-generic/device.h | 3 +++ include/linux/platform_device.h | 3 +++ 8 files changed, 24 insertions(+) (limited to 'include') diff --git a/arch/arm/include/asm/device.h b/arch/arm/include/asm/device.h index c61642b40603..9f390ce335cb 100644 --- a/arch/arm/include/asm/device.h +++ b/arch/arm/include/asm/device.h @@ -12,4 +12,7 @@ struct dev_archdata { #endif }; +struct pdev_archdata { +}; + #endif diff --git a/arch/ia64/include/asm/device.h b/arch/ia64/include/asm/device.h index 41ab85d66f33..d66d446b127c 100644 --- a/arch/ia64/include/asm/device.h +++ b/arch/ia64/include/asm/device.h @@ -15,4 +15,7 @@ struct dev_archdata { #endif }; +struct pdev_archdata { +}; + #endif /* _ASM_IA64_DEVICE_H */ diff --git a/arch/microblaze/include/asm/device.h b/arch/microblaze/include/asm/device.h index c042830793ed..30286db27c1c 100644 --- a/arch/microblaze/include/asm/device.h +++ b/arch/microblaze/include/asm/device.h @@ -16,6 +16,9 @@ struct dev_archdata { struct device_node *of_node; }; +struct pdev_archdata { +}; + #endif /* _ASM_MICROBLAZE_DEVICE_H */ diff --git a/arch/powerpc/include/asm/device.h b/arch/powerpc/include/asm/device.h index 7d2277cef09a..e3e06e0f7fc0 100644 --- a/arch/powerpc/include/asm/device.h +++ b/arch/powerpc/include/asm/device.h @@ -30,4 +30,7 @@ dev_archdata_get_node(const struct dev_archdata *ad) return ad->of_node; } +struct pdev_archdata { +}; + #endif /* _ASM_POWERPC_DEVICE_H */ diff --git a/arch/sparc/include/asm/device.h b/arch/sparc/include/asm/device.h index 3702e087df2c..f3b85b6b0b76 100644 --- a/arch/sparc/include/asm/device.h +++ b/arch/sparc/include/asm/device.h @@ -32,4 +32,7 @@ dev_archdata_get_node(const struct dev_archdata *ad) return ad->prom_node; } +struct pdev_archdata { +}; + #endif /* _ASM_SPARC_DEVICE_H */ diff --git a/arch/x86/include/asm/device.h b/arch/x86/include/asm/device.h index 4994a20acbcb..cee34e9ca45b 100644 --- a/arch/x86/include/asm/device.h +++ b/arch/x86/include/asm/device.h @@ -13,4 +13,7 @@ struct dma_map_ops *dma_ops; #endif }; +struct pdev_archdata { +}; + #endif /* _ASM_X86_DEVICE_H */ diff --git a/include/asm-generic/device.h b/include/asm-generic/device.h index c17c9600f220..d7c76bba640d 100644 --- a/include/asm-generic/device.h +++ b/include/asm-generic/device.h @@ -9,4 +9,7 @@ struct dev_archdata { }; +struct pdev_archdata { +}; + #endif /* _ASM_GENERIC_DEVICE_H */ diff --git a/include/linux/platform_device.h b/include/linux/platform_device.h index 8dc5123b6305..672a69849735 100644 --- a/include/linux/platform_device.h +++ b/include/linux/platform_device.h @@ -22,6 +22,9 @@ struct platform_device { struct resource * resource; struct platform_device_id *id_entry; + + /* arch specific additions */ + struct pdev_archdata archdata; }; #define platform_get_device_id(pdev) ((pdev)->id_entry) -- cgit v1.2.3 From 511647ff58fd0f1c1f415d2c757d841650edac91 Mon Sep 17 00:00:00 2001 From: Magnus Damm Date: Wed, 8 Jul 2009 13:23:07 +0200 Subject: PM: Remove platform device suspend_late()/resume_early() V2 This is V2 of the platform driver power management late/early callback removal patch. The callbacks ->suspend_late() and ->resume_early() are removed since all in-tree users now have been migrated to dev_pm_ops. Signed-off-by: Magnus Damm Acked-by: Greg Kroah-Hartman Acked-by: Pavel Machek Signed-off-by: Rafael J. Wysocki --- drivers/base/platform.c | 36 ------------------------------------ include/linux/platform_device.h | 2 -- 2 files changed, 38 deletions(-) (limited to 'include') diff --git a/drivers/base/platform.c b/drivers/base/platform.c index 81cb01bfc356..455e55971d0e 100644 --- a/drivers/base/platform.c +++ b/drivers/base/platform.c @@ -628,30 +628,6 @@ static int platform_legacy_suspend(struct device *dev, pm_message_t mesg) return ret; } -static int platform_legacy_suspend_late(struct device *dev, pm_message_t mesg) -{ - struct platform_driver *pdrv = to_platform_driver(dev->driver); - struct platform_device *pdev = to_platform_device(dev); - int ret = 0; - - if (dev->driver && pdrv->suspend_late) - ret = pdrv->suspend_late(pdev, mesg); - - return ret; -} - -static int platform_legacy_resume_early(struct device *dev) -{ - struct platform_driver *pdrv = to_platform_driver(dev->driver); - struct platform_device *pdev = to_platform_device(dev); - int ret = 0; - - if (dev->driver && pdrv->resume_early) - ret = pdrv->resume_early(pdev); - - return ret; -} - static int platform_legacy_resume(struct device *dev) { struct platform_driver *pdrv = to_platform_driver(dev->driver); @@ -714,8 +690,6 @@ static int platform_pm_suspend_noirq(struct device *dev) if (drv->pm) { if (drv->pm->suspend_noirq) ret = drv->pm->suspend_noirq(dev); - } else { - ret = platform_legacy_suspend_late(dev, PMSG_SUSPEND); } return ret; @@ -750,8 +724,6 @@ static int platform_pm_resume_noirq(struct device *dev) if (drv->pm) { if (drv->pm->resume_noirq) ret = drv->pm->resume_noirq(dev); - } else { - ret = platform_legacy_resume_early(dev); } return ret; @@ -797,8 +769,6 @@ static int platform_pm_freeze_noirq(struct device *dev) if (drv->pm) { if (drv->pm->freeze_noirq) ret = drv->pm->freeze_noirq(dev); - } else { - ret = platform_legacy_suspend_late(dev, PMSG_FREEZE); } return ret; @@ -833,8 +803,6 @@ static int platform_pm_thaw_noirq(struct device *dev) if (drv->pm) { if (drv->pm->thaw_noirq) ret = drv->pm->thaw_noirq(dev); - } else { - ret = platform_legacy_resume_early(dev); } return ret; @@ -869,8 +837,6 @@ static int platform_pm_poweroff_noirq(struct device *dev) if (drv->pm) { if (drv->pm->poweroff_noirq) ret = drv->pm->poweroff_noirq(dev); - } else { - ret = platform_legacy_suspend_late(dev, PMSG_HIBERNATE); } return ret; @@ -905,8 +871,6 @@ static int platform_pm_restore_noirq(struct device *dev) if (drv->pm) { if (drv->pm->restore_noirq) ret = drv->pm->restore_noirq(dev); - } else { - ret = platform_legacy_resume_early(dev); } return ret; diff --git a/include/linux/platform_device.h b/include/linux/platform_device.h index 672a69849735..3c6675c2444b 100644 --- a/include/linux/platform_device.h +++ b/include/linux/platform_device.h @@ -60,8 +60,6 @@ struct platform_driver { int (*remove)(struct platform_device *); void (*shutdown)(struct platform_device *); int (*suspend)(struct platform_device *, pm_message_t state); - int (*suspend_late)(struct platform_device *, pm_message_t state); - int (*resume_early)(struct platform_device *); int (*resume)(struct platform_device *); struct device_driver driver; struct platform_device_id *id_table; -- cgit v1.2.3 From 1f38ad8389bbca038d320c29d30aa1d6ed96b48d Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Wed, 22 Jul 2009 11:48:18 +0800 Subject: crypto: sha512 - Export struct sha512_state This patch renames struct sha512_ctx and exports it as struct sha512_state so that other sha512 implementations can use it as the reference structure for exporting their state. Signed-off-by: Herbert Xu --- crypto/sha512_generic.c | 20 +++++++------------- include/crypto/sha.h | 6 ++++++ 2 files changed, 13 insertions(+), 13 deletions(-) (limited to 'include') diff --git a/crypto/sha512_generic.c b/crypto/sha512_generic.c index 3bea38d12242..4fe95eb03226 100644 --- a/crypto/sha512_generic.c +++ b/crypto/sha512_generic.c @@ -21,12 +21,6 @@ #include #include -struct sha512_ctx { - u64 state[8]; - u32 count[4]; - u8 buf[128]; -}; - static DEFINE_PER_CPU(u64[80], msg_schedule); static inline u64 Ch(u64 x, u64 y, u64 z) @@ -141,7 +135,7 @@ sha512_transform(u64 *state, const u8 *input) static int sha512_init(struct shash_desc *desc) { - struct sha512_ctx *sctx = shash_desc_ctx(desc); + struct sha512_state *sctx = shash_desc_ctx(desc); sctx->state[0] = SHA512_H0; sctx->state[1] = SHA512_H1; sctx->state[2] = SHA512_H2; @@ -158,7 +152,7 @@ sha512_init(struct shash_desc *desc) static int sha384_init(struct shash_desc *desc) { - struct sha512_ctx *sctx = shash_desc_ctx(desc); + struct sha512_state *sctx = shash_desc_ctx(desc); sctx->state[0] = SHA384_H0; sctx->state[1] = SHA384_H1; sctx->state[2] = SHA384_H2; @@ -175,7 +169,7 @@ sha384_init(struct shash_desc *desc) static int sha512_update(struct shash_desc *desc, const u8 *data, unsigned int len) { - struct sha512_ctx *sctx = shash_desc_ctx(desc); + struct sha512_state *sctx = shash_desc_ctx(desc); unsigned int i, index, part_len; @@ -214,7 +208,7 @@ sha512_update(struct shash_desc *desc, const u8 *data, unsigned int len) static int sha512_final(struct shash_desc *desc, u8 *hash) { - struct sha512_ctx *sctx = shash_desc_ctx(desc); + struct sha512_state *sctx = shash_desc_ctx(desc); static u8 padding[128] = { 0x80, }; __be64 *dst = (__be64 *)hash; __be32 bits[4]; @@ -240,7 +234,7 @@ sha512_final(struct shash_desc *desc, u8 *hash) dst[i] = cpu_to_be64(sctx->state[i]); /* Zeroize sensitive information. */ - memset(sctx, 0, sizeof(struct sha512_ctx)); + memset(sctx, 0, sizeof(struct sha512_state)); return 0; } @@ -262,7 +256,7 @@ static struct shash_alg sha512 = { .init = sha512_init, .update = sha512_update, .final = sha512_final, - .descsize = sizeof(struct sha512_ctx), + .descsize = sizeof(struct sha512_state), .base = { .cra_name = "sha512", .cra_flags = CRYPTO_ALG_TYPE_SHASH, @@ -276,7 +270,7 @@ static struct shash_alg sha384 = { .init = sha384_init, .update = sha512_update, .final = sha384_final, - .descsize = sizeof(struct sha512_ctx), + .descsize = sizeof(struct sha512_state), .base = { .cra_name = "sha384", .cra_flags = CRYPTO_ALG_TYPE_SHASH, diff --git a/include/crypto/sha.h b/include/crypto/sha.h index 88ef5eb9514d..45b25ccf7cc6 100644 --- a/include/crypto/sha.h +++ b/include/crypto/sha.h @@ -76,4 +76,10 @@ struct sha256_state { u8 buf[SHA256_BLOCK_SIZE]; }; +struct sha512_state { + u64 state[8]; + u32 count[4]; + u8 buf[128]; +}; + #endif -- cgit v1.2.3 From 13887ed6888dad1608eb9530ebd83b6ba29db577 Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Wed, 22 Jul 2009 12:22:43 +0800 Subject: crypto: sha512_generic - Use 64-bit counters This patch replaces the 32-bit counters in sha512_generic with 64-bit counters. It also switches the bit count to the simpler byte count. Signed-off-by: Herbert Xu --- crypto/sha512_generic.c | 28 +++++++++++----------------- include/crypto/sha.h | 6 +++--- 2 files changed, 14 insertions(+), 20 deletions(-) (limited to 'include') diff --git a/crypto/sha512_generic.c b/crypto/sha512_generic.c index 4fe95eb03226..9ed9f60316e5 100644 --- a/crypto/sha512_generic.c +++ b/crypto/sha512_generic.c @@ -144,7 +144,7 @@ sha512_init(struct shash_desc *desc) sctx->state[5] = SHA512_H5; sctx->state[6] = SHA512_H6; sctx->state[7] = SHA512_H7; - sctx->count[0] = sctx->count[1] = sctx->count[2] = sctx->count[3] = 0; + sctx->count[0] = sctx->count[1] = 0; return 0; } @@ -161,7 +161,7 @@ sha384_init(struct shash_desc *desc) sctx->state[5] = SHA384_H5; sctx->state[6] = SHA384_H6; sctx->state[7] = SHA384_H7; - sctx->count[0] = sctx->count[1] = sctx->count[2] = sctx->count[3] = 0; + sctx->count[0] = sctx->count[1] = 0; return 0; } @@ -174,15 +174,11 @@ sha512_update(struct shash_desc *desc, const u8 *data, unsigned int len) unsigned int i, index, part_len; /* Compute number of bytes mod 128 */ - index = (unsigned int)((sctx->count[0] >> 3) & 0x7F); - - /* Update number of bits */ - if ((sctx->count[0] += (len << 3)) < (len << 3)) { - if ((sctx->count[1] += 1) < 1) - if ((sctx->count[2] += 1) < 1) - sctx->count[3]++; - sctx->count[1] += (len >> 29); - } + index = sctx->count[0] & 0x7f; + + /* Update number of bytes */ + if (!(sctx->count[0] += len)) + sctx->count[1]++; part_len = 128 - index; @@ -211,18 +207,16 @@ sha512_final(struct shash_desc *desc, u8 *hash) struct sha512_state *sctx = shash_desc_ctx(desc); static u8 padding[128] = { 0x80, }; __be64 *dst = (__be64 *)hash; - __be32 bits[4]; + __be64 bits[2]; unsigned int index, pad_len; int i; /* Save number of bits */ - bits[3] = cpu_to_be32(sctx->count[0]); - bits[2] = cpu_to_be32(sctx->count[1]); - bits[1] = cpu_to_be32(sctx->count[2]); - bits[0] = cpu_to_be32(sctx->count[3]); + bits[1] = cpu_to_be64(sctx->count[0] << 3); + bits[0] = cpu_to_be64(sctx->count[1] << 3 | sctx->count[0] >> 61); /* Pad out to 112 mod 128. */ - index = (sctx->count[0] >> 3) & 0x7f; + index = sctx->count[0] & 0x7f; pad_len = (index < 112) ? (112 - index) : ((128+112) - index); sha512_update(desc, padding, pad_len); diff --git a/include/crypto/sha.h b/include/crypto/sha.h index 45b25ccf7cc6..069e85ba97e1 100644 --- a/include/crypto/sha.h +++ b/include/crypto/sha.h @@ -77,9 +77,9 @@ struct sha256_state { }; struct sha512_state { - u64 state[8]; - u32 count[4]; - u8 buf[128]; + u64 count[2]; + u64 state[SHA512_DIGEST_SIZE / 8]; + u8 buf[SHA512_BLOCK_SIZE]; }; #endif -- cgit v1.2.3 From 9ba5f005c994ad28e266a0cd14ef29354be382c9 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Wed, 22 Jul 2009 14:18:35 +0200 Subject: softirq: introduce tasklet_hrtimer infrastructure commit ca109491f (hrtimer: removing all ur callback modes) moved all hrtimer callbacks into hard interrupt context when high resolution timers are active. That breaks code which relied on the assumption that the callback happens in softirq context. Provide a generic infrastructure which combines tasklets and hrtimers together to provide an in-softirq hrtimer experience. Signed-off-by: Peter Zijlstra Cc: torvalds@linux-foundation.org Cc: kaber@trash.net Cc: David Miller LKML-Reference: <1248265724.27058.1366.camel@twins> Signed-off-by: Thomas Gleixner --- include/linux/interrupt.h | 26 +++++++++++++++++++ kernel/softirq.c | 64 ++++++++++++++++++++++++++++++++++++++++++++++- 2 files changed, 89 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/interrupt.h b/include/linux/interrupt.h index 2721f07e9354..fd4c9c63c757 100644 --- a/include/linux/interrupt.h +++ b/include/linux/interrupt.h @@ -14,6 +14,7 @@ #include #include #include +#include #include #include @@ -517,6 +518,31 @@ extern void tasklet_kill_immediate(struct tasklet_struct *t, unsigned int cpu); extern void tasklet_init(struct tasklet_struct *t, void (*func)(unsigned long), unsigned long data); +struct tasklet_hrtimer { + struct hrtimer timer; + struct tasklet_struct tasklet; + enum hrtimer_restart (*function)(struct hrtimer *); +}; + +extern void +tasklet_hrtimer_init(struct tasklet_hrtimer *ttimer, + enum hrtimer_restart (*function)(struct hrtimer *), + clockid_t which_clock, enum hrtimer_mode mode); + +static inline +int tasklet_hrtimer_start(struct tasklet_hrtimer *ttimer, ktime_t time, + const enum hrtimer_mode mode) +{ + return hrtimer_start(&ttimer->timer, time, mode); +} + +static inline +void tasklet_hrtimer_cancel(struct tasklet_hrtimer *ttimer) +{ + hrtimer_cancel(&ttimer->timer); + tasklet_kill(&ttimer->tasklet); +} + /* * Autoprobing for irqs: * diff --git a/kernel/softirq.c b/kernel/softirq.c index 3a94905fa5d2..eb5e131a0485 100644 --- a/kernel/softirq.c +++ b/kernel/softirq.c @@ -345,7 +345,9 @@ void open_softirq(int nr, void (*action)(struct softirq_action *)) softirq_vec[nr].action = action; } -/* Tasklets */ +/* + * Tasklets + */ struct tasklet_head { struct tasklet_struct *head; @@ -493,6 +495,66 @@ void tasklet_kill(struct tasklet_struct *t) EXPORT_SYMBOL(tasklet_kill); +/* + * tasklet_hrtimer + */ + +/* + * The trampoline is called when the hrtimer expires. If this is + * called from the hrtimer interrupt then we schedule the tasklet as + * the timer callback function expects to run in softirq context. If + * it's called in softirq context anyway (i.e. high resolution timers + * disabled) then the hrtimer callback is called right away. + */ +static enum hrtimer_restart __hrtimer_tasklet_trampoline(struct hrtimer *timer) +{ + struct tasklet_hrtimer *ttimer = + container_of(timer, struct tasklet_hrtimer, timer); + + if (hrtimer_is_hres_active(timer)) { + tasklet_hi_schedule(&ttimer->tasklet); + return HRTIMER_NORESTART; + } + return ttimer->function(timer); +} + +/* + * Helper function which calls the hrtimer callback from + * tasklet/softirq context + */ +static void __tasklet_hrtimer_trampoline(unsigned long data) +{ + struct tasklet_hrtimer *ttimer = (void *)data; + enum hrtimer_restart restart; + + restart = ttimer->function(&ttimer->timer); + if (restart != HRTIMER_NORESTART) + hrtimer_restart(&ttimer->timer); +} + +/** + * tasklet_hrtimer_init - Init a tasklet/hrtimer combo for softirq callbacks + * @ttimer: tasklet_hrtimer which is initialized + * @function: hrtimer callback funtion which gets called from softirq context + * @which_clock: clock id (CLOCK_MONOTONIC/CLOCK_REALTIME) + * @mode: hrtimer mode (HRTIMER_MODE_ABS/HRTIMER_MODE_REL) + */ +void tasklet_hrtimer_init(struct tasklet_hrtimer *ttimer, + enum hrtimer_restart (*function)(struct hrtimer *), + clockid_t which_clock, enum hrtimer_mode mode) +{ + hrtimer_init(&ttimer->timer, which_clock, mode); + ttimer->timer.function = __hrtimer_tasklet_trampoline; + tasklet_init(&ttimer->tasklet, __tasklet_hrtimer_trampoline, + (unsigned long)ttimer); + ttimer->function = function; +} +EXPORT_SYMBOL_GPL(tasklet_hrtimer_init); + +/* + * Remote softirq bits + */ + DEFINE_PER_CPU(struct list_head [NR_SOFTIRQS], softirq_work_list); EXPORT_PER_CPU_SYMBOL(softirq_work_list); -- cgit v1.2.3 From fbd90375d7531927d312766b548376d909811b4d Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Wed, 22 Jul 2009 13:40:14 +0200 Subject: hrtimer: Remove cb_entry from struct hrtimer It's unused, remove it. Signed-off-by: Peter Zijlstra Signed-off-by: Thomas Gleixner LKML-Reference: --- include/linux/hrtimer.h | 2 -- kernel/hrtimer.c | 1 - 2 files changed, 3 deletions(-) (limited to 'include') diff --git a/include/linux/hrtimer.h b/include/linux/hrtimer.h index 54648e625efd..40e7d54fc424 100644 --- a/include/linux/hrtimer.h +++ b/include/linux/hrtimer.h @@ -91,7 +91,6 @@ enum hrtimer_restart { * @function: timer expiry callback function * @base: pointer to the timer base (per cpu and per clock) * @state: state information (See bit values above) - * @cb_entry: list head to enqueue an expired timer into the callback list * @start_site: timer statistics field to store the site where the timer * was started * @start_comm: timer statistics field to store the name of the process which @@ -108,7 +107,6 @@ struct hrtimer { enum hrtimer_restart (*function)(struct hrtimer *); struct hrtimer_clock_base *base; unsigned long state; - struct list_head cb_entry; #ifdef CONFIG_TIMER_STATS int start_pid; void *start_site; diff --git a/kernel/hrtimer.c b/kernel/hrtimer.c index 43d151f185b6..052a0f53e4eb 100644 --- a/kernel/hrtimer.c +++ b/kernel/hrtimer.c @@ -1092,7 +1092,6 @@ static void __hrtimer_init(struct hrtimer *timer, clockid_t clock_id, clock_id = CLOCK_MONOTONIC; timer->base = &cpu_base->clock_base[clock_id]; - INIT_LIST_HEAD(&timer->cb_entry); hrtimer_init_timer_hres(timer); #ifdef CONFIG_TIMER_STATS -- cgit v1.2.3 From 7f453c24b95a085fc7bd35d53b33abc4dc5a048b Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Tue, 21 Jul 2009 13:19:40 +0200 Subject: perf_counter: PERF_SAMPLE_ID and inherited counters Anton noted that for inherited counters the counter-id as provided by PERF_SAMPLE_ID isn't mappable to the id found through PERF_RECORD_ID because each inherited counter gets its own id. His suggestion was to always return the parent counter id, since that is the primary counter id as exposed. However, these inherited counters have a unique identifier so that events like PERF_EVENT_PERIOD and PERF_EVENT_THROTTLE can be specific about which counter gets modified, which is important when trying to normalize the sample streams. This patch removes PERF_EVENT_PERIOD in favour of PERF_SAMPLE_PERIOD, which is more useful anyway, since changing periods became a lot more common than initially thought -- rendering PERF_EVENT_PERIOD the less useful solution (also, PERF_SAMPLE_PERIOD reports the more accurate value, since it reports the value used to trigger the overflow, whereas PERF_EVENT_PERIOD simply reports the requested period changed, which might only take effect on the next cycle). This still leaves us PERF_EVENT_THROTTLE to consider, but since that _should_ be a rare occurrence, and linking it to a primary id is the most useful bit to diagnose the problem, we introduce a PERF_SAMPLE_STREAM_ID, for those few cases where the full reconstruction is important. [Does change the ABI a little, but I see no other way out] Suggested-by: Anton Blanchard Signed-off-by: Peter Zijlstra LKML-Reference: <1248095846.15751.8781.camel@twins> --- include/linux/perf_counter.h | 15 ++----- kernel/perf_counter.c | 92 +++++++++++++++---------------------------- tools/perf/builtin-annotate.c | 24 ----------- tools/perf/builtin-report.c | 24 ----------- 4 files changed, 35 insertions(+), 120 deletions(-) (limited to 'include') diff --git a/include/linux/perf_counter.h b/include/linux/perf_counter.h index 5e970c7d3fd5..bd15d7a5f5ce 100644 --- a/include/linux/perf_counter.h +++ b/include/linux/perf_counter.h @@ -120,8 +120,9 @@ enum perf_counter_sample_format { PERF_SAMPLE_ID = 1U << 6, PERF_SAMPLE_CPU = 1U << 7, PERF_SAMPLE_PERIOD = 1U << 8, + PERF_SAMPLE_STREAM_ID = 1U << 9, - PERF_SAMPLE_MAX = 1U << 9, /* non-ABI */ + PERF_SAMPLE_MAX = 1U << 10, /* non-ABI */ }; /* @@ -312,16 +313,7 @@ enum perf_event_type { * struct perf_event_header header; * u64 time; * u64 id; - * u64 sample_period; - * }; - */ - PERF_EVENT_PERIOD = 4, - - /* - * struct { - * struct perf_event_header header; - * u64 time; - * u64 id; + * u64 stream_id; * }; */ PERF_EVENT_THROTTLE = 5, @@ -356,6 +348,7 @@ enum perf_event_type { * { u64 time; } && PERF_SAMPLE_TIME * { u64 addr; } && PERF_SAMPLE_ADDR * { u64 id; } && PERF_SAMPLE_ID + * { u64 stream_id;} && PERF_SAMPLE_STREAM_ID * { u32 cpu, res; } && PERF_SAMPLE_CPU * { u64 period; } && PERF_SAMPLE_PERIOD * diff --git a/kernel/perf_counter.c b/kernel/perf_counter.c index e1d6a3aa1333..7530588fa5c5 100644 --- a/kernel/perf_counter.c +++ b/kernel/perf_counter.c @@ -154,6 +154,20 @@ static void unclone_ctx(struct perf_counter_context *ctx) } } +/* + * If we inherit counters we want to return the parent counter id + * to userspace. + */ +static u64 primary_counter_id(struct perf_counter *counter) +{ + u64 id = counter->id; + + if (counter->parent) + id = counter->parent->id; + + return id; +} + /* * Get the perf_counter_context for a task and lock it. * This has to cope with with the fact that until it is locked, @@ -1296,7 +1310,6 @@ static void perf_counter_cpu_sched_in(struct perf_cpu_context *cpuctx, int cpu) #define MAX_INTERRUPTS (~0ULL) static void perf_log_throttle(struct perf_counter *counter, int enable); -static void perf_log_period(struct perf_counter *counter, u64 period); static void perf_adjust_period(struct perf_counter *counter, u64 events) { @@ -1315,8 +1328,6 @@ static void perf_adjust_period(struct perf_counter *counter, u64 events) if (!sample_period) sample_period = 1; - perf_log_period(counter, sample_period); - hwc->sample_period = sample_period; } @@ -1705,7 +1716,7 @@ perf_read_hw(struct perf_counter *counter, char __user *buf, size_t count) values[n++] = counter->total_time_running + atomic64_read(&counter->child_total_time_running); if (counter->attr.read_format & PERF_FORMAT_ID) - values[n++] = counter->id; + values[n++] = primary_counter_id(counter); mutex_unlock(&counter->child_mutex); if (count < n * sizeof(u64)) @@ -1812,8 +1823,6 @@ static int perf_counter_period(struct perf_counter *counter, u64 __user *arg) counter->attr.sample_freq = value; } else { - perf_log_period(counter, value); - counter->attr.sample_period = value; counter->hw.sample_period = value; } @@ -2662,6 +2671,9 @@ static void perf_counter_output(struct perf_counter *counter, int nmi, if (sample_type & PERF_SAMPLE_ID) header.size += sizeof(u64); + if (sample_type & PERF_SAMPLE_STREAM_ID) + header.size += sizeof(u64); + if (sample_type & PERF_SAMPLE_CPU) { header.size += sizeof(cpu_entry); @@ -2705,7 +2717,13 @@ static void perf_counter_output(struct perf_counter *counter, int nmi, if (sample_type & PERF_SAMPLE_ADDR) perf_output_put(&handle, data->addr); - if (sample_type & PERF_SAMPLE_ID) + if (sample_type & PERF_SAMPLE_ID) { + u64 id = primary_counter_id(counter); + + perf_output_put(&handle, id); + } + + if (sample_type & PERF_SAMPLE_STREAM_ID) perf_output_put(&handle, counter->id); if (sample_type & PERF_SAMPLE_CPU) @@ -2728,7 +2746,7 @@ static void perf_counter_output(struct perf_counter *counter, int nmi, if (sub != counter) sub->pmu->read(sub); - group_entry.id = sub->id; + group_entry.id = primary_counter_id(sub); group_entry.counter = atomic64_read(&sub->count); perf_output_put(&handle, group_entry); @@ -2788,15 +2806,8 @@ perf_counter_read_event(struct perf_counter *counter, } if (counter->attr.read_format & PERF_FORMAT_ID) { - u64 id; - event.header.size += sizeof(u64); - if (counter->parent) - id = counter->parent->id; - else - id = counter->id; - - event.format[i++] = id; + event.format[i++] = primary_counter_id(counter); } ret = perf_output_begin(&handle, counter, event.header.size, 0, 0); @@ -3190,49 +3201,6 @@ void __perf_counter_mmap(struct vm_area_struct *vma) perf_counter_mmap_event(&mmap_event); } -/* - * Log sample_period changes so that analyzing tools can re-normalize the - * event flow. - */ - -struct freq_event { - struct perf_event_header header; - u64 time; - u64 id; - u64 period; -}; - -static void perf_log_period(struct perf_counter *counter, u64 period) -{ - struct perf_output_handle handle; - struct freq_event event; - int ret; - - if (counter->hw.sample_period == period) - return; - - if (counter->attr.sample_type & PERF_SAMPLE_PERIOD) - return; - - event = (struct freq_event) { - .header = { - .type = PERF_EVENT_PERIOD, - .misc = 0, - .size = sizeof(event), - }, - .time = sched_clock(), - .id = counter->id, - .period = period, - }; - - ret = perf_output_begin(&handle, counter, sizeof(event), 1, 0); - if (ret) - return; - - perf_output_put(&handle, event); - perf_output_end(&handle); -} - /* * IRQ throttle logging */ @@ -3246,14 +3214,16 @@ static void perf_log_throttle(struct perf_counter *counter, int enable) struct perf_event_header header; u64 time; u64 id; + u64 stream_id; } throttle_event = { .header = { .type = PERF_EVENT_THROTTLE + 1, .misc = 0, .size = sizeof(throttle_event), }, - .time = sched_clock(), - .id = counter->id, + .time = sched_clock(), + .id = primary_counter_id(counter), + .stream_id = counter->id, }; ret = perf_output_begin(&handle, counter, sizeof(throttle_event), 1, 0); diff --git a/tools/perf/builtin-annotate.c b/tools/perf/builtin-annotate.c index 5f9eefecc574..1dba568e1941 100644 --- a/tools/perf/builtin-annotate.c +++ b/tools/perf/builtin-annotate.c @@ -74,20 +74,12 @@ struct fork_event { u32 pid, ppid; }; -struct period_event { - struct perf_event_header header; - u64 time; - u64 id; - u64 sample_period; -}; - typedef union event_union { struct perf_event_header header; struct ip_event ip; struct mmap_event mmap; struct comm_event comm; struct fork_event fork; - struct period_event period; } event_t; @@ -997,19 +989,6 @@ process_fork_event(event_t *event, unsigned long offset, unsigned long head) return 0; } -static int -process_period_event(event_t *event, unsigned long offset, unsigned long head) -{ - dprintf("%p [%p]: PERF_EVENT_PERIOD: time:%Ld, id:%Ld: period:%Ld\n", - (void *)(offset + head), - (void *)(long)(event->header.size), - event->period.time, - event->period.id, - event->period.sample_period); - - return 0; -} - static int process_event(event_t *event, unsigned long offset, unsigned long head) { @@ -1025,9 +1004,6 @@ process_event(event_t *event, unsigned long offset, unsigned long head) case PERF_EVENT_FORK: return process_fork_event(event, offset, head); - - case PERF_EVENT_PERIOD: - return process_period_event(event, offset, head); /* * We dont process them right now but they are fine: */ diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c index a118bc77286d..b20a4b6e31b7 100644 --- a/tools/perf/builtin-report.c +++ b/tools/perf/builtin-report.c @@ -101,13 +101,6 @@ struct fork_event { u32 pid, ppid; }; -struct period_event { - struct perf_event_header header; - u64 time; - u64 id; - u64 sample_period; -}; - struct lost_event { struct perf_event_header header; u64 id; @@ -127,7 +120,6 @@ typedef union event_union { struct mmap_event mmap; struct comm_event comm; struct fork_event fork; - struct period_event period; struct lost_event lost; struct read_event read; } event_t; @@ -1635,19 +1627,6 @@ process_fork_event(event_t *event, unsigned long offset, unsigned long head) return 0; } -static int -process_period_event(event_t *event, unsigned long offset, unsigned long head) -{ - dprintf("%p [%p]: PERF_EVENT_PERIOD: time:%Ld, id:%Ld: period:%Ld\n", - (void *)(offset + head), - (void *)(long)(event->header.size), - event->period.time, - event->period.id, - event->period.sample_period); - - return 0; -} - static int process_lost_event(event_t *event, unsigned long offset, unsigned long head) { @@ -1729,9 +1708,6 @@ process_event(event_t *event, unsigned long offset, unsigned long head) case PERF_EVENT_FORK: return process_fork_event(event, offset, head); - case PERF_EVENT_PERIOD: - return process_period_event(event, offset, head); - case PERF_EVENT_LOST: return process_lost_event(event, offset, head); -- cgit v1.2.3 From 24c30dbbcdda9aeccb23b4eecb6bb8e538742ea4 Mon Sep 17 00:00:00 2001 From: Anton Vorontsov Date: Thu, 16 Jul 2009 21:31:31 +0000 Subject: of/mdio: Add support function for Ethernet fixed-link property Fixed-link support is broken for the ucc_eth, gianfar, and fs_enet device drivers. The "OF MDIO rework" patches removed most of the support. Instead of re-adding fixed-link stuff to the drivers, this patch adds a support function for parsing the fixed-link property and obtaining a dummy phy to match. Note: the dummy phy handling in arch/powerpc is a bit of a hack and needs to be reworked. This function is being added now to solve the regression in the Ethernet drivers, but it should be considered a temporary measure until the fixed link handling can be reworked. Signed-off-by: Anton Vorontsov Signed-off-by: Grant Likely Signed-off-by: David S. Miller --- drivers/of/of_mdio.c | 42 ++++++++++++++++++++++++++++++++++++++++++ include/linux/of_mdio.h | 3 +++ 2 files changed, 45 insertions(+) (limited to 'include') diff --git a/drivers/of/of_mdio.c b/drivers/of/of_mdio.c index aee967d7f760..bacaa536fd51 100644 --- a/drivers/of/of_mdio.c +++ b/drivers/of/of_mdio.c @@ -9,6 +9,10 @@ * out of the OpenFirmware device tree and using it to populate an mii_bus. */ +#include +#include +#include +#include #include #include #include @@ -137,3 +141,41 @@ struct phy_device *of_phy_connect(struct net_device *dev, return phy_connect_direct(dev, phy, hndlr, flags, iface) ? NULL : phy; } EXPORT_SYMBOL(of_phy_connect); + +/** + * of_phy_connect_fixed_link - Parse fixed-link property and return a dummy phy + * @dev: pointer to net_device claiming the phy + * @hndlr: Link state callback for the network device + * @iface: PHY data interface type + * + * This function is a temporary stop-gap and will be removed soon. It is + * only to support the fs_enet, ucc_geth and gianfar Ethernet drivers. Do + * not call this function from new drivers. + */ +struct phy_device *of_phy_connect_fixed_link(struct net_device *dev, + void (*hndlr)(struct net_device *), + phy_interface_t iface) +{ + struct device_node *net_np; + char bus_id[MII_BUS_ID_SIZE + 3]; + struct phy_device *phy; + const u32 *phy_id; + int sz; + + if (!dev->dev.parent) + return NULL; + + net_np = dev_archdata_get_node(&dev->dev.parent->archdata); + if (!net_np) + return NULL; + + phy_id = of_get_property(net_np, "fixed-link", &sz); + if (!phy_id || sz < sizeof(*phy_id)) + return NULL; + + sprintf(bus_id, PHY_ID_FMT, "0", phy_id[0]); + + phy = phy_connect(dev, bus_id, hndlr, 0, iface); + return IS_ERR(phy) ? NULL : phy; +} +EXPORT_SYMBOL(of_phy_connect_fixed_link); diff --git a/include/linux/of_mdio.h b/include/linux/of_mdio.h index c9663c690303..53b94e025c7c 100644 --- a/include/linux/of_mdio.h +++ b/include/linux/of_mdio.h @@ -18,5 +18,8 @@ extern struct phy_device *of_phy_connect(struct net_device *dev, struct device_node *phy_np, void (*hndlr)(struct net_device *), u32 flags, phy_interface_t iface); +extern struct phy_device *of_phy_connect_fixed_link(struct net_device *dev, + void (*hndlr)(struct net_device *), + phy_interface_t iface); #endif /* __LINUX_OF_MDIO_H */ -- cgit v1.2.3 From cf4f1e76c49dacfde0680b170b9a9b6a42f296bb Mon Sep 17 00:00:00 2001 From: Magnus Damm Date: Wed, 22 Jul 2009 14:32:03 +0000 Subject: usb: move r8a66597 register defines Move r8a66597 hardware register definitions from the host controller header file to the platform data header file. With this change in place we can easily share register definitions between the host controller driver and a future gadget driver. Signed-off-by: Magnus Damm Signed-off-by: Paul Mundt --- drivers/usb/host/r8a66597.h | 366 ------------------------------------------ include/linux/usb/r8a66597.h | 372 ++++++++++++++++++++++++++++++++++++++++++- 2 files changed, 370 insertions(+), 368 deletions(-) (limited to 'include') diff --git a/drivers/usb/host/r8a66597.h b/drivers/usb/host/r8a66597.h index eecbd917bc81..228e3fb23854 100644 --- a/drivers/usb/host/r8a66597.h +++ b/drivers/usb/host/r8a66597.h @@ -32,372 +32,6 @@ #include -#define SYSCFG0 0x00 -#define SYSCFG1 0x02 -#define SYSSTS0 0x04 -#define SYSSTS1 0x06 -#define DVSTCTR0 0x08 -#define DVSTCTR1 0x0A -#define TESTMODE 0x0C -#define PINCFG 0x0E -#define DMA0CFG 0x10 -#define DMA1CFG 0x12 -#define CFIFO 0x14 -#define D0FIFO 0x18 -#define D1FIFO 0x1C -#define CFIFOSEL 0x20 -#define CFIFOCTR 0x22 -#define CFIFOSIE 0x24 -#define D0FIFOSEL 0x28 -#define D0FIFOCTR 0x2A -#define D1FIFOSEL 0x2C -#define D1FIFOCTR 0x2E -#define INTENB0 0x30 -#define INTENB1 0x32 -#define INTENB2 0x34 -#define BRDYENB 0x36 -#define NRDYENB 0x38 -#define BEMPENB 0x3A -#define SOFCFG 0x3C -#define INTSTS0 0x40 -#define INTSTS1 0x42 -#define INTSTS2 0x44 -#define BRDYSTS 0x46 -#define NRDYSTS 0x48 -#define BEMPSTS 0x4A -#define FRMNUM 0x4C -#define UFRMNUM 0x4E -#define USBADDR 0x50 -#define USBREQ 0x54 -#define USBVAL 0x56 -#define USBINDX 0x58 -#define USBLENG 0x5A -#define DCPCFG 0x5C -#define DCPMAXP 0x5E -#define DCPCTR 0x60 -#define PIPESEL 0x64 -#define PIPECFG 0x68 -#define PIPEBUF 0x6A -#define PIPEMAXP 0x6C -#define PIPEPERI 0x6E -#define PIPE1CTR 0x70 -#define PIPE2CTR 0x72 -#define PIPE3CTR 0x74 -#define PIPE4CTR 0x76 -#define PIPE5CTR 0x78 -#define PIPE6CTR 0x7A -#define PIPE7CTR 0x7C -#define PIPE8CTR 0x7E -#define PIPE9CTR 0x80 -#define PIPE1TRE 0x90 -#define PIPE1TRN 0x92 -#define PIPE2TRE 0x94 -#define PIPE2TRN 0x96 -#define PIPE3TRE 0x98 -#define PIPE3TRN 0x9A -#define PIPE4TRE 0x9C -#define PIPE4TRN 0x9E -#define PIPE5TRE 0xA0 -#define PIPE5TRN 0xA2 -#define DEVADD0 0xD0 -#define DEVADD1 0xD2 -#define DEVADD2 0xD4 -#define DEVADD3 0xD6 -#define DEVADD4 0xD8 -#define DEVADD5 0xDA -#define DEVADD6 0xDC -#define DEVADD7 0xDE -#define DEVADD8 0xE0 -#define DEVADD9 0xE2 -#define DEVADDA 0xE4 - -/* System Configuration Control Register */ -#define XTAL 0xC000 /* b15-14: Crystal selection */ -#define XTAL48 0x8000 /* 48MHz */ -#define XTAL24 0x4000 /* 24MHz */ -#define XTAL12 0x0000 /* 12MHz */ -#define XCKE 0x2000 /* b13: External clock enable */ -#define PLLC 0x0800 /* b11: PLL control */ -#define SCKE 0x0400 /* b10: USB clock enable */ -#define PCSDIS 0x0200 /* b9: not CS wakeup */ -#define LPSME 0x0100 /* b8: Low power sleep mode */ -#define HSE 0x0080 /* b7: Hi-speed enable */ -#define DCFM 0x0040 /* b6: Controller function select */ -#define DRPD 0x0020 /* b5: D+/- pull down control */ -#define DPRPU 0x0010 /* b4: D+ pull up control */ -#define USBE 0x0001 /* b0: USB module operation enable */ - -/* System Configuration Status Register */ -#define OVCBIT 0x8000 /* b15-14: Over-current bit */ -#define OVCMON 0xC000 /* b15-14: Over-current monitor */ -#define SOFEA 0x0020 /* b5: SOF monitor */ -#define IDMON 0x0004 /* b3: ID-pin monitor */ -#define LNST 0x0003 /* b1-0: D+, D- line status */ -#define SE1 0x0003 /* SE1 */ -#define FS_KSTS 0x0002 /* Full-Speed K State */ -#define FS_JSTS 0x0001 /* Full-Speed J State */ -#define LS_JSTS 0x0002 /* Low-Speed J State */ -#define LS_KSTS 0x0001 /* Low-Speed K State */ -#define SE0 0x0000 /* SE0 */ - -/* Device State Control Register */ -#define EXTLP0 0x0400 /* b10: External port */ -#define VBOUT 0x0200 /* b9: VBUS output */ -#define WKUP 0x0100 /* b8: Remote wakeup */ -#define RWUPE 0x0080 /* b7: Remote wakeup sense */ -#define USBRST 0x0040 /* b6: USB reset enable */ -#define RESUME 0x0020 /* b5: Resume enable */ -#define UACT 0x0010 /* b4: USB bus enable */ -#define RHST 0x0007 /* b1-0: Reset handshake status */ -#define HSPROC 0x0004 /* HS handshake is processing */ -#define HSMODE 0x0003 /* Hi-Speed mode */ -#define FSMODE 0x0002 /* Full-Speed mode */ -#define LSMODE 0x0001 /* Low-Speed mode */ -#define UNDECID 0x0000 /* Undecided */ - -/* Test Mode Register */ -#define UTST 0x000F /* b3-0: Test select */ -#define H_TST_PACKET 0x000C /* HOST TEST Packet */ -#define H_TST_SE0_NAK 0x000B /* HOST TEST SE0 NAK */ -#define H_TST_K 0x000A /* HOST TEST K */ -#define H_TST_J 0x0009 /* HOST TEST J */ -#define H_TST_NORMAL 0x0000 /* HOST Normal Mode */ -#define P_TST_PACKET 0x0004 /* PERI TEST Packet */ -#define P_TST_SE0_NAK 0x0003 /* PERI TEST SE0 NAK */ -#define P_TST_K 0x0002 /* PERI TEST K */ -#define P_TST_J 0x0001 /* PERI TEST J */ -#define P_TST_NORMAL 0x0000 /* PERI Normal Mode */ - -/* Data Pin Configuration Register */ -#define LDRV 0x8000 /* b15: Drive Current Adjust */ -#define VIF1 0x0000 /* VIF = 1.8V */ -#define VIF3 0x8000 /* VIF = 3.3V */ -#define INTA 0x0001 /* b1: USB INT-pin active */ - -/* DMAx Pin Configuration Register */ -#define DREQA 0x4000 /* b14: Dreq active select */ -#define BURST 0x2000 /* b13: Burst mode */ -#define DACKA 0x0400 /* b10: Dack active select */ -#define DFORM 0x0380 /* b9-7: DMA mode select */ -#define CPU_ADR_RD_WR 0x0000 /* Address + RD/WR mode (CPU bus) */ -#define CPU_DACK_RD_WR 0x0100 /* DACK + RD/WR mode (CPU bus) */ -#define CPU_DACK_ONLY 0x0180 /* DACK only mode (CPU bus) */ -#define SPLIT_DACK_ONLY 0x0200 /* DACK only mode (SPLIT bus) */ -#define DENDA 0x0040 /* b6: Dend active select */ -#define PKTM 0x0020 /* b5: Packet mode */ -#define DENDE 0x0010 /* b4: Dend enable */ -#define OBUS 0x0004 /* b2: OUTbus mode */ - -/* CFIFO/DxFIFO Port Select Register */ -#define RCNT 0x8000 /* b15: Read count mode */ -#define REW 0x4000 /* b14: Buffer rewind */ -#define DCLRM 0x2000 /* b13: DMA buffer clear mode */ -#define DREQE 0x1000 /* b12: DREQ output enable */ -#define MBW_8 0x0000 /* 8bit */ -#define MBW_16 0x0400 /* 16bit */ -#define MBW_32 0x0800 /* 32bit */ -#define BIGEND 0x0100 /* b8: Big endian mode */ -#define BYTE_LITTLE 0x0000 /* little dendian */ -#define BYTE_BIG 0x0100 /* big endifan */ -#define ISEL 0x0020 /* b5: DCP FIFO port direction select */ -#define CURPIPE 0x000F /* b2-0: PIPE select */ - -/* CFIFO/DxFIFO Port Control Register */ -#define BVAL 0x8000 /* b15: Buffer valid flag */ -#define BCLR 0x4000 /* b14: Buffer clear */ -#define FRDY 0x2000 /* b13: FIFO ready */ -#define DTLN 0x0FFF /* b11-0: FIFO received data length */ - -/* Interrupt Enable Register 0 */ -#define VBSE 0x8000 /* b15: VBUS interrupt */ -#define RSME 0x4000 /* b14: Resume interrupt */ -#define SOFE 0x2000 /* b13: Frame update interrupt */ -#define DVSE 0x1000 /* b12: Device state transition interrupt */ -#define CTRE 0x0800 /* b11: Control transfer stage transition interrupt */ -#define BEMPE 0x0400 /* b10: Buffer empty interrupt */ -#define NRDYE 0x0200 /* b9: Buffer not ready interrupt */ -#define BRDYE 0x0100 /* b8: Buffer ready interrupt */ - -/* Interrupt Enable Register 1 */ -#define OVRCRE 0x8000 /* b15: Over-current interrupt */ -#define BCHGE 0x4000 /* b14: USB us chenge interrupt */ -#define DTCHE 0x1000 /* b12: Detach sense interrupt */ -#define ATTCHE 0x0800 /* b11: Attach sense interrupt */ -#define EOFERRE 0x0040 /* b6: EOF error interrupt */ -#define SIGNE 0x0020 /* b5: SETUP IGNORE interrupt */ -#define SACKE 0x0010 /* b4: SETUP ACK interrupt */ - -/* BRDY Interrupt Enable/Status Register */ -#define BRDY9 0x0200 /* b9: PIPE9 */ -#define BRDY8 0x0100 /* b8: PIPE8 */ -#define BRDY7 0x0080 /* b7: PIPE7 */ -#define BRDY6 0x0040 /* b6: PIPE6 */ -#define BRDY5 0x0020 /* b5: PIPE5 */ -#define BRDY4 0x0010 /* b4: PIPE4 */ -#define BRDY3 0x0008 /* b3: PIPE3 */ -#define BRDY2 0x0004 /* b2: PIPE2 */ -#define BRDY1 0x0002 /* b1: PIPE1 */ -#define BRDY0 0x0001 /* b1: PIPE0 */ - -/* NRDY Interrupt Enable/Status Register */ -#define NRDY9 0x0200 /* b9: PIPE9 */ -#define NRDY8 0x0100 /* b8: PIPE8 */ -#define NRDY7 0x0080 /* b7: PIPE7 */ -#define NRDY6 0x0040 /* b6: PIPE6 */ -#define NRDY5 0x0020 /* b5: PIPE5 */ -#define NRDY4 0x0010 /* b4: PIPE4 */ -#define NRDY3 0x0008 /* b3: PIPE3 */ -#define NRDY2 0x0004 /* b2: PIPE2 */ -#define NRDY1 0x0002 /* b1: PIPE1 */ -#define NRDY0 0x0001 /* b1: PIPE0 */ - -/* BEMP Interrupt Enable/Status Register */ -#define BEMP9 0x0200 /* b9: PIPE9 */ -#define BEMP8 0x0100 /* b8: PIPE8 */ -#define BEMP7 0x0080 /* b7: PIPE7 */ -#define BEMP6 0x0040 /* b6: PIPE6 */ -#define BEMP5 0x0020 /* b5: PIPE5 */ -#define BEMP4 0x0010 /* b4: PIPE4 */ -#define BEMP3 0x0008 /* b3: PIPE3 */ -#define BEMP2 0x0004 /* b2: PIPE2 */ -#define BEMP1 0x0002 /* b1: PIPE1 */ -#define BEMP0 0x0001 /* b0: PIPE0 */ - -/* SOF Pin Configuration Register */ -#define TRNENSEL 0x0100 /* b8: Select transaction enable period */ -#define BRDYM 0x0040 /* b6: BRDY clear timing */ -#define INTL 0x0020 /* b5: Interrupt sense select */ -#define EDGESTS 0x0010 /* b4: */ -#define SOFMODE 0x000C /* b3-2: SOF pin select */ -#define SOF_125US 0x0008 /* SOF OUT 125us Frame Signal */ -#define SOF_1MS 0x0004 /* SOF OUT 1ms Frame Signal */ -#define SOF_DISABLE 0x0000 /* SOF OUT Disable */ - -/* Interrupt Status Register 0 */ -#define VBINT 0x8000 /* b15: VBUS interrupt */ -#define RESM 0x4000 /* b14: Resume interrupt */ -#define SOFR 0x2000 /* b13: SOF frame update interrupt */ -#define DVST 0x1000 /* b12: Device state transition interrupt */ -#define CTRT 0x0800 /* b11: Control transfer stage transition interrupt */ -#define BEMP 0x0400 /* b10: Buffer empty interrupt */ -#define NRDY 0x0200 /* b9: Buffer not ready interrupt */ -#define BRDY 0x0100 /* b8: Buffer ready interrupt */ -#define VBSTS 0x0080 /* b7: VBUS input port */ -#define DVSQ 0x0070 /* b6-4: Device state */ -#define DS_SPD_CNFG 0x0070 /* Suspend Configured */ -#define DS_SPD_ADDR 0x0060 /* Suspend Address */ -#define DS_SPD_DFLT 0x0050 /* Suspend Default */ -#define DS_SPD_POWR 0x0040 /* Suspend Powered */ -#define DS_SUSP 0x0040 /* Suspend */ -#define DS_CNFG 0x0030 /* Configured */ -#define DS_ADDS 0x0020 /* Address */ -#define DS_DFLT 0x0010 /* Default */ -#define DS_POWR 0x0000 /* Powered */ -#define DVSQS 0x0030 /* b5-4: Device state */ -#define VALID 0x0008 /* b3: Setup packet detected flag */ -#define CTSQ 0x0007 /* b2-0: Control transfer stage */ -#define CS_SQER 0x0006 /* Sequence error */ -#define CS_WRND 0x0005 /* Control write nodata status stage */ -#define CS_WRSS 0x0004 /* Control write status stage */ -#define CS_WRDS 0x0003 /* Control write data stage */ -#define CS_RDSS 0x0002 /* Control read status stage */ -#define CS_RDDS 0x0001 /* Control read data stage */ -#define CS_IDST 0x0000 /* Idle or setup stage */ - -/* Interrupt Status Register 1 */ -#define OVRCR 0x8000 /* b15: Over-current interrupt */ -#define BCHG 0x4000 /* b14: USB bus chenge interrupt */ -#define DTCH 0x1000 /* b12: Detach sense interrupt */ -#define ATTCH 0x0800 /* b11: Attach sense interrupt */ -#define EOFERR 0x0040 /* b6: EOF-error interrupt */ -#define SIGN 0x0020 /* b5: Setup ignore interrupt */ -#define SACK 0x0010 /* b4: Setup acknowledge interrupt */ - -/* Frame Number Register */ -#define OVRN 0x8000 /* b15: Overrun error */ -#define CRCE 0x4000 /* b14: Received data error */ -#define FRNM 0x07FF /* b10-0: Frame number */ - -/* Micro Frame Number Register */ -#define UFRNM 0x0007 /* b2-0: Micro frame number */ - -/* Default Control Pipe Maxpacket Size Register */ -/* Pipe Maxpacket Size Register */ -#define DEVSEL 0xF000 /* b15-14: Device address select */ -#define MAXP 0x007F /* b6-0: Maxpacket size of default control pipe */ - -/* Default Control Pipe Control Register */ -#define BSTS 0x8000 /* b15: Buffer status */ -#define SUREQ 0x4000 /* b14: Send USB request */ -#define CSCLR 0x2000 /* b13: complete-split status clear */ -#define CSSTS 0x1000 /* b12: complete-split status */ -#define SUREQCLR 0x0800 /* b11: stop setup request */ -#define SQCLR 0x0100 /* b8: Sequence toggle bit clear */ -#define SQSET 0x0080 /* b7: Sequence toggle bit set */ -#define SQMON 0x0040 /* b6: Sequence toggle bit monitor */ -#define PBUSY 0x0020 /* b5: pipe busy */ -#define PINGE 0x0010 /* b4: ping enable */ -#define CCPL 0x0004 /* b2: Enable control transfer complete */ -#define PID 0x0003 /* b1-0: Response PID */ -#define PID_STALL11 0x0003 /* STALL */ -#define PID_STALL 0x0002 /* STALL */ -#define PID_BUF 0x0001 /* BUF */ -#define PID_NAK 0x0000 /* NAK */ - -/* Pipe Window Select Register */ -#define PIPENM 0x0007 /* b2-0: Pipe select */ - -/* Pipe Configuration Register */ -#define R8A66597_TYP 0xC000 /* b15-14: Transfer type */ -#define R8A66597_ISO 0xC000 /* Isochronous */ -#define R8A66597_INT 0x8000 /* Interrupt */ -#define R8A66597_BULK 0x4000 /* Bulk */ -#define R8A66597_BFRE 0x0400 /* b10: Buffer ready interrupt mode select */ -#define R8A66597_DBLB 0x0200 /* b9: Double buffer mode select */ -#define R8A66597_CNTMD 0x0100 /* b8: Continuous transfer mode select */ -#define R8A66597_SHTNAK 0x0080 /* b7: Transfer end NAK */ -#define R8A66597_DIR 0x0010 /* b4: Transfer direction select */ -#define R8A66597_EPNUM 0x000F /* b3-0: Eendpoint number select */ - -/* Pipe Buffer Configuration Register */ -#define BUFSIZE 0x7C00 /* b14-10: Pipe buffer size */ -#define BUFNMB 0x007F /* b6-0: Pipe buffer number */ -#define PIPE0BUF 256 -#define PIPExBUF 64 - -/* Pipe Maxpacket Size Register */ -#define MXPS 0x07FF /* b10-0: Maxpacket size */ - -/* Pipe Cycle Configuration Register */ -#define IFIS 0x1000 /* b12: Isochronous in-buffer flush mode select */ -#define IITV 0x0007 /* b2-0: Isochronous interval */ - -/* Pipex Control Register */ -#define BSTS 0x8000 /* b15: Buffer status */ -#define INBUFM 0x4000 /* b14: IN buffer monitor (Only for PIPE1 to 5) */ -#define CSCLR 0x2000 /* b13: complete-split status clear */ -#define CSSTS 0x1000 /* b12: complete-split status */ -#define ATREPM 0x0400 /* b10: Auto repeat mode */ -#define ACLRM 0x0200 /* b9: Out buffer auto clear mode */ -#define SQCLR 0x0100 /* b8: Sequence toggle bit clear */ -#define SQSET 0x0080 /* b7: Sequence toggle bit set */ -#define SQMON 0x0040 /* b6: Sequence toggle bit monitor */ -#define PBUSY 0x0020 /* b5: pipe busy */ -#define PID 0x0003 /* b1-0: Response PID */ - -/* PIPExTRE */ -#define TRENB 0x0200 /* b9: Transaction counter enable */ -#define TRCLR 0x0100 /* b8: Transaction counter clear */ - -/* PIPExTRN */ -#define TRNCNT 0xFFFF /* b15-0: Transaction counter */ - -/* DEVADDx */ -#define UPPHUB 0x7800 -#define HUBPORT 0x0700 -#define USBSPD 0x00C0 -#define RTPORT 0x0001 - #define R8A66597_MAX_NUM_PIPE 10 #define R8A66597_BUF_BSIZE 8 #define R8A66597_MAX_DEVICE 10 diff --git a/include/linux/usb/r8a66597.h b/include/linux/usb/r8a66597.h index 460ee3f6a2c6..26d216734057 100644 --- a/include/linux/usb/r8a66597.h +++ b/include/linux/usb/r8a66597.h @@ -28,7 +28,7 @@ #define R8A66597_PLATDATA_XTAL_48MHZ 0x03 struct r8a66597_platdata { - /* This ops can controll port power instead of DVSTCTR register. */ + /* This callback can control port power instead of DVSTCTR register. */ void (*port_power)(int port, int power); /* set one = on chip controller, set zero = external controller */ @@ -43,5 +43,373 @@ struct r8a66597_platdata { /* set one = big endian, set zero = little endian */ unsigned endian:1; }; -#endif + +/* Register definitions */ +#define SYSCFG0 0x00 +#define SYSCFG1 0x02 +#define SYSSTS0 0x04 +#define SYSSTS1 0x06 +#define DVSTCTR0 0x08 +#define DVSTCTR1 0x0A +#define TESTMODE 0x0C +#define PINCFG 0x0E +#define DMA0CFG 0x10 +#define DMA1CFG 0x12 +#define CFIFO 0x14 +#define D0FIFO 0x18 +#define D1FIFO 0x1C +#define CFIFOSEL 0x20 +#define CFIFOCTR 0x22 +#define CFIFOSIE 0x24 +#define D0FIFOSEL 0x28 +#define D0FIFOCTR 0x2A +#define D1FIFOSEL 0x2C +#define D1FIFOCTR 0x2E +#define INTENB0 0x30 +#define INTENB1 0x32 +#define INTENB2 0x34 +#define BRDYENB 0x36 +#define NRDYENB 0x38 +#define BEMPENB 0x3A +#define SOFCFG 0x3C +#define INTSTS0 0x40 +#define INTSTS1 0x42 +#define INTSTS2 0x44 +#define BRDYSTS 0x46 +#define NRDYSTS 0x48 +#define BEMPSTS 0x4A +#define FRMNUM 0x4C +#define UFRMNUM 0x4E +#define USBADDR 0x50 +#define USBREQ 0x54 +#define USBVAL 0x56 +#define USBINDX 0x58 +#define USBLENG 0x5A +#define DCPCFG 0x5C +#define DCPMAXP 0x5E +#define DCPCTR 0x60 +#define PIPESEL 0x64 +#define PIPECFG 0x68 +#define PIPEBUF 0x6A +#define PIPEMAXP 0x6C +#define PIPEPERI 0x6E +#define PIPE1CTR 0x70 +#define PIPE2CTR 0x72 +#define PIPE3CTR 0x74 +#define PIPE4CTR 0x76 +#define PIPE5CTR 0x78 +#define PIPE6CTR 0x7A +#define PIPE7CTR 0x7C +#define PIPE8CTR 0x7E +#define PIPE9CTR 0x80 +#define PIPE1TRE 0x90 +#define PIPE1TRN 0x92 +#define PIPE2TRE 0x94 +#define PIPE2TRN 0x96 +#define PIPE3TRE 0x98 +#define PIPE3TRN 0x9A +#define PIPE4TRE 0x9C +#define PIPE4TRN 0x9E +#define PIPE5TRE 0xA0 +#define PIPE5TRN 0xA2 +#define DEVADD0 0xD0 +#define DEVADD1 0xD2 +#define DEVADD2 0xD4 +#define DEVADD3 0xD6 +#define DEVADD4 0xD8 +#define DEVADD5 0xDA +#define DEVADD6 0xDC +#define DEVADD7 0xDE +#define DEVADD8 0xE0 +#define DEVADD9 0xE2 +#define DEVADDA 0xE4 + +/* System Configuration Control Register */ +#define XTAL 0xC000 /* b15-14: Crystal selection */ +#define XTAL48 0x8000 /* 48MHz */ +#define XTAL24 0x4000 /* 24MHz */ +#define XTAL12 0x0000 /* 12MHz */ +#define XCKE 0x2000 /* b13: External clock enable */ +#define PLLC 0x0800 /* b11: PLL control */ +#define SCKE 0x0400 /* b10: USB clock enable */ +#define PCSDIS 0x0200 /* b9: not CS wakeup */ +#define LPSME 0x0100 /* b8: Low power sleep mode */ +#define HSE 0x0080 /* b7: Hi-speed enable */ +#define DCFM 0x0040 /* b6: Controller function select */ +#define DRPD 0x0020 /* b5: D+/- pull down control */ +#define DPRPU 0x0010 /* b4: D+ pull up control */ +#define USBE 0x0001 /* b0: USB module operation enable */ + +/* System Configuration Status Register */ +#define OVCBIT 0x8000 /* b15-14: Over-current bit */ +#define OVCMON 0xC000 /* b15-14: Over-current monitor */ +#define SOFEA 0x0020 /* b5: SOF monitor */ +#define IDMON 0x0004 /* b3: ID-pin monitor */ +#define LNST 0x0003 /* b1-0: D+, D- line status */ +#define SE1 0x0003 /* SE1 */ +#define FS_KSTS 0x0002 /* Full-Speed K State */ +#define FS_JSTS 0x0001 /* Full-Speed J State */ +#define LS_JSTS 0x0002 /* Low-Speed J State */ +#define LS_KSTS 0x0001 /* Low-Speed K State */ +#define SE0 0x0000 /* SE0 */ + +/* Device State Control Register */ +#define EXTLP0 0x0400 /* b10: External port */ +#define VBOUT 0x0200 /* b9: VBUS output */ +#define WKUP 0x0100 /* b8: Remote wakeup */ +#define RWUPE 0x0080 /* b7: Remote wakeup sense */ +#define USBRST 0x0040 /* b6: USB reset enable */ +#define RESUME 0x0020 /* b5: Resume enable */ +#define UACT 0x0010 /* b4: USB bus enable */ +#define RHST 0x0007 /* b1-0: Reset handshake status */ +#define HSPROC 0x0004 /* HS handshake is processing */ +#define HSMODE 0x0003 /* Hi-Speed mode */ +#define FSMODE 0x0002 /* Full-Speed mode */ +#define LSMODE 0x0001 /* Low-Speed mode */ +#define UNDECID 0x0000 /* Undecided */ + +/* Test Mode Register */ +#define UTST 0x000F /* b3-0: Test select */ +#define H_TST_PACKET 0x000C /* HOST TEST Packet */ +#define H_TST_SE0_NAK 0x000B /* HOST TEST SE0 NAK */ +#define H_TST_K 0x000A /* HOST TEST K */ +#define H_TST_J 0x0009 /* HOST TEST J */ +#define H_TST_NORMAL 0x0000 /* HOST Normal Mode */ +#define P_TST_PACKET 0x0004 /* PERI TEST Packet */ +#define P_TST_SE0_NAK 0x0003 /* PERI TEST SE0 NAK */ +#define P_TST_K 0x0002 /* PERI TEST K */ +#define P_TST_J 0x0001 /* PERI TEST J */ +#define P_TST_NORMAL 0x0000 /* PERI Normal Mode */ + +/* Data Pin Configuration Register */ +#define LDRV 0x8000 /* b15: Drive Current Adjust */ +#define VIF1 0x0000 /* VIF = 1.8V */ +#define VIF3 0x8000 /* VIF = 3.3V */ +#define INTA 0x0001 /* b1: USB INT-pin active */ + +/* DMAx Pin Configuration Register */ +#define DREQA 0x4000 /* b14: Dreq active select */ +#define BURST 0x2000 /* b13: Burst mode */ +#define DACKA 0x0400 /* b10: Dack active select */ +#define DFORM 0x0380 /* b9-7: DMA mode select */ +#define CPU_ADR_RD_WR 0x0000 /* Address + RD/WR mode (CPU bus) */ +#define CPU_DACK_RD_WR 0x0100 /* DACK + RD/WR mode (CPU bus) */ +#define CPU_DACK_ONLY 0x0180 /* DACK only mode (CPU bus) */ +#define SPLIT_DACK_ONLY 0x0200 /* DACK only mode (SPLIT bus) */ +#define DENDA 0x0040 /* b6: Dend active select */ +#define PKTM 0x0020 /* b5: Packet mode */ +#define DENDE 0x0010 /* b4: Dend enable */ +#define OBUS 0x0004 /* b2: OUTbus mode */ + +/* CFIFO/DxFIFO Port Select Register */ +#define RCNT 0x8000 /* b15: Read count mode */ +#define REW 0x4000 /* b14: Buffer rewind */ +#define DCLRM 0x2000 /* b13: DMA buffer clear mode */ +#define DREQE 0x1000 /* b12: DREQ output enable */ +#define MBW_8 0x0000 /* 8bit */ +#define MBW_16 0x0400 /* 16bit */ +#define MBW_32 0x0800 /* 32bit */ +#define BIGEND 0x0100 /* b8: Big endian mode */ +#define BYTE_LITTLE 0x0000 /* little dendian */ +#define BYTE_BIG 0x0100 /* big endifan */ +#define ISEL 0x0020 /* b5: DCP FIFO port direction select */ +#define CURPIPE 0x000F /* b2-0: PIPE select */ + +/* CFIFO/DxFIFO Port Control Register */ +#define BVAL 0x8000 /* b15: Buffer valid flag */ +#define BCLR 0x4000 /* b14: Buffer clear */ +#define FRDY 0x2000 /* b13: FIFO ready */ +#define DTLN 0x0FFF /* b11-0: FIFO received data length */ + +/* Interrupt Enable Register 0 */ +#define VBSE 0x8000 /* b15: VBUS interrupt */ +#define RSME 0x4000 /* b14: Resume interrupt */ +#define SOFE 0x2000 /* b13: Frame update interrupt */ +#define DVSE 0x1000 /* b12: Device state transition interrupt */ +#define CTRE 0x0800 /* b11: Control transfer stage transition interrupt */ +#define BEMPE 0x0400 /* b10: Buffer empty interrupt */ +#define NRDYE 0x0200 /* b9: Buffer not ready interrupt */ +#define BRDYE 0x0100 /* b8: Buffer ready interrupt */ + +/* Interrupt Enable Register 1 */ +#define OVRCRE 0x8000 /* b15: Over-current interrupt */ +#define BCHGE 0x4000 /* b14: USB us chenge interrupt */ +#define DTCHE 0x1000 /* b12: Detach sense interrupt */ +#define ATTCHE 0x0800 /* b11: Attach sense interrupt */ +#define EOFERRE 0x0040 /* b6: EOF error interrupt */ +#define SIGNE 0x0020 /* b5: SETUP IGNORE interrupt */ +#define SACKE 0x0010 /* b4: SETUP ACK interrupt */ + +/* BRDY Interrupt Enable/Status Register */ +#define BRDY9 0x0200 /* b9: PIPE9 */ +#define BRDY8 0x0100 /* b8: PIPE8 */ +#define BRDY7 0x0080 /* b7: PIPE7 */ +#define BRDY6 0x0040 /* b6: PIPE6 */ +#define BRDY5 0x0020 /* b5: PIPE5 */ +#define BRDY4 0x0010 /* b4: PIPE4 */ +#define BRDY3 0x0008 /* b3: PIPE3 */ +#define BRDY2 0x0004 /* b2: PIPE2 */ +#define BRDY1 0x0002 /* b1: PIPE1 */ +#define BRDY0 0x0001 /* b1: PIPE0 */ + +/* NRDY Interrupt Enable/Status Register */ +#define NRDY9 0x0200 /* b9: PIPE9 */ +#define NRDY8 0x0100 /* b8: PIPE8 */ +#define NRDY7 0x0080 /* b7: PIPE7 */ +#define NRDY6 0x0040 /* b6: PIPE6 */ +#define NRDY5 0x0020 /* b5: PIPE5 */ +#define NRDY4 0x0010 /* b4: PIPE4 */ +#define NRDY3 0x0008 /* b3: PIPE3 */ +#define NRDY2 0x0004 /* b2: PIPE2 */ +#define NRDY1 0x0002 /* b1: PIPE1 */ +#define NRDY0 0x0001 /* b1: PIPE0 */ + +/* BEMP Interrupt Enable/Status Register */ +#define BEMP9 0x0200 /* b9: PIPE9 */ +#define BEMP8 0x0100 /* b8: PIPE8 */ +#define BEMP7 0x0080 /* b7: PIPE7 */ +#define BEMP6 0x0040 /* b6: PIPE6 */ +#define BEMP5 0x0020 /* b5: PIPE5 */ +#define BEMP4 0x0010 /* b4: PIPE4 */ +#define BEMP3 0x0008 /* b3: PIPE3 */ +#define BEMP2 0x0004 /* b2: PIPE2 */ +#define BEMP1 0x0002 /* b1: PIPE1 */ +#define BEMP0 0x0001 /* b0: PIPE0 */ + +/* SOF Pin Configuration Register */ +#define TRNENSEL 0x0100 /* b8: Select transaction enable period */ +#define BRDYM 0x0040 /* b6: BRDY clear timing */ +#define INTL 0x0020 /* b5: Interrupt sense select */ +#define EDGESTS 0x0010 /* b4: */ +#define SOFMODE 0x000C /* b3-2: SOF pin select */ +#define SOF_125US 0x0008 /* SOF OUT 125us Frame Signal */ +#define SOF_1MS 0x0004 /* SOF OUT 1ms Frame Signal */ +#define SOF_DISABLE 0x0000 /* SOF OUT Disable */ + +/* Interrupt Status Register 0 */ +#define VBINT 0x8000 /* b15: VBUS interrupt */ +#define RESM 0x4000 /* b14: Resume interrupt */ +#define SOFR 0x2000 /* b13: SOF frame update interrupt */ +#define DVST 0x1000 /* b12: Device state transition interrupt */ +#define CTRT 0x0800 /* b11: Control transfer stage transition interrupt */ +#define BEMP 0x0400 /* b10: Buffer empty interrupt */ +#define NRDY 0x0200 /* b9: Buffer not ready interrupt */ +#define BRDY 0x0100 /* b8: Buffer ready interrupt */ +#define VBSTS 0x0080 /* b7: VBUS input port */ +#define DVSQ 0x0070 /* b6-4: Device state */ +#define DS_SPD_CNFG 0x0070 /* Suspend Configured */ +#define DS_SPD_ADDR 0x0060 /* Suspend Address */ +#define DS_SPD_DFLT 0x0050 /* Suspend Default */ +#define DS_SPD_POWR 0x0040 /* Suspend Powered */ +#define DS_SUSP 0x0040 /* Suspend */ +#define DS_CNFG 0x0030 /* Configured */ +#define DS_ADDS 0x0020 /* Address */ +#define DS_DFLT 0x0010 /* Default */ +#define DS_POWR 0x0000 /* Powered */ +#define DVSQS 0x0030 /* b5-4: Device state */ +#define VALID 0x0008 /* b3: Setup packet detected flag */ +#define CTSQ 0x0007 /* b2-0: Control transfer stage */ +#define CS_SQER 0x0006 /* Sequence error */ +#define CS_WRND 0x0005 /* Control write nodata status stage */ +#define CS_WRSS 0x0004 /* Control write status stage */ +#define CS_WRDS 0x0003 /* Control write data stage */ +#define CS_RDSS 0x0002 /* Control read status stage */ +#define CS_RDDS 0x0001 /* Control read data stage */ +#define CS_IDST 0x0000 /* Idle or setup stage */ + +/* Interrupt Status Register 1 */ +#define OVRCR 0x8000 /* b15: Over-current interrupt */ +#define BCHG 0x4000 /* b14: USB bus chenge interrupt */ +#define DTCH 0x1000 /* b12: Detach sense interrupt */ +#define ATTCH 0x0800 /* b11: Attach sense interrupt */ +#define EOFERR 0x0040 /* b6: EOF-error interrupt */ +#define SIGN 0x0020 /* b5: Setup ignore interrupt */ +#define SACK 0x0010 /* b4: Setup acknowledge interrupt */ + +/* Frame Number Register */ +#define OVRN 0x8000 /* b15: Overrun error */ +#define CRCE 0x4000 /* b14: Received data error */ +#define FRNM 0x07FF /* b10-0: Frame number */ + +/* Micro Frame Number Register */ +#define UFRNM 0x0007 /* b2-0: Micro frame number */ + +/* Default Control Pipe Maxpacket Size Register */ +/* Pipe Maxpacket Size Register */ +#define DEVSEL 0xF000 /* b15-14: Device address select */ +#define MAXP 0x007F /* b6-0: Maxpacket size of default control pipe */ + +/* Default Control Pipe Control Register */ +#define BSTS 0x8000 /* b15: Buffer status */ +#define SUREQ 0x4000 /* b14: Send USB request */ +#define CSCLR 0x2000 /* b13: complete-split status clear */ +#define CSSTS 0x1000 /* b12: complete-split status */ +#define SUREQCLR 0x0800 /* b11: stop setup request */ +#define SQCLR 0x0100 /* b8: Sequence toggle bit clear */ +#define SQSET 0x0080 /* b7: Sequence toggle bit set */ +#define SQMON 0x0040 /* b6: Sequence toggle bit monitor */ +#define PBUSY 0x0020 /* b5: pipe busy */ +#define PINGE 0x0010 /* b4: ping enable */ +#define CCPL 0x0004 /* b2: Enable control transfer complete */ +#define PID 0x0003 /* b1-0: Response PID */ +#define PID_STALL11 0x0003 /* STALL */ +#define PID_STALL 0x0002 /* STALL */ +#define PID_BUF 0x0001 /* BUF */ +#define PID_NAK 0x0000 /* NAK */ + +/* Pipe Window Select Register */ +#define PIPENM 0x0007 /* b2-0: Pipe select */ + +/* Pipe Configuration Register */ +#define R8A66597_TYP 0xC000 /* b15-14: Transfer type */ +#define R8A66597_ISO 0xC000 /* Isochronous */ +#define R8A66597_INT 0x8000 /* Interrupt */ +#define R8A66597_BULK 0x4000 /* Bulk */ +#define R8A66597_BFRE 0x0400 /* b10: Buffer ready interrupt mode select */ +#define R8A66597_DBLB 0x0200 /* b9: Double buffer mode select */ +#define R8A66597_CNTMD 0x0100 /* b8: Continuous transfer mode select */ +#define R8A66597_SHTNAK 0x0080 /* b7: Transfer end NAK */ +#define R8A66597_DIR 0x0010 /* b4: Transfer direction select */ +#define R8A66597_EPNUM 0x000F /* b3-0: Eendpoint number select */ + +/* Pipe Buffer Configuration Register */ +#define BUFSIZE 0x7C00 /* b14-10: Pipe buffer size */ +#define BUFNMB 0x007F /* b6-0: Pipe buffer number */ +#define PIPE0BUF 256 +#define PIPExBUF 64 + +/* Pipe Maxpacket Size Register */ +#define MXPS 0x07FF /* b10-0: Maxpacket size */ + +/* Pipe Cycle Configuration Register */ +#define IFIS 0x1000 /* b12: Isochronous in-buffer flush mode select */ +#define IITV 0x0007 /* b2-0: Isochronous interval */ + +/* Pipex Control Register */ +#define BSTS 0x8000 /* b15: Buffer status */ +#define INBUFM 0x4000 /* b14: IN buffer monitor (Only for PIPE1 to 5) */ +#define CSCLR 0x2000 /* b13: complete-split status clear */ +#define CSSTS 0x1000 /* b12: complete-split status */ +#define ATREPM 0x0400 /* b10: Auto repeat mode */ +#define ACLRM 0x0200 /* b9: Out buffer auto clear mode */ +#define SQCLR 0x0100 /* b8: Sequence toggle bit clear */ +#define SQSET 0x0080 /* b7: Sequence toggle bit set */ +#define SQMON 0x0040 /* b6: Sequence toggle bit monitor */ +#define PBUSY 0x0020 /* b5: pipe busy */ +#define PID 0x0003 /* b1-0: Response PID */ + +/* PIPExTRE */ +#define TRENB 0x0200 /* b9: Transaction counter enable */ +#define TRCLR 0x0100 /* b8: Transaction counter clear */ + +/* PIPExTRN */ +#define TRNCNT 0xFFFF /* b15-0: Transaction counter */ + +/* DEVADDx */ +#define UPPHUB 0x7800 +#define HUBPORT 0x0700 +#define USBSPD 0x00C0 +#define RTPORT 0x0001 + +#endif /* __LINUX_USB_R8A66597_H */ -- cgit v1.2.3 From 2c59b0b70b9d5d61c726f179724660c4c2423f31 Mon Sep 17 00:00:00 2001 From: Magnus Damm Date: Wed, 22 Jul 2009 14:41:35 +0000 Subject: usb: m66592-udc platform data on_chip support Convert the m66592-udc driver to use the on_chip flag from platform data to enable on chip behaviour instead of relying on CONFIG_SUPERH_BUILT_IN_M66592 ugliness. This makes the code cleaner and also allows us to support both external and internal m66592 with the same kernel. It also makes the Kconfig part more future proof since we with this patch can add support for new processors with on-chip m66592 without modifying the Kconfig. The patch adds a m66592 header file for platform data and ties in platform data to the existing m66592 devices. Signed-off-by: Magnus Damm Signed-off-by: Paul Mundt --- arch/sh/boards/mach-highlander/setup.c | 7 + arch/sh/boards/mach-x3proto/setup.c | 7 + arch/sh/kernel/cpu/sh4a/setup-sh7722.c | 8 +- drivers/usb/gadget/Kconfig | 10 -- drivers/usb/gadget/m66592-udc.c | 252 +++++++++++++++++++-------------- drivers/usb/gadget/m66592-udc.h | 89 ++++++------ include/linux/usb/m66592.h | 44 ++++++ 7 files changed, 257 insertions(+), 160 deletions(-) create mode 100644 include/linux/usb/m66592.h (limited to 'include') diff --git a/arch/sh/boards/mach-highlander/setup.c b/arch/sh/boards/mach-highlander/setup.c index 1639f8915000..566e69d8d729 100644 --- a/arch/sh/boards/mach-highlander/setup.c +++ b/arch/sh/boards/mach-highlander/setup.c @@ -22,6 +22,7 @@ #include #include #include +#include #include #include #include @@ -60,6 +61,11 @@ static struct platform_device r8a66597_usb_host_device = { .resource = r8a66597_usb_host_resources, }; +static struct m66592_platdata usbf_platdata = { + .xtal = M66592_PLATDATA_XTAL_24MHZ, + .vif = 1, +}; + static struct resource m66592_usb_peripheral_resources[] = { [0] = { .name = "m66592_udc", @@ -81,6 +87,7 @@ static struct platform_device m66592_usb_peripheral_device = { .dev = { .dma_mask = NULL, /* don't use dma */ .coherent_dma_mask = 0xffffffff, + .platform_data = &usbf_platdata, }, .num_resources = ARRAY_SIZE(m66592_usb_peripheral_resources), .resource = m66592_usb_peripheral_resources, diff --git a/arch/sh/boards/mach-x3proto/setup.c b/arch/sh/boards/mach-x3proto/setup.c index 8913ae39a802..efe4cb9f8a77 100644 --- a/arch/sh/boards/mach-x3proto/setup.c +++ b/arch/sh/boards/mach-x3proto/setup.c @@ -17,6 +17,7 @@ #include #include #include +#include #include static struct resource heartbeat_resources[] = { @@ -89,6 +90,11 @@ static struct platform_device r8a66597_usb_host_device = { .resource = r8a66597_usb_host_resources, }; +static struct m66592_platdata usbf_platdata = { + .xtal = M66592_PLATDATA_XTAL_24MHZ, + .vif = 1, +}; + static struct resource m66592_usb_peripheral_resources[] = { [0] = { .name = "m66592_udc", @@ -109,6 +115,7 @@ static struct platform_device m66592_usb_peripheral_device = { .dev = { .dma_mask = NULL, /* don't use dma */ .coherent_dma_mask = 0xffffffff, + .platform_data = &usbf_platdata, }, .num_resources = ARRAY_SIZE(m66592_usb_peripheral_resources), .resource = m66592_usb_peripheral_resources, diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7722.c b/arch/sh/kernel/cpu/sh4a/setup-sh7722.c index ea524a2da3e4..0bad14a44238 100644 --- a/arch/sh/kernel/cpu/sh4a/setup-sh7722.c +++ b/arch/sh/kernel/cpu/sh4a/setup-sh7722.c @@ -13,6 +13,7 @@ #include #include #include +#include #include #include #include @@ -47,9 +48,13 @@ static struct platform_device rtc_device = { .resource = rtc_resources, }; +static struct m66592_platdata usbf_platdata = { + .on_chip = 1, +}; + static struct resource usbf_resources[] = { [0] = { - .name = "m66592_udc", + .name = "USBF", .start = 0x04480000, .end = 0x044800FF, .flags = IORESOURCE_MEM, @@ -67,6 +72,7 @@ static struct platform_device usbf_device = { .dev = { .dma_mask = NULL, .coherent_dma_mask = 0xffffffff, + .platform_data = &usbf_platdata, }, .num_resources = ARRAY_SIZE(usbf_resources), .resource = usbf_resources, diff --git a/drivers/usb/gadget/Kconfig b/drivers/usb/gadget/Kconfig index 7f8e83a954ac..b7f10bc25c2c 100644 --- a/drivers/usb/gadget/Kconfig +++ b/drivers/usb/gadget/Kconfig @@ -360,16 +360,6 @@ config USB_M66592 default USB_GADGET select USB_GADGET_SELECTED -config SUPERH_BUILT_IN_M66592 - boolean "Enable SuperH built-in USB like the M66592" - depends on USB_GADGET_M66592 && CPU_SUBTYPE_SH7722 - help - SH7722 has USB like the M66592. - - The transfer rate is very slow when use "Ethernet Gadget". - However, this problem is improved if change a value of - NET_IP_ALIGN to 4. - # # Controllers available only in discrete form (and all PCI controllers) # diff --git a/drivers/usb/gadget/m66592-udc.c b/drivers/usb/gadget/m66592-udc.c index 0dddd2f8ff35..a61c70caff12 100644 --- a/drivers/usb/gadget/m66592-udc.c +++ b/drivers/usb/gadget/m66592-udc.c @@ -31,38 +31,12 @@ #include "m66592-udc.h" - MODULE_DESCRIPTION("M66592 USB gadget driver"); MODULE_LICENSE("GPL"); MODULE_AUTHOR("Yoshihiro Shimoda"); MODULE_ALIAS("platform:m66592_udc"); -#define DRIVER_VERSION "26 Jun 2009" - -/* module parameters */ -#if defined(CONFIG_SUPERH_BUILT_IN_M66592) -static unsigned short endian = M66592_LITTLE; -module_param(endian, ushort, 0644); -MODULE_PARM_DESC(endian, "data endian: big=0, little=0 (default=0)"); -#else -static unsigned short clock = M66592_XTAL24; -module_param(clock, ushort, 0644); -MODULE_PARM_DESC(clock, "input clock: 48MHz=32768, 24MHz=16384, 12MHz=0 " - "(default=16384)"); - -static unsigned short vif = M66592_LDRV; -module_param(vif, ushort, 0644); -MODULE_PARM_DESC(vif, "input VIF: 3.3V=32768, 1.5V=0 (default=32768)"); - -static unsigned short endian; -module_param(endian, ushort, 0644); -MODULE_PARM_DESC(endian, "data endian: big=256, little=0 (default=0)"); - -static unsigned short irq_sense = M66592_INTL; -module_param(irq_sense, ushort, 0644); -MODULE_PARM_DESC(irq_sense, "IRQ sense: low level=2, falling edge=0 " - "(default=2)"); -#endif +#define DRIVER_VERSION "21 July 2009" static const char udc_name[] = "m66592_udc"; static const char *m66592_ep_name[] = { @@ -244,6 +218,7 @@ static inline int get_buffer_size(struct m66592 *m66592, u16 pipenum) static inline void pipe_change(struct m66592 *m66592, u16 pipenum) { struct m66592_ep *ep = m66592->pipenum2ep[pipenum]; + unsigned short mbw; if (ep->use_dma) return; @@ -252,7 +227,12 @@ static inline void pipe_change(struct m66592 *m66592, u16 pipenum) ndelay(450); - m66592_bset(m66592, M66592_MBW, ep->fifosel); + if (m66592->pdata->on_chip) + mbw = M66592_MBW_32; + else + mbw = M66592_MBW_16; + + m66592_bset(m66592, mbw, ep->fifosel); } static int pipe_buffer_setting(struct m66592 *m66592, @@ -332,6 +312,7 @@ static void pipe_buffer_release(struct m66592 *m66592, static void pipe_initialize(struct m66592_ep *ep) { struct m66592 *m66592 = ep->m66592; + unsigned short mbw; m66592_mdfy(m66592, 0, M66592_CURPIPE, ep->fifosel); @@ -343,7 +324,12 @@ static void pipe_initialize(struct m66592_ep *ep) ndelay(450); - m66592_bset(m66592, M66592_MBW, ep->fifosel); + if (m66592->pdata->on_chip) + mbw = M66592_MBW_32; + else + mbw = M66592_MBW_16; + + m66592_bset(m66592, mbw, ep->fifosel); } } @@ -359,15 +345,13 @@ static void m66592_ep_setting(struct m66592 *m66592, struct m66592_ep *ep, ep->fifosel = M66592_D0FIFOSEL; ep->fifoctr = M66592_D0FIFOCTR; ep->fifotrn = M66592_D0FIFOTRN; -#if !defined(CONFIG_SUPERH_BUILT_IN_M66592) - } else if (m66592->num_dma == 1) { + } else if (!m66592->pdata->on_chip && m66592->num_dma == 1) { m66592->num_dma++; ep->use_dma = 1; ep->fifoaddr = M66592_D1FIFO; ep->fifosel = M66592_D1FIFOSEL; ep->fifoctr = M66592_D1FIFOCTR; ep->fifotrn = M66592_D1FIFOTRN; -#endif } else { ep->use_dma = 0; ep->fifoaddr = M66592_CFIFO; @@ -612,76 +596,120 @@ static void start_ep0(struct m66592_ep *ep, struct m66592_request *req) } } -#if defined(CONFIG_SUPERH_BUILT_IN_M66592) static void init_controller(struct m66592 *m66592) { - m66592_bset(m66592, M66592_HSE, M66592_SYSCFG); /* High spd */ - m66592_bclr(m66592, M66592_USBE, M66592_SYSCFG); - m66592_bclr(m66592, M66592_DPRPU, M66592_SYSCFG); - m66592_bset(m66592, M66592_USBE, M66592_SYSCFG); + unsigned int endian; - /* This is a workaound for SH7722 2nd cut */ - m66592_bset(m66592, 0x8000, M66592_DVSTCTR); - m66592_bset(m66592, 0x1000, M66592_TESTMODE); - m66592_bclr(m66592, 0x8000, M66592_DVSTCTR); + if (m66592->pdata->on_chip) { + if (m66592->pdata->endian) + endian = 0; /* big endian */ + else + endian = M66592_LITTLE; /* little endian */ - m66592_bset(m66592, M66592_INTL, M66592_INTENB1); + m66592_bset(m66592, M66592_HSE, M66592_SYSCFG); /* High spd */ + m66592_bclr(m66592, M66592_USBE, M66592_SYSCFG); + m66592_bclr(m66592, M66592_DPRPU, M66592_SYSCFG); + m66592_bset(m66592, M66592_USBE, M66592_SYSCFG); - m66592_write(m66592, 0, M66592_CFBCFG); - m66592_write(m66592, 0, M66592_D0FBCFG); - m66592_bset(m66592, endian, M66592_CFBCFG); - m66592_bset(m66592, endian, M66592_D0FBCFG); -} -#else /* #if defined(CONFIG_SUPERH_BUILT_IN_M66592) */ -static void init_controller(struct m66592 *m66592) -{ - m66592_bset(m66592, (vif & M66592_LDRV) | (endian & M66592_BIGEND), - M66592_PINCFG); - m66592_bset(m66592, M66592_HSE, M66592_SYSCFG); /* High spd */ - m66592_mdfy(m66592, clock & M66592_XTAL, M66592_XTAL, M66592_SYSCFG); + /* This is a workaound for SH7722 2nd cut */ + m66592_bset(m66592, 0x8000, M66592_DVSTCTR); + m66592_bset(m66592, 0x1000, M66592_TESTMODE); + m66592_bclr(m66592, 0x8000, M66592_DVSTCTR); - m66592_bclr(m66592, M66592_USBE, M66592_SYSCFG); - m66592_bclr(m66592, M66592_DPRPU, M66592_SYSCFG); - m66592_bset(m66592, M66592_USBE, M66592_SYSCFG); + m66592_bset(m66592, M66592_INTL, M66592_INTENB1); + + m66592_write(m66592, 0, M66592_CFBCFG); + m66592_write(m66592, 0, M66592_D0FBCFG); + m66592_bset(m66592, endian, M66592_CFBCFG); + m66592_bset(m66592, endian, M66592_D0FBCFG); + } else { + unsigned int clock, vif, irq_sense; + + if (m66592->pdata->endian) + endian = M66592_BIGEND; /* big endian */ + else + endian = 0; /* little endian */ + + if (m66592->pdata->vif) + vif = M66592_LDRV; /* 3.3v */ + else + vif = 0; /* 1.5v */ + + switch (m66592->pdata->xtal) { + case M66592_PLATDATA_XTAL_12MHZ: + clock = M66592_XTAL12; + break; + case M66592_PLATDATA_XTAL_24MHZ: + clock = M66592_XTAL24; + break; + case M66592_PLATDATA_XTAL_48MHZ: + clock = M66592_XTAL48; + break; + default: + pr_warning("m66592-udc: xtal configuration error\n"); + clock = 0; + } - m66592_bset(m66592, M66592_XCKE, M66592_SYSCFG); + switch (m66592->irq_trigger) { + case IRQF_TRIGGER_LOW: + irq_sense = M66592_INTL; + break; + case IRQF_TRIGGER_FALLING: + irq_sense = 0; + break; + default: + pr_warning("m66592-udc: irq trigger config error\n"); + irq_sense = 0; + } - msleep(3); + m66592_bset(m66592, + (vif & M66592_LDRV) | (endian & M66592_BIGEND), + M66592_PINCFG); + m66592_bset(m66592, M66592_HSE, M66592_SYSCFG); /* High spd */ + m66592_mdfy(m66592, clock & M66592_XTAL, M66592_XTAL, + M66592_SYSCFG); + m66592_bclr(m66592, M66592_USBE, M66592_SYSCFG); + m66592_bclr(m66592, M66592_DPRPU, M66592_SYSCFG); + m66592_bset(m66592, M66592_USBE, M66592_SYSCFG); - m66592_bset(m66592, M66592_RCKE | M66592_PLLC, M66592_SYSCFG); + m66592_bset(m66592, M66592_XCKE, M66592_SYSCFG); + + msleep(3); - msleep(1); + m66592_bset(m66592, M66592_RCKE | M66592_PLLC, M66592_SYSCFG); - m66592_bset(m66592, M66592_SCKE, M66592_SYSCFG); + msleep(1); - m66592_bset(m66592, irq_sense & M66592_INTL, M66592_INTENB1); - m66592_write(m66592, M66592_BURST | M66592_CPU_ADR_RD_WR, - M66592_DMA0CFG); + m66592_bset(m66592, M66592_SCKE, M66592_SYSCFG); + + m66592_bset(m66592, irq_sense & M66592_INTL, M66592_INTENB1); + m66592_write(m66592, M66592_BURST | M66592_CPU_ADR_RD_WR, + M66592_DMA0CFG); + } } -#endif /* #if defined(CONFIG_SUPERH_BUILT_IN_M66592) */ static void disable_controller(struct m66592 *m66592) { -#if !defined(CONFIG_SUPERH_BUILT_IN_M66592) - m66592_bclr(m66592, M66592_SCKE, M66592_SYSCFG); - udelay(1); - m66592_bclr(m66592, M66592_PLLC, M66592_SYSCFG); - udelay(1); - m66592_bclr(m66592, M66592_RCKE, M66592_SYSCFG); - udelay(1); - m66592_bclr(m66592, M66592_XCKE, M66592_SYSCFG); -#endif + if (!m66592->pdata->on_chip) { + m66592_bclr(m66592, M66592_SCKE, M66592_SYSCFG); + udelay(1); + m66592_bclr(m66592, M66592_PLLC, M66592_SYSCFG); + udelay(1); + m66592_bclr(m66592, M66592_RCKE, M66592_SYSCFG); + udelay(1); + m66592_bclr(m66592, M66592_XCKE, M66592_SYSCFG); + } } static void m66592_start_xclock(struct m66592 *m66592) { -#if !defined(CONFIG_SUPERH_BUILT_IN_M66592) u16 tmp; - tmp = m66592_read(m66592, M66592_SYSCFG); - if (!(tmp & M66592_XCKE)) - m66592_bset(m66592, M66592_XCKE, M66592_SYSCFG); -#endif + if (!m66592->pdata->on_chip) { + tmp = m66592_read(m66592, M66592_SYSCFG); + if (!(tmp & M66592_XCKE)) + m66592_bset(m66592, M66592_XCKE, M66592_SYSCFG); + } } /*-------------------------------------------------------------------------*/ @@ -1169,8 +1197,7 @@ static irqreturn_t m66592_irq(int irq, void *_m66592) intsts0 = m66592_read(m66592, M66592_INTSTS0); intenb0 = m66592_read(m66592, M66592_INTENB0); -#if defined(CONFIG_SUPERH_BUILT_IN_M66592) - if (!intsts0 && !intenb0) { + if (m66592->pdata->on_chip && !intsts0 && !intenb0) { /* * When USB clock stops, it cannot read register. Even if a * clock stops, the interrupt occurs. So this driver turn on @@ -1180,7 +1207,6 @@ static irqreturn_t m66592_irq(int irq, void *_m66592) intsts0 = m66592_read(m66592, M66592_INTSTS0); intenb0 = m66592_read(m66592, M66592_INTENB0); } -#endif savepipe = m66592_read(m66592, M66592_CFIFOSEL); @@ -1526,9 +1552,11 @@ static int __exit m66592_remove(struct platform_device *pdev) iounmap(m66592->reg); free_irq(platform_get_irq(pdev, 0), m66592); m66592_free_request(&m66592->ep[0].ep, m66592->ep0_req); -#if defined(CONFIG_SUPERH_BUILT_IN_M66592) && defined(CONFIG_HAVE_CLK) - clk_disable(m66592->clk); - clk_put(m66592->clk); +#ifdef CONFIG_HAVE_CLK + if (m66592->pdata->on_chip) { + clk_disable(m66592->clk); + clk_put(m66592->clk); + } #endif kfree(m66592); return 0; @@ -1540,11 +1568,10 @@ static void nop_completion(struct usb_ep *ep, struct usb_request *r) static int __init m66592_probe(struct platform_device *pdev) { - struct resource *res; - int irq; + struct resource *res, *ires; void __iomem *reg = NULL; struct m66592 *m66592 = NULL; -#if defined(CONFIG_SUPERH_BUILT_IN_M66592) && defined(CONFIG_HAVE_CLK) +#ifdef CONFIG_HAVE_CLK char clk_name[8]; #endif int ret = 0; @@ -1557,10 +1584,11 @@ static int __init m66592_probe(struct platform_device *pdev) goto clean_up; } - irq = platform_get_irq(pdev, 0); - if (irq < 0) { + ires = platform_get_resource(pdev, IORESOURCE_IRQ, 0); + if (!ires) { ret = -ENODEV; - pr_err("platform_get_irq error.\n"); + dev_err(&pdev->dev, + "platform_get_resource IORESOURCE_IRQ error.\n"); goto clean_up; } @@ -1571,6 +1599,12 @@ static int __init m66592_probe(struct platform_device *pdev) goto clean_up; } + if (pdev->dev.platform_data == NULL) { + dev_err(&pdev->dev, "no platform data\n"); + ret = -ENODEV; + goto clean_up; + } + /* initialize ucd */ m66592 = kzalloc(sizeof(struct m66592), GFP_KERNEL); if (m66592 == NULL) { @@ -1578,6 +1612,9 @@ static int __init m66592_probe(struct platform_device *pdev) goto clean_up; } + m66592->pdata = pdev->dev.platform_data; + m66592->irq_trigger = ires->flags & IRQF_TRIGGER_MASK; + spin_lock_init(&m66592->lock); dev_set_drvdata(&pdev->dev, m66592); @@ -1595,22 +1632,25 @@ static int __init m66592_probe(struct platform_device *pdev) m66592->timer.data = (unsigned long)m66592; m66592->reg = reg; - ret = request_irq(irq, m66592_irq, IRQF_DISABLED | IRQF_SHARED, + ret = request_irq(ires->start, m66592_irq, IRQF_DISABLED | IRQF_SHARED, udc_name, m66592); if (ret < 0) { pr_err("request_irq error (%d)\n", ret); goto clean_up; } -#if defined(CONFIG_SUPERH_BUILT_IN_M66592) && defined(CONFIG_HAVE_CLK) - snprintf(clk_name, sizeof(clk_name), "usbf%d", pdev->id); - m66592->clk = clk_get(&pdev->dev, clk_name); - if (IS_ERR(m66592->clk)) { - dev_err(&pdev->dev, "cannot get clock \"%s\"\n", clk_name); - ret = PTR_ERR(m66592->clk); - goto clean_up2; +#ifdef CONFIG_HAVE_CLK + if (m66592->pdata->on_chip) { + snprintf(clk_name, sizeof(clk_name), "usbf%d", pdev->id); + m66592->clk = clk_get(&pdev->dev, clk_name); + if (IS_ERR(m66592->clk)) { + dev_err(&pdev->dev, "cannot get clock \"%s\"\n", + clk_name); + ret = PTR_ERR(m66592->clk); + goto clean_up2; + } + clk_enable(m66592->clk); } - clk_enable(m66592->clk); #endif INIT_LIST_HEAD(&m66592->gadget.ep_list); m66592->gadget.ep0 = &m66592->ep[0].ep; @@ -1652,12 +1692,14 @@ static int __init m66592_probe(struct platform_device *pdev) return 0; clean_up3: -#if defined(CONFIG_SUPERH_BUILT_IN_M66592) && defined(CONFIG_HAVE_CLK) - clk_disable(m66592->clk); - clk_put(m66592->clk); +#ifdef CONFIG_HAVE_CLK + if (m66592->pdata->on_chip) { + clk_disable(m66592->clk); + clk_put(m66592->clk); + } clean_up2: #endif - free_irq(irq, m66592); + free_irq(ires->start, m66592); clean_up: if (m66592) { if (m66592->ep0_req) diff --git a/drivers/usb/gadget/m66592-udc.h b/drivers/usb/gadget/m66592-udc.h index 9a9c2bf9fbd5..8b960deed680 100644 --- a/drivers/usb/gadget/m66592-udc.h +++ b/drivers/usb/gadget/m66592-udc.h @@ -23,10 +23,12 @@ #ifndef __M66592_UDC_H__ #define __M66592_UDC_H__ -#if defined(CONFIG_SUPERH_BUILT_IN_M66592) && defined(CONFIG_HAVE_CLK) +#ifdef CONFIG_HAVE_CLK #include #endif +#include + #define M66592_SYSCFG 0x00 #define M66592_XTAL 0xC000 /* b15-14: Crystal selection */ #define M66592_XTAL48 0x8000 /* 48MHz */ @@ -76,11 +78,11 @@ #define M66592_P_TST_J 0x0001 /* PERI TEST J */ #define M66592_P_TST_NORMAL 0x0000 /* PERI Normal Mode */ -#if defined(CONFIG_SUPERH_BUILT_IN_M66592) +/* built-in registers */ #define M66592_CFBCFG 0x0A #define M66592_D0FBCFG 0x0C #define M66592_LITTLE 0x0100 /* b8: Little endian mode */ -#else +/* external chip case */ #define M66592_PINCFG 0x0A #define M66592_LDRV 0x8000 /* b15: Drive Current Adjust */ #define M66592_BIGEND 0x0100 /* b8: Big endian mode */ @@ -100,8 +102,8 @@ #define M66592_PKTM 0x0020 /* b5: Packet mode */ #define M66592_DENDE 0x0010 /* b4: Dend enable */ #define M66592_OBUS 0x0004 /* b2: OUTbus mode */ -#endif /* #if defined(CONFIG_SUPERH_BUILT_IN_M66592) */ +/* common case */ #define M66592_CFIFO 0x10 #define M66592_D0FIFO 0x14 #define M66592_D1FIFO 0x18 @@ -113,13 +115,9 @@ #define M66592_REW 0x4000 /* b14: Buffer rewind */ #define M66592_DCLRM 0x2000 /* b13: DMA buffer clear mode */ #define M66592_DREQE 0x1000 /* b12: DREQ output enable */ -#if defined(CONFIG_SUPERH_BUILT_IN_M66592) -#define M66592_MBW 0x0800 /* b11: Maximum bit width for FIFO */ -#else -#define M66592_MBW 0x0400 /* b10: Maximum bit width for FIFO */ -#define M66592_MBW_8 0x0000 /* 8bit */ -#define M66592_MBW_16 0x0400 /* 16bit */ -#endif /* #if defined(CONFIG_SUPERH_BUILT_IN_M66592) */ +#define M66592_MBW_8 0x0000 /* 8bit */ +#define M66592_MBW_16 0x0400 /* 16bit */ +#define M66592_MBW_32 0x0800 /* 32bit */ #define M66592_TRENB 0x0200 /* b9: Transaction counter enable */ #define M66592_TRCLR 0x0100 /* b8: Transaction counter clear */ #define M66592_DEZPM 0x0080 /* b7: Zero-length packet mode */ @@ -480,9 +478,11 @@ struct m66592_ep { struct m66592 { spinlock_t lock; void __iomem *reg; -#if defined(CONFIG_SUPERH_BUILT_IN_M66592) && defined(CONFIG_HAVE_CLK) +#ifdef CONFIG_HAVE_CLK struct clk *clk; #endif + struct m66592_platdata *pdata; + unsigned long irq_trigger; struct usb_gadget gadget; struct usb_gadget_driver *driver; @@ -546,13 +546,13 @@ static inline void m66592_read_fifo(struct m66592 *m66592, { unsigned long fifoaddr = (unsigned long)m66592->reg + offset; -#if defined(CONFIG_SUPERH_BUILT_IN_M66592) - len = (len + 3) / 4; - insl(fifoaddr, buf, len); -#else - len = (len + 1) / 2; - insw(fifoaddr, buf, len); -#endif + if (m66592->pdata->on_chip) { + len = (len + 3) / 4; + insl(fifoaddr, buf, len); + } else { + len = (len + 1) / 2; + insw(fifoaddr, buf, len); + } } static inline void m66592_write(struct m66592 *m66592, u16 val, @@ -566,33 +566,34 @@ static inline void m66592_write_fifo(struct m66592 *m66592, void *buf, unsigned long len) { unsigned long fifoaddr = (unsigned long)m66592->reg + offset; -#if defined(CONFIG_SUPERH_BUILT_IN_M66592) - unsigned long count; - unsigned char *pb; - int i; - - count = len / 4; - outsl(fifoaddr, buf, count); - - if (len & 0x00000003) { - pb = buf + count * 4; - for (i = 0; i < (len & 0x00000003); i++) { - if (m66592_read(m66592, M66592_CFBCFG)) /* little */ - outb(pb[i], fifoaddr + (3 - i)); - else - outb(pb[i], fifoaddr + i); + + if (m66592->pdata->on_chip) { + unsigned long count; + unsigned char *pb; + int i; + + count = len / 4; + outsl(fifoaddr, buf, count); + + if (len & 0x00000003) { + pb = buf + count * 4; + for (i = 0; i < (len & 0x00000003); i++) { + if (m66592_read(m66592, M66592_CFBCFG)) /* le */ + outb(pb[i], fifoaddr + (3 - i)); + else + outb(pb[i], fifoaddr + i); + } + } + } else { + unsigned long odd = len & 0x0001; + + len = len / 2; + outsw(fifoaddr, buf, len); + if (odd) { + unsigned char *p = buf + len*2; + outb(*p, fifoaddr); } } -#else - unsigned long odd = len & 0x0001; - - len = len / 2; - outsw(fifoaddr, buf, len); - if (odd) { - unsigned char *p = buf + len*2; - outb(*p, fifoaddr); - } -#endif /* #if defined(CONFIG_SUPERH_BUILT_IN_M66592) */ } static inline void m66592_mdfy(struct m66592 *m66592, u16 val, u16 pat, diff --git a/include/linux/usb/m66592.h b/include/linux/usb/m66592.h new file mode 100644 index 000000000000..cda9625e7df0 --- /dev/null +++ b/include/linux/usb/m66592.h @@ -0,0 +1,44 @@ +/* + * M66592 driver platform data + * + * Copyright (C) 2009 Renesas Solutions Corp. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 of the License. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA + * + */ + +#ifndef __LINUX_USB_M66592_H +#define __LINUX_USB_M66592_H + +#define M66592_PLATDATA_XTAL_12MHZ 0x01 +#define M66592_PLATDATA_XTAL_24MHZ 0x02 +#define M66592_PLATDATA_XTAL_48MHZ 0x03 + +struct m66592_platdata { + /* one = on chip controller, zero = external controller */ + unsigned on_chip:1; + + /* one = big endian, zero = little endian */ + unsigned endian:1; + + /* (external controller only) M66592_PLATDATA_XTAL_nnMHZ */ + unsigned xtal:2; + + /* (external controller only) one = 3.3V, zero = 1.5V */ + unsigned vif:1; + +}; + +#endif /* __LINUX_USB_M66592_H */ + -- cgit v1.2.3 From 878fa89f97954337d1dc41f0ccc3a8b5f89cfbc7 Mon Sep 17 00:00:00 2001 From: Daniel Silverstone Date: Wed, 22 Jul 2009 18:51:24 +0200 Subject: IEEE80154: Add documentation to the IEEE80154 netlink and fakehard driver This adds some perfunctory documentation comments to the IEEE 802.15.4 fakehard.c driver (Fake hard MAC) and the nl802154.h (outgoing netlink messages) header. These comments are not necessarily complete, but they do reference the IEEE 802.15.4-2006 document where possible. Signed-off-by: Daniel Silverstone --- drivers/ieee802154/fakehard.c | 120 ++++++++++++++++++++++++++++++++++++++ include/net/ieee802154/nl802154.h | 76 ++++++++++++++++++++++++ 2 files changed, 196 insertions(+) (limited to 'include') diff --git a/drivers/ieee802154/fakehard.c b/drivers/ieee802154/fakehard.c index 0384144c0b34..9ec07e8552f2 100644 --- a/drivers/ieee802154/fakehard.c +++ b/drivers/ieee802154/fakehard.c @@ -31,6 +31,12 @@ #include #include +/** + * fake_get_pan_id - Retrieve the PAN ID of the device. + * @dev: The network device to retrieve the PAN of. + * + * Return the ID of the PAN from the PIB. + */ static u16 fake_get_pan_id(struct net_device *dev) { BUG_ON(dev->type != ARPHRD_IEEE802154); @@ -38,6 +44,14 @@ static u16 fake_get_pan_id(struct net_device *dev) return 0xeba1; } +/** + * fake_get_short_addr - Retrieve the short address of the device. + * @dev: The network device to retrieve the short address of. + * + * Returns the IEEE 802.15.4 short-form address cached for this + * device. If the device has not yet had a short address assigned + * then this should return 0xFFFF to indicate a lack of association. + */ static u16 fake_get_short_addr(struct net_device *dev) { BUG_ON(dev->type != ARPHRD_IEEE802154); @@ -45,6 +59,19 @@ static u16 fake_get_short_addr(struct net_device *dev) return 0x1; } +/** + * fake_get_dsn - Retrieve the DSN of the device. + * @dev: The network device to retrieve the DSN for. + * + * Returns the IEEE 802.15.4 DSN for the network device. + * The DSN is the sequence number which will be added to each + * packet or MAC command frame by the MAC during transmission. + * + * DSN means 'Data Sequence Number'. + * + * Note: This is in section 7.2.1.2 of the IEEE 802.15.4-2006 + * document. + */ static u8 fake_get_dsn(struct net_device *dev) { BUG_ON(dev->type != ARPHRD_IEEE802154); @@ -52,6 +79,19 @@ static u8 fake_get_dsn(struct net_device *dev) return 0x00; /* DSN are implemented in HW, so return just 0 */ } +/** + * fake_get_bsn - Retrieve the BSN of the device. + * @dev: The network device to retrieve the BSN for. + * + * Returns the IEEE 802.15.4 BSN for the network device. + * The BSN is the sequence number which will be added to each + * beacon frame sent by the MAC. + * + * BSN means 'Beacon Sequence Number'. + * + * Note: This is in section 7.2.1.2 of the IEEE 802.15.4-2006 + * document. + */ static u8 fake_get_bsn(struct net_device *dev) { BUG_ON(dev->type != ARPHRD_IEEE802154); @@ -59,6 +99,19 @@ static u8 fake_get_bsn(struct net_device *dev) return 0x00; /* BSN are implemented in HW, so return just 0 */ } +/** + * fake_assoc_req - Make an association request to the HW. + * @dev: The network device which we are associating to a network. + * @addr: The coordinator with which we wish to associate. + * @channel: The channel on which to associate. + * @cap: The capability information field to use in the association. + * + * Start an association with a coordinator. The coordinator's address + * and PAN ID can be found in @addr. + * + * Note: This is in section 7.3.1 and 7.5.3.1 of the IEEE + * 802.15.4-2006 document. + */ static int fake_assoc_req(struct net_device *dev, struct ieee802154_addr *addr, u8 channel, u8 cap) { @@ -67,18 +120,70 @@ static int fake_assoc_req(struct net_device *dev, IEEE802154_SUCCESS); } +/** + * fake_assoc_resp - Send an association response to a device. + * @dev: The network device on which to send the response. + * @addr: The address of the device to respond to. + * @short_addr: The assigned short address for the device (if any). + * @status: The result of the association request. + * + * Queue the association response of the coordinator to another + * device's attempt to associate with the network which we + * coordinate. This is then added to the indirect-send queue to be + * transmitted to the end device when it polls for data. + * + * Note: This is in section 7.3.2 and 7.5.3.1 of the IEEE + * 802.15.4-2006 document. + */ static int fake_assoc_resp(struct net_device *dev, struct ieee802154_addr *addr, u16 short_addr, u8 status) { return 0; } +/** + * fake_disassoc_req - Disassociate a device from a network. + * @dev: The network device on which we're disassociating a device. + * @addr: The device to disassociate from the network. + * @reason: The reason to give to the device for being disassociated. + * + * This sends a disassociation notification to the device being + * disassociated from the network. + * + * Note: This is in section 7.5.3.2 of the IEEE 802.15.4-2006 + * document, with the reason described in 7.3.3.2. + */ static int fake_disassoc_req(struct net_device *dev, struct ieee802154_addr *addr, u8 reason) { return ieee802154_nl_disassoc_confirm(dev, IEEE802154_SUCCESS); } +/** + * fake_start_req - Start an IEEE 802.15.4 PAN. + * @dev: The network device on which to start the PAN. + * @addr: The coordinator address to use when starting the PAN. + * @channel: The channel on which to start the PAN. + * @bcn_ord: Beacon order. + * @sf_ord: Superframe order. + * @pan_coord: Whether or not we are the PAN coordinator or just + * requesting a realignment perhaps? + * @blx: Battery Life Extension feature bitfield. + * @coord_realign: Something to realign something else. + * + * If pan_coord is non-zero then this starts a network with the + * provided parameters, otherwise it attempts a coordinator + * realignment of the stated network instead. + * + * Note: This is in section 7.5.2.3 of the IEEE 802.15.4-2006 + * document, with 7.3.8 describing coordinator realignment. + * + * Note: There is currently no way to notify the coordinator userland + * program of whether or not the PAN has started successfully. As + * such, the coordinator program cannot know when the MAC has + * completed starting the network and will simply have to assume + * completeness based on some form of time delay. + */ static int fake_start_req(struct net_device *dev, struct ieee802154_addr *addr, u8 channel, u8 bcn_ord, u8 sf_ord, u8 pan_coord, u8 blx, @@ -87,6 +192,21 @@ static int fake_start_req(struct net_device *dev, struct ieee802154_addr *addr, return 0; } +/** + * fake_scan_req - Start a channel scan. + * @dev: The network device on which to perform a channel scan. + * @type: The type of scan to perform. + * @channels: The channel bitmask to scan. + * @duration: How long to spend on each channel. + * + * This starts either a passive (energy) scan or an active (PAN) scan + * on the channels indicated in the @channels bitmask. The duration of + * the scan is measured in terms of superframe duration. Specifically, + * the scan will spend aBaseSuperFrameDuration * ((2^n) + 1) on each + * channel. + * + * Note: This is in section 7.5.2.1 of the IEEE 802.15.4-2006 document. + */ static int fake_scan_req(struct net_device *dev, u8 type, u32 channels, u8 duration) { diff --git a/include/net/ieee802154/nl802154.h b/include/net/ieee802154/nl802154.h index 78efcdf52b59..6096096f6d7d 100644 --- a/include/net/ieee802154/nl802154.h +++ b/include/net/ieee802154/nl802154.h @@ -24,17 +24,93 @@ struct net_device; struct ieee802154_addr; +/** + * ieee802154_nl_assoc_indic - Notify userland of an association request. + * @dev: The network device on which this association request was + * received. + * @addr: The address of the device requesting association. + * @cap: The capability information field from the device. + * + * This informs a userland coordinator of a device requesting to + * associate with the PAN controlled by the coordinator. + * + * Note: This is in section 7.3.1 of the IEEE 802.15.4-2006 document. + */ int ieee802154_nl_assoc_indic(struct net_device *dev, struct ieee802154_addr *addr, u8 cap); + +/** + * ieee802154_nl_assoc_confirm - Notify userland of association. + * @dev: The device which has completed association. + * @short_addr: The short address assigned to the device. + * @status: The status of the association. + * + * Inform userland of the result of an association request. If the + * association request included asking the coordinator to allocate + * a short address then it is returned in @short_addr. + * + * Note: This is in section 7.3.2 of the IEEE 802.15.4 document. + */ int ieee802154_nl_assoc_confirm(struct net_device *dev, u16 short_addr, u8 status); + +/** + * ieee802154_nl_disassoc_indic - Notify userland of disassociation. + * @dev: The device on which disassociation was indicated. + * @addr: The device which is disassociating. + * @reason: The reason for the disassociation. + * + * Inform userland that a device has disassociated from the network. + * + * Note: This is in section 7.3.3 of the IEEE 802.15.4 document. + */ int ieee802154_nl_disassoc_indic(struct net_device *dev, struct ieee802154_addr *addr, u8 reason); + +/** + * ieee802154_nl_disassoc_confirm - Notify userland of disassociation + * completion. + * @dev: The device on which disassociation was ordered. + * @status: The result of the disassociation. + * + * Inform userland of the result of requesting that a device + * disassociate, or the result of requesting that we disassociate from + * a PAN managed by another coordinator. + * + * Note: This is in section 7.1.4.3 of the IEEE 802.15.4 document. + */ int ieee802154_nl_disassoc_confirm(struct net_device *dev, u8 status); + +/** + * ieee802154_nl_scan_confirm - Notify userland of completion of scan. + * @dev: The device which was instructed to scan. + * @status: The status of the scan operation. + * @scan_type: What type of scan was performed. + * @unscanned: Any channels that the device was unable to scan. + * @edl: The energy levels (if a passive scan). + * + * + * Note: This is in section 7.1.11 of the IEEE 802.15.4 document. + * Note: This API does not permit the return of an active scan result. + */ int ieee802154_nl_scan_confirm(struct net_device *dev, u8 status, u8 scan_type, u32 unscanned, u8 *edl/*, struct list_head *pan_desc_list */); + +/** + * ieee802154_nl_beacon_indic - Notify userland of a received beacon. + * @dev: The device on which a beacon was received. + * @panid: The PAN of the coordinator. + * @coord_addr: The short address of the coordinator on that PAN. + * + * Note: This is in section 7.1.5 of the IEEE 802.15.4 document. + * Note: This API does not provide extended information such as what + * channel the PAN is on or what the LQI of the beacon frame was on + * receipt. + * Note: This API cannot indicate a beacon frame for a coordinator + * operating in long addressing mode. + */ int ieee802154_nl_beacon_indic(struct net_device *dev, u16 panid, u16 coord_addr); -- cgit v1.2.3 From f0166e5e3cdab66d5a31f796ce18e21fd3ce99dc Mon Sep 17 00:00:00 2001 From: Dmitry Eremin-Solenikov Date: Thu, 23 Jul 2009 16:56:29 +0400 Subject: ieee802154: move headers out of extra directory include/net/ieee802154/af_ieee802154.h (and others) naming seems to be too long and redundant. Drop one level of subdirectories. Signed-off-by: Dmitry Eremin-Solenikov --- drivers/ieee802154/fakehard.c | 8 +- include/net/af_ieee802154.h | 60 +++++++++++++ include/net/ieee802154.h | 160 +++++++++++++++++++++++++++++++++ include/net/ieee802154/af_ieee802154.h | 60 ------------- include/net/ieee802154/mac_def.h | 160 --------------------------------- include/net/ieee802154/netdevice.h | 115 ------------------------ include/net/ieee802154/nl802154.h | 117 ------------------------ include/net/ieee802154_netdev.h | 115 ++++++++++++++++++++++++ include/net/nl802154.h | 117 ++++++++++++++++++++++++ net/ieee802154/af_ieee802154.c | 4 +- net/ieee802154/dgram.c | 6 +- net/ieee802154/netlink.c | 6 +- net/ieee802154/raw.c | 2 +- 13 files changed, 465 insertions(+), 465 deletions(-) create mode 100644 include/net/af_ieee802154.h create mode 100644 include/net/ieee802154.h delete mode 100644 include/net/ieee802154/af_ieee802154.h delete mode 100644 include/net/ieee802154/mac_def.h delete mode 100644 include/net/ieee802154/netdevice.h delete mode 100644 include/net/ieee802154/nl802154.h create mode 100644 include/net/ieee802154_netdev.h create mode 100644 include/net/nl802154.h (limited to 'include') diff --git a/drivers/ieee802154/fakehard.c b/drivers/ieee802154/fakehard.c index 9ec07e8552f2..8a52e6efa239 100644 --- a/drivers/ieee802154/fakehard.c +++ b/drivers/ieee802154/fakehard.c @@ -26,10 +26,10 @@ #include #include -#include -#include -#include -#include +#include +#include +#include +#include /** * fake_get_pan_id - Retrieve the PAN ID of the device. diff --git a/include/net/af_ieee802154.h b/include/net/af_ieee802154.h new file mode 100644 index 000000000000..0d78605fb1a6 --- /dev/null +++ b/include/net/af_ieee802154.h @@ -0,0 +1,60 @@ +/* + * IEEE 802.15.4 inteface for userspace + * + * Copyright 2007, 2008 Siemens AG + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA. + * + * Written by: + * Sergey Lapin + * Dmitry Eremin-Solenikov + */ + +#ifndef _AF_IEEE802154_H +#define _AF_IEEE802154_H + +#include /* for sa_family_t */ + +enum { + IEEE802154_ADDR_NONE = 0x0, + /* RESERVED = 0x01, */ + IEEE802154_ADDR_SHORT = 0x2, /* 16-bit address + PANid */ + IEEE802154_ADDR_LONG = 0x3, /* 64-bit address + PANid */ +}; + +/* address length, octets */ +#define IEEE802154_ADDR_LEN 8 + +struct ieee802154_addr { + int addr_type; + u16 pan_id; + union { + u8 hwaddr[IEEE802154_ADDR_LEN]; + u16 short_addr; + }; +}; + +#define IEEE802154_PANID_BROADCAST 0xffff +#define IEEE802154_ADDR_BROADCAST 0xffff +#define IEEE802154_ADDR_UNDEF 0xfffe + +struct sockaddr_ieee802154 { + sa_family_t family; /* AF_IEEE802154 */ + struct ieee802154_addr addr; +}; + +/* master device */ +#define IEEE802154_SIOC_ADD_SLAVE (SIOCDEVPRIVATE + 0) + +#endif diff --git a/include/net/ieee802154.h b/include/net/ieee802154.h new file mode 100644 index 000000000000..d52685defb11 --- /dev/null +++ b/include/net/ieee802154.h @@ -0,0 +1,160 @@ +/* + * IEEE802.15.4-2003 specification + * + * Copyright (C) 2007, 2008 Siemens AG + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA. + * + * Written by: + * Pavel Smolenskiy + * Maxim Gorbachyov + * Maxim Osipov + * Dmitry Eremin-Solenikov + */ + +#ifndef NET_IEEE802154_H +#define NET_IEEE802154_H + +#define IEEE802154_FC_TYPE_BEACON 0x0 /* Frame is beacon */ +#define IEEE802154_FC_TYPE_DATA 0x1 /* Frame is data */ +#define IEEE802154_FC_TYPE_ACK 0x2 /* Frame is acknowledgment */ +#define IEEE802154_FC_TYPE_MAC_CMD 0x3 /* Frame is MAC command */ + +#define IEEE802154_FC_TYPE_SHIFT 0 +#define IEEE802154_FC_TYPE_MASK ((1 << 3) - 1) +#define IEEE802154_FC_TYPE(x) ((x & IEEE802154_FC_TYPE_MASK) >> IEEE802154_FC_TYPE_SHIFT) +#define IEEE802154_FC_SET_TYPE(v, x) do { \ + v = (((v) & ~IEEE802154_FC_TYPE_MASK) | \ + (((x) << IEEE802154_FC_TYPE_SHIFT) & IEEE802154_FC_TYPE_MASK)); \ + } while (0) + +#define IEEE802154_FC_SECEN (1 << 3) +#define IEEE802154_FC_FRPEND (1 << 4) +#define IEEE802154_FC_ACK_REQ (1 << 5) +#define IEEE802154_FC_INTRA_PAN (1 << 6) + +#define IEEE802154_FC_SAMODE_SHIFT 14 +#define IEEE802154_FC_SAMODE_MASK (3 << IEEE802154_FC_SAMODE_SHIFT) +#define IEEE802154_FC_DAMODE_SHIFT 10 +#define IEEE802154_FC_DAMODE_MASK (3 << IEEE802154_FC_DAMODE_SHIFT) + +#define IEEE802154_FC_SAMODE(x) \ + (((x) & IEEE802154_FC_SAMODE_MASK) >> IEEE802154_FC_SAMODE_SHIFT) + +#define IEEE802154_FC_DAMODE(x) \ + (((x) & IEEE802154_FC_DAMODE_MASK) >> IEEE802154_FC_DAMODE_SHIFT) + + +/* MAC's Command Frames Identifiers */ +#define IEEE802154_CMD_ASSOCIATION_REQ 0x01 +#define IEEE802154_CMD_ASSOCIATION_RESP 0x02 +#define IEEE802154_CMD_DISASSOCIATION_NOTIFY 0x03 +#define IEEE802154_CMD_DATA_REQ 0x04 +#define IEEE802154_CMD_PANID_CONFLICT_NOTIFY 0x05 +#define IEEE802154_CMD_ORPHAN_NOTIFY 0x06 +#define IEEE802154_CMD_BEACON_REQ 0x07 +#define IEEE802154_CMD_COORD_REALIGN_NOTIFY 0x08 +#define IEEE802154_CMD_GTS_REQ 0x09 + +/* + * The return values of MAC operations + */ +enum { + /* + * The requested operation was completed successfully. + * For a transmission request, this value indicates + * a successful transmission. + */ + IEEE802154_SUCCESS = 0x0, + + /* The beacon was lost following a synchronization request. */ + IEEE802154_BEACON_LOSS = 0xe0, + /* + * A transmission could not take place due to activity on the + * channel, i.e., the CSMA-CA mechanism has failed. + */ + IEEE802154_CHNL_ACCESS_FAIL = 0xe1, + /* The GTS request has been denied by the PAN coordinator. */ + IEEE802154_DENINED = 0xe2, + /* The attempt to disable the transceiver has failed. */ + IEEE802154_DISABLE_TRX_FAIL = 0xe3, + /* + * The received frame induces a failed security check according to + * the security suite. + */ + IEEE802154_FAILED_SECURITY_CHECK = 0xe4, + /* + * The frame resulting from secure processing has a length that is + * greater than aMACMaxFrameSize. + */ + IEEE802154_FRAME_TOO_LONG = 0xe5, + /* + * The requested GTS transmission failed because the specified GTS + * either did not have a transmit GTS direction or was not defined. + */ + IEEE802154_INVALID_GTS = 0xe6, + /* + * A request to purge an MSDU from the transaction queue was made using + * an MSDU handle that was not found in the transaction table. + */ + IEEE802154_INVALID_HANDLE = 0xe7, + /* A parameter in the primitive is out of the valid range.*/ + IEEE802154_INVALID_PARAMETER = 0xe8, + /* No acknowledgment was received after aMaxFrameRetries. */ + IEEE802154_NO_ACK = 0xe9, + /* A scan operation failed to find any network beacons.*/ + IEEE802154_NO_BEACON = 0xea, + /* No response data were available following a request. */ + IEEE802154_NO_DATA = 0xeb, + /* The operation failed because a short address was not allocated. */ + IEEE802154_NO_SHORT_ADDRESS = 0xec, + /* + * A receiver enable request was unsuccessful because it could not be + * completed within the CAP. + */ + IEEE802154_OUT_OF_CAP = 0xed, + /* + * A PAN identifier conflict has been detected and communicated to the + * PAN coordinator. + */ + IEEE802154_PANID_CONFLICT = 0xee, + /* A coordinator realignment command has been received. */ + IEEE802154_REALIGMENT = 0xef, + /* The transaction has expired and its information discarded. */ + IEEE802154_TRANSACTION_EXPIRED = 0xf0, + /* There is no capacity to store the transaction. */ + IEEE802154_TRANSACTION_OVERFLOW = 0xf1, + /* + * The transceiver was in the transmitter enabled state when the + * receiver was requested to be enabled. + */ + IEEE802154_TX_ACTIVE = 0xf2, + /* The appropriate key is not available in the ACL. */ + IEEE802154_UNAVAILABLE_KEY = 0xf3, + /* + * A SET/GET request was issued with the identifier of a PIB attribute + * that is not supported. + */ + IEEE802154_UNSUPPORTED_ATTR = 0xf4, + /* + * A request to perform a scan operation failed because the MLME was + * in the process of performing a previously initiated scan operation. + */ + IEEE802154_SCAN_IN_PROGRESS = 0xfc, +}; + + +#endif + + diff --git a/include/net/ieee802154/af_ieee802154.h b/include/net/ieee802154/af_ieee802154.h deleted file mode 100644 index 0d78605fb1a6..000000000000 --- a/include/net/ieee802154/af_ieee802154.h +++ /dev/null @@ -1,60 +0,0 @@ -/* - * IEEE 802.15.4 inteface for userspace - * - * Copyright 2007, 2008 Siemens AG - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License version 2 - * as published by the Free Software Foundation. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License along - * with this program; if not, write to the Free Software Foundation, Inc., - * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA. - * - * Written by: - * Sergey Lapin - * Dmitry Eremin-Solenikov - */ - -#ifndef _AF_IEEE802154_H -#define _AF_IEEE802154_H - -#include /* for sa_family_t */ - -enum { - IEEE802154_ADDR_NONE = 0x0, - /* RESERVED = 0x01, */ - IEEE802154_ADDR_SHORT = 0x2, /* 16-bit address + PANid */ - IEEE802154_ADDR_LONG = 0x3, /* 64-bit address + PANid */ -}; - -/* address length, octets */ -#define IEEE802154_ADDR_LEN 8 - -struct ieee802154_addr { - int addr_type; - u16 pan_id; - union { - u8 hwaddr[IEEE802154_ADDR_LEN]; - u16 short_addr; - }; -}; - -#define IEEE802154_PANID_BROADCAST 0xffff -#define IEEE802154_ADDR_BROADCAST 0xffff -#define IEEE802154_ADDR_UNDEF 0xfffe - -struct sockaddr_ieee802154 { - sa_family_t family; /* AF_IEEE802154 */ - struct ieee802154_addr addr; -}; - -/* master device */ -#define IEEE802154_SIOC_ADD_SLAVE (SIOCDEVPRIVATE + 0) - -#endif diff --git a/include/net/ieee802154/mac_def.h b/include/net/ieee802154/mac_def.h deleted file mode 100644 index 8cb684635650..000000000000 --- a/include/net/ieee802154/mac_def.h +++ /dev/null @@ -1,160 +0,0 @@ -/* - * IEEE802.15.4-2003 specification - * - * Copyright (C) 2007, 2008 Siemens AG - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License version 2 - * as published by the Free Software Foundation. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License along - * with this program; if not, write to the Free Software Foundation, Inc., - * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA. - * - * Written by: - * Pavel Smolenskiy - * Maxim Gorbachyov - * Maxim Osipov - * Dmitry Eremin-Solenikov - */ - -#ifndef IEEE802154_MAC_DEF_H -#define IEEE802154_MAC_DEF_H - -#define IEEE802154_FC_TYPE_BEACON 0x0 /* Frame is beacon */ -#define IEEE802154_FC_TYPE_DATA 0x1 /* Frame is data */ -#define IEEE802154_FC_TYPE_ACK 0x2 /* Frame is acknowledgment */ -#define IEEE802154_FC_TYPE_MAC_CMD 0x3 /* Frame is MAC command */ - -#define IEEE802154_FC_TYPE_SHIFT 0 -#define IEEE802154_FC_TYPE_MASK ((1 << 3) - 1) -#define IEEE802154_FC_TYPE(x) ((x & IEEE802154_FC_TYPE_MASK) >> IEEE802154_FC_TYPE_SHIFT) -#define IEEE802154_FC_SET_TYPE(v, x) do { \ - v = (((v) & ~IEEE802154_FC_TYPE_MASK) | \ - (((x) << IEEE802154_FC_TYPE_SHIFT) & IEEE802154_FC_TYPE_MASK)); \ - } while (0) - -#define IEEE802154_FC_SECEN (1 << 3) -#define IEEE802154_FC_FRPEND (1 << 4) -#define IEEE802154_FC_ACK_REQ (1 << 5) -#define IEEE802154_FC_INTRA_PAN (1 << 6) - -#define IEEE802154_FC_SAMODE_SHIFT 14 -#define IEEE802154_FC_SAMODE_MASK (3 << IEEE802154_FC_SAMODE_SHIFT) -#define IEEE802154_FC_DAMODE_SHIFT 10 -#define IEEE802154_FC_DAMODE_MASK (3 << IEEE802154_FC_DAMODE_SHIFT) - -#define IEEE802154_FC_SAMODE(x) \ - (((x) & IEEE802154_FC_SAMODE_MASK) >> IEEE802154_FC_SAMODE_SHIFT) - -#define IEEE802154_FC_DAMODE(x) \ - (((x) & IEEE802154_FC_DAMODE_MASK) >> IEEE802154_FC_DAMODE_SHIFT) - - -/* MAC's Command Frames Identifiers */ -#define IEEE802154_CMD_ASSOCIATION_REQ 0x01 -#define IEEE802154_CMD_ASSOCIATION_RESP 0x02 -#define IEEE802154_CMD_DISASSOCIATION_NOTIFY 0x03 -#define IEEE802154_CMD_DATA_REQ 0x04 -#define IEEE802154_CMD_PANID_CONFLICT_NOTIFY 0x05 -#define IEEE802154_CMD_ORPHAN_NOTIFY 0x06 -#define IEEE802154_CMD_BEACON_REQ 0x07 -#define IEEE802154_CMD_COORD_REALIGN_NOTIFY 0x08 -#define IEEE802154_CMD_GTS_REQ 0x09 - -/* - * The return values of MAC operations - */ -enum { - /* - * The requested operation was completed successfully. - * For a transmission request, this value indicates - * a successful transmission. - */ - IEEE802154_SUCCESS = 0x0, - - /* The beacon was lost following a synchronization request. */ - IEEE802154_BEACON_LOSS = 0xe0, - /* - * A transmission could not take place due to activity on the - * channel, i.e., the CSMA-CA mechanism has failed. - */ - IEEE802154_CHNL_ACCESS_FAIL = 0xe1, - /* The GTS request has been denied by the PAN coordinator. */ - IEEE802154_DENINED = 0xe2, - /* The attempt to disable the transceiver has failed. */ - IEEE802154_DISABLE_TRX_FAIL = 0xe3, - /* - * The received frame induces a failed security check according to - * the security suite. - */ - IEEE802154_FAILED_SECURITY_CHECK = 0xe4, - /* - * The frame resulting from secure processing has a length that is - * greater than aMACMaxFrameSize. - */ - IEEE802154_FRAME_TOO_LONG = 0xe5, - /* - * The requested GTS transmission failed because the specified GTS - * either did not have a transmit GTS direction or was not defined. - */ - IEEE802154_INVALID_GTS = 0xe6, - /* - * A request to purge an MSDU from the transaction queue was made using - * an MSDU handle that was not found in the transaction table. - */ - IEEE802154_INVALID_HANDLE = 0xe7, - /* A parameter in the primitive is out of the valid range.*/ - IEEE802154_INVALID_PARAMETER = 0xe8, - /* No acknowledgment was received after aMaxFrameRetries. */ - IEEE802154_NO_ACK = 0xe9, - /* A scan operation failed to find any network beacons.*/ - IEEE802154_NO_BEACON = 0xea, - /* No response data were available following a request. */ - IEEE802154_NO_DATA = 0xeb, - /* The operation failed because a short address was not allocated. */ - IEEE802154_NO_SHORT_ADDRESS = 0xec, - /* - * A receiver enable request was unsuccessful because it could not be - * completed within the CAP. - */ - IEEE802154_OUT_OF_CAP = 0xed, - /* - * A PAN identifier conflict has been detected and communicated to the - * PAN coordinator. - */ - IEEE802154_PANID_CONFLICT = 0xee, - /* A coordinator realignment command has been received. */ - IEEE802154_REALIGMENT = 0xef, - /* The transaction has expired and its information discarded. */ - IEEE802154_TRANSACTION_EXPIRED = 0xf0, - /* There is no capacity to store the transaction. */ - IEEE802154_TRANSACTION_OVERFLOW = 0xf1, - /* - * The transceiver was in the transmitter enabled state when the - * receiver was requested to be enabled. - */ - IEEE802154_TX_ACTIVE = 0xf2, - /* The appropriate key is not available in the ACL. */ - IEEE802154_UNAVAILABLE_KEY = 0xf3, - /* - * A SET/GET request was issued with the identifier of a PIB attribute - * that is not supported. - */ - IEEE802154_UNSUPPORTED_ATTR = 0xf4, - /* - * A request to perform a scan operation failed because the MLME was - * in the process of performing a previously initiated scan operation. - */ - IEEE802154_SCAN_IN_PROGRESS = 0xfc, -}; - - -#endif - - diff --git a/include/net/ieee802154/netdevice.h b/include/net/ieee802154/netdevice.h deleted file mode 100644 index e2506af3e7c8..000000000000 --- a/include/net/ieee802154/netdevice.h +++ /dev/null @@ -1,115 +0,0 @@ -/* - * An interface between IEEE802.15.4 device and rest of the kernel. - * - * Copyright (C) 2007, 2008, 2009 Siemens AG - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License version 2 - * as published by the Free Software Foundation. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License along - * with this program; if not, write to the Free Software Foundation, Inc., - * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA. - * - * Written by: - * Pavel Smolenskiy - * Maxim Gorbachyov - * Maxim Osipov - * Dmitry Eremin-Solenikov - */ - -#ifndef IEEE802154_NETDEVICE_H -#define IEEE802154_NETDEVICE_H - -/* - * A control block of skb passed between the ARPHRD_IEEE802154 device - * and other stack parts. - */ -struct ieee802154_mac_cb { - u8 lqi; - struct ieee802154_addr sa; - struct ieee802154_addr da; - u8 flags; - u8 seq; -}; - -static inline struct ieee802154_mac_cb *mac_cb(struct sk_buff *skb) -{ - return (struct ieee802154_mac_cb *)skb->cb; -} - -#define MAC_CB_FLAG_TYPEMASK ((1 << 3) - 1) - -#define MAC_CB_FLAG_ACKREQ (1 << 3) -#define MAC_CB_FLAG_SECEN (1 << 4) -#define MAC_CB_FLAG_INTRAPAN (1 << 5) - -static inline int mac_cb_is_ackreq(struct sk_buff *skb) -{ - return mac_cb(skb)->flags & MAC_CB_FLAG_ACKREQ; -} - -static inline int mac_cb_is_secen(struct sk_buff *skb) -{ - return mac_cb(skb)->flags & MAC_CB_FLAG_SECEN; -} - -static inline int mac_cb_is_intrapan(struct sk_buff *skb) -{ - return mac_cb(skb)->flags & MAC_CB_FLAG_INTRAPAN; -} - -static inline int mac_cb_type(struct sk_buff *skb) -{ - return mac_cb(skb)->flags & MAC_CB_FLAG_TYPEMASK; -} - -#define IEEE802154_MAC_SCAN_ED 0 -#define IEEE802154_MAC_SCAN_ACTIVE 1 -#define IEEE802154_MAC_SCAN_PASSIVE 2 -#define IEEE802154_MAC_SCAN_ORPHAN 3 - -/* - * This should be located at net_device->ml_priv - */ -struct ieee802154_mlme_ops { - int (*assoc_req)(struct net_device *dev, - struct ieee802154_addr *addr, - u8 channel, u8 cap); - int (*assoc_resp)(struct net_device *dev, - struct ieee802154_addr *addr, - u16 short_addr, u8 status); - int (*disassoc_req)(struct net_device *dev, - struct ieee802154_addr *addr, - u8 reason); - int (*start_req)(struct net_device *dev, - struct ieee802154_addr *addr, - u8 channel, u8 bcn_ord, u8 sf_ord, - u8 pan_coord, u8 blx, u8 coord_realign); - int (*scan_req)(struct net_device *dev, - u8 type, u32 channels, u8 duration); - - /* - * FIXME: these should become the part of PIB/MIB interface. - * However we still don't have IB interface of any kind - */ - u16 (*get_pan_id)(struct net_device *dev); - u16 (*get_short_addr)(struct net_device *dev); - u8 (*get_dsn)(struct net_device *dev); - u8 (*get_bsn)(struct net_device *dev); -}; - -static inline struct ieee802154_mlme_ops *ieee802154_mlme_ops( - struct net_device *dev) -{ - return dev->ml_priv; -} - -#endif - - diff --git a/include/net/ieee802154/nl802154.h b/include/net/ieee802154/nl802154.h deleted file mode 100644 index 6096096f6d7d..000000000000 --- a/include/net/ieee802154/nl802154.h +++ /dev/null @@ -1,117 +0,0 @@ -/* - * nl802154.h - * - * Copyright (C) 2007, 2008, 2009 Siemens AG - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License version 2 - * as published by the Free Software Foundation. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License along - * with this program; if not, write to the Free Software Foundation, Inc., - * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA. - * - */ - -#ifndef IEEE802154_NL_H -#define IEEE802154_NL_H - -struct net_device; -struct ieee802154_addr; - -/** - * ieee802154_nl_assoc_indic - Notify userland of an association request. - * @dev: The network device on which this association request was - * received. - * @addr: The address of the device requesting association. - * @cap: The capability information field from the device. - * - * This informs a userland coordinator of a device requesting to - * associate with the PAN controlled by the coordinator. - * - * Note: This is in section 7.3.1 of the IEEE 802.15.4-2006 document. - */ -int ieee802154_nl_assoc_indic(struct net_device *dev, - struct ieee802154_addr *addr, u8 cap); - -/** - * ieee802154_nl_assoc_confirm - Notify userland of association. - * @dev: The device which has completed association. - * @short_addr: The short address assigned to the device. - * @status: The status of the association. - * - * Inform userland of the result of an association request. If the - * association request included asking the coordinator to allocate - * a short address then it is returned in @short_addr. - * - * Note: This is in section 7.3.2 of the IEEE 802.15.4 document. - */ -int ieee802154_nl_assoc_confirm(struct net_device *dev, - u16 short_addr, u8 status); - -/** - * ieee802154_nl_disassoc_indic - Notify userland of disassociation. - * @dev: The device on which disassociation was indicated. - * @addr: The device which is disassociating. - * @reason: The reason for the disassociation. - * - * Inform userland that a device has disassociated from the network. - * - * Note: This is in section 7.3.3 of the IEEE 802.15.4 document. - */ -int ieee802154_nl_disassoc_indic(struct net_device *dev, - struct ieee802154_addr *addr, u8 reason); - -/** - * ieee802154_nl_disassoc_confirm - Notify userland of disassociation - * completion. - * @dev: The device on which disassociation was ordered. - * @status: The result of the disassociation. - * - * Inform userland of the result of requesting that a device - * disassociate, or the result of requesting that we disassociate from - * a PAN managed by another coordinator. - * - * Note: This is in section 7.1.4.3 of the IEEE 802.15.4 document. - */ -int ieee802154_nl_disassoc_confirm(struct net_device *dev, - u8 status); - -/** - * ieee802154_nl_scan_confirm - Notify userland of completion of scan. - * @dev: The device which was instructed to scan. - * @status: The status of the scan operation. - * @scan_type: What type of scan was performed. - * @unscanned: Any channels that the device was unable to scan. - * @edl: The energy levels (if a passive scan). - * - * - * Note: This is in section 7.1.11 of the IEEE 802.15.4 document. - * Note: This API does not permit the return of an active scan result. - */ -int ieee802154_nl_scan_confirm(struct net_device *dev, - u8 status, u8 scan_type, u32 unscanned, - u8 *edl/*, struct list_head *pan_desc_list */); - -/** - * ieee802154_nl_beacon_indic - Notify userland of a received beacon. - * @dev: The device on which a beacon was received. - * @panid: The PAN of the coordinator. - * @coord_addr: The short address of the coordinator on that PAN. - * - * Note: This is in section 7.1.5 of the IEEE 802.15.4 document. - * Note: This API does not provide extended information such as what - * channel the PAN is on or what the LQI of the beacon frame was on - * receipt. - * Note: This API cannot indicate a beacon frame for a coordinator - * operating in long addressing mode. - */ -int ieee802154_nl_beacon_indic(struct net_device *dev, u16 panid, - u16 coord_addr); - -#endif diff --git a/include/net/ieee802154_netdev.h b/include/net/ieee802154_netdev.h new file mode 100644 index 000000000000..e2506af3e7c8 --- /dev/null +++ b/include/net/ieee802154_netdev.h @@ -0,0 +1,115 @@ +/* + * An interface between IEEE802.15.4 device and rest of the kernel. + * + * Copyright (C) 2007, 2008, 2009 Siemens AG + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA. + * + * Written by: + * Pavel Smolenskiy + * Maxim Gorbachyov + * Maxim Osipov + * Dmitry Eremin-Solenikov + */ + +#ifndef IEEE802154_NETDEVICE_H +#define IEEE802154_NETDEVICE_H + +/* + * A control block of skb passed between the ARPHRD_IEEE802154 device + * and other stack parts. + */ +struct ieee802154_mac_cb { + u8 lqi; + struct ieee802154_addr sa; + struct ieee802154_addr da; + u8 flags; + u8 seq; +}; + +static inline struct ieee802154_mac_cb *mac_cb(struct sk_buff *skb) +{ + return (struct ieee802154_mac_cb *)skb->cb; +} + +#define MAC_CB_FLAG_TYPEMASK ((1 << 3) - 1) + +#define MAC_CB_FLAG_ACKREQ (1 << 3) +#define MAC_CB_FLAG_SECEN (1 << 4) +#define MAC_CB_FLAG_INTRAPAN (1 << 5) + +static inline int mac_cb_is_ackreq(struct sk_buff *skb) +{ + return mac_cb(skb)->flags & MAC_CB_FLAG_ACKREQ; +} + +static inline int mac_cb_is_secen(struct sk_buff *skb) +{ + return mac_cb(skb)->flags & MAC_CB_FLAG_SECEN; +} + +static inline int mac_cb_is_intrapan(struct sk_buff *skb) +{ + return mac_cb(skb)->flags & MAC_CB_FLAG_INTRAPAN; +} + +static inline int mac_cb_type(struct sk_buff *skb) +{ + return mac_cb(skb)->flags & MAC_CB_FLAG_TYPEMASK; +} + +#define IEEE802154_MAC_SCAN_ED 0 +#define IEEE802154_MAC_SCAN_ACTIVE 1 +#define IEEE802154_MAC_SCAN_PASSIVE 2 +#define IEEE802154_MAC_SCAN_ORPHAN 3 + +/* + * This should be located at net_device->ml_priv + */ +struct ieee802154_mlme_ops { + int (*assoc_req)(struct net_device *dev, + struct ieee802154_addr *addr, + u8 channel, u8 cap); + int (*assoc_resp)(struct net_device *dev, + struct ieee802154_addr *addr, + u16 short_addr, u8 status); + int (*disassoc_req)(struct net_device *dev, + struct ieee802154_addr *addr, + u8 reason); + int (*start_req)(struct net_device *dev, + struct ieee802154_addr *addr, + u8 channel, u8 bcn_ord, u8 sf_ord, + u8 pan_coord, u8 blx, u8 coord_realign); + int (*scan_req)(struct net_device *dev, + u8 type, u32 channels, u8 duration); + + /* + * FIXME: these should become the part of PIB/MIB interface. + * However we still don't have IB interface of any kind + */ + u16 (*get_pan_id)(struct net_device *dev); + u16 (*get_short_addr)(struct net_device *dev); + u8 (*get_dsn)(struct net_device *dev); + u8 (*get_bsn)(struct net_device *dev); +}; + +static inline struct ieee802154_mlme_ops *ieee802154_mlme_ops( + struct net_device *dev) +{ + return dev->ml_priv; +} + +#endif + + diff --git a/include/net/nl802154.h b/include/net/nl802154.h new file mode 100644 index 000000000000..6096096f6d7d --- /dev/null +++ b/include/net/nl802154.h @@ -0,0 +1,117 @@ +/* + * nl802154.h + * + * Copyright (C) 2007, 2008, 2009 Siemens AG + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA. + * + */ + +#ifndef IEEE802154_NL_H +#define IEEE802154_NL_H + +struct net_device; +struct ieee802154_addr; + +/** + * ieee802154_nl_assoc_indic - Notify userland of an association request. + * @dev: The network device on which this association request was + * received. + * @addr: The address of the device requesting association. + * @cap: The capability information field from the device. + * + * This informs a userland coordinator of a device requesting to + * associate with the PAN controlled by the coordinator. + * + * Note: This is in section 7.3.1 of the IEEE 802.15.4-2006 document. + */ +int ieee802154_nl_assoc_indic(struct net_device *dev, + struct ieee802154_addr *addr, u8 cap); + +/** + * ieee802154_nl_assoc_confirm - Notify userland of association. + * @dev: The device which has completed association. + * @short_addr: The short address assigned to the device. + * @status: The status of the association. + * + * Inform userland of the result of an association request. If the + * association request included asking the coordinator to allocate + * a short address then it is returned in @short_addr. + * + * Note: This is in section 7.3.2 of the IEEE 802.15.4 document. + */ +int ieee802154_nl_assoc_confirm(struct net_device *dev, + u16 short_addr, u8 status); + +/** + * ieee802154_nl_disassoc_indic - Notify userland of disassociation. + * @dev: The device on which disassociation was indicated. + * @addr: The device which is disassociating. + * @reason: The reason for the disassociation. + * + * Inform userland that a device has disassociated from the network. + * + * Note: This is in section 7.3.3 of the IEEE 802.15.4 document. + */ +int ieee802154_nl_disassoc_indic(struct net_device *dev, + struct ieee802154_addr *addr, u8 reason); + +/** + * ieee802154_nl_disassoc_confirm - Notify userland of disassociation + * completion. + * @dev: The device on which disassociation was ordered. + * @status: The result of the disassociation. + * + * Inform userland of the result of requesting that a device + * disassociate, or the result of requesting that we disassociate from + * a PAN managed by another coordinator. + * + * Note: This is in section 7.1.4.3 of the IEEE 802.15.4 document. + */ +int ieee802154_nl_disassoc_confirm(struct net_device *dev, + u8 status); + +/** + * ieee802154_nl_scan_confirm - Notify userland of completion of scan. + * @dev: The device which was instructed to scan. + * @status: The status of the scan operation. + * @scan_type: What type of scan was performed. + * @unscanned: Any channels that the device was unable to scan. + * @edl: The energy levels (if a passive scan). + * + * + * Note: This is in section 7.1.11 of the IEEE 802.15.4 document. + * Note: This API does not permit the return of an active scan result. + */ +int ieee802154_nl_scan_confirm(struct net_device *dev, + u8 status, u8 scan_type, u32 unscanned, + u8 *edl/*, struct list_head *pan_desc_list */); + +/** + * ieee802154_nl_beacon_indic - Notify userland of a received beacon. + * @dev: The device on which a beacon was received. + * @panid: The PAN of the coordinator. + * @coord_addr: The short address of the coordinator on that PAN. + * + * Note: This is in section 7.1.5 of the IEEE 802.15.4 document. + * Note: This API does not provide extended information such as what + * channel the PAN is on or what the LQI of the beacon frame was on + * receipt. + * Note: This API cannot indicate a beacon frame for a coordinator + * operating in long addressing mode. + */ +int ieee802154_nl_beacon_indic(struct net_device *dev, u16 panid, + u16 coord_addr); + +#endif diff --git a/net/ieee802154/af_ieee802154.c b/net/ieee802154/af_ieee802154.c index 3bb6bdb1dac1..69c8d9207aa7 100644 --- a/net/ieee802154/af_ieee802154.c +++ b/net/ieee802154/af_ieee802154.c @@ -34,8 +34,8 @@ #include #include -#include -#include +#include +#include #include "af802154.h" diff --git a/net/ieee802154/dgram.c b/net/ieee802154/dgram.c index 14d39840dd62..53dd912d52b4 100644 --- a/net/ieee802154/dgram.c +++ b/net/ieee802154/dgram.c @@ -26,9 +26,9 @@ #include #include #include -#include -#include -#include +#include +#include +#include #include diff --git a/net/ieee802154/netlink.c b/net/ieee802154/netlink.c index 27eda9fdf3c2..a615b9d13212 100644 --- a/net/ieee802154/netlink.c +++ b/net/ieee802154/netlink.c @@ -27,9 +27,9 @@ #include #include #include -#include -#include -#include +#include +#include +#include static unsigned int ieee802154_seq_num; diff --git a/net/ieee802154/raw.c b/net/ieee802154/raw.c index fca44d59f97e..ea8d1f15206e 100644 --- a/net/ieee802154/raw.c +++ b/net/ieee802154/raw.c @@ -26,7 +26,7 @@ #include #include #include -#include +#include #include "af802154.h" -- cgit v1.2.3 From 5dea271b6d87bd1d79a59c1d5baac2596a841c37 Mon Sep 17 00:00:00 2001 From: Mike Snitzer Date: Thu, 23 Jul 2009 20:30:42 +0100 Subject: dm table: pass correct dev area size to device_area_is_valid Incorrect device area lengths are being passed to device_area_is_valid(). The regression appeared in 2.6.31-rc1 through commit 754c5fc7ebb417b23601a6222a6005cc2e7f2913. With the dm-stripe target, the size of the target (ti->len) was used instead of the stripe_width (ti->len/#stripes). An example of a consequent incorrect error message is: device-mapper: table: 254:0: sdb too small for target Signed-off-by: Mike Snitzer Signed-off-by: Alasdair G Kergon --- drivers/md/dm-crypt.c | 2 +- drivers/md/dm-delay.c | 4 ++-- drivers/md/dm-linear.c | 2 +- drivers/md/dm-mpath.c | 2 +- drivers/md/dm-raid1.c | 2 +- drivers/md/dm-stripe.c | 7 ++++--- drivers/md/dm-table.c | 10 +++++----- include/linux/device-mapper.h | 4 ++-- 8 files changed, 17 insertions(+), 16 deletions(-) (limited to 'include') diff --git a/drivers/md/dm-crypt.c b/drivers/md/dm-crypt.c index 529e2ba505c3..ed1038164019 100644 --- a/drivers/md/dm-crypt.c +++ b/drivers/md/dm-crypt.c @@ -1318,7 +1318,7 @@ static int crypt_iterate_devices(struct dm_target *ti, { struct crypt_config *cc = ti->private; - return fn(ti, cc->dev, cc->start, data); + return fn(ti, cc->dev, cc->start, ti->len, data); } static struct target_type crypt_target = { diff --git a/drivers/md/dm-delay.c b/drivers/md/dm-delay.c index 4e5b843cd4d7..ebe7381f47c8 100644 --- a/drivers/md/dm-delay.c +++ b/drivers/md/dm-delay.c @@ -324,12 +324,12 @@ static int delay_iterate_devices(struct dm_target *ti, struct delay_c *dc = ti->private; int ret = 0; - ret = fn(ti, dc->dev_read, dc->start_read, data); + ret = fn(ti, dc->dev_read, dc->start_read, ti->len, data); if (ret) goto out; if (dc->dev_write) - ret = fn(ti, dc->dev_write, dc->start_write, data); + ret = fn(ti, dc->dev_write, dc->start_write, ti->len, data); out: return ret; diff --git a/drivers/md/dm-linear.c b/drivers/md/dm-linear.c index 9184b6deb868..82f7d6e6b1ea 100644 --- a/drivers/md/dm-linear.c +++ b/drivers/md/dm-linear.c @@ -139,7 +139,7 @@ static int linear_iterate_devices(struct dm_target *ti, { struct linear_c *lc = ti->private; - return fn(ti, lc->dev, lc->start, data); + return fn(ti, lc->dev, lc->start, ti->len, data); } static struct target_type linear_target = { diff --git a/drivers/md/dm-mpath.c b/drivers/md/dm-mpath.c index c70604a20897..6f0d90d4a541 100644 --- a/drivers/md/dm-mpath.c +++ b/drivers/md/dm-mpath.c @@ -1453,7 +1453,7 @@ static int multipath_iterate_devices(struct dm_target *ti, list_for_each_entry(pg, &m->priority_groups, list) { list_for_each_entry(p, &pg->pgpaths, list) { - ret = fn(ti, p->path.dev, ti->begin, data); + ret = fn(ti, p->path.dev, ti->begin, ti->len, data); if (ret) goto out; } diff --git a/drivers/md/dm-raid1.c b/drivers/md/dm-raid1.c index 42a3e4341d60..9726577cde49 100644 --- a/drivers/md/dm-raid1.c +++ b/drivers/md/dm-raid1.c @@ -1293,7 +1293,7 @@ static int mirror_iterate_devices(struct dm_target *ti, for (i = 0; !ret && i < ms->nr_mirrors; i++) ret = fn(ti, ms->mirror[i].dev, - ms->mirror[i].offset, data); + ms->mirror[i].offset, ti->len, data); return ret; } diff --git a/drivers/md/dm-stripe.c b/drivers/md/dm-stripe.c index b240e85ae39a..4e0e5937e42a 100644 --- a/drivers/md/dm-stripe.c +++ b/drivers/md/dm-stripe.c @@ -320,10 +320,11 @@ static int stripe_iterate_devices(struct dm_target *ti, int ret = 0; unsigned i = 0; - do + do { ret = fn(ti, sc->stripe[i].dev, - sc->stripe[i].physical_start, data); - while (!ret && ++i < sc->stripes); + sc->stripe[i].physical_start, + sc->stripe_width, data); + } while (!ret && ++i < sc->stripes); return ret; } diff --git a/drivers/md/dm-table.c b/drivers/md/dm-table.c index 5d16d06613aa..d952b3441913 100644 --- a/drivers/md/dm-table.c +++ b/drivers/md/dm-table.c @@ -346,7 +346,7 @@ static void close_dev(struct dm_dev_internal *d, struct mapped_device *md) * If possible, this checks an area of a destination device is valid. */ static int device_area_is_valid(struct dm_target *ti, struct dm_dev *dev, - sector_t start, void *data) + sector_t start, sector_t len, void *data) { struct queue_limits *limits = data; struct block_device *bdev = dev->bdev; @@ -359,7 +359,7 @@ static int device_area_is_valid(struct dm_target *ti, struct dm_dev *dev, if (!dev_size) return 1; - if ((start >= dev_size) || (start + ti->len > dev_size)) { + if ((start >= dev_size) || (start + len > dev_size)) { DMWARN("%s: %s too small for target", dm_device_name(ti->table->md), bdevname(bdev, b)); return 0; @@ -377,11 +377,11 @@ static int device_area_is_valid(struct dm_target *ti, struct dm_dev *dev, return 0; } - if (ti->len & (logical_block_size_sectors - 1)) { + if (len & (logical_block_size_sectors - 1)) { DMWARN("%s: len=%llu not aligned to h/w " "logical block size %hu of %s", dm_device_name(ti->table->md), - (unsigned long long)ti->len, + (unsigned long long)len, limits->logical_block_size, bdevname(bdev, b)); return 0; } @@ -482,7 +482,7 @@ static int __table_get_device(struct dm_table *t, struct dm_target *ti, #define min_not_zero(l, r) (l == 0) ? r : ((r == 0) ? l : min(l, r)) int dm_set_device_limits(struct dm_target *ti, struct dm_dev *dev, - sector_t start, void *data) + sector_t start, sector_t len, void *data) { struct queue_limits *limits = data; struct block_device *bdev = dev->bdev; diff --git a/include/linux/device-mapper.h b/include/linux/device-mapper.h index 0d6310657f32..655e7721580a 100644 --- a/include/linux/device-mapper.h +++ b/include/linux/device-mapper.h @@ -84,7 +84,7 @@ typedef int (*dm_merge_fn) (struct dm_target *ti, struct bvec_merge_data *bvm, typedef int (*iterate_devices_callout_fn) (struct dm_target *ti, struct dm_dev *dev, - sector_t physical_start, + sector_t start, sector_t len, void *data); typedef int (*dm_iterate_devices_fn) (struct dm_target *ti, @@ -104,7 +104,7 @@ void dm_error(const char *message); * Combine device limits. */ int dm_set_device_limits(struct dm_target *ti, struct dm_dev *dev, - sector_t start, void *data); + sector_t start, sector_t len, void *data); struct dm_dev { struct block_device *bdev; -- cgit v1.2.3 From d9ab77161d811ffb0bccf396f7155cc905c1b9e1 Mon Sep 17 00:00:00 2001 From: Dmitry Torokhov Date: Wed, 22 Jul 2009 00:37:25 +0200 Subject: Driver Core: Make PM operations a const pointer They are not supposed to be modified by drivers, so make them const. Signed-off-by: Dmitry Torokhov Acked-by: Greg Kroah-Hartman Signed-off-by: Rafael J. Wysocki --- drivers/base/platform.c | 2 +- drivers/base/power/main.c | 8 +++++--- drivers/pci/pci-driver.c | 16 ++++++++-------- include/linux/device.h | 9 +++++---- 4 files changed, 19 insertions(+), 16 deletions(-) (limited to 'include') diff --git a/drivers/base/platform.c b/drivers/base/platform.c index 455e55971d0e..ae5c4aa6d269 100644 --- a/drivers/base/platform.c +++ b/drivers/base/platform.c @@ -889,7 +889,7 @@ static int platform_pm_restore_noirq(struct device *dev) #endif /* !CONFIG_HIBERNATION */ -static struct dev_pm_ops platform_dev_pm_ops = { +static const struct dev_pm_ops platform_dev_pm_ops = { .prepare = platform_pm_prepare, .complete = platform_pm_complete, .suspend = platform_pm_suspend, diff --git a/drivers/base/power/main.c b/drivers/base/power/main.c index 58a3e572f2c9..1b1a786b7dec 100644 --- a/drivers/base/power/main.c +++ b/drivers/base/power/main.c @@ -157,8 +157,9 @@ void device_pm_move_last(struct device *dev) * @ops: PM operations to choose from. * @state: PM transition of the system being carried out. */ -static int pm_op(struct device *dev, struct dev_pm_ops *ops, - pm_message_t state) +static int pm_op(struct device *dev, + const struct dev_pm_ops *ops, + pm_message_t state) { int error = 0; @@ -220,7 +221,8 @@ static int pm_op(struct device *dev, struct dev_pm_ops *ops, * The operation is executed with interrupts disabled by the only remaining * functional CPU in the system. */ -static int pm_noirq_op(struct device *dev, struct dev_pm_ops *ops, +static int pm_noirq_op(struct device *dev, + const struct dev_pm_ops *ops, pm_message_t state) { int error = 0; diff --git a/drivers/pci/pci-driver.c b/drivers/pci/pci-driver.c index d76c4c85367e..0c2ea44ae5e1 100644 --- a/drivers/pci/pci-driver.c +++ b/drivers/pci/pci-driver.c @@ -575,7 +575,7 @@ static void pci_pm_complete(struct device *dev) static int pci_pm_suspend(struct device *dev) { struct pci_dev *pci_dev = to_pci_dev(dev); - struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL; + const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL; if (pci_has_legacy_pm_support(pci_dev)) return pci_legacy_suspend(dev, PMSG_SUSPEND); @@ -613,7 +613,7 @@ static int pci_pm_suspend(struct device *dev) static int pci_pm_suspend_noirq(struct device *dev) { struct pci_dev *pci_dev = to_pci_dev(dev); - struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL; + const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL; if (pci_has_legacy_pm_support(pci_dev)) return pci_legacy_suspend_late(dev, PMSG_SUSPEND); @@ -672,7 +672,7 @@ static int pci_pm_resume_noirq(struct device *dev) static int pci_pm_resume(struct device *dev) { struct pci_dev *pci_dev = to_pci_dev(dev); - struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL; + const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL; int error = 0; /* @@ -711,7 +711,7 @@ static int pci_pm_resume(struct device *dev) static int pci_pm_freeze(struct device *dev) { struct pci_dev *pci_dev = to_pci_dev(dev); - struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL; + const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL; if (pci_has_legacy_pm_support(pci_dev)) return pci_legacy_suspend(dev, PMSG_FREEZE); @@ -780,7 +780,7 @@ static int pci_pm_thaw_noirq(struct device *dev) static int pci_pm_thaw(struct device *dev) { struct pci_dev *pci_dev = to_pci_dev(dev); - struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL; + const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL; int error = 0; if (pci_has_legacy_pm_support(pci_dev)) @@ -799,7 +799,7 @@ static int pci_pm_thaw(struct device *dev) static int pci_pm_poweroff(struct device *dev) { struct pci_dev *pci_dev = to_pci_dev(dev); - struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL; + const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL; if (pci_has_legacy_pm_support(pci_dev)) return pci_legacy_suspend(dev, PMSG_HIBERNATE); @@ -872,7 +872,7 @@ static int pci_pm_restore_noirq(struct device *dev) static int pci_pm_restore(struct device *dev) { struct pci_dev *pci_dev = to_pci_dev(dev); - struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL; + const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL; int error = 0; /* @@ -910,7 +910,7 @@ static int pci_pm_restore(struct device *dev) #endif /* !CONFIG_HIBERNATION */ -struct dev_pm_ops pci_dev_pm_ops = { +const struct dev_pm_ops pci_dev_pm_ops = { .prepare = pci_pm_prepare, .complete = pci_pm_complete, .suspend = pci_pm_suspend, diff --git a/include/linux/device.h b/include/linux/device.h index aebb81036db2..a28642975053 100644 --- a/include/linux/device.h +++ b/include/linux/device.h @@ -62,7 +62,7 @@ struct bus_type { int (*suspend)(struct device *dev, pm_message_t state); int (*resume)(struct device *dev); - struct dev_pm_ops *pm; + const struct dev_pm_ops *pm; struct bus_type_private *p; }; @@ -132,7 +132,7 @@ struct device_driver { int (*resume) (struct device *dev); struct attribute_group **groups; - struct dev_pm_ops *pm; + const struct dev_pm_ops *pm; struct driver_private *p; }; @@ -200,7 +200,8 @@ struct class { int (*suspend)(struct device *dev, pm_message_t state); int (*resume)(struct device *dev); - struct dev_pm_ops *pm; + const struct dev_pm_ops *pm; + struct class_private *p; }; @@ -291,7 +292,7 @@ struct device_type { char *(*nodename)(struct device *dev); void (*release)(struct device *dev); - struct dev_pm_ops *pm; + const struct dev_pm_ops *pm; }; /* interface for exporting device attributes */ -- cgit v1.2.3 From c1dc13e9d0bc35a8d85bf4238c48c1b627d48f35 Mon Sep 17 00:00:00 2001 From: Rémi Denis-Courmont Date: Tue, 21 Jul 2009 01:57:57 +0000 Subject: Phonet: sockets list through proc_fs MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This provides a list of sockets with their Phonet bind addresses and some socket debug informations through /proc/net/phonet. Signed-off-by: Rémi Denis-Courmont Signed-off-by: David S. Miller --- include/net/phonet/pn_dev.h | 2 + net/phonet/pn_dev.c | 7 ++++ net/phonet/socket.c | 96 +++++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 105 insertions(+) (limited to 'include') diff --git a/include/net/phonet/pn_dev.h b/include/net/phonet/pn_dev.h index 29d126736611..44c923c9e21d 100644 --- a/include/net/phonet/pn_dev.h +++ b/include/net/phonet/pn_dev.h @@ -49,4 +49,6 @@ void phonet_address_notify(int event, struct net_device *dev, u8 addr); #define PN_NO_ADDR 0xff +extern const struct file_operations pn_sock_seq_fops; + #endif diff --git a/net/phonet/pn_dev.c b/net/phonet/pn_dev.c index b0d6ddd82a9d..5107b7949c9c 100644 --- a/net/phonet/pn_dev.c +++ b/net/phonet/pn_dev.c @@ -218,6 +218,11 @@ static int phonet_init_net(struct net *net) if (!pnn) return -ENOMEM; + if (!proc_net_fops_create(net, "phonet", 0, &pn_sock_seq_fops)) { + kfree(pnn); + return -ENOMEM; + } + INIT_LIST_HEAD(&pnn->pndevs.list); spin_lock_init(&pnn->pndevs.lock); net_assign_generic(net, phonet_net_id, pnn); @@ -233,6 +238,8 @@ static void phonet_exit_net(struct net *net) for_each_netdev(net, dev) phonet_device_destroy(dev); rtnl_unlock(); + + proc_net_remove(net, "phonet"); kfree(pnn); } diff --git a/net/phonet/socket.c b/net/phonet/socket.c index ada2a35bf7a2..aa1617a7f265 100644 --- a/net/phonet/socket.c +++ b/net/phonet/socket.c @@ -412,3 +412,99 @@ found: return 0; } EXPORT_SYMBOL(pn_sock_get_port); + +static struct sock *pn_sock_get_idx(struct seq_file *seq, loff_t pos) +{ + struct net *net = seq_file_net(seq); + struct hlist_node *node; + struct sock *sknode; + + sk_for_each(sknode, node, &pnsocks.hlist) { + if (!net_eq(net, sock_net(sknode))) + continue; + if (!pos) + return sknode; + pos--; + } + return NULL; +} + +static struct sock *pn_sock_get_next(struct seq_file *seq, struct sock *sk) +{ + struct net *net = seq_file_net(seq); + + do + sk = sk_next(sk); + while (sk && !net_eq(net, sock_net(sk))); + + return sk; +} + +static void *pn_sock_seq_start(struct seq_file *seq, loff_t *pos) + __acquires(pnsocks.lock) +{ + spin_lock_bh(&pnsocks.lock); + return *pos ? pn_sock_get_idx(seq, *pos - 1) : SEQ_START_TOKEN; +} + +static void *pn_sock_seq_next(struct seq_file *seq, void *v, loff_t *pos) +{ + struct sock *sk; + + if (v == SEQ_START_TOKEN) + sk = pn_sock_get_idx(seq, 0); + else + sk = pn_sock_get_next(seq, v); + (*pos)++; + return sk; +} + +static void pn_sock_seq_stop(struct seq_file *seq, void *v) + __releases(pnsocks.lock) +{ + spin_unlock_bh(&pnsocks.lock); +} + +static int pn_sock_seq_show(struct seq_file *seq, void *v) +{ + int len; + + if (v == SEQ_START_TOKEN) + seq_printf(seq, "%s%n", "pt loc rem rs st tx_queue rx_queue " + " uid inode ref pointer drops", &len); + else { + struct sock *sk = v; + struct pn_sock *pn = pn_sk(sk); + + seq_printf(seq, "%2d %04X:%04X:%02X %02X %08X:%08X %5d %lu " + "%d %p %d%n", + sk->sk_protocol, pn->sobject, 0, pn->resource, + sk->sk_state, + sk_wmem_alloc_get(sk), sk_rmem_alloc_get(sk), + sock_i_uid(sk), sock_i_ino(sk), + atomic_read(&sk->sk_refcnt), sk, + atomic_read(&sk->sk_drops), &len); + } + seq_printf(seq, "%*s\n", 127 - len, ""); + return 0; +} + +static const struct seq_operations pn_sock_seq_ops = { + .start = pn_sock_seq_start, + .next = pn_sock_seq_next, + .stop = pn_sock_seq_stop, + .show = pn_sock_seq_show, +}; + +static int pn_sock_open(struct inode *inode, struct file *file) +{ + return seq_open(file, &pn_sock_seq_ops); +} + +const struct file_operations pn_sock_seq_fops = { + .owner = THIS_MODULE, + .open = pn_sock_open, + .read = seq_read, + .llseek = seq_lseek, + .release = seq_release, +}; -- cgit v1.2.3 From ab300465676b0c0559af62d57ec9a902f5680b03 Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Fri, 24 Jul 2009 15:26:15 +0800 Subject: crypto: api - Fix aligned ctx helper The aligned ctx helper was using a bogus alignment value thas was one off the correct value. Fortunately the current users do not require anything beyond the natural alignment of the platform so this hasn't caused a problem. This patch fixes that and also removes the unnecessary minimum check since if the alignment is less than the natural alignment then the subsequent ALIGN operation should be a noop. Signed-off-by: Herbert Xu --- include/crypto/algapi.h | 8 ++------ 1 file changed, 2 insertions(+), 6 deletions(-) (limited to 'include') diff --git a/include/crypto/algapi.h b/include/crypto/algapi.h index 9de6c38f4069..3f388146a914 100644 --- a/include/crypto/algapi.h +++ b/include/crypto/algapi.h @@ -175,12 +175,8 @@ int blkcipher_walk_virt_block(struct blkcipher_desc *desc, static inline void *crypto_tfm_ctx_aligned(struct crypto_tfm *tfm) { - unsigned long addr = (unsigned long)crypto_tfm_ctx(tfm); - unsigned long align = crypto_tfm_alg_alignmask(tfm); - - if (align <= crypto_tfm_ctx_alignment()) - align = 1; - return (void *)ALIGN(addr, align); + return PTR_ALIGN(crypto_tfm_ctx(tfm), + crypto_tfm_alg_alignmask(tfm) + 1); } static inline struct crypto_instance *crypto_tfm_alg_instance( -- cgit v1.2.3 From c94aa5ca3088018d2a7a9bd3258aefffe29df265 Mon Sep 17 00:00:00 2001 From: Ming Lei Date: Thu, 16 Jul 2009 15:44:29 +0200 Subject: lockdep: Print the shortest dependency chain if finding a circle Currently lockdep will print the 1st circle detected if it exists when acquiring a new (next) lock. This patch prints the shortest path from the next lock to be acquired to the previous held lock if a circle is found. The patch still uses the current method to check circle, and once the circle is found, breadth-first search algorithem is used to compute the shortest path from the next lock to the previous lock in the forward lock dependency graph. Printing the shortest path will shorten the dependency chain, and make troubleshooting for possible circular locking easier. Signed-off-by: Ming Lei Signed-off-by: Peter Zijlstra LKML-Reference: <1246201486-7308-2-git-send-email-tom.leiming@gmail.com> Signed-off-by: Ingo Molnar --- include/linux/lockdep.h | 6 +++ kernel/lockdep.c | 115 +++++++++++++++++++++++++++++++++++++++++---- kernel/lockdep_internals.h | 83 ++++++++++++++++++++++++++++++++ 3 files changed, 195 insertions(+), 9 deletions(-) (limited to 'include') diff --git a/include/linux/lockdep.h b/include/linux/lockdep.h index b25d1b53df0d..9ec026f8d09e 100644 --- a/include/linux/lockdep.h +++ b/include/linux/lockdep.h @@ -149,6 +149,12 @@ struct lock_list { struct lock_class *class; struct stack_trace trace; int distance; + + /*The parent field is used to implement breadth-first search,and + *the bit 0 is reused to indicate if the lock has been accessed + *in BFS. + */ + struct lock_list *parent; }; /* diff --git a/kernel/lockdep.c b/kernel/lockdep.c index 8bbeef996c76..93dc70d18cdf 100644 --- a/kernel/lockdep.c +++ b/kernel/lockdep.c @@ -897,6 +897,79 @@ static int add_lock_to_list(struct lock_class *class, struct lock_class *this, return 1; } +static struct circular_queue lock_cq; +static int __search_shortest_path(struct lock_list *source_entry, + struct lock_class *target, + struct lock_list **target_entry, + int forward) +{ + struct lock_list *entry; + struct circular_queue *cq = &lock_cq; + int ret = 1; + + __cq_init(cq); + + mark_lock_accessed(source_entry, NULL); + if (source_entry->class == target) { + *target_entry = source_entry; + ret = 0; + goto exit; + } + + __cq_enqueue(cq, (unsigned long)source_entry); + + while (!__cq_empty(cq)) { + struct lock_list *lock; + struct list_head *head; + + __cq_dequeue(cq, (unsigned long *)&lock); + + if (!lock->class) { + ret = -2; + goto exit; + } + + if (forward) + head = &lock->class->locks_after; + else + head = &lock->class->locks_before; + + list_for_each_entry(entry, head, entry) { + if (!lock_accessed(entry)) { + mark_lock_accessed(entry, lock); + if (entry->class == target) { + *target_entry = entry; + ret = 0; + goto exit; + } + + if (__cq_enqueue(cq, (unsigned long)entry)) { + ret = -1; + goto exit; + } + } + } + } +exit: + return ret; +} + +static inline int __search_forward_shortest_path(struct lock_list *src_entry, + struct lock_class *target, + struct lock_list **target_entry) +{ + return __search_shortest_path(src_entry, target, target_entry, 1); + +} + +static inline int __search_backward_shortest_path(struct lock_list *src_entry, + struct lock_class *target, + struct lock_list **target_entry) +{ + return __search_shortest_path(src_entry, target, target_entry, 0); + +} + /* * Recursive, forwards-direction lock-dependency checking, used for * both noncyclic checking and for hardirq-unsafe/softirq-unsafe @@ -934,7 +1007,7 @@ print_circular_bug_header(struct lock_list *entry, unsigned int depth) { struct task_struct *curr = current; - if (!debug_locks_off_graph_unlock() || debug_locks_silent) + if (debug_locks_silent) return 0; printk("\n=======================================================\n"); @@ -954,19 +1027,41 @@ print_circular_bug_header(struct lock_list *entry, unsigned int depth) return 0; } -static noinline int print_circular_bug_tail(void) +static noinline int print_circular_bug(void) { struct task_struct *curr = current; struct lock_list this; + struct lock_list *target; + struct lock_list *parent; + int result; + unsigned long depth; - if (debug_locks_silent) + if (!debug_locks_off_graph_unlock() || debug_locks_silent) return 0; this.class = hlock_class(check_source); if (!save_trace(&this.trace)) return 0; - print_circular_bug_entry(&this, 0); + result = __search_forward_shortest_path(&this, + hlock_class(check_target), + &target); + if (result) { + printk("\n%s:search shortest path failed:%d\n", __func__, + result); + return 0; + } + + depth = get_lock_depth(target); + + print_circular_bug_header(target, depth); + + parent = get_lock_parent(target); + + while (parent) { + print_circular_bug_entry(parent, --depth); + parent = get_lock_parent(parent); + } printk("\nother info that might help us debug this:\n\n"); lockdep_print_held_locks(curr); @@ -1072,14 +1167,15 @@ check_noncircular(struct lock_class *source, unsigned int depth) */ list_for_each_entry(entry, &source->locks_after, entry) { if (entry->class == hlock_class(check_target)) - return print_circular_bug_header(entry, depth+1); + return 2; debug_atomic_inc(&nr_cyclic_checks); - if (!check_noncircular(entry->class, depth+1)) - return print_circular_bug_entry(entry, depth+1); + if (check_noncircular(entry->class, depth+1) == 2) + return 2; } return 1; } + #if defined(CONFIG_TRACE_IRQFLAGS) && defined(CONFIG_PROVE_LOCKING) /* * Forwards and backwards subgraph searching, for the purposes of @@ -1484,8 +1580,9 @@ check_prev_add(struct task_struct *curr, struct held_lock *prev, */ check_source = next; check_target = prev; - if (!(check_noncircular(hlock_class(next), 0))) - return print_circular_bug_tail(); + if (check_noncircular(hlock_class(next), 0) == 2) + return print_circular_bug(); + if (!check_prev_add_irq(curr, prev, next)) return 0; diff --git a/kernel/lockdep_internals.h b/kernel/lockdep_internals.h index 699a2ac3a0d7..6f48d37d5be2 100644 --- a/kernel/lockdep_internals.h +++ b/kernel/lockdep_internals.h @@ -136,3 +136,86 @@ extern atomic_t nr_find_usage_backwards_recursions; # define debug_atomic_dec(ptr) do { } while (0) # define debug_atomic_read(ptr) 0 #endif + +/* The circular_queue and helpers is used to implement the + * breadth-first search(BFS)algorithem, by which we can build + * the shortest path from the next lock to be acquired to the + * previous held lock if there is a circular between them. + * */ +#define MAX_CIRCULAR_QUE_SIZE 4096UL +struct circular_queue{ + unsigned long element[MAX_CIRCULAR_QUE_SIZE]; + unsigned int front, rear; +}; + +#define LOCK_ACCESSED 1UL +#define LOCK_ACCESSED_MASK (~LOCK_ACCESSED) + +static inline void __cq_init(struct circular_queue *cq) +{ + cq->front = cq->rear = 0; +} + +static inline int __cq_empty(struct circular_queue *cq) +{ + return (cq->front == cq->rear); +} + +static inline int __cq_full(struct circular_queue *cq) +{ + return ((cq->rear + 1)%MAX_CIRCULAR_QUE_SIZE) == cq->front; +} + +static inline int __cq_enqueue(struct circular_queue *cq, unsigned long elem) +{ + if (__cq_full(cq)) + return -1; + + cq->element[cq->rear] = elem; + cq->rear = (cq->rear + 1)%MAX_CIRCULAR_QUE_SIZE; + return 0; +} + +static inline int __cq_dequeue(struct circular_queue *cq, unsigned long *elem) +{ + if (__cq_empty(cq)) + return -1; + + *elem = cq->element[cq->front]; + cq->front = (cq->front + 1)%MAX_CIRCULAR_QUE_SIZE; + return 0; +} + +static inline int __cq_get_elem_count(struct circular_queue *cq) +{ + return (cq->rear - cq->front)%MAX_CIRCULAR_QUE_SIZE; +} + +static inline void mark_lock_accessed(struct lock_list *lock, + struct lock_list *parent) +{ + lock->parent = (void *) parent + LOCK_ACCESSED; +} + +static inline unsigned long lock_accessed(struct lock_list *lock) +{ + return (unsigned long)lock->parent & LOCK_ACCESSED; +} + +static inline struct lock_list *get_lock_parent(struct lock_list *child) +{ + return (struct lock_list *) + ((unsigned long)child->parent & LOCK_ACCESSED_MASK); +} + +static inline unsigned long get_lock_depth(struct lock_list *child) +{ + unsigned long depth = 0; + struct lock_list *parent; + + while ((parent = get_lock_parent(child))) { + child = parent; + depth++; + } + return depth; +} -- cgit v1.2.3 From af012961450949ea297b209e091bd1a3805b8a0a Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Thu, 16 Jul 2009 15:44:29 +0200 Subject: lockdep: BFS cleanup Some cleanups of the lockdep code after the BFS series: - Remove the last traces of the generation id - Fixup comment style - Move the bfs routines into lockdep.c - Cleanup the bfs routines [ tom.leiming@gmail.com: Fix crash ] Signed-off-by: Peter Zijlstra LKML-Reference: <1246201486-7308-11-git-send-email-tom.leiming@gmail.com> Signed-off-by: Ingo Molnar --- include/linux/lockdep.h | 7 +- kernel/lockdep.c | 284 +++++++++++++++++++++++++++------------------ kernel/lockdep_internals.h | 97 +--------------- 3 files changed, 175 insertions(+), 213 deletions(-) (limited to 'include') diff --git a/include/linux/lockdep.h b/include/linux/lockdep.h index 9ec026f8d09e..12aabfcb45f6 100644 --- a/include/linux/lockdep.h +++ b/include/linux/lockdep.h @@ -58,7 +58,6 @@ struct lock_class { struct lockdep_subclass_key *key; unsigned int subclass; - unsigned int dep_gen_id; /* * IRQ/softirq usage tracking bits: @@ -150,9 +149,9 @@ struct lock_list { struct stack_trace trace; int distance; - /*The parent field is used to implement breadth-first search,and - *the bit 0 is reused to indicate if the lock has been accessed - *in BFS. + /* + * The parent field is used to implement breadth-first search, and the + * bit 0 is reused to indicate if the lock has been accessed in BFS. */ struct lock_list *parent; }; diff --git a/kernel/lockdep.c b/kernel/lockdep.c index 744da6265d99..1cedb00e3e7a 100644 --- a/kernel/lockdep.c +++ b/kernel/lockdep.c @@ -43,6 +43,7 @@ #include #include #include + #include #include "lockdep_internals.h" @@ -118,7 +119,7 @@ static inline int debug_locks_off_graph_unlock(void) static int lockdep_initialized; unsigned long nr_list_entries; -struct lock_list list_entries[MAX_LOCKDEP_ENTRIES]; +static struct lock_list list_entries[MAX_LOCKDEP_ENTRIES]; /* * All data structures here are protected by the global debug_lock. @@ -390,19 +391,6 @@ unsigned int nr_process_chains; unsigned int max_lockdep_depth; unsigned int max_recursion_depth; -static unsigned int lockdep_dependency_gen_id; - -static bool lockdep_dependency_visit(struct lock_class *source, - unsigned int depth) -{ - if (!depth) - lockdep_dependency_gen_id++; - if (source->dep_gen_id == lockdep_dependency_gen_id) - return true; - source->dep_gen_id = lockdep_dependency_gen_id; - return false; -} - #ifdef CONFIG_DEBUG_LOCKDEP /* * We cannot printk in early bootup code. Not even early_printk() @@ -551,88 +539,6 @@ static void lockdep_print_held_locks(struct task_struct *curr) } } -static void print_lock_class_header(struct lock_class *class, int depth) -{ - int bit; - - printk("%*s->", depth, ""); - print_lock_name(class); - printk(" ops: %lu", class->ops); - printk(" {\n"); - - for (bit = 0; bit < LOCK_USAGE_STATES; bit++) { - if (class->usage_mask & (1 << bit)) { - int len = depth; - - len += printk("%*s %s", depth, "", usage_str[bit]); - len += printk(" at:\n"); - print_stack_trace(class->usage_traces + bit, len); - } - } - printk("%*s }\n", depth, ""); - - printk("%*s ... key at: ",depth,""); - print_ip_sym((unsigned long)class->key); -} - -/* - * printk the shortest lock dependencies from @start to @end in reverse order: - */ -static void __used -print_shortest_lock_dependencies(struct lock_list *leaf, - struct lock_list *root) -{ - struct lock_list *entry = leaf; - int depth; - - /*compute depth from generated tree by BFS*/ - depth = get_lock_depth(leaf); - - do { - print_lock_class_header(entry->class, depth); - printk("%*s ... acquired at:\n", depth, ""); - print_stack_trace(&entry->trace, 2); - printk("\n"); - - if (depth == 0 && (entry != root)) { - printk("lockdep:%s bad BFS generated tree\n", __func__); - break; - } - - entry = get_lock_parent(entry); - depth--; - } while (entry && (depth >= 0)); - - return; -} -/* - * printk all lock dependencies starting at : - */ -static void __used -print_lock_dependencies(struct lock_class *class, int depth) -{ - struct lock_list *entry; - - if (lockdep_dependency_visit(class, depth)) - return; - - if (DEBUG_LOCKS_WARN_ON(depth >= 20)) - return; - - print_lock_class_header(class, depth); - - list_for_each_entry(entry, &class->locks_after, entry) { - if (DEBUG_LOCKS_WARN_ON(!entry->class)) - return; - - print_lock_dependencies(entry->class, depth + 1); - - printk("%*s ... acquired at:\n",depth,""); - print_stack_trace(&entry->trace, 2); - printk("\n"); - } -} - static void print_kernel_version(void) { printk("%s %.*s\n", init_utsname()->release, @@ -927,14 +833,106 @@ static int add_lock_to_list(struct lock_class *class, struct lock_class *this, return 1; } -unsigned long bfs_accessed[BITS_TO_LONGS(MAX_LOCKDEP_ENTRIES)]; -static struct circular_queue lock_cq; +/*For good efficiency of modular, we use power of 2*/ +#define MAX_CIRCULAR_QUEUE_SIZE 4096UL +#define CQ_MASK (MAX_CIRCULAR_QUEUE_SIZE-1) + +/* The circular_queue and helpers is used to implement the + * breadth-first search(BFS)algorithem, by which we can build + * the shortest path from the next lock to be acquired to the + * previous held lock if there is a circular between them. + * */ +struct circular_queue { + unsigned long element[MAX_CIRCULAR_QUEUE_SIZE]; + unsigned int front, rear; +}; + +static struct circular_queue lock_cq; +static unsigned long bfs_accessed[BITS_TO_LONGS(MAX_LOCKDEP_ENTRIES)]; + unsigned int max_bfs_queue_depth; + +static inline void __cq_init(struct circular_queue *cq) +{ + cq->front = cq->rear = 0; + bitmap_zero(bfs_accessed, MAX_LOCKDEP_ENTRIES); +} + +static inline int __cq_empty(struct circular_queue *cq) +{ + return (cq->front == cq->rear); +} + +static inline int __cq_full(struct circular_queue *cq) +{ + return ((cq->rear + 1) & CQ_MASK) == cq->front; +} + +static inline int __cq_enqueue(struct circular_queue *cq, unsigned long elem) +{ + if (__cq_full(cq)) + return -1; + + cq->element[cq->rear] = elem; + cq->rear = (cq->rear + 1) & CQ_MASK; + return 0; +} + +static inline int __cq_dequeue(struct circular_queue *cq, unsigned long *elem) +{ + if (__cq_empty(cq)) + return -1; + + *elem = cq->element[cq->front]; + cq->front = (cq->front + 1) & CQ_MASK; + return 0; +} + +static inline unsigned int __cq_get_elem_count(struct circular_queue *cq) +{ + return (cq->rear - cq->front) & CQ_MASK; +} + +static inline void mark_lock_accessed(struct lock_list *lock, + struct lock_list *parent) +{ + unsigned long nr; + nr = lock - list_entries; + WARN_ON(nr >= nr_list_entries); + lock->parent = parent; + set_bit(nr, bfs_accessed); +} + +static inline unsigned long lock_accessed(struct lock_list *lock) +{ + unsigned long nr; + nr = lock - list_entries; + WARN_ON(nr >= nr_list_entries); + return test_bit(nr, bfs_accessed); +} + +static inline struct lock_list *get_lock_parent(struct lock_list *child) +{ + return child->parent; +} + +static inline int get_lock_depth(struct lock_list *child) +{ + int depth = 0; + struct lock_list *parent; + + while ((parent = get_lock_parent(child))) { + child = parent; + depth++; + } + return depth; +} + static int __bfs(struct lock_list *source_entry, - void *data, - int (*match)(struct lock_list *entry, void *data), - struct lock_list **target_entry, - int forward) + void *data, + int (*match)(struct lock_list *entry, void *data), + struct lock_list **target_entry, + int forward) { struct lock_list *entry; struct list_head *head; @@ -1202,14 +1200,6 @@ check_noncircular(struct lock_list *root, struct lock_class *target, * without creating any illegal irq-safe -> irq-unsafe lock dependency. */ - -#define BFS_PROCESS_RET(ret) do { \ - if (ret < 0) \ - return print_bfs_bug(ret); \ - if (ret == 1) \ - return 1; \ - } while (0) - static inline int usage_match(struct lock_list *entry, void *bit) { return entry->class->usage_mask & (1 << (enum lock_usage_bit)bit); @@ -1263,6 +1253,60 @@ find_usage_backwards(struct lock_list *root, enum lock_usage_bit bit, return result; } +static void print_lock_class_header(struct lock_class *class, int depth) +{ + int bit; + + printk("%*s->", depth, ""); + print_lock_name(class); + printk(" ops: %lu", class->ops); + printk(" {\n"); + + for (bit = 0; bit < LOCK_USAGE_STATES; bit++) { + if (class->usage_mask & (1 << bit)) { + int len = depth; + + len += printk("%*s %s", depth, "", usage_str[bit]); + len += printk(" at:\n"); + print_stack_trace(class->usage_traces + bit, len); + } + } + printk("%*s }\n", depth, ""); + + printk("%*s ... key at: ",depth,""); + print_ip_sym((unsigned long)class->key); +} + +/* + * printk the shortest lock dependencies from @start to @end in reverse order: + */ +static void __used +print_shortest_lock_dependencies(struct lock_list *leaf, + struct lock_list *root) +{ + struct lock_list *entry = leaf; + int depth; + + /*compute depth from generated tree by BFS*/ + depth = get_lock_depth(leaf); + + do { + print_lock_class_header(entry->class, depth); + printk("%*s ... acquired at:\n", depth, ""); + print_stack_trace(&entry->trace, 2); + printk("\n"); + + if (depth == 0 && (entry != root)) { + printk("lockdep:%s bad BFS generated tree\n", __func__); + break; + } + + entry = get_lock_parent(entry); + depth--; + } while (entry && (depth >= 0)); + + return; +} static int print_bad_irq_dependency(struct task_struct *curr, @@ -1349,12 +1393,18 @@ check_usage(struct task_struct *curr, struct held_lock *prev, this.class = hlock_class(prev); ret = find_usage_backwards(&this, bit_backwards, &target_entry); - BFS_PROCESS_RET(ret); + if (ret < 0) + return print_bfs_bug(ret); + if (ret == 1) + return ret; that.parent = NULL; that.class = hlock_class(next); ret = find_usage_forwards(&that, bit_forwards, &target_entry1); - BFS_PROCESS_RET(ret); + if (ret < 0) + return print_bfs_bug(ret); + if (ret == 1) + return ret; return print_bad_irq_dependency(curr, &this, &that, target_entry, target_entry1, @@ -2038,7 +2088,10 @@ check_usage_forwards(struct task_struct *curr, struct held_lock *this, root.parent = NULL; root.class = hlock_class(this); ret = find_usage_forwards(&root, bit, &target_entry); - BFS_PROCESS_RET(ret); + if (ret < 0) + return print_bfs_bug(ret); + if (ret == 1) + return ret; return print_irq_inversion_bug(curr, &root, target_entry, this, 1, irqclass); @@ -2059,7 +2112,10 @@ check_usage_backwards(struct task_struct *curr, struct held_lock *this, root.parent = NULL; root.class = hlock_class(this); ret = find_usage_backwards(&root, bit, &target_entry); - BFS_PROCESS_RET(ret); + if (ret < 0) + return print_bfs_bug(ret); + if (ret == 1) + return ret; return print_irq_inversion_bug(curr, &root, target_entry, this, 1, irqclass); diff --git a/kernel/lockdep_internals.h b/kernel/lockdep_internals.h index 6baa8807efdd..a2ee95ad1313 100644 --- a/kernel/lockdep_internals.h +++ b/kernel/lockdep_internals.h @@ -91,6 +91,8 @@ extern unsigned int nr_process_chains; extern unsigned int max_lockdep_depth; extern unsigned int max_recursion_depth; +extern unsigned int max_bfs_queue_depth; + #ifdef CONFIG_PROVE_LOCKING extern unsigned long lockdep_count_forward_deps(struct lock_class *); extern unsigned long lockdep_count_backward_deps(struct lock_class *); @@ -136,98 +138,3 @@ extern atomic_t nr_find_usage_backwards_recursions; # define debug_atomic_dec(ptr) do { } while (0) # define debug_atomic_read(ptr) 0 #endif - - -extern unsigned int max_bfs_queue_depth; -extern unsigned long nr_list_entries; -extern struct lock_list list_entries[MAX_LOCKDEP_ENTRIES]; -extern unsigned long bfs_accessed[]; - -/*For good efficiency of modular, we use power of 2*/ -#define MAX_CIRCULAR_QUE_SIZE 4096UL - -/* The circular_queue and helpers is used to implement the - * breadth-first search(BFS)algorithem, by which we can build - * the shortest path from the next lock to be acquired to the - * previous held lock if there is a circular between them. - * */ -struct circular_queue{ - unsigned long element[MAX_CIRCULAR_QUE_SIZE]; - unsigned int front, rear; -}; - -static inline void __cq_init(struct circular_queue *cq) -{ - cq->front = cq->rear = 0; - bitmap_zero(bfs_accessed, MAX_LOCKDEP_ENTRIES); -} - -static inline int __cq_empty(struct circular_queue *cq) -{ - return (cq->front == cq->rear); -} - -static inline int __cq_full(struct circular_queue *cq) -{ - return ((cq->rear + 1)&(MAX_CIRCULAR_QUE_SIZE-1)) == cq->front; -} - -static inline int __cq_enqueue(struct circular_queue *cq, unsigned long elem) -{ - if (__cq_full(cq)) - return -1; - - cq->element[cq->rear] = elem; - cq->rear = (cq->rear + 1)&(MAX_CIRCULAR_QUE_SIZE-1); - return 0; -} - -static inline int __cq_dequeue(struct circular_queue *cq, unsigned long *elem) -{ - if (__cq_empty(cq)) - return -1; - - *elem = cq->element[cq->front]; - cq->front = (cq->front + 1)&(MAX_CIRCULAR_QUE_SIZE-1); - return 0; -} - -static inline unsigned int __cq_get_elem_count(struct circular_queue *cq) -{ - return (cq->rear - cq->front)&(MAX_CIRCULAR_QUE_SIZE-1); -} - -static inline void mark_lock_accessed(struct lock_list *lock, - struct lock_list *parent) -{ - unsigned long nr; - nr = lock - list_entries; - WARN_ON(nr >= nr_list_entries); - lock->parent = parent; - set_bit(nr, bfs_accessed); -} - -static inline unsigned long lock_accessed(struct lock_list *lock) -{ - unsigned long nr; - nr = lock - list_entries; - WARN_ON(nr >= nr_list_entries); - return test_bit(nr, bfs_accessed); -} - -static inline struct lock_list *get_lock_parent(struct lock_list *child) -{ - return child->parent; -} - -static inline int get_lock_depth(struct lock_list *child) -{ - int depth = 0; - struct lock_list *parent; - - while ((parent = get_lock_parent(child))) { - child = parent; - depth++; - } - return depth; -} -- cgit v1.2.3 From 26e744b6b61066203fd57de0d3962353621e06f8 Mon Sep 17 00:00:00 2001 From: Brian Johnson Date: Sun, 19 Jul 2009 05:52:58 -0300 Subject: V4L/DVB (12283): gspca - sn9c20x: New subdriver for sn9c201 and sn9c202 bridges. Signed-off-by: Brian Johnson Signed-off-by: Jean-Francois Moine Signed-off-by: Mauro Carvalho Chehab --- Documentation/video4linux/gspca.txt | 32 + drivers/media/video/gspca/Kconfig | 16 + drivers/media/video/gspca/Makefile | 2 + drivers/media/video/gspca/sn9c20x.c | 2433 +++++++++++++++++++++++++++++++++++ include/linux/videodev2.h | 1 + include/media/v4l2-chip-ident.h | 12 + 6 files changed, 2496 insertions(+) create mode 100644 drivers/media/video/gspca/sn9c20x.c (limited to 'include') diff --git a/Documentation/video4linux/gspca.txt b/Documentation/video4linux/gspca.txt index 2bcf78896e22..573f95b58807 100644 --- a/Documentation/video4linux/gspca.txt +++ b/Documentation/video4linux/gspca.txt @@ -44,7 +44,9 @@ zc3xx 0458:7007 Genius VideoCam V2 zc3xx 0458:700c Genius VideoCam V3 zc3xx 0458:700f Genius VideoCam Web V2 sonixj 0458:7025 Genius Eye 311Q +sn9c20x 0458:7029 Genius Look 320s sonixj 0458:702e Genius Slim 310 NB +sn9c20x 045e:00f4 LifeCam VX-6000 (SN9C20x + OV9650) sonixj 045e:00f5 MicroSoft VX3000 sonixj 045e:00f7 MicroSoft VX1000 ov519 045e:028c Micro$oft xbox cam @@ -282,6 +284,28 @@ sonixj 0c45:613a Microdia Sonix PC Camera sonixj 0c45:613b Surfer SN-206 sonixj 0c45:613c Sonix Pccam168 sonixj 0c45:6143 Sonix Pccam168 +sn9c20x 0c45:6240 PC Camera (SN9C201 + MT9M001) +sn9c20x 0c45:6242 PC Camera (SN9C201 + MT9M111) +sn9c20x 0c45:6248 PC Camera (SN9C201 + OV9655) +sn9c20x 0c45:624e PC Camera (SN9C201 + SOI968) +sn9c20x 0c45:624f PC Camera (SN9C201 + OV9650) +sn9c20x 0c45:6251 PC Camera (SN9C201 + OV9650) +sn9c20x 0c45:6253 PC Camera (SN9C201 + OV9650) +sn9c20x 0c45:6260 PC Camera (SN9C201 + OV7670) +sn9c20x 0c45:6270 PC Camera (SN9C201 + MT9V011/MT9V111/MT9V112) +sn9c20x 0c45:627b PC Camera (SN9C201 + OV7660) +sn9c20x 0c45:627c PC Camera (SN9C201 + HV7131R) +sn9c20x 0c45:627f PC Camera (SN9C201 + OV9650) +sn9c20x 0c45:6280 PC Camera (SN9C202 + MT9M001) +sn9c20x 0c45:6282 PC Camera (SN9C202 + MT9M111) +sn9c20x 0c45:6288 PC Camera (SN9C202 + OV9655) +sn9c20x 0c45:628e PC Camera (SN9C202 + SOI968) +sn9c20x 0c45:628f PC Camera (SN9C202 + OV9650) +sn9c20x 0c45:62a0 PC Camera (SN9C202 + OV7670) +sn9c20x 0c45:62b0 PC Camera (SN9C202 + MT9V011/MT9V111/MT9V112) +sn9c20x 0c45:62b3 PC Camera (SN9C202 + OV9655) +sn9c20x 0c45:62bb PC Camera (SN9C202 + OV7660) +sn9c20x 0c45:62bc PC Camera (SN9C202 + HV7131R) sunplus 0d64:0303 Sunplus FashionCam DXG etoms 102c:6151 Qcam Sangha CIF etoms 102c:6251 Qcam xxxxxx VGA @@ -290,6 +314,7 @@ spca561 10fd:7e50 FlyCam Usb 100 zc3xx 10fd:8050 Typhoon Webshot II USB 300k ov534 1415:2000 Sony HD Eye for PS3 (SLEH 00201) pac207 145f:013a Trust WB-1300N +sn9c20x 145f:013d Trust WB-3600R vc032x 15b8:6001 HP 2.0 Megapixel vc032x 15b8:6002 HP 2.0 Megapixel rz406aa spca501 1776:501c Arowana 300K CMOS Camera @@ -300,4 +325,11 @@ spca500 2899:012c Toptro Industrial spca508 8086:0110 Intel Easy PC Camera spca500 8086:0630 Intel Pocket PC Camera spca506 99fa:8988 Grandtec V.cap +sn9c20x a168:0610 Dino-Lite Digital Microscope (SN9C201 + HV7131R) +sn9c20x a168:0611 Dino-Lite Digital Microscope (SN9C201 + HV7131R) +sn9c20x a168:0613 Dino-Lite Digital Microscope (SN9C201 + HV7131R) +sn9c20x a168:0618 Dino-Lite Digital Microscope (SN9C201 + HV7131R) +sn9c20x a168:0614 Dino-Lite Digital Microscope (SN9C201 + MT9M111) +sn9c20x a168:0615 Dino-Lite Digital Microscope (SN9C201 + MT9M111) +sn9c20x a168:0617 Dino-Lite Digital Microscope (SN9C201 + MT9M111) spca561 abcd:cdee Petcam diff --git a/drivers/media/video/gspca/Kconfig b/drivers/media/video/gspca/Kconfig index 578dc4ffc965..34f46f2bc040 100644 --- a/drivers/media/video/gspca/Kconfig +++ b/drivers/media/video/gspca/Kconfig @@ -102,6 +102,22 @@ config USB_GSPCA_PAC7311 To compile this driver as a module, choose M here: the module will be called gspca_pac7311. +config USB_GSPCA_SN9C20X + tristate "SN9C20X USB Camera Driver" + depends on VIDEO_V4L2 && USB_GSPCA + help + Say Y here if you want support for cameras based on the + sn9c20x chips (SN9C201 and SN9C202). + + To compile this driver as a module, choose M here: the + module will be called gspca_sn9c20x. + +config USB_GSPCA_SN9C20X_EVDEV + bool "Enable evdev support" + depends on USB_GSPCA_SN9C20X + ---help--- + Say Y here in order to enable evdev support for sn9c20x webcam button. + config USB_GSPCA_SONIXB tristate "SONIX Bayer USB Camera Driver" depends on VIDEO_V4L2 && USB_GSPCA diff --git a/drivers/media/video/gspca/Makefile b/drivers/media/video/gspca/Makefile index 8a6643e8eb96..f6d3b86e9ad5 100644 --- a/drivers/media/video/gspca/Makefile +++ b/drivers/media/video/gspca/Makefile @@ -8,6 +8,7 @@ obj-$(CONFIG_USB_GSPCA_OV519) += gspca_ov519.o obj-$(CONFIG_USB_GSPCA_OV534) += gspca_ov534.o obj-$(CONFIG_USB_GSPCA_PAC207) += gspca_pac207.o obj-$(CONFIG_USB_GSPCA_PAC7311) += gspca_pac7311.o +obj-$(CONFIG_USB_GSPCA_SN9C20X) += gspca_sn9c20x.o obj-$(CONFIG_USB_GSPCA_SONIXB) += gspca_sonixb.o obj-$(CONFIG_USB_GSPCA_SONIXJ) += gspca_sonixj.o obj-$(CONFIG_USB_GSPCA_SPCA500) += gspca_spca500.o @@ -35,6 +36,7 @@ gspca_ov519-objs := ov519.o gspca_ov534-objs := ov534.o gspca_pac207-objs := pac207.o gspca_pac7311-objs := pac7311.o +gspca_sn9c20x-objs := sn9c20x.o gspca_sonixb-objs := sonixb.o gspca_sonixj-objs := sonixj.o gspca_spca500-objs := spca500.o diff --git a/drivers/media/video/gspca/sn9c20x.c b/drivers/media/video/gspca/sn9c20x.c new file mode 100644 index 000000000000..78ab26ceb90e --- /dev/null +++ b/drivers/media/video/gspca/sn9c20x.c @@ -0,0 +1,2433 @@ +/* + * Sonix sn9c201 sn9c202 library + * Copyright (C) 2008-2009 microdia project + * Copyright (C) 2009 Brian Johnson + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#include "gspca.h" +#include "jpeg.h" + +#include +#ifdef CONFIG_USB_GSPCA_SN9C20X_EVDEV +#include +#include +#include +#include +#endif + +MODULE_AUTHOR("Brian Johnson , " + "microdia project "); +MODULE_DESCRIPTION("GSPCA/SN9C20X USB Camera Driver"); +MODULE_LICENSE("GPL"); + +#define MODULE_NAME "sn9c20x" + +#define MODE_RAW 0x10 +#define MODE_JPEG 0x20 +#define MODE_SXGA 0x80 + +#define SENSOR_OV9650 0 +#define SENSOR_OV9655 1 +#define SENSOR_SOI968 2 +#define SENSOR_OV7660 3 +#define SENSOR_OV7670 4 +#define SENSOR_MT9V011 5 +#define SENSOR_MT9V111 6 +#define SENSOR_MT9V112 7 +#define SENSOR_MT9M001 8 +#define SENSOR_MT9M111 9 +#define SENSOR_HV7131R 10 +#define SENSOR_MT9VPRB 20 + +/* specific webcam descriptor */ +struct sd { + struct gspca_dev gspca_dev; + +#define MIN_AVG_LUM 80 +#define MAX_AVG_LUM 130 + atomic_t avg_lum; + u8 old_step; + u8 older_step; + u8 exposure_step; + + u8 brightness; + u8 contrast; + u8 saturation; + s16 hue; + u8 gamma; + u8 red; + u8 blue; + + u8 hflip; + u8 vflip; + u8 gain; + u16 exposure; + u8 auto_exposure; + + u8 i2c_addr; + u8 sensor; + u8 hstart; + u8 vstart; + + u8 *jpeg_hdr; + u8 quality; + +#ifdef CONFIG_USB_GSPCA_SN9C20X_EVDEV + struct input_dev *input_dev; + u8 input_gpio; + struct task_struct *input_task; +#endif +}; + +static int sd_setbrightness(struct gspca_dev *gspca_dev, s32 val); +static int sd_getbrightness(struct gspca_dev *gspca_dev, s32 *val); +static int sd_setcontrast(struct gspca_dev *gspca_dev, s32 val); +static int sd_getcontrast(struct gspca_dev *gspca_dev, s32 *val); +static int sd_setsaturation(struct gspca_dev *gspca_dev, s32 val); +static int sd_getsaturation(struct gspca_dev *gspca_dev, s32 *val); +static int sd_sethue(struct gspca_dev *gspca_dev, s32 val); +static int sd_gethue(struct gspca_dev *gspca_dev, s32 *val); +static int sd_setgamma(struct gspca_dev *gspca_dev, s32 val); +static int sd_getgamma(struct gspca_dev *gspca_dev, s32 *val); +static int sd_setredbalance(struct gspca_dev *gspca_dev, s32 val); +static int sd_getredbalance(struct gspca_dev *gspca_dev, s32 *val); +static int sd_setbluebalance(struct gspca_dev *gspca_dev, s32 val); +static int sd_getbluebalance(struct gspca_dev *gspca_dev, s32 *val); +static int sd_setvflip(struct gspca_dev *gspca_dev, s32 val); +static int sd_getvflip(struct gspca_dev *gspca_dev, s32 *val); +static int sd_sethflip(struct gspca_dev *gspca_dev, s32 val); +static int sd_gethflip(struct gspca_dev *gspca_dev, s32 *val); +static int sd_setgain(struct gspca_dev *gspca_dev, s32 val); +static int sd_getgain(struct gspca_dev *gspca_dev, s32 *val); +static int sd_setexposure(struct gspca_dev *gspca_dev, s32 val); +static int sd_getexposure(struct gspca_dev *gspca_dev, s32 *val); +static int sd_setautoexposure(struct gspca_dev *gspca_dev, s32 val); +static int sd_getautoexposure(struct gspca_dev *gspca_dev, s32 *val); + +static struct ctrl sd_ctrls[] = { + { +#define BRIGHTNESS_IDX 0 + { + .id = V4L2_CID_BRIGHTNESS, + .type = V4L2_CTRL_TYPE_INTEGER, + .name = "Brightness", + .minimum = 0, + .maximum = 0xff, + .step = 1, +#define BRIGHTNESS_DEFAULT 0x7f + .default_value = BRIGHTNESS_DEFAULT, + }, + .set = sd_setbrightness, + .get = sd_getbrightness, + }, + { +#define CONTRAST_IDX 1 + { + .id = V4L2_CID_CONTRAST, + .type = V4L2_CTRL_TYPE_INTEGER, + .name = "Contrast", + .minimum = 0, + .maximum = 0xff, + .step = 1, +#define CONTRAST_DEFAULT 0x7f + .default_value = CONTRAST_DEFAULT, + }, + .set = sd_setcontrast, + .get = sd_getcontrast, + }, + { +#define SATURATION_IDX 2 + { + .id = V4L2_CID_SATURATION, + .type = V4L2_CTRL_TYPE_INTEGER, + .name = "Saturation", + .minimum = 0, + .maximum = 0xff, + .step = 1, +#define SATURATION_DEFAULT 0x7f + .default_value = SATURATION_DEFAULT, + }, + .set = sd_setsaturation, + .get = sd_getsaturation, + }, + { +#define HUE_IDX 3 + { + .id = V4L2_CID_HUE, + .type = V4L2_CTRL_TYPE_INTEGER, + .name = "Hue", + .minimum = -180, + .maximum = 180, + .step = 1, +#define HUE_DEFAULT 0 + .default_value = HUE_DEFAULT, + }, + .set = sd_sethue, + .get = sd_gethue, + }, + { +#define GAMMA_IDX 4 + { + .id = V4L2_CID_GAMMA, + .type = V4L2_CTRL_TYPE_INTEGER, + .name = "Gamma", + .minimum = 0, + .maximum = 0xff, + .step = 1, +#define GAMMA_DEFAULT 0x10 + .default_value = GAMMA_DEFAULT, + }, + .set = sd_setgamma, + .get = sd_getgamma, + }, + { +#define BLUE_IDX 5 + { + .id = V4L2_CID_BLUE_BALANCE, + .type = V4L2_CTRL_TYPE_INTEGER, + .name = "Blue Balance", + .minimum = 0, + .maximum = 0x7f, + .step = 1, +#define BLUE_DEFAULT 0x28 + .default_value = BLUE_DEFAULT, + }, + .set = sd_setbluebalance, + .get = sd_getbluebalance, + }, + { +#define RED_IDX 6 + { + .id = V4L2_CID_RED_BALANCE, + .type = V4L2_CTRL_TYPE_INTEGER, + .name = "Red Balance", + .minimum = 0, + .maximum = 0x7f, + .step = 1, +#define RED_DEFAULT 0x28 + .default_value = RED_DEFAULT, + }, + .set = sd_setredbalance, + .get = sd_getredbalance, + }, + { +#define HFLIP_IDX 7 + { + .id = V4L2_CID_HFLIP, + .type = V4L2_CTRL_TYPE_BOOLEAN, + .name = "Horizontal Flip", + .minimum = 0, + .maximum = 1, + .step = 1, +#define HFLIP_DEFAULT 0 + .default_value = HFLIP_DEFAULT, + }, + .set = sd_sethflip, + .get = sd_gethflip, + }, + { +#define VFLIP_IDX 8 + { + .id = V4L2_CID_VFLIP, + .type = V4L2_CTRL_TYPE_BOOLEAN, + .name = "Vertical Flip", + .minimum = 0, + .maximum = 1, + .step = 1, +#define VFLIP_DEFAULT 0 + .default_value = VFLIP_DEFAULT, + }, + .set = sd_setvflip, + .get = sd_getvflip, + }, + { +#define EXPOSURE_IDX 9 + { + .id = V4L2_CID_EXPOSURE, + .type = V4L2_CTRL_TYPE_INTEGER, + .name = "Exposure", + .minimum = 0, + .maximum = 0x1780, + .step = 1, +#define EXPOSURE_DEFAULT 0x33 + .default_value = EXPOSURE_DEFAULT, + }, + .set = sd_setexposure, + .get = sd_getexposure, + }, + { +#define GAIN_IDX 10 + { + .id = V4L2_CID_GAIN, + .type = V4L2_CTRL_TYPE_INTEGER, + .name = "Gain", + .minimum = 0, + .maximum = 28, + .step = 1, +#define GAIN_DEFAULT 0x00 + .default_value = GAIN_DEFAULT, + }, + .set = sd_setgain, + .get = sd_getgain, + }, + { +#define AUTOGAIN_IDX 11 + { + .id = V4L2_CID_AUTOGAIN, + .type = V4L2_CTRL_TYPE_BOOLEAN, + .name = "Auto Exposure", + .minimum = 0, + .maximum = 1, + .step = 1, +#define AUTO_EXPOSURE_DEFAULT 1 + .default_value = AUTO_EXPOSURE_DEFAULT, + }, + .set = sd_setautoexposure, + .get = sd_getautoexposure, + }, +}; + +static const struct v4l2_pix_format vga_mode[] = { + {160, 120, V4L2_PIX_FMT_JPEG, V4L2_FIELD_NONE, + .bytesperline = 240, + .sizeimage = 240 * 120, + .colorspace = V4L2_COLORSPACE_JPEG, + .priv = 0 | MODE_JPEG}, + {160, 120, V4L2_PIX_FMT_SBGGR8, V4L2_FIELD_NONE, + .bytesperline = 160, + .sizeimage = 160 * 120, + .colorspace = V4L2_COLORSPACE_SRGB, + .priv = 0 | MODE_RAW}, + {160, 120, V4L2_PIX_FMT_SN9C20X_I420, V4L2_FIELD_NONE, + .bytesperline = 240, + .sizeimage = 240 * 120, + .colorspace = V4L2_COLORSPACE_SRGB, + .priv = 0}, + {320, 240, V4L2_PIX_FMT_JPEG, V4L2_FIELD_NONE, + .bytesperline = 480, + .sizeimage = 480 * 240 , + .colorspace = V4L2_COLORSPACE_JPEG, + .priv = 1 | MODE_JPEG}, + {320, 240, V4L2_PIX_FMT_SBGGR8, V4L2_FIELD_NONE, + .bytesperline = 320, + .sizeimage = 320 * 240 , + .colorspace = V4L2_COLORSPACE_SRGB, + .priv = 1 | MODE_RAW}, + {320, 240, V4L2_PIX_FMT_SN9C20X_I420, V4L2_FIELD_NONE, + .bytesperline = 480, + .sizeimage = 480 * 240 , + .colorspace = V4L2_COLORSPACE_SRGB, + .priv = 1}, + {640, 480, V4L2_PIX_FMT_JPEG, V4L2_FIELD_NONE, + .bytesperline = 960, + .sizeimage = 960 * 480, + .colorspace = V4L2_COLORSPACE_JPEG, + .priv = 2 | MODE_JPEG}, + {640, 480, V4L2_PIX_FMT_SBGGR8, V4L2_FIELD_NONE, + .bytesperline = 640, + .sizeimage = 640 * 480, + .colorspace = V4L2_COLORSPACE_SRGB, + .priv = 2 | MODE_RAW}, + {640, 480, V4L2_PIX_FMT_SN9C20X_I420, V4L2_FIELD_NONE, + .bytesperline = 960, + .sizeimage = 960 * 480, + .colorspace = V4L2_COLORSPACE_SRGB, + .priv = 2}, +}; + +static const struct v4l2_pix_format sxga_mode[] = { + {160, 120, V4L2_PIX_FMT_JPEG, V4L2_FIELD_NONE, + .bytesperline = 240, + .sizeimage = 240 * 120, + .colorspace = V4L2_COLORSPACE_JPEG, + .priv = 0 | MODE_JPEG}, + {160, 120, V4L2_PIX_FMT_SBGGR8, V4L2_FIELD_NONE, + .bytesperline = 160, + .sizeimage = 160 * 120, + .colorspace = V4L2_COLORSPACE_SRGB, + .priv = 0 | MODE_RAW}, + {160, 120, V4L2_PIX_FMT_SN9C20X_I420, V4L2_FIELD_NONE, + .bytesperline = 240, + .sizeimage = 240 * 120, + .colorspace = V4L2_COLORSPACE_SRGB, + .priv = 0}, + {320, 240, V4L2_PIX_FMT_JPEG, V4L2_FIELD_NONE, + .bytesperline = 480, + .sizeimage = 480 * 240 , + .colorspace = V4L2_COLORSPACE_JPEG, + .priv = 1 | MODE_JPEG}, + {320, 240, V4L2_PIX_FMT_SBGGR8, V4L2_FIELD_NONE, + .bytesperline = 320, + .sizeimage = 320 * 240 , + .colorspace = V4L2_COLORSPACE_SRGB, + .priv = 1 | MODE_RAW}, + {320, 240, V4L2_PIX_FMT_SN9C20X_I420, V4L2_FIELD_NONE, + .bytesperline = 480, + .sizeimage = 480 * 240 , + .colorspace = V4L2_COLORSPACE_SRGB, + .priv = 1}, + {640, 480, V4L2_PIX_FMT_JPEG, V4L2_FIELD_NONE, + .bytesperline = 960, + .sizeimage = 960 * 480, + .colorspace = V4L2_COLORSPACE_JPEG, + .priv = 2 | MODE_JPEG}, + {640, 480, V4L2_PIX_FMT_SBGGR8, V4L2_FIELD_NONE, + .bytesperline = 640, + .sizeimage = 640 * 480, + .colorspace = V4L2_COLORSPACE_SRGB, + .priv = 2 | MODE_RAW}, + {640, 480, V4L2_PIX_FMT_SN9C20X_I420, V4L2_FIELD_NONE, + .bytesperline = 960, + .sizeimage = 960 * 480, + .colorspace = V4L2_COLORSPACE_SRGB, + .priv = 2}, + {1280, 1024, V4L2_PIX_FMT_SBGGR8, V4L2_FIELD_NONE, + .bytesperline = 1280, + .sizeimage = (1280 * 1024) + 64, + .colorspace = V4L2_COLORSPACE_SRGB, + .priv = 3 | MODE_RAW | MODE_SXGA}, +}; + +static const int hsv_red_x[] = { + 41, 44, 46, 48, 50, 52, 54, 56, + 58, 60, 62, 64, 66, 68, 70, 72, + 74, 76, 78, 80, 81, 83, 85, 87, + 88, 90, 92, 93, 95, 97, 98, 100, + 101, 102, 104, 105, 107, 108, 109, 110, + 112, 113, 114, 115, 116, 117, 118, 119, + 120, 121, 122, 123, 123, 124, 125, 125, + 126, 127, 127, 128, 128, 129, 129, 129, + 130, 130, 130, 130, 131, 131, 131, 131, + 131, 131, 131, 131, 130, 130, 130, 130, + 129, 129, 129, 128, 128, 127, 127, 126, + 125, 125, 124, 123, 122, 122, 121, 120, + 119, 118, 117, 116, 115, 114, 112, 111, + 110, 109, 107, 106, 105, 103, 102, 101, + 99, 98, 96, 94, 93, 91, 90, 88, + 86, 84, 83, 81, 79, 77, 75, 74, + 72, 70, 68, 66, 64, 62, 60, 58, + 56, 54, 52, 49, 47, 45, 43, 41, + 39, 36, 34, 32, 30, 28, 25, 23, + 21, 19, 16, 14, 12, 9, 7, 5, + 3, 0, -1, -3, -6, -8, -10, -12, + -15, -17, -19, -22, -24, -26, -28, -30, + -33, -35, -37, -39, -41, -44, -46, -48, + -50, -52, -54, -56, -58, -60, -62, -64, + -66, -68, -70, -72, -74, -76, -78, -80, + -81, -83, -85, -87, -88, -90, -92, -93, + -95, -97, -98, -100, -101, -102, -104, -105, + -107, -108, -109, -110, -112, -113, -114, -115, + -116, -117, -118, -119, -120, -121, -122, -123, + -123, -124, -125, -125, -126, -127, -127, -128, + -128, -128, -128, -128, -128, -128, -128, -128, + -128, -128, -128, -128, -128, -128, -128, -128, + -128, -128, -128, -128, -128, -128, -128, -128, + -128, -127, -127, -126, -125, -125, -124, -123, + -122, -122, -121, -120, -119, -118, -117, -116, + -115, -114, -112, -111, -110, -109, -107, -106, + -105, -103, -102, -101, -99, -98, -96, -94, + -93, -91, -90, -88, -86, -84, -83, -81, + -79, -77, -75, -74, -72, -70, -68, -66, + -64, -62, -60, -58, -56, -54, -52, -49, + -47, -45, -43, -41, -39, -36, -34, -32, + -30, -28, -25, -23, -21, -19, -16, -14, + -12, -9, -7, -5, -3, 0, 1, 3, + 6, 8, 10, 12, 15, 17, 19, 22, + 24, 26, 28, 30, 33, 35, 37, 39, 41 +}; + +static const int hsv_red_y[] = { + 82, 80, 78, 76, 74, 73, 71, 69, + 67, 65, 63, 61, 58, 56, 54, 52, + 50, 48, 46, 44, 41, 39, 37, 35, + 32, 30, 28, 26, 23, 21, 19, 16, + 14, 12, 10, 7, 5, 3, 0, -1, + -3, -6, -8, -10, -13, -15, -17, -19, + -22, -24, -26, -29, -31, -33, -35, -38, + -40, -42, -44, -46, -48, -51, -53, -55, + -57, -59, -61, -63, -65, -67, -69, -71, + -73, -75, -77, -79, -81, -82, -84, -86, + -88, -89, -91, -93, -94, -96, -98, -99, + -101, -102, -104, -105, -106, -108, -109, -110, + -112, -113, -114, -115, -116, -117, -119, -120, + -120, -121, -122, -123, -124, -125, -126, -126, + -127, -128, -128, -128, -128, -128, -128, -128, + -128, -128, -128, -128, -128, -128, -128, -128, + -128, -128, -128, -128, -128, -128, -128, -128, + -128, -128, -128, -128, -128, -128, -128, -128, + -127, -127, -126, -125, -125, -124, -123, -122, + -121, -120, -119, -118, -117, -116, -115, -114, + -113, -111, -110, -109, -107, -106, -105, -103, + -102, -100, -99, -97, -96, -94, -92, -91, + -89, -87, -85, -84, -82, -80, -78, -76, + -74, -73, -71, -69, -67, -65, -63, -61, + -58, -56, -54, -52, -50, -48, -46, -44, + -41, -39, -37, -35, -32, -30, -28, -26, + -23, -21, -19, -16, -14, -12, -10, -7, + -5, -3, 0, 1, 3, 6, 8, 10, + 13, 15, 17, 19, 22, 24, 26, 29, + 31, 33, 35, 38, 40, 42, 44, 46, + 48, 51, 53, 55, 57, 59, 61, 63, + 65, 67, 69, 71, 73, 75, 77, 79, + 81, 82, 84, 86, 88, 89, 91, 93, + 94, 96, 98, 99, 101, 102, 104, 105, + 106, 108, 109, 110, 112, 113, 114, 115, + 116, 117, 119, 120, 120, 121, 122, 123, + 124, 125, 126, 126, 127, 128, 128, 129, + 129, 130, 130, 131, 131, 131, 131, 132, + 132, 132, 132, 132, 132, 132, 132, 132, + 132, 132, 132, 131, 131, 131, 130, 130, + 130, 129, 129, 128, 127, 127, 126, 125, + 125, 124, 123, 122, 121, 120, 119, 118, + 117, 116, 115, 114, 113, 111, 110, 109, + 107, 106, 105, 103, 102, 100, 99, 97, + 96, 94, 92, 91, 89, 87, 85, 84, 82 +}; + +static const int hsv_green_x[] = { + -124, -124, -125, -125, -125, -125, -125, -125, + -125, -126, -126, -125, -125, -125, -125, -125, + -125, -124, -124, -124, -123, -123, -122, -122, + -121, -121, -120, -120, -119, -118, -117, -117, + -116, -115, -114, -113, -112, -111, -110, -109, + -108, -107, -105, -104, -103, -102, -100, -99, + -98, -96, -95, -93, -92, -91, -89, -87, + -86, -84, -83, -81, -79, -77, -76, -74, + -72, -70, -69, -67, -65, -63, -61, -59, + -57, -55, -53, -51, -49, -47, -45, -43, + -41, -39, -37, -35, -33, -30, -28, -26, + -24, -22, -20, -18, -15, -13, -11, -9, + -7, -4, -2, 0, 1, 3, 6, 8, + 10, 12, 14, 17, 19, 21, 23, 25, + 27, 29, 32, 34, 36, 38, 40, 42, + 44, 46, 48, 50, 52, 54, 56, 58, + 60, 62, 64, 66, 68, 70, 71, 73, + 75, 77, 78, 80, 82, 83, 85, 87, + 88, 90, 91, 93, 94, 96, 97, 98, + 100, 101, 102, 104, 105, 106, 107, 108, + 109, 111, 112, 113, 113, 114, 115, 116, + 117, 118, 118, 119, 120, 120, 121, 122, + 122, 123, 123, 124, 124, 124, 125, 125, + 125, 125, 125, 125, 125, 126, 126, 125, + 125, 125, 125, 125, 125, 124, 124, 124, + 123, 123, 122, 122, 121, 121, 120, 120, + 119, 118, 117, 117, 116, 115, 114, 113, + 112, 111, 110, 109, 108, 107, 105, 104, + 103, 102, 100, 99, 98, 96, 95, 93, + 92, 91, 89, 87, 86, 84, 83, 81, + 79, 77, 76, 74, 72, 70, 69, 67, + 65, 63, 61, 59, 57, 55, 53, 51, + 49, 47, 45, 43, 41, 39, 37, 35, + 33, 30, 28, 26, 24, 22, 20, 18, + 15, 13, 11, 9, 7, 4, 2, 0, + -1, -3, -6, -8, -10, -12, -14, -17, + -19, -21, -23, -25, -27, -29, -32, -34, + -36, -38, -40, -42, -44, -46, -48, -50, + -52, -54, -56, -58, -60, -62, -64, -66, + -68, -70, -71, -73, -75, -77, -78, -80, + -82, -83, -85, -87, -88, -90, -91, -93, + -94, -96, -97, -98, -100, -101, -102, -104, + -105, -106, -107, -108, -109, -111, -112, -113, + -113, -114, -115, -116, -117, -118, -118, -119, + -120, -120, -121, -122, -122, -123, -123, -124, -124 +}; + +static const int hsv_green_y[] = { + -100, -99, -98, -97, -95, -94, -93, -91, + -90, -89, -87, -86, -84, -83, -81, -80, + -78, -76, -75, -73, -71, -70, -68, -66, + -64, -63, -61, -59, -57, -55, -53, -51, + -49, -48, -46, -44, -42, -40, -38, -36, + -34, -32, -30, -27, -25, -23, -21, -19, + -17, -15, -13, -11, -9, -7, -4, -2, + 0, 1, 3, 5, 7, 9, 11, 14, + 16, 18, 20, 22, 24, 26, 28, 30, + 32, 34, 36, 38, 40, 42, 44, 46, + 48, 50, 52, 54, 56, 58, 59, 61, + 63, 65, 67, 68, 70, 72, 74, 75, + 77, 78, 80, 82, 83, 85, 86, 88, + 89, 90, 92, 93, 95, 96, 97, 98, + 100, 101, 102, 103, 104, 105, 106, 107, + 108, 109, 110, 111, 112, 112, 113, 114, + 115, 115, 116, 116, 117, 117, 118, 118, + 119, 119, 119, 120, 120, 120, 120, 120, + 121, 121, 121, 121, 121, 121, 120, 120, + 120, 120, 120, 119, 119, 119, 118, 118, + 117, 117, 116, 116, 115, 114, 114, 113, + 112, 111, 111, 110, 109, 108, 107, 106, + 105, 104, 103, 102, 100, 99, 98, 97, + 95, 94, 93, 91, 90, 89, 87, 86, + 84, 83, 81, 80, 78, 76, 75, 73, + 71, 70, 68, 66, 64, 63, 61, 59, + 57, 55, 53, 51, 49, 48, 46, 44, + 42, 40, 38, 36, 34, 32, 30, 27, + 25, 23, 21, 19, 17, 15, 13, 11, + 9, 7, 4, 2, 0, -1, -3, -5, + -7, -9, -11, -14, -16, -18, -20, -22, + -24, -26, -28, -30, -32, -34, -36, -38, + -40, -42, -44, -46, -48, -50, -52, -54, + -56, -58, -59, -61, -63, -65, -67, -68, + -70, -72, -74, -75, -77, -78, -80, -82, + -83, -85, -86, -88, -89, -90, -92, -93, + -95, -96, -97, -98, -100, -101, -102, -103, + -104, -105, -106, -107, -108, -109, -110, -111, + -112, -112, -113, -114, -115, -115, -116, -116, + -117, -117, -118, -118, -119, -119, -119, -120, + -120, -120, -120, -120, -121, -121, -121, -121, + -121, -121, -120, -120, -120, -120, -120, -119, + -119, -119, -118, -118, -117, -117, -116, -116, + -115, -114, -114, -113, -112, -111, -111, -110, + -109, -108, -107, -106, -105, -104, -103, -102, -100 +}; + +static const int hsv_blue_x[] = { + 112, 113, 114, 114, 115, 116, 117, 117, + 118, 118, 119, 119, 120, 120, 120, 121, + 121, 121, 122, 122, 122, 122, 122, 122, + 122, 122, 122, 122, 122, 122, 121, 121, + 121, 120, 120, 120, 119, 119, 118, 118, + 117, 116, 116, 115, 114, 113, 113, 112, + 111, 110, 109, 108, 107, 106, 105, 104, + 103, 102, 100, 99, 98, 97, 95, 94, + 93, 91, 90, 88, 87, 85, 84, 82, + 80, 79, 77, 76, 74, 72, 70, 69, + 67, 65, 63, 61, 60, 58, 56, 54, + 52, 50, 48, 46, 44, 42, 40, 38, + 36, 34, 32, 30, 28, 26, 24, 22, + 19, 17, 15, 13, 11, 9, 7, 5, + 2, 0, -1, -3, -5, -7, -9, -12, + -14, -16, -18, -20, -22, -24, -26, -28, + -31, -33, -35, -37, -39, -41, -43, -45, + -47, -49, -51, -53, -54, -56, -58, -60, + -62, -64, -66, -67, -69, -71, -73, -74, + -76, -78, -79, -81, -83, -84, -86, -87, + -89, -90, -92, -93, -94, -96, -97, -98, + -99, -101, -102, -103, -104, -105, -106, -107, + -108, -109, -110, -111, -112, -113, -114, -114, + -115, -116, -117, -117, -118, -118, -119, -119, + -120, -120, -120, -121, -121, -121, -122, -122, + -122, -122, -122, -122, -122, -122, -122, -122, + -122, -122, -121, -121, -121, -120, -120, -120, + -119, -119, -118, -118, -117, -116, -116, -115, + -114, -113, -113, -112, -111, -110, -109, -108, + -107, -106, -105, -104, -103, -102, -100, -99, + -98, -97, -95, -94, -93, -91, -90, -88, + -87, -85, -84, -82, -80, -79, -77, -76, + -74, -72, -70, -69, -67, -65, -63, -61, + -60, -58, -56, -54, -52, -50, -48, -46, + -44, -42, -40, -38, -36, -34, -32, -30, + -28, -26, -24, -22, -19, -17, -15, -13, + -11, -9, -7, -5, -2, 0, 1, 3, + 5, 7, 9, 12, 14, 16, 18, 20, + 22, 24, 26, 28, 31, 33, 35, 37, + 39, 41, 43, 45, 47, 49, 51, 53, + 54, 56, 58, 60, 62, 64, 66, 67, + 69, 71, 73, 74, 76, 78, 79, 81, + 83, 84, 86, 87, 89, 90, 92, 93, + 94, 96, 97, 98, 99, 101, 102, 103, + 104, 105, 106, 107, 108, 109, 110, 111, 112 +}; + +static const int hsv_blue_y[] = { + -11, -13, -15, -17, -19, -21, -23, -25, + -27, -29, -31, -33, -35, -37, -39, -41, + -43, -45, -46, -48, -50, -52, -54, -55, + -57, -59, -61, -62, -64, -66, -67, -69, + -71, -72, -74, -75, -77, -78, -80, -81, + -83, -84, -86, -87, -88, -90, -91, -92, + -93, -95, -96, -97, -98, -99, -100, -101, + -102, -103, -104, -105, -106, -106, -107, -108, + -109, -109, -110, -111, -111, -112, -112, -113, + -113, -114, -114, -114, -115, -115, -115, -115, + -116, -116, -116, -116, -116, -116, -116, -116, + -116, -115, -115, -115, -115, -114, -114, -114, + -113, -113, -112, -112, -111, -111, -110, -110, + -109, -108, -108, -107, -106, -105, -104, -103, + -102, -101, -100, -99, -98, -97, -96, -95, + -94, -93, -91, -90, -89, -88, -86, -85, + -84, -82, -81, -79, -78, -76, -75, -73, + -71, -70, -68, -67, -65, -63, -62, -60, + -58, -56, -55, -53, -51, -49, -47, -45, + -44, -42, -40, -38, -36, -34, -32, -30, + -28, -26, -24, -22, -20, -18, -16, -14, + -12, -10, -8, -6, -4, -2, 0, 1, + 3, 5, 7, 9, 11, 13, 15, 17, + 19, 21, 23, 25, 27, 29, 31, 33, + 35, 37, 39, 41, 43, 45, 46, 48, + 50, 52, 54, 55, 57, 59, 61, 62, + 64, 66, 67, 69, 71, 72, 74, 75, + 77, 78, 80, 81, 83, 84, 86, 87, + 88, 90, 91, 92, 93, 95, 96, 97, + 98, 99, 100, 101, 102, 103, 104, 105, + 106, 106, 107, 108, 109, 109, 110, 111, + 111, 112, 112, 113, 113, 114, 114, 114, + 115, 115, 115, 115, 116, 116, 116, 116, + 116, 116, 116, 116, 116, 115, 115, 115, + 115, 114, 114, 114, 113, 113, 112, 112, + 111, 111, 110, 110, 109, 108, 108, 107, + 106, 105, 104, 103, 102, 101, 100, 99, + 98, 97, 96, 95, 94, 93, 91, 90, + 89, 88, 86, 85, 84, 82, 81, 79, + 78, 76, 75, 73, 71, 70, 68, 67, + 65, 63, 62, 60, 58, 56, 55, 53, + 51, 49, 47, 45, 44, 42, 40, 38, + 36, 34, 32, 30, 28, 26, 24, 22, + 20, 18, 16, 14, 12, 10, 8, 6, + 4, 2, 0, -1, -3, -5, -7, -9, -11 +}; + +static u16 i2c_ident[] = { + V4L2_IDENT_OV9650, + V4L2_IDENT_OV9655, + V4L2_IDENT_SOI968, + V4L2_IDENT_OV7660, + V4L2_IDENT_OV7670, + V4L2_IDENT_MT9V011, + V4L2_IDENT_MT9V111, + V4L2_IDENT_MT9V112, + V4L2_IDENT_MT9M001C12ST, + V4L2_IDENT_MT9M111, + V4L2_IDENT_HV7131R, +}; + +static u16 bridge_init[][2] = { + {0x1000, 0x78}, {0x1001, 0x40}, {0x1002, 0x1c}, + {0x1020, 0x80}, {0x1061, 0x01}, {0x1067, 0x40}, + {0x1068, 0x30}, {0x1069, 0x20}, {0x106a, 0x10}, + {0x106b, 0x08}, {0x1188, 0x87}, {0x11a1, 0x00}, + {0x11a2, 0x00}, {0x11a3, 0x6a}, {0x11a4, 0x50}, + {0x11ab, 0x00}, {0x11ac, 0x00}, {0x11ad, 0x50}, + {0x11ae, 0x3c}, {0x118a, 0x04}, {0x0395, 0x04}, + {0x11b8, 0x3a}, {0x118b, 0x0e}, {0x10f7, 0x05}, + {0x10f8, 0x14}, {0x10fa, 0xff}, {0x10f9, 0x00}, + {0x11ba, 0x0a}, {0x11a5, 0x2d}, {0x11a6, 0x2d}, + {0x11a7, 0x3a}, {0x11a8, 0x05}, {0x11a9, 0x04}, + {0x11aa, 0x3f}, {0x11af, 0x28}, {0x11b0, 0xd8}, + {0x11b1, 0x14}, {0x11b2, 0xec}, {0x11b3, 0x32}, + {0x11b4, 0xdd}, {0x11b5, 0x32}, {0x11b6, 0xdd}, + {0x10e0, 0x2c}, {0x11bc, 0x40}, {0x11bd, 0x01}, + {0x11be, 0xf0}, {0x11bf, 0x00}, {0x118c, 0x1f}, + {0x118d, 0x1f}, {0x118e, 0x1f}, {0x118f, 0x1f}, + {0x1180, 0x01}, {0x1181, 0x00}, {0x1182, 0x01}, + {0x1183, 0x00}, {0x1184, 0x50}, {0x1185, 0x80} +}; + +/* Gain = (bit[3:0] / 16 + 1) * (bit[4] + 1) * (bit[5] + 1) * (bit[6] + 1) */ +static u8 ov_gain[] = { + 0x00 /* 1x */, 0x04 /* 1.25x */, 0x08 /* 1.5x */, 0x0c /* 1.75x */, + 0x10 /* 2x */, 0x12 /* 2.25x */, 0x14 /* 2.5x */, 0x16 /* 2.75x */, + 0x18 /* 3x */, 0x1a /* 3.25x */, 0x1c /* 3.5x */, 0x1e /* 3.75x */, + 0x30 /* 4x */, 0x31 /* 4.25x */, 0x32 /* 4.5x */, 0x33 /* 4.75x */, + 0x34 /* 5x */, 0x35 /* 5.25x */, 0x36 /* 5.5x */, 0x37 /* 5.75x */, + 0x38 /* 6x */, 0x39 /* 6.25x */, 0x3a /* 6.5x */, 0x3b /* 6.75x */, + 0x3c /* 7x */, 0x3d /* 7.25x */, 0x3e /* 7.5x */, 0x3f /* 7.75x */, + 0x70 /* 8x */ +}; + +/* Gain = (bit[8] + 1) * (bit[7] + 1) * (bit[6:0] * 0.03125) */ +static u16 micron1_gain[] = { + /* 1x 1.25x 1.5x 1.75x */ + 0x0020, 0x0028, 0x0030, 0x0038, + /* 2x 2.25x 2.5x 2.75x */ + 0x00a0, 0x00a4, 0x00a8, 0x00ac, + /* 3x 3.25x 3.5x 3.75x */ + 0x00b0, 0x00b4, 0x00b8, 0x00bc, + /* 4x 4.25x 4.5x 4.75x */ + 0x00c0, 0x00c4, 0x00c8, 0x00cc, + /* 5x 5.25x 5.5x 5.75x */ + 0x00d0, 0x00d4, 0x00d8, 0x00dc, + /* 6x 6.25x 6.5x 6.75x */ + 0x00e0, 0x00e4, 0x00e8, 0x00ec, + /* 7x 7.25x 7.5x 7.75x */ + 0x00f0, 0x00f4, 0x00f8, 0x00fc, + /* 8x */ + 0x01c0 +}; + +/* mt9m001 sensor uses a different gain formula then other micron sensors */ +/* Gain = (bit[6] + 1) * (bit[5-0] * 0.125) */ +static u16 micron2_gain[] = { + /* 1x 1.25x 1.5x 1.75x */ + 0x0008, 0x000a, 0x000c, 0x000e, + /* 2x 2.25x 2.5x 2.75x */ + 0x0010, 0x0012, 0x0014, 0x0016, + /* 3x 3.25x 3.5x 3.75x */ + 0x0018, 0x001a, 0x001c, 0x001e, + /* 4x 4.25x 4.5x 4.75x */ + 0x0020, 0x0051, 0x0052, 0x0053, + /* 5x 5.25x 5.5x 5.75x */ + 0x0054, 0x0055, 0x0056, 0x0057, + /* 6x 6.25x 6.5x 6.75x */ + 0x0058, 0x0059, 0x005a, 0x005b, + /* 7x 7.25x 7.5x 7.75x */ + 0x005c, 0x005d, 0x005e, 0x005f, + /* 8x */ + 0x0060 +}; + +/* Gain = .5 + bit[7:0] / 16 */ +static u8 hv7131r_gain[] = { + 0x08 /* 1x */, 0x0c /* 1.25x */, 0x10 /* 1.5x */, 0x14 /* 1.75x */, + 0x18 /* 2x */, 0x1c /* 2.25x */, 0x20 /* 2.5x */, 0x24 /* 2.75x */, + 0x28 /* 3x */, 0x2c /* 3.25x */, 0x30 /* 3.5x */, 0x34 /* 3.75x */, + 0x38 /* 4x */, 0x3c /* 4.25x */, 0x40 /* 4.5x */, 0x44 /* 4.75x */, + 0x48 /* 5x */, 0x4c /* 5.25x */, 0x50 /* 5.5x */, 0x54 /* 5.75x */, + 0x58 /* 6x */, 0x5c /* 6.25x */, 0x60 /* 6.5x */, 0x64 /* 6.75x */, + 0x68 /* 7x */, 0x6c /* 7.25x */, 0x70 /* 7.5x */, 0x74 /* 7.75x */, + 0x78 /* 8x */ +}; + +static u8 soi968_init[][2] = { + {0x12, 0x80}, {0x0c, 0x00}, {0x0f, 0x1f}, + {0x11, 0x80}, {0x38, 0x52}, {0x1e, 0x00}, + {0x33, 0x08}, {0x35, 0x8c}, {0x36, 0x0c}, + {0x37, 0x04}, {0x45, 0x04}, {0x47, 0xff}, + {0x3e, 0x00}, {0x3f, 0x00}, {0x3b, 0x20}, + {0x3a, 0x96}, {0x3d, 0x0a}, {0x14, 0x8e}, + {0x13, 0x8a}, {0x12, 0x40}, {0x17, 0x13}, + {0x18, 0x63}, {0x19, 0x01}, {0x1a, 0x79}, + {0x32, 0x24}, {0x03, 0x00}, {0x11, 0x40}, + {0x2a, 0x10}, {0x2b, 0xe0}, {0x10, 0x32}, + {0x00, 0x00}, {0x01, 0x80}, {0x02, 0x80}, +}; + +static u8 ov7660_init[][2] = { + {0x0e, 0x80}, {0x0d, 0x08}, {0x0f, 0xc3}, + {0x04, 0xc3}, {0x10, 0x40}, {0x11, 0x40}, + {0x12, 0x05}, {0x13, 0xba}, {0x14, 0x2a}, + {0x37, 0x0f}, {0x38, 0x02}, {0x39, 0x43}, + {0x3a, 0x00}, {0x69, 0x90}, {0x2d, 0xf6}, + {0x2e, 0x0b}, {0x01, 0x78}, {0x02, 0x50}, +}; + +static u8 ov7670_init[][2] = { + {0x12, 0x80}, {0x11, 0x80}, {0x3a, 0x04}, {0x12, 0x01}, + {0x32, 0xb6}, {0x03, 0x0a}, {0x0c, 0x00}, {0x3e, 0x00}, + {0x70, 0x3a}, {0x71, 0x35}, {0x72, 0x11}, {0x73, 0xf0}, + {0xa2, 0x02}, {0x13, 0xe0}, {0x00, 0x00}, {0x10, 0x00}, + {0x0d, 0x40}, {0x14, 0x28}, {0xa5, 0x05}, {0xab, 0x07}, + {0x24, 0x95}, {0x25, 0x33}, {0x26, 0xe3}, {0x9f, 0x75}, + {0xa0, 0x65}, {0xa1, 0x0b}, {0xa6, 0xd8}, {0xa7, 0xd8}, + {0xa8, 0xf0}, {0xa9, 0x90}, {0xaa, 0x94}, {0x13, 0xe5}, + {0x0e, 0x61}, {0x0f, 0x4b}, {0x16, 0x02}, {0x1e, 0x27}, + {0x21, 0x02}, {0x22, 0x91}, {0x29, 0x07}, {0x33, 0x0b}, + {0x35, 0x0b}, {0x37, 0x1d}, {0x38, 0x71}, {0x39, 0x2a}, + {0x3c, 0x78}, {0x4d, 0x40}, {0x4e, 0x20}, {0x69, 0x00}, + {0x74, 0x19}, {0x8d, 0x4f}, {0x8e, 0x00}, {0x8f, 0x00}, + {0x90, 0x00}, {0x91, 0x00}, {0x96, 0x00}, {0x9a, 0x80}, + {0xb0, 0x84}, {0xb1, 0x0c}, {0xb2, 0x0e}, {0xb3, 0x82}, + {0xb8, 0x0a}, {0x43, 0x0a}, {0x44, 0xf0}, {0x45, 0x20}, + {0x46, 0x7d}, {0x47, 0x29}, {0x48, 0x4a}, {0x59, 0x8c}, + {0x5a, 0xa5}, {0x5b, 0xde}, {0x5c, 0x96}, {0x5d, 0x66}, + {0x5e, 0x10}, {0x6c, 0x0a}, {0x6d, 0x55}, {0x6e, 0x11}, + {0x6f, 0x9e}, {0x6a, 0x40}, {0x01, 0x40}, {0x02, 0x40}, + {0x13, 0xe7}, {0x4f, 0x6e}, {0x50, 0x70}, {0x51, 0x02}, + {0x52, 0x1d}, {0x53, 0x56}, {0x54, 0x73}, {0x55, 0x0a}, + {0x56, 0x55}, {0x57, 0x80}, {0x58, 0x9e}, {0x41, 0x08}, + {0x3f, 0x02}, {0x75, 0x03}, {0x76, 0x63}, {0x4c, 0x04}, + {0x77, 0x06}, {0x3d, 0x02}, {0x4b, 0x09}, {0xc9, 0x30}, + {0x41, 0x08}, {0x56, 0x48}, {0x34, 0x11}, {0xa4, 0x88}, + {0x96, 0x00}, {0x97, 0x30}, {0x98, 0x20}, {0x99, 0x30}, + {0x9a, 0x84}, {0x9b, 0x29}, {0x9c, 0x03}, {0x9d, 0x99}, + {0x9e, 0x7f}, {0x78, 0x04}, {0x79, 0x01}, {0xc8, 0xf0}, + {0x79, 0x0f}, {0xc8, 0x00}, {0x79, 0x10}, {0xc8, 0x7e}, + {0x79, 0x0a}, {0xc8, 0x80}, {0x79, 0x0b}, {0xc8, 0x01}, + {0x79, 0x0c}, {0xc8, 0x0f}, {0x79, 0x0d}, {0xc8, 0x20}, + {0x79, 0x09}, {0xc8, 0x80}, {0x79, 0x02}, {0xc8, 0xc0}, + {0x79, 0x03}, {0xc8, 0x40}, {0x79, 0x05}, {0xc8, 0x30}, + {0x79, 0x26}, {0x62, 0x20}, {0x63, 0x00}, {0x64, 0x06}, + {0x65, 0x00}, {0x66, 0x05}, {0x94, 0x05}, {0x95, 0x0a}, + {0x17, 0x13}, {0x18, 0x01}, {0x19, 0x02}, {0x1a, 0x7a}, + {0x46, 0x59}, {0x47, 0x30}, {0x58, 0x9a}, {0x59, 0x84}, + {0x5a, 0x91}, {0x5b, 0x57}, {0x5c, 0x75}, {0x5d, 0x6d}, + {0x5e, 0x13}, {0x64, 0x07}, {0x94, 0x07}, {0x95, 0x0d}, + {0xa6, 0xdf}, {0xa7, 0xdf}, {0x48, 0x4d}, {0x51, 0x00}, + {0x6b, 0x0a}, {0x11, 0x80}, {0x2a, 0x00}, {0x2b, 0x00}, + {0x92, 0x00}, {0x93, 0x00}, {0x55, 0x0a}, {0x56, 0x60}, + {0x4f, 0x6e}, {0x50, 0x70}, {0x51, 0x00}, {0x52, 0x1d}, + {0x53, 0x56}, {0x54, 0x73}, {0x58, 0x9a}, {0x4f, 0x6e}, + {0x50, 0x70}, {0x51, 0x00}, {0x52, 0x1d}, {0x53, 0x56}, + {0x54, 0x73}, {0x58, 0x9a}, {0x3f, 0x01}, {0x7b, 0x03}, + {0x7c, 0x09}, {0x7d, 0x16}, {0x7e, 0x38}, {0x7f, 0x47}, + {0x80, 0x53}, {0x81, 0x5e}, {0x82, 0x6a}, {0x83, 0x74}, + {0x84, 0x80}, {0x85, 0x8c}, {0x86, 0x9b}, {0x87, 0xb2}, + {0x88, 0xcc}, {0x89, 0xe5}, {0x7a, 0x24}, {0x3b, 0x00}, + {0x9f, 0x76}, {0xa0, 0x65}, {0x13, 0xe2}, {0x6b, 0x0a}, + {0x11, 0x80}, {0x2a, 0x00}, {0x2b, 0x00}, {0x92, 0x00}, + {0x93, 0x00}, +}; + +static u8 ov9650_init[][2] = { + {0x12, 0x80}, {0x00, 0x00}, {0x01, 0x78}, + {0x02, 0x78}, {0x03, 0x36}, {0x04, 0x03}, + {0x05, 0x00}, {0x06, 0x00}, {0x08, 0x00}, + {0x09, 0x01}, {0x0c, 0x00}, {0x0d, 0x00}, + {0x0e, 0xa0}, {0x0f, 0x52}, {0x10, 0x7c}, + {0x11, 0x80}, {0x12, 0x45}, {0x13, 0xc2}, + {0x14, 0x2e}, {0x15, 0x00}, {0x16, 0x07}, + {0x17, 0x24}, {0x18, 0xc5}, {0x19, 0x00}, + {0x1a, 0x3c}, {0x1b, 0x00}, {0x1e, 0x04}, + {0x1f, 0x00}, {0x24, 0x78}, {0x25, 0x68}, + {0x26, 0xd4}, {0x27, 0x80}, {0x28, 0x80}, + {0x29, 0x30}, {0x2a, 0x00}, {0x2b, 0x00}, + {0x2c, 0x80}, {0x2d, 0x00}, {0x2e, 0x00}, + {0x2f, 0x00}, {0x30, 0x08}, {0x31, 0x30}, + {0x32, 0x84}, {0x33, 0xe2}, {0x34, 0xbf}, + {0x35, 0x81}, {0x36, 0xf9}, {0x37, 0x00}, + {0x38, 0x93}, {0x39, 0x50}, {0x3a, 0x01}, + {0x3b, 0x01}, {0x3c, 0x73}, {0x3d, 0x19}, + {0x3e, 0x0b}, {0x3f, 0x80}, {0x40, 0xc1}, + {0x41, 0x00}, {0x42, 0x08}, {0x67, 0x80}, + {0x68, 0x80}, {0x69, 0x40}, {0x6a, 0x00}, + {0x6b, 0x0a}, {0x8b, 0x06}, {0x8c, 0x20}, + {0x8d, 0x00}, {0x8e, 0x00}, {0x8f, 0xdf}, + {0x92, 0x00}, {0x93, 0x00}, {0x94, 0x88}, + {0x95, 0x88}, {0x96, 0x04}, {0xa1, 0x00}, + {0xa5, 0x80}, {0xa8, 0x80}, {0xa9, 0xb8}, + {0xaa, 0x92}, {0xab, 0x0a}, +}; + +static u8 ov9655_init[][2] = { + {0x12, 0x80}, {0x12, 0x01}, {0x0d, 0x00}, {0x0e, 0x61}, + {0x11, 0x80}, {0x13, 0xba}, {0x14, 0x2e}, {0x16, 0x24}, + {0x1e, 0x04}, {0x1e, 0x04}, {0x1e, 0x04}, {0x27, 0x08}, + {0x28, 0x08}, {0x29, 0x15}, {0x2c, 0x08}, {0x32, 0xbf}, + {0x34, 0x3d}, {0x35, 0x00}, {0x36, 0xf8}, {0x38, 0x12}, + {0x39, 0x57}, {0x3a, 0x00}, {0x3b, 0xcc}, {0x3c, 0x0c}, + {0x3d, 0x19}, {0x3e, 0x0c}, {0x3f, 0x01}, {0x41, 0x40}, + {0x42, 0x80}, {0x45, 0x46}, {0x46, 0x62}, {0x47, 0x2a}, + {0x48, 0x3c}, {0x4a, 0xf0}, {0x4b, 0xdc}, {0x4c, 0xdc}, + {0x4d, 0xdc}, {0x4e, 0xdc}, {0x69, 0x02}, {0x6c, 0x04}, + {0x6f, 0x9e}, {0x70, 0x05}, {0x71, 0x78}, {0x77, 0x02}, + {0x8a, 0x23}, {0x8c, 0x0d}, {0x90, 0x7e}, {0x91, 0x7c}, + {0x9f, 0x6e}, {0xa0, 0x6e}, {0xa5, 0x68}, {0xa6, 0x60}, + {0xa8, 0xc1}, {0xa9, 0xfa}, {0xaa, 0x92}, {0xab, 0x04}, + {0xac, 0x80}, {0xad, 0x80}, {0xae, 0x80}, {0xaf, 0x80}, + {0xb2, 0xf2}, {0xb3, 0x20}, {0xb5, 0x00}, {0xb6, 0xaf}, + {0xbb, 0xae}, {0xbc, 0x44}, {0xbd, 0x44}, {0xbe, 0x3b}, + {0xbf, 0x3a}, {0xc0, 0xe2}, {0xc1, 0xc8}, {0xc2, 0x01}, + {0xc4, 0x00}, {0xc6, 0x85}, {0xc7, 0x81}, {0xc9, 0xe0}, + {0xca, 0xe8}, {0xcc, 0xd8}, {0xcd, 0x93}, {0x12, 0x61}, + {0x36, 0xfa}, {0x8c, 0x8d}, {0xc0, 0xaa}, {0x69, 0x0a}, + {0x03, 0x12}, {0x17, 0x14}, {0x18, 0x00}, {0x19, 0x01}, + {0x1a, 0x3d}, {0x32, 0xbf}, {0x11, 0x80}, {0x2a, 0x10}, + {0x2b, 0x0a}, {0x92, 0x00}, {0x93, 0x00}, {0x1e, 0x04}, + {0x1e, 0x04}, {0x10, 0x7c}, {0x04, 0x03}, {0xa1, 0x00}, + {0x2d, 0x00}, {0x2e, 0x00}, {0x00, 0x00}, {0x01, 0x80}, + {0x02, 0x80}, {0x12, 0x61}, {0x36, 0xfa}, {0x8c, 0x8d}, + {0xc0, 0xaa}, {0x69, 0x0a}, {0x03, 0x12}, {0x17, 0x14}, + {0x18, 0x00}, {0x19, 0x01}, {0x1a, 0x3d}, {0x32, 0xbf}, + {0x11, 0x80}, {0x2a, 0x10}, {0x2b, 0x0a}, {0x92, 0x00}, + {0x93, 0x00}, {0x04, 0x01}, {0x10, 0x1f}, {0xa1, 0x00}, + {0x00, 0x0a}, {0xa1, 0x00}, {0x10, 0x5d}, {0x04, 0x03}, + {0x00, 0x01}, {0xa1, 0x00}, {0x10, 0x7c}, {0x04, 0x03}, + {0x00, 0x03}, {0x00, 0x0a}, {0x00, 0x10}, {0x00, 0x13}, +}; + +static u16 mt9v112_init[][2] = { + {0xf0, 0x0000}, {0x0d, 0x0021}, {0x0d, 0x0020}, + {0x34, 0xc019}, {0x0a, 0x0011}, {0x0b, 0x000b}, + {0x20, 0x0703}, {0x35, 0x2022}, {0xf0, 0x0001}, + {0x05, 0x0000}, {0x06, 0x340c}, {0x3b, 0x042a}, + {0x3c, 0x0400}, {0xf0, 0x0002}, {0x2e, 0x0c58}, + {0x5b, 0x0001}, {0xc8, 0x9f0b}, {0xf0, 0x0001}, + {0x9b, 0x5300}, {0xf0, 0x0000}, {0x2b, 0x0020}, + {0x2c, 0x002a}, {0x2d, 0x0032}, {0x2e, 0x0020}, + {0x09, 0x01dc}, {0x01, 0x000c}, {0x02, 0x0020}, + {0x03, 0x01e0}, {0x04, 0x0280}, {0x06, 0x000c}, + {0x05, 0x0098}, {0x20, 0x0703}, {0x09, 0x01f2}, + {0x2b, 0x00a0}, {0x2c, 0x00a0}, {0x2d, 0x00a0}, + {0x2e, 0x00a0}, {0x01, 0x000c}, {0x02, 0x0020}, + {0x03, 0x01e0}, {0x04, 0x0280}, {0x06, 0x000c}, + {0x05, 0x0098}, {0x09, 0x01c1}, {0x2b, 0x00ae}, + {0x2c, 0x00ae}, {0x2d, 0x00ae}, {0x2e, 0x00ae}, +}; + +static u16 mt9v111_init[][2] = { + {0x01, 0x0004}, {0x0d, 0x0001}, {0x0d, 0x0000}, + {0x01, 0x0001}, {0x02, 0x0016}, {0x03, 0x01e1}, + {0x04, 0x0281}, {0x05, 0x0004}, {0x07, 0x3002}, + {0x21, 0x0000}, {0x25, 0x4024}, {0x26, 0xff03}, + {0x27, 0xff10}, {0x2b, 0x7828}, {0x2c, 0xb43c}, + {0x2d, 0xf0a0}, {0x2e, 0x0c64}, {0x2f, 0x0064}, + {0x67, 0x4010}, {0x06, 0x301e}, {0x08, 0x0480}, + {0x01, 0x0004}, {0x02, 0x0016}, {0x03, 0x01e6}, + {0x04, 0x0286}, {0x05, 0x0004}, {0x06, 0x0000}, + {0x07, 0x3002}, {0x08, 0x0008}, {0x0c, 0x0000}, + {0x0d, 0x0000}, {0x0e, 0x0000}, {0x0f, 0x0000}, + {0x10, 0x0000}, {0x11, 0x0000}, {0x12, 0x00b0}, + {0x13, 0x007c}, {0x14, 0x0000}, {0x15, 0x0000}, + {0x16, 0x0000}, {0x17, 0x0000}, {0x18, 0x0000}, + {0x19, 0x0000}, {0x1a, 0x0000}, {0x1b, 0x0000}, + {0x1c, 0x0000}, {0x1d, 0x0000}, {0x30, 0x0000}, + {0x30, 0x0005}, {0x31, 0x0000}, {0x02, 0x0016}, + {0x03, 0x01e1}, {0x04, 0x0281}, {0x05, 0x0004}, + {0x06, 0x0000}, {0x07, 0x3002}, {0x06, 0x002d}, + {0x05, 0x0004}, {0x09, 0x0064}, {0x2b, 0x00a0}, + {0x2c, 0x00a0}, {0x2d, 0x00a0}, {0x2e, 0x00a0}, + {0x02, 0x0016}, {0x03, 0x01e1}, {0x04, 0x0281}, + {0x05, 0x0004}, {0x06, 0x002d}, {0x07, 0x3002}, + {0x0e, 0x0008}, {0x06, 0x002d}, {0x05, 0x0004}, +}; + +static u16 mt9v011_init[][2] = { + {0x07, 0x0002}, {0x0d, 0x0001}, {0x0d, 0x0000}, + {0x01, 0x0008}, {0x02, 0x0016}, {0x03, 0x01e1}, + {0x04, 0x0281}, {0x05, 0x0083}, {0x06, 0x0006}, + {0x0d, 0x0002}, {0x0a, 0x0000}, {0x0b, 0x0000}, + {0x0c, 0x0000}, {0x0d, 0x0000}, {0x0e, 0x0000}, + {0x0f, 0x0000}, {0x10, 0x0000}, {0x11, 0x0000}, + {0x12, 0x0000}, {0x13, 0x0000}, {0x14, 0x0000}, + {0x15, 0x0000}, {0x16, 0x0000}, {0x17, 0x0000}, + {0x18, 0x0000}, {0x19, 0x0000}, {0x1a, 0x0000}, + {0x1b, 0x0000}, {0x1c, 0x0000}, {0x1d, 0x0000}, + {0x32, 0x0000}, {0x20, 0x1101}, {0x21, 0x0000}, + {0x22, 0x0000}, {0x23, 0x0000}, {0x24, 0x0000}, + {0x25, 0x0000}, {0x26, 0x0000}, {0x27, 0x0024}, + {0x2f, 0xf7b0}, {0x30, 0x0005}, {0x31, 0x0000}, + {0x32, 0x0000}, {0x33, 0x0000}, {0x34, 0x0100}, + {0x3d, 0x068f}, {0x40, 0x01e0}, {0x41, 0x00d1}, + {0x44, 0x0082}, {0x5a, 0x0000}, {0x5b, 0x0000}, + {0x5c, 0x0000}, {0x5d, 0x0000}, {0x5e, 0x0000}, + {0x5f, 0xa31d}, {0x62, 0x0611}, {0x0a, 0x0000}, + {0x06, 0x0029}, {0x05, 0x0009}, {0x20, 0x1101}, + {0x20, 0x1101}, {0x09, 0x0064}, {0x07, 0x0003}, + {0x2b, 0x0033}, {0x2c, 0x00a0}, {0x2d, 0x00a0}, + {0x2e, 0x0033}, {0x07, 0x0002}, {0x06, 0x0000}, + {0x06, 0x0029}, {0x05, 0x0009}, +}; + +static u16 mt9m001_init[][2] = { + {0x0d, 0x0001}, {0x0d, 0x0000}, {0x01, 0x000e}, + {0x02, 0x0014}, {0x03, 0x03c1}, {0x04, 0x0501}, + {0x05, 0x0083}, {0x06, 0x0006}, {0x0d, 0x0002}, + {0x0a, 0x0000}, {0x0c, 0x0000}, {0x11, 0x0000}, + {0x1e, 0x8000}, {0x5f, 0x8904}, {0x60, 0x0000}, + {0x61, 0x0000}, {0x62, 0x0498}, {0x63, 0x0000}, + {0x64, 0x0000}, {0x20, 0x111d}, {0x06, 0x00f2}, + {0x05, 0x0013}, {0x09, 0x10f2}, {0x07, 0x0003}, + {0x2b, 0x002a}, {0x2d, 0x002a}, {0x2c, 0x002a}, + {0x2e, 0x0029}, {0x07, 0x0002}, +}; + +static u16 mt9m111_init[][2] = { + {0xf0, 0x0000}, {0x0d, 0x0008}, {0x0d, 0x0009}, + {0x0d, 0x0008}, {0xf0, 0x0001}, {0x3a, 0x4300}, + {0x9b, 0x4300}, {0xa1, 0x0280}, {0xa4, 0x0200}, + {0x06, 0x308e}, {0xf0, 0x0000}, +}; + +static u8 hv7131r_init[][2] = { + {0x02, 0x08}, {0x02, 0x00}, {0x01, 0x08}, + {0x02, 0x00}, {0x20, 0x00}, {0x21, 0xd0}, + {0x22, 0x00}, {0x23, 0x09}, {0x01, 0x08}, + {0x01, 0x08}, {0x01, 0x08}, {0x25, 0x07}, + {0x26, 0xc3}, {0x27, 0x50}, {0x30, 0x62}, + {0x31, 0x10}, {0x32, 0x06}, {0x33, 0x10}, + {0x20, 0x00}, {0x21, 0xd0}, {0x22, 0x00}, + {0x23, 0x09}, {0x01, 0x08}, +}; + +int reg_r(struct gspca_dev *gspca_dev, u16 reg, u16 length) +{ + struct usb_device *dev = gspca_dev->dev; + int result; + result = usb_control_msg(dev, usb_rcvctrlpipe(dev, 0), + 0x00, + USB_DIR_IN | USB_TYPE_VENDOR | USB_RECIP_INTERFACE, + reg, + 0x00, + gspca_dev->usb_buf, + length, + 500); + if (unlikely(result < 0 || result != length)) { + err("Read register failed 0x%02X", reg); + return -EIO; + } + return 0; +} + +int reg_w(struct gspca_dev *gspca_dev, u16 reg, const u8 *buffer, int length) +{ + struct usb_device *dev = gspca_dev->dev; + int result; + memcpy(gspca_dev->usb_buf, buffer, length); + result = usb_control_msg(dev, usb_sndctrlpipe(dev, 0), + 0x08, + USB_DIR_OUT | USB_TYPE_VENDOR | USB_RECIP_INTERFACE, + reg, + 0x00, + gspca_dev->usb_buf, + length, + 500); + if (unlikely(result < 0 || result != length)) { + err("Write register failed index 0x%02X", reg); + return -EIO; + } + return 0; +} + +int reg_w1(struct gspca_dev *gspca_dev, u16 reg, const u8 value) +{ + u8 data[1] = {value}; + return reg_w(gspca_dev, reg, data, 1); +} + +int i2c_w(struct gspca_dev *gspca_dev, const u8 *buffer) +{ + int i; + reg_w(gspca_dev, 0x10c0, buffer, 8); + for (i = 0; i < 5; i++) { + reg_r(gspca_dev, 0x10c0, 1); + if (gspca_dev->usb_buf[0] & 0x04) { + if (gspca_dev->usb_buf[0] & 0x08) + return -1; + return 0; + } + msleep(1); + } + return -1; +} + +int i2c_w1(struct gspca_dev *gspca_dev, u8 reg, u8 val) +{ + struct sd *sd = (struct sd *) gspca_dev; + + u8 row[8]; + + /* + * from the point of view of the bridge, the length + * includes the address + */ + row[0] = 0x81 | (2 << 4); + row[1] = sd->i2c_addr; + row[2] = reg; + row[3] = val; + row[4] = 0x00; + row[5] = 0x00; + row[6] = 0x00; + row[7] = 0x10; + + return i2c_w(gspca_dev, row); +} + +int i2c_w2(struct gspca_dev *gspca_dev, u8 reg, u16 val) +{ + struct sd *sd = (struct sd *) gspca_dev; + u8 row[8]; + + /* + * from the point of view of the bridge, the length + * includes the address + */ + row[0] = 0x81 | (3 << 4); + row[1] = sd->i2c_addr; + row[2] = reg; + row[3] = (val >> 8) & 0xff; + row[4] = val & 0xff; + row[5] = 0x00; + row[6] = 0x00; + row[7] = 0x10; + + return i2c_w(gspca_dev, row); +} + +int i2c_r1(struct gspca_dev *gspca_dev, u8 reg, u8 *val) +{ + struct sd *sd = (struct sd *) gspca_dev; + u8 row[8]; + + row[0] = 0x81 | 0x10; + row[1] = sd->i2c_addr; + row[2] = reg; + row[3] = 0; + row[4] = 0; + row[5] = 0; + row[6] = 0; + row[7] = 0x10; + reg_w(gspca_dev, 0x10c0, row, 8); + msleep(1); + row[0] = 0x81 | (2 << 4) | 0x02; + row[2] = 0; + reg_w(gspca_dev, 0x10c0, row, 8); + msleep(1); + reg_r(gspca_dev, 0x10c2, 5); + *val = gspca_dev->usb_buf[3]; + return 0; +} + +int i2c_r2(struct gspca_dev *gspca_dev, u8 reg, u16 *val) +{ + struct sd *sd = (struct sd *) gspca_dev; + u8 row[8]; + + row[0] = 0x81 | 0x10; + row[1] = sd->i2c_addr; + row[2] = reg; + row[3] = 0; + row[4] = 0; + row[5] = 0; + row[6] = 0; + row[7] = 0x10; + reg_w(gspca_dev, 0x10c0, row, 8); + msleep(1); + row[0] = 0x81 | (3 << 4) | 0x02; + row[2] = 0; + reg_w(gspca_dev, 0x10c0, row, 8); + msleep(1); + reg_r(gspca_dev, 0x10c2, 5); + *val = (gspca_dev->usb_buf[2] << 8) | gspca_dev->usb_buf[3]; + return 0; +} + +static int ov9650_init_sensor(struct gspca_dev *gspca_dev) +{ + int i; + struct sd *sd = (struct sd *) gspca_dev; + + for (i = 0; i < ARRAY_SIZE(ov9650_init); i++) { + if (i2c_w1(gspca_dev, ov9650_init[i][0], + ov9650_init[i][1]) < 0) { + err("OV9650 sensor initialization failed"); + return -ENODEV; + } + } + sd->hstart = 1; + sd->vstart = 7; + return 0; +} + +static int ov9655_init_sensor(struct gspca_dev *gspca_dev) +{ + int i; + struct sd *sd = (struct sd *) gspca_dev; + + for (i = 0; i < ARRAY_SIZE(ov9655_init); i++) { + if (i2c_w1(gspca_dev, ov9655_init[i][0], + ov9655_init[i][1]) < 0) { + err("OV9655 sensor initialization failed"); + return -ENODEV; + } + } + /* disable hflip and vflip */ + gspca_dev->ctrl_dis = (1 << HFLIP_IDX) | (1 << VFLIP_IDX); + sd->hstart = 0; + sd->vstart = 7; + return 0; +} + +static int soi968_init_sensor(struct gspca_dev *gspca_dev) +{ + int i; + struct sd *sd = (struct sd *) gspca_dev; + + for (i = 0; i < ARRAY_SIZE(soi968_init); i++) { + if (i2c_w1(gspca_dev, soi968_init[i][0], + soi968_init[i][1]) < 0) { + err("SOI968 sensor initialization failed"); + return -ENODEV; + } + } + /* disable hflip and vflip */ + gspca_dev->ctrl_dis = (1 << HFLIP_IDX) | (1 << VFLIP_IDX); + sd->hstart = 60; + sd->vstart = 11; + return 0; +} + +static int ov7660_init_sensor(struct gspca_dev *gspca_dev) +{ + int i; + struct sd *sd = (struct sd *) gspca_dev; + + for (i = 0; i < ARRAY_SIZE(ov7660_init); i++) { + if (i2c_w1(gspca_dev, ov7660_init[i][0], + ov7660_init[i][1]) < 0) { + err("OV7660 sensor initialization failed"); + return -ENODEV; + } + } + /* disable hflip and vflip */ + gspca_dev->ctrl_dis = (1 << HFLIP_IDX) | (1 << VFLIP_IDX); + sd->hstart = 1; + sd->vstart = 1; + return 0; +} + +static int ov7670_init_sensor(struct gspca_dev *gspca_dev) +{ + int i; + struct sd *sd = (struct sd *) gspca_dev; + + for (i = 0; i < ARRAY_SIZE(ov7670_init); i++) { + if (i2c_w1(gspca_dev, ov7670_init[i][0], + ov7670_init[i][1]) < 0) { + err("OV7670 sensor initialization failed"); + return -ENODEV; + } + } + /* disable hflip and vflip */ + gspca_dev->ctrl_dis = (1 << HFLIP_IDX) | (1 << VFLIP_IDX); + sd->hstart = 0; + sd->vstart = 1; + return 0; +} + +static int mt9v_init_sensor(struct gspca_dev *gspca_dev) +{ + struct sd *sd = (struct sd *) gspca_dev; + int i; + u16 value; + int ret; + + sd->i2c_addr = 0x5d; + ret = i2c_r2(gspca_dev, 0xff, &value); + if ((ret == 0) && (value == 0x8243)) { + for (i = 0; i < ARRAY_SIZE(mt9v011_init); i++) { + if (i2c_w2(gspca_dev, mt9v011_init[i][0], + mt9v011_init[i][1]) < 0) { + err("MT9V011 sensor initialization failed"); + return -ENODEV; + } + } + sd->hstart = 2; + sd->vstart = 2; + sd->sensor = SENSOR_MT9V011; + info("MT9V011 sensor detected"); + return 0; + } + + sd->i2c_addr = 0x5c; + i2c_w2(gspca_dev, 0x01, 0x0004); + ret = i2c_r2(gspca_dev, 0xff, &value); + if ((ret == 0) && (value == 0x823a)) { + for (i = 0; i < ARRAY_SIZE(mt9v111_init); i++) { + if (i2c_w2(gspca_dev, mt9v111_init[i][0], + mt9v111_init[i][1]) < 0) { + err("MT9V111 sensor initialization failed"); + return -ENODEV; + } + } + sd->hstart = 2; + sd->vstart = 2; + sd->sensor = SENSOR_MT9V111; + info("MT9V111 sensor detected"); + return 0; + } + + sd->i2c_addr = 0x5d; + ret = i2c_w2(gspca_dev, 0xf0, 0x0000); + if (ret < 0) { + sd->i2c_addr = 0x48; + i2c_w2(gspca_dev, 0xf0, 0x0000); + } + ret = i2c_r2(gspca_dev, 0x00, &value); + if ((ret == 0) && (value == 0x1229)) { + for (i = 0; i < ARRAY_SIZE(mt9v112_init); i++) { + if (i2c_w2(gspca_dev, mt9v112_init[i][0], + mt9v112_init[i][1]) < 0) { + err("MT9V112 sensor initialization failed"); + return -ENODEV; + } + } + sd->hstart = 6; + sd->vstart = 2; + sd->sensor = SENSOR_MT9V112; + info("MT9V112 sensor detected"); + return 0; + } + + return -ENODEV; +} + +static int mt9m111_init_sensor(struct gspca_dev *gspca_dev) +{ + struct sd *sd = (struct sd *) gspca_dev; + int i; + for (i = 0; i < ARRAY_SIZE(mt9m111_init); i++) { + if (i2c_w2(gspca_dev, mt9m111_init[i][0], + mt9m111_init[i][1]) < 0) { + err("MT9M111 sensor initialization failed"); + return -ENODEV; + } + } + sd->hstart = 0; + sd->vstart = 2; + return 0; +} + +static int mt9m001_init_sensor(struct gspca_dev *gspca_dev) +{ + struct sd *sd = (struct sd *) gspca_dev; + int i; + for (i = 0; i < ARRAY_SIZE(mt9m001_init); i++) { + if (i2c_w2(gspca_dev, mt9m001_init[i][0], + mt9m001_init[i][1]) < 0) { + err("MT9M001 sensor initialization failed"); + return -ENODEV; + } + } + /* disable hflip and vflip */ + gspca_dev->ctrl_dis = (1 << HFLIP_IDX) | (1 << VFLIP_IDX); + sd->hstart = 2; + sd->vstart = 2; + return 0; +} + +static int hv7131r_init_sensor(struct gspca_dev *gspca_dev) +{ + int i; + struct sd *sd = (struct sd *) gspca_dev; + + for (i = 0; i < ARRAY_SIZE(hv7131r_init); i++) { + if (i2c_w1(gspca_dev, hv7131r_init[i][0], + hv7131r_init[i][1]) < 0) { + err("HV7131R Sensor initialization failed"); + return -ENODEV; + } + } + sd->hstart = 0; + sd->vstart = 1; + return 0; +} + +#ifdef CONFIG_USB_GSPCA_SN9C20X_EVDEV +static int input_kthread(void *data) +{ + struct gspca_dev *gspca_dev = (struct gspca_dev *)data; + struct sd *sd = (struct sd *) gspca_dev; + + DECLARE_WAIT_QUEUE_HEAD(wait); + set_freezable(); + for (;;) { + if (kthread_should_stop()) + break; + + if (reg_r(gspca_dev, 0x1005, 1) < 0) + continue; + + input_report_key(sd->input_dev, + KEY_CAMERA, + gspca_dev->usb_buf[0] & sd->input_gpio); + input_sync(sd->input_dev); + + wait_event_freezable_timeout(wait, + kthread_should_stop(), + msecs_to_jiffies(100)); + } + return 0; +} + + +static int sn9c20x_input_init(struct gspca_dev *gspca_dev) +{ + struct sd *sd = (struct sd *) gspca_dev; + if (sd->input_gpio == 0) + return 0; + + sd->input_dev = input_allocate_device(); + if (!sd->input_dev) + return -ENOMEM; + + sd->input_dev->name = "SN9C20X Webcam"; + + sd->input_dev->phys = kasprintf(GFP_KERNEL, "usb-%s-%s", + gspca_dev->dev->bus->bus_name, + gspca_dev->dev->devpath); + + if (!sd->input_dev->phys) + return -ENOMEM; + + usb_to_input_id(gspca_dev->dev, &sd->input_dev->id); + sd->input_dev->dev.parent = &gspca_dev->dev->dev; + + set_bit(EV_KEY, sd->input_dev->evbit); + set_bit(KEY_CAMERA, sd->input_dev->keybit); + + if (input_register_device(sd->input_dev)) + return -EINVAL; + + sd->input_task = kthread_run(input_kthread, gspca_dev, "sn9c20x/%d", + gspca_dev->vdev.minor); + + if (IS_ERR(sd->input_task)) + return -EINVAL; + + return 0; +} + +static void sn9c20x_input_cleanup(struct gspca_dev *gspca_dev) +{ + struct sd *sd = (struct sd *) gspca_dev; + if (sd->input_task != NULL && !IS_ERR(sd->input_task)) + kthread_stop(sd->input_task); + + if (sd->input_dev != NULL) { + input_unregister_device(sd->input_dev); + kfree(sd->input_dev->phys); + input_free_device(sd->input_dev); + sd->input_dev = NULL; + } +} +#endif + +static int set_cmatrix(struct gspca_dev *gspca_dev) +{ + struct sd *sd = (struct sd *) gspca_dev; + s32 hue_coord, hue_index = 180 + sd->hue; + u8 cmatrix[21]; + memset(cmatrix, 0, 21); + + cmatrix[2] = (sd->contrast * 0x25 / 0x100) + 0x26; + cmatrix[0] = 0x13 + (cmatrix[2] - 0x26) * 0x13 / 0x25; + cmatrix[4] = 0x07 + (cmatrix[2] - 0x26) * 0x07 / 0x25; + cmatrix[18] = sd->brightness - 0x80; + + hue_coord = (hsv_red_x[hue_index] * sd->saturation) >> 8; + cmatrix[6] = (unsigned char)(hue_coord & 0xff); + cmatrix[7] = (unsigned char)((hue_coord >> 8) & 0x0f); + + hue_coord = (hsv_red_y[hue_index] * sd->saturation) >> 8; + cmatrix[8] = (unsigned char)(hue_coord & 0xff); + cmatrix[9] = (unsigned char)((hue_coord >> 8) & 0x0f); + + hue_coord = (hsv_green_x[hue_index] * sd->saturation) >> 8; + cmatrix[10] = (unsigned char)(hue_coord & 0xff); + cmatrix[11] = (unsigned char)((hue_coord >> 8) & 0x0f); + + hue_coord = (hsv_green_y[hue_index] * sd->saturation) >> 8; + cmatrix[12] = (unsigned char)(hue_coord & 0xff); + cmatrix[13] = (unsigned char)((hue_coord >> 8) & 0x0f); + + hue_coord = (hsv_blue_x[hue_index] * sd->saturation) >> 8; + cmatrix[14] = (unsigned char)(hue_coord & 0xff); + cmatrix[15] = (unsigned char)((hue_coord >> 8) & 0x0f); + + hue_coord = (hsv_blue_y[hue_index] * sd->saturation) >> 8; + cmatrix[16] = (unsigned char)(hue_coord & 0xff); + cmatrix[17] = (unsigned char)((hue_coord >> 8) & 0x0f); + + return reg_w(gspca_dev, 0x10e1, cmatrix, 21); +} + +static int set_gamma(struct gspca_dev *gspca_dev) +{ + struct sd *sd = (struct sd *) gspca_dev; + u8 gamma[17]; + u8 gval = sd->gamma * 0xb8 / 0x100; + + + gamma[0] = 0x0a; + gamma[1] = 0x13 + (gval * (0xcb - 0x13) / 0xb8); + gamma[2] = 0x25 + (gval * (0xee - 0x25) / 0xb8); + gamma[3] = 0x37 + (gval * (0xfa - 0x37) / 0xb8); + gamma[4] = 0x45 + (gval * (0xfc - 0x45) / 0xb8); + gamma[5] = 0x55 + (gval * (0xfb - 0x55) / 0xb8); + gamma[6] = 0x65 + (gval * (0xfc - 0x65) / 0xb8); + gamma[7] = 0x74 + (gval * (0xfd - 0x74) / 0xb8); + gamma[8] = 0x83 + (gval * (0xfe - 0x83) / 0xb8); + gamma[9] = 0x92 + (gval * (0xfc - 0x92) / 0xb8); + gamma[10] = 0xa1 + (gval * (0xfc - 0xa1) / 0xb8); + gamma[11] = 0xb0 + (gval * (0xfc - 0xb0) / 0xb8); + gamma[12] = 0xbf + (gval * (0xfb - 0xbf) / 0xb8); + gamma[13] = 0xce + (gval * (0xfb - 0xce) / 0xb8); + gamma[14] = 0xdf + (gval * (0xfd - 0xdf) / 0xb8); + gamma[15] = 0xea + (gval * (0xf9 - 0xea) / 0xb8); + gamma[16] = 0xf5; + + return reg_w(gspca_dev, 0x1190, gamma, 17); +} + +static int set_redblue(struct gspca_dev *gspca_dev) +{ + struct sd *sd = (struct sd *) gspca_dev; + reg_w1(gspca_dev, 0x118c, sd->red); + reg_w1(gspca_dev, 0x118f, sd->blue); + return 0; +} + +static int set_hvflip(struct gspca_dev *gspca_dev) +{ + u8 value, tslb; + u16 value2; + struct sd *sd = (struct sd *) gspca_dev; + switch (sd->sensor) { + case SENSOR_OV9650: + i2c_r1(gspca_dev, 0x1e, &value); + value &= ~0x30; + tslb = 0x01; + if (sd->hflip) + value |= 0x20; + if (sd->vflip) { + value |= 0x10; + tslb = 0x49; + } + i2c_w1(gspca_dev, 0x1e, value); + i2c_w1(gspca_dev, 0x3a, tslb); + break; + case SENSOR_MT9V111: + case SENSOR_MT9V011: + i2c_r2(gspca_dev, 0x20, &value2); + value2 &= ~0xc0a0; + if (sd->hflip) + value2 |= 0x8080; + if (sd->vflip) + value2 |= 0x4020; + i2c_w2(gspca_dev, 0x20, value2); + break; + case SENSOR_MT9M111: + case SENSOR_MT9V112: + i2c_r2(gspca_dev, 0x20, &value2); + value2 &= ~0x0003; + if (sd->hflip) + value2 |= 0x0002; + if (sd->vflip) + value2 |= 0x0001; + i2c_w2(gspca_dev, 0x20, value2); + break; + case SENSOR_HV7131R: + i2c_r1(gspca_dev, 0x01, &value); + value &= ~0x03; + if (sd->vflip) + value |= 0x01; + if (sd->hflip) + value |= 0x02; + i2c_w1(gspca_dev, 0x01, value); + break; + } + return 0; +} + +static int set_exposure(struct gspca_dev *gspca_dev) +{ + struct sd *sd = (struct sd *) gspca_dev; + u8 exp[8] = {0x81, sd->i2c_addr, 0x00, 0x00, 0x00, 0x00, 0x00, 0x1e}; + switch (sd->sensor) { + case SENSOR_OV7660: + case SENSOR_OV7670: + case SENSOR_SOI968: + case SENSOR_OV9655: + case SENSOR_OV9650: + exp[0] |= (3 << 4); + exp[2] = 0x2d; + exp[3] = sd->exposure & 0xff; + exp[4] = sd->exposure >> 8; + break; + case SENSOR_MT9M001: + case SENSOR_MT9M111: + case SENSOR_MT9V112: + case SENSOR_MT9V111: + case SENSOR_MT9V011: + exp[0] |= (3 << 4); + exp[2] = 0x09; + exp[3] = sd->exposure >> 8; + exp[4] = sd->exposure & 0xff; + break; + case SENSOR_HV7131R: + exp[0] |= (4 << 4); + exp[2] = 0x25; + exp[3] = ((sd->exposure * 0xffffff) / 0xffff) >> 16; + exp[4] = ((sd->exposure * 0xffffff) / 0xffff) >> 8; + exp[5] = ((sd->exposure * 0xffffff) / 0xffff) & 0xff; + break; + } + i2c_w(gspca_dev, exp); + return 0; +} + +static int set_gain(struct gspca_dev *gspca_dev) +{ + struct sd *sd = (struct sd *) gspca_dev; + u8 gain[8] = {0x81, sd->i2c_addr, 0x00, 0x00, 0x00, 0x00, 0x00, 0x1d}; + switch (sd->sensor) { + case SENSOR_OV7660: + case SENSOR_OV7670: + case SENSOR_SOI968: + case SENSOR_OV9655: + case SENSOR_OV9650: + gain[0] |= (2 << 4); + gain[3] = ov_gain[sd->gain]; + break; + case SENSOR_MT9V011: + case SENSOR_MT9V111: + gain[0] |= (3 << 4); + gain[2] = 0x35; + gain[3] = micron1_gain[sd->gain] >> 8; + gain[4] = micron1_gain[sd->gain] & 0xff; + break; + case SENSOR_MT9V112: + case SENSOR_MT9M111: + gain[0] |= (3 << 4); + gain[2] = 0x2f; + gain[3] = micron1_gain[sd->gain] >> 8; + gain[4] = micron1_gain[sd->gain] & 0xff; + break; + case SENSOR_MT9M001: + gain[0] |= (3 << 4); + gain[2] = 0x2f; + gain[3] = micron2_gain[sd->gain] >> 8; + gain[4] = micron2_gain[sd->gain] & 0xff; + break; + case SENSOR_HV7131R: + gain[0] |= (2 << 4); + gain[2] = 0x30; + gain[3] = hv7131r_gain[sd->gain]; + break; + } + i2c_w(gspca_dev, gain); + return 0; +} + +static int sd_setbrightness(struct gspca_dev *gspca_dev, s32 val) +{ + struct sd *sd = (struct sd *) gspca_dev; + + sd->brightness = val; + if (gspca_dev->streaming) + return set_cmatrix(gspca_dev); + return 0; +} + +static int sd_getbrightness(struct gspca_dev *gspca_dev, s32 *val) +{ + struct sd *sd = (struct sd *) gspca_dev; + *val = sd->brightness; + return 0; +} + + +static int sd_setcontrast(struct gspca_dev *gspca_dev, s32 val) +{ + struct sd *sd = (struct sd *) gspca_dev; + + sd->contrast = val; + if (gspca_dev->streaming) + return set_cmatrix(gspca_dev); + return 0; +} + +static int sd_getcontrast(struct gspca_dev *gspca_dev, s32 *val) +{ + struct sd *sd = (struct sd *) gspca_dev; + *val = sd->contrast; + return 0; +} + +static int sd_setsaturation(struct gspca_dev *gspca_dev, s32 val) +{ + struct sd *sd = (struct sd *) gspca_dev; + + sd->saturation = val; + if (gspca_dev->streaming) + return set_cmatrix(gspca_dev); + return 0; +} + +static int sd_getsaturation(struct gspca_dev *gspca_dev, s32 *val) +{ + struct sd *sd = (struct sd *) gspca_dev; + *val = sd->saturation; + return 0; +} + +static int sd_sethue(struct gspca_dev *gspca_dev, s32 val) +{ + struct sd *sd = (struct sd *) gspca_dev; + + sd->hue = val; + if (gspca_dev->streaming) + return set_cmatrix(gspca_dev); + return 0; +} + +static int sd_gethue(struct gspca_dev *gspca_dev, s32 *val) +{ + struct sd *sd = (struct sd *) gspca_dev; + *val = sd->hue; + return 0; +} + +static int sd_setgamma(struct gspca_dev *gspca_dev, s32 val) +{ + struct sd *sd = (struct sd *) gspca_dev; + + sd->gamma = val; + if (gspca_dev->streaming) + return set_gamma(gspca_dev); + return 0; +} + +static int sd_getgamma(struct gspca_dev *gspca_dev, s32 *val) +{ + struct sd *sd = (struct sd *) gspca_dev; + *val = sd->gamma; + return 0; +} + +static int sd_setredbalance(struct gspca_dev *gspca_dev, s32 val) +{ + struct sd *sd = (struct sd *) gspca_dev; + + sd->red = val; + if (gspca_dev->streaming) + return set_redblue(gspca_dev); + return 0; +} + +static int sd_getredbalance(struct gspca_dev *gspca_dev, s32 *val) +{ + struct sd *sd = (struct sd *) gspca_dev; + *val = sd->red; + return 0; +} + +static int sd_setbluebalance(struct gspca_dev *gspca_dev, s32 val) +{ + struct sd *sd = (struct sd *) gspca_dev; + + sd->blue = val; + if (gspca_dev->streaming) + return set_redblue(gspca_dev); + return 0; +} + +static int sd_getbluebalance(struct gspca_dev *gspca_dev, s32 *val) +{ + struct sd *sd = (struct sd *) gspca_dev; + *val = sd->blue; + return 0; +} + +static int sd_sethflip(struct gspca_dev *gspca_dev, s32 val) +{ + struct sd *sd = (struct sd *) gspca_dev; + + sd->hflip = val; + if (gspca_dev->streaming) + return set_hvflip(gspca_dev); + return 0; +} + +static int sd_gethflip(struct gspca_dev *gspca_dev, s32 *val) +{ + struct sd *sd = (struct sd *) gspca_dev; + *val = sd->hflip; + return 0; +} + +static int sd_setvflip(struct gspca_dev *gspca_dev, s32 val) +{ + struct sd *sd = (struct sd *) gspca_dev; + + sd->vflip = val; + if (gspca_dev->streaming) + return set_hvflip(gspca_dev); + return 0; +} + +static int sd_getvflip(struct gspca_dev *gspca_dev, s32 *val) +{ + struct sd *sd = (struct sd *) gspca_dev; + *val = sd->vflip; + return 0; +} + +static int sd_setexposure(struct gspca_dev *gspca_dev, s32 val) +{ + struct sd *sd = (struct sd *) gspca_dev; + + sd->exposure = val; + if (gspca_dev->streaming) + return set_exposure(gspca_dev); + return 0; +} + +static int sd_getexposure(struct gspca_dev *gspca_dev, s32 *val) +{ + struct sd *sd = (struct sd *) gspca_dev; + *val = sd->exposure; + return 0; +} + +static int sd_setgain(struct gspca_dev *gspca_dev, s32 val) +{ + struct sd *sd = (struct sd *) gspca_dev; + + sd->gain = val; + if (gspca_dev->streaming) + return set_gain(gspca_dev); + return 0; +} + +static int sd_getgain(struct gspca_dev *gspca_dev, s32 *val) +{ + struct sd *sd = (struct sd *) gspca_dev; + *val = sd->gain; + return 0; +} + +static int sd_setautoexposure(struct gspca_dev *gspca_dev, s32 val) +{ + struct sd *sd = (struct sd *) gspca_dev; + sd->auto_exposure = val; + return 0; +} + +static int sd_getautoexposure(struct gspca_dev *gspca_dev, s32 *val) +{ + struct sd *sd = (struct sd *) gspca_dev; + *val = sd->auto_exposure; + return 0; +} + +#ifdef CONFIG_VIDEO_ADV_DEBUG +static int sd_dbg_g_register(struct gspca_dev *gspca_dev, + struct v4l2_dbg_register *reg) +{ + struct sd *sd = (struct sd *) gspca_dev; + switch (reg->match.type) { + case V4L2_CHIP_MATCH_HOST: + if (reg->match.addr != 0) + return -EINVAL; + if (reg->reg < 0x1000 || reg->reg > 0x11ff) + return -EINVAL; + if (reg_r(gspca_dev, reg->reg, 1) < 0) + return -EINVAL; + reg->val = gspca_dev->usb_buf[0]; + return 0; + case V4L2_CHIP_MATCH_I2C_ADDR: + if (reg->match.addr != sd->i2c_addr) + return -EINVAL; + if (sd->sensor >= SENSOR_MT9V011 && + sd->sensor <= SENSOR_MT9M111) { + if (i2c_r2(gspca_dev, reg->reg, (u16 *)®->val) < 0) + return -EINVAL; + } else { + if (i2c_r1(gspca_dev, reg->reg, (u8 *)®->val) < 0) + return -EINVAL; + } + return 0; + } + return -EINVAL; +} + +static int sd_dbg_s_register(struct gspca_dev *gspca_dev, + struct v4l2_dbg_register *reg) +{ + struct sd *sd = (struct sd *) gspca_dev; + switch (reg->match.type) { + case V4L2_CHIP_MATCH_HOST: + if (reg->match.addr != 0) + return -EINVAL; + if (reg->reg < 0x1000 || reg->reg > 0x11ff) + return -EINVAL; + if (reg_w1(gspca_dev, reg->reg, reg->val) < 0) + return -EINVAL; + return 0; + case V4L2_CHIP_MATCH_I2C_ADDR: + if (reg->match.addr != sd->i2c_addr) + return -EINVAL; + if (sd->sensor >= SENSOR_MT9V011 && + sd->sensor <= SENSOR_MT9M111) { + if (i2c_w2(gspca_dev, reg->reg, reg->val) < 0) + return -EINVAL; + } else { + if (i2c_w1(gspca_dev, reg->reg, reg->val) < 0) + return -EINVAL; + } + return 0; + } + return -EINVAL; +} +#endif + +static int sd_chip_ident(struct gspca_dev *gspca_dev, + struct v4l2_dbg_chip_ident *chip) +{ + struct sd *sd = (struct sd *) gspca_dev; + + switch (chip->match.type) { + case V4L2_CHIP_MATCH_HOST: + if (chip->match.addr != 0) + return -EINVAL; + chip->revision = 0; + chip->ident = V4L2_IDENT_SN9C20X; + return 0; + case V4L2_CHIP_MATCH_I2C_ADDR: + if (chip->match.addr != sd->i2c_addr) + return -EINVAL; + chip->revision = 0; + chip->ident = i2c_ident[sd->sensor]; + return 0; + } + return -EINVAL; +} + +static int sd_config(struct gspca_dev *gspca_dev, + const struct usb_device_id *id) +{ + struct sd *sd = (struct sd *) gspca_dev; + struct cam *cam; + + cam = &gspca_dev->cam; + + sd->sensor = (id->driver_info >> 8) & 0xff; + sd->i2c_addr = id->driver_info & 0xff; + + switch (sd->sensor) { + case SENSOR_OV9650: + cam->cam_mode = sxga_mode; + cam->nmodes = ARRAY_SIZE(sxga_mode); + break; + default: + cam->cam_mode = vga_mode; + cam->nmodes = ARRAY_SIZE(vga_mode); + } + + sd->old_step = 0; + sd->older_step = 0; + sd->exposure_step = 16; + + sd->brightness = BRIGHTNESS_DEFAULT; + sd->contrast = CONTRAST_DEFAULT; + sd->saturation = SATURATION_DEFAULT; + sd->hue = HUE_DEFAULT; + sd->gamma = GAMMA_DEFAULT; + sd->red = RED_DEFAULT; + sd->blue = BLUE_DEFAULT; + + sd->hflip = HFLIP_DEFAULT; + sd->vflip = VFLIP_DEFAULT; + sd->exposure = EXPOSURE_DEFAULT; + sd->gain = GAIN_DEFAULT; + sd->auto_exposure = AUTO_EXPOSURE_DEFAULT; + + sd->quality = 95; + +#ifdef CONFIG_USB_GSPCA_SN9C20X_EVDEV + sd->input_gpio = (id->driver_info >> 16) & 0xff; + if (sn9c20x_input_init(gspca_dev) < 0) + return -ENODEV; +#endif + return 0; +} + +static int sd_init(struct gspca_dev *gspca_dev) +{ + struct sd *sd = (struct sd *) gspca_dev; + int i; + u8 value; + u8 i2c_init[9] = + {0x80, sd->i2c_addr, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x03}; + + for (i = 0; i < ARRAY_SIZE(bridge_init); i++) { + value = bridge_init[i][1]; + if (reg_w(gspca_dev, bridge_init[i][0], &value, 1) < 0) { + err("Device initialization failed"); + return -ENODEV; + } + } + + if (reg_w(gspca_dev, 0x10c0, i2c_init, 9) < 0) { + err("Device initialization failed"); + return -ENODEV; + } + + switch (sd->sensor) { + case SENSOR_OV9650: + if (ov9650_init_sensor(gspca_dev) < 0) + return -ENODEV; + info("OV9650 sensor detected"); + break; + case SENSOR_OV9655: + if (ov9655_init_sensor(gspca_dev) < 0) + return -ENODEV; + info("OV9655 sensor detected"); + break; + case SENSOR_SOI968: + if (soi968_init_sensor(gspca_dev) < 0) + return -ENODEV; + info("SOI968 sensor detected"); + break; + case SENSOR_OV7660: + if (ov7660_init_sensor(gspca_dev) < 0) + return -ENODEV; + info("OV7660 sensor detected"); + break; + case SENSOR_OV7670: + if (ov7670_init_sensor(gspca_dev) < 0) + return -ENODEV; + info("OV7670 sensor detected"); + break; + case SENSOR_MT9VPRB: + if (mt9v_init_sensor(gspca_dev) < 0) + return -ENODEV; + break; + case SENSOR_MT9M111: + if (mt9m111_init_sensor(gspca_dev) < 0) + return -ENODEV; + info("MT9M111 sensor detected"); + break; + case SENSOR_MT9M001: + if (mt9m001_init_sensor(gspca_dev) < 0) + return -ENODEV; + info("MT9M001 sensor detected"); + break; + case SENSOR_HV7131R: + if (hv7131r_init_sensor(gspca_dev) < 0) + return -ENODEV; + info("HV7131R sensor detected"); + break; + default: + info("Unsupported Sensor"); + return -ENODEV; + } + + return 0; +} + +static void configure_sensor_output(struct gspca_dev *gspca_dev, int mode) +{ + struct sd *sd = (struct sd *) gspca_dev; + u8 value; + switch (sd->sensor) { + case SENSOR_OV9650: + if (mode & MODE_SXGA) { + i2c_w1(gspca_dev, 0x17, 0x1b); + i2c_w1(gspca_dev, 0x18, 0xbc); + i2c_w1(gspca_dev, 0x19, 0x01); + i2c_w1(gspca_dev, 0x1a, 0x82); + i2c_r1(gspca_dev, 0x12, &value); + i2c_w1(gspca_dev, 0x12, value & 0x07); + } else { + i2c_w1(gspca_dev, 0x17, 0x24); + i2c_w1(gspca_dev, 0x18, 0xc5); + i2c_w1(gspca_dev, 0x19, 0x00); + i2c_w1(gspca_dev, 0x1a, 0x3c); + i2c_r1(gspca_dev, 0x12, &value); + i2c_w1(gspca_dev, 0x12, (value & 0x7) | 0x40); + } + break; + } +} + +#define HW_WIN(mode, hstart, vstart) \ +((const u8 []){hstart & 0xff, hstart >> 8, \ +vstart & 0xff, vstart >> 8, \ +(mode & MODE_SXGA ? 1280 >> 4 : 640 >> 4), \ +(mode & MODE_SXGA ? 1024 >> 3 : 480 >> 3)}) + +#define CLR_WIN(width, height) \ +((const u8 [])\ +{0, width >> 2, 0, height >> 1,\ +((width >> 10) & 0x01) | ((height >> 8) & 0x6)}) + +static int sd_start(struct gspca_dev *gspca_dev) +{ + struct sd *sd = (struct sd *) gspca_dev; + int mode = gspca_dev->cam.cam_mode[(int) gspca_dev->curr_mode].priv; + int width = gspca_dev->width; + int height = gspca_dev->height; + u8 fmt, scale = 0; + + sd->jpeg_hdr = kmalloc(JPEG_HDR_SZ, GFP_KERNEL); + if (sd->jpeg_hdr == NULL) + return -ENOMEM; + + jpeg_define(sd->jpeg_hdr, height, width, + 0x21); + jpeg_set_qual(sd->jpeg_hdr, sd->quality); + + if (mode & MODE_RAW) + fmt = 0x2d; + else if (mode & MODE_JPEG) + fmt = 0x2c; + else + fmt = 0x2f; + + switch (mode & 0x0f) { + case 3: + scale = 0xc0; + info("Set 1280x1024"); + break; + case 2: + scale = 0x80; + info("Set 640x480"); + break; + case 1: + scale = 0x90; + info("Set 320x240"); + break; + case 0: + scale = 0xa0; + info("Set 160x120"); + break; + } + + configure_sensor_output(gspca_dev, mode); + reg_w(gspca_dev, 0x1100, sd->jpeg_hdr + JPEG_QT0_OFFSET, 64); + reg_w(gspca_dev, 0x1140, sd->jpeg_hdr + JPEG_QT1_OFFSET, 64); + reg_w(gspca_dev, 0x10fb, CLR_WIN(width, height), 5); + reg_w(gspca_dev, 0x1180, HW_WIN(mode, sd->hstart, sd->vstart), 6); + reg_w1(gspca_dev, 0x1189, scale); + reg_w1(gspca_dev, 0x10e0, fmt); + + set_cmatrix(gspca_dev); + set_gamma(gspca_dev); + set_redblue(gspca_dev); + set_gain(gspca_dev); + set_exposure(gspca_dev); + set_hvflip(gspca_dev); + + reg_r(gspca_dev, 0x1061, 1); + reg_w1(gspca_dev, 0x1061, gspca_dev->usb_buf[0] | 0x02); + return 0; +} + +static void sd_stopN(struct gspca_dev *gspca_dev) +{ + reg_r(gspca_dev, 0x1061, 1); + reg_w1(gspca_dev, 0x1061, gspca_dev->usb_buf[0] & ~0x02); +} + +static void sd_stop0(struct gspca_dev *gspca_dev) +{ + struct sd *sd = (struct sd *) gspca_dev; + kfree(sd->jpeg_hdr); +} + +static void do_autoexposure(struct gspca_dev *gspca_dev) +{ + struct sd *sd = (struct sd *) gspca_dev; + int avg_lum, new_exp; + + if (!sd->auto_exposure) + return; + + avg_lum = atomic_read(&sd->avg_lum); + + /* + * some hardcoded values are present + * like those for maximal/minimal exposure + * and exposure steps + */ + if (avg_lum < MIN_AVG_LUM) { + if (sd->exposure > 0x1770) + return; + + new_exp = sd->exposure + sd->exposure_step; + if (new_exp > 0x1770) + new_exp = 0x1770; + if (new_exp < 0x10) + new_exp = 0x10; + sd->exposure = new_exp; + set_exposure(gspca_dev); + + sd->older_step = sd->old_step; + sd->old_step = 1; + + if (sd->old_step ^ sd->older_step) + sd->exposure_step /= 2; + else + sd->exposure_step += 2; + } + if (avg_lum > MAX_AVG_LUM) { + if (sd->exposure < 0x10) + return; + new_exp = sd->exposure - sd->exposure_step; + if (new_exp > 0x1700) + new_exp = 0x1770; + if (new_exp < 0x10) + new_exp = 0x10; + sd->exposure = new_exp; + set_exposure(gspca_dev); + sd->older_step = sd->old_step; + sd->old_step = 0; + + if (sd->old_step ^ sd->older_step) + sd->exposure_step /= 2; + else + sd->exposure_step += 2; + } +} + +static void sd_pkt_scan(struct gspca_dev *gspca_dev, + struct gspca_frame *frame, /* target */ + u8 *data, /* isoc packet */ + int len) /* iso packet length */ +{ + struct sd *sd = (struct sd *) gspca_dev; + int avg_lum; + static unsigned char frame_header[] = + {0xff, 0xff, 0x00, 0xc4, 0xc4, 0x96}; + if (len == 64 && memcmp(data, frame_header, 6) == 0) { + avg_lum = ((data[35] >> 2) & 3) | + (data[20] << 2) | + (data[19] << 10); + avg_lum += ((data[35] >> 4) & 3) | + (data[22] << 2) | + (data[21] << 10); + avg_lum += ((data[35] >> 6) & 3) | + (data[24] << 2) | + (data[23] << 10); + avg_lum += (data[36] & 3) | + (data[26] << 2) | + (data[25] << 10); + avg_lum += ((data[36] >> 2) & 3) | + (data[28] << 2) | + (data[27] << 10); + avg_lum += ((data[36] >> 4) & 3) | + (data[30] << 2) | + (data[29] << 10); + avg_lum += ((data[36] >> 6) & 3) | + (data[32] << 2) | + (data[31] << 10); + avg_lum += ((data[44] >> 4) & 3) | + (data[34] << 2) | + (data[33] << 10); + avg_lum >>= 9; + atomic_set(&sd->avg_lum, avg_lum); + gspca_frame_add(gspca_dev, LAST_PACKET, + frame, data, len); + return; + } + if (gspca_dev->last_packet_type == LAST_PACKET) { + if (gspca_dev->cam.cam_mode[(int) gspca_dev->curr_mode].priv + & MODE_JPEG) { + gspca_frame_add(gspca_dev, FIRST_PACKET, frame, + sd->jpeg_hdr, JPEG_HDR_SZ); + gspca_frame_add(gspca_dev, INTER_PACKET, frame, + data, len); + } else { + gspca_frame_add(gspca_dev, FIRST_PACKET, frame, + data, len); + } + } else { + gspca_frame_add(gspca_dev, INTER_PACKET, frame, data, len); + } +} + +/* sub-driver description */ +static const struct sd_desc sd_desc = { + .name = MODULE_NAME, + .ctrls = sd_ctrls, + .nctrls = ARRAY_SIZE(sd_ctrls), + .config = sd_config, + .init = sd_init, + .start = sd_start, + .stopN = sd_stopN, + .stop0 = sd_stop0, + .pkt_scan = sd_pkt_scan, + .dq_callback = do_autoexposure, +#ifdef CONFIG_VIDEO_ADV_DEBUG + .set_register = sd_dbg_s_register, + .get_register = sd_dbg_g_register, +#endif + .get_chip_ident = sd_chip_ident, +}; + +#define SN9C20X(sensor, i2c_addr, button_mask) \ + .driver_info = (button_mask << 16) \ + | (SENSOR_ ## sensor << 8) \ + | (i2c_addr) + +static const __devinitdata struct usb_device_id device_table[] = { + {USB_DEVICE(0x0c45, 0x6240), SN9C20X(MT9M001, 0x5d, 0)}, + {USB_DEVICE(0x0c45, 0x6242), SN9C20X(MT9M111, 0x5d, 0)}, + {USB_DEVICE(0x0c45, 0x6248), SN9C20X(OV9655, 0x30, 0)}, + {USB_DEVICE(0x0c45, 0x624e), SN9C20X(SOI968, 0x30, 0x10)}, + {USB_DEVICE(0x0c45, 0x624f), SN9C20X(OV9650, 0x30, 0)}, + {USB_DEVICE(0x0c45, 0x6251), SN9C20X(OV9650, 0x30, 0)}, + {USB_DEVICE(0x0c45, 0x6253), SN9C20X(OV9650, 0x30, 0)}, + {USB_DEVICE(0x0c45, 0x6260), SN9C20X(OV7670, 0x21, 0)}, + {USB_DEVICE(0x0c45, 0x6270), SN9C20X(MT9VPRB, 0x00, 0)}, + {USB_DEVICE(0x0c45, 0x627b), SN9C20X(OV7660, 0x21, 0)}, + {USB_DEVICE(0x0c45, 0x627c), SN9C20X(HV7131R, 0x11, 0)}, + {USB_DEVICE(0x0c45, 0x627f), SN9C20X(OV9650, 0x30, 0)}, + {USB_DEVICE(0x0c45, 0x6280), SN9C20X(MT9M001, 0x5d, 0)}, + {USB_DEVICE(0x0c45, 0x6282), SN9C20X(MT9M111, 0x5d, 0)}, + {USB_DEVICE(0x0c45, 0x6288), SN9C20X(OV9655, 0x30, 0)}, + {USB_DEVICE(0x0c45, 0x628e), SN9C20X(SOI968, 0x30, 0)}, + {USB_DEVICE(0x0c45, 0x628f), SN9C20X(OV9650, 0x30, 0)}, + {USB_DEVICE(0x0c45, 0x62a0), SN9C20X(OV7670, 0x21, 0)}, + {USB_DEVICE(0x0c45, 0x62b0), SN9C20X(MT9VPRB, 0x00, 0)}, + {USB_DEVICE(0x0c45, 0x62b3), SN9C20X(OV9655, 0x30, 0)}, + {USB_DEVICE(0x0c45, 0x62bb), SN9C20X(OV7660, 0x21, 0)}, + {USB_DEVICE(0x0c45, 0x62bc), SN9C20X(HV7131R, 0x11, 0)}, + {USB_DEVICE(0x045e, 0x00f4), SN9C20X(OV9650, 0x30, 0)}, + {USB_DEVICE(0x145f, 0x013d), SN9C20X(OV7660, 0x21, 0)}, + {USB_DEVICE(0x0458, 0x7029), SN9C20X(HV7131R, 0x11, 0)}, + {USB_DEVICE(0xa168, 0x0610), SN9C20X(HV7131R, 0x11, 0)}, + {USB_DEVICE(0xa168, 0x0611), SN9C20X(HV7131R, 0x11, 0)}, + {USB_DEVICE(0xa168, 0x0613), SN9C20X(HV7131R, 0x11, 0)}, + {USB_DEVICE(0xa168, 0x0618), SN9C20X(HV7131R, 0x11, 0)}, + {USB_DEVICE(0xa168, 0x0614), SN9C20X(MT9M111, 0x5d, 0)}, + {USB_DEVICE(0xa168, 0x0615), SN9C20X(MT9M111, 0x5d, 0)}, + {USB_DEVICE(0xa168, 0x0617), SN9C20X(MT9M111, 0x5d, 0)}, + {} +}; +MODULE_DEVICE_TABLE(usb, device_table); + +/* -- device connect -- */ +static int sd_probe(struct usb_interface *intf, + const struct usb_device_id *id) +{ + return gspca_dev_probe(intf, id, &sd_desc, sizeof(struct sd), + THIS_MODULE); +} + +static void sd_disconnect(struct usb_interface *intf) +{ +#ifdef CONFIG_USB_GSPCA_SN9C20X_EVDEV + struct gspca_dev *gspca_dev = usb_get_intfdata(intf); + + sn9c20x_input_cleanup(gspca_dev); +#endif + + gspca_disconnect(intf); +} + +static struct usb_driver sd_driver = { + .name = MODULE_NAME, + .id_table = device_table, + .probe = sd_probe, + .disconnect = sd_disconnect, +#ifdef CONFIG_PM + .suspend = gspca_suspend, + .resume = gspca_resume, + .reset_resume = gspca_resume, +#endif +}; + +/* -- module insert / remove -- */ +static int __init sd_mod_init(void) +{ + int ret; + ret = usb_register(&sd_driver); + if (ret < 0) + return ret; + info("registered"); + return 0; +} +static void __exit sd_mod_exit(void) +{ + usb_deregister(&sd_driver); + info("deregistered"); +} + +module_init(sd_mod_init); +module_exit(sd_mod_exit); diff --git a/include/linux/videodev2.h b/include/linux/videodev2.h index 95846d988011..74f16876f38d 100644 --- a/include/linux/videodev2.h +++ b/include/linux/videodev2.h @@ -338,6 +338,7 @@ struct v4l2_pix_format { /* Vendor-specific formats */ #define V4L2_PIX_FMT_WNVA v4l2_fourcc('W', 'N', 'V', 'A') /* Winnov hw compress */ #define V4L2_PIX_FMT_SN9C10X v4l2_fourcc('S', '9', '1', '0') /* SN9C10x compression */ +#define V4L2_PIX_FMT_SN9C20X_I420 v4l2_fourcc('S', '9', '2', '0') /* SN9C20x YUV 4:2:0 */ #define V4L2_PIX_FMT_PWC1 v4l2_fourcc('P', 'W', 'C', '1') /* pwc older webcam */ #define V4L2_PIX_FMT_PWC2 v4l2_fourcc('P', 'W', 'C', '2') /* pwc newer webcam */ #define V4L2_PIX_FMT_ET61X251 v4l2_fourcc('E', '6', '2', '5') /* ET61X251 compression */ diff --git a/include/media/v4l2-chip-ident.h b/include/media/v4l2-chip-ident.h index 11a4a2d3e364..94e908c0d7a0 100644 --- a/include/media/v4l2-chip-ident.h +++ b/include/media/v4l2-chip-ident.h @@ -60,6 +60,10 @@ enum { V4L2_IDENT_OV7670 = 250, V4L2_IDENT_OV7720 = 251, V4L2_IDENT_OV7725 = 252, + V4L2_IDENT_OV7660 = 253, + V4L2_IDENT_OV9650 = 254, + V4L2_IDENT_OV9655 = 255, + V4L2_IDENT_SOI968 = 256, /* module saa7146: reserved range 300-309 */ V4L2_IDENT_SAA7146 = 300, @@ -161,6 +165,9 @@ enum { /* module tw9910: just ident 9910 */ V4L2_IDENT_TW9910 = 9910, + /* module sn9c20x: just ident 10000 */ + V4L2_IDENT_SN9C20X = 10000, + /* module msp3400: reserved range 34000-34999 and 44000-44999 */ V4L2_IDENT_MSPX4XX = 34000, /* generic MSPX4XX identifier, only use internally (tveeprom.c). */ @@ -237,6 +244,11 @@ enum { V4L2_IDENT_MT9V022IX7ATC = 45010, /* No way to detect "normal" I77ATx */ V4L2_IDENT_MT9V022IX7ATM = 45015, /* and "lead free" IA7ATx chips */ V4L2_IDENT_MT9T031 = 45020, + V4L2_IDENT_MT9V111 = 45031, + V4L2_IDENT_MT9V112 = 45032, + + /* HV7131R CMOS sensor: just ident 46000 */ + V4L2_IDENT_HV7131R = 46000, /* module cs53132a: just ident 53132 */ V4L2_IDENT_CS53l32A = 53132, -- cgit v1.2.3 From b9454e83cac42fcdc90bfbfba479132bd6629455 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Wed, 8 Jul 2009 13:29:08 +0200 Subject: nl80211: introduce new key attributes We will soon want to nest key attributes into some new attribute for configuring static WEP keys at connect() and ibss_join() time, so we need nested attributes for that. However, key attributes right now are 'global'. This patch thus introduces new nested attributes for the key settings and functions for parsing them. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- include/linux/nl80211.h | 36 +++++++++ net/wireless/nl80211.c | 203 +++++++++++++++++++++++++++++++++++++----------- 2 files changed, 192 insertions(+), 47 deletions(-) (limited to 'include') diff --git a/include/linux/nl80211.h b/include/linux/nl80211.h index e496a2daf7ef..48e0913c2209 100644 --- a/include/linux/nl80211.h +++ b/include/linux/nl80211.h @@ -567,6 +567,9 @@ enum nl80211_commands { * @NL80211_ATTR_PREV_BSSID: previous BSSID, to be used by in ASSOCIATE * commands to specify using a reassociate frame * + * @NL80211_ATTR_KEY: key information in a nested attribute with + * %NL80211_KEY_* sub-attributes + * * @NL80211_ATTR_MAX: highest attribute number currently defined * @__NL80211_ATTR_AFTER_LAST: internal use */ @@ -692,6 +695,8 @@ enum nl80211_attrs { NL80211_ATTR_PREV_BSSID, + NL80211_ATTR_KEY, + /* add attributes here, update the policy in nl80211.c */ __NL80211_ATTR_AFTER_LAST, @@ -720,6 +725,7 @@ enum nl80211_attrs { #define NL80211_ATTR_CIPHER_SUITE_GROUP NL80211_ATTR_CIPHER_SUITE_GROUP #define NL80211_ATTR_WPA_VERSIONS NL80211_ATTR_WPA_VERSIONS #define NL80211_ATTR_AKM_SUITES NL80211_ATTR_AKM_SUITES +#define NL80211_ATTR_KEY NL80211_ATTR_KEY #define NL80211_MAX_SUPP_RATES 32 #define NL80211_MAX_SUPP_REG_RULES 32 @@ -1320,4 +1326,34 @@ enum nl80211_wpa_versions { NL80211_WPA_VERSION_2 = 1 << 1, }; +/** + * enum nl80211_key_attributes - key attributes + * @__NL80211_KEY_INVALID: invalid + * @NL80211_KEY_DATA: (temporal) key data; for TKIP this consists of + * 16 bytes encryption key followed by 8 bytes each for TX and RX MIC + * keys + * @NL80211_KEY_IDX: key ID (u8, 0-3) + * @NL80211_KEY_CIPHER: key cipher suite (u32, as defined by IEEE 802.11 + * section 7.3.2.25.1, e.g. 0x000FAC04) + * @NL80211_KEY_SEQ: transmit key sequence number (IV/PN) for TKIP and + * CCMP keys, each six bytes in little endian + * @NL80211_KEY_DEFAULT: flag indicating default key + * @NL80211_KEY_DEFAULT_MGMT: flag indicating default management key + * @__NL80211_KEY_AFTER_LAST: internal + * @NL80211_KEY_MAX: highest key attribute + */ +enum nl80211_key_attributes { + __NL80211_KEY_INVALID, + NL80211_KEY_DATA, + NL80211_KEY_IDX, + NL80211_KEY_CIPHER, + NL80211_KEY_SEQ, + NL80211_KEY_DEFAULT, + NL80211_KEY_DEFAULT_MGMT, + + /* keep last */ + __NL80211_KEY_AFTER_LAST, + NL80211_KEY_MAX = __NL80211_KEY_AFTER_LAST - 1 +}; + #endif /* __LINUX_NL80211_H */ diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index a00fd644370b..50cf59316292 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -73,6 +73,7 @@ static struct nla_policy nl80211_policy[NL80211_ATTR_MAX+1] __read_mostly = { [NL80211_ATTR_MAC] = { .type = NLA_BINARY, .len = ETH_ALEN }, [NL80211_ATTR_PREV_BSSID] = { .type = NLA_BINARY, .len = ETH_ALEN }, + [NL80211_ATTR_KEY] = { .type = NLA_NESTED, }, [NL80211_ATTR_KEY_DATA] = { .type = NLA_BINARY, .len = WLAN_MAX_KEY_LEN }, [NL80211_ATTR_KEY_IDX] = { .type = NLA_U8 }, @@ -134,6 +135,18 @@ static struct nla_policy nl80211_policy[NL80211_ATTR_MAX+1] __read_mostly = { [NL80211_ATTR_WPA_VERSIONS] = { .type = NLA_U32 }, }; +/* policy for the attributes */ +static struct nla_policy +nl80211_key_policy[NL80211_KEY_MAX + 1] __read_mostly = { + [NL80211_KEY_DATA] = { .type = NLA_BINARY, + .len = WLAN_MAX_KEY_LEN }, + [NL80211_KEY_IDX] = { .type = NLA_U8 }, + [NL80211_KEY_CIPHER] = { .type = NLA_U32 }, + [NL80211_KEY_SEQ] = { .type = NLA_BINARY, .len = 8 }, + [NL80211_KEY_DEFAULT] = { .type = NLA_FLAG }, + [NL80211_KEY_DEFAULT_MGMT] = { .type = NLA_FLAG }, +}; + /* IE validation */ static bool is_valid_ie_attr(const struct nlattr *attr) { @@ -198,6 +211,100 @@ static int nl80211_msg_put_channel(struct sk_buff *msg, /* netlink command implementations */ +struct key_parse { + struct key_params p; + int idx; + bool def, defmgmt; +}; + +static int nl80211_parse_key_new(struct nlattr *key, struct key_parse *k) +{ + struct nlattr *tb[NL80211_KEY_MAX + 1]; + int err = nla_parse_nested(tb, NL80211_KEY_MAX, key, + nl80211_key_policy); + if (err) + return err; + + k->def = !!tb[NL80211_KEY_DEFAULT]; + k->defmgmt = !!tb[NL80211_KEY_DEFAULT_MGMT]; + + if (tb[NL80211_KEY_IDX]) + k->idx = nla_get_u8(tb[NL80211_KEY_IDX]); + + if (tb[NL80211_KEY_DATA]) { + k->p.key = nla_data(tb[NL80211_KEY_DATA]); + k->p.key_len = nla_len(tb[NL80211_KEY_DATA]); + } + + if (tb[NL80211_KEY_SEQ]) { + k->p.seq = nla_data(tb[NL80211_KEY_SEQ]); + k->p.seq_len = nla_len(tb[NL80211_KEY_SEQ]); + } + + if (tb[NL80211_KEY_CIPHER]) + k->p.cipher = nla_get_u32(tb[NL80211_KEY_CIPHER]); + + return 0; +} + +static int nl80211_parse_key_old(struct genl_info *info, struct key_parse *k) +{ + if (info->attrs[NL80211_ATTR_KEY_DATA]) { + k->p.key = nla_data(info->attrs[NL80211_ATTR_KEY_DATA]); + k->p.key_len = nla_len(info->attrs[NL80211_ATTR_KEY_DATA]); + } + + if (info->attrs[NL80211_ATTR_KEY_SEQ]) { + k->p.seq = nla_data(info->attrs[NL80211_ATTR_KEY_SEQ]); + k->p.seq_len = nla_len(info->attrs[NL80211_ATTR_KEY_SEQ]); + } + + if (info->attrs[NL80211_ATTR_KEY_IDX]) + k->idx = nla_get_u8(info->attrs[NL80211_ATTR_KEY_IDX]); + + if (info->attrs[NL80211_ATTR_KEY_CIPHER]) + k->p.cipher = nla_get_u32(info->attrs[NL80211_ATTR_KEY_CIPHER]); + + k->def = !!info->attrs[NL80211_ATTR_KEY_DEFAULT]; + k->defmgmt = !!info->attrs[NL80211_ATTR_KEY_DEFAULT_MGMT]; + + return 0; +} + +static int nl80211_parse_key(struct genl_info *info, struct key_parse *k) +{ + int err; + + memset(k, 0, sizeof(*k)); + k->idx = -1; + + if (info->attrs[NL80211_ATTR_KEY]) + err = nl80211_parse_key_new(info->attrs[NL80211_ATTR_KEY], k); + else + err = nl80211_parse_key_old(info, k); + + if (err) + return err; + + if (k->def && k->defmgmt) + return -EINVAL; + + if (k->idx != -1) { + if (k->defmgmt) { + if (k->idx < 4 || k->idx > 5) + return -EINVAL; + } else if (k->def) { + if (k->idx < 0 || k->idx > 3) + return -EINVAL; + } else { + if (k->idx < 0 || k->idx > 5) + return -EINVAL; + } + } + + return 0; +} + static int nl80211_send_wiphy(struct sk_buff *msg, u32 pid, u32 seq, int flags, struct cfg80211_registered_device *dev) { @@ -943,10 +1050,12 @@ static int nl80211_del_interface(struct sk_buff *skb, struct genl_info *info) struct get_key_cookie { struct sk_buff *msg; int error; + int idx; }; static void get_key_callback(void *c, struct key_params *params) { + struct nlattr *key; struct get_key_cookie *cookie = c; if (params->key) @@ -961,6 +1070,26 @@ static void get_key_callback(void *c, struct key_params *params) NLA_PUT_U32(cookie->msg, NL80211_ATTR_KEY_CIPHER, params->cipher); + key = nla_nest_start(cookie->msg, NL80211_ATTR_KEY); + if (!key) + goto nla_put_failure; + + if (params->key) + NLA_PUT(cookie->msg, NL80211_KEY_DATA, + params->key_len, params->key); + + if (params->seq) + NLA_PUT(cookie->msg, NL80211_KEY_SEQ, + params->seq_len, params->seq); + + if (params->cipher) + NLA_PUT_U32(cookie->msg, NL80211_KEY_CIPHER, + params->cipher); + + NLA_PUT_U8(cookie->msg, NL80211_ATTR_KEY_IDX, cookie->idx); + + nla_nest_end(cookie->msg, key); + return; nla_put_failure: cookie->error = 1; @@ -1014,6 +1143,7 @@ static int nl80211_get_key(struct sk_buff *skb, struct genl_info *info) } cookie.msg = msg; + cookie.idx = key_idx; NLA_PUT_U32(msg, NL80211_ATTR_IFINDEX, dev->ifindex); NLA_PUT_U8(msg, NL80211_ATTR_KEY_IDX, key_idx); @@ -1049,26 +1179,21 @@ static int nl80211_get_key(struct sk_buff *skb, struct genl_info *info) static int nl80211_set_key(struct sk_buff *skb, struct genl_info *info) { struct cfg80211_registered_device *rdev; + struct key_parse key; int err; struct net_device *dev; - u8 key_idx; int (*func)(struct wiphy *wiphy, struct net_device *netdev, u8 key_index); - if (!info->attrs[NL80211_ATTR_KEY_IDX]) - return -EINVAL; - - key_idx = nla_get_u8(info->attrs[NL80211_ATTR_KEY_IDX]); + err = nl80211_parse_key(info, &key); + if (err) + return err; - if (info->attrs[NL80211_ATTR_KEY_DEFAULT_MGMT]) { - if (key_idx < 4 || key_idx > 5) - return -EINVAL; - } else if (key_idx > 3) + if (key.idx < 0) return -EINVAL; - /* currently only support setting default key */ - if (!info->attrs[NL80211_ATTR_KEY_DEFAULT] && - !info->attrs[NL80211_ATTR_KEY_DEFAULT_MGMT]) + /* only support setting default key */ + if (!key.def && !key.defmgmt) return -EINVAL; rtnl_lock(); @@ -1077,7 +1202,7 @@ static int nl80211_set_key(struct sk_buff *skb, struct genl_info *info) if (err) goto unlock_rtnl; - if (info->attrs[NL80211_ATTR_KEY_DEFAULT]) + if (key.def) func = rdev->ops->set_default_key; else func = rdev->ops->set_default_mgmt_key; @@ -1087,13 +1212,13 @@ static int nl80211_set_key(struct sk_buff *skb, struct genl_info *info) goto out; } - err = func(&rdev->wiphy, dev, key_idx); + err = func(&rdev->wiphy, dev, key.idx); #ifdef CONFIG_WIRELESS_EXT if (!err) { if (func == rdev->ops->set_default_key) - dev->ieee80211_ptr->wext.default_key = key_idx; + dev->ieee80211_ptr->wext.default_key = key.idx; else - dev->ieee80211_ptr->wext.default_mgmt_key = key_idx; + dev->ieee80211_ptr->wext.default_mgmt_key = key.idx; } #endif @@ -1112,34 +1237,20 @@ static int nl80211_new_key(struct sk_buff *skb, struct genl_info *info) struct cfg80211_registered_device *rdev; int err, i; struct net_device *dev; - struct key_params params; - u8 key_idx = 0; + struct key_parse key; u8 *mac_addr = NULL; - memset(¶ms, 0, sizeof(params)); + err = nl80211_parse_key(info, &key); + if (err) + return err; - if (!info->attrs[NL80211_ATTR_KEY_CIPHER]) + if (!key.p.key) return -EINVAL; - if (info->attrs[NL80211_ATTR_KEY_DATA]) { - params.key = nla_data(info->attrs[NL80211_ATTR_KEY_DATA]); - params.key_len = nla_len(info->attrs[NL80211_ATTR_KEY_DATA]); - } - - if (info->attrs[NL80211_ATTR_KEY_SEQ]) { - params.seq = nla_data(info->attrs[NL80211_ATTR_KEY_SEQ]); - params.seq_len = nla_len(info->attrs[NL80211_ATTR_KEY_SEQ]); - } - - if (info->attrs[NL80211_ATTR_KEY_IDX]) - key_idx = nla_get_u8(info->attrs[NL80211_ATTR_KEY_IDX]); - - params.cipher = nla_get_u32(info->attrs[NL80211_ATTR_KEY_CIPHER]); - if (info->attrs[NL80211_ATTR_MAC]) mac_addr = nla_data(info->attrs[NL80211_ATTR_MAC]); - if (cfg80211_validate_key_settings(¶ms, key_idx, mac_addr)) + if (cfg80211_validate_key_settings(&key.p, key.idx, mac_addr)) return -EINVAL; rtnl_lock(); @@ -1149,7 +1260,7 @@ static int nl80211_new_key(struct sk_buff *skb, struct genl_info *info) goto unlock_rtnl; for (i = 0; i < rdev->wiphy.n_cipher_suites; i++) - if (params.cipher == rdev->wiphy.cipher_suites[i]) + if (key.p.cipher == rdev->wiphy.cipher_suites[i]) break; if (i == rdev->wiphy.n_cipher_suites) { err = -EINVAL; @@ -1161,7 +1272,7 @@ static int nl80211_new_key(struct sk_buff *skb, struct genl_info *info) goto out; } - err = rdev->ops->add_key(&rdev->wiphy, dev, key_idx, mac_addr, ¶ms); + err = rdev->ops->add_key(&rdev->wiphy, dev, key.idx, mac_addr, &key.p); out: cfg80211_unlock_rdev(rdev); @@ -1177,14 +1288,12 @@ static int nl80211_del_key(struct sk_buff *skb, struct genl_info *info) struct cfg80211_registered_device *rdev; int err; struct net_device *dev; - u8 key_idx = 0; u8 *mac_addr = NULL; + struct key_parse key; - if (info->attrs[NL80211_ATTR_KEY_IDX]) - key_idx = nla_get_u8(info->attrs[NL80211_ATTR_KEY_IDX]); - - if (key_idx > 5) - return -EINVAL; + err = nl80211_parse_key(info, &key); + if (err) + return err; if (info->attrs[NL80211_ATTR_MAC]) mac_addr = nla_data(info->attrs[NL80211_ATTR_MAC]); @@ -1200,13 +1309,13 @@ static int nl80211_del_key(struct sk_buff *skb, struct genl_info *info) goto out; } - err = rdev->ops->del_key(&rdev->wiphy, dev, key_idx, mac_addr); + err = rdev->ops->del_key(&rdev->wiphy, dev, key.idx, mac_addr); #ifdef CONFIG_WIRELESS_EXT if (!err) { - if (key_idx == dev->ieee80211_ptr->wext.default_key) + if (key.idx == dev->ieee80211_ptr->wext.default_key) dev->ieee80211_ptr->wext.default_key = -1; - else if (key_idx == dev->ieee80211_ptr->wext.default_mgmt_key) + else if (key.idx == dev->ieee80211_ptr->wext.default_mgmt_key) dev->ieee80211_ptr->wext.default_mgmt_key = -1; } #endif -- cgit v1.2.3 From fffd0934b9390f34bec45762192b7edd3b12b4b5 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Wed, 8 Jul 2009 14:22:54 +0200 Subject: cfg80211: rework key operation This reworks the key operation in cfg80211, and now only allows, from userspace, configuring keys (via nl80211) after the connection has been established (in managed mode), the IBSS been joined (in IBSS mode), at any time (in AP[_VLAN] modes) or never for all the other modes. In order to do shared key authentication correctly, it is now possible to give a WEP key to the AUTH command. To configure static WEP keys, these are given to the CONNECT or IBSS_JOIN command directly, for a userspace SME it is assumed it will configure it properly after the connection has been established. Since mac80211 used to check the default key in IBSS mode to see whether or not the network is protected, it needs an update in that area, as well as an update to make use of the WEP key passed to auth() for shared key authentication. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- include/linux/nl80211.h | 5 ++ include/net/cfg80211.h | 18 ++++- net/mac80211/ibss.c | 9 +-- net/mac80211/ieee80211_i.h | 8 ++- net/mac80211/mlme.c | 11 ++- net/mac80211/util.c | 16 +++-- net/mac80211/wep.c | 6 +- net/mac80211/wep.h | 3 + net/wireless/core.c | 11 ++- net/wireless/core.h | 32 +++++++-- net/wireless/ibss.c | 79 +++++++++++++++++---- net/wireless/mlme.c | 16 ++++- net/wireless/nl80211.c | 170 ++++++++++++++++++++++++++++++++++++++++----- net/wireless/sme.c | 97 +++++++++++++++++++------- net/wireless/util.c | 41 ++++++++++- net/wireless/wext-compat.c | 163 ++++++++++++++++++++++++++----------------- net/wireless/wext-sme.c | 30 ++++++-- 17 files changed, 554 insertions(+), 161 deletions(-) (limited to 'include') diff --git a/include/linux/nl80211.h b/include/linux/nl80211.h index 48e0913c2209..b043b78dd2c3 100644 --- a/include/linux/nl80211.h +++ b/include/linux/nl80211.h @@ -569,6 +569,9 @@ enum nl80211_commands { * * @NL80211_ATTR_KEY: key information in a nested attribute with * %NL80211_KEY_* sub-attributes + * @NL80211_ATTR_KEYS: array of keys for static WEP keys for connect() + * and join_ibss(), key information is in a nested attribute each + * with %NL80211_KEY_* sub-attributes * * @NL80211_ATTR_MAX: highest attribute number currently defined * @__NL80211_ATTR_AFTER_LAST: internal use @@ -696,6 +699,7 @@ enum nl80211_attrs { NL80211_ATTR_PREV_BSSID, NL80211_ATTR_KEY, + NL80211_ATTR_KEYS, /* add attributes here, update the policy in nl80211.c */ @@ -726,6 +730,7 @@ enum nl80211_attrs { #define NL80211_ATTR_WPA_VERSIONS NL80211_ATTR_WPA_VERSIONS #define NL80211_ATTR_AKM_SUITES NL80211_ATTR_AKM_SUITES #define NL80211_ATTR_KEY NL80211_ATTR_KEY +#define NL80211_ATTR_KEYS NL80211_ATTR_KEYS #define NL80211_MAX_SUPP_RATES 32 #define NL80211_MAX_SUPP_REG_RULES 32 diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index 83c2c727d71e..65a5cbcb5d14 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -647,12 +647,17 @@ struct cfg80211_crypto_settings { * @auth_type: Authentication type (algorithm) * @ie: Extra IEs to add to Authentication frame or %NULL * @ie_len: Length of ie buffer in octets + * @key_len: length of WEP key for shared key authentication + * @key_idx: index of WEP key for shared key authentication + * @key: WEP key for shared key authentication */ struct cfg80211_auth_request { struct cfg80211_bss *bss; const u8 *ie; size_t ie_len; enum nl80211_auth_type auth_type; + const u8 *key; + u8 key_len, key_idx; }; /** @@ -727,6 +732,8 @@ struct cfg80211_disassoc_request { * @ie: information element(s) to include in the beacon * @ie_len: length of that * @beacon_interval: beacon interval to use + * @privacy: this is a protected network, keys will be configured + * after joining */ struct cfg80211_ibss_params { u8 *ssid; @@ -736,6 +743,7 @@ struct cfg80211_ibss_params { u8 ssid_len, ie_len; u16 beacon_interval; bool channel_fixed; + bool privacy; }; /** @@ -755,6 +763,9 @@ struct cfg80211_ibss_params { * @assoc_ie_len: Length of assoc_ie in octets * @privacy: indicates whether privacy-enabled APs should be used * @crypto: crypto settings + * @key_len: length of WEP key for shared key authentication + * @key_idx: index of WEP key for shared key authentication + * @key: WEP key for shared key authentication */ struct cfg80211_connect_params { struct ieee80211_channel *channel; @@ -766,6 +777,8 @@ struct cfg80211_connect_params { size_t ie_len; bool privacy; struct cfg80211_crypto_settings crypto; + const u8 *key; + u8 key_len, key_idx; }; /** @@ -1223,9 +1236,10 @@ extern void wiphy_unregister(struct wiphy *wiphy); */ extern void wiphy_free(struct wiphy *wiphy); -/* internal struct */ +/* internal structs */ struct cfg80211_conn; struct cfg80211_internal_bss; +struct cfg80211_cached_keys; #define MAX_AUTH_BSSES 4 @@ -1267,6 +1281,7 @@ struct wireless_dev { CFG80211_SME_CONNECTED, } sme_state; struct cfg80211_conn *conn; + struct cfg80211_cached_keys *connect_keys; struct list_head event_list; spinlock_t event_lock; @@ -1280,6 +1295,7 @@ struct wireless_dev { struct { struct cfg80211_ibss_params ibss; struct cfg80211_connect_params connect; + struct cfg80211_cached_keys *keys; u8 *ie; size_t ie_len; u8 bssid[ETH_ALEN]; diff --git a/net/mac80211/ibss.c b/net/mac80211/ibss.c index 15d5a53b59a8..8e2220000e5c 100644 --- a/net/mac80211/ibss.c +++ b/net/mac80211/ibss.c @@ -57,7 +57,7 @@ static void ieee80211_rx_mgmt_auth_ibss(struct ieee80211_sub_if_data *sdata, */ if (auth_alg == WLAN_AUTH_OPEN && auth_transaction == 1) ieee80211_send_auth(sdata, 2, WLAN_AUTH_OPEN, NULL, 0, - sdata->u.ibss.bssid, 0); + sdata->u.ibss.bssid, NULL, 0, 0); } static void __ieee80211_sta_join_ibss(struct ieee80211_sub_if_data *sdata, @@ -494,7 +494,7 @@ static void ieee80211_sta_create_ibss(struct ieee80211_sub_if_data *sdata) capability = WLAN_CAPABILITY_IBSS; - if (sdata->default_key) + if (ifibss->privacy) capability |= WLAN_CAPABILITY_PRIVACY; else sdata->drop_unencrypted = 0; @@ -524,9 +524,8 @@ static void ieee80211_sta_find_ibss(struct ieee80211_sub_if_data *sdata) return; capability = WLAN_CAPABILITY_IBSS; - if (sdata->default_key) + if (ifibss->privacy) capability |= WLAN_CAPABILITY_PRIVACY; - if (ifibss->fixed_bssid) bssid = ifibss->bssid; if (ifibss->fixed_channel) @@ -872,6 +871,8 @@ int ieee80211_ibss_join(struct ieee80211_sub_if_data *sdata, } else sdata->u.ibss.fixed_bssid = false; + sdata->u.ibss.privacy = params->privacy; + sdata->vif.bss_conf.beacon_int = params->beacon_interval; sdata->u.ibss.channel = params->channel; diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h index 327aabc07abe..06b3411530f2 100644 --- a/net/mac80211/ieee80211_i.h +++ b/net/mac80211/ieee80211_i.h @@ -247,6 +247,9 @@ struct ieee80211_mgd_work { int tries; + u8 key[WLAN_KEY_LEN_WEP104]; + u8 key_len, key_idx; + /* must be last */ u8 ie[0]; /* for auth or assoc frame, not probe */ }; @@ -321,6 +324,7 @@ struct ieee80211_if_ibss { bool fixed_bssid; bool fixed_channel; + bool privacy; u8 bssid[ETH_ALEN]; u8 ssid[IEEE80211_MAX_SSID_LEN]; @@ -1093,8 +1097,8 @@ int ieee80211_add_pending_skbs(struct ieee80211_local *local, void ieee80211_send_auth(struct ieee80211_sub_if_data *sdata, u16 transaction, u16 auth_alg, - u8 *extra, size_t extra_len, - const u8 *bssid, int encrypt); + u8 *extra, size_t extra_len, const u8 *bssid, + const u8 *key, u8 key_len, u8 key_idx); int ieee80211_build_preq_ies(struct ieee80211_local *local, u8 *buffer, const u8 *ie, size_t ie_len); void ieee80211_send_probe_req(struct ieee80211_sub_if_data *sdata, u8 *dst, diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index c9db9646025b..8e4a60497bba 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -954,7 +954,7 @@ ieee80211_authenticate(struct ieee80211_sub_if_data *sdata, sdata->dev->name, wk->bss->cbss.bssid, wk->tries); ieee80211_send_auth(sdata, 1, wk->auth_alg, wk->ie, wk->ie_len, - wk->bss->cbss.bssid, 0); + wk->bss->cbss.bssid, NULL, 0, 0); wk->auth_transaction = 2; wk->timeout = jiffies + IEEE80211_AUTH_TIMEOUT; @@ -1176,7 +1176,8 @@ static void ieee80211_auth_challenge(struct ieee80211_sub_if_data *sdata, return; ieee80211_send_auth(sdata, 3, wk->auth_alg, elems.challenge - 2, elems.challenge_len + 2, - wk->bss->cbss.bssid, 1); + wk->bss->cbss.bssid, + wk->key, wk->key_len, wk->key_idx); wk->auth_transaction = 4; } @@ -2175,6 +2176,12 @@ int ieee80211_mgd_auth(struct ieee80211_sub_if_data *sdata, wk->ie_len = req->ie_len; } + if (req->key && req->key_len) { + wk->key_len = req->key_len; + wk->key_idx = req->key_idx; + memcpy(wk->key, req->key, req->key_len); + } + ssid = ieee80211_bss_get_ie(req->bss, WLAN_EID_SSID); memcpy(wk->ssid, ssid + 2, ssid[1]); wk->ssid_len = ssid[1]; diff --git a/net/mac80211/util.c b/net/mac80211/util.c index 915e77769312..dbf66b52d38c 100644 --- a/net/mac80211/util.c +++ b/net/mac80211/util.c @@ -31,6 +31,7 @@ #include "mesh.h" #include "wme.h" #include "led.h" +#include "wep.h" /* privid for wiphys to determine whether they belong to us or not */ void *mac80211_wiphy_privid = &mac80211_wiphy_privid; @@ -804,12 +805,13 @@ u32 ieee80211_mandatory_rates(struct ieee80211_local *local, void ieee80211_send_auth(struct ieee80211_sub_if_data *sdata, u16 transaction, u16 auth_alg, - u8 *extra, size_t extra_len, - const u8 *bssid, int encrypt) + u8 *extra, size_t extra_len, const u8 *bssid, + const u8 *key, u8 key_len, u8 key_idx) { struct ieee80211_local *local = sdata->local; struct sk_buff *skb; struct ieee80211_mgmt *mgmt; + int err; skb = dev_alloc_skb(local->hw.extra_tx_headroom + sizeof(*mgmt) + 6 + extra_len); @@ -824,8 +826,6 @@ void ieee80211_send_auth(struct ieee80211_sub_if_data *sdata, memset(mgmt, 0, 24 + 6); mgmt->frame_control = cpu_to_le16(IEEE80211_FTYPE_MGMT | IEEE80211_STYPE_AUTH); - if (encrypt) - mgmt->frame_control |= cpu_to_le16(IEEE80211_FCTL_PROTECTED); memcpy(mgmt->da, bssid, ETH_ALEN); memcpy(mgmt->sa, sdata->dev->dev_addr, ETH_ALEN); memcpy(mgmt->bssid, bssid, ETH_ALEN); @@ -835,7 +835,13 @@ void ieee80211_send_auth(struct ieee80211_sub_if_data *sdata, if (extra) memcpy(skb_put(skb, extra_len), extra, extra_len); - ieee80211_tx_skb(sdata, skb, encrypt); + if (auth_alg == WLAN_AUTH_SHARED_KEY && transaction == 3) { + mgmt->frame_control |= cpu_to_le16(IEEE80211_FCTL_PROTECTED); + err = ieee80211_wep_encrypt(local, skb, key, key_len, key_idx); + WARN_ON(err); + } + + ieee80211_tx_skb(sdata, skb, 0); } int ieee80211_build_preq_ies(struct ieee80211_local *local, u8 *buffer, diff --git a/net/mac80211/wep.c b/net/mac80211/wep.c index 4fafb2d27c84..8a980f136941 100644 --- a/net/mac80211/wep.c +++ b/net/mac80211/wep.c @@ -144,9 +144,9 @@ void ieee80211_wep_encrypt_data(struct crypto_blkcipher *tfm, u8 *rc4key, * * WEP frame payload: IV + TX key idx, RC4(data), ICV = RC4(CRC32(data)) */ -static int ieee80211_wep_encrypt(struct ieee80211_local *local, - struct sk_buff *skb, - const u8 *key, int keylen, int keyidx) +int ieee80211_wep_encrypt(struct ieee80211_local *local, + struct sk_buff *skb, + const u8 *key, int keylen, int keyidx) { u8 *iv; size_t len; diff --git a/net/mac80211/wep.h b/net/mac80211/wep.h index 85219ded8703..fe29d7e5759f 100644 --- a/net/mac80211/wep.h +++ b/net/mac80211/wep.h @@ -20,6 +20,9 @@ int ieee80211_wep_init(struct ieee80211_local *local); void ieee80211_wep_free(struct ieee80211_local *local); void ieee80211_wep_encrypt_data(struct crypto_blkcipher *tfm, u8 *rc4key, size_t klen, u8 *data, size_t data_len); +int ieee80211_wep_encrypt(struct ieee80211_local *local, + struct sk_buff *skb, + const u8 *key, int keylen, int keyidx); int ieee80211_wep_decrypt_data(struct crypto_blkcipher *tfm, u8 *rc4key, size_t klen, u8 *data, size_t data_len); bool ieee80211_wep_is_weak_iv(struct sk_buff *skb, struct ieee80211_key *key); diff --git a/net/wireless/core.c b/net/wireless/core.c index 1a78b3c70cf2..97cc5968b7d6 100644 --- a/net/wireless/core.c +++ b/net/wireless/core.c @@ -666,14 +666,10 @@ static int cfg80211_netdev_notifier_call(struct notifier_block * nb, wdev_lock(wdev); switch (wdev->iftype) { case NL80211_IFTYPE_ADHOC: - if (wdev->wext.ibss.ssid_len) - __cfg80211_join_ibss(rdev, dev, - &wdev->wext.ibss); + cfg80211_ibss_wext_join(rdev, wdev); break; case NL80211_IFTYPE_STATION: - if (wdev->wext.connect.ssid_len) - __cfg80211_connect(rdev, dev, - &wdev->wext.connect); + cfg80211_mgd_wext_connect(rdev, wdev); break; default: break; @@ -690,6 +686,9 @@ static int cfg80211_netdev_notifier_call(struct notifier_block * nb, } mutex_unlock(&rdev->devlist_mtx); mutex_destroy(&wdev->mtx); +#ifdef CONFIG_WIRELESS_EXT + kfree(wdev->wext.keys); +#endif break; case NETDEV_PRE_UP: if (!(wdev->wiphy->interface_modes & BIT(wdev->iftype))) diff --git a/net/wireless/core.h b/net/wireless/core.h index e46cd6eb61d7..2ec8ddbe57de 100644 --- a/net/wireless/core.h +++ b/net/wireless/core.h @@ -238,6 +238,12 @@ struct cfg80211_event { }; }; +struct cfg80211_cached_keys { + struct key_params params[6]; + u8 data[6][WLAN_MAX_KEY_LEN]; + int def, defmgmt; +}; + /* free object */ extern void cfg80211_dev_free(struct cfg80211_registered_device *rdev); @@ -256,14 +262,18 @@ void cfg80211_bss_age(struct cfg80211_registered_device *dev, /* IBSS */ int __cfg80211_join_ibss(struct cfg80211_registered_device *rdev, struct net_device *dev, - struct cfg80211_ibss_params *params); + struct cfg80211_ibss_params *params, + struct cfg80211_cached_keys *connkeys); int cfg80211_join_ibss(struct cfg80211_registered_device *rdev, struct net_device *dev, - struct cfg80211_ibss_params *params); + struct cfg80211_ibss_params *params, + struct cfg80211_cached_keys *connkeys); void cfg80211_clear_ibss(struct net_device *dev, bool nowext); int cfg80211_leave_ibss(struct cfg80211_registered_device *rdev, struct net_device *dev, bool nowext); void __cfg80211_ibss_joined(struct net_device *dev, const u8 *bssid); +int cfg80211_ibss_wext_join(struct cfg80211_registered_device *rdev, + struct wireless_dev *wdev); /* MLME */ int __cfg80211_mlme_auth(struct cfg80211_registered_device *rdev, @@ -272,12 +282,14 @@ int __cfg80211_mlme_auth(struct cfg80211_registered_device *rdev, enum nl80211_auth_type auth_type, const u8 *bssid, const u8 *ssid, int ssid_len, - const u8 *ie, int ie_len); + const u8 *ie, int ie_len, + const u8 *key, int key_len, int key_idx); int cfg80211_mlme_auth(struct cfg80211_registered_device *rdev, struct net_device *dev, struct ieee80211_channel *chan, enum nl80211_auth_type auth_type, const u8 *bssid, const u8 *ssid, int ssid_len, - const u8 *ie, int ie_len); + const u8 *ie, int ie_len, + const u8 *key, int key_len, int key_idx); int __cfg80211_mlme_assoc(struct cfg80211_registered_device *rdev, struct net_device *dev, struct ieee80211_channel *chan, @@ -310,10 +322,12 @@ void __cfg80211_connect_result(struct net_device *dev, const u8 *bssid, /* SME */ int __cfg80211_connect(struct cfg80211_registered_device *rdev, struct net_device *dev, - struct cfg80211_connect_params *connect); + struct cfg80211_connect_params *connect, + struct cfg80211_cached_keys *connkeys); int cfg80211_connect(struct cfg80211_registered_device *rdev, struct net_device *dev, - struct cfg80211_connect_params *connect); + struct cfg80211_connect_params *connect, + struct cfg80211_cached_keys *connkeys); int __cfg80211_disconnect(struct cfg80211_registered_device *rdev, struct net_device *dev, u16 reason, bool wextev); @@ -323,11 +337,14 @@ int cfg80211_disconnect(struct cfg80211_registered_device *rdev, void __cfg80211_roamed(struct wireless_dev *wdev, const u8 *bssid, const u8 *req_ie, size_t req_ie_len, const u8 *resp_ie, size_t resp_ie_len); +int cfg80211_mgd_wext_connect(struct cfg80211_registered_device *rdev, + struct wireless_dev *wdev); void cfg80211_conn_work(struct work_struct *work); /* internal helpers */ -int cfg80211_validate_key_settings(struct key_params *params, int key_idx, +int cfg80211_validate_key_settings(struct cfg80211_registered_device *rdev, + struct key_params *params, int key_idx, const u8 *mac_addr); void __cfg80211_disconnected(struct net_device *dev, const u8 *ie, size_t ie_len, u16 reason, bool from_ap); @@ -335,5 +352,6 @@ void cfg80211_sme_scan_done(struct net_device *dev); void cfg80211_sme_rx_auth(struct net_device *dev, const u8 *buf, size_t len); void cfg80211_sme_disassoc(struct net_device *dev, int idx); void __cfg80211_scan_done(struct work_struct *wk); +void cfg80211_upload_connect_keys(struct wireless_dev *wdev); #endif /* __NET_WIRELESS_CORE_H */ diff --git a/net/wireless/ibss.c b/net/wireless/ibss.c index 99ef9364b7e8..9394e78cd11f 100644 --- a/net/wireless/ibss.c +++ b/net/wireless/ibss.c @@ -39,6 +39,8 @@ void __cfg80211_ibss_joined(struct net_device *dev, const u8 *bssid) cfg80211_hold_bss(bss_from_pub(bss)); wdev->current_bss = bss_from_pub(bss); + cfg80211_upload_connect_keys(wdev); + nl80211_send_ibss_bssid(wiphy_to_dev(wdev->wiphy), dev, bssid, GFP_KERNEL); #ifdef CONFIG_WIRELESS_EXT @@ -71,7 +73,8 @@ EXPORT_SYMBOL(cfg80211_ibss_joined); int __cfg80211_join_ibss(struct cfg80211_registered_device *rdev, struct net_device *dev, - struct cfg80211_ibss_params *params) + struct cfg80211_ibss_params *params, + struct cfg80211_cached_keys *connkeys) { struct wireless_dev *wdev = dev->ieee80211_ptr; int err; @@ -81,13 +84,18 @@ int __cfg80211_join_ibss(struct cfg80211_registered_device *rdev, if (wdev->ssid_len) return -EALREADY; + if (WARN_ON(wdev->connect_keys)) + kfree(wdev->connect_keys); + wdev->connect_keys = connkeys; + #ifdef CONFIG_WIRELESS_EXT wdev->wext.ibss.channel = params->channel; #endif err = rdev->ops->join_ibss(&rdev->wiphy, dev, params); - - if (err) + if (err) { + wdev->connect_keys = NULL; return err; + } memcpy(wdev->ssid, params->ssid, params->ssid_len); wdev->ssid_len = params->ssid_len; @@ -97,13 +105,14 @@ int __cfg80211_join_ibss(struct cfg80211_registered_device *rdev, int cfg80211_join_ibss(struct cfg80211_registered_device *rdev, struct net_device *dev, - struct cfg80211_ibss_params *params) + struct cfg80211_ibss_params *params, + struct cfg80211_cached_keys *connkeys) { struct wireless_dev *wdev = dev->ieee80211_ptr; int err; wdev_lock(wdev); - err = __cfg80211_join_ibss(rdev, dev, params); + err = __cfg80211_join_ibss(rdev, dev, params, connkeys); wdev_unlock(wdev); return err; @@ -112,9 +121,22 @@ int cfg80211_join_ibss(struct cfg80211_registered_device *rdev, static void __cfg80211_clear_ibss(struct net_device *dev, bool nowext) { struct wireless_dev *wdev = dev->ieee80211_ptr; + struct cfg80211_registered_device *rdev = wiphy_to_dev(wdev->wiphy); + int i; ASSERT_WDEV_LOCK(wdev); + kfree(wdev->connect_keys); + wdev->connect_keys = NULL; + + /* + * Delete all the keys ... pairwise keys can't really + * exist any more anyway, but default keys might. + */ + if (rdev->ops->del_key) + for (i = 0; i < 6; i++) + rdev->ops->del_key(wdev->wiphy, dev, i, NULL); + if (wdev->current_bss) { cfg80211_unhold_bss(wdev->current_bss); cfg80211_put_bss(&wdev->current_bss->pub); @@ -172,11 +194,14 @@ int cfg80211_leave_ibss(struct cfg80211_registered_device *rdev, } #ifdef CONFIG_WIRELESS_EXT -static int cfg80211_ibss_wext_join(struct cfg80211_registered_device *rdev, - struct wireless_dev *wdev) +int cfg80211_ibss_wext_join(struct cfg80211_registered_device *rdev, + struct wireless_dev *wdev) { + struct cfg80211_cached_keys *ck = NULL; enum ieee80211_band band; - int i; + int i, err; + + ASSERT_WDEV_LOCK(wdev); if (!wdev->wext.ibss.beacon_interval) wdev->wext.ibss.beacon_interval = 100; @@ -216,8 +241,24 @@ static int cfg80211_ibss_wext_join(struct cfg80211_registered_device *rdev, if (!netif_running(wdev->netdev)) return 0; - return cfg80211_join_ibss(wiphy_to_dev(wdev->wiphy), - wdev->netdev, &wdev->wext.ibss); + if (wdev->wext.keys) + wdev->wext.keys->def = wdev->wext.default_key; + + wdev->wext.ibss.privacy = wdev->wext.default_key != -1; + + if (wdev->wext.keys) { + ck = kmemdup(wdev->wext.keys, sizeof(*ck), GFP_KERNEL); + if (!ck) + return -ENOMEM; + for (i = 0; i < 6; i++) + ck->params[i].key = ck->data[i]; + } + err = __cfg80211_join_ibss(rdev, wdev->netdev, + &wdev->wext.ibss, ck); + if (err) + kfree(ck); + + return err; } int cfg80211_ibss_wext_siwfreq(struct net_device *dev, @@ -265,7 +306,11 @@ int cfg80211_ibss_wext_siwfreq(struct net_device *dev, wdev->wext.ibss.channel_fixed = false; } - return cfg80211_ibss_wext_join(wiphy_to_dev(wdev->wiphy), wdev); + wdev_lock(wdev); + err = cfg80211_ibss_wext_join(wiphy_to_dev(wdev->wiphy), wdev); + wdev_unlock(wdev); + + return err; } /* temporary symbol - mark GPL - in the future the handler won't be */ EXPORT_SYMBOL_GPL(cfg80211_ibss_wext_siwfreq); @@ -333,7 +378,11 @@ int cfg80211_ibss_wext_siwessid(struct net_device *dev, memcpy(wdev->wext.ibss.ssid, ssid, len); wdev->wext.ibss.ssid_len = len; - return cfg80211_ibss_wext_join(wiphy_to_dev(wdev->wiphy), wdev); + wdev_lock(wdev); + err = cfg80211_ibss_wext_join(wiphy_to_dev(wdev->wiphy), wdev); + wdev_unlock(wdev); + + return err; } /* temporary symbol - mark GPL - in the future the handler won't be */ EXPORT_SYMBOL_GPL(cfg80211_ibss_wext_siwessid); @@ -414,7 +463,11 @@ int cfg80211_ibss_wext_siwap(struct net_device *dev, } else wdev->wext.ibss.bssid = NULL; - return cfg80211_ibss_wext_join(wiphy_to_dev(wdev->wiphy), wdev); + wdev_lock(wdev); + err = cfg80211_ibss_wext_join(wiphy_to_dev(wdev->wiphy), wdev); + wdev_unlock(wdev); + + return err; } /* temporary symbol - mark GPL - in the future the handler won't be */ EXPORT_SYMBOL_GPL(cfg80211_ibss_wext_siwap); diff --git a/net/wireless/mlme.c b/net/wireless/mlme.c index 1b2ca1fea7a1..8e4ce2fdf862 100644 --- a/net/wireless/mlme.c +++ b/net/wireless/mlme.c @@ -328,7 +328,8 @@ int __cfg80211_mlme_auth(struct cfg80211_registered_device *rdev, enum nl80211_auth_type auth_type, const u8 *bssid, const u8 *ssid, int ssid_len, - const u8 *ie, int ie_len) + const u8 *ie, int ie_len, + const u8 *key, int key_len, int key_idx) { struct wireless_dev *wdev = dev->ieee80211_ptr; struct cfg80211_auth_request req; @@ -337,6 +338,10 @@ int __cfg80211_mlme_auth(struct cfg80211_registered_device *rdev, ASSERT_WDEV_LOCK(wdev); + if (auth_type == NL80211_AUTHTYPE_SHARED_KEY) + if (!key || !key_len || key_idx < 0 || key_idx > 4) + return -EINVAL; + if (wdev->current_bss && memcmp(bssid, wdev->current_bss->pub.bssid, ETH_ALEN) == 0) return -EALREADY; @@ -359,6 +364,9 @@ int __cfg80211_mlme_auth(struct cfg80211_registered_device *rdev, req.auth_type = auth_type; req.bss = cfg80211_get_bss(&rdev->wiphy, chan, bssid, ssid, ssid_len, WLAN_CAPABILITY_ESS, WLAN_CAPABILITY_ESS); + req.key = key; + req.key_len = key_len; + req.key_idx = key_idx; if (!req.bss) return -ENOENT; @@ -396,13 +404,15 @@ int cfg80211_mlme_auth(struct cfg80211_registered_device *rdev, struct net_device *dev, struct ieee80211_channel *chan, enum nl80211_auth_type auth_type, const u8 *bssid, const u8 *ssid, int ssid_len, - const u8 *ie, int ie_len) + const u8 *ie, int ie_len, + const u8 *key, int key_len, int key_idx) { int err; wdev_lock(dev->ieee80211_ptr); err = __cfg80211_mlme_auth(rdev, dev, chan, auth_type, bssid, - ssid, ssid_len, ie, ie_len); + ssid, ssid_len, ie, ie_len, + key, key_len, key_idx); wdev_unlock(dev->ieee80211_ptr); return err; diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index 50cf59316292..45c5f9c8e51b 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -138,8 +138,7 @@ static struct nla_policy nl80211_policy[NL80211_ATTR_MAX+1] __read_mostly = { /* policy for the attributes */ static struct nla_policy nl80211_key_policy[NL80211_KEY_MAX + 1] __read_mostly = { - [NL80211_KEY_DATA] = { .type = NLA_BINARY, - .len = WLAN_MAX_KEY_LEN }, + [NL80211_KEY_DATA] = { .type = NLA_BINARY, .len = WLAN_MAX_KEY_LEN }, [NL80211_KEY_IDX] = { .type = NLA_U8 }, [NL80211_KEY_CIPHER] = { .type = NLA_U32 }, [NL80211_KEY_SEQ] = { .type = NLA_BINARY, .len = 8 }, @@ -305,6 +304,83 @@ static int nl80211_parse_key(struct genl_info *info, struct key_parse *k) return 0; } +static struct cfg80211_cached_keys * +nl80211_parse_connkeys(struct cfg80211_registered_device *rdev, + struct nlattr *keys) +{ + struct key_parse parse; + struct nlattr *key; + struct cfg80211_cached_keys *result; + int rem, err, def = 0; + + result = kzalloc(sizeof(*result), GFP_KERNEL); + if (!result) + return ERR_PTR(-ENOMEM); + + result->def = -1; + result->defmgmt = -1; + + nla_for_each_nested(key, keys, rem) { + memset(&parse, 0, sizeof(parse)); + parse.idx = -1; + + err = nl80211_parse_key_new(key, &parse); + if (err) + goto error; + err = -EINVAL; + if (!parse.p.key) + goto error; + if (parse.idx < 0 || parse.idx > 4) + goto error; + if (parse.def) { + if (def) + goto error; + def = 1; + result->def = parse.idx; + } else if (parse.defmgmt) + goto error; + err = cfg80211_validate_key_settings(rdev, &parse.p, + parse.idx, NULL); + if (err) + goto error; + result->params[parse.idx].cipher = parse.p.cipher; + result->params[parse.idx].key_len = parse.p.key_len; + result->params[parse.idx].key = result->data[parse.idx]; + memcpy(result->data[parse.idx], parse.p.key, parse.p.key_len); + } + + return result; + error: + kfree(result); + return ERR_PTR(err); +} + +static int nl80211_key_allowed(struct wireless_dev *wdev) +{ + ASSERT_WDEV_LOCK(wdev); + + if (!netif_running(wdev->netdev)) + return -ENETDOWN; + + switch (wdev->iftype) { + case NL80211_IFTYPE_AP: + case NL80211_IFTYPE_AP_VLAN: + break; + case NL80211_IFTYPE_ADHOC: + if (!wdev->current_bss) + return -ENOLINK; + break; + case NL80211_IFTYPE_STATION: + if (wdev->sme_state != CFG80211_SME_CONNECTED) + return -ENOLINK; + break; + default: + return -EINVAL; + } + + return 0; +} + static int nl80211_send_wiphy(struct sk_buff *msg, u32 pid, u32 seq, int flags, struct cfg80211_registered_device *dev) { @@ -1212,7 +1288,11 @@ static int nl80211_set_key(struct sk_buff *skb, struct genl_info *info) goto out; } - err = func(&rdev->wiphy, dev, key.idx); + wdev_lock(dev->ieee80211_ptr); + err = nl80211_key_allowed(dev->ieee80211_ptr); + if (!err) + err = func(&rdev->wiphy, dev, key.idx); + #ifdef CONFIG_WIRELESS_EXT if (!err) { if (func == rdev->ops->set_default_key) @@ -1221,6 +1301,7 @@ static int nl80211_set_key(struct sk_buff *skb, struct genl_info *info) dev->ieee80211_ptr->wext.default_mgmt_key = key.idx; } #endif + wdev_unlock(dev->ieee80211_ptr); out: cfg80211_unlock_rdev(rdev); @@ -1235,7 +1316,7 @@ static int nl80211_set_key(struct sk_buff *skb, struct genl_info *info) static int nl80211_new_key(struct sk_buff *skb, struct genl_info *info) { struct cfg80211_registered_device *rdev; - int err, i; + int err; struct net_device *dev; struct key_parse key; u8 *mac_addr = NULL; @@ -1250,29 +1331,28 @@ static int nl80211_new_key(struct sk_buff *skb, struct genl_info *info) if (info->attrs[NL80211_ATTR_MAC]) mac_addr = nla_data(info->attrs[NL80211_ATTR_MAC]); - if (cfg80211_validate_key_settings(&key.p, key.idx, mac_addr)) - return -EINVAL; - rtnl_lock(); err = get_rdev_dev_by_info_ifindex(info->attrs, &rdev, &dev); if (err) goto unlock_rtnl; - for (i = 0; i < rdev->wiphy.n_cipher_suites; i++) - if (key.p.cipher == rdev->wiphy.cipher_suites[i]) - break; - if (i == rdev->wiphy.n_cipher_suites) { - err = -EINVAL; + if (!rdev->ops->add_key) { + err = -EOPNOTSUPP; goto out; } - if (!rdev->ops->add_key) { - err = -EOPNOTSUPP; + if (cfg80211_validate_key_settings(rdev, &key.p, key.idx, mac_addr)) { + err = -EINVAL; goto out; } - err = rdev->ops->add_key(&rdev->wiphy, dev, key.idx, mac_addr, &key.p); + wdev_lock(dev->ieee80211_ptr); + err = nl80211_key_allowed(dev->ieee80211_ptr); + if (!err) + err = rdev->ops->add_key(&rdev->wiphy, dev, key.idx, + mac_addr, &key.p); + wdev_unlock(dev->ieee80211_ptr); out: cfg80211_unlock_rdev(rdev); @@ -1309,7 +1389,10 @@ static int nl80211_del_key(struct sk_buff *skb, struct genl_info *info) goto out; } - err = rdev->ops->del_key(&rdev->wiphy, dev, key.idx, mac_addr); + wdev_lock(dev->ieee80211_ptr); + err = nl80211_key_allowed(dev->ieee80211_ptr); + if (!err) + err = rdev->ops->del_key(&rdev->wiphy, dev, key.idx, mac_addr); #ifdef CONFIG_WIRELESS_EXT if (!err) { @@ -1319,6 +1402,7 @@ static int nl80211_del_key(struct sk_buff *skb, struct genl_info *info) dev->ieee80211_ptr->wext.default_mgmt_key = -1; } #endif + wdev_unlock(dev->ieee80211_ptr); out: cfg80211_unlock_rdev(rdev); @@ -3159,6 +3243,7 @@ static int nl80211_authenticate(struct sk_buff *skb, struct genl_info *info) const u8 *bssid, *ssid, *ie = NULL; int err, ssid_len, ie_len = 0; enum nl80211_auth_type auth_type; + struct key_parse key; if (!is_valid_ie_attr(info->attrs[NL80211_ATTR_IE])) return -EINVAL; @@ -3175,6 +3260,25 @@ static int nl80211_authenticate(struct sk_buff *skb, struct genl_info *info) if (!info->attrs[NL80211_ATTR_WIPHY_FREQ]) return -EINVAL; + err = nl80211_parse_key(info, &key); + if (err) + return err; + + if (key.idx >= 0) { + if (!key.p.key || !key.p.key_len) + return -EINVAL; + if ((key.p.cipher != WLAN_CIPHER_SUITE_WEP40 || + key.p.key_len != WLAN_KEY_LEN_WEP40) && + (key.p.cipher != WLAN_CIPHER_SUITE_WEP104 || + key.p.key_len != WLAN_KEY_LEN_WEP104)) + return -EINVAL; + if (key.idx > 4) + return -EINVAL; + } else { + key.p.key_len = 0; + key.p.key = NULL; + } + rtnl_lock(); err = get_rdev_dev_by_info_ifindex(info->attrs, &rdev, &dev); @@ -3219,7 +3323,8 @@ static int nl80211_authenticate(struct sk_buff *skb, struct genl_info *info) } err = cfg80211_mlme_auth(rdev, dev, chan, auth_type, bssid, - ssid, ssid_len, ie, ie_len); + ssid, ssid_len, ie, ie_len, + key.p.key, key.p.key_len, key.idx); out: cfg80211_unlock_rdev(rdev); @@ -3506,6 +3611,7 @@ static int nl80211_join_ibss(struct sk_buff *skb, struct genl_info *info) struct net_device *dev; struct cfg80211_ibss_params ibss; struct wiphy *wiphy; + struct cfg80211_cached_keys *connkeys = NULL; int err; memset(&ibss, 0, sizeof(ibss)); @@ -3570,13 +3676,26 @@ static int nl80211_join_ibss(struct sk_buff *skb, struct genl_info *info) } ibss.channel_fixed = !!info->attrs[NL80211_ATTR_FREQ_FIXED]; + ibss.privacy = !!info->attrs[NL80211_ATTR_PRIVACY]; + + if (ibss.privacy && info->attrs[NL80211_ATTR_KEYS]) { + connkeys = nl80211_parse_connkeys(rdev, + info->attrs[NL80211_ATTR_KEYS]); + if (IS_ERR(connkeys)) { + err = PTR_ERR(connkeys); + connkeys = NULL; + goto out; + } + } - err = cfg80211_join_ibss(rdev, dev, &ibss); + err = cfg80211_join_ibss(rdev, dev, &ibss, connkeys); out: cfg80211_unlock_rdev(rdev); dev_put(dev); unlock_rtnl: + if (err) + kfree(connkeys); rtnl_unlock(); return err; } @@ -3746,6 +3865,7 @@ static int nl80211_connect(struct sk_buff *skb, struct genl_info *info) struct net_device *dev; struct cfg80211_connect_params connect; struct wiphy *wiphy; + struct cfg80211_cached_keys *connkeys = NULL; int err; memset(&connect, 0, sizeof(connect)); @@ -3810,12 +3930,24 @@ static int nl80211_connect(struct sk_buff *skb, struct genl_info *info) } } - err = cfg80211_connect(rdev, dev, &connect); + if (connect.privacy && info->attrs[NL80211_ATTR_KEYS]) { + connkeys = nl80211_parse_connkeys(rdev, + info->attrs[NL80211_ATTR_KEYS]); + if (IS_ERR(connkeys)) { + err = PTR_ERR(connkeys); + connkeys = NULL; + goto out; + } + } + + err = cfg80211_connect(rdev, dev, &connect, connkeys); out: cfg80211_unlock_rdev(rdev); dev_put(dev); unlock_rtnl: + if (err) + kfree(connkeys); rtnl_unlock(); return err; } diff --git a/net/wireless/sme.c b/net/wireless/sme.c index 79ca56cbfd36..d635a99dba51 100644 --- a/net/wireless/sme.c +++ b/net/wireless/sme.c @@ -125,7 +125,9 @@ static int cfg80211_conn_do_work(struct wireless_dev *wdev) params->channel, params->auth_type, params->bssid, params->ssid, params->ssid_len, - NULL, 0); + NULL, 0, + params->key, params->key_len, + params->key_idx); case CFG80211_CONN_ASSOCIATE_NEXT: BUG_ON(!rdev->ops->assoc); wdev->conn->state = CFG80211_CONN_ASSOCIATING; @@ -279,8 +281,12 @@ void cfg80211_sme_rx_auth(struct net_device *dev, /* select automatically between only open, shared, leap */ switch (wdev->conn->params.auth_type) { case NL80211_AUTHTYPE_OPEN_SYSTEM: - wdev->conn->params.auth_type = - NL80211_AUTHTYPE_SHARED_KEY; + if (wdev->connect_keys) + wdev->conn->params.auth_type = + NL80211_AUTHTYPE_SHARED_KEY; + else + wdev->conn->params.auth_type = + NL80211_AUTHTYPE_NETWORK_EAP; break; case NL80211_AUTHTYPE_SHARED_KEY: wdev->conn->params.auth_type = @@ -353,10 +359,8 @@ void __cfg80211_connect_result(struct net_device *dev, const u8 *bssid, #endif if (status == WLAN_STATUS_SUCCESS && - wdev->sme_state == CFG80211_SME_IDLE) { - wdev->sme_state = CFG80211_SME_CONNECTED; - return; - } + wdev->sme_state == CFG80211_SME_IDLE) + goto success; if (wdev->sme_state != CFG80211_SME_CONNECTING) return; @@ -370,24 +374,29 @@ void __cfg80211_connect_result(struct net_device *dev, const u8 *bssid, if (wdev->conn) wdev->conn->state = CFG80211_CONN_IDLE; - if (status == WLAN_STATUS_SUCCESS) { - bss = cfg80211_get_bss(wdev->wiphy, NULL, bssid, - wdev->ssid, wdev->ssid_len, - WLAN_CAPABILITY_ESS, - WLAN_CAPABILITY_ESS); - - if (WARN_ON(!bss)) - return; - - cfg80211_hold_bss(bss_from_pub(bss)); - wdev->current_bss = bss_from_pub(bss); - - wdev->sme_state = CFG80211_SME_CONNECTED; - } else { + if (status != WLAN_STATUS_SUCCESS) { wdev->sme_state = CFG80211_SME_IDLE; kfree(wdev->conn); wdev->conn = NULL; + kfree(wdev->connect_keys); + wdev->connect_keys = NULL; + return; } + + bss = cfg80211_get_bss(wdev->wiphy, NULL, bssid, + wdev->ssid, wdev->ssid_len, + WLAN_CAPABILITY_ESS, + WLAN_CAPABILITY_ESS); + + if (WARN_ON(!bss)) + return; + + cfg80211_hold_bss(bss_from_pub(bss)); + wdev->current_bss = bss_from_pub(bss); + + success: + wdev->sme_state = CFG80211_SME_CONNECTED; + cfg80211_upload_connect_keys(wdev); } void cfg80211_connect_result(struct net_device *dev, const u8 *bssid, @@ -516,6 +525,8 @@ void __cfg80211_disconnected(struct net_device *dev, const u8 *ie, size_t ie_len, u16 reason, bool from_ap) { struct wireless_dev *wdev = dev->ieee80211_ptr; + struct cfg80211_registered_device *rdev = wiphy_to_dev(wdev->wiphy); + int i; #ifdef CONFIG_WIRELESS_EXT union iwreq_data wrqu; #endif @@ -543,8 +554,15 @@ void __cfg80211_disconnected(struct net_device *dev, const u8 *ie, wdev->conn = NULL; } - nl80211_send_disconnected(wiphy_to_dev(wdev->wiphy), dev, - reason, ie, ie_len, from_ap); + nl80211_send_disconnected(rdev, dev, reason, ie, ie_len, from_ap); + + /* + * Delete all the keys ... pairwise keys can't really + * exist any more anyway, but default keys might. + */ + if (rdev->ops->del_key) + for (i = 0; i < 6; i++) + rdev->ops->del_key(wdev->wiphy, dev, i, NULL); #ifdef CONFIG_WIRELESS_EXT memset(&wrqu, 0, sizeof(wrqu)); @@ -580,7 +598,8 @@ EXPORT_SYMBOL(cfg80211_disconnected); int __cfg80211_connect(struct cfg80211_registered_device *rdev, struct net_device *dev, - struct cfg80211_connect_params *connect) + struct cfg80211_connect_params *connect, + struct cfg80211_cached_keys *connkeys) { struct wireless_dev *wdev = dev->ieee80211_ptr; int err; @@ -590,6 +609,24 @@ int __cfg80211_connect(struct cfg80211_registered_device *rdev, if (wdev->sme_state != CFG80211_SME_IDLE) return -EALREADY; + if (WARN_ON(wdev->connect_keys)) { + kfree(wdev->connect_keys); + wdev->connect_keys = NULL; + } + + if (connkeys && connkeys->def >= 0) { + int idx; + + idx = connkeys->def; + /* If given a WEP key we may need it for shared key auth */ + if (connkeys->params[idx].cipher == WLAN_CIPHER_SUITE_WEP40 || + connkeys->params[idx].cipher == WLAN_CIPHER_SUITE_WEP104) { + connect->key_idx = idx; + connect->key = connkeys->params[idx].key; + connect->key_len = connkeys->params[idx].key_len; + } + } + if (!rdev->ops->connect) { if (!rdev->ops->auth || !rdev->ops->assoc) return -EOPNOTSUPP; @@ -640,6 +677,7 @@ int __cfg80211_connect(struct cfg80211_registered_device *rdev, cfg80211_get_conn_bss(wdev); wdev->sme_state = CFG80211_SME_CONNECTING; + wdev->connect_keys = connkeys; /* we're good if we have both BSSID and channel */ if (wdev->conn->params.bssid && wdev->conn->params.channel) { @@ -662,13 +700,16 @@ int __cfg80211_connect(struct cfg80211_registered_device *rdev, kfree(wdev->conn); wdev->conn = NULL; wdev->sme_state = CFG80211_SME_IDLE; + wdev->connect_keys = NULL; } return err; } else { wdev->sme_state = CFG80211_SME_CONNECTING; + wdev->connect_keys = connkeys; err = rdev->ops->connect(&rdev->wiphy, dev, connect); if (err) { + wdev->connect_keys = NULL; wdev->sme_state = CFG80211_SME_IDLE; return err; } @@ -682,12 +723,13 @@ int __cfg80211_connect(struct cfg80211_registered_device *rdev, int cfg80211_connect(struct cfg80211_registered_device *rdev, struct net_device *dev, - struct cfg80211_connect_params *connect) + struct cfg80211_connect_params *connect, + struct cfg80211_cached_keys *connkeys) { int err; wdev_lock(dev->ieee80211_ptr); - err = __cfg80211_connect(rdev, dev, connect); + err = __cfg80211_connect(rdev, dev, connect, connkeys); wdev_unlock(dev->ieee80211_ptr); return err; @@ -704,6 +746,9 @@ int __cfg80211_disconnect(struct cfg80211_registered_device *rdev, if (wdev->sme_state == CFG80211_SME_IDLE) return -EINVAL; + kfree(wdev->connect_keys); + wdev->connect_keys = NULL; + if (!rdev->ops->disconnect) { if (!rdev->ops->deauth) return -EOPNOTSUPP; diff --git a/net/wireless/util.c b/net/wireless/util.c index 28f8f96801d4..4bab380a1204 100644 --- a/net/wireless/util.c +++ b/net/wireless/util.c @@ -141,9 +141,12 @@ void ieee80211_set_bitrate_flags(struct wiphy *wiphy) set_mandatory_flags_band(wiphy->bands[band], band); } -int cfg80211_validate_key_settings(struct key_params *params, int key_idx, +int cfg80211_validate_key_settings(struct cfg80211_registered_device *rdev, + struct key_params *params, int key_idx, const u8 *mac_addr) { + int i; + if (key_idx > 5) return -EINVAL; @@ -197,6 +200,12 @@ int cfg80211_validate_key_settings(struct key_params *params, int key_idx, } } + for (i = 0; i < rdev->wiphy.n_cipher_suites; i++) + if (params->cipher == rdev->wiphy.cipher_suites[i]) + break; + if (i == rdev->wiphy.n_cipher_suites) + return -EINVAL; + return 0; } @@ -523,3 +532,33 @@ const u8 *ieee80211_bss_get_ie(struct cfg80211_bss *bss, u8 ie) return NULL; } EXPORT_SYMBOL(ieee80211_bss_get_ie); + +void cfg80211_upload_connect_keys(struct wireless_dev *wdev) +{ + struct cfg80211_registered_device *rdev = wiphy_to_dev(wdev->wiphy); + struct net_device *dev = wdev->netdev; + int i; + + if (!wdev->connect_keys) + return; + + for (i = 0; i < 6; i++) { + if (!wdev->connect_keys->params[i].cipher) + continue; + if (rdev->ops->add_key(wdev->wiphy, dev, i, NULL, + &wdev->connect_keys->params[i])) + printk(KERN_ERR "%s: failed to set key %d\n", + dev->name, i); + if (wdev->connect_keys->def == i) + if (rdev->ops->set_default_key(wdev->wiphy, dev, i)) + printk(KERN_ERR "%s: failed to set defkey %d\n", + dev->name, i); + if (wdev->connect_keys->defmgmt == i) + if (rdev->ops->set_default_mgmt_key(wdev->wiphy, dev, i)) + printk(KERN_ERR "%s: failed to set mgtdef %d\n", + dev->name, i); + } + + kfree(wdev->connect_keys); + wdev->connect_keys = NULL; +} diff --git a/net/wireless/wext-compat.c b/net/wireless/wext-compat.c index 5088d89a30fc..5d0176338539 100644 --- a/net/wireless/wext-compat.c +++ b/net/wireless/wext-compat.c @@ -453,15 +453,32 @@ int cfg80211_wext_giwretry(struct net_device *dev, } EXPORT_SYMBOL_GPL(cfg80211_wext_giwretry); -static int cfg80211_set_encryption(struct cfg80211_registered_device *rdev, - struct net_device *dev, const u8 *addr, - bool remove, bool tx_key, int idx, - struct key_params *params) +static int __cfg80211_set_encryption(struct cfg80211_registered_device *rdev, + struct net_device *dev, const u8 *addr, + bool remove, bool tx_key, int idx, + struct key_params *params) { struct wireless_dev *wdev = dev->ieee80211_ptr; - int err; + int err, i; + + if (!wdev->wext.keys) { + wdev->wext.keys = kzalloc(sizeof(*wdev->wext.keys), + GFP_KERNEL); + if (!wdev->wext.keys) + return -ENOMEM; + for (i = 0; i < 6; i++) + wdev->wext.keys->params[i].key = + wdev->wext.keys->data[i]; + } + + if (wdev->iftype != NL80211_IFTYPE_ADHOC && + wdev->iftype != NL80211_IFTYPE_STATION) + return -EOPNOTSUPP; if (params->cipher == WLAN_CIPHER_SUITE_AES_CMAC) { + if (!wdev->current_bss) + return -ENOLINK; + if (!rdev->ops->set_default_mgmt_key) return -EOPNOTSUPP; @@ -471,8 +488,14 @@ static int cfg80211_set_encryption(struct cfg80211_registered_device *rdev, return -EINVAL; if (remove) { - err = rdev->ops->del_key(&rdev->wiphy, dev, idx, addr); + err = 0; + if (wdev->current_bss) + err = rdev->ops->del_key(&rdev->wiphy, dev, idx, addr); if (!err) { + if (!addr) { + wdev->wext.keys->params[idx].key_len = 0; + wdev->wext.keys->params[idx].cipher = 0; + } if (idx == wdev->wext.default_key) wdev->wext.default_key = -1; else if (idx == wdev->wext.default_mgmt_key) @@ -486,36 +509,64 @@ static int cfg80211_set_encryption(struct cfg80211_registered_device *rdev, return 0; return err; - } else { - if (addr) - tx_key = false; + } - if (cfg80211_validate_key_settings(params, idx, addr)) - return -EINVAL; + if (addr) + tx_key = false; + if (cfg80211_validate_key_settings(rdev, params, idx, addr)) + return -EINVAL; + + err = 0; + if (wdev->current_bss) err = rdev->ops->add_key(&rdev->wiphy, dev, idx, addr, params); - if (err) - return err; + if (err) + return err; + + if (!addr) { + wdev->wext.keys->params[idx] = *params; + memcpy(wdev->wext.keys->data[idx], + params->key, params->key_len); + wdev->wext.keys->params[idx].key = + wdev->wext.keys->data[idx]; + } - if (tx_key || (!addr && wdev->wext.default_key == -1)) { + if (params->cipher != WLAN_CIPHER_SUITE_AES_CMAC && + (tx_key || (!addr && wdev->wext.default_key == -1))) { + if (wdev->current_bss) err = rdev->ops->set_default_key(&rdev->wiphy, dev, idx); - if (!err) - wdev->wext.default_key = idx; - return err; - } + if (!err) + wdev->wext.default_key = idx; + return err; + } - if (params->cipher == WLAN_CIPHER_SUITE_AES_CMAC && - (tx_key || (!addr && wdev->wext.default_mgmt_key == -1))) { + if (params->cipher == WLAN_CIPHER_SUITE_AES_CMAC && + (tx_key || (!addr && wdev->wext.default_mgmt_key == -1))) { + if (wdev->current_bss) err = rdev->ops->set_default_mgmt_key(&rdev->wiphy, dev, idx); - if (!err) - wdev->wext.default_mgmt_key = idx; - return err; - } - - return 0; + if (!err) + wdev->wext.default_mgmt_key = idx; + return err; } + + return 0; +} + +static int cfg80211_set_encryption(struct cfg80211_registered_device *rdev, + struct net_device *dev, const u8 *addr, + bool remove, bool tx_key, int idx, + struct key_params *params) +{ + int err; + + wdev_lock(dev->ieee80211_ptr); + err = __cfg80211_set_encryption(rdev, dev, addr, remove, + tx_key, idx, params); + wdev_unlock(dev->ieee80211_ptr); + + return err; } int cfg80211_wext_siwencode(struct net_device *dev, @@ -528,6 +579,10 @@ int cfg80211_wext_siwencode(struct net_device *dev, bool remove = false; struct key_params params; + if (wdev->iftype != NL80211_IFTYPE_STATION && + wdev->iftype != NL80211_IFTYPE_ADHOC) + return -EOPNOTSUPP; + /* no use -- only MFP (set_default_mgmt_key) is optional */ if (!rdev->ops->del_key || !rdev->ops->add_key || @@ -548,9 +603,14 @@ int cfg80211_wext_siwencode(struct net_device *dev, remove = true; else if (erq->length == 0) { /* No key data - just set the default TX key index */ - err = rdev->ops->set_default_key(&rdev->wiphy, dev, idx); + err = 0; + wdev_lock(wdev); + if (wdev->current_bss) + err = rdev->ops->set_default_key(&rdev->wiphy, + dev, idx); if (!err) wdev->wext.default_key = idx; + wdev_unlock(wdev); return err; } @@ -583,6 +643,10 @@ int cfg80211_wext_siwencodeext(struct net_device *dev, struct key_params params; u32 cipher; + if (wdev->iftype != NL80211_IFTYPE_STATION && + wdev->iftype != NL80211_IFTYPE_ADHOC) + return -EOPNOTSUPP; + /* no use -- only MFP (set_default_mgmt_key) is optional */ if (!rdev->ops->del_key || !rdev->ops->add_key || @@ -656,37 +720,15 @@ int cfg80211_wext_siwencodeext(struct net_device *dev, } EXPORT_SYMBOL_GPL(cfg80211_wext_siwencodeext); -struct giwencode_cookie { - size_t buflen; - char *keybuf; -}; - -static void giwencode_get_key_cb(void *cookie, struct key_params *params) -{ - struct giwencode_cookie *data = cookie; - - if (!params->key) { - data->buflen = 0; - return; - } - - data->buflen = min_t(size_t, data->buflen, params->key_len); - memcpy(data->keybuf, params->key, data->buflen); -} - int cfg80211_wext_giwencode(struct net_device *dev, struct iw_request_info *info, struct iw_point *erq, char *keybuf) { struct wireless_dev *wdev = dev->ieee80211_ptr; - struct cfg80211_registered_device *rdev = wiphy_to_dev(wdev->wiphy); - int idx, err; - struct giwencode_cookie data = { - .keybuf = keybuf, - .buflen = erq->length, - }; + int idx; - if (!rdev->ops->get_key) + if (wdev->iftype != NL80211_IFTYPE_STATION && + wdev->iftype != NL80211_IFTYPE_ADHOC) return -EOPNOTSUPP; idx = erq->flags & IW_ENCODE_INDEX; @@ -701,21 +743,18 @@ int cfg80211_wext_giwencode(struct net_device *dev, erq->flags = idx + 1; - err = rdev->ops->get_key(&rdev->wiphy, dev, idx, NULL, &data, - giwencode_get_key_cb); - if (!err) { - erq->length = data.buflen; - erq->flags |= IW_ENCODE_ENABLED; - return 0; - } - - if (err == -ENOENT) { + if (!wdev->wext.keys || !wdev->wext.keys->params[idx].cipher) { erq->flags |= IW_ENCODE_DISABLED; erq->length = 0; return 0; } - return err; + erq->length = min_t(size_t, erq->length, + wdev->wext.keys->params[idx].key_len); + memcpy(keybuf, wdev->wext.keys->params[idx].key, erq->length); + erq->flags |= IW_ENCODE_ENABLED; + + return 0; } EXPORT_SYMBOL_GPL(cfg80211_wext_giwencode); diff --git a/net/wireless/wext-sme.c b/net/wireless/wext-sme.c index 6f75aaa7f795..c33ea9a5de78 100644 --- a/net/wireless/wext-sme.c +++ b/net/wireless/wext-sme.c @@ -10,10 +10,11 @@ #include #include "nl80211.h" -static int cfg80211_mgd_wext_connect(struct cfg80211_registered_device *rdev, - struct wireless_dev *wdev) +int cfg80211_mgd_wext_connect(struct cfg80211_registered_device *rdev, + struct wireless_dev *wdev) { - int err; + struct cfg80211_cached_keys *ck = NULL; + int err, i; ASSERT_RDEV_LOCK(rdev); ASSERT_WDEV_LOCK(wdev); @@ -25,10 +26,25 @@ static int cfg80211_mgd_wext_connect(struct cfg80211_registered_device *rdev, wdev->wext.connect.ie_len = wdev->wext.ie_len; wdev->wext.connect.privacy = wdev->wext.default_key != -1; - err = 0; - if (wdev->wext.connect.ssid_len != 0) - err = __cfg80211_connect(rdev, wdev->netdev, - &wdev->wext.connect); + if (wdev->wext.keys) { + wdev->wext.keys->def = wdev->wext.default_key; + wdev->wext.keys->defmgmt = wdev->wext.default_mgmt_key; + } + + if (!wdev->wext.connect.ssid_len) + return 0; + + if (wdev->wext.keys) { + ck = kmemdup(wdev->wext.keys, sizeof(*ck), GFP_KERNEL); + if (!ck) + return -ENOMEM; + for (i = 0; i < 6; i++) + ck->params[i].key = ck->data[i]; + } + err = __cfg80211_connect(rdev, wdev->netdev, + &wdev->wext.connect, ck); + if (err) + kfree(ck); return err; } -- cgit v1.2.3 From ca3dbc20d47ae43c201c215259d078e227bfcf01 Mon Sep 17 00:00:00 2001 From: Helmut Schaa Date: Fri, 10 Jul 2009 14:54:58 +0200 Subject: cfg80211: update misleading comment In cfg80211_scan_request n_channels refers to the total number of channels to scan. Update the misleading comment accordingly. Signed-off-by: Helmut Schaa Signed-off-by: John W. Linville --- include/net/cfg80211.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index 65a5cbcb5d14..a981ca8a5701 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -538,7 +538,7 @@ struct cfg80211_ssid { * @ssids: SSIDs to scan for (active scan only) * @n_ssids: number of SSIDs * @channels: channels to scan on. - * @n_channels: number of channels for each band + * @n_channels: total number of channels to scan * @ie: optional information element(s) to add into Probe Request or %NULL * @ie_len: length of ie in octets * @wiphy: the wiphy this was for -- cgit v1.2.3 From 48ab905d1a81b7df33a33def04a890e4e0c51460 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Fri, 10 Jul 2009 18:42:31 +0200 Subject: nl80211: report BSS status When connected to a BSS, or joined to an IBSS, we'll want to know in userspace without using wireless extensions, so report the BSS status in the BSS list. Userspace can query the BSS list, display all the information and retrieve the station information as well. For example (from hwsim): $ iw dev wlan1 scan dump BSS 02:00:00:00:00:00 (on wlan1) -- associated freq: 2462 beacon interval: 100 capability: ESS ShortSlotTime (0x0401) signal: -50.00 dBm SSID: j Supported rates: 1.0* 2.0* 5.5* 11.0* 6.0 9.0 12.0 18.0 DS Paramater set: channel 11 ERP: Extended supported rates: 24.0 36.0 48.0 54.0 Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- include/linux/nl80211.h | 11 +++++++++ net/wireless/nl80211.c | 65 ++++++++++++++++++++++++++++++++++++------------- 2 files changed, 59 insertions(+), 17 deletions(-) (limited to 'include') diff --git a/include/linux/nl80211.h b/include/linux/nl80211.h index b043b78dd2c3..962e2232a074 100644 --- a/include/linux/nl80211.h +++ b/include/linux/nl80211.h @@ -1260,6 +1260,7 @@ enum nl80211_channel_type { * in mBm (100 * dBm) (s32) * @NL80211_BSS_SIGNAL_UNSPEC: signal strength of the probe response/beacon * in unspecified units, scaled to 0..100 (u8) + * @NL80211_BSS_STATUS: status, if this BSS is "used" * @__NL80211_BSS_AFTER_LAST: internal * @NL80211_BSS_MAX: highest BSS attribute */ @@ -1273,12 +1274,22 @@ enum nl80211_bss { NL80211_BSS_INFORMATION_ELEMENTS, NL80211_BSS_SIGNAL_MBM, NL80211_BSS_SIGNAL_UNSPEC, + NL80211_BSS_STATUS, /* keep last */ __NL80211_BSS_AFTER_LAST, NL80211_BSS_MAX = __NL80211_BSS_AFTER_LAST - 1 }; +/** + * enum nl80211_bss_status - BSS "status" + */ +enum nl80211_bss_status { + NL80211_BSS_STATUS_AUTHENTICATED, + NL80211_BSS_STATUS_ASSOCIATED, + NL80211_BSS_STATUS_IBSS_JOINED, +}; + /** * enum nl80211_auth_type - AuthenticationType * diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index 45c5f9c8e51b..da450ef1fc7e 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -3094,11 +3094,15 @@ static int nl80211_trigger_scan(struct sk_buff *skb, struct genl_info *info) static int nl80211_send_bss(struct sk_buff *msg, u32 pid, u32 seq, int flags, struct cfg80211_registered_device *rdev, - struct net_device *dev, - struct cfg80211_bss *res) + struct wireless_dev *wdev, + struct cfg80211_internal_bss *intbss) { + struct cfg80211_bss *res = &intbss->pub; void *hdr; struct nlattr *bss; + int i; + + ASSERT_WDEV_LOCK(wdev); hdr = nl80211hdr_put(msg, pid, seq, flags, NL80211_CMD_NEW_SCAN_RESULTS); @@ -3107,7 +3111,7 @@ static int nl80211_send_bss(struct sk_buff *msg, u32 pid, u32 seq, int flags, NLA_PUT_U32(msg, NL80211_ATTR_SCAN_GENERATION, rdev->bss_generation); - NLA_PUT_U32(msg, NL80211_ATTR_IFINDEX, dev->ifindex); + NLA_PUT_U32(msg, NL80211_ATTR_IFINDEX, wdev->netdev->ifindex); bss = nla_nest_start(msg, NL80211_ATTR_BSS); if (!bss) @@ -3136,6 +3140,28 @@ static int nl80211_send_bss(struct sk_buff *msg, u32 pid, u32 seq, int flags, break; } + switch (wdev->iftype) { + case NL80211_IFTYPE_STATION: + if (intbss == wdev->current_bss) + NLA_PUT_U32(msg, NL80211_BSS_STATUS, + NL80211_BSS_STATUS_ASSOCIATED); + else for (i = 0; i < MAX_AUTH_BSSES; i++) { + if (intbss != wdev->auth_bsses[i]) + continue; + NLA_PUT_U32(msg, NL80211_BSS_STATUS, + NL80211_BSS_STATUS_AUTHENTICATED); + break; + } + break; + case NL80211_IFTYPE_ADHOC: + if (intbss == wdev->current_bss) + NLA_PUT_U32(msg, NL80211_BSS_STATUS, + NL80211_BSS_STATUS_IBSS_JOINED); + break; + default: + break; + } + nla_nest_end(msg, bss); return genlmsg_end(msg, hdr); @@ -3148,9 +3174,10 @@ static int nl80211_send_bss(struct sk_buff *msg, u32 pid, u32 seq, int flags, static int nl80211_dump_scan(struct sk_buff *skb, struct netlink_callback *cb) { - struct cfg80211_registered_device *dev; - struct net_device *netdev; + struct cfg80211_registered_device *rdev; + struct net_device *dev; struct cfg80211_internal_bss *scan; + struct wireless_dev *wdev; int ifidx = cb->args[0]; int start = cb->args[1], idx = 0; int err; @@ -3171,39 +3198,43 @@ static int nl80211_dump_scan(struct sk_buff *skb, cb->args[0] = ifidx; } - netdev = dev_get_by_index(&init_net, ifidx); - if (!netdev) + dev = dev_get_by_index(&init_net, ifidx); + if (!dev) return -ENODEV; - dev = cfg80211_get_dev_from_ifindex(ifidx); - if (IS_ERR(dev)) { - err = PTR_ERR(dev); + rdev = cfg80211_get_dev_from_ifindex(ifidx); + if (IS_ERR(rdev)) { + err = PTR_ERR(rdev); goto out_put_netdev; } - spin_lock_bh(&dev->bss_lock); - cfg80211_bss_expire(dev); + wdev = dev->ieee80211_ptr; - list_for_each_entry(scan, &dev->bss_list, list) { + wdev_lock(wdev); + spin_lock_bh(&rdev->bss_lock); + cfg80211_bss_expire(rdev); + + list_for_each_entry(scan, &rdev->bss_list, list) { if (++idx <= start) continue; if (nl80211_send_bss(skb, NETLINK_CB(cb->skb).pid, cb->nlh->nlmsg_seq, NLM_F_MULTI, - dev, netdev, &scan->pub) < 0) { + rdev, wdev, scan) < 0) { idx--; goto out; } } out: - spin_unlock_bh(&dev->bss_lock); + spin_unlock_bh(&rdev->bss_lock); + wdev_unlock(wdev); cb->args[1] = idx; err = skb->len; - cfg80211_unlock_rdev(dev); + cfg80211_unlock_rdev(rdev); out_put_netdev: - dev_put(netdev); + dev_put(dev); return err; } -- cgit v1.2.3 From b770b43e95a66587fbd8c1841de83da87fbf23ea Mon Sep 17 00:00:00 2001 From: "Luis R. Rodriguez" Date: Thu, 16 Jul 2009 10:15:09 -0700 Subject: mac80211: drop frames for sta with no valid rate When we're associated we should be able to send data to target sta. If we cannot we may be trying to use the incorrect band to talk to the sta. Lets catch any such cases, warn, and drop the frames to not invalidate assumptions being made on rate control algorithms when they have a valid sta to communicate with. Any such cases should be handled and fixed. Signed-off-by: Luis R. Rodriguez Signed-off-by: John W. Linville --- include/net/mac80211.h | 11 +++++++++++ net/mac80211/tx.c | 20 ++++++++++++++++++++ 2 files changed, 31 insertions(+) (limited to 'include') diff --git a/include/net/mac80211.h b/include/net/mac80211.h index ce7cb1b5d453..d98fac54577b 100644 --- a/include/net/mac80211.h +++ b/include/net/mac80211.h @@ -2110,6 +2110,17 @@ rate_lowest_index(struct ieee80211_supported_band *sband, return 0; } +static inline +bool rate_usable_index_exists(struct ieee80211_supported_band *sband, + struct ieee80211_sta *sta) +{ + unsigned int i; + + for (i = 0; i < sband->n_bitrates; i++) + if (rate_supported(sta, sband->band, i)) + return true; + return false; +} int ieee80211_rate_control_register(struct rate_control_ops *ops); void ieee80211_rate_control_unregister(struct rate_control_ops *ops); diff --git a/net/mac80211/tx.c b/net/mac80211/tx.c index 60ae086995b1..f3efd4f16e91 100644 --- a/net/mac80211/tx.c +++ b/net/mac80211/tx.c @@ -512,6 +512,7 @@ ieee80211_tx_h_rate_ctrl(struct ieee80211_tx_data *tx) int i, len; bool inval = false, rts = false, short_preamble = false; struct ieee80211_tx_rate_control txrc; + u32 sta_flags; memset(&txrc, 0, sizeof(txrc)); @@ -544,7 +545,26 @@ ieee80211_tx_h_rate_ctrl(struct ieee80211_tx_data *tx) (tx->sta && test_sta_flags(tx->sta, WLAN_STA_SHORT_PREAMBLE)))) txrc.short_preamble = short_preamble = true; + sta_flags = tx->sta ? get_sta_flags(tx->sta) : 0; + + /* + * Lets not bother rate control if we're associated and cannot + * talk to the sta. This should not happen. + */ + if (WARN((tx->local->sw_scanning) && + (sta_flags & WLAN_STA_ASSOC) && + !rate_usable_index_exists(sband, &tx->sta->sta), + "%s: Dropped data frame as no usable bitrate found while " + "scanning and associated. Target station: " + "%pM on %d GHz band\n", + tx->dev->name, hdr->addr1, + tx->channel->band ? 5 : 2)) + return TX_DROP; + /* + * If we're associated with the sta at this point we know we can at + * least send the frame at the lowest bit rate. + */ rate_control_get_rate(tx->sdata, tx->sta, &txrc); if (unlikely(info->control.rates[0].idx < 0)) -- cgit v1.2.3 From 4c6d4f5c33fbe19b134c1af43af166fee79eb986 Mon Sep 17 00:00:00 2001 From: "Luis R. Rodriguez" Date: Thu, 16 Jul 2009 10:05:41 -0700 Subject: mac80211: add helper for management / no-ack frame rate decision All current rate control algorithms agree to send management and no-ack frames at the lowest rate. They also agree to do this when sta and the private rate control data is NULL. We add a hlper to mac80211 for this and simplify the rate control algorithm code. Developers wishing to make enhancements to rate control algorithms are for broadcast/multicast can opt to not use this in their gate_rate() mac80211 callback. Cc: Zhu Yi Acked-by: Reinette Chatre Cc: ipw3945-devel@lists.sourceforge.net Cc: Gabor Juhos Acked-by: Felix Fietkau Cc: Derek Smithies Cc: Chittajit Mitra Signed-off-by: Luis R. Rodriguez Signed-off-by: John W. Linville --- drivers/net/wireless/ath/ath9k/rc.c | 14 +------------- drivers/net/wireless/iwlwifi/iwl-3945-rs.c | 13 +++---------- drivers/net/wireless/iwlwifi/iwl-agn-rs.c | 7 +------ include/net/mac80211.h | 23 +++++++++++++++++++++++ net/mac80211/rate.c | 29 +++++++++++++++++++++++++++++ net/mac80211/rc80211_minstrel.c | 22 +--------------------- net/mac80211/rc80211_pid_algo.c | 11 +---------- 7 files changed, 59 insertions(+), 60 deletions(-) (limited to 'include') diff --git a/drivers/net/wireless/ath/ath9k/rc.c b/drivers/net/wireless/ath/ath9k/rc.c index 2c72901adbee..630fcf46e0dd 100644 --- a/drivers/net/wireless/ath/ath9k/rc.c +++ b/drivers/net/wireless/ath/ath9k/rc.c @@ -1518,23 +1518,11 @@ exit: static void ath_get_rate(void *priv, struct ieee80211_sta *sta, void *priv_sta, struct ieee80211_tx_rate_control *txrc) { - struct ieee80211_supported_band *sband = txrc->sband; - struct sk_buff *skb = txrc->skb; - struct ieee80211_hdr *hdr = (struct ieee80211_hdr *)skb->data; - struct ieee80211_tx_info *tx_info = IEEE80211_SKB_CB(skb); struct ath_softc *sc = priv; struct ath_rate_priv *ath_rc_priv = priv_sta; - __le16 fc = hdr->frame_control; - /* lowest rate for management and NO_ACK frames */ - if (!ieee80211_is_data(fc) || - tx_info->flags & IEEE80211_TX_CTL_NO_ACK || !sta) { - tx_info->control.rates[0].idx = rate_lowest_index(sband, sta); - tx_info->control.rates[0].count = - (tx_info->flags & IEEE80211_TX_CTL_NO_ACK) ? - 1 : ATH_MGT_TXMAXTRY; + if (rate_control_send_low(sta, priv_sta, txrc)) return; - } /* Find tx rate for unicast frames */ ath_rc_ratefind(sc, ath_rc_priv, txrc); diff --git a/drivers/net/wireless/iwlwifi/iwl-3945-rs.c b/drivers/net/wireless/iwlwifi/iwl-3945-rs.c index 2b776924d5e7..a16bd4147eac 100644 --- a/drivers/net/wireless/iwlwifi/iwl-3945-rs.c +++ b/drivers/net/wireless/iwlwifi/iwl-3945-rs.c @@ -673,7 +673,6 @@ static void rs_get_rate(void *priv_r, struct ieee80211_sta *sta, s8 scale_action = 0; unsigned long flags; struct ieee80211_hdr *hdr = (struct ieee80211_hdr *)skb->data; - __le16 fc; u16 rate_mask = sta ? sta->supp_rates[sband->band] : 0; s8 max_rate_idx = -1; struct iwl_priv *priv = (struct iwl_priv *)priv_r; @@ -681,16 +680,10 @@ static void rs_get_rate(void *priv_r, struct ieee80211_sta *sta, IWL_DEBUG_RATE(priv, "enter\n"); - /* Send management frames and NO_ACK data using lowest rate. */ - fc = hdr->frame_control; - if (!ieee80211_is_data(fc) || info->flags & IEEE80211_TX_CTL_NO_ACK || - !sta || !priv_sta) { - IWL_DEBUG_RATE(priv, "leave: No STA priv data to update!\n"); - info->control.rates[0].idx = rate_lowest_index(sband, sta); - if (info->flags & IEEE80211_TX_CTL_NO_ACK) - info->control.rates[0].count = 1; + if (rate_control_send_low(sta, priv_sta, txrc)) return; - } + + rate_mask = sta->supp_rates[sband->band]; /* get user max rate if set */ max_rate_idx = txrc->max_rate_idx; diff --git a/drivers/net/wireless/iwlwifi/iwl-agn-rs.c b/drivers/net/wireless/iwlwifi/iwl-agn-rs.c index 3fea027f35d1..63280411fd58 100644 --- a/drivers/net/wireless/iwlwifi/iwl-agn-rs.c +++ b/drivers/net/wireless/iwlwifi/iwl-agn-rs.c @@ -2481,13 +2481,8 @@ static void rs_get_rate(void *priv_r, struct ieee80211_sta *sta, void *priv_sta, } /* Send management frames and NO_ACK data using lowest rate. */ - if (!ieee80211_is_data(hdr->frame_control) || - info->flags & IEEE80211_TX_CTL_NO_ACK || !sta || !lq_sta) { - info->control.rates[0].idx = rate_lowest_index(sband, sta); - if (info->flags & IEEE80211_TX_CTL_NO_ACK) - info->control.rates[0].count = 1; + if (rate_control_send_low(sta, priv_sta, txrc)) return; - } rate_idx = lq_sta->last_txrate_idx; diff --git a/include/net/mac80211.h b/include/net/mac80211.h index d98fac54577b..a861259c3050 100644 --- a/include/net/mac80211.h +++ b/include/net/mac80211.h @@ -2094,6 +2094,29 @@ static inline int rate_supported(struct ieee80211_sta *sta, return (sta == NULL || sta->supp_rates[band] & BIT(index)); } +/** + * rate_control_send_low - helper for drivers for management/no-ack frames + * + * Rate control algorithms that agree to use the lowest rate to + * send management frames and NO_ACK data with the respective hw + * retries should use this in the beginning of their mac80211 get_rate + * callback. If true is returned the rate control can simply return. + * If false is returned we guarantee that sta and sta and priv_sta is + * not null. + * + * Rate control algorithms wishing to do more intelligent selection of + * rate for multicast/broadcast frames may choose to not use this. + * + * @sta: &struct ieee80211_sta pointer to the target destination. Note + * that this may be null. + * @priv_sta: private rate control structure. This may be null. + * @txrc: rate control information we sholud populate for mac80211. + */ +bool rate_control_send_low(struct ieee80211_sta *sta, + void *priv_sta, + struct ieee80211_tx_rate_control *txrc); + + static inline s8 rate_lowest_index(struct ieee80211_supported_band *sband, struct ieee80211_sta *sta) diff --git a/net/mac80211/rate.c b/net/mac80211/rate.c index 4641f00a1e5c..8ac7a984d886 100644 --- a/net/mac80211/rate.c +++ b/net/mac80211/rate.c @@ -198,6 +198,35 @@ static void rate_control_release(struct kref *kref) kfree(ctrl_ref); } +static bool rc_no_data_or_no_ack(struct ieee80211_tx_rate_control *txrc) +{ + struct sk_buff *skb = txrc->skb; + struct ieee80211_hdr *hdr = (struct ieee80211_hdr *) skb->data; + struct ieee80211_tx_info *info = IEEE80211_SKB_CB(skb); + __le16 fc; + + fc = hdr->frame_control; + + return ((info->flags & IEEE80211_TX_CTL_NO_ACK) || !ieee80211_is_data(fc)); +} + +bool rate_control_send_low(struct ieee80211_sta *sta, + void *priv_sta, + struct ieee80211_tx_rate_control *txrc) +{ + struct ieee80211_tx_info *info = IEEE80211_SKB_CB(txrc->skb); + + if (!sta || !priv_sta || rc_no_data_or_no_ack(txrc)) { + info->control.rates[0].idx = rate_lowest_index(txrc->sband, sta); + info->control.rates[0].count = + (info->flags & IEEE80211_TX_CTL_NO_ACK) ? + 1 : txrc->hw->max_rate_tries; + return true; + } + return false; +} +EXPORT_SYMBOL(rate_control_send_low); + void rate_control_get_rate(struct ieee80211_sub_if_data *sdata, struct sta_info *sta, struct ieee80211_tx_rate_control *txrc) diff --git a/net/mac80211/rc80211_minstrel.c b/net/mac80211/rc80211_minstrel.c index 5bdce0c951dd..7c5142988bbb 100644 --- a/net/mac80211/rc80211_minstrel.c +++ b/net/mac80211/rc80211_minstrel.c @@ -70,19 +70,6 @@ rix_to_ndx(struct minstrel_sta_info *mi, int rix) return i; } -static inline bool -use_low_rate(struct sk_buff *skb) -{ - struct ieee80211_hdr *hdr = (struct ieee80211_hdr *) skb->data; - struct ieee80211_tx_info *info = IEEE80211_SKB_CB(skb); - __le16 fc; - - fc = hdr->frame_control; - - return ((info->flags & IEEE80211_TX_CTL_NO_ACK) || !ieee80211_is_data(fc)); -} - - static void minstrel_update_stats(struct minstrel_priv *mp, struct minstrel_sta_info *mi) { @@ -231,7 +218,6 @@ minstrel_get_rate(void *priv, struct ieee80211_sta *sta, void *priv_sta, struct ieee80211_tx_rate_control *txrc) { struct sk_buff *skb = txrc->skb; - struct ieee80211_supported_band *sband = txrc->sband; struct ieee80211_tx_info *info = IEEE80211_SKB_CB(skb); struct minstrel_sta_info *mi = priv_sta; struct minstrel_priv *mp = priv; @@ -244,14 +230,8 @@ minstrel_get_rate(void *priv, struct ieee80211_sta *sta, int mrr_ndx[3]; int sample_rate; - if (!sta || !mi || use_low_rate(skb)) { - ar[0].idx = rate_lowest_index(sband, sta); - if (info->flags & IEEE80211_TX_CTL_NO_ACK) - ar[0].count = 1; - else - ar[0].count = mp->max_retry; + if (rate_control_send_low(sta, priv_sta, txrc)) return; - } mrr = mp->has_mrr && !txrc->rts && !txrc->bss_conf->use_cts_prot; diff --git a/net/mac80211/rc80211_pid_algo.c b/net/mac80211/rc80211_pid_algo.c index 549607703bd8..8c053be9dc24 100644 --- a/net/mac80211/rc80211_pid_algo.c +++ b/net/mac80211/rc80211_pid_algo.c @@ -276,11 +276,9 @@ rate_control_pid_get_rate(void *priv, struct ieee80211_sta *sta, { struct sk_buff *skb = txrc->skb; struct ieee80211_supported_band *sband = txrc->sband; - struct ieee80211_hdr *hdr = (struct ieee80211_hdr *) skb->data; struct ieee80211_tx_info *info = IEEE80211_SKB_CB(skb); struct rc_pid_sta_info *spinfo = priv_sta; int rateidx; - __le16 fc; if (txrc->rts) info->control.rates[0].count = @@ -290,15 +288,8 @@ rate_control_pid_get_rate(void *priv, struct ieee80211_sta *sta, txrc->hw->conf.short_frame_max_tx_count; /* Send management frames and NO_ACK data using lowest rate. */ - fc = hdr->frame_control; - if (!sta || !spinfo || !ieee80211_is_data(fc) || - info->flags & IEEE80211_TX_CTL_NO_ACK) { - info->control.rates[0].idx = rate_lowest_index(sband, sta); - if (info->flags & IEEE80211_TX_CTL_NO_ACK) - info->control.rates[0].count = 1; - + if (rate_control_send_low(sta, priv_sta, txrc)) return; - } rateidx = spinfo->txrate_idx; -- cgit v1.2.3 From 3b8d81e020f77c9da8b85b0685c8cd2ca7c7b150 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Wed, 17 Jun 2009 17:43:56 +0200 Subject: mac80211: remove master netdev With the internal 'pending' queue system in place, we can simply put packets there instead of pushing them off to the master dev, getting rid of the master interface completely. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- include/net/mac80211.h | 3 + net/mac80211/agg-tx.c | 3 - net/mac80211/debugfs.c | 2 +- net/mac80211/ieee80211_i.h | 19 ++- net/mac80211/iface.c | 43 ++++--- net/mac80211/main.c | 120 +----------------- net/mac80211/rate.c | 2 +- net/mac80211/rx.c | 16 ++- net/mac80211/scan.c | 19 ++- net/mac80211/tx.c | 299 +++++++++++++++++++-------------------------- net/mac80211/util.c | 52 ++------ net/mac80211/wme.c | 6 +- net/mac80211/wme.h | 3 +- 13 files changed, 195 insertions(+), 392 deletions(-) (limited to 'include') diff --git a/include/net/mac80211.h b/include/net/mac80211.h index a861259c3050..7dd67a1ff4d5 100644 --- a/include/net/mac80211.h +++ b/include/net/mac80211.h @@ -241,6 +241,8 @@ struct ieee80211_bss_conf { * it can be sent out. * @IEEE80211_TX_INTFL_RETRIED: completely internal to mac80211, * used to indicate that a frame was already retried due to PS + * @IEEE80211_TX_INTFL_DONT_ENCRYPT: completely internal to mac80211, + * used to indicate frame should not be encrypted */ enum mac80211_tx_control_flags { IEEE80211_TX_CTL_REQ_TX_STATUS = BIT(0), @@ -259,6 +261,7 @@ enum mac80211_tx_control_flags { IEEE80211_TX_INTFL_RCALGO = BIT(13), IEEE80211_TX_INTFL_NEED_TXPROCESSING = BIT(14), IEEE80211_TX_INTFL_RETRIED = BIT(15), + IEEE80211_TX_INTFL_DONT_ENCRYPT = BIT(16), }; /** diff --git a/net/mac80211/agg-tx.c b/net/mac80211/agg-tx.c index 9e5762ad307d..1958c7c42cd9 100644 --- a/net/mac80211/agg-tx.c +++ b/net/mac80211/agg-tx.c @@ -383,9 +383,6 @@ static void ieee80211_agg_splice_packets(struct ieee80211_local *local, if (!skb_queue_empty(&sta->ampdu_mlme.tid_tx[tid]->pending)) { spin_lock_irqsave(&local->queue_stop_reason_lock, flags); - /* mark queue as pending, it is stopped already */ - __set_bit(IEEE80211_QUEUE_STOP_REASON_PENDING, - &local->queue_stop_reasons[queue]); /* copy over remaining packets */ skb_queue_splice_tail_init( &sta->ampdu_mlme.tid_tx[tid]->pending, diff --git a/net/mac80211/debugfs.c b/net/mac80211/debugfs.c index 6c439cd5ccea..96991b68f048 100644 --- a/net/mac80211/debugfs.c +++ b/net/mac80211/debugfs.c @@ -175,7 +175,7 @@ static ssize_t queues_read(struct file *file, char __user *user_buf, for (q = 0; q < local->hw.queues; q++) res += sprintf(buf + res, "%02d: %#.8lx/%d\n", q, local->queue_stop_reasons[q], - __netif_subqueue_stopped(local->mdev, q)); + skb_queue_len(&local->pending[q])); spin_unlock_irqrestore(&local->queue_stop_reason_lock, flags); return simple_read_from_buffer(user_buf, count, ppos, buf, res); diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h index a34bca2dc52f..6a0177137dd5 100644 --- a/net/mac80211/ieee80211_i.h +++ b/net/mac80211/ieee80211_i.h @@ -567,14 +567,9 @@ enum queue_stop_reason { IEEE80211_QUEUE_STOP_REASON_CSA, IEEE80211_QUEUE_STOP_REASON_AGGREGATION, IEEE80211_QUEUE_STOP_REASON_SUSPEND, - IEEE80211_QUEUE_STOP_REASON_PENDING, IEEE80211_QUEUE_STOP_REASON_SKB_ADD, }; -struct ieee80211_master_priv { - struct ieee80211_local *local; -}; - struct ieee80211_local { /* embed the driver visible part. * don't cast (use the static inlines below), but we keep @@ -587,13 +582,20 @@ struct ieee80211_local { /* also used to protect ampdu_ac_queue and amdpu_ac_stop_refcnt */ spinlock_t queue_stop_reason_lock; - struct net_device *mdev; /* wmaster# - "master" 802.11 device */ int open_count; int monitors, cooked_mntrs; /* number of interfaces with corresponding FIF_ flags */ int fif_fcsfail, fif_plcpfail, fif_control, fif_other_bss; unsigned int filter_flags; /* FIF_* */ struct iw_statistics wstats; + + /* protects the aggregated multicast list and filter calls */ + spinlock_t filter_lock; + + /* aggregated multicast list */ + struct dev_addr_list *mc_list; + int mc_count; + bool tim_in_locked_section; /* see ieee80211_beacon_get() */ /* @@ -813,10 +815,6 @@ struct ieee80211_local { static inline struct ieee80211_sub_if_data * IEEE80211_DEV_TO_SUB_IF(struct net_device *dev) { - struct ieee80211_local *local = wdev_priv(dev->ieee80211_ptr); - - BUG_ON(!local || local->mdev == dev); - return netdev_priv(dev); } @@ -996,7 +994,6 @@ void ieee80211_recalc_idle(struct ieee80211_local *local); /* tx handling */ void ieee80211_clear_tx_pending(struct ieee80211_local *local); void ieee80211_tx_pending(unsigned long data); -int ieee80211_master_start_xmit(struct sk_buff *skb, struct net_device *dev); int ieee80211_monitor_start_xmit(struct sk_buff *skb, struct net_device *dev); int ieee80211_subif_start_xmit(struct sk_buff *skb, struct net_device *dev); diff --git a/net/mac80211/iface.c b/net/mac80211/iface.c index 090aa5a47182..2f797a86ced5 100644 --- a/net/mac80211/iface.c +++ b/net/mac80211/iface.c @@ -190,10 +190,6 @@ static int ieee80211_open(struct net_device *dev) ETH_ALEN); } - if (compare_ether_addr(null_addr, local->mdev->dev_addr) == 0) - memcpy(local->mdev->dev_addr, local->hw.wiphy->perm_addr, - ETH_ALEN); - /* * Validate the MAC address for this device. */ @@ -229,9 +225,9 @@ static int ieee80211_open(struct net_device *dev) if (sdata->u.mntr_flags & MONITOR_FLAG_OTHER_BSS) local->fif_other_bss++; - netif_addr_lock_bh(local->mdev); + spin_lock_bh(&local->filter_lock); ieee80211_configure_filter(local); - netif_addr_unlock_bh(local->mdev); + spin_unlock_bh(&local->filter_lock); break; default: conf.vif = &sdata->vif; @@ -243,9 +239,9 @@ static int ieee80211_open(struct net_device *dev) if (ieee80211_vif_is_mesh(&sdata->vif)) { local->fif_other_bss++; - netif_addr_lock_bh(local->mdev); + spin_lock_bh(&local->filter_lock); ieee80211_configure_filter(local); - netif_addr_unlock_bh(local->mdev); + spin_unlock_bh(&local->filter_lock); ieee80211_start_mesh(sdata); } @@ -279,10 +275,6 @@ static int ieee80211_open(struct net_device *dev) } if (local->open_count == 0) { - res = dev_open(local->mdev); - WARN_ON(res); - if (res) - goto err_del_interface; tasklet_enable(&local->tx_pending_tasklet); tasklet_enable(&local->tasklet); } @@ -393,7 +385,14 @@ static int ieee80211_stop(struct net_device *dev) if (sdata->flags & IEEE80211_SDATA_PROMISC) atomic_dec(&local->iff_promiscs); - dev_mc_unsync(local->mdev, dev); + netif_addr_lock_bh(dev); + spin_lock_bh(&local->filter_lock); + __dev_addr_unsync(&local->mc_list, &local->mc_count, + &dev->mc_list, &dev->mc_count); + ieee80211_configure_filter(local); + spin_unlock_bh(&local->filter_lock); + netif_addr_unlock_bh(dev); + del_timer_sync(&local->dynamic_ps_timer); cancel_work_sync(&local->dynamic_ps_enable_work); @@ -442,9 +441,9 @@ static int ieee80211_stop(struct net_device *dev) if (sdata->u.mntr_flags & MONITOR_FLAG_OTHER_BSS) local->fif_other_bss--; - netif_addr_lock_bh(local->mdev); + spin_lock_bh(&local->filter_lock); ieee80211_configure_filter(local); - netif_addr_unlock_bh(local->mdev); + spin_unlock_bh(&local->filter_lock); break; case NL80211_IFTYPE_STATION: del_timer_sync(&sdata->u.mgd.chswitch_timer); @@ -487,9 +486,9 @@ static int ieee80211_stop(struct net_device *dev) local->fif_other_bss--; atomic_dec(&local->iff_allmultis); - netif_addr_lock_bh(local->mdev); + spin_lock_bh(&local->filter_lock); ieee80211_configure_filter(local); - netif_addr_unlock_bh(local->mdev); + spin_unlock_bh(&local->filter_lock); ieee80211_stop_mesh(sdata); } @@ -535,9 +534,6 @@ static int ieee80211_stop(struct net_device *dev) ieee80211_recalc_ps(local, -1); if (local->open_count == 0) { - if (netif_running(local->mdev)) - dev_close(local->mdev); - drv_stop(local); ieee80211_led_radio(local, false); @@ -584,8 +580,11 @@ static void ieee80211_set_multicast_list(struct net_device *dev) atomic_dec(&local->iff_promiscs); sdata->flags ^= IEEE80211_SDATA_PROMISC; } - - dev_mc_sync(local->mdev, dev); + spin_lock_bh(&local->filter_lock); + __dev_addr_sync(&local->mc_list, &local->mc_count, + &dev->mc_list, &dev->mc_count); + ieee80211_configure_filter(local); + spin_unlock_bh(&local->filter_lock); } /* diff --git a/net/mac80211/main.c b/net/mac80211/main.c index 5b69f5f07299..3234f3751d22 100644 --- a/net/mac80211/main.c +++ b/net/mac80211/main.c @@ -83,75 +83,14 @@ void ieee80211_configure_filter(struct ieee80211_local *local) new_flags |= (1<<31); drv_configure_filter(local, changed_flags, &new_flags, - local->mdev->mc_count, - local->mdev->mc_list); + local->mc_count, + local->mc_list); WARN_ON(new_flags & (1<<31)); local->filter_flags = new_flags & ~(1<<31); } -/* master interface */ - -static int header_parse_80211(const struct sk_buff *skb, unsigned char *haddr) -{ - memcpy(haddr, skb_mac_header(skb) + 10, ETH_ALEN); /* addr2 */ - return ETH_ALEN; -} - -static const struct header_ops ieee80211_header_ops = { - .create = eth_header, - .parse = header_parse_80211, - .rebuild = eth_rebuild_header, - .cache = eth_header_cache, - .cache_update = eth_header_cache_update, -}; - -static int ieee80211_master_open(struct net_device *dev) -{ - struct ieee80211_master_priv *mpriv = netdev_priv(dev); - struct ieee80211_local *local = mpriv->local; - struct ieee80211_sub_if_data *sdata; - int res = -EOPNOTSUPP; - - /* we hold the RTNL here so can safely walk the list */ - list_for_each_entry(sdata, &local->interfaces, list) { - if (netif_running(sdata->dev)) { - res = 0; - break; - } - } - - if (res) - return res; - - netif_tx_start_all_queues(local->mdev); - - return 0; -} - -static int ieee80211_master_stop(struct net_device *dev) -{ - struct ieee80211_master_priv *mpriv = netdev_priv(dev); - struct ieee80211_local *local = mpriv->local; - struct ieee80211_sub_if_data *sdata; - - /* we hold the RTNL here so can safely walk the list */ - list_for_each_entry(sdata, &local->interfaces, list) - if (netif_running(sdata->dev)) - dev_close(sdata->dev); - - return 0; -} - -static void ieee80211_master_set_multicast_list(struct net_device *dev) -{ - struct ieee80211_master_priv *mpriv = netdev_priv(dev); - struct ieee80211_local *local = mpriv->local; - - ieee80211_configure_filter(local); -} - int ieee80211_hw_config(struct ieee80211_local *local, u32 changed) { struct ieee80211_channel *chan, *scan_chan; @@ -310,7 +249,6 @@ void ieee80211_tx_status_irqsafe(struct ieee80211_hw *hw, struct ieee80211_tx_info *info = IEEE80211_SKB_CB(skb); int tmp; - skb->dev = local->mdev; skb->pkt_type = IEEE80211_TX_STATUS_MSG; skb_queue_tail(info->flags & IEEE80211_TX_CTL_REQ_TX_STATUS ? &local->skb_queue : &local->skb_queue_unreliable, skb); @@ -716,7 +654,7 @@ struct ieee80211_hw *ieee80211_alloc_hw(size_t priv_data_len, mutex_init(&local->scan_mtx); spin_lock_init(&local->key_lock); - + spin_lock_init(&local->filter_lock); spin_lock_init(&local->queue_stop_reason_lock); INIT_DELAYED_WORK(&local->scan_work, ieee80211_scan_work); @@ -752,30 +690,11 @@ struct ieee80211_hw *ieee80211_alloc_hw(size_t priv_data_len, } EXPORT_SYMBOL(ieee80211_alloc_hw); -static const struct net_device_ops ieee80211_master_ops = { - .ndo_start_xmit = ieee80211_master_start_xmit, - .ndo_open = ieee80211_master_open, - .ndo_stop = ieee80211_master_stop, - .ndo_set_multicast_list = ieee80211_master_set_multicast_list, - .ndo_select_queue = ieee80211_select_queue, -}; - -static void ieee80211_master_setup(struct net_device *mdev) -{ - mdev->type = ARPHRD_IEEE80211; - mdev->netdev_ops = &ieee80211_master_ops; - mdev->header_ops = &ieee80211_header_ops; - mdev->tx_queue_len = 1000; - mdev->addr_len = ETH_ALEN; -} - int ieee80211_register_hw(struct ieee80211_hw *hw) { struct ieee80211_local *local = hw_to_local(hw); int result; enum ieee80211_band band; - struct net_device *mdev; - struct ieee80211_master_priv *mpriv; int channels, i, j, max_bitrates; bool supp_ht; static const u32 cipher_suites[] = { @@ -874,16 +793,6 @@ int ieee80211_register_hw(struct ieee80211_hw *hw) if (hw->queues > IEEE80211_MAX_QUEUES) hw->queues = IEEE80211_MAX_QUEUES; - mdev = alloc_netdev_mq(sizeof(struct ieee80211_master_priv), - "wmaster%d", ieee80211_master_setup, - hw->queues); - if (!mdev) - goto fail_mdev_alloc; - - mpriv = netdev_priv(mdev); - mpriv->local = local; - local->mdev = mdev; - local->hw.workqueue = create_singlethread_workqueue(wiphy_name(local->hw.wiphy)); if (!local->hw.workqueue) { @@ -918,17 +827,6 @@ int ieee80211_register_hw(struct ieee80211_hw *hw) } rtnl_lock(); - result = dev_alloc_name(local->mdev, local->mdev->name); - if (result < 0) - goto fail_dev; - - memcpy(local->mdev->dev_addr, local->hw.wiphy->perm_addr, ETH_ALEN); - SET_NETDEV_DEV(local->mdev, wiphy_dev(local->hw.wiphy)); - local->mdev->features |= NETIF_F_NETNS_LOCAL; - - result = register_netdevice(local->mdev); - if (result < 0) - goto fail_dev; result = ieee80211_init_rate_ctrl_alg(local, hw->rate_control_algorithm); @@ -981,9 +879,6 @@ int ieee80211_register_hw(struct ieee80211_hw *hw) ieee80211_led_exit(local); ieee80211_remove_interfaces(local); fail_rate: - unregister_netdevice(local->mdev); - local->mdev = NULL; - fail_dev: rtnl_unlock(); ieee80211_wep_free(local); fail_wep: @@ -992,9 +887,6 @@ int ieee80211_register_hw(struct ieee80211_hw *hw) debugfs_hw_del(local); destroy_workqueue(local->hw.workqueue); fail_workqueue: - if (local->mdev) - free_netdev(local->mdev); - fail_mdev_alloc: wiphy_unregister(local->hw.wiphy); fail_wiphy_register: kfree(local->int_scan_req.channels); @@ -1019,13 +911,8 @@ void ieee80211_unregister_hw(struct ieee80211_hw *hw) * because the driver cannot be handing us frames any * more and the tasklet is killed. */ - - /* First, we remove all virtual interfaces. */ ieee80211_remove_interfaces(local); - /* then, finally, remove the master interface */ - unregister_netdevice(local->mdev); - rtnl_unlock(); ieee80211_clear_tx_pending(local); @@ -1044,7 +931,6 @@ void ieee80211_unregister_hw(struct ieee80211_hw *hw) wiphy_unregister(local->hw.wiphy); ieee80211_wep_free(local); ieee80211_led_exit(local); - free_netdev(local->mdev); kfree(local->int_scan_req.channels); } EXPORT_SYMBOL(ieee80211_unregister_hw); diff --git a/net/mac80211/rate.c b/net/mac80211/rate.c index 8ac7a984d886..b33efc4fc267 100644 --- a/net/mac80211/rate.c +++ b/net/mac80211/rate.c @@ -287,7 +287,7 @@ int ieee80211_init_rate_ctrl_alg(struct ieee80211_local *local, struct rate_control_ref *ref, *old; ASSERT_RTNL(); - if (local->open_count || netif_running(local->mdev)) + if (local->open_count) return -EBUSY; ref = rate_control_alloc(name, local); diff --git a/net/mac80211/rx.c b/net/mac80211/rx.c index b513fb791153..7f33f775c5df 100644 --- a/net/mac80211/rx.c +++ b/net/mac80211/rx.c @@ -1478,6 +1478,7 @@ ieee80211_rx_h_mesh_fwding(struct ieee80211_rx_data *rx) struct ieee80211s_hdr *mesh_hdr; unsigned int hdrlen; struct sk_buff *skb = rx->skb, *fwd_skb; + struct ieee80211_local *local = rx->local; hdr = (struct ieee80211_hdr *) skb->data; hdrlen = ieee80211_hdrlen(hdr->frame_control); @@ -1520,6 +1521,8 @@ ieee80211_rx_h_mesh_fwding(struct ieee80211_rx_data *rx) dropped_frames_ttl); else { struct ieee80211_hdr *fwd_hdr; + struct ieee80211_tx_info *info; + fwd_skb = skb_copy(skb, GFP_ATOMIC); if (!fwd_skb && net_ratelimit()) @@ -1533,9 +1536,12 @@ ieee80211_rx_h_mesh_fwding(struct ieee80211_rx_data *rx) */ memcpy(fwd_hdr->addr1, fwd_hdr->addr2, ETH_ALEN); memcpy(fwd_hdr->addr2, rx->dev->dev_addr, ETH_ALEN); - fwd_skb->dev = rx->local->mdev; + info = IEEE80211_SKB_CB(fwd_skb); + memset(info, 0, sizeof(*info)); + info->flags |= IEEE80211_TX_INTFL_NEED_TXPROCESSING; fwd_skb->iif = rx->dev->ifindex; - dev_queue_xmit(fwd_skb); + ieee80211_select_queue(local, fwd_skb); + ieee80211_add_pending_skb(local, fwd_skb); } } @@ -1803,8 +1809,7 @@ ieee80211_rx_h_mgmt(struct ieee80211_rx_data *rx) return RX_DROP_MONITOR; } -static void ieee80211_rx_michael_mic_report(struct net_device *dev, - struct ieee80211_hdr *hdr, +static void ieee80211_rx_michael_mic_report(struct ieee80211_hdr *hdr, struct ieee80211_rx_data *rx) { int keyidx; @@ -2114,7 +2119,7 @@ static void __ieee80211_rx_handle_packet(struct ieee80211_hw *hw, } if ((status->flag & RX_FLAG_MMIC_ERROR)) { - ieee80211_rx_michael_mic_report(local->mdev, hdr, &rx); + ieee80211_rx_michael_mic_report(hdr, &rx); return; } @@ -2483,7 +2488,6 @@ void ieee80211_rx_irqsafe(struct ieee80211_hw *hw, struct sk_buff *skb) BUILD_BUG_ON(sizeof(struct ieee80211_rx_status) > sizeof(skb->cb)); - skb->dev = local->mdev; skb->pkt_type = IEEE80211_RX_MSG; skb_queue_tail(&local->skb_queue, skb); tasklet_schedule(&local->tasklet); diff --git a/net/mac80211/scan.c b/net/mac80211/scan.c index 5f4f7869d050..74820656dc89 100644 --- a/net/mac80211/scan.c +++ b/net/mac80211/scan.c @@ -294,16 +294,13 @@ void ieee80211_scan_completed(struct ieee80211_hw *hw, bool aborted) if (was_hw_scan) goto done; - netif_tx_lock_bh(local->mdev); - netif_addr_lock(local->mdev); + spin_lock_bh(&local->filter_lock); local->filter_flags &= ~FIF_BCN_PRBRESP_PROMISC; drv_configure_filter(local, FIF_BCN_PRBRESP_PROMISC, &local->filter_flags, - local->mdev->mc_count, - local->mdev->mc_list); - - netif_addr_unlock(local->mdev); - netif_tx_unlock_bh(local->mdev); + local->mc_count, + local->mc_list); + spin_unlock_bh(&local->filter_lock); drv_sw_scan_complete(local); @@ -382,13 +379,13 @@ static int ieee80211_start_sw_scan(struct ieee80211_local *local) local->scan_state = SCAN_SET_CHANNEL; local->scan_channel_idx = 0; - netif_addr_lock_bh(local->mdev); + spin_lock_bh(&local->filter_lock); local->filter_flags |= FIF_BCN_PRBRESP_PROMISC; drv_configure_filter(local, FIF_BCN_PRBRESP_PROMISC, &local->filter_flags, - local->mdev->mc_count, - local->mdev->mc_list); - netif_addr_unlock_bh(local->mdev); + local->mc_count, + local->mc_list); + spin_unlock_bh(&local->filter_lock); /* TODO: start scan as soon as all nullfunc frames are ACKed */ queue_delayed_work(local->hw.workqueue, &local->scan_work, diff --git a/net/mac80211/tx.c b/net/mac80211/tx.c index f3efd4f16e91..7adaeb2c53e8 100644 --- a/net/mac80211/tx.c +++ b/net/mac80211/tx.c @@ -451,7 +451,7 @@ ieee80211_tx_h_select_key(struct ieee80211_tx_data *tx) struct ieee80211_tx_info *info = IEEE80211_SKB_CB(tx->skb); struct ieee80211_hdr *hdr = (struct ieee80211_hdr *)tx->skb->data; - if (unlikely(tx->skb->do_not_encrypt)) + if (unlikely(info->flags & IEEE80211_TX_INTFL_DONT_ENCRYPT)) tx->key = NULL; else if (tx->sta && (key = rcu_dereference(tx->sta->key))) tx->key = key; @@ -497,7 +497,7 @@ ieee80211_tx_h_select_key(struct ieee80211_tx_data *tx) } if (!tx->key || !(tx->key->flags & KEY_FLAG_UPLOADED_TO_HARDWARE)) - tx->skb->do_not_encrypt = 1; + info->flags |= IEEE80211_TX_INTFL_DONT_ENCRYPT; return TX_CONTINUE; } @@ -774,9 +774,7 @@ static int ieee80211_fragment(struct ieee80211_local *local, memcpy(tmp->cb, skb->cb, sizeof(tmp->cb)); skb_copy_queue_mapping(tmp, skb); tmp->priority = skb->priority; - tmp->do_not_encrypt = skb->do_not_encrypt; tmp->dev = skb->dev; - tmp->iif = skb->iif; /* copy header and data */ memcpy(skb_put(tmp, hdrlen), skb->data, hdrlen); @@ -804,7 +802,7 @@ ieee80211_tx_h_fragment(struct ieee80211_tx_data *tx) /* * Warn when submitting a fragmented A-MPDU frame and drop it. - * This scenario is handled in __ieee80211_tx_prepare but extra + * This scenario is handled in ieee80211_tx_prepare but extra * caution taken here as fragmented ampdu may cause Tx stop. */ if (WARN_ON(info->flags & IEEE80211_TX_CTL_AMPDU)) @@ -943,11 +941,12 @@ static bool __ieee80211_parse_tx_radiotap(struct ieee80211_tx_data *tx, struct ieee80211_radiotap_header *rthdr = (struct ieee80211_radiotap_header *) skb->data; struct ieee80211_supported_band *sband; + struct ieee80211_tx_info *info = IEEE80211_SKB_CB(skb); int ret = ieee80211_radiotap_iterator_init(&iterator, rthdr, skb->len); sband = tx->local->hw.wiphy->bands[tx->channel->band]; - skb->do_not_encrypt = 1; + info->flags |= IEEE80211_TX_INTFL_DONT_ENCRYPT; tx->flags &= ~IEEE80211_TX_FRAGMENTED; /* @@ -985,7 +984,7 @@ static bool __ieee80211_parse_tx_radiotap(struct ieee80211_tx_data *tx, skb_trim(skb, skb->len - FCS_LEN); } if (*iterator.this_arg & IEEE80211_RADIOTAP_F_WEP) - tx->skb->do_not_encrypt = 0; + info->flags &= ~IEEE80211_TX_INTFL_DONT_ENCRYPT; if (*iterator.this_arg & IEEE80211_RADIOTAP_F_FRAG) tx->flags |= IEEE80211_TX_FRAGMENTED; break; @@ -1018,13 +1017,12 @@ static bool __ieee80211_parse_tx_radiotap(struct ieee80211_tx_data *tx, * initialises @tx */ static ieee80211_tx_result -__ieee80211_tx_prepare(struct ieee80211_tx_data *tx, - struct sk_buff *skb, - struct net_device *dev) +ieee80211_tx_prepare(struct ieee80211_sub_if_data *sdata, + struct ieee80211_tx_data *tx, + struct sk_buff *skb) { - struct ieee80211_local *local = wdev_priv(dev->ieee80211_ptr); + struct ieee80211_local *local = sdata->local; struct ieee80211_hdr *hdr; - struct ieee80211_sub_if_data *sdata; struct ieee80211_tx_info *info = IEEE80211_SKB_CB(skb); int hdrlen, tid; u8 *qc, *state; @@ -1032,9 +1030,9 @@ __ieee80211_tx_prepare(struct ieee80211_tx_data *tx, memset(tx, 0, sizeof(*tx)); tx->skb = skb; - tx->dev = dev; /* use original interface */ + tx->dev = sdata->dev; /* use original interface */ tx->local = local; - tx->sdata = IEEE80211_DEV_TO_SUB_IF(dev); + tx->sdata = sdata; tx->channel = local->hw.conf.channel; /* * Set this flag (used below to indicate "automatic fragmentation"), @@ -1043,7 +1041,6 @@ __ieee80211_tx_prepare(struct ieee80211_tx_data *tx, tx->flags |= IEEE80211_TX_FRAGMENTED; /* process and remove the injection radiotap header */ - sdata = IEEE80211_DEV_TO_SUB_IF(dev); if (unlikely(info->flags & IEEE80211_TX_CTL_INJECTED)) { if (!__ieee80211_parse_tx_radiotap(tx, skb)) return TX_DROP; @@ -1139,50 +1136,28 @@ __ieee80211_tx_prepare(struct ieee80211_tx_data *tx, return TX_CONTINUE; } -/* - * NB: @tx is uninitialised when passed in here - */ -static int ieee80211_tx_prepare(struct ieee80211_local *local, - struct ieee80211_tx_data *tx, - struct sk_buff *skb) -{ - struct net_device *dev; - - dev = dev_get_by_index(&init_net, skb->iif); - if (unlikely(dev && !is_ieee80211_device(local, dev))) { - dev_put(dev); - dev = NULL; - } - if (unlikely(!dev)) - return -ENODEV; - /* - * initialises tx with control - * - * return value is safe to ignore here because this function - * can only be invoked for multicast frames - * - * XXX: clean up - */ - __ieee80211_tx_prepare(tx, skb, dev); - dev_put(dev); - return 0; -} - static int __ieee80211_tx(struct ieee80211_local *local, struct sk_buff **skbp, - struct sta_info *sta) + struct sta_info *sta, + bool txpending) { struct sk_buff *skb = *skbp, *next; struct ieee80211_tx_info *info; + unsigned long flags; int ret, len; bool fragm = false; - local->mdev->trans_start = jiffies; - while (skb) { - if (ieee80211_queue_stopped(&local->hw, - skb_get_queue_mapping(skb))) - return IEEE80211_TX_PENDING; + int q = skb_get_queue_mapping(skb); + + spin_lock_irqsave(&local->queue_stop_reason_lock, flags); + ret = IEEE80211_TX_OK; + if (local->queue_stop_reasons[q] || + (!txpending && !skb_queue_empty(&local->pending[q]))) + ret = IEEE80211_TX_PENDING; + spin_unlock_irqrestore(&local->queue_stop_reason_lock, flags); + if (ret != IEEE80211_TX_OK) + return ret; info = IEEE80211_SKB_CB(skb); @@ -1254,10 +1229,10 @@ static int invoke_tx_handlers(struct ieee80211_tx_data *tx) return 0; } -static void ieee80211_tx(struct net_device *dev, struct sk_buff *skb, - bool txpending) +static void ieee80211_tx(struct ieee80211_sub_if_data *sdata, + struct sk_buff *skb, bool txpending) { - struct ieee80211_local *local = wdev_priv(dev->ieee80211_ptr); + struct ieee80211_local *local = sdata->local; struct ieee80211_tx_data tx; ieee80211_tx_result res_prepare; struct ieee80211_tx_info *info = IEEE80211_SKB_CB(skb); @@ -1268,8 +1243,6 @@ static void ieee80211_tx(struct net_device *dev, struct sk_buff *skb, queue = skb_get_queue_mapping(skb); - WARN_ON(!txpending && !skb_queue_empty(&local->pending[queue])); - if (unlikely(skb->len < 10)) { dev_kfree_skb(skb); return; @@ -1278,7 +1251,7 @@ static void ieee80211_tx(struct net_device *dev, struct sk_buff *skb, rcu_read_lock(); /* initialises tx */ - res_prepare = __ieee80211_tx_prepare(&tx, skb, dev); + res_prepare = ieee80211_tx_prepare(sdata, &tx, skb); if (unlikely(res_prepare == TX_DROP)) { dev_kfree_skb(skb); @@ -1297,7 +1270,7 @@ static void ieee80211_tx(struct net_device *dev, struct sk_buff *skb, retries = 0; retry: - ret = __ieee80211_tx(local, &tx.skb, tx.sta); + ret = __ieee80211_tx(local, &tx.skb, tx.sta, txpending); switch (ret) { case IEEE80211_TX_OK: break; @@ -1315,34 +1288,35 @@ static void ieee80211_tx(struct net_device *dev, struct sk_buff *skb, spin_lock_irqsave(&local->queue_stop_reason_lock, flags); - if (__netif_subqueue_stopped(local->mdev, queue)) { + if (local->queue_stop_reasons[queue] || + !skb_queue_empty(&local->pending[queue])) { + /* + * if queue is stopped, queue up frames for later + * transmission from the tasklet + */ do { next = skb->next; skb->next = NULL; if (unlikely(txpending)) - skb_queue_head(&local->pending[queue], - skb); + __skb_queue_head(&local->pending[queue], + skb); else - skb_queue_tail(&local->pending[queue], - skb); + __skb_queue_tail(&local->pending[queue], + skb); } while ((skb = next)); - /* - * Make sure nobody will enable the queue on us - * (without going through the tasklet) nor disable the - * netdev queue underneath the pending handling code. - */ - __set_bit(IEEE80211_QUEUE_STOP_REASON_PENDING, - &local->queue_stop_reasons[queue]); - spin_unlock_irqrestore(&local->queue_stop_reason_lock, flags); } else { + /* + * otherwise retry, but this is a race condition or + * a driver bug (which we warn about if it persists) + */ spin_unlock_irqrestore(&local->queue_stop_reason_lock, flags); retries++; - if (WARN(retries > 10, "tx refused but queue active")) + if (WARN(retries > 10, "tx refused but queue active\n")) goto drop; goto retry; } @@ -1403,14 +1377,13 @@ static int ieee80211_skb_resize(struct ieee80211_local *local, return 0; } -int ieee80211_master_start_xmit(struct sk_buff *skb, struct net_device *dev) +static void ieee80211_xmit(struct ieee80211_sub_if_data *sdata, + struct sk_buff *skb) { - struct ieee80211_master_priv *mpriv = netdev_priv(dev); - struct ieee80211_local *local = mpriv->local; + struct ieee80211_local *local = sdata->local; struct ieee80211_tx_info *info = IEEE80211_SKB_CB(skb); struct ieee80211_hdr *hdr = (struct ieee80211_hdr *) skb->data; - struct net_device *odev = NULL; - struct ieee80211_sub_if_data *osdata; + struct ieee80211_sub_if_data *tmp_sdata; int headroom; bool may_encrypt; enum { @@ -1419,20 +1392,7 @@ int ieee80211_master_start_xmit(struct sk_buff *skb, struct net_device *dev) UNKNOWN_ADDRESS, } monitor_iface = NOT_MONITOR; - if (skb->iif) - odev = dev_get_by_index(&init_net, skb->iif); - if (unlikely(odev && !is_ieee80211_device(local, odev))) { - dev_put(odev); - odev = NULL; - } - if (unlikely(!odev)) { -#ifdef CONFIG_MAC80211_VERBOSE_DEBUG - printk(KERN_DEBUG "%s: Discarded packet with nonexistent " - "originating device\n", dev->name); -#endif - dev_kfree_skb(skb); - return NETDEV_TX_OK; - } + dev_hold(sdata->dev); if ((local->hw.flags & IEEE80211_HW_PS_NULLFUNC_STACK) && local->hw.conf.dynamic_ps_timeout > 0 && @@ -1448,26 +1408,21 @@ int ieee80211_master_start_xmit(struct sk_buff *skb, struct net_device *dev) msecs_to_jiffies(local->hw.conf.dynamic_ps_timeout)); } - memset(info, 0, sizeof(*info)); - info->flags |= IEEE80211_TX_CTL_REQ_TX_STATUS; - osdata = IEEE80211_DEV_TO_SUB_IF(odev); - - if (ieee80211_vif_is_mesh(&osdata->vif) && + if (ieee80211_vif_is_mesh(&sdata->vif) && ieee80211_is_data(hdr->frame_control)) { if (is_multicast_ether_addr(hdr->addr3)) memcpy(hdr->addr1, hdr->addr3, ETH_ALEN); else - if (mesh_nexthop_lookup(skb, osdata)) { - dev_put(odev); - return NETDEV_TX_OK; + if (mesh_nexthop_lookup(skb, sdata)) { + dev_put(sdata->dev); + return; } - if (memcmp(odev->dev_addr, hdr->addr4, ETH_ALEN) != 0) - IEEE80211_IFSTA_MESH_CTR_INC(&osdata->u.mesh, - fwded_frames); - } else if (unlikely(osdata->vif.type == NL80211_IFTYPE_MONITOR)) { - struct ieee80211_sub_if_data *sdata; + if (memcmp(sdata->dev->dev_addr, hdr->addr4, ETH_ALEN) != 0) + IEEE80211_IFSTA_MESH_CTR_INC(&sdata->u.mesh, + fwded_frames); + } else if (unlikely(sdata->vif.type == NL80211_IFTYPE_MONITOR)) { int hdrlen; u16 len_rthdr; @@ -1491,19 +1446,17 @@ int ieee80211_master_start_xmit(struct sk_buff *skb, struct net_device *dev) */ rcu_read_lock(); - list_for_each_entry_rcu(sdata, &local->interfaces, + list_for_each_entry_rcu(tmp_sdata, &local->interfaces, list) { - if (!netif_running(sdata->dev)) + if (!netif_running(tmp_sdata->dev)) continue; - if (sdata->vif.type != NL80211_IFTYPE_AP) + if (tmp_sdata->vif.type != NL80211_IFTYPE_AP) continue; - if (compare_ether_addr(sdata->dev->dev_addr, + if (compare_ether_addr(tmp_sdata->dev->dev_addr, hdr->addr2)) { - dev_hold(sdata->dev); - dev_put(odev); - osdata = sdata; - odev = osdata->dev; - skb->iif = sdata->dev->ifindex; + dev_hold(tmp_sdata->dev); + dev_put(sdata->dev); + sdata = tmp_sdata; monitor_iface = FOUND_SDATA; break; } @@ -1512,31 +1465,31 @@ int ieee80211_master_start_xmit(struct sk_buff *skb, struct net_device *dev) } } - may_encrypt = !skb->do_not_encrypt; + may_encrypt = !(info->flags & IEEE80211_TX_INTFL_DONT_ENCRYPT); - headroom = osdata->local->tx_headroom; + headroom = local->tx_headroom; if (may_encrypt) headroom += IEEE80211_ENCRYPT_HEADROOM; headroom -= skb_headroom(skb); headroom = max_t(int, 0, headroom); - if (ieee80211_skb_resize(osdata->local, skb, headroom, may_encrypt)) { + if (ieee80211_skb_resize(local, skb, headroom, may_encrypt)) { dev_kfree_skb(skb); - dev_put(odev); - return NETDEV_TX_OK; + dev_put(sdata->dev); + return; } - if (osdata->vif.type == NL80211_IFTYPE_AP_VLAN) - osdata = container_of(osdata->bss, - struct ieee80211_sub_if_data, - u.ap); + tmp_sdata = sdata; + if (sdata->vif.type == NL80211_IFTYPE_AP_VLAN) + tmp_sdata = container_of(sdata->bss, + struct ieee80211_sub_if_data, + u.ap); if (likely(monitor_iface != UNKNOWN_ADDRESS)) - info->control.vif = &osdata->vif; - - ieee80211_tx(odev, skb, false); - dev_put(odev); + info->control.vif = &tmp_sdata->vif; - return NETDEV_TX_OK; + ieee80211_select_queue(local, skb); + ieee80211_tx(sdata, skb, false); + dev_put(sdata->dev); } int ieee80211_monitor_start_xmit(struct sk_buff *skb, @@ -1546,6 +1499,7 @@ int ieee80211_monitor_start_xmit(struct sk_buff *skb, struct ieee80211_channel *chan = local->hw.conf.channel; struct ieee80211_radiotap_header *prthdr = (struct ieee80211_radiotap_header *)skb->data; + struct ieee80211_tx_info *info = IEEE80211_SKB_CB(skb); u16 len_rthdr; /* @@ -1583,15 +1537,9 @@ int ieee80211_monitor_start_xmit(struct sk_buff *skb, if (unlikely(skb->len < len_rthdr)) goto fail; /* skb too short for claimed rt header extent */ - skb->dev = local->mdev; - /* needed because we set skb device to master */ skb->iif = dev->ifindex; - /* sometimes we do encrypt injected frames, will be fixed - * up in radiotap parser if not wanted */ - skb->do_not_encrypt = 0; - /* * fix up the pointers accounting for the radiotap * header still being in there. We are being given @@ -1606,8 +1554,10 @@ int ieee80211_monitor_start_xmit(struct sk_buff *skb, skb_set_network_header(skb, len_rthdr); skb_set_transport_header(skb, len_rthdr); - /* pass the radiotap header up to the next stage intact */ - dev_queue_xmit(skb); + memset(info, 0, sizeof(*info)); + + /* pass the radiotap header up to xmit */ + ieee80211_xmit(IEEE80211_DEV_TO_SUB_IF(dev), skb); return NETDEV_TX_OK; fail: @@ -1635,6 +1585,7 @@ int ieee80211_subif_start_xmit(struct sk_buff *skb, { struct ieee80211_sub_if_data *sdata = IEEE80211_DEV_TO_SUB_IF(dev); struct ieee80211_local *local = sdata->local; + struct ieee80211_tx_info *info = IEEE80211_SKB_CB(skb); int ret = NETDEV_TX_BUSY, head_need; u16 ethertype, hdrlen, meshhdrlen = 0; __le16 fc; @@ -1864,7 +1815,6 @@ int ieee80211_subif_start_xmit(struct sk_buff *skb, skb->iif = dev->ifindex; - skb->dev = local->mdev; dev->stats.tx_packets++; dev->stats.tx_bytes += skb->len; @@ -1875,8 +1825,10 @@ int ieee80211_subif_start_xmit(struct sk_buff *skb, skb_set_network_header(skb, nh_pos); skb_set_transport_header(skb, h_pos); + memset(info, 0, sizeof(*info)); + dev->trans_start = jiffies; - dev_queue_xmit(skb); + ieee80211_xmit(sdata, skb); return NETDEV_TX_OK; @@ -1918,7 +1870,6 @@ static bool ieee80211_tx_pending_skb(struct ieee80211_local *local, return true; } - /* validate info->control.vif against skb->iif */ sdata = IEEE80211_DEV_TO_SUB_IF(dev); if (sdata->vif.type == NL80211_IFTYPE_AP_VLAN) sdata = container_of(sdata->bss, @@ -1932,12 +1883,13 @@ static bool ieee80211_tx_pending_skb(struct ieee80211_local *local, } if (info->flags & IEEE80211_TX_INTFL_NEED_TXPROCESSING) { - ieee80211_tx(dev, skb, true); + /* do not use sdata, it may have been changed above */ + ieee80211_tx(IEEE80211_DEV_TO_SUB_IF(dev), skb, true); } else { hdr = (struct ieee80211_hdr *)skb->data; sta = sta_info_get(local, hdr->addr1); - ret = __ieee80211_tx(local, &skb, sta); + ret = __ieee80211_tx(local, &skb, sta, true); if (ret != IEEE80211_TX_OK) result = false; } @@ -1949,59 +1901,43 @@ static bool ieee80211_tx_pending_skb(struct ieee80211_local *local, } /* - * Transmit all pending packets. Called from tasklet, locks master device - * TX lock so that no new packets can come in. + * Transmit all pending packets. Called from tasklet. */ void ieee80211_tx_pending(unsigned long data) { struct ieee80211_local *local = (struct ieee80211_local *)data; - struct net_device *dev = local->mdev; unsigned long flags; int i; - bool next; + bool txok; rcu_read_lock(); - netif_tx_lock_bh(dev); + spin_lock_irqsave(&local->queue_stop_reason_lock, flags); for (i = 0; i < local->hw.queues; i++) { /* * If queue is stopped by something other than due to pending * frames, or we have no pending frames, proceed to next queue. */ - spin_lock_irqsave(&local->queue_stop_reason_lock, flags); - next = false; - if (local->queue_stop_reasons[i] != - BIT(IEEE80211_QUEUE_STOP_REASON_PENDING) || + if (local->queue_stop_reasons[i] || skb_queue_empty(&local->pending[i])) - next = true; - spin_unlock_irqrestore(&local->queue_stop_reason_lock, flags); - - if (next) continue; - /* - * start the queue now to allow processing our packets, - * we're under the tx lock here anyway so nothing will - * happen as a result of this - */ - netif_start_subqueue(local->mdev, i); - while (!skb_queue_empty(&local->pending[i])) { - struct sk_buff *skb = skb_dequeue(&local->pending[i]); - - if (!ieee80211_tx_pending_skb(local, skb)) { - skb_queue_head(&local->pending[i], skb); + struct sk_buff *skb = __skb_dequeue(&local->pending[i]); + spin_unlock_irqrestore(&local->queue_stop_reason_lock, + flags); + + txok = ieee80211_tx_pending_skb(local, skb); + if (!txok) + __skb_queue_head(&local->pending[i], skb); + spin_lock_irqsave(&local->queue_stop_reason_lock, + flags); + if (!txok) break; - } } - - /* Start regular packet processing again. */ - if (skb_queue_empty(&local->pending[i])) - ieee80211_wake_queue_by_reason(&local->hw, i, - IEEE80211_QUEUE_STOP_REASON_PENDING); } + spin_unlock_irqrestore(&local->queue_stop_reason_lock, flags); - netif_tx_unlock_bh(dev); rcu_read_unlock(); } @@ -2176,8 +2112,7 @@ struct sk_buff *ieee80211_beacon_get(struct ieee80211_hw *hw, info = IEEE80211_SKB_CB(skb); - skb->do_not_encrypt = 1; - + info->flags |= IEEE80211_TX_INTFL_DONT_ENCRYPT; info->band = band; /* * XXX: For now, always use the lowest rate @@ -2248,9 +2183,6 @@ ieee80211_get_buffered_bc(struct ieee80211_hw *hw, sdata = vif_to_sdata(vif); bss = &sdata->u.ap; - if (!bss) - return NULL; - rcu_read_lock(); beacon = rcu_dereference(bss->beacon); @@ -2276,7 +2208,7 @@ ieee80211_get_buffered_bc(struct ieee80211_hw *hw, cpu_to_le16(IEEE80211_FCTL_MOREDATA); } - if (!ieee80211_tx_prepare(local, &tx, skb)) + if (!ieee80211_tx_prepare(sdata, &tx, skb)) break; dev_kfree_skb_any(skb); } @@ -2296,3 +2228,18 @@ ieee80211_get_buffered_bc(struct ieee80211_hw *hw, return skb; } EXPORT_SYMBOL(ieee80211_get_buffered_bc); + +void ieee80211_tx_skb(struct ieee80211_sub_if_data *sdata, struct sk_buff *skb, + int encrypt) +{ + struct ieee80211_tx_info *info = IEEE80211_SKB_CB(skb); + skb_set_mac_header(skb, 0); + skb_set_network_header(skb, 0); + skb_set_transport_header(skb, 0); + + skb->iif = sdata->dev->ifindex; + if (!encrypt) + info->flags |= IEEE80211_TX_INTFL_DONT_ENCRYPT; + + ieee80211_xmit(sdata, skb); +} diff --git a/net/mac80211/util.c b/net/mac80211/util.c index dbf66b52d38c..7fc55846d601 100644 --- a/net/mac80211/util.c +++ b/net/mac80211/util.c @@ -275,16 +275,12 @@ static void __ieee80211_wake_queue(struct ieee80211_hw *hw, int queue, __clear_bit(reason, &local->queue_stop_reasons[queue]); - if (!skb_queue_empty(&local->pending[queue]) && - local->queue_stop_reasons[queue] == - BIT(IEEE80211_QUEUE_STOP_REASON_PENDING)) - tasklet_schedule(&local->tx_pending_tasklet); - if (local->queue_stop_reasons[queue] != 0) /* someone still has this queue stopped */ return; - netif_wake_subqueue(local->mdev, queue); + if (!skb_queue_empty(&local->pending[queue])) + tasklet_schedule(&local->tx_pending_tasklet); } void ieee80211_wake_queue_by_reason(struct ieee80211_hw *hw, int queue, @@ -313,14 +309,6 @@ static void __ieee80211_stop_queue(struct ieee80211_hw *hw, int queue, if (WARN_ON(queue >= hw->queues)) return; - /* - * Only stop if it was previously running, this is necessary - * for correct pending packets handling because there we may - * start (but not wake) the queue and rely on that. - */ - if (!local->queue_stop_reasons[queue]) - netif_stop_subqueue(local->mdev, queue); - __set_bit(reason, &local->queue_stop_reasons[queue]); } @@ -351,8 +339,7 @@ void ieee80211_add_pending_skb(struct ieee80211_local *local, spin_lock_irqsave(&local->queue_stop_reason_lock, flags); __ieee80211_stop_queue(hw, queue, IEEE80211_QUEUE_STOP_REASON_SKB_ADD); - __ieee80211_stop_queue(hw, queue, IEEE80211_QUEUE_STOP_REASON_PENDING); - skb_queue_tail(&local->pending[queue], skb); + __skb_queue_tail(&local->pending[queue], skb); __ieee80211_wake_queue(hw, queue, IEEE80211_QUEUE_STOP_REASON_SKB_ADD); spin_unlock_irqrestore(&local->queue_stop_reason_lock, flags); } @@ -373,16 +360,12 @@ int ieee80211_add_pending_skbs(struct ieee80211_local *local, while ((skb = skb_dequeue(skbs))) { ret++; queue = skb_get_queue_mapping(skb); - skb_queue_tail(&local->pending[queue], skb); + __skb_queue_tail(&local->pending[queue], skb); } - for (i = 0; i < hw->queues; i++) { - if (ret) - __ieee80211_stop_queue(hw, i, - IEEE80211_QUEUE_STOP_REASON_PENDING); + for (i = 0; i < hw->queues; i++) __ieee80211_wake_queue(hw, i, IEEE80211_QUEUE_STOP_REASON_SKB_ADD); - } spin_unlock_irqrestore(&local->queue_stop_reason_lock, flags); return ret; @@ -413,11 +396,16 @@ EXPORT_SYMBOL(ieee80211_stop_queues); int ieee80211_queue_stopped(struct ieee80211_hw *hw, int queue) { struct ieee80211_local *local = hw_to_local(hw); + unsigned long flags; + int ret; if (WARN_ON(queue >= hw->queues)) return true; - return __netif_subqueue_stopped(local->mdev, queue); + spin_lock_irqsave(&local->queue_stop_reason_lock, flags); + ret = !!local->queue_stop_reasons[queue]; + spin_unlock_irqrestore(&local->queue_stop_reason_lock, flags); + return ret; } EXPORT_SYMBOL(ieee80211_queue_stopped); @@ -761,20 +749,6 @@ void ieee80211_sta_def_wmm_params(struct ieee80211_sub_if_data *sdata, ieee80211_set_wmm_default(sdata); } -void ieee80211_tx_skb(struct ieee80211_sub_if_data *sdata, struct sk_buff *skb, - int encrypt) -{ - skb->dev = sdata->local->mdev; - skb_set_mac_header(skb, 0); - skb_set_network_header(skb, 0); - skb_set_transport_header(skb, 0); - - skb->iif = sdata->dev->ifindex; - skb->do_not_encrypt = !encrypt; - - dev_queue_xmit(skb); -} - u32 ieee80211_mandatory_rates(struct ieee80211_local *local, enum ieee80211_band band) { @@ -1049,9 +1023,9 @@ int ieee80211_reconfig(struct ieee80211_local *local) /* reconfigure hardware */ ieee80211_hw_config(local, ~0); - netif_addr_lock_bh(local->mdev); + spin_lock_bh(&local->filter_lock); ieee80211_configure_filter(local); - netif_addr_unlock_bh(local->mdev); + spin_unlock_bh(&local->filter_lock); /* Finally also reconfigure all the BSS information */ list_for_each_entry(sdata, &local->interfaces, list) { diff --git a/net/mac80211/wme.c b/net/mac80211/wme.c index 116a923b14d6..b19b7696f3a2 100644 --- a/net/mac80211/wme.c +++ b/net/mac80211/wme.c @@ -85,10 +85,8 @@ static u16 classify80211(struct ieee80211_local *local, struct sk_buff *skb) return ieee802_1d_to_ac[skb->priority]; } -u16 ieee80211_select_queue(struct net_device *dev, struct sk_buff *skb) +void ieee80211_select_queue(struct ieee80211_local *local, struct sk_buff *skb) { - struct ieee80211_master_priv *mpriv = netdev_priv(dev); - struct ieee80211_local *local = mpriv->local; struct ieee80211_hdr *hdr = (struct ieee80211_hdr *) skb->data; u16 queue; u8 tid; @@ -113,5 +111,5 @@ u16 ieee80211_select_queue(struct net_device *dev, struct sk_buff *skb) *p = 0; } - return queue; + skb_set_queue_mapping(skb, queue); } diff --git a/net/mac80211/wme.h b/net/mac80211/wme.h index 7520d2e014dc..d4fd87ca5118 100644 --- a/net/mac80211/wme.h +++ b/net/mac80211/wme.h @@ -20,6 +20,7 @@ extern const int ieee802_1d_to_ac[8]; -u16 ieee80211_select_queue(struct net_device *dev, struct sk_buff *skb); +void ieee80211_select_queue(struct ieee80211_local *local, + struct sk_buff *skb); #endif /* _WME_H */ -- cgit v1.2.3 From 72bce62775db0315511474e8d8f8e25d25b48366 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Wed, 17 Jun 2009 17:45:28 +0200 Subject: net: remove unused skb->do_not_encrypt mac80211 required this due to the master netdev, but now it can put all information into skb->cb and this can go. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- include/linux/skbuff.h | 6 +----- net/core/skbuff.c | 3 --- 2 files changed, 1 insertion(+), 8 deletions(-) (limited to 'include') diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index f2c69a2cca17..df7b23ac66e6 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -304,7 +304,6 @@ typedef unsigned char *sk_buff_data_t; * @tc_index: Traffic control index * @tc_verd: traffic control verdict * @ndisc_nodetype: router type (from link layer) - * @do_not_encrypt: set to prevent encryption of this frame * @dma_cookie: a cookie to one of several possible DMA operations * done by skb DMA functions * @secmark: security marking @@ -379,13 +378,10 @@ struct sk_buff { kmemcheck_bitfield_begin(flags2); #ifdef CONFIG_IPV6_NDISC_NODETYPE __u8 ndisc_nodetype:2; -#endif -#if defined(CONFIG_MAC80211) || defined(CONFIG_MAC80211_MODULE) - __u8 do_not_encrypt:1; #endif kmemcheck_bitfield_end(flags2); - /* 0/13/14 bit hole */ + /* 0/14 bit hole */ #ifdef CONFIG_NET_DMA dma_cookie_t dma_cookie; diff --git a/net/core/skbuff.c b/net/core/skbuff.c index 9e0597d189b0..80a96166df39 100644 --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@ -559,9 +559,6 @@ static void __copy_skb_header(struct sk_buff *new, const struct sk_buff *old) #endif #endif new->vlan_tci = old->vlan_tci; -#if defined(CONFIG_MAC80211) || defined(CONFIG_MAC80211_MODULE) - new->do_not_encrypt = old->do_not_encrypt; -#endif skb_copy_secmark(new, old); } -- cgit v1.2.3 From 8150f32b90f630ad3e460f026ce338cb81685bc9 Mon Sep 17 00:00:00 2001 From: Dmitry Torokhov Date: Fri, 24 Jul 2009 22:11:32 -0700 Subject: Driver Core: Make PM operations a const pointer They are not supposed to be modified by drivers, so make them const. Signed-off-by: Dmitry Torokhov Acked-by: Greg Kroah-Hartman Signed-off-by: Rafael J. Wysocki --- drivers/base/platform.c | 2 +- drivers/base/power/main.c | 8 +++++--- drivers/pci/pci-driver.c | 16 ++++++++-------- include/linux/device.h | 9 +++++---- 4 files changed, 19 insertions(+), 16 deletions(-) (limited to 'include') diff --git a/drivers/base/platform.c b/drivers/base/platform.c index 81cb01bfc356..c4b3fcee17b2 100644 --- a/drivers/base/platform.c +++ b/drivers/base/platform.c @@ -925,7 +925,7 @@ static int platform_pm_restore_noirq(struct device *dev) #endif /* !CONFIG_HIBERNATION */ -static struct dev_pm_ops platform_dev_pm_ops = { +static const struct dev_pm_ops platform_dev_pm_ops = { .prepare = platform_pm_prepare, .complete = platform_pm_complete, .suspend = platform_pm_suspend, diff --git a/drivers/base/power/main.c b/drivers/base/power/main.c index 58a3e572f2c9..1b1a786b7dec 100644 --- a/drivers/base/power/main.c +++ b/drivers/base/power/main.c @@ -157,8 +157,9 @@ void device_pm_move_last(struct device *dev) * @ops: PM operations to choose from. * @state: PM transition of the system being carried out. */ -static int pm_op(struct device *dev, struct dev_pm_ops *ops, - pm_message_t state) +static int pm_op(struct device *dev, + const struct dev_pm_ops *ops, + pm_message_t state) { int error = 0; @@ -220,7 +221,8 @@ static int pm_op(struct device *dev, struct dev_pm_ops *ops, * The operation is executed with interrupts disabled by the only remaining * functional CPU in the system. */ -static int pm_noirq_op(struct device *dev, struct dev_pm_ops *ops, +static int pm_noirq_op(struct device *dev, + const struct dev_pm_ops *ops, pm_message_t state) { int error = 0; diff --git a/drivers/pci/pci-driver.c b/drivers/pci/pci-driver.c index d76c4c85367e..0c2ea44ae5e1 100644 --- a/drivers/pci/pci-driver.c +++ b/drivers/pci/pci-driver.c @@ -575,7 +575,7 @@ static void pci_pm_complete(struct device *dev) static int pci_pm_suspend(struct device *dev) { struct pci_dev *pci_dev = to_pci_dev(dev); - struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL; + const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL; if (pci_has_legacy_pm_support(pci_dev)) return pci_legacy_suspend(dev, PMSG_SUSPEND); @@ -613,7 +613,7 @@ static int pci_pm_suspend(struct device *dev) static int pci_pm_suspend_noirq(struct device *dev) { struct pci_dev *pci_dev = to_pci_dev(dev); - struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL; + const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL; if (pci_has_legacy_pm_support(pci_dev)) return pci_legacy_suspend_late(dev, PMSG_SUSPEND); @@ -672,7 +672,7 @@ static int pci_pm_resume_noirq(struct device *dev) static int pci_pm_resume(struct device *dev) { struct pci_dev *pci_dev = to_pci_dev(dev); - struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL; + const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL; int error = 0; /* @@ -711,7 +711,7 @@ static int pci_pm_resume(struct device *dev) static int pci_pm_freeze(struct device *dev) { struct pci_dev *pci_dev = to_pci_dev(dev); - struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL; + const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL; if (pci_has_legacy_pm_support(pci_dev)) return pci_legacy_suspend(dev, PMSG_FREEZE); @@ -780,7 +780,7 @@ static int pci_pm_thaw_noirq(struct device *dev) static int pci_pm_thaw(struct device *dev) { struct pci_dev *pci_dev = to_pci_dev(dev); - struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL; + const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL; int error = 0; if (pci_has_legacy_pm_support(pci_dev)) @@ -799,7 +799,7 @@ static int pci_pm_thaw(struct device *dev) static int pci_pm_poweroff(struct device *dev) { struct pci_dev *pci_dev = to_pci_dev(dev); - struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL; + const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL; if (pci_has_legacy_pm_support(pci_dev)) return pci_legacy_suspend(dev, PMSG_HIBERNATE); @@ -872,7 +872,7 @@ static int pci_pm_restore_noirq(struct device *dev) static int pci_pm_restore(struct device *dev) { struct pci_dev *pci_dev = to_pci_dev(dev); - struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL; + const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL; int error = 0; /* @@ -910,7 +910,7 @@ static int pci_pm_restore(struct device *dev) #endif /* !CONFIG_HIBERNATION */ -struct dev_pm_ops pci_dev_pm_ops = { +const struct dev_pm_ops pci_dev_pm_ops = { .prepare = pci_pm_prepare, .complete = pci_pm_complete, .suspend = pci_pm_suspend, diff --git a/include/linux/device.h b/include/linux/device.h index aebb81036db2..a28642975053 100644 --- a/include/linux/device.h +++ b/include/linux/device.h @@ -62,7 +62,7 @@ struct bus_type { int (*suspend)(struct device *dev, pm_message_t state); int (*resume)(struct device *dev); - struct dev_pm_ops *pm; + const struct dev_pm_ops *pm; struct bus_type_private *p; }; @@ -132,7 +132,7 @@ struct device_driver { int (*resume) (struct device *dev); struct attribute_group **groups; - struct dev_pm_ops *pm; + const struct dev_pm_ops *pm; struct driver_private *p; }; @@ -200,7 +200,8 @@ struct class { int (*suspend)(struct device *dev, pm_message_t state); int (*resume)(struct device *dev); - struct dev_pm_ops *pm; + const struct dev_pm_ops *pm; + struct class_private *p; }; @@ -291,7 +292,7 @@ struct device_type { char *(*nodename)(struct device *dev); void (*release)(struct device *dev); - struct dev_pm_ops *pm; + const struct dev_pm_ops *pm; }; /* interface for exporting device attributes */ -- cgit v1.2.3 From cb3824bade2549d7ad059d5802da43312540fdee Mon Sep 17 00:00:00 2001 From: Karsten Keil Date: Wed, 8 Jul 2009 14:21:12 +0200 Subject: ISDN: Make isdnhdlc usable for other ISDN drivers isdnhdlc is useful for other ISDN drivers as well. Move the include file to a central location and the source to the central isdn location. Signed-off-by: Karsten Keil --- drivers/isdn/Kconfig | 6 +- drivers/isdn/hisax/Kconfig | 6 +- drivers/isdn/hisax/Makefile | 4 - drivers/isdn/hisax/isdnhdlc.c | 603 ------------------------------------------ drivers/isdn/hisax/isdnhdlc.h | 70 ----- drivers/isdn/hisax/st5481.h | 2 +- drivers/isdn/i4l/Kconfig | 11 + drivers/isdn/i4l/Makefile | 1 + drivers/isdn/i4l/isdnhdlc.c | 603 ++++++++++++++++++++++++++++++++++++++++++ include/linux/isdn/hdlc.h | 70 +++++ 10 files changed, 689 insertions(+), 687 deletions(-) delete mode 100644 drivers/isdn/hisax/isdnhdlc.c delete mode 100644 drivers/isdn/hisax/isdnhdlc.h create mode 100644 drivers/isdn/i4l/isdnhdlc.c create mode 100644 include/linux/isdn/hdlc.h (limited to 'include') diff --git a/drivers/isdn/Kconfig b/drivers/isdn/Kconfig index 02bdca6f95c3..022a19452953 100644 --- a/drivers/isdn/Kconfig +++ b/drivers/isdn/Kconfig @@ -21,8 +21,6 @@ menuconfig ISDN if ISDN -source "drivers/isdn/mISDN/Kconfig" - menuconfig ISDN_I4L tristate "Old ISDN4Linux (deprecated)" ---help--- @@ -41,9 +39,9 @@ menuconfig ISDN_I4L It is still available, though, for use with adapters that are not supported by the new CAPI subsystem yet. -if ISDN_I4L +source "drivers/isdn/mISDN/Kconfig" + source "drivers/isdn/i4l/Kconfig" -endif menuconfig ISDN_CAPI tristate "CAPI 2.0 subsystem" diff --git a/drivers/isdn/hisax/Kconfig b/drivers/isdn/hisax/Kconfig index 7832d8ba8e44..3464ebc4cdbc 100644 --- a/drivers/isdn/hisax/Kconfig +++ b/drivers/isdn/hisax/Kconfig @@ -391,6 +391,7 @@ comment "HiSax sub driver modules" config HISAX_ST5481 tristate "ST5481 USB ISDN modem (EXPERIMENTAL)" depends on USB && EXPERIMENTAL + select ISDN_HDLC select CRC_CCITT select BITREVERSE help @@ -418,11 +419,6 @@ config HISAX_FRITZ_PCIPNP (the latter also needs you to select "ISA Plug and Play support" from the menu "Plug and Play configuration") -config HISAX_HDLC - bool - depends on HISAX_ST5481 - default y - config HISAX_AVM_A1_PCMCIA bool depends on HISAX_AVM_A1_CS diff --git a/drivers/isdn/hisax/Makefile b/drivers/isdn/hisax/Makefile index c7a3794bdae4..ab638b083df9 100644 --- a/drivers/isdn/hisax/Makefile +++ b/drivers/isdn/hisax/Makefile @@ -16,10 +16,6 @@ obj-$(CONFIG_HISAX_HFCUSB) += hfc_usb.o obj-$(CONFIG_HISAX_HFC4S8S) += hfc4s8s_l1.o obj-$(CONFIG_HISAX_FRITZ_PCIPNP) += hisax_isac.o hisax_fcpcipnp.o -ifdef CONFIG_HISAX_HDLC -obj-$(CONFIG_ISDN_DRV_HISAX) += isdnhdlc.o -endif - # Multipart objects. hisax_st5481-y := st5481_init.o st5481_usb.o st5481_d.o \ diff --git a/drivers/isdn/hisax/isdnhdlc.c b/drivers/isdn/hisax/isdnhdlc.c deleted file mode 100644 index c69a77a80062..000000000000 --- a/drivers/isdn/hisax/isdnhdlc.c +++ /dev/null @@ -1,603 +0,0 @@ -/* - * isdnhdlc.c -- General purpose ISDN HDLC decoder. - * - *Copyright (C) 2002 Wolfgang Mües - * 2001 Frode Isaksen - * 2001 Kai Germaschewski - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. - */ - -#include -#include -#include -#include "isdnhdlc.h" - -/*-------------------------------------------------------------------*/ - -MODULE_AUTHOR("Wolfgang Mües , " - "Frode Isaksen , " - "Kai Germaschewski "); -MODULE_DESCRIPTION("General purpose ISDN HDLC decoder"); -MODULE_LICENSE("GPL"); - -/*-------------------------------------------------------------------*/ - -enum { - HDLC_FAST_IDLE,HDLC_GET_FLAG_B0,HDLC_GETFLAG_B1A6,HDLC_GETFLAG_B7, - HDLC_GET_DATA,HDLC_FAST_FLAG -}; - -enum { - HDLC_SEND_DATA,HDLC_SEND_CRC1,HDLC_SEND_FAST_FLAG, - HDLC_SEND_FIRST_FLAG,HDLC_SEND_CRC2,HDLC_SEND_CLOSING_FLAG, - HDLC_SEND_IDLE1,HDLC_SEND_FAST_IDLE,HDLC_SENDFLAG_B0, - HDLC_SENDFLAG_B1A6,HDLC_SENDFLAG_B7,STOPPED -}; - -void isdnhdlc_rcv_init (struct isdnhdlc_vars *hdlc, int do_adapt56) -{ - hdlc->bit_shift = 0; - hdlc->hdlc_bits1 = 0; - hdlc->data_bits = 0; - hdlc->ffbit_shift = 0; - hdlc->data_received = 0; - hdlc->state = HDLC_GET_DATA; - hdlc->do_adapt56 = do_adapt56; - hdlc->dchannel = 0; - hdlc->crc = 0; - hdlc->cbin = 0; - hdlc->shift_reg = 0; - hdlc->ffvalue = 0; - hdlc->dstpos = 0; -} - -void isdnhdlc_out_init (struct isdnhdlc_vars *hdlc, int is_d_channel, int do_adapt56) -{ - hdlc->bit_shift = 0; - hdlc->hdlc_bits1 = 0; - hdlc->data_bits = 0; - hdlc->ffbit_shift = 0; - hdlc->data_received = 0; - hdlc->do_closing = 0; - hdlc->ffvalue = 0; - if (is_d_channel) { - hdlc->dchannel = 1; - hdlc->state = HDLC_SEND_FIRST_FLAG; - } else { - hdlc->dchannel = 0; - hdlc->state = HDLC_SEND_FAST_FLAG; - hdlc->ffvalue = 0x7e; - } - hdlc->cbin = 0x7e; - hdlc->bit_shift = 0; - if(do_adapt56){ - hdlc->do_adapt56 = 1; - hdlc->data_bits = 0; - hdlc->state = HDLC_SENDFLAG_B0; - } else { - hdlc->do_adapt56 = 0; - hdlc->data_bits = 8; - } - hdlc->shift_reg = 0; -} - -/* - isdnhdlc_decode - decodes HDLC frames from a transparent bit stream. - - The source buffer is scanned for valid HDLC frames looking for - flags (01111110) to indicate the start of a frame. If the start of - the frame is found, the bit stuffing is removed (0 after 5 1's). - When a new flag is found, the complete frame has been received - and the CRC is checked. - If a valid frame is found, the function returns the frame length - excluding the CRC with the bit HDLC_END_OF_FRAME set. - If the beginning of a valid frame is found, the function returns - the length. - If a framing error is found (too many 1s and not a flag) the function - returns the length with the bit HDLC_FRAMING_ERROR set. - If a CRC error is found the function returns the length with the - bit HDLC_CRC_ERROR set. - If the frame length exceeds the destination buffer size, the function - returns the length with the bit HDLC_LENGTH_ERROR set. - - src - source buffer - slen - source buffer length - count - number of bytes removed (decoded) from the source buffer - dst _ destination buffer - dsize - destination buffer size - returns - number of decoded bytes in the destination buffer and status - flag. - */ -int isdnhdlc_decode (struct isdnhdlc_vars *hdlc, const unsigned char *src, - int slen, int *count, unsigned char *dst, int dsize) -{ - int status=0; - - static const unsigned char fast_flag[]={ - 0x00,0x00,0x00,0x20,0x30,0x38,0x3c,0x3e,0x3f - }; - - static const unsigned char fast_flag_value[]={ - 0x00,0x7e,0xfc,0xf9,0xf3,0xe7,0xcf,0x9f,0x3f - }; - - static const unsigned char fast_abort[]={ - 0x00,0x00,0x80,0xc0,0xe0,0xf0,0xf8,0xfc,0xfe,0xff - }; - - *count = slen; - - while(slen > 0){ - if(hdlc->bit_shift==0){ - hdlc->cbin = *src++; - slen--; - hdlc->bit_shift = 8; - if(hdlc->do_adapt56){ - hdlc->bit_shift --; - } - } - - switch(hdlc->state){ - case STOPPED: - return 0; - case HDLC_FAST_IDLE: - if(hdlc->cbin == 0xff){ - hdlc->bit_shift = 0; - break; - } - hdlc->state = HDLC_GET_FLAG_B0; - hdlc->hdlc_bits1 = 0; - hdlc->bit_shift = 8; - break; - case HDLC_GET_FLAG_B0: - if(!(hdlc->cbin & 0x80)) { - hdlc->state = HDLC_GETFLAG_B1A6; - hdlc->hdlc_bits1 = 0; - } else { - if(!hdlc->do_adapt56){ - if(++hdlc->hdlc_bits1 >=8 ) if(hdlc->bit_shift==1) - hdlc->state = HDLC_FAST_IDLE; - } - } - hdlc->cbin<<=1; - hdlc->bit_shift --; - break; - case HDLC_GETFLAG_B1A6: - if(hdlc->cbin & 0x80){ - hdlc->hdlc_bits1++; - if(hdlc->hdlc_bits1==6){ - hdlc->state = HDLC_GETFLAG_B7; - } - } else { - hdlc->hdlc_bits1 = 0; - } - hdlc->cbin<<=1; - hdlc->bit_shift --; - break; - case HDLC_GETFLAG_B7: - if(hdlc->cbin & 0x80) { - hdlc->state = HDLC_GET_FLAG_B0; - } else { - hdlc->state = HDLC_GET_DATA; - hdlc->crc = 0xffff; - hdlc->shift_reg = 0; - hdlc->hdlc_bits1 = 0; - hdlc->data_bits = 0; - hdlc->data_received = 0; - } - hdlc->cbin<<=1; - hdlc->bit_shift --; - break; - case HDLC_GET_DATA: - if(hdlc->cbin & 0x80){ - hdlc->hdlc_bits1++; - switch(hdlc->hdlc_bits1){ - case 6: - break; - case 7: - if(hdlc->data_received) { - // bad frame - status = -HDLC_FRAMING_ERROR; - } - if(!hdlc->do_adapt56){ - if(hdlc->cbin==fast_abort[hdlc->bit_shift+1]){ - hdlc->state = HDLC_FAST_IDLE; - hdlc->bit_shift=1; - break; - } - } else { - hdlc->state = HDLC_GET_FLAG_B0; - } - break; - default: - hdlc->shift_reg>>=1; - hdlc->shift_reg |= 0x80; - hdlc->data_bits++; - break; - } - } else { - switch(hdlc->hdlc_bits1){ - case 5: - break; - case 6: - if(hdlc->data_received){ - if (hdlc->dstpos < 2) { - status = -HDLC_FRAMING_ERROR; - } else if (hdlc->crc != 0xf0b8){ - // crc error - status = -HDLC_CRC_ERROR; - } else { - // remove CRC - hdlc->dstpos -= 2; - // good frame - status = hdlc->dstpos; - } - } - hdlc->crc = 0xffff; - hdlc->shift_reg = 0; - hdlc->data_bits = 0; - if(!hdlc->do_adapt56){ - if(hdlc->cbin==fast_flag[hdlc->bit_shift]){ - hdlc->ffvalue = fast_flag_value[hdlc->bit_shift]; - hdlc->state = HDLC_FAST_FLAG; - hdlc->ffbit_shift = hdlc->bit_shift; - hdlc->bit_shift = 1; - } else { - hdlc->state = HDLC_GET_DATA; - hdlc->data_received = 0; - } - } else { - hdlc->state = HDLC_GET_DATA; - hdlc->data_received = 0; - } - break; - default: - hdlc->shift_reg>>=1; - hdlc->data_bits++; - break; - } - hdlc->hdlc_bits1 = 0; - } - if (status) { - hdlc->dstpos = 0; - *count -= slen; - hdlc->cbin <<= 1; - hdlc->bit_shift--; - return status; - } - if(hdlc->data_bits==8){ - hdlc->data_bits = 0; - hdlc->data_received = 1; - hdlc->crc = crc_ccitt_byte(hdlc->crc, hdlc->shift_reg); - - // good byte received - if (hdlc->dstpos < dsize) { - dst[hdlc->dstpos++] = hdlc->shift_reg; - } else { - // frame too long - status = -HDLC_LENGTH_ERROR; - hdlc->dstpos = 0; - } - } - hdlc->cbin <<= 1; - hdlc->bit_shift--; - break; - case HDLC_FAST_FLAG: - if(hdlc->cbin==hdlc->ffvalue){ - hdlc->bit_shift = 0; - break; - } else { - if(hdlc->cbin == 0xff){ - hdlc->state = HDLC_FAST_IDLE; - hdlc->bit_shift=0; - } else if(hdlc->ffbit_shift==8){ - hdlc->state = HDLC_GETFLAG_B7; - break; - } else { - hdlc->shift_reg = fast_abort[hdlc->ffbit_shift-1]; - hdlc->hdlc_bits1 = hdlc->ffbit_shift-2; - if(hdlc->hdlc_bits1<0)hdlc->hdlc_bits1 = 0; - hdlc->data_bits = hdlc->ffbit_shift-1; - hdlc->state = HDLC_GET_DATA; - hdlc->data_received = 0; - } - } - break; - default: - break; - } - } - *count -= slen; - return 0; -} - -/* - isdnhdlc_encode - encodes HDLC frames to a transparent bit stream. - - The bit stream starts with a beginning flag (01111110). After - that each byte is added to the bit stream with bit stuffing added - (0 after 5 1's). - When the last byte has been removed from the source buffer, the - CRC (2 bytes is added) and the frame terminates with the ending flag. - For the dchannel, the idle character (all 1's) is also added at the end. - If this function is called with empty source buffer (slen=0), flags or - idle character will be generated. - - src - source buffer - slen - source buffer length - count - number of bytes removed (encoded) from source buffer - dst _ destination buffer - dsize - destination buffer size - returns - number of encoded bytes in the destination buffer -*/ -int isdnhdlc_encode(struct isdnhdlc_vars *hdlc, const unsigned char *src, - unsigned short slen, int *count, - unsigned char *dst, int dsize) -{ - static const unsigned char xfast_flag_value[] = { - 0x7e,0x3f,0x9f,0xcf,0xe7,0xf3,0xf9,0xfc,0x7e - }; - - int len = 0; - - *count = slen; - - while (dsize > 0) { - if(hdlc->bit_shift==0){ - if(slen && !hdlc->do_closing){ - hdlc->shift_reg = *src++; - slen--; - if (slen == 0) - hdlc->do_closing = 1; /* closing sequence, CRC + flag(s) */ - hdlc->bit_shift = 8; - } else { - if(hdlc->state == HDLC_SEND_DATA){ - if(hdlc->data_received){ - hdlc->state = HDLC_SEND_CRC1; - hdlc->crc ^= 0xffff; - hdlc->bit_shift = 8; - hdlc->shift_reg = hdlc->crc & 0xff; - } else if(!hdlc->do_adapt56){ - hdlc->state = HDLC_SEND_FAST_FLAG; - } else { - hdlc->state = HDLC_SENDFLAG_B0; - } - } - - } - } - - switch(hdlc->state){ - case STOPPED: - while (dsize--) - *dst++ = 0xff; - - return dsize; - case HDLC_SEND_FAST_FLAG: - hdlc->do_closing = 0; - if(slen == 0){ - *dst++ = hdlc->ffvalue; - len++; - dsize--; - break; - } - if(hdlc->bit_shift==8){ - hdlc->cbin = hdlc->ffvalue>>(8-hdlc->data_bits); - hdlc->state = HDLC_SEND_DATA; - hdlc->crc = 0xffff; - hdlc->hdlc_bits1 = 0; - hdlc->data_received = 1; - } - break; - case HDLC_SENDFLAG_B0: - hdlc->do_closing = 0; - hdlc->cbin <<= 1; - hdlc->data_bits++; - hdlc->hdlc_bits1 = 0; - hdlc->state = HDLC_SENDFLAG_B1A6; - break; - case HDLC_SENDFLAG_B1A6: - hdlc->cbin <<= 1; - hdlc->data_bits++; - hdlc->cbin++; - if(++hdlc->hdlc_bits1 == 6) - hdlc->state = HDLC_SENDFLAG_B7; - break; - case HDLC_SENDFLAG_B7: - hdlc->cbin <<= 1; - hdlc->data_bits++; - if(slen == 0){ - hdlc->state = HDLC_SENDFLAG_B0; - break; - } - if(hdlc->bit_shift==8){ - hdlc->state = HDLC_SEND_DATA; - hdlc->crc = 0xffff; - hdlc->hdlc_bits1 = 0; - hdlc->data_received = 1; - } - break; - case HDLC_SEND_FIRST_FLAG: - hdlc->data_received = 1; - if(hdlc->data_bits==8){ - hdlc->state = HDLC_SEND_DATA; - hdlc->crc = 0xffff; - hdlc->hdlc_bits1 = 0; - break; - } - hdlc->cbin <<= 1; - hdlc->data_bits++; - if(hdlc->shift_reg & 0x01) - hdlc->cbin++; - hdlc->shift_reg >>= 1; - hdlc->bit_shift--; - if(hdlc->bit_shift==0){ - hdlc->state = HDLC_SEND_DATA; - hdlc->crc = 0xffff; - hdlc->hdlc_bits1 = 0; - } - break; - case HDLC_SEND_DATA: - hdlc->cbin <<= 1; - hdlc->data_bits++; - if(hdlc->hdlc_bits1 == 5){ - hdlc->hdlc_bits1 = 0; - break; - } - if(hdlc->bit_shift==8){ - hdlc->crc = crc_ccitt_byte(hdlc->crc, hdlc->shift_reg); - } - if(hdlc->shift_reg & 0x01){ - hdlc->hdlc_bits1++; - hdlc->cbin++; - hdlc->shift_reg >>= 1; - hdlc->bit_shift--; - } else { - hdlc->hdlc_bits1 = 0; - hdlc->shift_reg >>= 1; - hdlc->bit_shift--; - } - break; - case HDLC_SEND_CRC1: - hdlc->cbin <<= 1; - hdlc->data_bits++; - if(hdlc->hdlc_bits1 == 5){ - hdlc->hdlc_bits1 = 0; - break; - } - if(hdlc->shift_reg & 0x01){ - hdlc->hdlc_bits1++; - hdlc->cbin++; - hdlc->shift_reg >>= 1; - hdlc->bit_shift--; - } else { - hdlc->hdlc_bits1 = 0; - hdlc->shift_reg >>= 1; - hdlc->bit_shift--; - } - if(hdlc->bit_shift==0){ - hdlc->shift_reg = (hdlc->crc >> 8); - hdlc->state = HDLC_SEND_CRC2; - hdlc->bit_shift = 8; - } - break; - case HDLC_SEND_CRC2: - hdlc->cbin <<= 1; - hdlc->data_bits++; - if(hdlc->hdlc_bits1 == 5){ - hdlc->hdlc_bits1 = 0; - break; - } - if(hdlc->shift_reg & 0x01){ - hdlc->hdlc_bits1++; - hdlc->cbin++; - hdlc->shift_reg >>= 1; - hdlc->bit_shift--; - } else { - hdlc->hdlc_bits1 = 0; - hdlc->shift_reg >>= 1; - hdlc->bit_shift--; - } - if(hdlc->bit_shift==0){ - hdlc->shift_reg = 0x7e; - hdlc->state = HDLC_SEND_CLOSING_FLAG; - hdlc->bit_shift = 8; - } - break; - case HDLC_SEND_CLOSING_FLAG: - hdlc->cbin <<= 1; - hdlc->data_bits++; - if(hdlc->hdlc_bits1 == 5){ - hdlc->hdlc_bits1 = 0; - break; - } - if(hdlc->shift_reg & 0x01){ - hdlc->cbin++; - } - hdlc->shift_reg >>= 1; - hdlc->bit_shift--; - if(hdlc->bit_shift==0){ - hdlc->ffvalue = xfast_flag_value[hdlc->data_bits]; - if(hdlc->dchannel){ - hdlc->ffvalue = 0x7e; - hdlc->state = HDLC_SEND_IDLE1; - hdlc->bit_shift = 8-hdlc->data_bits; - if(hdlc->bit_shift==0) - hdlc->state = HDLC_SEND_FAST_IDLE; - } else { - if(!hdlc->do_adapt56){ - hdlc->state = HDLC_SEND_FAST_FLAG; - hdlc->data_received = 0; - } else { - hdlc->state = HDLC_SENDFLAG_B0; - hdlc->data_received = 0; - } - // Finished with this frame, send flags - if (dsize > 1) dsize = 1; - } - } - break; - case HDLC_SEND_IDLE1: - hdlc->do_closing = 0; - hdlc->cbin <<= 1; - hdlc->cbin++; - hdlc->data_bits++; - hdlc->bit_shift--; - if(hdlc->bit_shift==0){ - hdlc->state = HDLC_SEND_FAST_IDLE; - hdlc->bit_shift = 0; - } - break; - case HDLC_SEND_FAST_IDLE: - hdlc->do_closing = 0; - hdlc->cbin = 0xff; - hdlc->data_bits = 8; - if(hdlc->bit_shift == 8){ - hdlc->cbin = 0x7e; - hdlc->state = HDLC_SEND_FIRST_FLAG; - } else { - *dst++ = hdlc->cbin; - hdlc->bit_shift = hdlc->data_bits = 0; - len++; - dsize = 0; - } - break; - default: - break; - } - if(hdlc->do_adapt56){ - if(hdlc->data_bits==7){ - hdlc->cbin <<= 1; - hdlc->cbin++; - hdlc->data_bits++; - } - } - if(hdlc->data_bits==8){ - *dst++ = hdlc->cbin; - hdlc->data_bits = 0; - len++; - dsize--; - } - } - *count -= slen; - - return len; -} - -EXPORT_SYMBOL(isdnhdlc_rcv_init); -EXPORT_SYMBOL(isdnhdlc_decode); -EXPORT_SYMBOL(isdnhdlc_out_init); -EXPORT_SYMBOL(isdnhdlc_encode); diff --git a/drivers/isdn/hisax/isdnhdlc.h b/drivers/isdn/hisax/isdnhdlc.h deleted file mode 100644 index cf0a95a24015..000000000000 --- a/drivers/isdn/hisax/isdnhdlc.h +++ /dev/null @@ -1,70 +0,0 @@ -/* - * isdnhdlc.h -- General purpose ISDN HDLC decoder. - * - * Implementation of a HDLC decoder/encoder in software. - * Neccessary because some ISDN devices don't have HDLC - * controllers. Also included: a bit reversal table. - * - *Copyright (C) 2002 Wolfgang Mües - * 2001 Frode Isaksen - * 2001 Kai Germaschewski - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. - */ - -#ifndef __ISDNHDLC_H__ -#define __ISDNHDLC_H__ - -struct isdnhdlc_vars { - int bit_shift; - int hdlc_bits1; - int data_bits; - int ffbit_shift; // encoding only - int state; - int dstpos; - - unsigned short crc; - - unsigned char cbin; - unsigned char shift_reg; - unsigned char ffvalue; - - unsigned int data_received:1; // set if transferring data - unsigned int dchannel:1; // set if D channel (send idle instead of flags) - unsigned int do_adapt56:1; // set if 56K adaptation - unsigned int do_closing:1; // set if in closing phase (need to send CRC + flag -}; - - -/* - The return value from isdnhdlc_decode is - the frame length, 0 if no complete frame was decoded, - or a negative error number -*/ -#define HDLC_FRAMING_ERROR 1 -#define HDLC_CRC_ERROR 2 -#define HDLC_LENGTH_ERROR 3 - -extern void isdnhdlc_rcv_init (struct isdnhdlc_vars *hdlc, int do_adapt56); - -extern int isdnhdlc_decode (struct isdnhdlc_vars *hdlc, const unsigned char *src, int slen,int *count, - unsigned char *dst, int dsize); - -extern void isdnhdlc_out_init (struct isdnhdlc_vars *hdlc,int is_d_channel,int do_adapt56); - -extern int isdnhdlc_encode (struct isdnhdlc_vars *hdlc,const unsigned char *src,unsigned short slen,int *count, - unsigned char *dst,int dsize); - -#endif /* __ISDNHDLC_H__ */ diff --git a/drivers/isdn/hisax/st5481.h b/drivers/isdn/hisax/st5481.h index cff7a6354334..64f78a8c28c5 100644 --- a/drivers/isdn/hisax/st5481.h +++ b/drivers/isdn/hisax/st5481.h @@ -226,7 +226,7 @@ printk(KERN_WARNING "%s:%s: " format "\n" , __FILE__, __func__ , ## arg) #define INFO(format, arg...) \ printk(KERN_INFO "%s:%s: " format "\n" , __FILE__, __func__ , ## arg) -#include "isdnhdlc.h" +#include #include "fsm.h" #include "hisax_if.h" #include diff --git a/drivers/isdn/i4l/Kconfig b/drivers/isdn/i4l/Kconfig index ed3510f273d8..dd744ffd240b 100644 --- a/drivers/isdn/i4l/Kconfig +++ b/drivers/isdn/i4l/Kconfig @@ -2,6 +2,8 @@ # Old ISDN4Linux config # +if ISDN_I4L + config ISDN_PPP bool "Support synchronous PPP" depends on INET @@ -135,3 +137,12 @@ source "drivers/isdn/act2000/Kconfig" source "drivers/isdn/hysdn/Kconfig" endmenu +# end ISDN_I4L +endif + +config ISDN_HDLC + tristate + depends on HISAX_ST5481 + select CRC_CCITT + select BITREVERSE + diff --git a/drivers/isdn/i4l/Makefile b/drivers/isdn/i4l/Makefile index 49a06c0005dd..cb9d3bb9fae0 100644 --- a/drivers/isdn/i4l/Makefile +++ b/drivers/isdn/i4l/Makefile @@ -4,6 +4,7 @@ obj-$(CONFIG_ISDN_I4L) += isdn.o obj-$(CONFIG_ISDN_PPP_BSDCOMP) += isdn_bsdcomp.o +obj-$(CONFIG_ISDN_HDLC) += isdnhdlc.o # Multipart objects. diff --git a/drivers/isdn/i4l/isdnhdlc.c b/drivers/isdn/i4l/isdnhdlc.c new file mode 100644 index 000000000000..44ec7418496b --- /dev/null +++ b/drivers/isdn/i4l/isdnhdlc.c @@ -0,0 +1,603 @@ +/* + * isdnhdlc.c -- General purpose ISDN HDLC decoder. + * + *Copyright (C) 2002 Wolfgang Mües + * 2001 Frode Isaksen + * 2001 Kai Germaschewski + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. + */ + +#include +#include +#include +#include + +/*-------------------------------------------------------------------*/ + +MODULE_AUTHOR("Wolfgang Mües , " + "Frode Isaksen , " + "Kai Germaschewski "); +MODULE_DESCRIPTION("General purpose ISDN HDLC decoder"); +MODULE_LICENSE("GPL"); + +/*-------------------------------------------------------------------*/ + +enum { + HDLC_FAST_IDLE,HDLC_GET_FLAG_B0,HDLC_GETFLAG_B1A6,HDLC_GETFLAG_B7, + HDLC_GET_DATA,HDLC_FAST_FLAG +}; + +enum { + HDLC_SEND_DATA,HDLC_SEND_CRC1,HDLC_SEND_FAST_FLAG, + HDLC_SEND_FIRST_FLAG,HDLC_SEND_CRC2,HDLC_SEND_CLOSING_FLAG, + HDLC_SEND_IDLE1,HDLC_SEND_FAST_IDLE,HDLC_SENDFLAG_B0, + HDLC_SENDFLAG_B1A6,HDLC_SENDFLAG_B7,STOPPED +}; + +void isdnhdlc_rcv_init (struct isdnhdlc_vars *hdlc, int do_adapt56) +{ + hdlc->bit_shift = 0; + hdlc->hdlc_bits1 = 0; + hdlc->data_bits = 0; + hdlc->ffbit_shift = 0; + hdlc->data_received = 0; + hdlc->state = HDLC_GET_DATA; + hdlc->do_adapt56 = do_adapt56; + hdlc->dchannel = 0; + hdlc->crc = 0; + hdlc->cbin = 0; + hdlc->shift_reg = 0; + hdlc->ffvalue = 0; + hdlc->dstpos = 0; +} + +void isdnhdlc_out_init (struct isdnhdlc_vars *hdlc, int is_d_channel, int do_adapt56) +{ + hdlc->bit_shift = 0; + hdlc->hdlc_bits1 = 0; + hdlc->data_bits = 0; + hdlc->ffbit_shift = 0; + hdlc->data_received = 0; + hdlc->do_closing = 0; + hdlc->ffvalue = 0; + if (is_d_channel) { + hdlc->dchannel = 1; + hdlc->state = HDLC_SEND_FIRST_FLAG; + } else { + hdlc->dchannel = 0; + hdlc->state = HDLC_SEND_FAST_FLAG; + hdlc->ffvalue = 0x7e; + } + hdlc->cbin = 0x7e; + hdlc->bit_shift = 0; + if(do_adapt56){ + hdlc->do_adapt56 = 1; + hdlc->data_bits = 0; + hdlc->state = HDLC_SENDFLAG_B0; + } else { + hdlc->do_adapt56 = 0; + hdlc->data_bits = 8; + } + hdlc->shift_reg = 0; +} + +/* + isdnhdlc_decode - decodes HDLC frames from a transparent bit stream. + + The source buffer is scanned for valid HDLC frames looking for + flags (01111110) to indicate the start of a frame. If the start of + the frame is found, the bit stuffing is removed (0 after 5 1's). + When a new flag is found, the complete frame has been received + and the CRC is checked. + If a valid frame is found, the function returns the frame length + excluding the CRC with the bit HDLC_END_OF_FRAME set. + If the beginning of a valid frame is found, the function returns + the length. + If a framing error is found (too many 1s and not a flag) the function + returns the length with the bit HDLC_FRAMING_ERROR set. + If a CRC error is found the function returns the length with the + bit HDLC_CRC_ERROR set. + If the frame length exceeds the destination buffer size, the function + returns the length with the bit HDLC_LENGTH_ERROR set. + + src - source buffer + slen - source buffer length + count - number of bytes removed (decoded) from the source buffer + dst _ destination buffer + dsize - destination buffer size + returns - number of decoded bytes in the destination buffer and status + flag. + */ +int isdnhdlc_decode (struct isdnhdlc_vars *hdlc, const unsigned char *src, + int slen, int *count, unsigned char *dst, int dsize) +{ + int status=0; + + static const unsigned char fast_flag[]={ + 0x00,0x00,0x00,0x20,0x30,0x38,0x3c,0x3e,0x3f + }; + + static const unsigned char fast_flag_value[]={ + 0x00,0x7e,0xfc,0xf9,0xf3,0xe7,0xcf,0x9f,0x3f + }; + + static const unsigned char fast_abort[]={ + 0x00,0x00,0x80,0xc0,0xe0,0xf0,0xf8,0xfc,0xfe,0xff + }; + + *count = slen; + + while(slen > 0){ + if(hdlc->bit_shift==0){ + hdlc->cbin = *src++; + slen--; + hdlc->bit_shift = 8; + if(hdlc->do_adapt56){ + hdlc->bit_shift --; + } + } + + switch(hdlc->state){ + case STOPPED: + return 0; + case HDLC_FAST_IDLE: + if(hdlc->cbin == 0xff){ + hdlc->bit_shift = 0; + break; + } + hdlc->state = HDLC_GET_FLAG_B0; + hdlc->hdlc_bits1 = 0; + hdlc->bit_shift = 8; + break; + case HDLC_GET_FLAG_B0: + if(!(hdlc->cbin & 0x80)) { + hdlc->state = HDLC_GETFLAG_B1A6; + hdlc->hdlc_bits1 = 0; + } else { + if(!hdlc->do_adapt56){ + if(++hdlc->hdlc_bits1 >=8 ) if(hdlc->bit_shift==1) + hdlc->state = HDLC_FAST_IDLE; + } + } + hdlc->cbin<<=1; + hdlc->bit_shift --; + break; + case HDLC_GETFLAG_B1A6: + if(hdlc->cbin & 0x80){ + hdlc->hdlc_bits1++; + if(hdlc->hdlc_bits1==6){ + hdlc->state = HDLC_GETFLAG_B7; + } + } else { + hdlc->hdlc_bits1 = 0; + } + hdlc->cbin<<=1; + hdlc->bit_shift --; + break; + case HDLC_GETFLAG_B7: + if(hdlc->cbin & 0x80) { + hdlc->state = HDLC_GET_FLAG_B0; + } else { + hdlc->state = HDLC_GET_DATA; + hdlc->crc = 0xffff; + hdlc->shift_reg = 0; + hdlc->hdlc_bits1 = 0; + hdlc->data_bits = 0; + hdlc->data_received = 0; + } + hdlc->cbin<<=1; + hdlc->bit_shift --; + break; + case HDLC_GET_DATA: + if(hdlc->cbin & 0x80){ + hdlc->hdlc_bits1++; + switch(hdlc->hdlc_bits1){ + case 6: + break; + case 7: + if(hdlc->data_received) { + // bad frame + status = -HDLC_FRAMING_ERROR; + } + if(!hdlc->do_adapt56){ + if(hdlc->cbin==fast_abort[hdlc->bit_shift+1]){ + hdlc->state = HDLC_FAST_IDLE; + hdlc->bit_shift=1; + break; + } + } else { + hdlc->state = HDLC_GET_FLAG_B0; + } + break; + default: + hdlc->shift_reg>>=1; + hdlc->shift_reg |= 0x80; + hdlc->data_bits++; + break; + } + } else { + switch(hdlc->hdlc_bits1){ + case 5: + break; + case 6: + if(hdlc->data_received){ + if (hdlc->dstpos < 2) { + status = -HDLC_FRAMING_ERROR; + } else if (hdlc->crc != 0xf0b8){ + // crc error + status = -HDLC_CRC_ERROR; + } else { + // remove CRC + hdlc->dstpos -= 2; + // good frame + status = hdlc->dstpos; + } + } + hdlc->crc = 0xffff; + hdlc->shift_reg = 0; + hdlc->data_bits = 0; + if(!hdlc->do_adapt56){ + if(hdlc->cbin==fast_flag[hdlc->bit_shift]){ + hdlc->ffvalue = fast_flag_value[hdlc->bit_shift]; + hdlc->state = HDLC_FAST_FLAG; + hdlc->ffbit_shift = hdlc->bit_shift; + hdlc->bit_shift = 1; + } else { + hdlc->state = HDLC_GET_DATA; + hdlc->data_received = 0; + } + } else { + hdlc->state = HDLC_GET_DATA; + hdlc->data_received = 0; + } + break; + default: + hdlc->shift_reg>>=1; + hdlc->data_bits++; + break; + } + hdlc->hdlc_bits1 = 0; + } + if (status) { + hdlc->dstpos = 0; + *count -= slen; + hdlc->cbin <<= 1; + hdlc->bit_shift--; + return status; + } + if(hdlc->data_bits==8){ + hdlc->data_bits = 0; + hdlc->data_received = 1; + hdlc->crc = crc_ccitt_byte(hdlc->crc, hdlc->shift_reg); + + // good byte received + if (hdlc->dstpos < dsize) { + dst[hdlc->dstpos++] = hdlc->shift_reg; + } else { + // frame too long + status = -HDLC_LENGTH_ERROR; + hdlc->dstpos = 0; + } + } + hdlc->cbin <<= 1; + hdlc->bit_shift--; + break; + case HDLC_FAST_FLAG: + if(hdlc->cbin==hdlc->ffvalue){ + hdlc->bit_shift = 0; + break; + } else { + if(hdlc->cbin == 0xff){ + hdlc->state = HDLC_FAST_IDLE; + hdlc->bit_shift=0; + } else if(hdlc->ffbit_shift==8){ + hdlc->state = HDLC_GETFLAG_B7; + break; + } else { + hdlc->shift_reg = fast_abort[hdlc->ffbit_shift-1]; + hdlc->hdlc_bits1 = hdlc->ffbit_shift-2; + if(hdlc->hdlc_bits1<0)hdlc->hdlc_bits1 = 0; + hdlc->data_bits = hdlc->ffbit_shift-1; + hdlc->state = HDLC_GET_DATA; + hdlc->data_received = 0; + } + } + break; + default: + break; + } + } + *count -= slen; + return 0; +} + +/* + isdnhdlc_encode - encodes HDLC frames to a transparent bit stream. + + The bit stream starts with a beginning flag (01111110). After + that each byte is added to the bit stream with bit stuffing added + (0 after 5 1's). + When the last byte has been removed from the source buffer, the + CRC (2 bytes is added) and the frame terminates with the ending flag. + For the dchannel, the idle character (all 1's) is also added at the end. + If this function is called with empty source buffer (slen=0), flags or + idle character will be generated. + + src - source buffer + slen - source buffer length + count - number of bytes removed (encoded) from source buffer + dst _ destination buffer + dsize - destination buffer size + returns - number of encoded bytes in the destination buffer +*/ +int isdnhdlc_encode(struct isdnhdlc_vars *hdlc, const unsigned char *src, + unsigned short slen, int *count, + unsigned char *dst, int dsize) +{ + static const unsigned char xfast_flag_value[] = { + 0x7e,0x3f,0x9f,0xcf,0xe7,0xf3,0xf9,0xfc,0x7e + }; + + int len = 0; + + *count = slen; + + while (dsize > 0) { + if(hdlc->bit_shift==0){ + if(slen && !hdlc->do_closing){ + hdlc->shift_reg = *src++; + slen--; + if (slen == 0) + hdlc->do_closing = 1; /* closing sequence, CRC + flag(s) */ + hdlc->bit_shift = 8; + } else { + if(hdlc->state == HDLC_SEND_DATA){ + if(hdlc->data_received){ + hdlc->state = HDLC_SEND_CRC1; + hdlc->crc ^= 0xffff; + hdlc->bit_shift = 8; + hdlc->shift_reg = hdlc->crc & 0xff; + } else if(!hdlc->do_adapt56){ + hdlc->state = HDLC_SEND_FAST_FLAG; + } else { + hdlc->state = HDLC_SENDFLAG_B0; + } + } + + } + } + + switch(hdlc->state){ + case STOPPED: + while (dsize--) + *dst++ = 0xff; + + return dsize; + case HDLC_SEND_FAST_FLAG: + hdlc->do_closing = 0; + if(slen == 0){ + *dst++ = hdlc->ffvalue; + len++; + dsize--; + break; + } + if(hdlc->bit_shift==8){ + hdlc->cbin = hdlc->ffvalue>>(8-hdlc->data_bits); + hdlc->state = HDLC_SEND_DATA; + hdlc->crc = 0xffff; + hdlc->hdlc_bits1 = 0; + hdlc->data_received = 1; + } + break; + case HDLC_SENDFLAG_B0: + hdlc->do_closing = 0; + hdlc->cbin <<= 1; + hdlc->data_bits++; + hdlc->hdlc_bits1 = 0; + hdlc->state = HDLC_SENDFLAG_B1A6; + break; + case HDLC_SENDFLAG_B1A6: + hdlc->cbin <<= 1; + hdlc->data_bits++; + hdlc->cbin++; + if(++hdlc->hdlc_bits1 == 6) + hdlc->state = HDLC_SENDFLAG_B7; + break; + case HDLC_SENDFLAG_B7: + hdlc->cbin <<= 1; + hdlc->data_bits++; + if(slen == 0){ + hdlc->state = HDLC_SENDFLAG_B0; + break; + } + if(hdlc->bit_shift==8){ + hdlc->state = HDLC_SEND_DATA; + hdlc->crc = 0xffff; + hdlc->hdlc_bits1 = 0; + hdlc->data_received = 1; + } + break; + case HDLC_SEND_FIRST_FLAG: + hdlc->data_received = 1; + if(hdlc->data_bits==8){ + hdlc->state = HDLC_SEND_DATA; + hdlc->crc = 0xffff; + hdlc->hdlc_bits1 = 0; + break; + } + hdlc->cbin <<= 1; + hdlc->data_bits++; + if(hdlc->shift_reg & 0x01) + hdlc->cbin++; + hdlc->shift_reg >>= 1; + hdlc->bit_shift--; + if(hdlc->bit_shift==0){ + hdlc->state = HDLC_SEND_DATA; + hdlc->crc = 0xffff; + hdlc->hdlc_bits1 = 0; + } + break; + case HDLC_SEND_DATA: + hdlc->cbin <<= 1; + hdlc->data_bits++; + if(hdlc->hdlc_bits1 == 5){ + hdlc->hdlc_bits1 = 0; + break; + } + if(hdlc->bit_shift==8){ + hdlc->crc = crc_ccitt_byte(hdlc->crc, hdlc->shift_reg); + } + if(hdlc->shift_reg & 0x01){ + hdlc->hdlc_bits1++; + hdlc->cbin++; + hdlc->shift_reg >>= 1; + hdlc->bit_shift--; + } else { + hdlc->hdlc_bits1 = 0; + hdlc->shift_reg >>= 1; + hdlc->bit_shift--; + } + break; + case HDLC_SEND_CRC1: + hdlc->cbin <<= 1; + hdlc->data_bits++; + if(hdlc->hdlc_bits1 == 5){ + hdlc->hdlc_bits1 = 0; + break; + } + if(hdlc->shift_reg & 0x01){ + hdlc->hdlc_bits1++; + hdlc->cbin++; + hdlc->shift_reg >>= 1; + hdlc->bit_shift--; + } else { + hdlc->hdlc_bits1 = 0; + hdlc->shift_reg >>= 1; + hdlc->bit_shift--; + } + if(hdlc->bit_shift==0){ + hdlc->shift_reg = (hdlc->crc >> 8); + hdlc->state = HDLC_SEND_CRC2; + hdlc->bit_shift = 8; + } + break; + case HDLC_SEND_CRC2: + hdlc->cbin <<= 1; + hdlc->data_bits++; + if(hdlc->hdlc_bits1 == 5){ + hdlc->hdlc_bits1 = 0; + break; + } + if(hdlc->shift_reg & 0x01){ + hdlc->hdlc_bits1++; + hdlc->cbin++; + hdlc->shift_reg >>= 1; + hdlc->bit_shift--; + } else { + hdlc->hdlc_bits1 = 0; + hdlc->shift_reg >>= 1; + hdlc->bit_shift--; + } + if(hdlc->bit_shift==0){ + hdlc->shift_reg = 0x7e; + hdlc->state = HDLC_SEND_CLOSING_FLAG; + hdlc->bit_shift = 8; + } + break; + case HDLC_SEND_CLOSING_FLAG: + hdlc->cbin <<= 1; + hdlc->data_bits++; + if(hdlc->hdlc_bits1 == 5){ + hdlc->hdlc_bits1 = 0; + break; + } + if(hdlc->shift_reg & 0x01){ + hdlc->cbin++; + } + hdlc->shift_reg >>= 1; + hdlc->bit_shift--; + if(hdlc->bit_shift==0){ + hdlc->ffvalue = xfast_flag_value[hdlc->data_bits]; + if(hdlc->dchannel){ + hdlc->ffvalue = 0x7e; + hdlc->state = HDLC_SEND_IDLE1; + hdlc->bit_shift = 8-hdlc->data_bits; + if(hdlc->bit_shift==0) + hdlc->state = HDLC_SEND_FAST_IDLE; + } else { + if(!hdlc->do_adapt56){ + hdlc->state = HDLC_SEND_FAST_FLAG; + hdlc->data_received = 0; + } else { + hdlc->state = HDLC_SENDFLAG_B0; + hdlc->data_received = 0; + } + // Finished with this frame, send flags + if (dsize > 1) dsize = 1; + } + } + break; + case HDLC_SEND_IDLE1: + hdlc->do_closing = 0; + hdlc->cbin <<= 1; + hdlc->cbin++; + hdlc->data_bits++; + hdlc->bit_shift--; + if(hdlc->bit_shift==0){ + hdlc->state = HDLC_SEND_FAST_IDLE; + hdlc->bit_shift = 0; + } + break; + case HDLC_SEND_FAST_IDLE: + hdlc->do_closing = 0; + hdlc->cbin = 0xff; + hdlc->data_bits = 8; + if(hdlc->bit_shift == 8){ + hdlc->cbin = 0x7e; + hdlc->state = HDLC_SEND_FIRST_FLAG; + } else { + *dst++ = hdlc->cbin; + hdlc->bit_shift = hdlc->data_bits = 0; + len++; + dsize = 0; + } + break; + default: + break; + } + if(hdlc->do_adapt56){ + if(hdlc->data_bits==7){ + hdlc->cbin <<= 1; + hdlc->cbin++; + hdlc->data_bits++; + } + } + if(hdlc->data_bits==8){ + *dst++ = hdlc->cbin; + hdlc->data_bits = 0; + len++; + dsize--; + } + } + *count -= slen; + + return len; +} + +EXPORT_SYMBOL(isdnhdlc_rcv_init); +EXPORT_SYMBOL(isdnhdlc_decode); +EXPORT_SYMBOL(isdnhdlc_out_init); +EXPORT_SYMBOL(isdnhdlc_encode); diff --git a/include/linux/isdn/hdlc.h b/include/linux/isdn/hdlc.h new file mode 100644 index 000000000000..cf0a95a24015 --- /dev/null +++ b/include/linux/isdn/hdlc.h @@ -0,0 +1,70 @@ +/* + * isdnhdlc.h -- General purpose ISDN HDLC decoder. + * + * Implementation of a HDLC decoder/encoder in software. + * Neccessary because some ISDN devices don't have HDLC + * controllers. Also included: a bit reversal table. + * + *Copyright (C) 2002 Wolfgang Mües + * 2001 Frode Isaksen + * 2001 Kai Germaschewski + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. + */ + +#ifndef __ISDNHDLC_H__ +#define __ISDNHDLC_H__ + +struct isdnhdlc_vars { + int bit_shift; + int hdlc_bits1; + int data_bits; + int ffbit_shift; // encoding only + int state; + int dstpos; + + unsigned short crc; + + unsigned char cbin; + unsigned char shift_reg; + unsigned char ffvalue; + + unsigned int data_received:1; // set if transferring data + unsigned int dchannel:1; // set if D channel (send idle instead of flags) + unsigned int do_adapt56:1; // set if 56K adaptation + unsigned int do_closing:1; // set if in closing phase (need to send CRC + flag +}; + + +/* + The return value from isdnhdlc_decode is + the frame length, 0 if no complete frame was decoded, + or a negative error number +*/ +#define HDLC_FRAMING_ERROR 1 +#define HDLC_CRC_ERROR 2 +#define HDLC_LENGTH_ERROR 3 + +extern void isdnhdlc_rcv_init (struct isdnhdlc_vars *hdlc, int do_adapt56); + +extern int isdnhdlc_decode (struct isdnhdlc_vars *hdlc, const unsigned char *src, int slen,int *count, + unsigned char *dst, int dsize); + +extern void isdnhdlc_out_init (struct isdnhdlc_vars *hdlc,int is_d_channel,int do_adapt56); + +extern int isdnhdlc_encode (struct isdnhdlc_vars *hdlc,const unsigned char *src,unsigned short slen,int *count, + unsigned char *dst,int dsize); + +#endif /* __ISDNHDLC_H__ */ -- cgit v1.2.3 From 6bd4bcd3cd8affc09eaee7efbc037f65f4a71501 Mon Sep 17 00:00:00 2001 From: Karsten Keil Date: Wed, 8 Jul 2009 19:11:09 +0200 Subject: ISDN: Clean up isdnhdlc code Clean up isdnhdlc to meet current code standard. Remove hint to already removed bit reversal table. Signed-off-by: Karsten Keil --- drivers/isdn/i4l/isdnhdlc.c | 366 +++++++++++++++++++++++--------------------- include/linux/isdn/hdlc.h | 68 ++++---- 2 files changed, 231 insertions(+), 203 deletions(-) (limited to 'include') diff --git a/drivers/isdn/i4l/isdnhdlc.c b/drivers/isdn/i4l/isdnhdlc.c index 44ec7418496b..b80e55ab8914 100644 --- a/drivers/isdn/i4l/isdnhdlc.c +++ b/drivers/isdn/i4l/isdnhdlc.c @@ -1,23 +1,24 @@ /* * isdnhdlc.c -- General purpose ISDN HDLC decoder. * - *Copyright (C) 2002 Wolfgang Mües - * 2001 Frode Isaksen - * 2001 Kai Germaschewski + * Copyright (C) + * 2002 Wolfgang Mües + * 2001 Frode Isaksen + * 2001 Kai Germaschewski * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. */ #include @@ -36,20 +37,20 @@ MODULE_LICENSE("GPL"); /*-------------------------------------------------------------------*/ enum { - HDLC_FAST_IDLE,HDLC_GET_FLAG_B0,HDLC_GETFLAG_B1A6,HDLC_GETFLAG_B7, - HDLC_GET_DATA,HDLC_FAST_FLAG + HDLC_FAST_IDLE, HDLC_GET_FLAG_B0, HDLC_GETFLAG_B1A6, HDLC_GETFLAG_B7, + HDLC_GET_DATA, HDLC_FAST_FLAG }; enum { - HDLC_SEND_DATA,HDLC_SEND_CRC1,HDLC_SEND_FAST_FLAG, - HDLC_SEND_FIRST_FLAG,HDLC_SEND_CRC2,HDLC_SEND_CLOSING_FLAG, - HDLC_SEND_IDLE1,HDLC_SEND_FAST_IDLE,HDLC_SENDFLAG_B0, - HDLC_SENDFLAG_B1A6,HDLC_SENDFLAG_B7,STOPPED + HDLC_SEND_DATA, HDLC_SEND_CRC1, HDLC_SEND_FAST_FLAG, + HDLC_SEND_FIRST_FLAG, HDLC_SEND_CRC2, HDLC_SEND_CLOSING_FLAG, + HDLC_SEND_IDLE1, HDLC_SEND_FAST_IDLE, HDLC_SENDFLAG_B0, + HDLC_SENDFLAG_B1A6, HDLC_SENDFLAG_B7, STOPPED }; -void isdnhdlc_rcv_init (struct isdnhdlc_vars *hdlc, int do_adapt56) +void isdnhdlc_rcv_init(struct isdnhdlc_vars *hdlc, int do_adapt56) { - hdlc->bit_shift = 0; + hdlc->bit_shift = 0; hdlc->hdlc_bits1 = 0; hdlc->data_bits = 0; hdlc->ffbit_shift = 0; @@ -63,10 +64,12 @@ void isdnhdlc_rcv_init (struct isdnhdlc_vars *hdlc, int do_adapt56) hdlc->ffvalue = 0; hdlc->dstpos = 0; } +EXPORT_SYMBOL(isdnhdlc_out_init); -void isdnhdlc_out_init (struct isdnhdlc_vars *hdlc, int is_d_channel, int do_adapt56) +void isdnhdlc_out_init(struct isdnhdlc_vars *hdlc, int is_d_channel, + int do_adapt56) { - hdlc->bit_shift = 0; + hdlc->bit_shift = 0; hdlc->hdlc_bits1 = 0; hdlc->data_bits = 0; hdlc->ffbit_shift = 0; @@ -83,7 +86,7 @@ void isdnhdlc_out_init (struct isdnhdlc_vars *hdlc, int is_d_channel, int do_ada } hdlc->cbin = 0x7e; hdlc->bit_shift = 0; - if(do_adapt56){ + if (do_adapt56) { hdlc->do_adapt56 = 1; hdlc->data_bits = 0; hdlc->state = HDLC_SENDFLAG_B0; @@ -93,6 +96,25 @@ void isdnhdlc_out_init (struct isdnhdlc_vars *hdlc, int is_d_channel, int do_ada } hdlc->shift_reg = 0; } +EXPORT_SYMBOL(isdnhdlc_rcv_init); + +static int +check_frame(struct isdnhdlc_vars *hdlc) +{ + int status; + + if (hdlc->dstpos < 2) /* too small - framing error */ + status = -HDLC_FRAMING_ERROR; + else if (hdlc->crc != 0xf0b8) /* crc error */ + status = -HDLC_CRC_ERROR; + else { + /* remove CRC */ + hdlc->dstpos -= 2; + /* good frame */ + status = hdlc->dstpos; + } + return status; +} /* isdnhdlc_decode - decodes HDLC frames from a transparent bit stream. @@ -121,40 +143,63 @@ void isdnhdlc_out_init (struct isdnhdlc_vars *hdlc, int is_d_channel, int do_ada returns - number of decoded bytes in the destination buffer and status flag. */ -int isdnhdlc_decode (struct isdnhdlc_vars *hdlc, const unsigned char *src, - int slen, int *count, unsigned char *dst, int dsize) +int isdnhdlc_decode(struct isdnhdlc_vars *hdlc, const u8 *src, int slen, + int *count, u8 *dst, int dsize) { - int status=0; + int status = 0; - static const unsigned char fast_flag[]={ - 0x00,0x00,0x00,0x20,0x30,0x38,0x3c,0x3e,0x3f + static const unsigned char fast_flag[] = { + 0x00, 0x00, 0x00, 0x20, 0x30, 0x38, 0x3c, 0x3e, 0x3f }; - static const unsigned char fast_flag_value[]={ - 0x00,0x7e,0xfc,0xf9,0xf3,0xe7,0xcf,0x9f,0x3f + static const unsigned char fast_flag_value[] = { + 0x00, 0x7e, 0xfc, 0xf9, 0xf3, 0xe7, 0xcf, 0x9f, 0x3f }; - static const unsigned char fast_abort[]={ - 0x00,0x00,0x80,0xc0,0xe0,0xf0,0xf8,0xfc,0xfe,0xff + static const unsigned char fast_abort[] = { + 0x00, 0x00, 0x80, 0xc0, 0xe0, 0xf0, 0xf8, 0xfc, 0xfe, 0xff }; +#define handle_fast_flag(h) \ + do {\ + if (h->cbin == fast_flag[h->bit_shift]) {\ + h->ffvalue = fast_flag_value[h->bit_shift];\ + h->state = HDLC_FAST_FLAG;\ + h->ffbit_shift = h->bit_shift;\ + h->bit_shift = 1;\ + } else {\ + h->state = HDLC_GET_DATA;\ + h->data_received = 0;\ + } \ + } while (0) + +#define handle_abort(h) \ + do {\ + h->shift_reg = fast_abort[h->ffbit_shift - 1];\ + h->hdlc_bits1 = h->ffbit_shift - 2;\ + if (h->hdlc_bits1 < 0)\ + h->hdlc_bits1 = 0;\ + h->data_bits = h->ffbit_shift - 1;\ + h->state = HDLC_GET_DATA;\ + h->data_received = 0;\ + } while (0) + *count = slen; - while(slen > 0){ - if(hdlc->bit_shift==0){ + while (slen > 0) { + if (hdlc->bit_shift == 0) { hdlc->cbin = *src++; slen--; hdlc->bit_shift = 8; - if(hdlc->do_adapt56){ - hdlc->bit_shift --; - } + if (hdlc->do_adapt56) + hdlc->bit_shift--; } - switch(hdlc->state){ + switch (hdlc->state) { case STOPPED: return 0; case HDLC_FAST_IDLE: - if(hdlc->cbin == 0xff){ + if (hdlc->cbin == 0xff) { hdlc->bit_shift = 0; break; } @@ -163,32 +208,30 @@ int isdnhdlc_decode (struct isdnhdlc_vars *hdlc, const unsigned char *src, hdlc->bit_shift = 8; break; case HDLC_GET_FLAG_B0: - if(!(hdlc->cbin & 0x80)) { + if (!(hdlc->cbin & 0x80)) { hdlc->state = HDLC_GETFLAG_B1A6; hdlc->hdlc_bits1 = 0; } else { - if(!hdlc->do_adapt56){ - if(++hdlc->hdlc_bits1 >=8 ) if(hdlc->bit_shift==1) + if ((!hdlc->do_adapt56) && + (++hdlc->hdlc_bits1 >= 8) && + (hdlc->bit_shift == 1)) hdlc->state = HDLC_FAST_IDLE; - } } - hdlc->cbin<<=1; - hdlc->bit_shift --; + hdlc->cbin <<= 1; + hdlc->bit_shift--; break; case HDLC_GETFLAG_B1A6: - if(hdlc->cbin & 0x80){ + if (hdlc->cbin & 0x80) { hdlc->hdlc_bits1++; - if(hdlc->hdlc_bits1==6){ + if (hdlc->hdlc_bits1 == 6) hdlc->state = HDLC_GETFLAG_B7; - } - } else { + } else hdlc->hdlc_bits1 = 0; - } - hdlc->cbin<<=1; - hdlc->bit_shift --; + hdlc->cbin <<= 1; + hdlc->bit_shift--; break; case HDLC_GETFLAG_B7: - if(hdlc->cbin & 0x80) { + if (hdlc->cbin & 0x80) { hdlc->state = HDLC_GET_FLAG_B0; } else { hdlc->state = HDLC_GET_DATA; @@ -198,74 +241,55 @@ int isdnhdlc_decode (struct isdnhdlc_vars *hdlc, const unsigned char *src, hdlc->data_bits = 0; hdlc->data_received = 0; } - hdlc->cbin<<=1; - hdlc->bit_shift --; + hdlc->cbin <<= 1; + hdlc->bit_shift--; break; case HDLC_GET_DATA: - if(hdlc->cbin & 0x80){ + if (hdlc->cbin & 0x80) { hdlc->hdlc_bits1++; - switch(hdlc->hdlc_bits1){ + switch (hdlc->hdlc_bits1) { case 6: break; case 7: - if(hdlc->data_received) { - // bad frame + if (hdlc->data_received) + /* bad frame */ status = -HDLC_FRAMING_ERROR; - } - if(!hdlc->do_adapt56){ - if(hdlc->cbin==fast_abort[hdlc->bit_shift+1]){ - hdlc->state = HDLC_FAST_IDLE; - hdlc->bit_shift=1; + if (!hdlc->do_adapt56) { + if (hdlc->cbin == fast_abort + [hdlc->bit_shift + 1]) { + hdlc->state = + HDLC_FAST_IDLE; + hdlc->bit_shift = 1; break; } - } else { + } else hdlc->state = HDLC_GET_FLAG_B0; - } break; default: - hdlc->shift_reg>>=1; + hdlc->shift_reg >>= 1; hdlc->shift_reg |= 0x80; hdlc->data_bits++; break; } } else { - switch(hdlc->hdlc_bits1){ + switch (hdlc->hdlc_bits1) { case 5: break; case 6: - if(hdlc->data_received){ - if (hdlc->dstpos < 2) { - status = -HDLC_FRAMING_ERROR; - } else if (hdlc->crc != 0xf0b8){ - // crc error - status = -HDLC_CRC_ERROR; - } else { - // remove CRC - hdlc->dstpos -= 2; - // good frame - status = hdlc->dstpos; - } - } + if (hdlc->data_received) + status = check_frame(hdlc); hdlc->crc = 0xffff; hdlc->shift_reg = 0; hdlc->data_bits = 0; - if(!hdlc->do_adapt56){ - if(hdlc->cbin==fast_flag[hdlc->bit_shift]){ - hdlc->ffvalue = fast_flag_value[hdlc->bit_shift]; - hdlc->state = HDLC_FAST_FLAG; - hdlc->ffbit_shift = hdlc->bit_shift; - hdlc->bit_shift = 1; - } else { - hdlc->state = HDLC_GET_DATA; - hdlc->data_received = 0; - } - } else { + if (!hdlc->do_adapt56) + handle_fast_flag(hdlc); + else { hdlc->state = HDLC_GET_DATA; hdlc->data_received = 0; } break; default: - hdlc->shift_reg>>=1; + hdlc->shift_reg >>= 1; hdlc->data_bits++; break; } @@ -278,16 +302,17 @@ int isdnhdlc_decode (struct isdnhdlc_vars *hdlc, const unsigned char *src, hdlc->bit_shift--; return status; } - if(hdlc->data_bits==8){ + if (hdlc->data_bits == 8) { hdlc->data_bits = 0; hdlc->data_received = 1; - hdlc->crc = crc_ccitt_byte(hdlc->crc, hdlc->shift_reg); + hdlc->crc = crc_ccitt_byte(hdlc->crc, + hdlc->shift_reg); - // good byte received - if (hdlc->dstpos < dsize) { + /* good byte received */ + if (hdlc->dstpos < dsize) dst[hdlc->dstpos++] = hdlc->shift_reg; - } else { - // frame too long + else { + /* frame too long */ status = -HDLC_LENGTH_ERROR; hdlc->dstpos = 0; } @@ -296,24 +321,18 @@ int isdnhdlc_decode (struct isdnhdlc_vars *hdlc, const unsigned char *src, hdlc->bit_shift--; break; case HDLC_FAST_FLAG: - if(hdlc->cbin==hdlc->ffvalue){ + if (hdlc->cbin == hdlc->ffvalue) { hdlc->bit_shift = 0; break; } else { - if(hdlc->cbin == 0xff){ + if (hdlc->cbin == 0xff) { hdlc->state = HDLC_FAST_IDLE; - hdlc->bit_shift=0; - } else if(hdlc->ffbit_shift==8){ + hdlc->bit_shift = 0; + } else if (hdlc->ffbit_shift == 8) { hdlc->state = HDLC_GETFLAG_B7; break; - } else { - hdlc->shift_reg = fast_abort[hdlc->ffbit_shift-1]; - hdlc->hdlc_bits1 = hdlc->ffbit_shift-2; - if(hdlc->hdlc_bits1<0)hdlc->hdlc_bits1 = 0; - hdlc->data_bits = hdlc->ffbit_shift-1; - hdlc->state = HDLC_GET_DATA; - hdlc->data_received = 0; - } + } else + handle_abort(hdlc); } break; default: @@ -323,7 +342,7 @@ int isdnhdlc_decode (struct isdnhdlc_vars *hdlc, const unsigned char *src, *count -= slen; return 0; } - +EXPORT_SYMBOL(isdnhdlc_decode); /* isdnhdlc_encode - encodes HDLC frames to a transparent bit stream. @@ -343,12 +362,11 @@ int isdnhdlc_decode (struct isdnhdlc_vars *hdlc, const unsigned char *src, dsize - destination buffer size returns - number of encoded bytes in the destination buffer */ -int isdnhdlc_encode(struct isdnhdlc_vars *hdlc, const unsigned char *src, - unsigned short slen, int *count, - unsigned char *dst, int dsize) +int isdnhdlc_encode(struct isdnhdlc_vars *hdlc, const u8 *src, u16 slen, + int *count, u8 *dst, int dsize) { static const unsigned char xfast_flag_value[] = { - 0x7e,0x3f,0x9f,0xcf,0xe7,0xf3,0xf9,0xfc,0x7e + 0x7e, 0x3f, 0x9f, 0xcf, 0xe7, 0xf3, 0xf9, 0xfc, 0x7e }; int len = 0; @@ -356,31 +374,34 @@ int isdnhdlc_encode(struct isdnhdlc_vars *hdlc, const unsigned char *src, *count = slen; while (dsize > 0) { - if(hdlc->bit_shift==0){ - if(slen && !hdlc->do_closing){ + if (hdlc->bit_shift == 0) { + if (slen && !hdlc->do_closing) { hdlc->shift_reg = *src++; slen--; if (slen == 0) - hdlc->do_closing = 1; /* closing sequence, CRC + flag(s) */ + /* closing sequence, CRC + flag(s) */ + hdlc->do_closing = 1; hdlc->bit_shift = 8; } else { - if(hdlc->state == HDLC_SEND_DATA){ - if(hdlc->data_received){ + if (hdlc->state == HDLC_SEND_DATA) { + if (hdlc->data_received) { hdlc->state = HDLC_SEND_CRC1; hdlc->crc ^= 0xffff; hdlc->bit_shift = 8; - hdlc->shift_reg = hdlc->crc & 0xff; - } else if(!hdlc->do_adapt56){ - hdlc->state = HDLC_SEND_FAST_FLAG; - } else { - hdlc->state = HDLC_SENDFLAG_B0; - } + hdlc->shift_reg = + hdlc->crc & 0xff; + } else if (!hdlc->do_adapt56) + hdlc->state = + HDLC_SEND_FAST_FLAG; + else + hdlc->state = + HDLC_SENDFLAG_B0; } } } - switch(hdlc->state){ + switch (hdlc->state) { case STOPPED: while (dsize--) *dst++ = 0xff; @@ -388,14 +409,15 @@ int isdnhdlc_encode(struct isdnhdlc_vars *hdlc, const unsigned char *src, return dsize; case HDLC_SEND_FAST_FLAG: hdlc->do_closing = 0; - if(slen == 0){ + if (slen == 0) { *dst++ = hdlc->ffvalue; len++; dsize--; break; } - if(hdlc->bit_shift==8){ - hdlc->cbin = hdlc->ffvalue>>(8-hdlc->data_bits); + if (hdlc->bit_shift == 8) { + hdlc->cbin = hdlc->ffvalue >> + (8 - hdlc->data_bits); hdlc->state = HDLC_SEND_DATA; hdlc->crc = 0xffff; hdlc->hdlc_bits1 = 0; @@ -413,17 +435,17 @@ int isdnhdlc_encode(struct isdnhdlc_vars *hdlc, const unsigned char *src, hdlc->cbin <<= 1; hdlc->data_bits++; hdlc->cbin++; - if(++hdlc->hdlc_bits1 == 6) + if (++hdlc->hdlc_bits1 == 6) hdlc->state = HDLC_SENDFLAG_B7; break; case HDLC_SENDFLAG_B7: hdlc->cbin <<= 1; hdlc->data_bits++; - if(slen == 0){ + if (slen == 0) { hdlc->state = HDLC_SENDFLAG_B0; break; } - if(hdlc->bit_shift==8){ + if (hdlc->bit_shift == 8) { hdlc->state = HDLC_SEND_DATA; hdlc->crc = 0xffff; hdlc->hdlc_bits1 = 0; @@ -432,7 +454,7 @@ int isdnhdlc_encode(struct isdnhdlc_vars *hdlc, const unsigned char *src, break; case HDLC_SEND_FIRST_FLAG: hdlc->data_received = 1; - if(hdlc->data_bits==8){ + if (hdlc->data_bits == 8) { hdlc->state = HDLC_SEND_DATA; hdlc->crc = 0xffff; hdlc->hdlc_bits1 = 0; @@ -440,11 +462,11 @@ int isdnhdlc_encode(struct isdnhdlc_vars *hdlc, const unsigned char *src, } hdlc->cbin <<= 1; hdlc->data_bits++; - if(hdlc->shift_reg & 0x01) + if (hdlc->shift_reg & 0x01) hdlc->cbin++; hdlc->shift_reg >>= 1; hdlc->bit_shift--; - if(hdlc->bit_shift==0){ + if (hdlc->bit_shift == 0) { hdlc->state = HDLC_SEND_DATA; hdlc->crc = 0xffff; hdlc->hdlc_bits1 = 0; @@ -453,14 +475,14 @@ int isdnhdlc_encode(struct isdnhdlc_vars *hdlc, const unsigned char *src, case HDLC_SEND_DATA: hdlc->cbin <<= 1; hdlc->data_bits++; - if(hdlc->hdlc_bits1 == 5){ + if (hdlc->hdlc_bits1 == 5) { hdlc->hdlc_bits1 = 0; break; } - if(hdlc->bit_shift==8){ - hdlc->crc = crc_ccitt_byte(hdlc->crc, hdlc->shift_reg); - } - if(hdlc->shift_reg & 0x01){ + if (hdlc->bit_shift == 8) + hdlc->crc = crc_ccitt_byte(hdlc->crc, + hdlc->shift_reg); + if (hdlc->shift_reg & 0x01) { hdlc->hdlc_bits1++; hdlc->cbin++; hdlc->shift_reg >>= 1; @@ -474,11 +496,11 @@ int isdnhdlc_encode(struct isdnhdlc_vars *hdlc, const unsigned char *src, case HDLC_SEND_CRC1: hdlc->cbin <<= 1; hdlc->data_bits++; - if(hdlc->hdlc_bits1 == 5){ + if (hdlc->hdlc_bits1 == 5) { hdlc->hdlc_bits1 = 0; break; } - if(hdlc->shift_reg & 0x01){ + if (hdlc->shift_reg & 0x01) { hdlc->hdlc_bits1++; hdlc->cbin++; hdlc->shift_reg >>= 1; @@ -488,7 +510,7 @@ int isdnhdlc_encode(struct isdnhdlc_vars *hdlc, const unsigned char *src, hdlc->shift_reg >>= 1; hdlc->bit_shift--; } - if(hdlc->bit_shift==0){ + if (hdlc->bit_shift == 0) { hdlc->shift_reg = (hdlc->crc >> 8); hdlc->state = HDLC_SEND_CRC2; hdlc->bit_shift = 8; @@ -497,11 +519,11 @@ int isdnhdlc_encode(struct isdnhdlc_vars *hdlc, const unsigned char *src, case HDLC_SEND_CRC2: hdlc->cbin <<= 1; hdlc->data_bits++; - if(hdlc->hdlc_bits1 == 5){ + if (hdlc->hdlc_bits1 == 5) { hdlc->hdlc_bits1 = 0; break; } - if(hdlc->shift_reg & 0x01){ + if (hdlc->shift_reg & 0x01) { hdlc->hdlc_bits1++; hdlc->cbin++; hdlc->shift_reg >>= 1; @@ -511,7 +533,7 @@ int isdnhdlc_encode(struct isdnhdlc_vars *hdlc, const unsigned char *src, hdlc->shift_reg >>= 1; hdlc->bit_shift--; } - if(hdlc->bit_shift==0){ + if (hdlc->bit_shift == 0) { hdlc->shift_reg = 0x7e; hdlc->state = HDLC_SEND_CLOSING_FLAG; hdlc->bit_shift = 8; @@ -520,33 +542,36 @@ int isdnhdlc_encode(struct isdnhdlc_vars *hdlc, const unsigned char *src, case HDLC_SEND_CLOSING_FLAG: hdlc->cbin <<= 1; hdlc->data_bits++; - if(hdlc->hdlc_bits1 == 5){ + if (hdlc->hdlc_bits1 == 5) { hdlc->hdlc_bits1 = 0; break; } - if(hdlc->shift_reg & 0x01){ + if (hdlc->shift_reg & 0x01) hdlc->cbin++; - } hdlc->shift_reg >>= 1; hdlc->bit_shift--; - if(hdlc->bit_shift==0){ - hdlc->ffvalue = xfast_flag_value[hdlc->data_bits]; - if(hdlc->dchannel){ + if (hdlc->bit_shift == 0) { + hdlc->ffvalue = + xfast_flag_value[hdlc->data_bits]; + if (hdlc->dchannel) { hdlc->ffvalue = 0x7e; hdlc->state = HDLC_SEND_IDLE1; hdlc->bit_shift = 8-hdlc->data_bits; - if(hdlc->bit_shift==0) - hdlc->state = HDLC_SEND_FAST_IDLE; + if (hdlc->bit_shift == 0) + hdlc->state = + HDLC_SEND_FAST_IDLE; } else { - if(!hdlc->do_adapt56){ - hdlc->state = HDLC_SEND_FAST_FLAG; + if (!hdlc->do_adapt56) { + hdlc->state = + HDLC_SEND_FAST_FLAG; hdlc->data_received = 0; } else { hdlc->state = HDLC_SENDFLAG_B0; hdlc->data_received = 0; } - // Finished with this frame, send flags - if (dsize > 1) dsize = 1; + /* Finished this frame, send flags */ + if (dsize > 1) + dsize = 1; } } break; @@ -556,7 +581,7 @@ int isdnhdlc_encode(struct isdnhdlc_vars *hdlc, const unsigned char *src, hdlc->cbin++; hdlc->data_bits++; hdlc->bit_shift--; - if(hdlc->bit_shift==0){ + if (hdlc->bit_shift == 0) { hdlc->state = HDLC_SEND_FAST_IDLE; hdlc->bit_shift = 0; } @@ -565,12 +590,13 @@ int isdnhdlc_encode(struct isdnhdlc_vars *hdlc, const unsigned char *src, hdlc->do_closing = 0; hdlc->cbin = 0xff; hdlc->data_bits = 8; - if(hdlc->bit_shift == 8){ + if (hdlc->bit_shift == 8) { hdlc->cbin = 0x7e; hdlc->state = HDLC_SEND_FIRST_FLAG; } else { *dst++ = hdlc->cbin; - hdlc->bit_shift = hdlc->data_bits = 0; + hdlc->bit_shift = 0; + hdlc->data_bits = 0; len++; dsize = 0; } @@ -578,14 +604,14 @@ int isdnhdlc_encode(struct isdnhdlc_vars *hdlc, const unsigned char *src, default: break; } - if(hdlc->do_adapt56){ - if(hdlc->data_bits==7){ + if (hdlc->do_adapt56) { + if (hdlc->data_bits == 7) { hdlc->cbin <<= 1; hdlc->cbin++; hdlc->data_bits++; } } - if(hdlc->data_bits==8){ + if (hdlc->data_bits == 8) { *dst++ = hdlc->cbin; hdlc->data_bits = 0; len++; @@ -596,8 +622,4 @@ int isdnhdlc_encode(struct isdnhdlc_vars *hdlc, const unsigned char *src, return len; } - -EXPORT_SYMBOL(isdnhdlc_rcv_init); -EXPORT_SYMBOL(isdnhdlc_decode); -EXPORT_SYMBOL(isdnhdlc_out_init); EXPORT_SYMBOL(isdnhdlc_encode); diff --git a/include/linux/isdn/hdlc.h b/include/linux/isdn/hdlc.h index cf0a95a24015..8f3540c7f692 100644 --- a/include/linux/isdn/hdlc.h +++ b/include/linux/isdn/hdlc.h @@ -1,27 +1,28 @@ /* - * isdnhdlc.h -- General purpose ISDN HDLC decoder. + * hdlc.h -- General purpose ISDN HDLC decoder. * * Implementation of a HDLC decoder/encoder in software. * Neccessary because some ISDN devices don't have HDLC - * controllers. Also included: a bit reversal table. + * controllers. * - *Copyright (C) 2002 Wolfgang Mües - * 2001 Frode Isaksen - * 2001 Kai Germaschewski + * Copyright (C) + * 2002 Wolfgang Mües + * 2001 Frode Isaksen + * 2001 Kai Germaschewski * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. */ #ifndef __ISDNHDLC_H__ @@ -31,20 +32,24 @@ struct isdnhdlc_vars { int bit_shift; int hdlc_bits1; int data_bits; - int ffbit_shift; // encoding only + int ffbit_shift; /* encoding only */ int state; int dstpos; - unsigned short crc; + u16 crc; - unsigned char cbin; - unsigned char shift_reg; - unsigned char ffvalue; + u8 cbin; + u8 shift_reg; + u8 ffvalue; - unsigned int data_received:1; // set if transferring data - unsigned int dchannel:1; // set if D channel (send idle instead of flags) - unsigned int do_adapt56:1; // set if 56K adaptation - unsigned int do_closing:1; // set if in closing phase (need to send CRC + flag + /* set if transferring data */ + u32 data_received:1; + /* set if D channel (send idle instead of flags) */ + u32 dchannel:1; + /* set if 56K adaptation */ + u32 do_adapt56:1; + /* set if in closing phase (need to send CRC + flag) */ + u32 do_closing:1; }; @@ -57,14 +62,15 @@ struct isdnhdlc_vars { #define HDLC_CRC_ERROR 2 #define HDLC_LENGTH_ERROR 3 -extern void isdnhdlc_rcv_init (struct isdnhdlc_vars *hdlc, int do_adapt56); +extern void isdnhdlc_rcv_init(struct isdnhdlc_vars *hdlc, int do_adapt56); -extern int isdnhdlc_decode (struct isdnhdlc_vars *hdlc, const unsigned char *src, int slen,int *count, - unsigned char *dst, int dsize); +extern int isdnhdlc_decode(struct isdnhdlc_vars *hdlc, const u8 *src, + int slen, int *count, u8 *dst, int dsize); -extern void isdnhdlc_out_init (struct isdnhdlc_vars *hdlc,int is_d_channel,int do_adapt56); +extern void isdnhdlc_out_init(struct isdnhdlc_vars *hdlc, int is_d_channel, + int do_adapt56); -extern int isdnhdlc_encode (struct isdnhdlc_vars *hdlc,const unsigned char *src,unsigned short slen,int *count, - unsigned char *dst,int dsize); +extern int isdnhdlc_encode(struct isdnhdlc_vars *hdlc, const u8 *src, + u16 slen, int *count, u8 *dst, int dsize); #endif /* __ISDNHDLC_H__ */ -- cgit v1.2.3 From c38fc3bc2ecddd4f5278131603e6964cbed071b2 Mon Sep 17 00:00:00 2001 From: Karsten Keil Date: Wed, 8 Jul 2009 20:31:42 +0200 Subject: ISDN: Add support for none reverse bitstreams to isdnhdc The original isdnhdlc code was developed for devices which had reversed bitorder in the byte stream. Adding code to handle normal bitstreams as well. Signed-off-by: Karsten Keil --- drivers/isdn/hisax/st5481_b.c | 5 ++- drivers/isdn/hisax/st5481_d.c | 2 +- drivers/isdn/hisax/st5481_usb.c | 11 ++++--- drivers/isdn/i4l/isdnhdlc.c | 70 ++++++++++++++++++++--------------------- include/linux/isdn/hdlc.h | 12 +++++-- 5 files changed, 56 insertions(+), 44 deletions(-) (limited to 'include') diff --git a/drivers/isdn/hisax/st5481_b.c b/drivers/isdn/hisax/st5481_b.c index 0074b600a0ef..95b1cdd97958 100644 --- a/drivers/isdn/hisax/st5481_b.c +++ b/drivers/isdn/hisax/st5481_b.c @@ -218,7 +218,10 @@ static void st5481B_mode(struct st5481_bcs *bcs, int mode) if (bcs->mode != L1_MODE_NULL) { // Open the B channel if (bcs->mode != L1_MODE_TRANS) { - isdnhdlc_out_init(&b_out->hdlc_state, 0, bcs->mode == L1_MODE_HDLC_56K); + u32 features = HDLC_BITREVERSE; + if (bcs->mode == L1_MODE_HDLC_56K) + features |= HDLC_56KBIT; + isdnhdlc_out_init(&b_out->hdlc_state, features); } st5481_usb_pipe_reset(adapter, (bcs->channel+1)*2, NULL, NULL); diff --git a/drivers/isdn/hisax/st5481_d.c b/drivers/isdn/hisax/st5481_d.c index 077991c1cd05..39e8e49cfd2d 100644 --- a/drivers/isdn/hisax/st5481_d.c +++ b/drivers/isdn/hisax/st5481_d.c @@ -417,7 +417,7 @@ static void dout_start_xmit(struct FsmInst *fsm, int event, void *arg) DBG(2,"len=%d",skb->len); - isdnhdlc_out_init(&d_out->hdlc_state, 1, 0); + isdnhdlc_out_init(&d_out->hdlc_state, HDLC_DCHANNEL | HDLC_BITREVERSE); if (test_and_set_bit(buf_nr, &d_out->busy)) { WARNING("ep %d urb %d busy %#lx", EP_D_OUT, buf_nr, d_out->busy); diff --git a/drivers/isdn/hisax/st5481_usb.c b/drivers/isdn/hisax/st5481_usb.c index 2b3a055059ea..10d41c5d73ed 100644 --- a/drivers/isdn/hisax/st5481_usb.c +++ b/drivers/isdn/hisax/st5481_usb.c @@ -637,10 +637,13 @@ void st5481_in_mode(struct st5481_in *in, int mode) usb_unlink_urb(in->urb[1]); if (in->mode != L1_MODE_NULL) { - if (in->mode != L1_MODE_TRANS) - isdnhdlc_rcv_init(&in->hdlc_state, - in->mode == L1_MODE_HDLC_56K); - + if (in->mode != L1_MODE_TRANS) { + u32 features = HDLC_BITREVERSE; + + if (in->mode == L1_MODE_HDLC_56K) + features |= HDLC_56KBIT; + isdnhdlc_rcv_init(&in->hdlc_state, features); + } st5481_usb_pipe_reset(in->adapter, in->ep, NULL, NULL); st5481_usb_device_ctrl_msg(in->adapter, in->counter, in->packet_size, diff --git a/drivers/isdn/i4l/isdnhdlc.c b/drivers/isdn/i4l/isdnhdlc.c index b80e55ab8914..df345ce73f48 100644 --- a/drivers/isdn/i4l/isdnhdlc.c +++ b/drivers/isdn/i4l/isdnhdlc.c @@ -2,6 +2,7 @@ * isdnhdlc.c -- General purpose ISDN HDLC decoder. * * Copyright (C) + * 2009 Karsten Keil * 2002 Wolfgang Mües * 2001 Frode Isaksen * 2001 Kai Germaschewski @@ -25,6 +26,7 @@ #include #include #include +#include /*-------------------------------------------------------------------*/ @@ -48,35 +50,21 @@ enum { HDLC_SENDFLAG_B1A6, HDLC_SENDFLAG_B7, STOPPED }; -void isdnhdlc_rcv_init(struct isdnhdlc_vars *hdlc, int do_adapt56) +void isdnhdlc_rcv_init(struct isdnhdlc_vars *hdlc, u32 features) { - hdlc->bit_shift = 0; - hdlc->hdlc_bits1 = 0; - hdlc->data_bits = 0; - hdlc->ffbit_shift = 0; - hdlc->data_received = 0; + memset(hdlc, 0, sizeof(struct isdnhdlc_vars)); hdlc->state = HDLC_GET_DATA; - hdlc->do_adapt56 = do_adapt56; - hdlc->dchannel = 0; - hdlc->crc = 0; - hdlc->cbin = 0; - hdlc->shift_reg = 0; - hdlc->ffvalue = 0; - hdlc->dstpos = 0; + if (features & HDLC_56KBIT) + hdlc->do_adapt56 = 1; + if (features & HDLC_BITREVERSE) + hdlc->do_bitreverse = 1; } EXPORT_SYMBOL(isdnhdlc_out_init); -void isdnhdlc_out_init(struct isdnhdlc_vars *hdlc, int is_d_channel, - int do_adapt56) +void isdnhdlc_out_init(struct isdnhdlc_vars *hdlc, u32 features) { - hdlc->bit_shift = 0; - hdlc->hdlc_bits1 = 0; - hdlc->data_bits = 0; - hdlc->ffbit_shift = 0; - hdlc->data_received = 0; - hdlc->do_closing = 0; - hdlc->ffvalue = 0; - if (is_d_channel) { + memset(hdlc, 0, sizeof(struct isdnhdlc_vars)); + if (features & HDLC_DCHANNEL) { hdlc->dchannel = 1; hdlc->state = HDLC_SEND_FIRST_FLAG; } else { @@ -85,16 +73,13 @@ void isdnhdlc_out_init(struct isdnhdlc_vars *hdlc, int is_d_channel, hdlc->ffvalue = 0x7e; } hdlc->cbin = 0x7e; - hdlc->bit_shift = 0; - if (do_adapt56) { + if (features & HDLC_56KBIT) { hdlc->do_adapt56 = 1; - hdlc->data_bits = 0; hdlc->state = HDLC_SENDFLAG_B0; - } else { - hdlc->do_adapt56 = 0; + } else hdlc->data_bits = 8; - } - hdlc->shift_reg = 0; + if (features & HDLC_BITREVERSE) + hdlc->do_bitreverse = 1; } EXPORT_SYMBOL(isdnhdlc_rcv_init); @@ -188,7 +173,11 @@ int isdnhdlc_decode(struct isdnhdlc_vars *hdlc, const u8 *src, int slen, while (slen > 0) { if (hdlc->bit_shift == 0) { - hdlc->cbin = *src++; + /* the code is for bitreverse streams */ + if (hdlc->do_bitreverse == 0) + hdlc->cbin = bitrev8(*src++); + else + hdlc->cbin = *src++; slen--; hdlc->bit_shift = 8; if (hdlc->do_adapt56) @@ -405,12 +394,15 @@ int isdnhdlc_encode(struct isdnhdlc_vars *hdlc, const u8 *src, u16 slen, case STOPPED: while (dsize--) *dst++ = 0xff; - return dsize; case HDLC_SEND_FAST_FLAG: hdlc->do_closing = 0; if (slen == 0) { - *dst++ = hdlc->ffvalue; + /* the code is for bitreverse streams */ + if (hdlc->do_bitreverse == 0) + *dst++ = bitrev8(hdlc->ffvalue); + else + *dst++ = hdlc->ffvalue; len++; dsize--; break; @@ -594,7 +586,11 @@ int isdnhdlc_encode(struct isdnhdlc_vars *hdlc, const u8 *src, u16 slen, hdlc->cbin = 0x7e; hdlc->state = HDLC_SEND_FIRST_FLAG; } else { - *dst++ = hdlc->cbin; + /* the code is for bitreverse streams */ + if (hdlc->do_bitreverse == 0) + *dst++ = bitrev8(hdlc->cbin); + else + *dst++ = hdlc->cbin; hdlc->bit_shift = 0; hdlc->data_bits = 0; len++; @@ -612,7 +608,11 @@ int isdnhdlc_encode(struct isdnhdlc_vars *hdlc, const u8 *src, u16 slen, } } if (hdlc->data_bits == 8) { - *dst++ = hdlc->cbin; + /* the code is for bitreverse streams */ + if (hdlc->do_bitreverse == 0) + *dst++ = bitrev8(hdlc->cbin); + else + *dst++ = hdlc->cbin; hdlc->data_bits = 0; len++; dsize--; diff --git a/include/linux/isdn/hdlc.h b/include/linux/isdn/hdlc.h index 8f3540c7f692..4b3ecc40889a 100644 --- a/include/linux/isdn/hdlc.h +++ b/include/linux/isdn/hdlc.h @@ -6,6 +6,7 @@ * controllers. * * Copyright (C) + * 2009 Karsten Keil * 2002 Wolfgang Mües * 2001 Frode Isaksen * 2001 Kai Germaschewski @@ -50,8 +51,14 @@ struct isdnhdlc_vars { u32 do_adapt56:1; /* set if in closing phase (need to send CRC + flag) */ u32 do_closing:1; + /* set if data is bitreverse */ + u32 do_bitreverse:1; }; +/* Feature Flags */ +#define HDLC_56KBIT 0x01 +#define HDLC_DCHANNEL 0x02 +#define HDLC_BITREVERSE 0x04 /* The return value from isdnhdlc_decode is @@ -62,13 +69,12 @@ struct isdnhdlc_vars { #define HDLC_CRC_ERROR 2 #define HDLC_LENGTH_ERROR 3 -extern void isdnhdlc_rcv_init(struct isdnhdlc_vars *hdlc, int do_adapt56); +extern void isdnhdlc_rcv_init(struct isdnhdlc_vars *hdlc, u32 features); extern int isdnhdlc_decode(struct isdnhdlc_vars *hdlc, const u8 *src, int slen, int *count, u8 *dst, int dsize); -extern void isdnhdlc_out_init(struct isdnhdlc_vars *hdlc, int is_d_channel, - int do_adapt56); +extern void isdnhdlc_out_init(struct isdnhdlc_vars *hdlc, u32 features); extern int isdnhdlc_encode(struct isdnhdlc_vars *hdlc, const u8 *src, u16 slen, int *count, u8 *dst, int dsize); -- cgit v1.2.3 From fb286f0471a04ef646c8e5c79750ae6718183745 Mon Sep 17 00:00:00 2001 From: Karsten Keil Date: Thu, 9 Jul 2009 10:02:29 +0200 Subject: mISDN: Make clearing B-channel a common function Clearing B-channel is needed in every driver, so it makes sense to have it as common function. Signed-off-by: Karsten Keil --- drivers/isdn/hardware/mISDN/hfcmulti.c | 16 +--------------- drivers/isdn/hardware/mISDN/hfcpci.c | 16 +--------------- drivers/isdn/hardware/mISDN/hfcsusb.c | 16 +--------------- drivers/isdn/mISDN/hwchannel.c | 15 +++++++++++++-- include/linux/mISDNhw.h | 1 + 5 files changed, 17 insertions(+), 47 deletions(-) (limited to 'include') diff --git a/drivers/isdn/hardware/mISDN/hfcmulti.c b/drivers/isdn/hardware/mISDN/hfcmulti.c index e1dab30aed30..fd77bb15d790 100644 --- a/drivers/isdn/hardware/mISDN/hfcmulti.c +++ b/drivers/isdn/hardware/mISDN/hfcmulti.c @@ -3416,22 +3416,8 @@ deactivate_bchannel(struct bchannel *bch) u_long flags; spin_lock_irqsave(&hc->lock, flags); - if (test_and_clear_bit(FLG_TX_NEXT, &bch->Flags)) { - dev_kfree_skb(bch->next_skb); - bch->next_skb = NULL; - } - if (bch->tx_skb) { - dev_kfree_skb(bch->tx_skb); - bch->tx_skb = NULL; - } - bch->tx_idx = 0; - if (bch->rx_skb) { - dev_kfree_skb(bch->rx_skb); - bch->rx_skb = NULL; - } + mISDN_clear_bchannel(bch); hc->chan[bch->slot].coeff_count = 0; - test_and_clear_bit(FLG_ACTIVE, &bch->Flags); - test_and_clear_bit(FLG_TX_BUSY, &bch->Flags); hc->chan[bch->slot].rx_off = 0; hc->chan[bch->slot].conf = -1; mode_hfcmulti(hc, bch->slot, ISDN_P_NONE, -1, 0, -1, 0); diff --git a/drivers/isdn/hardware/mISDN/hfcpci.c b/drivers/isdn/hardware/mISDN/hfcpci.c index 228ffbed1286..70e6b0e01121 100644 --- a/drivers/isdn/hardware/mISDN/hfcpci.c +++ b/drivers/isdn/hardware/mISDN/hfcpci.c @@ -1522,22 +1522,8 @@ deactivate_bchannel(struct bchannel *bch) u_long flags; spin_lock_irqsave(&hc->lock, flags); - if (test_and_clear_bit(FLG_TX_NEXT, &bch->Flags)) { - dev_kfree_skb(bch->next_skb); - bch->next_skb = NULL; - } - if (bch->tx_skb) { - dev_kfree_skb(bch->tx_skb); - bch->tx_skb = NULL; - } - bch->tx_idx = 0; - if (bch->rx_skb) { - dev_kfree_skb(bch->rx_skb); - bch->rx_skb = NULL; - } + mISDN_clear_bchannel(bch); mode_hfcpci(bch, bch->nr, ISDN_P_NONE); - test_and_clear_bit(FLG_ACTIVE, &bch->Flags); - test_and_clear_bit(FLG_TX_BUSY, &bch->Flags); spin_unlock_irqrestore(&hc->lock, flags); } diff --git a/drivers/isdn/hardware/mISDN/hfcsusb.c b/drivers/isdn/hardware/mISDN/hfcsusb.c index 6b7704c41b94..fc46a26cb14f 100644 --- a/drivers/isdn/hardware/mISDN/hfcsusb.c +++ b/drivers/isdn/hardware/mISDN/hfcsusb.c @@ -1809,21 +1809,7 @@ deactivate_bchannel(struct bchannel *bch) hw->name, __func__, bch->nr); spin_lock_irqsave(&hw->lock, flags); - if (test_and_clear_bit(FLG_TX_NEXT, &bch->Flags)) { - dev_kfree_skb(bch->next_skb); - bch->next_skb = NULL; - } - if (bch->tx_skb) { - dev_kfree_skb(bch->tx_skb); - bch->tx_skb = NULL; - } - bch->tx_idx = 0; - if (bch->rx_skb) { - dev_kfree_skb(bch->rx_skb); - bch->rx_skb = NULL; - } - clear_bit(FLG_ACTIVE, &bch->Flags); - clear_bit(FLG_TX_BUSY, &bch->Flags); + mISDN_clear_bchannel(bch); spin_unlock_irqrestore(&hw->lock, flags); hfcsusb_setup_bch(bch, ISDN_P_NONE); hfcsusb_stop_endpoint(hw, bch->nr); diff --git a/drivers/isdn/mISDN/hwchannel.c b/drivers/isdn/mISDN/hwchannel.c index 0481a0cdf6db..e8049be552aa 100644 --- a/drivers/isdn/mISDN/hwchannel.c +++ b/drivers/isdn/mISDN/hwchannel.c @@ -114,13 +114,14 @@ mISDN_freedchannel(struct dchannel *ch) } EXPORT_SYMBOL(mISDN_freedchannel); -int -mISDN_freebchannel(struct bchannel *ch) +void +mISDN_clear_bchannel(struct bchannel *ch) { if (ch->tx_skb) { dev_kfree_skb(ch->tx_skb); ch->tx_skb = NULL; } + ch->tx_idx = 0; if (ch->rx_skb) { dev_kfree_skb(ch->rx_skb); ch->rx_skb = NULL; @@ -129,6 +130,16 @@ mISDN_freebchannel(struct bchannel *ch) dev_kfree_skb(ch->next_skb); ch->next_skb = NULL; } + test_and_clear_bit(FLG_TX_BUSY, &ch->Flags); + test_and_clear_bit(FLG_TX_NEXT, &ch->Flags); + test_and_clear_bit(FLG_ACTIVE, &ch->Flags); +} +EXPORT_SYMBOL(mISDN_clear_bchannel); + +int +mISDN_freebchannel(struct bchannel *ch) +{ + mISDN_clear_bchannel(ch); skb_queue_purge(&ch->rqueue); ch->rcount = 0; flush_scheduled_work(); diff --git a/include/linux/mISDNhw.h b/include/linux/mISDNhw.h index 7f9831da847f..4af841408fb5 100644 --- a/include/linux/mISDNhw.h +++ b/include/linux/mISDNhw.h @@ -168,6 +168,7 @@ struct bchannel { extern int mISDN_initdchannel(struct dchannel *, int, void *); extern int mISDN_initbchannel(struct bchannel *, int); extern int mISDN_freedchannel(struct dchannel *); +extern void mISDN_clear_bchannel(struct bchannel *); extern int mISDN_freebchannel(struct bchannel *); extern void queue_ch_frame(struct mISDNchannel *, u_int, int, struct sk_buff *); -- cgit v1.2.3 From da2272c91ae81b41ae6fa6ebdc767a6cef73b770 Mon Sep 17 00:00:00 2001 From: Karsten Keil Date: Wed, 22 Jul 2009 20:01:59 +0200 Subject: mISDN: Add support for Speedfax+ cards Add support for the Siemens ISAR DSP chip and cards based on it, including analog modem protocols. Signed-off-by: Karsten Keil --- drivers/isdn/hardware/mISDN/Kconfig | 13 + drivers/isdn/hardware/mISDN/Makefile | 2 + drivers/isdn/hardware/mISDN/isar.h | 269 +++++ drivers/isdn/hardware/mISDN/mISDNisar.c | 1726 +++++++++++++++++++++++++++++++ drivers/isdn/hardware/mISDN/speedfax.c | 526 ++++++++++ include/linux/mISDNif.h | 16 +- 6 files changed, 2551 insertions(+), 1 deletion(-) create mode 100644 drivers/isdn/hardware/mISDN/isar.h create mode 100644 drivers/isdn/hardware/mISDN/mISDNisar.c create mode 100644 drivers/isdn/hardware/mISDN/speedfax.c (limited to 'include') diff --git a/drivers/isdn/hardware/mISDN/Kconfig b/drivers/isdn/hardware/mISDN/Kconfig index 212dee6adec6..2600534be07e 100644 --- a/drivers/isdn/hardware/mISDN/Kconfig +++ b/drivers/isdn/hardware/mISDN/Kconfig @@ -47,6 +47,15 @@ config MISDN_AVMFRITZ help Enable support for AVMs FRITZ!CARD PCI cards +config MISDN_SPEEDFAX + tristate "Support for Sedlbauer Speedfax+" + depends on MISDN + depends on PCI + select MISDN_IPAC + select MISDN_ISAR + help + Enable support for Sedlbauer Speedfax+. + config MISDN_INFINEON tristate "Support for cards with Infineon chipset" depends on MISDN @@ -61,3 +70,7 @@ config MISDN_IPAC tristate depends on MISDN +config MISDN_ISAR + tristate + depends on MISDN + diff --git a/drivers/isdn/hardware/mISDN/Makefile b/drivers/isdn/hardware/mISDN/Makefile index 8204d84af70a..e18f964e185b 100644 --- a/drivers/isdn/hardware/mISDN/Makefile +++ b/drivers/isdn/hardware/mISDN/Makefile @@ -7,6 +7,8 @@ obj-$(CONFIG_MISDN_HFCPCI) += hfcpci.o obj-$(CONFIG_MISDN_HFCMULTI) += hfcmulti.o obj-$(CONFIG_MISDN_HFCUSB) += hfcsusb.o obj-$(CONFIG_MISDN_AVMFRITZ) += avmfritz.o +obj-$(CONFIG_MISDN_SPEEDFAX) += speedfax.o obj-$(CONFIG_MISDN_INFINEON) += mISDNinfineon.o # chip modules obj-$(CONFIG_MISDN_IPAC) += mISDNipac.o +obj-$(CONFIG_MISDN_ISAR) += mISDNisar.o diff --git a/drivers/isdn/hardware/mISDN/isar.h b/drivers/isdn/hardware/mISDN/isar.h new file mode 100644 index 000000000000..092351acd2ab --- /dev/null +++ b/drivers/isdn/hardware/mISDN/isar.h @@ -0,0 +1,269 @@ +/* + * + * isar.h ISAR (Siemens PSB 7110) specific defines + * + * Author Karsten Keil (keil@isdn4linux.de) + * + * Copyright 2009 by Karsten Keil + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. + * + */ + +#include "iohelper.h" + +struct isar_hw; + +struct isar_ch { + struct bchannel bch; + struct isar_hw *is; + struct timer_list ftimer; + u8 nr; + u8 dpath; + u8 mml; + u8 state; + u8 cmd; + u8 mod; + u8 newcmd; + u8 newmod; + u8 try_mod; + u8 conmsg[16]; +}; + +struct isar_hw { + struct isar_ch ch[2]; + void *hw; + spinlock_t *hwlock; /* lock HW acccess */ + char *name; + struct module *owner; + read_reg_t *read_reg; + write_reg_t *write_reg; + fifo_func_t *read_fifo; + fifo_func_t *write_fifo; + int (*ctrl)(void *, u32, u_long); + void (*release)(struct isar_hw *); + int (*init)(struct isar_hw *); + int (*open)(struct isar_hw *, struct channel_req *); + int (*firmware)(struct isar_hw *, const u8 *, int); + unsigned long Flags; + int version; + u8 bstat; + u8 iis; + u8 cmsb; + u8 clsb; + u8 buf[256]; + u8 log[256]; +}; + +#define ISAR_IRQMSK 0x04 +#define ISAR_IRQSTA 0x04 +#define ISAR_IRQBIT 0x75 +#define ISAR_CTRL_H 0x61 +#define ISAR_CTRL_L 0x60 +#define ISAR_IIS 0x58 +#define ISAR_IIA 0x58 +#define ISAR_HIS 0x50 +#define ISAR_HIA 0x50 +#define ISAR_MBOX 0x4c +#define ISAR_WADR 0x4a +#define ISAR_RADR 0x48 + +#define ISAR_HIS_VNR 0x14 +#define ISAR_HIS_DKEY 0x02 +#define ISAR_HIS_FIRM 0x1e +#define ISAR_HIS_STDSP 0x08 +#define ISAR_HIS_DIAG 0x05 +#define ISAR_HIS_P0CFG 0x3c +#define ISAR_HIS_P12CFG 0x24 +#define ISAR_HIS_SARTCFG 0x25 +#define ISAR_HIS_PUMPCFG 0x26 +#define ISAR_HIS_PUMPCTRL 0x2a +#define ISAR_HIS_IOM2CFG 0x27 +#define ISAR_HIS_IOM2REQ 0x07 +#define ISAR_HIS_IOM2CTRL 0x2b +#define ISAR_HIS_BSTREQ 0x0c +#define ISAR_HIS_PSTREQ 0x0e +#define ISAR_HIS_SDATA 0x20 +#define ISAR_HIS_DPS1 0x40 +#define ISAR_HIS_DPS2 0x80 +#define SET_DPS(x) ((x<<6) & 0xc0) + +#define ISAR_IIS_MSCMSD 0x3f +#define ISAR_IIS_VNR 0x15 +#define ISAR_IIS_DKEY 0x03 +#define ISAR_IIS_FIRM 0x1f +#define ISAR_IIS_STDSP 0x09 +#define ISAR_IIS_DIAG 0x25 +#define ISAR_IIS_GSTEV 0x00 +#define ISAR_IIS_BSTEV 0x28 +#define ISAR_IIS_BSTRSP 0x2c +#define ISAR_IIS_PSTRSP 0x2e +#define ISAR_IIS_PSTEV 0x2a +#define ISAR_IIS_IOM2RSP 0x27 +#define ISAR_IIS_RDATA 0x20 +#define ISAR_IIS_INVMSG 0x3f + +#define ISAR_CTRL_SWVER 0x10 +#define ISAR_CTRL_STST 0x40 + +#define ISAR_MSG_HWVER 0x20 + +#define ISAR_DP1_USE 1 +#define ISAR_DP2_USE 2 +#define ISAR_RATE_REQ 3 + +#define PMOD_DISABLE 0 +#define PMOD_FAX 1 +#define PMOD_DATAMODEM 2 +#define PMOD_HALFDUPLEX 3 +#define PMOD_V110 4 +#define PMOD_DTMF 5 +#define PMOD_DTMF_TRANS 6 +#define PMOD_BYPASS 7 + +#define PCTRL_ORIG 0x80 +#define PV32P2_V23R 0x40 +#define PV32P2_V22A 0x20 +#define PV32P2_V22B 0x10 +#define PV32P2_V22C 0x08 +#define PV32P2_V21 0x02 +#define PV32P2_BEL 0x01 + +/* LSB MSB in ISAR doc wrong !!! Arghhh */ +#define PV32P3_AMOD 0x80 +#define PV32P3_V32B 0x02 +#define PV32P3_V23B 0x01 +#define PV32P4_48 0x11 +#define PV32P5_48 0x05 +#define PV32P4_UT48 0x11 +#define PV32P5_UT48 0x0d +#define PV32P4_96 0x11 +#define PV32P5_96 0x03 +#define PV32P4_UT96 0x11 +#define PV32P5_UT96 0x0f +#define PV32P4_B96 0x91 +#define PV32P5_B96 0x0b +#define PV32P4_UTB96 0xd1 +#define PV32P5_UTB96 0x0f +#define PV32P4_120 0xb1 +#define PV32P5_120 0x09 +#define PV32P4_UT120 0xf1 +#define PV32P5_UT120 0x0f +#define PV32P4_144 0x99 +#define PV32P5_144 0x09 +#define PV32P4_UT144 0xf9 +#define PV32P5_UT144 0x0f +#define PV32P6_CTN 0x01 +#define PV32P6_ATN 0x02 + +#define PFAXP2_CTN 0x01 +#define PFAXP2_ATN 0x04 + +#define PSEV_10MS_TIMER 0x02 +#define PSEV_CON_ON 0x18 +#define PSEV_CON_OFF 0x19 +#define PSEV_V24_OFF 0x20 +#define PSEV_CTS_ON 0x21 +#define PSEV_CTS_OFF 0x22 +#define PSEV_DCD_ON 0x23 +#define PSEV_DCD_OFF 0x24 +#define PSEV_DSR_ON 0x25 +#define PSEV_DSR_OFF 0x26 +#define PSEV_REM_RET 0xcc +#define PSEV_REM_REN 0xcd +#define PSEV_GSTN_CLR 0xd4 + +#define PSEV_RSP_READY 0xbc +#define PSEV_LINE_TX_H 0xb3 +#define PSEV_LINE_TX_B 0xb2 +#define PSEV_LINE_RX_H 0xb1 +#define PSEV_LINE_RX_B 0xb0 +#define PSEV_RSP_CONN 0xb5 +#define PSEV_RSP_DISC 0xb7 +#define PSEV_RSP_FCERR 0xb9 +#define PSEV_RSP_SILDET 0xbe +#define PSEV_RSP_SILOFF 0xab +#define PSEV_FLAGS_DET 0xba + +#define PCTRL_CMD_TDTMF 0x5a + +#define PCTRL_CMD_FTH 0xa7 +#define PCTRL_CMD_FRH 0xa5 +#define PCTRL_CMD_FTM 0xa8 +#define PCTRL_CMD_FRM 0xa6 +#define PCTRL_CMD_SILON 0xac +#define PCTRL_CMD_CONT 0xa2 +#define PCTRL_CMD_ESC 0xa4 +#define PCTRL_CMD_SILOFF 0xab +#define PCTRL_CMD_HALT 0xa9 + +#define PCTRL_LOC_RET 0xcf +#define PCTRL_LOC_REN 0xce + +#define SMODE_DISABLE 0 +#define SMODE_V14 2 +#define SMODE_HDLC 3 +#define SMODE_BINARY 4 +#define SMODE_FSK_V14 5 + +#define SCTRL_HDMC_BOTH 0x00 +#define SCTRL_HDMC_DTX 0x80 +#define SCTRL_HDMC_DRX 0x40 +#define S_P1_OVSP 0x40 +#define S_P1_SNP 0x20 +#define S_P1_EOP 0x10 +#define S_P1_EDP 0x08 +#define S_P1_NSB 0x04 +#define S_P1_CHS_8 0x03 +#define S_P1_CHS_7 0x02 +#define S_P1_CHS_6 0x01 +#define S_P1_CHS_5 0x00 + +#define S_P2_BFT_DEF 0x10 + +#define IOM_CTRL_ENA 0x80 +#define IOM_CTRL_NOPCM 0x00 +#define IOM_CTRL_ALAW 0x02 +#define IOM_CTRL_ULAW 0x04 +#define IOM_CTRL_RCV 0x01 + +#define IOM_P1_TXD 0x10 + +#define HDLC_FED 0x40 +#define HDLC_FSD 0x20 +#define HDLC_FST 0x20 +#define HDLC_ERROR 0x1c +#define HDLC_ERR_FAD 0x10 +#define HDLC_ERR_RER 0x08 +#define HDLC_ERR_CER 0x04 +#define SART_NMD 0x01 + +#define BSTAT_RDM0 0x1 +#define BSTAT_RDM1 0x2 +#define BSTAT_RDM2 0x4 +#define BSTAT_RDM3 0x8 +#define BSTEV_TBO 0x1f +#define BSTEV_RBO 0x2f + +/* FAX State Machine */ +#define STFAX_NULL 0 +#define STFAX_READY 1 +#define STFAX_LINE 2 +#define STFAX_CONT 3 +#define STFAX_ACTIV 4 +#define STFAX_ESCAPE 5 +#define STFAX_SILDET 6 + +extern u32 mISDNisar_init(struct isar_hw *, void *); +extern void mISDNisar_irq(struct isar_hw *); diff --git a/drivers/isdn/hardware/mISDN/mISDNisar.c b/drivers/isdn/hardware/mISDN/mISDNisar.c new file mode 100644 index 000000000000..de352a17673a --- /dev/null +++ b/drivers/isdn/hardware/mISDN/mISDNisar.c @@ -0,0 +1,1726 @@ +/* + * mISDNisar.c ISAR (Siemens PSB 7110) specific functions + * + * Author Karsten Keil (keil@isdn4linux.de) + * + * Copyright 2009 by Karsten Keil + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. + * + */ + +/* define this to enable static debug messages, if you kernel supports + * dynamic debugging, you should use debugfs for this + */ +/* #define DEBUG */ + +#include +#include +#include +#include "isar.h" + +#define ISAR_REV "2.1" + +MODULE_AUTHOR("Karsten Keil"); +MODULE_LICENSE("GPL v2"); +MODULE_VERSION(ISAR_REV); + +#define DEBUG_HW_FIRMWARE_FIFO 0x10000 + +static const u8 faxmodulation_s[] = "3,24,48,72,73,74,96,97,98,121,122,145,146"; +static const u8 faxmodulation[] = {3, 24, 48, 72, 73, 74, 96, 97, 98, 121, + 122, 145, 146}; +#define FAXMODCNT 13 + +static void isar_setup(struct isar_hw *); + +static inline int +waitforHIA(struct isar_hw *isar, int timeout) +{ + int t = timeout; + u8 val = isar->read_reg(isar->hw, ISAR_HIA); + + while ((val & 1) && t) { + udelay(1); + t--; + val = isar->read_reg(isar->hw, ISAR_HIA); + } + pr_debug("%s: HIA after %dus\n", isar->name, timeout - t); + return timeout; +} + +/* + * send msg to ISAR mailbox + * if msg is NULL use isar->buf + */ +static int +send_mbox(struct isar_hw *isar, u8 his, u8 creg, u8 len, u8 *msg) +{ + if (!waitforHIA(isar, 1000)) + return 0; + pr_debug("send_mbox(%02x,%02x,%d)\n", his, creg, len); + isar->write_reg(isar->hw, ISAR_CTRL_H, creg); + isar->write_reg(isar->hw, ISAR_CTRL_L, len); + isar->write_reg(isar->hw, ISAR_WADR, 0); + if (!msg) + msg = isar->buf; + if (msg && len) { + isar->write_fifo(isar->hw, ISAR_MBOX, msg, len); + if (isar->ch[0].bch.debug & DEBUG_HW_BFIFO) { + int l = 0; + + while (l < (int)len) { + hex_dump_to_buffer(msg + l, len - l, 32, 1, + isar->log, 256, 1); + pr_debug("%s: %s %02x: %s\n", isar->name, + __func__, l, isar->log); + l += 32; + } + } + } + isar->write_reg(isar->hw, ISAR_HIS, his); + waitforHIA(isar, 1000); + return 1; +} + +/* + * receive message from ISAR mailbox + * if msg is NULL use isar->buf + */ +static void +rcv_mbox(struct isar_hw *isar, u8 *msg) +{ + if (!msg) + msg = isar->buf; + isar->write_reg(isar->hw, ISAR_RADR, 0); + if (msg && isar->clsb) { + isar->read_fifo(isar->hw, ISAR_MBOX, msg, isar->clsb); + if (isar->ch[0].bch.debug & DEBUG_HW_BFIFO) { + int l = 0; + + while (l < (int)isar->clsb) { + hex_dump_to_buffer(msg + l, isar->clsb - l, 32, + 1, isar->log, 256, 1); + pr_debug("%s: %s %02x: %s\n", isar->name, + __func__, l, isar->log); + l += 32; + } + } + } + isar->write_reg(isar->hw, ISAR_IIA, 0); +} + +static inline void +get_irq_infos(struct isar_hw *isar) +{ + isar->iis = isar->read_reg(isar->hw, ISAR_IIS); + isar->cmsb = isar->read_reg(isar->hw, ISAR_CTRL_H); + isar->clsb = isar->read_reg(isar->hw, ISAR_CTRL_L); + pr_debug("%s: rcv_mbox(%02x,%02x,%d)\n", isar->name, + isar->iis, isar->cmsb, isar->clsb); +} + +/* + * poll answer message from ISAR mailbox + * should be used only with ISAR IRQs disabled before DSP was started + * + */ +static int +poll_mbox(struct isar_hw *isar, int maxdelay) +{ + int t = maxdelay; + u8 irq; + + irq = isar->read_reg(isar->hw, ISAR_IRQBIT); + while (t && !(irq & ISAR_IRQSTA)) { + udelay(1); + t--; + } + if (t) { + get_irq_infos(isar); + rcv_mbox(isar, NULL); + } + pr_debug("%s: pulled %d bytes after %d us\n", + isar->name, isar->clsb, maxdelay - t); + return t; +} + +static int +ISARVersion(struct isar_hw *isar) +{ + int ver; + + /* disable ISAR IRQ */ + isar->write_reg(isar->hw, ISAR_IRQBIT, 0); + isar->buf[0] = ISAR_MSG_HWVER; + isar->buf[1] = 0; + isar->buf[2] = 1; + if (!send_mbox(isar, ISAR_HIS_VNR, 0, 3, NULL)) + return -1; + if (!poll_mbox(isar, 1000)) + return -2; + if (isar->iis == ISAR_IIS_VNR) { + if (isar->clsb == 1) { + ver = isar->buf[0] & 0xf; + return ver; + } + return -3; + } + return -4; +} + +static int +load_firmware(struct isar_hw *isar, const u8 *buf, int size) +{ + u32 saved_debug = isar->ch[0].bch.debug; + int ret, cnt; + u8 nom, noc; + u16 left, val, *sp = (u16 *)buf; + u8 *mp; + u_long flags; + + struct { + u16 sadr; + u16 len; + u16 d_key; + } blk_head; + + if (1 != isar->version) { + pr_err("%s: ISAR wrong version %d firmware download aborted\n", + isar->name, isar->version); + return -EINVAL; + } + if (!(saved_debug & DEBUG_HW_FIRMWARE_FIFO)) + isar->ch[0].bch.debug &= ~DEBUG_HW_BFIFO; + pr_debug("%s: load firmware %d words (%d bytes)\n", + isar->name, size/2, size); + cnt = 0; + size /= 2; + /* disable ISAR IRQ */ + spin_lock_irqsave(isar->hwlock, flags); + isar->write_reg(isar->hw, ISAR_IRQBIT, 0); + spin_unlock_irqrestore(isar->hwlock, flags); + while (cnt < size) { + blk_head.sadr = le16_to_cpu(*sp++); + blk_head.len = le16_to_cpu(*sp++); + blk_head.d_key = le16_to_cpu(*sp++); + cnt += 3; + pr_debug("ISAR firmware block (%#x,%d,%#x)\n", + blk_head.sadr, blk_head.len, blk_head.d_key & 0xff); + left = blk_head.len; + if (cnt + left > size) { + pr_info("%s: firmware error have %d need %d words\n", + isar->name, size, cnt + left); + ret = -EINVAL; + goto reterrflg; + } + spin_lock_irqsave(isar->hwlock, flags); + if (!send_mbox(isar, ISAR_HIS_DKEY, blk_head.d_key & 0xff, + 0, NULL)) { + pr_info("ISAR send_mbox dkey failed\n"); + ret = -ETIME; + goto reterror; + } + if (!poll_mbox(isar, 1000)) { + pr_warning("ISAR poll_mbox dkey failed\n"); + ret = -ETIME; + goto reterror; + } + spin_unlock_irqrestore(isar->hwlock, flags); + if ((isar->iis != ISAR_IIS_DKEY) || isar->cmsb || isar->clsb) { + pr_info("ISAR wrong dkey response (%x,%x,%x)\n", + isar->iis, isar->cmsb, isar->clsb); + ret = 1; + goto reterrflg; + } + while (left > 0) { + if (left > 126) + noc = 126; + else + noc = left; + nom = (2 * noc) + 3; + mp = isar->buf; + /* the ISAR is big endian */ + *mp++ = blk_head.sadr >> 8; + *mp++ = blk_head.sadr & 0xFF; + left -= noc; + cnt += noc; + *mp++ = noc; + pr_debug("%s: load %3d words at %04x\n", isar->name, + noc, blk_head.sadr); + blk_head.sadr += noc; + while (noc) { + val = le16_to_cpu(*sp++); + *mp++ = val >> 8; + *mp++ = val & 0xFF;; + noc--; + } + spin_lock_irqsave(isar->hwlock, flags); + if (!send_mbox(isar, ISAR_HIS_FIRM, 0, nom, NULL)) { + pr_info("ISAR send_mbox prog failed\n"); + ret = -ETIME; + goto reterror; + } + if (!poll_mbox(isar, 1000)) { + pr_info("ISAR poll_mbox prog failed\n"); + ret = -ETIME; + goto reterror; + } + spin_unlock_irqrestore(isar->hwlock, flags); + if ((isar->iis != ISAR_IIS_FIRM) || + isar->cmsb || isar->clsb) { + pr_info("ISAR wrong prog response (%x,%x,%x)\n", + isar->iis, isar->cmsb, isar->clsb); + ret = -EIO; + goto reterrflg; + } + } + pr_debug("%s: ISAR firmware block %d words loaded\n", + isar->name, blk_head.len); + } + isar->ch[0].bch.debug = saved_debug; + /* 10ms delay */ + cnt = 10; + while (cnt--) + mdelay(1); + isar->buf[0] = 0xff; + isar->buf[1] = 0xfe; + isar->bstat = 0; + spin_lock_irqsave(isar->hwlock, flags); + if (!send_mbox(isar, ISAR_HIS_STDSP, 0, 2, NULL)) { + pr_info("ISAR send_mbox start dsp failed\n"); + ret = -ETIME; + goto reterror; + } + if (!poll_mbox(isar, 1000)) { + pr_info("ISAR poll_mbox start dsp failed\n"); + ret = -ETIME; + goto reterror; + } + if ((isar->iis != ISAR_IIS_STDSP) || isar->cmsb || isar->clsb) { + pr_info("ISAR wrong start dsp response (%x,%x,%x)\n", + isar->iis, isar->cmsb, isar->clsb); + ret = -EIO; + goto reterror; + } else + pr_debug("%s: ISAR start dsp success\n", isar->name); + + /* NORMAL mode entered */ + /* Enable IRQs of ISAR */ + isar->write_reg(isar->hw, ISAR_IRQBIT, ISAR_IRQSTA); + spin_unlock_irqrestore(isar->hwlock, flags); + cnt = 1000; /* max 1s */ + while ((!isar->bstat) && cnt) { + mdelay(1); + cnt--; + } + if (!cnt) { + pr_info("ISAR no general status event received\n"); + ret = -ETIME; + goto reterrflg; + } else + pr_debug("%s: ISAR general status event %x\n", + isar->name, isar->bstat); + /* 10ms delay */ + cnt = 10; + while (cnt--) + mdelay(1); + isar->iis = 0; + spin_lock_irqsave(isar->hwlock, flags); + if (!send_mbox(isar, ISAR_HIS_DIAG, ISAR_CTRL_STST, 0, NULL)) { + pr_info("ISAR send_mbox self tst failed\n"); + ret = -ETIME; + goto reterror; + } + spin_unlock_irqrestore(isar->hwlock, flags); + cnt = 10000; /* max 100 ms */ + while ((isar->iis != ISAR_IIS_DIAG) && cnt) { + udelay(10); + cnt--; + } + mdelay(1); + if (!cnt) { + pr_info("ISAR no self tst response\n"); + ret = -ETIME; + goto reterrflg; + } + if ((isar->cmsb == ISAR_CTRL_STST) && (isar->clsb == 1) + && (isar->buf[0] == 0)) + pr_debug("%s: ISAR selftest OK\n", isar->name); + else { + pr_info("ISAR selftest not OK %x/%x/%x\n", + isar->cmsb, isar->clsb, isar->buf[0]); + ret = -EIO; + goto reterrflg; + } + spin_lock_irqsave(isar->hwlock, flags); + isar->iis = 0; + if (!send_mbox(isar, ISAR_HIS_DIAG, ISAR_CTRL_SWVER, 0, NULL)) { + pr_info("ISAR RQST SVN failed\n"); + ret = -ETIME; + goto reterror; + } + spin_unlock_irqrestore(isar->hwlock, flags); + cnt = 30000; /* max 300 ms */ + while ((isar->iis != ISAR_IIS_DIAG) && cnt) { + udelay(10); + cnt--; + } + mdelay(1); + if (!cnt) { + pr_info("ISAR no SVN response\n"); + ret = -ETIME; + goto reterrflg; + } else { + if ((isar->cmsb == ISAR_CTRL_SWVER) && (isar->clsb == 1)) { + pr_notice("%s: ISAR software version %#x\n", + isar->name, isar->buf[0]); + } else { + pr_info("%s: ISAR wrong swver response (%x,%x)" + " cnt(%d)\n", isar->name, isar->cmsb, + isar->clsb, cnt); + ret = -EIO; + goto reterrflg; + } + } + spin_lock_irqsave(isar->hwlock, flags); + isar_setup(isar); + spin_unlock_irqrestore(isar->hwlock, flags); + ret = 0; +reterrflg: + spin_lock_irqsave(isar->hwlock, flags); +reterror: + isar->ch[0].bch.debug = saved_debug; + if (ret) + /* disable ISAR IRQ */ + isar->write_reg(isar->hw, ISAR_IRQBIT, 0); + spin_unlock_irqrestore(isar->hwlock, flags); + return ret; +} + +static inline void +deliver_status(struct isar_ch *ch, int status) +{ + pr_debug("%s: HL->LL FAXIND %x\n", ch->is->name, status); + _queue_data(&ch->bch.ch, PH_CONTROL_IND, status, 0, NULL, GFP_ATOMIC); +} + +static inline void +isar_rcv_frame(struct isar_ch *ch) +{ + u8 *ptr; + + if (!ch->is->clsb) { + pr_debug("%s; ISAR zero len frame\n", ch->is->name); + ch->is->write_reg(ch->is->hw, ISAR_IIA, 0); + return; + } + switch (ch->bch.state) { + case ISDN_P_NONE: + pr_debug("%s: ISAR protocol 0 spurious IIS_RDATA %x/%x/%x\n", + ch->is->name, ch->is->iis, ch->is->cmsb, ch->is->clsb); + ch->is->write_reg(ch->is->hw, ISAR_IIA, 0); + break; + case ISDN_P_B_RAW: + case ISDN_P_B_L2DTMF: + case ISDN_P_B_MODEM_ASYNC: + if (!ch->bch.rx_skb) { + ch->bch.rx_skb = mI_alloc_skb(ch->bch.maxlen, + GFP_ATOMIC); + if (unlikely(!ch->bch.rx_skb)) { + pr_info("%s: B receive out of memory\n", + ch->is->name); + ch->is->write_reg(ch->is->hw, ISAR_IIA, 0); + break; + } + } + rcv_mbox(ch->is, skb_put(ch->bch.rx_skb, ch->is->clsb)); + recv_Bchannel(&ch->bch, 0); + break; + case ISDN_P_B_HDLC: + if (!ch->bch.rx_skb) { + ch->bch.rx_skb = mI_alloc_skb(ch->bch.maxlen, + GFP_ATOMIC); + if (unlikely(!ch->bch.rx_skb)) { + pr_info("%s: B receive out of memory\n", + ch->is->name); + ch->is->write_reg(ch->is->hw, ISAR_IIA, 0); + break; + } + } + if ((ch->bch.rx_skb->len + ch->is->clsb) > + (ch->bch.maxlen + 2)) { + pr_debug("%s: incoming packet too large\n", + ch->is->name); + ch->is->write_reg(ch->is->hw, ISAR_IIA, 0); + skb_trim(ch->bch.rx_skb, 0); + break; + } + if (ch->is->cmsb & HDLC_ERROR) { + pr_debug("%s: ISAR frame error %x len %d\n", + ch->is->name, ch->is->cmsb, ch->is->clsb); +#ifdef ERROR_STATISTIC + if (ch->is->cmsb & HDLC_ERR_RER) + ch->bch.err_inv++; + if (ch->is->cmsb & HDLC_ERR_CER) + ch->bch.err_crc++; +#endif + skb_trim(ch->bch.rx_skb, 0); + ch->is->write_reg(ch->is->hw, ISAR_IIA, 0); + break; + } + if (ch->is->cmsb & HDLC_FSD) + skb_trim(ch->bch.rx_skb, 0); + ptr = skb_put(ch->bch.rx_skb, ch->is->clsb); + rcv_mbox(ch->is, ptr); + if (ch->is->cmsb & HDLC_FED) { + if (ch->bch.rx_skb->len < 3) { /* last 2 are the FCS */ + pr_debug("%s: ISAR frame to short %d\n", + ch->is->name, ch->bch.rx_skb->len); + skb_trim(ch->bch.rx_skb, 0); + break; + } + skb_trim(ch->bch.rx_skb, ch->bch.rx_skb->len - 2); + recv_Bchannel(&ch->bch, 0); + } + break; + case ISDN_P_B_T30_FAX: + if (ch->state != STFAX_ACTIV) { + pr_debug("%s: isar_rcv_frame: not ACTIV\n", + ch->is->name); + ch->is->write_reg(ch->is->hw, ISAR_IIA, 0); + if (ch->bch.rx_skb) + skb_trim(ch->bch.rx_skb, 0); + break; + } + if (!ch->bch.rx_skb) { + ch->bch.rx_skb = mI_alloc_skb(ch->bch.maxlen, + GFP_ATOMIC); + if (unlikely(!ch->bch.rx_skb)) { + pr_info("%s: B receive out of memory\n", + __func__); + ch->is->write_reg(ch->is->hw, ISAR_IIA, 0); + break; + } + } + if (ch->cmd == PCTRL_CMD_FRM) { + rcv_mbox(ch->is, skb_put(ch->bch.rx_skb, ch->is->clsb)); + pr_debug("%s: isar_rcv_frame: %d\n", + ch->is->name, ch->bch.rx_skb->len); + if (ch->is->cmsb & SART_NMD) { /* ABORT */ + pr_debug("%s: isar_rcv_frame: no more data\n", + ch->is->name); + ch->is->write_reg(ch->is->hw, ISAR_IIA, 0); + send_mbox(ch->is, SET_DPS(ch->dpath) | + ISAR_HIS_PUMPCTRL, PCTRL_CMD_ESC, + 0, NULL); + ch->state = STFAX_ESCAPE; + /* set_skb_flag(skb, DF_NOMOREDATA); */ + } + recv_Bchannel(&ch->bch, 0); + if (ch->is->cmsb & SART_NMD) + deliver_status(ch, HW_MOD_NOCARR); + break; + } + if (ch->cmd != PCTRL_CMD_FRH) { + pr_debug("%s: isar_rcv_frame: unknown fax mode %x\n", + ch->is->name, ch->cmd); + ch->is->write_reg(ch->is->hw, ISAR_IIA, 0); + if (ch->bch.rx_skb) + skb_trim(ch->bch.rx_skb, 0); + break; + } + /* PCTRL_CMD_FRH */ + if ((ch->bch.rx_skb->len + ch->is->clsb) > + (ch->bch.maxlen + 2)) { + pr_info("%s: %s incoming packet too large\n", + ch->is->name, __func__); + ch->is->write_reg(ch->is->hw, ISAR_IIA, 0); + skb_trim(ch->bch.rx_skb, 0); + break; + } else if (ch->is->cmsb & HDLC_ERROR) { + pr_info("%s: ISAR frame error %x len %d\n", + ch->is->name, ch->is->cmsb, ch->is->clsb); + skb_trim(ch->bch.rx_skb, 0); + ch->is->write_reg(ch->is->hw, ISAR_IIA, 0); + break; + } + if (ch->is->cmsb & HDLC_FSD) + skb_trim(ch->bch.rx_skb, 0); + ptr = skb_put(ch->bch.rx_skb, ch->is->clsb); + rcv_mbox(ch->is, ptr); + if (ch->is->cmsb & HDLC_FED) { + if (ch->bch.rx_skb->len < 3) { /* last 2 are the FCS */ + pr_info("%s: ISAR frame to short %d\n", + ch->is->name, ch->bch.rx_skb->len); + skb_trim(ch->bch.rx_skb, 0); + break; + } + skb_trim(ch->bch.rx_skb, ch->bch.rx_skb->len - 2); + recv_Bchannel(&ch->bch, 0); + } + if (ch->is->cmsb & SART_NMD) { /* ABORT */ + pr_debug("%s: isar_rcv_frame: no more data\n", + ch->is->name); + ch->is->write_reg(ch->is->hw, ISAR_IIA, 0); + if (ch->bch.rx_skb) + skb_trim(ch->bch.rx_skb, 0); + send_mbox(ch->is, SET_DPS(ch->dpath) | + ISAR_HIS_PUMPCTRL, PCTRL_CMD_ESC, 0, NULL); + ch->state = STFAX_ESCAPE; + deliver_status(ch, HW_MOD_NOCARR); + } + break; + default: + pr_info("isar_rcv_frame protocol (%x)error\n", ch->bch.state); + ch->is->write_reg(ch->is->hw, ISAR_IIA, 0); + break; + } +} + +static void +isar_fill_fifo(struct isar_ch *ch) +{ + int count; + u8 msb; + u8 *ptr; + + pr_debug("%s: ch%d tx_skb %p tx_idx %d\n", + ch->is->name, ch->bch.nr, ch->bch.tx_skb, ch->bch.tx_idx); + if (!ch->bch.tx_skb) + return; + count = ch->bch.tx_skb->len - ch->bch.tx_idx; + if (count <= 0) + return; + if (!(ch->is->bstat & + (ch->dpath == 1 ? BSTAT_RDM1 : BSTAT_RDM2))) + return; + if (count > ch->mml) { + msb = 0; + count = ch->mml; + } else { + msb = HDLC_FED; + } + ptr = ch->bch.tx_skb->data + ch->bch.tx_idx; + if (!ch->bch.tx_idx) { + pr_debug("%s: frame start\n", ch->is->name); + if ((ch->bch.state == ISDN_P_B_T30_FAX) && + (ch->cmd == PCTRL_CMD_FTH)) { + if (count > 1) { + if ((ptr[0] == 0xff) && (ptr[1] == 0x13)) { + /* last frame */ + test_and_set_bit(FLG_LASTDATA, + &ch->bch.Flags); + pr_debug("%s: set LASTDATA\n", + ch->is->name); + if (msb == HDLC_FED) + test_and_set_bit(FLG_DLEETX, + &ch->bch.Flags); + } + } + } + msb |= HDLC_FST; + } + ch->bch.tx_idx += count; + switch (ch->bch.state) { + case ISDN_P_NONE: + pr_info("%s: wrong protocol 0\n", __func__); + break; + case ISDN_P_B_RAW: + case ISDN_P_B_L2DTMF: + case ISDN_P_B_MODEM_ASYNC: + send_mbox(ch->is, SET_DPS(ch->dpath) | ISAR_HIS_SDATA, + 0, count, ptr); + break; + case ISDN_P_B_HDLC: + send_mbox(ch->is, SET_DPS(ch->dpath) | ISAR_HIS_SDATA, + msb, count, ptr); + break; + case ISDN_P_B_T30_FAX: + if (ch->state != STFAX_ACTIV) + pr_debug("%s: not ACTIV\n", ch->is->name); + else if (ch->cmd == PCTRL_CMD_FTH) + send_mbox(ch->is, SET_DPS(ch->dpath) | ISAR_HIS_SDATA, + msb, count, ptr); + else if (ch->cmd == PCTRL_CMD_FTM) + send_mbox(ch->is, SET_DPS(ch->dpath) | ISAR_HIS_SDATA, + 0, count, ptr); + else + pr_debug("%s: not FTH/FTM\n", ch->is->name); + break; + default: + pr_info("%s: protocol(%x) error\n", + __func__, ch->bch.state); + break; + } +} + +static inline struct isar_ch * +sel_bch_isar(struct isar_hw *isar, u8 dpath) +{ + struct isar_ch *base = &isar->ch[0]; + + if ((!dpath) || (dpath > 2)) + return NULL; + if (base->dpath == dpath) + return base; + base++; + if (base->dpath == dpath) + return base; + return NULL; +} + +static void +send_next(struct isar_ch *ch) +{ + pr_debug("%s: %s ch%d tx_skb %p tx_idx %d\n", + ch->is->name, __func__, ch->bch.nr, + ch->bch.tx_skb, ch->bch.tx_idx); + if (ch->bch.state == ISDN_P_B_T30_FAX) { + if (ch->cmd == PCTRL_CMD_FTH) { + if (test_bit(FLG_LASTDATA, &ch->bch.Flags)) { + pr_debug("set NMD_DATA\n"); + test_and_set_bit(FLG_NMD_DATA, &ch->bch.Flags); + } + } else if (ch->cmd == PCTRL_CMD_FTM) { + if (test_bit(FLG_DLEETX, &ch->bch.Flags)) { + test_and_set_bit(FLG_LASTDATA, &ch->bch.Flags); + test_and_set_bit(FLG_NMD_DATA, &ch->bch.Flags); + } + } + } + if (ch->bch.tx_skb) { + /* send confirm, on trans, free on hdlc. */ + if (test_bit(FLG_TRANSPARENT, &ch->bch.Flags)) + confirm_Bsend(&ch->bch); + dev_kfree_skb(ch->bch.tx_skb); + } + if (get_next_bframe(&ch->bch)) + isar_fill_fifo(ch); + else { + if (test_and_clear_bit(FLG_DLEETX, &ch->bch.Flags)) { + if (test_and_clear_bit(FLG_LASTDATA, + &ch->bch.Flags)) { + if (test_and_clear_bit(FLG_NMD_DATA, + &ch->bch.Flags)) { + u8 zd = 0; + send_mbox(ch->is, SET_DPS(ch->dpath) | + ISAR_HIS_SDATA, 0x01, 1, &zd); + } + test_and_set_bit(FLG_LL_OK, &ch->bch.Flags); + } else { + deliver_status(ch, HW_MOD_CONNECT); + } + } + } +} + +static void +check_send(struct isar_hw *isar, u8 rdm) +{ + struct isar_ch *ch; + + pr_debug("%s: rdm %x\n", isar->name, rdm); + if (rdm & BSTAT_RDM1) { + ch = sel_bch_isar(isar, 1); + if (ch && test_bit(FLG_ACTIVE, &ch->bch.Flags)) { + if (ch->bch.tx_skb && (ch->bch.tx_skb->len > + ch->bch.tx_idx)) + isar_fill_fifo(ch); + else + send_next(ch); + } + } + if (rdm & BSTAT_RDM2) { + ch = sel_bch_isar(isar, 2); + if (ch && test_bit(FLG_ACTIVE, &ch->bch.Flags)) { + if (ch->bch.tx_skb && (ch->bch.tx_skb->len > + ch->bch.tx_idx)) + isar_fill_fifo(ch); + else + send_next(ch); + } + } +} + +const char *dmril[] = {"NO SPEED", "1200/75", "NODEF2", "75/1200", "NODEF4", + "300", "600", "1200", "2400", "4800", "7200", + "9600nt", "9600t", "12000", "14400", "WRONG"}; +const char *dmrim[] = {"NO MOD", "NO DEF", "V32/V32b", "V22", "V21", + "Bell103", "V23", "Bell202", "V17", "V29", "V27ter"}; + +static void +isar_pump_status_rsp(struct isar_ch *ch) { + u8 ril = ch->is->buf[0]; + u8 rim; + + if (!test_and_clear_bit(ISAR_RATE_REQ, &ch->is->Flags)) + return; + if (ril > 14) { + pr_info("%s: wrong pstrsp ril=%d\n", ch->is->name, ril); + ril = 15; + } + switch (ch->is->buf[1]) { + case 0: + rim = 0; + break; + case 0x20: + rim = 2; + break; + case 0x40: + rim = 3; + break; + case 0x41: + rim = 4; + break; + case 0x51: + rim = 5; + break; + case 0x61: + rim = 6; + break; + case 0x71: + rim = 7; + break; + case 0x82: + rim = 8; + break; + case 0x92: + rim = 9; + break; + case 0xa2: + rim = 10; + break; + default: + rim = 1; + break; + } + sprintf(ch->conmsg, "%s %s", dmril[ril], dmrim[rim]); + pr_debug("%s: pump strsp %s\n", ch->is->name, ch->conmsg); +} + +static void +isar_pump_statev_modem(struct isar_ch *ch, u8 devt) { + u8 dps = SET_DPS(ch->dpath); + + switch (devt) { + case PSEV_10MS_TIMER: + pr_debug("%s: pump stev TIMER\n", ch->is->name); + break; + case PSEV_CON_ON: + pr_debug("%s: pump stev CONNECT\n", ch->is->name); + deliver_status(ch, HW_MOD_CONNECT); + break; + case PSEV_CON_OFF: + pr_debug("%s: pump stev NO CONNECT\n", ch->is->name); + send_mbox(ch->is, dps | ISAR_HIS_PSTREQ, 0, 0, NULL); + deliver_status(ch, HW_MOD_NOCARR); + break; + case PSEV_V24_OFF: + pr_debug("%s: pump stev V24 OFF\n", ch->is->name); + break; + case PSEV_CTS_ON: + pr_debug("%s: pump stev CTS ON\n", ch->is->name); + break; + case PSEV_CTS_OFF: + pr_debug("%s pump stev CTS OFF\n", ch->is->name); + break; + case PSEV_DCD_ON: + pr_debug("%s: pump stev CARRIER ON\n", ch->is->name); + test_and_set_bit(ISAR_RATE_REQ, &ch->is->Flags); + send_mbox(ch->is, dps | ISAR_HIS_PSTREQ, 0, 0, NULL); + break; + case PSEV_DCD_OFF: + pr_debug("%s: pump stev CARRIER OFF\n", ch->is->name); + break; + case PSEV_DSR_ON: + pr_debug("%s: pump stev DSR ON\n", ch->is->name); + break; + case PSEV_DSR_OFF: + pr_debug("%s: pump stev DSR_OFF\n", ch->is->name); + break; + case PSEV_REM_RET: + pr_debug("%s: pump stev REMOTE RETRAIN\n", ch->is->name); + break; + case PSEV_REM_REN: + pr_debug("%s: pump stev REMOTE RENEGOTIATE\n", ch->is->name); + break; + case PSEV_GSTN_CLR: + pr_debug("%s: pump stev GSTN CLEAR\n", ch->is->name); + break; + default: + pr_info("u%s: nknown pump stev %x\n", ch->is->name, devt); + break; + } +} + +static void +isar_pump_statev_fax(struct isar_ch *ch, u8 devt) { + u8 dps = SET_DPS(ch->dpath); + u8 p1; + + switch (devt) { + case PSEV_10MS_TIMER: + pr_debug("%s: pump stev TIMER\n", ch->is->name); + break; + case PSEV_RSP_READY: + pr_debug("%s: pump stev RSP_READY\n", ch->is->name); + ch->state = STFAX_READY; + deliver_status(ch, HW_MOD_READY); +#ifdef AUTOCON + if (test_bit(BC_FLG_ORIG, &ch->bch.Flags)) + isar_pump_cmd(bch, HW_MOD_FRH, 3); + else + isar_pump_cmd(bch, HW_MOD_FTH, 3); +#endif + break; + case PSEV_LINE_TX_H: + if (ch->state == STFAX_LINE) { + pr_debug("%s: pump stev LINE_TX_H\n", ch->is->name); + ch->state = STFAX_CONT; + send_mbox(ch->is, dps | ISAR_HIS_PUMPCTRL, + PCTRL_CMD_CONT, 0, NULL); + } else { + pr_debug("%s: pump stev LINE_TX_H wrong st %x\n", + ch->is->name, ch->state); + } + break; + case PSEV_LINE_RX_H: + if (ch->state == STFAX_LINE) { + pr_debug("%s: pump stev LINE_RX_H\n", ch->is->name); + ch->state = STFAX_CONT; + send_mbox(ch->is, dps | ISAR_HIS_PUMPCTRL, + PCTRL_CMD_CONT, 0, NULL); + } else { + pr_debug("%s: pump stev LINE_RX_H wrong st %x\n", + ch->is->name, ch->state); + } + break; + case PSEV_LINE_TX_B: + if (ch->state == STFAX_LINE) { + pr_debug("%s: pump stev LINE_TX_B\n", ch->is->name); + ch->state = STFAX_CONT; + send_mbox(ch->is, dps | ISAR_HIS_PUMPCTRL, + PCTRL_CMD_CONT, 0, NULL); + } else { + pr_debug("%s: pump stev LINE_TX_B wrong st %x\n", + ch->is->name, ch->state); + } + break; + case PSEV_LINE_RX_B: + if (ch->state == STFAX_LINE) { + pr_debug("%s: pump stev LINE_RX_B\n", ch->is->name); + ch->state = STFAX_CONT; + send_mbox(ch->is, dps | ISAR_HIS_PUMPCTRL, + PCTRL_CMD_CONT, 0, NULL); + } else { + pr_debug("%s: pump stev LINE_RX_B wrong st %x\n", + ch->is->name, ch->state); + } + break; + case PSEV_RSP_CONN: + if (ch->state == STFAX_CONT) { + pr_debug("%s: pump stev RSP_CONN\n", ch->is->name); + ch->state = STFAX_ACTIV; + test_and_set_bit(ISAR_RATE_REQ, &ch->is->Flags); + send_mbox(ch->is, dps | ISAR_HIS_PSTREQ, 0, 0, NULL); + if (ch->cmd == PCTRL_CMD_FTH) { + int delay = (ch->mod == 3) ? 1000 : 200; + /* 1s (200 ms) Flags before data */ + if (test_and_set_bit(FLG_FTI_RUN, + &ch->bch.Flags)) + del_timer(&ch->ftimer); + ch->ftimer.expires = + jiffies + ((delay * HZ)/1000); + test_and_set_bit(FLG_LL_CONN, + &ch->bch.Flags); + add_timer(&ch->ftimer); + } else { + deliver_status(ch, HW_MOD_CONNECT); + } + } else { + pr_debug("%s: pump stev RSP_CONN wrong st %x\n", + ch->is->name, ch->state); + } + break; + case PSEV_FLAGS_DET: + pr_debug("%s: pump stev FLAGS_DET\n", ch->is->name); + break; + case PSEV_RSP_DISC: + pr_debug("%s: pump stev RSP_DISC state(%d)\n", + ch->is->name, ch->state); + if (ch->state == STFAX_ESCAPE) { + p1 = 5; + switch (ch->newcmd) { + case 0: + ch->state = STFAX_READY; + break; + case PCTRL_CMD_FTM: + p1 = 2; + case PCTRL_CMD_FTH: + send_mbox(ch->is, dps | ISAR_HIS_PUMPCTRL, + PCTRL_CMD_SILON, 1, &p1); + ch->state = STFAX_SILDET; + break; + case PCTRL_CMD_FRH: + case PCTRL_CMD_FRM: + ch->mod = ch->newmod; + p1 = ch->newmod; + ch->newmod = 0; + ch->cmd = ch->newcmd; + ch->newcmd = 0; + send_mbox(ch->is, dps | ISAR_HIS_PUMPCTRL, + ch->cmd, 1, &p1); + ch->state = STFAX_LINE; + ch->try_mod = 3; + break; + default: + pr_debug("%s: RSP_DISC unknown newcmd %x\n", + ch->is->name, ch->newcmd); + break; + } + } else if (ch->state == STFAX_ACTIV) { + if (test_and_clear_bit(FLG_LL_OK, &ch->bch.Flags)) + deliver_status(ch, HW_MOD_OK); + else if (ch->cmd == PCTRL_CMD_FRM) + deliver_status(ch, HW_MOD_NOCARR); + else + deliver_status(ch, HW_MOD_FCERROR); + ch->state = STFAX_READY; + } else if (ch->state != STFAX_SILDET) { + /* ignore in STFAX_SILDET */ + ch->state = STFAX_READY; + deliver_status(ch, HW_MOD_FCERROR); + } + break; + case PSEV_RSP_SILDET: + pr_debug("%s: pump stev RSP_SILDET\n", ch->is->name); + if (ch->state == STFAX_SILDET) { + ch->mod = ch->newmod; + p1 = ch->newmod; + ch->newmod = 0; + ch->cmd = ch->newcmd; + ch->newcmd = 0; + send_mbox(ch->is, dps | ISAR_HIS_PUMPCTRL, + ch->cmd, 1, &p1); + ch->state = STFAX_LINE; + ch->try_mod = 3; + } + break; + case PSEV_RSP_SILOFF: + pr_debug("%s: pump stev RSP_SILOFF\n", ch->is->name); + break; + case PSEV_RSP_FCERR: + if (ch->state == STFAX_LINE) { + pr_debug("%s: pump stev RSP_FCERR try %d\n", + ch->is->name, ch->try_mod); + if (ch->try_mod--) { + send_mbox(ch->is, dps | ISAR_HIS_PUMPCTRL, + ch->cmd, 1, &ch->mod); + break; + } + } + pr_debug("%s: pump stev RSP_FCERR\n", ch->is->name); + ch->state = STFAX_ESCAPE; + send_mbox(ch->is, dps | ISAR_HIS_PUMPCTRL, PCTRL_CMD_ESC, + 0, NULL); + deliver_status(ch, HW_MOD_FCERROR); + break; + default: + break; + } +} + +void +mISDNisar_irq(struct isar_hw *isar) +{ + struct isar_ch *ch; + + get_irq_infos(isar); + switch (isar->iis & ISAR_IIS_MSCMSD) { + case ISAR_IIS_RDATA: + ch = sel_bch_isar(isar, isar->iis >> 6); + if (ch) + isar_rcv_frame(ch); + else { + pr_debug("%s: ISAR spurious IIS_RDATA %x/%x/%x\n", + isar->name, isar->iis, isar->cmsb, + isar->clsb); + isar->write_reg(isar->hw, ISAR_IIA, 0); + } + break; + case ISAR_IIS_GSTEV: + isar->write_reg(isar->hw, ISAR_IIA, 0); + isar->bstat |= isar->cmsb; + check_send(isar, isar->cmsb); + break; + case ISAR_IIS_BSTEV: +#ifdef ERROR_STATISTIC + ch = sel_bch_isar(isar, isar->iis >> 6); + if (ch) { + if (isar->cmsb == BSTEV_TBO) + ch->bch.err_tx++; + if (isar->cmsb == BSTEV_RBO) + ch->bch.err_rdo++; + } +#endif + pr_debug("%s: Buffer STEV dpath%d msb(%x)\n", + isar->name, isar->iis>>6, isar->cmsb); + isar->write_reg(isar->hw, ISAR_IIA, 0); + break; + case ISAR_IIS_PSTEV: + ch = sel_bch_isar(isar, isar->iis >> 6); + if (ch) { + rcv_mbox(isar, NULL); + if (ch->bch.state == ISDN_P_B_MODEM_ASYNC) + isar_pump_statev_modem(ch, isar->cmsb); + else if (ch->bch.state == ISDN_P_B_T30_FAX) + isar_pump_statev_fax(ch, isar->cmsb); + else if (ch->bch.state == ISDN_P_B_RAW) { + int tt; + tt = isar->cmsb | 0x30; + if (tt == 0x3e) + tt = '*'; + else if (tt == 0x3f) + tt = '#'; + else if (tt > '9') + tt += 7; + tt |= DTMF_TONE_VAL; + _queue_data(&ch->bch.ch, PH_CONTROL_IND, + MISDN_ID_ANY, sizeof(tt), &tt, + GFP_ATOMIC); + } else + pr_debug("%s: ISAR IIS_PSTEV pm %d sta %x\n", + isar->name, ch->bch.state, + isar->cmsb); + } else { + pr_debug("%s: ISAR spurious IIS_PSTEV %x/%x/%x\n", + isar->name, isar->iis, isar->cmsb, + isar->clsb); + isar->write_reg(isar->hw, ISAR_IIA, 0); + } + break; + case ISAR_IIS_PSTRSP: + ch = sel_bch_isar(isar, isar->iis >> 6); + if (ch) { + rcv_mbox(isar, NULL); + isar_pump_status_rsp(ch); + } else { + pr_debug("%s: ISAR spurious IIS_PSTRSP %x/%x/%x\n", + isar->name, isar->iis, isar->cmsb, + isar->clsb); + isar->write_reg(isar->hw, ISAR_IIA, 0); + } + break; + case ISAR_IIS_DIAG: + case ISAR_IIS_BSTRSP: + case ISAR_IIS_IOM2RSP: + rcv_mbox(isar, NULL); + break; + case ISAR_IIS_INVMSG: + rcv_mbox(isar, NULL); + pr_debug("%s: invalid msg his:%x\n", isar->name, isar->cmsb); + break; + default: + rcv_mbox(isar, NULL); + pr_debug("%s: unhandled msg iis(%x) ctrl(%x/%x)\n", + isar->name, isar->iis, isar->cmsb, isar->clsb); + break; + } +} +EXPORT_SYMBOL(mISDNisar_irq); + +static void +ftimer_handler(unsigned long data) +{ + struct isar_ch *ch = (struct isar_ch *)data; + + pr_debug("%s: ftimer flags %lx\n", ch->is->name, ch->bch.Flags); + test_and_clear_bit(FLG_FTI_RUN, &ch->bch.Flags); + if (test_and_clear_bit(FLG_LL_CONN, &ch->bch.Flags)) + deliver_status(ch, HW_MOD_CONNECT); +} + +static void +setup_pump(struct isar_ch *ch) { + u8 dps = SET_DPS(ch->dpath); + u8 ctrl, param[6]; + + switch (ch->bch.state) { + case ISDN_P_NONE: + case ISDN_P_B_RAW: + case ISDN_P_B_HDLC: + send_mbox(ch->is, dps | ISAR_HIS_PUMPCFG, PMOD_BYPASS, 0, NULL); + break; + case ISDN_P_B_L2DTMF: + if (test_bit(FLG_DTMFSEND, &ch->bch.Flags)) { + param[0] = 5; /* TOA 5 db */ + send_mbox(ch->is, dps | ISAR_HIS_PUMPCFG, + PMOD_DTMF_TRANS, 1, param); + } else { + param[0] = 40; /* REL -46 dbm */ + send_mbox(ch->is, dps | ISAR_HIS_PUMPCFG, + PMOD_DTMF, 1, param); + } + case ISDN_P_B_MODEM_ASYNC: + ctrl = PMOD_DATAMODEM; + if (test_bit(FLG_ORIGIN, &ch->bch.Flags)) { + ctrl |= PCTRL_ORIG; + param[5] = PV32P6_CTN; + } else { + param[5] = PV32P6_ATN; + } + param[0] = 6; /* 6 db */ + param[1] = PV32P2_V23R | PV32P2_V22A | PV32P2_V22B | + PV32P2_V22C | PV32P2_V21 | PV32P2_BEL; + param[2] = PV32P3_AMOD | PV32P3_V32B | PV32P3_V23B; + param[3] = PV32P4_UT144; + param[4] = PV32P5_UT144; + send_mbox(ch->is, dps | ISAR_HIS_PUMPCFG, ctrl, 6, param); + break; + case ISDN_P_B_T30_FAX: + ctrl = PMOD_FAX; + if (test_bit(FLG_ORIGIN, &ch->bch.Flags)) { + ctrl |= PCTRL_ORIG; + param[1] = PFAXP2_CTN; + } else { + param[1] = PFAXP2_ATN; + } + param[0] = 6; /* 6 db */ + send_mbox(ch->is, dps | ISAR_HIS_PUMPCFG, ctrl, 2, param); + ch->state = STFAX_NULL; + ch->newcmd = 0; + ch->newmod = 0; + test_and_set_bit(FLG_FTI_RUN, &ch->bch.Flags); + break; + } + udelay(1000); + send_mbox(ch->is, dps | ISAR_HIS_PSTREQ, 0, 0, NULL); + udelay(1000); +} + +static void +setup_sart(struct isar_ch *ch) { + u8 dps = SET_DPS(ch->dpath); + u8 ctrl, param[2] = {0, 0}; + + switch (ch->bch.state) { + case ISDN_P_NONE: + send_mbox(ch->is, dps | ISAR_HIS_SARTCFG, SMODE_DISABLE, + 0, NULL); + break; + case ISDN_P_B_RAW: + case ISDN_P_B_L2DTMF: + send_mbox(ch->is, dps | ISAR_HIS_SARTCFG, SMODE_BINARY, + 2, param); + break; + case ISDN_P_B_HDLC: + case ISDN_P_B_T30_FAX: + send_mbox(ch->is, dps | ISAR_HIS_SARTCFG, SMODE_HDLC, + 1, param); + break; + case ISDN_P_B_MODEM_ASYNC: + ctrl = SMODE_V14 | SCTRL_HDMC_BOTH; + param[0] = S_P1_CHS_8; + param[1] = S_P2_BFT_DEF; + send_mbox(ch->is, dps | ISAR_HIS_SARTCFG, ctrl, 2, param); + break; + } + udelay(1000); + send_mbox(ch->is, dps | ISAR_HIS_BSTREQ, 0, 0, NULL); + udelay(1000); +} + +static void +setup_iom2(struct isar_ch *ch) { + u8 dps = SET_DPS(ch->dpath); + u8 cmsb = IOM_CTRL_ENA, msg[5] = {IOM_P1_TXD, 0, 0, 0, 0}; + + if (ch->bch.nr == 2) { + msg[1] = 1; + msg[3] = 1; + } + switch (ch->bch.state) { + case ISDN_P_NONE: + cmsb = 0; + /* dummy slot */ + msg[1] = ch->dpath + 2; + msg[3] = ch->dpath + 2; + break; + case ISDN_P_B_RAW: + case ISDN_P_B_HDLC: + break; + case ISDN_P_B_MODEM_ASYNC: + case ISDN_P_B_T30_FAX: + cmsb |= IOM_CTRL_RCV; + case ISDN_P_B_L2DTMF: + if (test_bit(FLG_DTMFSEND, &ch->bch.Flags)) + cmsb |= IOM_CTRL_RCV; + cmsb |= IOM_CTRL_ALAW; + break; + } + send_mbox(ch->is, dps | ISAR_HIS_IOM2CFG, cmsb, 5, msg); + udelay(1000); + send_mbox(ch->is, dps | ISAR_HIS_IOM2REQ, 0, 0, NULL); + udelay(1000); +} + +static int +modeisar(struct isar_ch *ch, u32 bprotocol) +{ + /* Here we are selecting the best datapath for requested protocol */ + if (ch->bch.state == ISDN_P_NONE) { /* New Setup */ + switch (bprotocol) { + case ISDN_P_NONE: /* init */ + if (!ch->dpath) + /* no init for dpath 0 */ + return 0; + test_and_clear_bit(FLG_HDLC, &ch->bch.Flags); + test_and_clear_bit(FLG_TRANSPARENT, &ch->bch.Flags); + break; + case ISDN_P_B_RAW: + case ISDN_P_B_HDLC: + /* best is datapath 2 */ + if (!test_and_set_bit(ISAR_DP2_USE, &ch->is->Flags)) + ch->dpath = 2; + else if (!test_and_set_bit(ISAR_DP1_USE, + &ch->is->Flags)) + ch->dpath = 1; + else { + pr_info("modeisar both pathes in use\n"); + return -EBUSY; + } + if (bprotocol == ISDN_P_B_HDLC) + test_and_set_bit(FLG_HDLC, &ch->bch.Flags); + else + test_and_set_bit(FLG_TRANSPARENT, + &ch->bch.Flags); + break; + case ISDN_P_B_MODEM_ASYNC: + case ISDN_P_B_T30_FAX: + case ISDN_P_B_L2DTMF: + /* only datapath 1 */ + if (!test_and_set_bit(ISAR_DP1_USE, &ch->is->Flags)) + ch->dpath = 1; + else { + pr_info("%s: ISAR modeisar analog functions" + "only with DP1\n", ch->is->name); + return -EBUSY; + } + break; + default: + pr_info("%s: protocol not known %x\n", ch->is->name, + bprotocol); + return -ENOPROTOOPT; + } + } + pr_debug("%s: ISAR ch%d dp%d protocol %x->%x\n", ch->is->name, + ch->bch.nr, ch->dpath, ch->bch.state, bprotocol); + ch->bch.state = bprotocol; + setup_pump(ch); + setup_iom2(ch); + setup_sart(ch); + if (ch->bch.state == ISDN_P_NONE) { + /* Clear resources */ + if (ch->dpath == 1) + test_and_clear_bit(ISAR_DP1_USE, &ch->is->Flags); + else if (ch->dpath == 2) + test_and_clear_bit(ISAR_DP2_USE, &ch->is->Flags); + ch->dpath = 0; + ch->is->ctrl(ch->is->hw, HW_DEACT_IND, ch->bch.nr); + } else + ch->is->ctrl(ch->is->hw, HW_ACTIVATE_IND, ch->bch.nr); + return 0; +} + +static void +isar_pump_cmd(struct isar_ch *ch, u32 cmd, u8 para) +{ + u8 dps = SET_DPS(ch->dpath); + u8 ctrl = 0, nom = 0, p1 = 0; + + pr_debug("%s: isar_pump_cmd %x/%x state(%x)\n", + ch->is->name, cmd, para, ch->bch.state); + switch (cmd) { + case HW_MOD_FTM: + if (ch->state == STFAX_READY) { + p1 = para; + ctrl = PCTRL_CMD_FTM; + nom = 1; + ch->state = STFAX_LINE; + ch->cmd = ctrl; + ch->mod = para; + ch->newmod = 0; + ch->newcmd = 0; + ch->try_mod = 3; + } else if ((ch->state == STFAX_ACTIV) && + (ch->cmd == PCTRL_CMD_FTM) && (ch->mod == para)) + deliver_status(ch, HW_MOD_CONNECT); + else { + ch->newmod = para; + ch->newcmd = PCTRL_CMD_FTM; + nom = 0; + ctrl = PCTRL_CMD_ESC; + ch->state = STFAX_ESCAPE; + } + break; + case HW_MOD_FTH: + if (ch->state == STFAX_READY) { + p1 = para; + ctrl = PCTRL_CMD_FTH; + nom = 1; + ch->state = STFAX_LINE; + ch->cmd = ctrl; + ch->mod = para; + ch->newmod = 0; + ch->newcmd = 0; + ch->try_mod = 3; + } else if ((ch->state == STFAX_ACTIV) && + (ch->cmd == PCTRL_CMD_FTH) && (ch->mod == para)) + deliver_status(ch, HW_MOD_CONNECT); + else { + ch->newmod = para; + ch->newcmd = PCTRL_CMD_FTH; + nom = 0; + ctrl = PCTRL_CMD_ESC; + ch->state = STFAX_ESCAPE; + } + break; + case HW_MOD_FRM: + if (ch->state == STFAX_READY) { + p1 = para; + ctrl = PCTRL_CMD_FRM; + nom = 1; + ch->state = STFAX_LINE; + ch->cmd = ctrl; + ch->mod = para; + ch->newmod = 0; + ch->newcmd = 0; + ch->try_mod = 3; + } else if ((ch->state == STFAX_ACTIV) && + (ch->cmd == PCTRL_CMD_FRM) && (ch->mod == para)) + deliver_status(ch, HW_MOD_CONNECT); + else { + ch->newmod = para; + ch->newcmd = PCTRL_CMD_FRM; + nom = 0; + ctrl = PCTRL_CMD_ESC; + ch->state = STFAX_ESCAPE; + } + break; + case HW_MOD_FRH: + if (ch->state == STFAX_READY) { + p1 = para; + ctrl = PCTRL_CMD_FRH; + nom = 1; + ch->state = STFAX_LINE; + ch->cmd = ctrl; + ch->mod = para; + ch->newmod = 0; + ch->newcmd = 0; + ch->try_mod = 3; + } else if ((ch->state == STFAX_ACTIV) && + (ch->cmd == PCTRL_CMD_FRH) && (ch->mod == para)) + deliver_status(ch, HW_MOD_CONNECT); + else { + ch->newmod = para; + ch->newcmd = PCTRL_CMD_FRH; + nom = 0; + ctrl = PCTRL_CMD_ESC; + ch->state = STFAX_ESCAPE; + } + break; + case PCTRL_CMD_TDTMF: + p1 = para; + nom = 1; + ctrl = PCTRL_CMD_TDTMF; + break; + } + if (ctrl) + send_mbox(ch->is, dps | ISAR_HIS_PUMPCTRL, ctrl, nom, &p1); +} + +static void +isar_setup(struct isar_hw *isar) +{ + u8 msg; + int i; + + /* Dpath 1, 2 */ + msg = 61; + for (i = 0; i < 2; i++) { + /* Buffer Config */ + send_mbox(isar, (i ? ISAR_HIS_DPS2 : ISAR_HIS_DPS1) | + ISAR_HIS_P12CFG, 4, 1, &msg); + isar->ch[i].mml = msg; + isar->ch[i].bch.state = 0; + isar->ch[i].dpath = i + 1; + modeisar(&isar->ch[i], ISDN_P_NONE); + } +} + +static int +isar_l2l1(struct mISDNchannel *ch, struct sk_buff *skb) +{ + struct bchannel *bch = container_of(ch, struct bchannel, ch); + struct isar_ch *ich = container_of(bch, struct isar_ch, bch); + int ret = -EINVAL; + struct mISDNhead *hh = mISDN_HEAD_P(skb); + u32 id, *val; + u_long flags; + + switch (hh->prim) { + case PH_DATA_REQ: + spin_lock_irqsave(ich->is->hwlock, flags); + ret = bchannel_senddata(bch, skb); + if (ret > 0) { /* direct TX */ + id = hh->id; /* skb can be freed */ + ret = 0; + isar_fill_fifo(ich); + spin_unlock_irqrestore(ich->is->hwlock, flags); + if (!test_bit(FLG_TRANSPARENT, &bch->Flags)) + queue_ch_frame(ch, PH_DATA_CNF, id, NULL); + } else + spin_unlock_irqrestore(ich->is->hwlock, flags); + return ret; + case PH_ACTIVATE_REQ: + spin_lock_irqsave(ich->is->hwlock, flags); + if (!test_and_set_bit(FLG_ACTIVE, &bch->Flags)) + ret = modeisar(ich, ch->protocol); + else + ret = 0; + spin_unlock_irqrestore(ich->is->hwlock, flags); + if (!ret) + _queue_data(ch, PH_ACTIVATE_IND, MISDN_ID_ANY, 0, + NULL, GFP_KERNEL); + break; + case PH_DEACTIVATE_REQ: + spin_lock_irqsave(ich->is->hwlock, flags); + mISDN_clear_bchannel(bch); + modeisar(ich, ISDN_P_NONE); + spin_unlock_irqrestore(ich->is->hwlock, flags); + _queue_data(ch, PH_DEACTIVATE_IND, MISDN_ID_ANY, 0, + NULL, GFP_KERNEL); + ret = 0; + break; + case PH_CONTROL_REQ: + val = (u32 *)skb->data; + pr_debug("%s: PH_CONTROL | REQUEST %x/%x\n", ich->is->name, + hh->id, *val); + if ((hh->id == 0) && ((*val & ~DTMF_TONE_MASK) == + DTMF_TONE_VAL)) { + if (bch->state == ISDN_P_B_L2DTMF) { + char tt = *val & DTMF_TONE_MASK; + + if (tt == '*') + tt = 0x1e; + else if (tt == '#') + tt = 0x1f; + else if (tt > '9') + tt -= 7; + tt &= 0x1f; + spin_lock_irqsave(ich->is->hwlock, flags); + isar_pump_cmd(ich, PCTRL_CMD_TDTMF, tt); + spin_unlock_irqrestore(ich->is->hwlock, flags); + } else { + pr_info("%s: DTMF send wrong protocol %x\n", + __func__, bch->state); + return -EINVAL; + } + } else if ((hh->id == HW_MOD_FRM) || (hh->id == HW_MOD_FRH) || + (hh->id == HW_MOD_FTM) || (hh->id == HW_MOD_FTH)) { + for (id = 0; id < FAXMODCNT; id++) + if (faxmodulation[id] == *val) + break; + if ((FAXMODCNT > id) && + test_bit(FLG_INITIALIZED, &bch->Flags)) { + pr_debug("%s: isar: new mod\n", ich->is->name); + isar_pump_cmd(ich, hh->id, *val); + ret = 0; + } else { + pr_info("%s: wrong modulation\n", + ich->is->name); + ret = -EINVAL; + } + } else if (hh->id == HW_MOD_LASTDATA) + test_and_set_bit(FLG_DLEETX, &bch->Flags); + else { + pr_info("%s: unknown PH_CONTROL_REQ %x\n", + ich->is->name, hh->id); + ret = -EINVAL; + } + default: + pr_info("%s: %s unknown prim(%x,%x)\n", + ich->is->name, __func__, hh->prim, hh->id); + ret = -EINVAL; + } + if (!ret) + dev_kfree_skb(skb); + return ret; +} + +static int +channel_bctrl(struct bchannel *bch, struct mISDN_ctrl_req *cq) +{ + int ret = 0; + + switch (cq->op) { + case MISDN_CTRL_GETOP: + cq->op = 0; + break; + /* Nothing implemented yet */ + case MISDN_CTRL_FILL_EMPTY: + default: + pr_info("%s: unknown Op %x\n", __func__, cq->op); + ret = -EINVAL; + break; + } + return ret; +} + +static int +isar_bctrl(struct mISDNchannel *ch, u32 cmd, void *arg) +{ + struct bchannel *bch = container_of(ch, struct bchannel, ch); + struct isar_ch *ich = container_of(bch, struct isar_ch, bch); + int ret = -EINVAL; + u_long flags; + + pr_debug("%s: %s cmd:%x %p\n", ich->is->name, __func__, cmd, arg); + switch (cmd) { + case CLOSE_CHANNEL: + test_and_clear_bit(FLG_OPEN, &bch->Flags); + if (test_bit(FLG_ACTIVE, &bch->Flags)) { + spin_lock_irqsave(ich->is->hwlock, flags); + mISDN_freebchannel(bch); + modeisar(ich, ISDN_P_NONE); + spin_unlock_irqrestore(ich->is->hwlock, flags); + } else { + skb_queue_purge(&bch->rqueue); + bch->rcount = 0; + } + ch->protocol = ISDN_P_NONE; + ch->peer = NULL; + module_put(ich->is->owner); + ret = 0; + break; + case CONTROL_CHANNEL: + ret = channel_bctrl(bch, arg); + break; + default: + pr_info("%s: %s unknown prim(%x)\n", + ich->is->name, __func__, cmd); + } + return ret; +} + +static void +free_isar(struct isar_hw *isar) +{ + modeisar(&isar->ch[0], ISDN_P_NONE); + modeisar(&isar->ch[1], ISDN_P_NONE); + del_timer(&isar->ch[0].ftimer); + del_timer(&isar->ch[1].ftimer); + test_and_clear_bit(FLG_INITIALIZED, &isar->ch[0].bch.Flags); + test_and_clear_bit(FLG_INITIALIZED, &isar->ch[1].bch.Flags); +} + +static int +init_isar(struct isar_hw *isar) +{ + int cnt = 3; + + while (cnt--) { + isar->version = ISARVersion(isar); + if (isar->ch[0].bch.debug & DEBUG_HW) + pr_notice("%s: Testing version %d (%d time)\n", + isar->name, isar->version, 3 - cnt); + if (isar->version == 1) + break; + isar->ctrl(isar->hw, HW_RESET_REQ, 0); + } + if (isar->version != 1) + return -EINVAL; + isar->ch[0].ftimer.function = &ftimer_handler; + isar->ch[0].ftimer.data = (long)&isar->ch[0]; + init_timer(&isar->ch[0].ftimer); + test_and_set_bit(FLG_INITIALIZED, &isar->ch[0].bch.Flags); + isar->ch[1].ftimer.function = &ftimer_handler; + isar->ch[1].ftimer.data = (long)&isar->ch[1]; + init_timer(&isar->ch[1].ftimer); + test_and_set_bit(FLG_INITIALIZED, &isar->ch[1].bch.Flags); + return 0; +} + +static int +isar_open(struct isar_hw *isar, struct channel_req *rq) +{ + struct bchannel *bch; + + if (rq->adr.channel > 2) + return -EINVAL; + if (rq->protocol == ISDN_P_NONE) + return -EINVAL; + bch = &isar->ch[rq->adr.channel - 1].bch; + if (test_and_set_bit(FLG_OPEN, &bch->Flags)) + return -EBUSY; /* b-channel can be only open once */ + test_and_clear_bit(FLG_FILLEMPTY, &bch->Flags); + bch->ch.protocol = rq->protocol; + rq->ch = &bch->ch; + return 0; +} + +u32 +mISDNisar_init(struct isar_hw *isar, void *hw) +{ + u32 ret, i; + + isar->hw = hw; + for (i = 0; i < 2; i++) { + isar->ch[i].bch.nr = i + 1; + mISDN_initbchannel(&isar->ch[i].bch, MAX_DATA_MEM); + isar->ch[i].bch.ch.nr = i + 1; + isar->ch[i].bch.ch.send = &isar_l2l1; + isar->ch[i].bch.ch.ctrl = isar_bctrl; + isar->ch[i].bch.hw = hw; + isar->ch[i].is = isar; + } + + isar->init = &init_isar; + isar->release = &free_isar; + isar->firmware = &load_firmware; + isar->open = &isar_open; + + ret = (1 << (ISDN_P_B_RAW & ISDN_P_B_MASK)) | + (1 << (ISDN_P_B_HDLC & ISDN_P_B_MASK)) | + (1 << (ISDN_P_B_L2DTMF & ISDN_P_B_MASK)) | + (1 << (ISDN_P_B_MODEM_ASYNC & ISDN_P_B_MASK)) | + (1 << (ISDN_P_B_T30_FAX & ISDN_P_B_MASK)); + + return ret; +} +EXPORT_SYMBOL(mISDNisar_init); + +static int isar_mod_init(void) +{ + pr_notice("mISDN: ISAR driver Rev. %s\n", ISAR_REV); + return 0; +} + +static void isar_mod_cleanup(void) +{ + pr_notice("mISDN: ISAR module unloaded\n"); +} +module_init(isar_mod_init); +module_exit(isar_mod_cleanup); diff --git a/drivers/isdn/hardware/mISDN/speedfax.c b/drivers/isdn/hardware/mISDN/speedfax.c new file mode 100644 index 000000000000..ff3a4e290da3 --- /dev/null +++ b/drivers/isdn/hardware/mISDN/speedfax.c @@ -0,0 +1,526 @@ +/* + * speedfax.c low level stuff for Sedlbauer Speedfax+ cards + * based on the ISAR DSP + * Thanks to Sedlbauer AG for informations and HW + * + * Author Karsten Keil + * + * Copyright 2009 by Karsten Keil + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. + * + */ + +#include +#include +#include +#include +#include +#include "ipac.h" +#include "isar.h" + +#define SPEEDFAX_REV "2.0" + +#define PCI_SUBVENDOR_SPEEDFAX_PYRAMID 0x51 +#define PCI_SUBVENDOR_SPEEDFAX_PCI 0x54 +#define PCI_SUB_ID_SEDLBAUER 0x01 + +#define SFAX_PCI_ADDR 0xc8 +#define SFAX_PCI_ISAC 0xd0 +#define SFAX_PCI_ISAR 0xe0 + +/* TIGER 100 Registers */ + +#define TIGER_RESET_ADDR 0x00 +#define TIGER_EXTERN_RESET_ON 0x01 +#define TIGER_EXTERN_RESET_OFF 0x00 +#define TIGER_AUX_CTRL 0x02 +#define TIGER_AUX_DATA 0x03 +#define TIGER_AUX_IRQMASK 0x05 +#define TIGER_AUX_STATUS 0x07 + +/* Tiger AUX BITs */ +#define SFAX_AUX_IOMASK 0xdd /* 1 and 5 are inputs */ +#define SFAX_ISAR_RESET_BIT_OFF 0x00 +#define SFAX_ISAR_RESET_BIT_ON 0x01 +#define SFAX_TIGER_IRQ_BIT 0x02 +#define SFAX_LED1_BIT 0x08 +#define SFAX_LED2_BIT 0x10 + +#define SFAX_PCI_RESET_ON (SFAX_ISAR_RESET_BIT_ON) +#define SFAX_PCI_RESET_OFF (SFAX_LED1_BIT | SFAX_LED2_BIT) + +static int sfax_cnt; +static u32 debug; +static u32 irqloops = 4; + +struct sfax_hw { + struct list_head list; + struct pci_dev *pdev; + char name[MISDN_MAX_IDLEN]; + u32 irq; + u32 irqcnt; + u32 cfg; + struct _ioport p_isac; + struct _ioport p_isar; + u8 aux_data; + spinlock_t lock; /* HW access lock */ + struct isac_hw isac; + struct isar_hw isar; +}; + +static LIST_HEAD(Cards); +static DEFINE_RWLOCK(card_lock); /* protect Cards */ + +static void +_set_debug(struct sfax_hw *card) +{ + card->isac.dch.debug = debug; + card->isar.ch[0].bch.debug = debug; + card->isar.ch[1].bch.debug = debug; +} + +static int +set_debug(const char *val, struct kernel_param *kp) +{ + int ret; + struct sfax_hw *card; + + ret = param_set_uint(val, kp); + if (!ret) { + read_lock(&card_lock); + list_for_each_entry(card, &Cards, list) + _set_debug(card); + read_unlock(&card_lock); + } + return ret; +} + +MODULE_AUTHOR("Karsten Keil"); +MODULE_LICENSE("GPL v2"); +MODULE_VERSION(SPEEDFAX_REV); +module_param_call(debug, set_debug, param_get_uint, &debug, S_IRUGO | S_IWUSR); +MODULE_PARM_DESC(debug, "Speedfax debug mask"); +module_param(irqloops, uint, S_IRUGO | S_IWUSR); +MODULE_PARM_DESC(irqloops, "Speedfax maximal irqloops (default 4)"); + +IOFUNC_IND(ISAC, sfax_hw, p_isac) +IOFUNC_IND(ISAR, sfax_hw, p_isar) + +static irqreturn_t +speedfax_irq(int intno, void *dev_id) +{ + struct sfax_hw *sf = dev_id; + u8 val; + int cnt = irqloops; + + spin_lock(&sf->lock); + val = inb(sf->cfg + TIGER_AUX_STATUS); + if (val & SFAX_TIGER_IRQ_BIT) { /* for us or shared ? */ + spin_unlock(&sf->lock); + return IRQ_NONE; /* shared */ + } + sf->irqcnt++; + val = ReadISAR_IND(sf, ISAR_IRQBIT); +Start_ISAR: + if (val & ISAR_IRQSTA) + mISDNisar_irq(&sf->isar); + val = ReadISAC_IND(sf, ISAC_ISTA); + if (val) + mISDNisac_irq(&sf->isac, val); + val = ReadISAR_IND(sf, ISAR_IRQBIT); + if ((val & ISAR_IRQSTA) && cnt--) + goto Start_ISAR; + if (cnt < irqloops) + pr_debug("%s: %d irqloops cpu%d\n", sf->name, + irqloops - cnt, smp_processor_id()); + if (irqloops && !cnt) + pr_notice("%s: %d IRQ LOOP cpu%d\n", sf->name, + irqloops, smp_processor_id()); + spin_unlock(&sf->lock); + return IRQ_HANDLED; +} + +static void +enable_hwirq(struct sfax_hw *sf) +{ + WriteISAC_IND(sf, ISAC_MASK, 0); + WriteISAR_IND(sf, ISAR_IRQBIT, ISAR_IRQMSK); + outb(SFAX_TIGER_IRQ_BIT, sf->cfg + TIGER_AUX_IRQMASK); +} + +static void +disable_hwirq(struct sfax_hw *sf) +{ + WriteISAC_IND(sf, ISAC_MASK, 0xFF); + WriteISAR_IND(sf, ISAR_IRQBIT, 0); + outb(0, sf->cfg + TIGER_AUX_IRQMASK); +} + +static void +reset_speedfax(struct sfax_hw *sf) +{ + + pr_debug("%s: resetting card\n", sf->name); + outb(TIGER_EXTERN_RESET_ON, sf->cfg + TIGER_RESET_ADDR); + outb(SFAX_PCI_RESET_ON, sf->cfg + TIGER_AUX_DATA); + mdelay(1); + outb(TIGER_EXTERN_RESET_OFF, sf->cfg + TIGER_RESET_ADDR); + sf->aux_data = SFAX_PCI_RESET_OFF; + outb(sf->aux_data, sf->cfg + TIGER_AUX_DATA); + mdelay(1); +} + +static int +sfax_ctrl(struct sfax_hw *sf, u32 cmd, u_long arg) +{ + int ret = 0; + + switch (cmd) { + case HW_RESET_REQ: + reset_speedfax(sf); + break; + case HW_ACTIVATE_IND: + if (arg & 1) + sf->aux_data &= ~SFAX_LED1_BIT; + if (arg & 2) + sf->aux_data &= ~SFAX_LED2_BIT; + outb(sf->aux_data, sf->cfg + TIGER_AUX_DATA); + break; + case HW_DEACT_IND: + if (arg & 1) + sf->aux_data |= SFAX_LED1_BIT; + if (arg & 2) + sf->aux_data |= SFAX_LED2_BIT; + outb(sf->aux_data, sf->cfg + TIGER_AUX_DATA); + break; + default: + pr_info("%s: %s unknown command %x %lx\n", + sf->name, __func__, cmd, arg); + ret = -EINVAL; + break; + } + return ret; +} + +static int +channel_ctrl(struct sfax_hw *sf, struct mISDN_ctrl_req *cq) +{ + int ret = 0; + + switch (cq->op) { + case MISDN_CTRL_GETOP: + cq->op = MISDN_CTRL_LOOP; + break; + case MISDN_CTRL_LOOP: + /* cq->channel: 0 disable, 1 B1 loop 2 B2 loop, 3 both */ + if (cq->channel < 0 || cq->channel > 3) { + ret = -EINVAL; + break; + } + ret = sf->isac.ctrl(&sf->isac, HW_TESTLOOP, cq->channel); + break; + default: + pr_info("%s: unknown Op %x\n", sf->name, cq->op); + ret = -EINVAL; + break; + } + return ret; +} + +static int +sfax_dctrl(struct mISDNchannel *ch, u32 cmd, void *arg) +{ + struct mISDNdevice *dev = container_of(ch, struct mISDNdevice, D); + struct dchannel *dch = container_of(dev, struct dchannel, dev); + struct sfax_hw *sf = dch->hw; + struct channel_req *rq; + int err = 0; + + pr_debug("%s: cmd:%x %p\n", sf->name, cmd, arg); + switch (cmd) { + case OPEN_CHANNEL: + rq = arg; + if (rq->protocol == ISDN_P_TE_S0) + err = sf->isac.open(&sf->isac, rq); + else + err = sf->isar.open(&sf->isar, rq); + if (err) + break; + if (!try_module_get(THIS_MODULE)) + pr_info("%s: cannot get module\n", sf->name); + break; + case CLOSE_CHANNEL: + pr_debug("%s: dev(%d) close from %p\n", sf->name, + dch->dev.id, __builtin_return_address(0)); + module_put(THIS_MODULE); + break; + case CONTROL_CHANNEL: + err = channel_ctrl(sf, arg); + break; + default: + pr_debug("%s: unknown command %x\n", sf->name, cmd); + return -EINVAL; + } + return err; +} + +static int __devinit +init_card(struct sfax_hw *sf) +{ + int ret, cnt = 3; + u_long flags; + + ret = request_irq(sf->irq, speedfax_irq, IRQF_SHARED, sf->name, sf); + if (ret) { + pr_info("%s: couldn't get interrupt %d\n", sf->name, sf->irq); + return ret; + } + while (cnt--) { + spin_lock_irqsave(&sf->lock, flags); + ret = sf->isac.init(&sf->isac); + if (ret) { + spin_unlock_irqrestore(&sf->lock, flags); + pr_info("%s: ISAC init failed with %d\n", + sf->name, ret); + break; + } + enable_hwirq(sf); + /* RESET Receiver and Transmitter */ + WriteISAC_IND(sf, ISAC_CMDR, 0x41); + spin_unlock_irqrestore(&sf->lock, flags); + msleep_interruptible(10); + if (debug & DEBUG_HW) + pr_notice("%s: IRQ %d count %d\n", sf->name, + sf->irq, sf->irqcnt); + if (!sf->irqcnt) { + pr_info("%s: IRQ(%d) got no requests during init %d\n", + sf->name, sf->irq, 3 - cnt); + } else + return 0; + } + free_irq(sf->irq, sf); + return -EIO; +} + + +static int __devinit +setup_speedfax(struct sfax_hw *sf) +{ + u_long flags; + + if (!request_region(sf->cfg, 256, sf->name)) { + pr_info("mISDN: %s config port %x-%x already in use\n", + sf->name, sf->cfg, sf->cfg + 255); + return -EIO; + } + outb(0xff, sf->cfg); + outb(0, sf->cfg); + outb(0xdd, sf->cfg + TIGER_AUX_CTRL); + outb(0, sf->cfg + TIGER_AUX_IRQMASK); + + sf->isac.type = IPAC_TYPE_ISAC; + sf->p_isac.ale = sf->cfg + SFAX_PCI_ADDR; + sf->p_isac.port = sf->cfg + SFAX_PCI_ISAC; + sf->p_isar.ale = sf->cfg + SFAX_PCI_ADDR; + sf->p_isar.port = sf->cfg + SFAX_PCI_ISAR; + ASSIGN_FUNC(IND, ISAC, sf->isac); + ASSIGN_FUNC(IND, ISAR, sf->isar); + spin_lock_irqsave(&sf->lock, flags); + reset_speedfax(sf); + disable_hwirq(sf); + spin_unlock_irqrestore(&sf->lock, flags); + return 0; +} + +static void +release_card(struct sfax_hw *card) { + u_long flags; + + spin_lock_irqsave(&card->lock, flags); + disable_hwirq(card); + spin_unlock_irqrestore(&card->lock, flags); + card->isac.release(&card->isac); + free_irq(card->irq, card); + card->isar.release(&card->isar); + mISDN_unregister_device(&card->isac.dch.dev); + release_region(card->cfg, 256); + pci_disable_device(card->pdev); + pci_set_drvdata(card->pdev, NULL); + write_lock_irqsave(&card_lock, flags); + list_del(&card->list); + write_unlock_irqrestore(&card_lock, flags); + kfree(card); + sfax_cnt--; +} + +static int __devinit +setup_instance(struct sfax_hw *card) +{ + const struct firmware *firmware; + int i, err; + u_long flags; + + snprintf(card->name, MISDN_MAX_IDLEN - 1, "Speedfax.%d", sfax_cnt + 1); + write_lock_irqsave(&card_lock, flags); + list_add_tail(&card->list, &Cards); + write_unlock_irqrestore(&card_lock, flags); + _set_debug(card); + spin_lock_init(&card->lock); + card->isac.hwlock = &card->lock; + card->isar.hwlock = &card->lock; + card->isar.ctrl = (void *)&sfax_ctrl; + card->isac.name = card->name; + card->isar.name = card->name; + card->isar.owner = THIS_MODULE; + + err = request_firmware(&firmware, "isdn/ISAR.BIN", &card->pdev->dev); + if (err < 0) { + pr_info("%s: firmware request failed %d\n", + card->name, err); + goto error_fw; + } + if (debug & DEBUG_HW) + pr_notice("%s: got firmware %zu bytes\n", + card->name, firmware->size); + + mISDNisac_init(&card->isac, card); + + card->isac.dch.dev.D.ctrl = sfax_dctrl; + card->isac.dch.dev.Bprotocols = + mISDNisar_init(&card->isar, card); + for (i = 0; i < 2; i++) { + set_channelmap(i + 1, card->isac.dch.dev.channelmap); + list_add(&card->isar.ch[i].bch.ch.list, + &card->isac.dch.dev.bchannels); + } + + err = setup_speedfax(card); + if (err) + goto error_setup; + err = card->isar.init(&card->isar); + if (err) + goto error; + err = mISDN_register_device(&card->isac.dch.dev, + &card->pdev->dev, card->name); + if (err) + goto error; + err = init_card(card); + if (err) + goto error_init; + err = card->isar.firmware(&card->isar, firmware->data, firmware->size); + if (!err) { + release_firmware(firmware); + sfax_cnt++; + pr_notice("SpeedFax %d cards installed\n", sfax_cnt); + return 0; + } + disable_hwirq(card); + free_irq(card->irq, card); +error_init: + mISDN_unregister_device(&card->isac.dch.dev); +error: + release_region(card->cfg, 256); +error_setup: + card->isac.release(&card->isac); + card->isar.release(&card->isar); + release_firmware(firmware); +error_fw: + pci_disable_device(card->pdev); + write_lock_irqsave(&card_lock, flags); + list_del(&card->list); + write_unlock_irqrestore(&card_lock, flags); + kfree(card); + return err; +} + +static int __devinit +sfaxpci_probe(struct pci_dev *pdev, const struct pci_device_id *ent) +{ + int err = -ENOMEM; + struct sfax_hw *card = kzalloc(sizeof(struct sfax_hw), GFP_KERNEL); + + if (!card) { + pr_info("No memory for Speedfax+ PCI\n"); + return err; + } + card->pdev = pdev; + err = pci_enable_device(pdev); + if (err) { + kfree(card); + return err; + } + + pr_notice("mISDN: Speedfax found adapter %s at %s\n", + (char *)ent->driver_data, pci_name(pdev)); + + card->cfg = pci_resource_start(pdev, 0); + card->irq = pdev->irq; + pci_set_drvdata(pdev, card); + err = setup_instance(card); + if (err) + pci_set_drvdata(pdev, NULL); + return err; +} + +static void __devexit +sfax_remove_pci(struct pci_dev *pdev) +{ + struct sfax_hw *card = pci_get_drvdata(pdev); + + if (card) + release_card(card); + else + pr_debug("%s: drvdata allready removed\n", __func__); +} + +static struct pci_device_id sfaxpci_ids[] __devinitdata = { + { PCI_VENDOR_ID_TIGERJET, PCI_DEVICE_ID_TIGERJET_100, + PCI_SUBVENDOR_SPEEDFAX_PYRAMID, PCI_SUB_ID_SEDLBAUER, + 0, 0, (unsigned long) "Pyramid Speedfax + PCI" + }, + { PCI_VENDOR_ID_TIGERJET, PCI_DEVICE_ID_TIGERJET_100, + PCI_SUBVENDOR_SPEEDFAX_PCI, PCI_SUB_ID_SEDLBAUER, + 0, 0, (unsigned long) "Sedlbauer Speedfax + PCI" + }, + { } +}; +MODULE_DEVICE_TABLE(pci, sfaxpci_ids); + +static struct pci_driver sfaxpci_driver = { + .name = "speedfax+ pci", + .probe = sfaxpci_probe, + .remove = __devexit_p(sfax_remove_pci), + .id_table = sfaxpci_ids, +}; + +static int __init +Speedfax_init(void) +{ + int err; + + pr_notice("Sedlbauer Speedfax+ Driver Rev. %s\n", + SPEEDFAX_REV); + err = pci_register_driver(&sfaxpci_driver); + return err; +} + +static void __exit +Speedfax_cleanup(void) +{ + pci_unregister_driver(&sfaxpci_driver); +} + +module_init(Speedfax_init); +module_exit(Speedfax_cleanup); diff --git a/include/linux/mISDNif.h b/include/linux/mISDNif.h index 45100b39a7cf..536ca12442ca 100644 --- a/include/linux/mISDNif.h +++ b/include/linux/mISDNif.h @@ -37,7 +37,7 @@ */ #define MISDN_MAJOR_VERSION 1 #define MISDN_MINOR_VERSION 1 -#define MISDN_RELEASE 20 +#define MISDN_RELEASE 21 /* primitives for information exchange * generell format @@ -153,6 +153,18 @@ #define HFC_VOL_CHANGE_RX 0x2602 #define HFC_SPL_LOOP_ON 0x2603 #define HFC_SPL_LOOP_OFF 0x2604 +/* for T30 FAX and analog modem */ +#define HW_MOD_FRM 0x4000 +#define HW_MOD_FRH 0x4001 +#define HW_MOD_FTM 0x4002 +#define HW_MOD_FTH 0x4003 +#define HW_MOD_FTS 0x4004 +#define HW_MOD_CONNECT 0x4010 +#define HW_MOD_OK 0x4011 +#define HW_MOD_NOCARR 0x4012 +#define HW_MOD_FCERROR 0x4013 +#define HW_MOD_READY 0x4014 +#define HW_MOD_LASTDATA 0x4015 /* DSP_TONE_PATT_ON parameter */ #define TONE_OFF 0x0000 @@ -224,6 +236,8 @@ #define ISDN_P_B_L2DTMF 0x24 #define ISDN_P_B_L2DSP 0x25 #define ISDN_P_B_L2DSPHDLC 0x26 +#define ISDN_P_B_T30_FAX 0x27 +#define ISDN_P_B_MODEM_ASYNC 0x28 #define OPTION_L2_PMX 1 #define OPTION_L2_PTP 2 -- cgit v1.2.3 From 633aae23ff31bef692a70772652e753a0ae59b81 Mon Sep 17 00:00:00 2001 From: Dmitry Torokhov Date: Wed, 22 Jul 2009 21:51:36 -0700 Subject: Input: serio - switch to using dev_pm_ops Signed-off-by: Dmitry Torokhov --- drivers/input/serio/serio.c | 23 +++++++++++------------ include/linux/serio.h | 2 -- 2 files changed, 11 insertions(+), 14 deletions(-) (limited to 'include') diff --git a/drivers/input/serio/serio.c b/drivers/input/serio/serio.c index d66f4944f2a0..0236f0d5fd91 100644 --- a/drivers/input/serio/serio.c +++ b/drivers/input/serio/serio.c @@ -931,15 +931,11 @@ static int serio_uevent(struct device *dev, struct kobj_uevent_env *env) #endif /* CONFIG_HOTPLUG */ #ifdef CONFIG_PM -static int serio_suspend(struct device *dev, pm_message_t state) +static int serio_suspend(struct device *dev) { struct serio *serio = to_serio_port(dev); - if (!serio->suspended && state.event == PM_EVENT_SUSPEND) - serio_cleanup(serio); - - serio->suspended = state.event == PM_EVENT_SUSPEND || - state.event == PM_EVENT_FREEZE; + serio_cleanup(serio); return 0; } @@ -952,13 +948,17 @@ static int serio_resume(struct device *dev) * Driver reconnect can take a while, so better let kseriod * deal with it. */ - if (serio->suspended) { - serio->suspended = false; - serio_queue_event(serio, NULL, SERIO_RECONNECT_PORT); - } + serio_queue_event(serio, NULL, SERIO_RECONNECT_PORT); return 0; } + +static const struct dev_pm_ops serio_pm_ops = { + .suspend = serio_suspend, + .resume = serio_resume, + .poweroff = serio_suspend, + .restore = serio_resume, +}; #endif /* CONFIG_PM */ /* called from serio_driver->connect/disconnect methods under serio_mutex */ @@ -1015,8 +1015,7 @@ static struct bus_type serio_bus = { .remove = serio_driver_remove, .shutdown = serio_shutdown, #ifdef CONFIG_PM - .suspend = serio_suspend, - .resume = serio_resume, + .pm = &serio_pm_ops, #endif }; diff --git a/include/linux/serio.h b/include/linux/serio.h index 126d24c9eaa8..a640bc2afe76 100644 --- a/include/linux/serio.h +++ b/include/linux/serio.h @@ -31,8 +31,6 @@ struct serio { bool manual_bind; bool registered; /* port has been fully registered with driver core */ - bool suspended; /* port is suspended */ - struct serio_device_id id; -- cgit v1.2.3 From dcf777f6ed9799c5ac90ac17a5c369e6b73ca92e Mon Sep 17 00:00:00 2001 From: Ralf Baechle Date: Sun, 26 Jul 2009 19:11:14 -0700 Subject: NET: ROSE: Don't use static buffer. The use of a static buffer in rose2asc() to return its result is not threadproof and can result in corruption if multiple threads are trying to use one of the procfs files based on rose2asc(). Signed-off-by: Ralf Baechle Signed-off-by: David S. Miller --- include/net/rose.h | 2 +- net/rose/af_rose.c | 18 ++++++++---------- net/rose/rose_route.c | 23 ++++++++++++----------- 3 files changed, 21 insertions(+), 22 deletions(-) (limited to 'include') diff --git a/include/net/rose.h b/include/net/rose.h index cbd5364b2c8a..5ba9f02731eb 100644 --- a/include/net/rose.h +++ b/include/net/rose.h @@ -156,7 +156,7 @@ extern int sysctl_rose_maximum_vcs; extern int sysctl_rose_window_size; extern int rosecmp(rose_address *, rose_address *); extern int rosecmpm(rose_address *, rose_address *, unsigned short); -extern const char *rose2asc(const rose_address *); +extern char *rose2asc(char *buf, const rose_address *); extern struct sock *rose_find_socket(unsigned int, struct rose_neigh *); extern void rose_kill_by_neigh(struct rose_neigh *); extern unsigned int rose_new_lci(struct rose_neigh *); diff --git a/net/rose/af_rose.c b/net/rose/af_rose.c index 6bd8e93869ed..f0a76f6bca71 100644 --- a/net/rose/af_rose.c +++ b/net/rose/af_rose.c @@ -92,23 +92,21 @@ static void rose_set_lockdep_key(struct net_device *dev) /* * Convert a ROSE address into text. */ -const char *rose2asc(const rose_address *addr) +char *rose2asc(char *buf, const rose_address *addr) { - static char buffer[11]; - if (addr->rose_addr[0] == 0x00 && addr->rose_addr[1] == 0x00 && addr->rose_addr[2] == 0x00 && addr->rose_addr[3] == 0x00 && addr->rose_addr[4] == 0x00) { - strcpy(buffer, "*"); + strcpy(buf, "*"); } else { - sprintf(buffer, "%02X%02X%02X%02X%02X", addr->rose_addr[0] & 0xFF, + sprintf(buf, "%02X%02X%02X%02X%02X", addr->rose_addr[0] & 0xFF, addr->rose_addr[1] & 0xFF, addr->rose_addr[2] & 0xFF, addr->rose_addr[3] & 0xFF, addr->rose_addr[4] & 0xFF); } - return buffer; + return buf; } /* @@ -1437,7 +1435,7 @@ static void rose_info_stop(struct seq_file *seq, void *v) static int rose_info_show(struct seq_file *seq, void *v) { - char buf[11]; + char buf[11], rsbuf[11]; if (v == SEQ_START_TOKEN) seq_puts(seq, @@ -1455,8 +1453,8 @@ static int rose_info_show(struct seq_file *seq, void *v) devname = dev->name; seq_printf(seq, "%-10s %-9s ", - rose2asc(&rose->dest_addr), - ax2asc(buf, &rose->dest_call)); + rose2asc(rsbuf, &rose->dest_addr), + ax2asc(buf, &rose->dest_call)); if (ax25cmp(&rose->source_call, &null_ax25_address) == 0) callsign = "??????-?"; @@ -1465,7 +1463,7 @@ static int rose_info_show(struct seq_file *seq, void *v) seq_printf(seq, "%-10s %-9s %-5s %3.3X %05d %d %d %d %d %3lu %3lu %3lu %3lu %3lu %3lu/%03lu %5d %5d %ld\n", - rose2asc(&rose->source_addr), + rose2asc(rsbuf, &rose->source_addr), callsign, devname, rose->lci & 0x0FFF, diff --git a/net/rose/rose_route.c b/net/rose/rose_route.c index a81066a1010a..9478d9b3d977 100644 --- a/net/rose/rose_route.c +++ b/net/rose/rose_route.c @@ -1104,6 +1104,7 @@ static void rose_node_stop(struct seq_file *seq, void *v) static int rose_node_show(struct seq_file *seq, void *v) { + char rsbuf[11]; int i; if (v == SEQ_START_TOKEN) @@ -1112,13 +1113,13 @@ static int rose_node_show(struct seq_file *seq, void *v) const struct rose_node *rose_node = v; /* if (rose_node->loopback) { seq_printf(seq, "%-10s %04d 1 loopback\n", - rose2asc(&rose_node->address), - rose_node->mask); + rose2asc(rsbuf, &rose_node->address), + rose_node->mask); } else { */ seq_printf(seq, "%-10s %04d %d", - rose2asc(&rose_node->address), - rose_node->mask, - rose_node->count); + rose2asc(rsbuf, &rose_node->address), + rose_node->mask, + rose_node->count); for (i = 0; i < rose_node->count; i++) seq_printf(seq, " %05d", @@ -1267,7 +1268,7 @@ static void rose_route_stop(struct seq_file *seq, void *v) static int rose_route_show(struct seq_file *seq, void *v) { - char buf[11]; + char buf[11], rsbuf[11]; if (v == SEQ_START_TOKEN) seq_puts(seq, @@ -1279,7 +1280,7 @@ static int rose_route_show(struct seq_file *seq, void *v) seq_printf(seq, "%3.3X %-10s %-9s %05d ", rose_route->lci1, - rose2asc(&rose_route->src_addr), + rose2asc(rsbuf, &rose_route->src_addr), ax2asc(buf, &rose_route->src_call), rose_route->neigh1->number); else @@ -1289,10 +1290,10 @@ static int rose_route_show(struct seq_file *seq, void *v) if (rose_route->neigh2) seq_printf(seq, "%3.3X %-10s %-9s %05d\n", - rose_route->lci2, - rose2asc(&rose_route->dest_addr), - ax2asc(buf, &rose_route->dest_call), - rose_route->neigh2->number); + rose_route->lci2, + rose2asc(rsbuf, &rose_route->dest_addr), + ax2asc(buf, &rose_route->dest_call), + rose_route->neigh2->number); else seq_puts(seq, "000 * * 00000\n"); -- cgit v1.2.3 From b5eb0589937eae2d58fca17fa45ed44152e772ed Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Wed, 15 Jul 2009 05:23:23 +0000 Subject: net: deprecate print_mac We've had %pM for long enough now, time to deprecate print_mac() and remove the __maybe_unused attribute from DECLARE_MAC_BUF so that variables declared with that can be found and removed. Otherwise people are putting in new users of print_mac(). Signed-off-by: Johannes Berg Signed-off-by: David S. Miller --- include/linux/if_ether.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/include/linux/if_ether.h b/include/linux/if_ether.h index 70fdba2bbf71..580b6004d00e 100644 --- a/include/linux/if_ether.h +++ b/include/linux/if_ether.h @@ -139,10 +139,10 @@ extern ssize_t sysfs_format_mac(char *buf, const unsigned char *addr, int len); /* * Display a 6 byte device address (MAC) in a readable format. */ -extern char *print_mac(char *buf, const unsigned char *addr); +extern char *print_mac(char *buf, const unsigned char *addr) __deprecated; #define MAC_FMT "%02x:%02x:%02x:%02x:%02x:%02x" #define MAC_BUF_SIZE 18 -#define DECLARE_MAC_BUF(var) char var[MAC_BUF_SIZE] __maybe_unused +#define DECLARE_MAC_BUF(var) char var[MAC_BUF_SIZE] #endif -- cgit v1.2.3 From 51def0bea92629dff02ff1de40603eb90c609c55 Mon Sep 17 00:00:00 2001 From: Tomas Winkler Date: Wed, 22 Jul 2009 14:06:56 +0000 Subject: imwc3200: move iwmc3200 SDIO ids to sdio_ids.h 1. add intel's sdio vendor id to sdio_ids.h 2. move iwmc3200 sdio devices' ids to sdio_ids.h Signed-off-by: Tomas Winkler Signed-off-by: David S. Miller --- drivers/net/wimax/i2400m/sdio.c | 12 +++++------- drivers/net/wireless/iwmc3200wifi/sdio.c | 4 +++- drivers/net/wireless/iwmc3200wifi/sdio.h | 3 --- include/linux/mmc/sdio_ids.h | 6 ++++++ 4 files changed, 14 insertions(+), 11 deletions(-) (limited to 'include') diff --git a/drivers/net/wimax/i2400m/sdio.c b/drivers/net/wimax/i2400m/sdio.c index 2538825d1c66..ea7b29034aab 100644 --- a/drivers/net/wimax/i2400m/sdio.c +++ b/drivers/net/wimax/i2400m/sdio.c @@ -58,6 +58,7 @@ */ #include +#include #include #include #include "i2400m-sdio.h" @@ -501,15 +502,12 @@ void i2400ms_remove(struct sdio_func *func) d_fnend(3, dev, "SDIO func %p\n", func); } -enum { - I2400MS_INTEL_VID = 0x89, -}; - static const struct sdio_device_id i2400ms_sdio_ids[] = { - /* Intel: i2400m WiMAX over SDIO */ - { SDIO_DEVICE(I2400MS_INTEL_VID, 0x1402) }, - { }, /* end: all zeroes */ + /* Intel: i2400m WiMAX (iwmc3200) over SDIO */ + { SDIO_DEVICE(SDIO_VENDOR_ID_INTEL, + SDIO_DEVICE_ID_INTEL_IWMC3200WIMAX) }, + { /* end: all zeroes */ }, }; MODULE_DEVICE_TABLE(sdio, i2400ms_sdio_ids); diff --git a/drivers/net/wireless/iwmc3200wifi/sdio.c b/drivers/net/wireless/iwmc3200wifi/sdio.c index b93f620ee4f1..8b1de84003ca 100644 --- a/drivers/net/wireless/iwmc3200wifi/sdio.c +++ b/drivers/net/wireless/iwmc3200wifi/sdio.c @@ -65,6 +65,7 @@ #include #include #include +#include #include #include @@ -492,7 +493,8 @@ static void iwm_sdio_remove(struct sdio_func *func) } static const struct sdio_device_id iwm_sdio_ids[] = { - { SDIO_DEVICE(SDIO_VENDOR_ID_INTEL, SDIO_DEVICE_ID_IWM) }, + { SDIO_DEVICE(SDIO_VENDOR_ID_INTEL, + SDIO_DEVICE_ID_INTEL_IWMC3200WIFI) }, { /* end: all zeroes */ }, }; MODULE_DEVICE_TABLE(sdio, iwm_sdio_ids); diff --git a/drivers/net/wireless/iwmc3200wifi/sdio.h b/drivers/net/wireless/iwmc3200wifi/sdio.h index b3c156b08dda..aab6b6892e45 100644 --- a/drivers/net/wireless/iwmc3200wifi/sdio.h +++ b/drivers/net/wireless/iwmc3200wifi/sdio.h @@ -39,9 +39,6 @@ #ifndef __IWM_SDIO_H__ #define __IWM_SDIO_H__ -#define SDIO_VENDOR_ID_INTEL 0x89 -#define SDIO_DEVICE_ID_IWM 0x1403 - #define IWM_SDIO_DATA_ADDR 0x0 #define IWM_SDIO_INTR_ENABLE_ADDR 0x14 #define IWM_SDIO_INTR_STATUS_ADDR 0x13 diff --git a/include/linux/mmc/sdio_ids.h b/include/linux/mmc/sdio_ids.h index 39751c8cde9c..2dbfb5a05994 100644 --- a/include/linux/mmc/sdio_ids.h +++ b/include/linux/mmc/sdio_ids.h @@ -22,6 +22,12 @@ /* * Vendors and devices. Sort key: vendor first, device next. */ +#define SDIO_VENDOR_ID_INTEL 0x0089 +#define SDIO_DEVICE_ID_INTEL_IWMC3200WIMAX 0x1402 +#define SDIO_DEVICE_ID_INTEL_IWMC3200WIFI 0x1403 +#define SDIO_DEVICE_ID_INTEL_IWMC3200TOP 0x1404 +#define SDIO_DEVICE_ID_INTEL_IWMC3200GPS 0x1405 +#define SDIO_DEVICE_ID_INTEL_IWMC3200BT 0x1406 #define SDIO_VENDOR_ID_MARVELL 0x02df #define SDIO_DEVICE_ID_MARVELL_LIBERTAS 0x9103 -- cgit v1.2.3 From 8a729fce76f7af50d8b622f2fb26adce9c8df743 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Sun, 26 Jul 2009 23:18:11 +0000 Subject: net: ethtool_op_get_rx_csum() should be public and exported Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- include/linux/ethtool.h | 1 + net/core/ethtool.c | 3 ++- 2 files changed, 3 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/ethtool.h b/include/linux/ethtool.h index 9b660bd2e2b3..90c4a3616d94 100644 --- a/include/linux/ethtool.h +++ b/include/linux/ethtool.h @@ -368,6 +368,7 @@ struct net_device; /* Some generic methods drivers may use in their ethtool_ops */ u32 ethtool_op_get_link(struct net_device *dev); +u32 ethtool_op_get_rx_csum(struct net_device *dev); u32 ethtool_op_get_tx_csum(struct net_device *dev); int ethtool_op_set_tx_csum(struct net_device *dev, u32 data); int ethtool_op_set_tx_hw_csum(struct net_device *dev, u32 data); diff --git a/net/core/ethtool.c b/net/core/ethtool.c index cf36ff44ebb2..44e571111d3a 100644 --- a/net/core/ethtool.c +++ b/net/core/ethtool.c @@ -34,11 +34,13 @@ u32 ethtool_op_get_rx_csum(struct net_device *dev) { return (dev->features & NETIF_F_ALL_CSUM) != 0; } +EXPORT_SYMBOL(ethtool_op_get_rx_csum); u32 ethtool_op_get_tx_csum(struct net_device *dev) { return (dev->features & NETIF_F_ALL_CSUM) != 0; } +EXPORT_SYMBOL(ethtool_op_get_tx_csum); int ethtool_op_set_tx_csum(struct net_device *dev, u32 data) { @@ -1125,7 +1127,6 @@ int dev_ethtool(struct net *net, struct ifreq *ifr) EXPORT_SYMBOL(ethtool_op_get_link); EXPORT_SYMBOL(ethtool_op_get_sg); EXPORT_SYMBOL(ethtool_op_get_tso); -EXPORT_SYMBOL(ethtool_op_get_tx_csum); EXPORT_SYMBOL(ethtool_op_set_sg); EXPORT_SYMBOL(ethtool_op_set_tso); EXPORT_SYMBOL(ethtool_op_set_tx_csum); -- cgit v1.2.3 From 9e1b32caa525cb236e80e9c671e179bcecccc657 Mon Sep 17 00:00:00 2001 From: Benjamin Herrenschmidt Date: Wed, 22 Jul 2009 15:44:28 +1000 Subject: mm: Pass virtual address to [__]p{te,ud,md}_free_tlb() mm: Pass virtual address to [__]p{te,ud,md}_free_tlb() Upcoming paches to support the new 64-bit "BookE" powerpc architecture will need to have the virtual address corresponding to PTE page when freeing it, due to the way the HW table walker works. Basically, the TLB can be loaded with "large" pages that cover the whole virtual space (well, sort-of, half of it actually) represented by a PTE page, and which contain an "indirect" bit indicating that this TLB entry RPN points to an array of PTEs from which the TLB can then create direct entries. Thus, in order to invalidate those when PTE pages are deleted, we need the virtual address to pass to tlbilx or tlbivax instructions. The old trick of sticking it somewhere in the PTE page struct page sucks too much, the address is almost readily available in all call sites and almost everybody implemets these as macros, so we may as well add the argument everywhere. I added it to the pmd and pud variants for consistency. Signed-off-by: Benjamin Herrenschmidt Acked-by: David Howells [MN10300 & FRV] Acked-by: Nick Piggin Acked-by: Martin Schwidefsky [s390] Signed-off-by: Linus Torvalds --- arch/alpha/include/asm/tlb.h | 4 ++-- arch/arm/include/asm/tlb.h | 4 ++-- arch/avr32/include/asm/pgalloc.h | 2 +- arch/cris/include/asm/pgalloc.h | 2 +- arch/frv/include/asm/pgalloc.h | 4 ++-- arch/frv/include/asm/pgtable.h | 2 +- arch/ia64/include/asm/pgalloc.h | 6 +++--- arch/ia64/include/asm/tlb.h | 12 ++++++------ arch/m32r/include/asm/pgalloc.h | 4 ++-- arch/m68k/include/asm/motorola_pgalloc.h | 6 ++++-- arch/m68k/include/asm/sun3_pgalloc.h | 4 ++-- arch/microblaze/include/asm/pgalloc.h | 4 ++-- arch/mips/include/asm/pgalloc.h | 6 +++--- arch/mn10300/include/asm/pgalloc.h | 2 +- arch/parisc/include/asm/tlb.h | 4 ++-- arch/powerpc/include/asm/pgalloc-32.h | 2 +- arch/powerpc/include/asm/pgalloc-64.h | 4 ++-- arch/powerpc/include/asm/pgalloc.h | 6 +++--- arch/powerpc/mm/hugetlbpage.c | 4 ++-- arch/s390/include/asm/tlb.h | 9 ++++++--- arch/sh/include/asm/pgalloc.h | 4 ++-- arch/sh/include/asm/tlb.h | 6 +++--- arch/sparc/include/asm/pgalloc_32.h | 8 ++++---- arch/sparc/include/asm/tlb_64.h | 6 +++--- arch/um/include/asm/pgalloc.h | 4 ++-- arch/um/include/asm/tlb.h | 6 +++--- arch/x86/include/asm/pgalloc.h | 25 ++++++++++++++++++++++--- arch/x86/mm/pgtable.c | 6 +++--- arch/xtensa/include/asm/tlb.h | 2 +- include/asm-generic/4level-fixup.h | 4 ++-- include/asm-generic/pgtable-nopmd.h | 2 +- include/asm-generic/pgtable-nopud.h | 2 +- include/asm-generic/tlb.h | 12 ++++++------ mm/memory.c | 11 ++++++----- 34 files changed, 107 insertions(+), 82 deletions(-) (limited to 'include') diff --git a/arch/alpha/include/asm/tlb.h b/arch/alpha/include/asm/tlb.h index c13636575fba..42866759f3fa 100644 --- a/arch/alpha/include/asm/tlb.h +++ b/arch/alpha/include/asm/tlb.h @@ -9,7 +9,7 @@ #include -#define __pte_free_tlb(tlb, pte) pte_free((tlb)->mm, pte) -#define __pmd_free_tlb(tlb, pmd) pmd_free((tlb)->mm, pmd) +#define __pte_free_tlb(tlb, pte, address) pte_free((tlb)->mm, pte) +#define __pmd_free_tlb(tlb, pmd, address) pmd_free((tlb)->mm, pmd) #endif diff --git a/arch/arm/include/asm/tlb.h b/arch/arm/include/asm/tlb.h index 321c83e43a1e..f41a6f57cd12 100644 --- a/arch/arm/include/asm/tlb.h +++ b/arch/arm/include/asm/tlb.h @@ -102,8 +102,8 @@ tlb_end_vma(struct mmu_gather *tlb, struct vm_area_struct *vma) } #define tlb_remove_page(tlb,page) free_page_and_swap_cache(page) -#define pte_free_tlb(tlb, ptep) pte_free((tlb)->mm, ptep) -#define pmd_free_tlb(tlb, pmdp) pmd_free((tlb)->mm, pmdp) +#define pte_free_tlb(tlb, ptep, addr) pte_free((tlb)->mm, ptep) +#define pmd_free_tlb(tlb, pmdp, addr) pmd_free((tlb)->mm, pmdp) #define tlb_migrate_finish(mm) do { } while (0) diff --git a/arch/avr32/include/asm/pgalloc.h b/arch/avr32/include/asm/pgalloc.h index 640821323943..92ecd8446ef8 100644 --- a/arch/avr32/include/asm/pgalloc.h +++ b/arch/avr32/include/asm/pgalloc.h @@ -83,7 +83,7 @@ static inline void pte_free(struct mm_struct *mm, pgtable_t pte) quicklist_free_page(QUICK_PT, NULL, pte); } -#define __pte_free_tlb(tlb,pte) \ +#define __pte_free_tlb(tlb,pte,addr) \ do { \ pgtable_page_dtor(pte); \ tlb_remove_page((tlb), pte); \ diff --git a/arch/cris/include/asm/pgalloc.h b/arch/cris/include/asm/pgalloc.h index a1ba761d0573..6da975db112f 100644 --- a/arch/cris/include/asm/pgalloc.h +++ b/arch/cris/include/asm/pgalloc.h @@ -47,7 +47,7 @@ static inline void pte_free(struct mm_struct *mm, pgtable_t pte) __free_page(pte); } -#define __pte_free_tlb(tlb,pte) \ +#define __pte_free_tlb(tlb,pte,address) \ do { \ pgtable_page_dtor(pte); \ tlb_remove_page((tlb), pte); \ diff --git a/arch/frv/include/asm/pgalloc.h b/arch/frv/include/asm/pgalloc.h index 971e6addb009..416d19a632f2 100644 --- a/arch/frv/include/asm/pgalloc.h +++ b/arch/frv/include/asm/pgalloc.h @@ -49,7 +49,7 @@ static inline void pte_free(struct mm_struct *mm, pgtable_t pte) __free_page(pte); } -#define __pte_free_tlb(tlb,pte) \ +#define __pte_free_tlb(tlb,pte,address) \ do { \ pgtable_page_dtor(pte); \ tlb_remove_page((tlb),(pte)); \ @@ -62,7 +62,7 @@ do { \ */ #define pmd_alloc_one(mm, addr) ({ BUG(); ((pmd_t *) 2); }) #define pmd_free(mm, x) do { } while (0) -#define __pmd_free_tlb(tlb,x) do { } while (0) +#define __pmd_free_tlb(tlb,x,a) do { } while (0) #endif /* CONFIG_MMU */ diff --git a/arch/frv/include/asm/pgtable.h b/arch/frv/include/asm/pgtable.h index 33233011b1c1..22c60692b551 100644 --- a/arch/frv/include/asm/pgtable.h +++ b/arch/frv/include/asm/pgtable.h @@ -225,7 +225,7 @@ static inline pud_t *pud_offset(pgd_t *pgd, unsigned long address) */ #define pud_alloc_one(mm, address) NULL #define pud_free(mm, x) do { } while (0) -#define __pud_free_tlb(tlb, x) do { } while (0) +#define __pud_free_tlb(tlb, x, address) do { } while (0) /* * The "pud_xxx()" functions here are trivial for a folded two-level diff --git a/arch/ia64/include/asm/pgalloc.h b/arch/ia64/include/asm/pgalloc.h index b9ac1a6fc216..96a8d927db28 100644 --- a/arch/ia64/include/asm/pgalloc.h +++ b/arch/ia64/include/asm/pgalloc.h @@ -48,7 +48,7 @@ static inline void pud_free(struct mm_struct *mm, pud_t *pud) { quicklist_free(0, NULL, pud); } -#define __pud_free_tlb(tlb, pud) pud_free((tlb)->mm, pud) +#define __pud_free_tlb(tlb, pud, address) pud_free((tlb)->mm, pud) #endif /* CONFIG_PGTABLE_4 */ static inline void @@ -67,7 +67,7 @@ static inline void pmd_free(struct mm_struct *mm, pmd_t *pmd) quicklist_free(0, NULL, pmd); } -#define __pmd_free_tlb(tlb, pmd) pmd_free((tlb)->mm, pmd) +#define __pmd_free_tlb(tlb, pmd, address) pmd_free((tlb)->mm, pmd) static inline void pmd_populate(struct mm_struct *mm, pmd_t * pmd_entry, pgtable_t pte) @@ -117,6 +117,6 @@ static inline void check_pgt_cache(void) quicklist_trim(0, NULL, 25, 16); } -#define __pte_free_tlb(tlb, pte) pte_free((tlb)->mm, pte) +#define __pte_free_tlb(tlb, pte, address) pte_free((tlb)->mm, pte) #endif /* _ASM_IA64_PGALLOC_H */ diff --git a/arch/ia64/include/asm/tlb.h b/arch/ia64/include/asm/tlb.h index 20d8a39680c2..85d965cb19a0 100644 --- a/arch/ia64/include/asm/tlb.h +++ b/arch/ia64/include/asm/tlb.h @@ -236,22 +236,22 @@ do { \ __tlb_remove_tlb_entry(tlb, ptep, addr); \ } while (0) -#define pte_free_tlb(tlb, ptep) \ +#define pte_free_tlb(tlb, ptep, address) \ do { \ tlb->need_flush = 1; \ - __pte_free_tlb(tlb, ptep); \ + __pte_free_tlb(tlb, ptep, address); \ } while (0) -#define pmd_free_tlb(tlb, ptep) \ +#define pmd_free_tlb(tlb, ptep, address) \ do { \ tlb->need_flush = 1; \ - __pmd_free_tlb(tlb, ptep); \ + __pmd_free_tlb(tlb, ptep, address); \ } while (0) -#define pud_free_tlb(tlb, pudp) \ +#define pud_free_tlb(tlb, pudp, address) \ do { \ tlb->need_flush = 1; \ - __pud_free_tlb(tlb, pudp); \ + __pud_free_tlb(tlb, pudp, address); \ } while (0) #endif /* _ASM_IA64_TLB_H */ diff --git a/arch/m32r/include/asm/pgalloc.h b/arch/m32r/include/asm/pgalloc.h index f11a2b909cdb..0fc736198979 100644 --- a/arch/m32r/include/asm/pgalloc.h +++ b/arch/m32r/include/asm/pgalloc.h @@ -58,7 +58,7 @@ static inline void pte_free(struct mm_struct *mm, pgtable_t pte) __free_page(pte); } -#define __pte_free_tlb(tlb, pte) pte_free((tlb)->mm, (pte)) +#define __pte_free_tlb(tlb, pte, addr) pte_free((tlb)->mm, (pte)) /* * allocating and freeing a pmd is trivial: the 1-entry pmd is @@ -68,7 +68,7 @@ static inline void pte_free(struct mm_struct *mm, pgtable_t pte) #define pmd_alloc_one(mm, addr) ({ BUG(); ((pmd_t *)2); }) #define pmd_free(mm, x) do { } while (0) -#define __pmd_free_tlb(tlb, x) do { } while (0) +#define __pmd_free_tlb(tlb, x, addr) do { } while (0) #define pgd_populate(mm, pmd, pte) BUG() #define check_pgt_cache() do { } while (0) diff --git a/arch/m68k/include/asm/motorola_pgalloc.h b/arch/m68k/include/asm/motorola_pgalloc.h index d08bf6261df8..15ee4c74a9f0 100644 --- a/arch/m68k/include/asm/motorola_pgalloc.h +++ b/arch/m68k/include/asm/motorola_pgalloc.h @@ -54,7 +54,8 @@ static inline void pte_free(struct mm_struct *mm, pgtable_t page) __free_page(page); } -static inline void __pte_free_tlb(struct mmu_gather *tlb, pgtable_t page) +static inline void __pte_free_tlb(struct mmu_gather *tlb, pgtable_t page, + unsigned long address) { pgtable_page_dtor(page); cache_page(kmap(page)); @@ -73,7 +74,8 @@ static inline int pmd_free(struct mm_struct *mm, pmd_t *pmd) return free_pointer_table(pmd); } -static inline int __pmd_free_tlb(struct mmu_gather *tlb, pmd_t *pmd) +static inline int __pmd_free_tlb(struct mmu_gather *tlb, pmd_t *pmd, + unsigned long address) { return free_pointer_table(pmd); } diff --git a/arch/m68k/include/asm/sun3_pgalloc.h b/arch/m68k/include/asm/sun3_pgalloc.h index d4c83f143816..48d80d5a666f 100644 --- a/arch/m68k/include/asm/sun3_pgalloc.h +++ b/arch/m68k/include/asm/sun3_pgalloc.h @@ -32,7 +32,7 @@ static inline void pte_free(struct mm_struct *mm, pgtable_t page) __free_page(page); } -#define __pte_free_tlb(tlb,pte) \ +#define __pte_free_tlb(tlb,pte,addr) \ do { \ pgtable_page_dtor(pte); \ tlb_remove_page((tlb), pte); \ @@ -80,7 +80,7 @@ static inline void pmd_populate(struct mm_struct *mm, pmd_t *pmd, pgtable_t page * inside the pgd, so has no extra memory associated with it. */ #define pmd_free(mm, x) do { } while (0) -#define __pmd_free_tlb(tlb, x) do { } while (0) +#define __pmd_free_tlb(tlb, x, addr) do { } while (0) static inline void pgd_free(struct mm_struct *mm, pgd_t *pgd) { diff --git a/arch/microblaze/include/asm/pgalloc.h b/arch/microblaze/include/asm/pgalloc.h index 59a757e46ba5..b0131da1387b 100644 --- a/arch/microblaze/include/asm/pgalloc.h +++ b/arch/microblaze/include/asm/pgalloc.h @@ -180,7 +180,7 @@ extern inline void pte_free(struct mm_struct *mm, struct page *ptepage) __free_page(ptepage); } -#define __pte_free_tlb(tlb, pte) pte_free((tlb)->mm, (pte)) +#define __pte_free_tlb(tlb, pte, addr) pte_free((tlb)->mm, (pte)) #define pmd_populate(mm, pmd, pte) (pmd_val(*(pmd)) = page_address(pte)) @@ -193,7 +193,7 @@ extern inline void pte_free(struct mm_struct *mm, struct page *ptepage) */ #define pmd_alloc_one(mm, address) ({ BUG(); ((pmd_t *)2); }) /*#define pmd_free(mm, x) do { } while (0)*/ -#define __pmd_free_tlb(tlb, x) do { } while (0) +#define __pmd_free_tlb(tlb, x, addr) do { } while (0) #define pgd_populate(mm, pmd, pte) BUG() extern int do_check_pgt_cache(int, int); diff --git a/arch/mips/include/asm/pgalloc.h b/arch/mips/include/asm/pgalloc.h index 1275831dda29..f705735feefc 100644 --- a/arch/mips/include/asm/pgalloc.h +++ b/arch/mips/include/asm/pgalloc.h @@ -98,7 +98,7 @@ static inline void pte_free(struct mm_struct *mm, pgtable_t pte) __free_pages(pte, PTE_ORDER); } -#define __pte_free_tlb(tlb,pte) \ +#define __pte_free_tlb(tlb,pte,address) \ do { \ pgtable_page_dtor(pte); \ tlb_remove_page((tlb), pte); \ @@ -111,7 +111,7 @@ do { \ * inside the pgd, so has no extra memory associated with it. */ #define pmd_free(mm, x) do { } while (0) -#define __pmd_free_tlb(tlb, x) do { } while (0) +#define __pmd_free_tlb(tlb, x, addr) do { } while (0) #endif @@ -132,7 +132,7 @@ static inline void pmd_free(struct mm_struct *mm, pmd_t *pmd) free_pages((unsigned long)pmd, PMD_ORDER); } -#define __pmd_free_tlb(tlb, x) pmd_free((tlb)->mm, x) +#define __pmd_free_tlb(tlb, x, addr) pmd_free((tlb)->mm, x) #endif diff --git a/arch/mn10300/include/asm/pgalloc.h b/arch/mn10300/include/asm/pgalloc.h index ec057e1bd4cf..a19f11327cd8 100644 --- a/arch/mn10300/include/asm/pgalloc.h +++ b/arch/mn10300/include/asm/pgalloc.h @@ -51,6 +51,6 @@ static inline void pte_free(struct mm_struct *mm, struct page *pte) } -#define __pte_free_tlb(tlb, pte) tlb_remove_page((tlb), (pte)) +#define __pte_free_tlb(tlb, pte, addr) tlb_remove_page((tlb), (pte)) #endif /* _ASM_PGALLOC_H */ diff --git a/arch/parisc/include/asm/tlb.h b/arch/parisc/include/asm/tlb.h index 383b1db310ee..07924903989e 100644 --- a/arch/parisc/include/asm/tlb.h +++ b/arch/parisc/include/asm/tlb.h @@ -21,7 +21,7 @@ do { if (!(tlb)->fullmm) \ #include -#define __pmd_free_tlb(tlb, pmd) pmd_free((tlb)->mm, pmd) -#define __pte_free_tlb(tlb, pte) pte_free((tlb)->mm, pte) +#define __pmd_free_tlb(tlb, pmd, addr) pmd_free((tlb)->mm, pmd) +#define __pte_free_tlb(tlb, pte, addr) pte_free((tlb)->mm, pte) #endif diff --git a/arch/powerpc/include/asm/pgalloc-32.h b/arch/powerpc/include/asm/pgalloc-32.h index 0815eb40acae..c9500d666a1d 100644 --- a/arch/powerpc/include/asm/pgalloc-32.h +++ b/arch/powerpc/include/asm/pgalloc-32.h @@ -16,7 +16,7 @@ extern void pgd_free(struct mm_struct *mm, pgd_t *pgd); */ /* #define pmd_alloc_one(mm,address) ({ BUG(); ((pmd_t *)2); }) */ #define pmd_free(mm, x) do { } while (0) -#define __pmd_free_tlb(tlb,x) do { } while (0) +#define __pmd_free_tlb(tlb,x,a) do { } while (0) /* #define pgd_populate(mm, pmd, pte) BUG() */ #ifndef CONFIG_BOOKE diff --git a/arch/powerpc/include/asm/pgalloc-64.h b/arch/powerpc/include/asm/pgalloc-64.h index afda2bdd860f..e6f069c4f713 100644 --- a/arch/powerpc/include/asm/pgalloc-64.h +++ b/arch/powerpc/include/asm/pgalloc-64.h @@ -118,11 +118,11 @@ static inline void pgtable_free(pgtable_free_t pgf) kmem_cache_free(pgtable_cache[cachenum], p); } -#define __pmd_free_tlb(tlb, pmd) \ +#define __pmd_free_tlb(tlb, pmd,addr) \ pgtable_free_tlb(tlb, pgtable_free_cache(pmd, \ PMD_CACHE_NUM, PMD_TABLE_SIZE-1)) #ifndef CONFIG_PPC_64K_PAGES -#define __pud_free_tlb(tlb, pud) \ +#define __pud_free_tlb(tlb, pud, addr) \ pgtable_free_tlb(tlb, pgtable_free_cache(pud, \ PUD_CACHE_NUM, PUD_TABLE_SIZE-1)) #endif /* CONFIG_PPC_64K_PAGES */ diff --git a/arch/powerpc/include/asm/pgalloc.h b/arch/powerpc/include/asm/pgalloc.h index 5d8480265a77..1730e5e298d6 100644 --- a/arch/powerpc/include/asm/pgalloc.h +++ b/arch/powerpc/include/asm/pgalloc.h @@ -38,14 +38,14 @@ static inline pgtable_free_t pgtable_free_cache(void *p, int cachenum, extern void pgtable_free_tlb(struct mmu_gather *tlb, pgtable_free_t pgf); #ifdef CONFIG_SMP -#define __pte_free_tlb(tlb,ptepage) \ +#define __pte_free_tlb(tlb,ptepage,address) \ do { \ pgtable_page_dtor(ptepage); \ pgtable_free_tlb(tlb, pgtable_free_cache(page_address(ptepage), \ - PTE_NONCACHE_NUM, PTE_TABLE_SIZE-1)); \ + PTE_NONCACHE_NUM, PTE_TABLE_SIZE-1)); \ } while (0) #else -#define __pte_free_tlb(tlb, pte) pte_free((tlb)->mm, (pte)) +#define __pte_free_tlb(tlb, pte, address) pte_free((tlb)->mm, (pte)) #endif diff --git a/arch/powerpc/mm/hugetlbpage.c b/arch/powerpc/mm/hugetlbpage.c index 9920d6a7cf29..c46ef2ffa3d9 100644 --- a/arch/powerpc/mm/hugetlbpage.c +++ b/arch/powerpc/mm/hugetlbpage.c @@ -305,7 +305,7 @@ static void hugetlb_free_pmd_range(struct mmu_gather *tlb, pud_t *pud, pmd = pmd_offset(pud, start); pud_clear(pud); - pmd_free_tlb(tlb, pmd); + pmd_free_tlb(tlb, pmd, start); } static void hugetlb_free_pud_range(struct mmu_gather *tlb, pgd_t *pgd, @@ -348,7 +348,7 @@ static void hugetlb_free_pud_range(struct mmu_gather *tlb, pgd_t *pgd, pud = pud_offset(pgd, start); pgd_clear(pgd); - pud_free_tlb(tlb, pud); + pud_free_tlb(tlb, pud, start); } /* diff --git a/arch/s390/include/asm/tlb.h b/arch/s390/include/asm/tlb.h index 3d8a96d39d9d..81150b053689 100644 --- a/arch/s390/include/asm/tlb.h +++ b/arch/s390/include/asm/tlb.h @@ -96,7 +96,8 @@ static inline void tlb_remove_page(struct mmu_gather *tlb, struct page *page) * pte_free_tlb frees a pte table and clears the CRSTE for the * page table from the tlb. */ -static inline void pte_free_tlb(struct mmu_gather *tlb, pgtable_t pte) +static inline void pte_free_tlb(struct mmu_gather *tlb, pgtable_t pte, + unsigned long address) { if (!tlb->fullmm) { tlb->array[tlb->nr_ptes++] = pte; @@ -113,7 +114,8 @@ static inline void pte_free_tlb(struct mmu_gather *tlb, pgtable_t pte) * as the pgd. pmd_free_tlb checks the asce_limit against 2GB * to avoid the double free of the pmd in this case. */ -static inline void pmd_free_tlb(struct mmu_gather *tlb, pmd_t *pmd) +static inline void pmd_free_tlb(struct mmu_gather *tlb, pmd_t *pmd, + unsigned long address) { #ifdef __s390x__ if (tlb->mm->context.asce_limit <= (1UL << 31)) @@ -134,7 +136,8 @@ static inline void pmd_free_tlb(struct mmu_gather *tlb, pmd_t *pmd) * as the pgd. pud_free_tlb checks the asce_limit against 4TB * to avoid the double free of the pud in this case. */ -static inline void pud_free_tlb(struct mmu_gather *tlb, pud_t *pud) +static inline void pud_free_tlb(struct mmu_gather *tlb, pud_t *pud, + unsigned long address) { #ifdef __s390x__ if (tlb->mm->context.asce_limit <= (1UL << 42)) diff --git a/arch/sh/include/asm/pgalloc.h b/arch/sh/include/asm/pgalloc.h index 84dd2db7104c..89a482750a5b 100644 --- a/arch/sh/include/asm/pgalloc.h +++ b/arch/sh/include/asm/pgalloc.h @@ -73,7 +73,7 @@ static inline void pte_free(struct mm_struct *mm, pgtable_t pte) quicklist_free_page(QUICK_PT, NULL, pte); } -#define __pte_free_tlb(tlb,pte) \ +#define __pte_free_tlb(tlb,pte,addr) \ do { \ pgtable_page_dtor(pte); \ tlb_remove_page((tlb), (pte)); \ @@ -85,7 +85,7 @@ do { \ */ #define pmd_free(mm, x) do { } while (0) -#define __pmd_free_tlb(tlb,x) do { } while (0) +#define __pmd_free_tlb(tlb,x,addr) do { } while (0) static inline void check_pgt_cache(void) { diff --git a/arch/sh/include/asm/tlb.h b/arch/sh/include/asm/tlb.h index 9c16f737074a..da8fe7ab8728 100644 --- a/arch/sh/include/asm/tlb.h +++ b/arch/sh/include/asm/tlb.h @@ -91,9 +91,9 @@ tlb_end_vma(struct mmu_gather *tlb, struct vm_area_struct *vma) } #define tlb_remove_page(tlb,page) free_page_and_swap_cache(page) -#define pte_free_tlb(tlb, ptep) pte_free((tlb)->mm, ptep) -#define pmd_free_tlb(tlb, pmdp) pmd_free((tlb)->mm, pmdp) -#define pud_free_tlb(tlb, pudp) pud_free((tlb)->mm, pudp) +#define pte_free_tlb(tlb, ptep, addr) pte_free((tlb)->mm, ptep) +#define pmd_free_tlb(tlb, pmdp, addr) pmd_free((tlb)->mm, pmdp) +#define pud_free_tlb(tlb, pudp, addr) pud_free((tlb)->mm, pudp) #define tlb_migrate_finish(mm) do { } while (0) diff --git a/arch/sparc/include/asm/pgalloc_32.h b/arch/sparc/include/asm/pgalloc_32.h index 681582d26969..ca2b34456c4b 100644 --- a/arch/sparc/include/asm/pgalloc_32.h +++ b/arch/sparc/include/asm/pgalloc_32.h @@ -44,8 +44,8 @@ BTFIXUPDEF_CALL(pmd_t *, pmd_alloc_one, struct mm_struct *, unsigned long) BTFIXUPDEF_CALL(void, free_pmd_fast, pmd_t *) #define free_pmd_fast(pmd) BTFIXUP_CALL(free_pmd_fast)(pmd) -#define pmd_free(mm, pmd) free_pmd_fast(pmd) -#define __pmd_free_tlb(tlb, pmd) pmd_free((tlb)->mm, pmd) +#define pmd_free(mm, pmd) free_pmd_fast(pmd) +#define __pmd_free_tlb(tlb, pmd, addr) pmd_free((tlb)->mm, pmd) BTFIXUPDEF_CALL(void, pmd_populate, pmd_t *, struct page *) #define pmd_populate(MM, PMD, PTE) BTFIXUP_CALL(pmd_populate)(PMD, PTE) @@ -62,7 +62,7 @@ BTFIXUPDEF_CALL(void, free_pte_fast, pte_t *) #define pte_free_kernel(mm, pte) BTFIXUP_CALL(free_pte_fast)(pte) BTFIXUPDEF_CALL(void, pte_free, pgtable_t ) -#define pte_free(mm, pte) BTFIXUP_CALL(pte_free)(pte) -#define __pte_free_tlb(tlb, pte) pte_free((tlb)->mm, pte) +#define pte_free(mm, pte) BTFIXUP_CALL(pte_free)(pte) +#define __pte_free_tlb(tlb, pte, addr) pte_free((tlb)->mm, pte) #endif /* _SPARC_PGALLOC_H */ diff --git a/arch/sparc/include/asm/tlb_64.h b/arch/sparc/include/asm/tlb_64.h index ee38e731bfa6..dca406b9b6fc 100644 --- a/arch/sparc/include/asm/tlb_64.h +++ b/arch/sparc/include/asm/tlb_64.h @@ -100,9 +100,9 @@ static inline void tlb_remove_page(struct mmu_gather *mp, struct page *page) } #define tlb_remove_tlb_entry(mp,ptep,addr) do { } while (0) -#define pte_free_tlb(mp, ptepage) pte_free((mp)->mm, ptepage) -#define pmd_free_tlb(mp, pmdp) pmd_free((mp)->mm, pmdp) -#define pud_free_tlb(tlb,pudp) __pud_free_tlb(tlb,pudp) +#define pte_free_tlb(mp, ptepage, addr) pte_free((mp)->mm, ptepage) +#define pmd_free_tlb(mp, pmdp, addr) pmd_free((mp)->mm, pmdp) +#define pud_free_tlb(tlb,pudp, addr) __pud_free_tlb(tlb,pudp,addr) #define tlb_migrate_finish(mm) do { } while (0) #define tlb_start_vma(tlb, vma) do { } while (0) diff --git a/arch/um/include/asm/pgalloc.h b/arch/um/include/asm/pgalloc.h index 718984359f8c..32c8ce4e1515 100644 --- a/arch/um/include/asm/pgalloc.h +++ b/arch/um/include/asm/pgalloc.h @@ -40,7 +40,7 @@ static inline void pte_free(struct mm_struct *mm, pgtable_t pte) __free_page(pte); } -#define __pte_free_tlb(tlb,pte) \ +#define __pte_free_tlb(tlb,pte, address) \ do { \ pgtable_page_dtor(pte); \ tlb_remove_page((tlb),(pte)); \ @@ -53,7 +53,7 @@ static inline void pmd_free(struct mm_struct *mm, pmd_t *pmd) free_page((unsigned long)pmd); } -#define __pmd_free_tlb(tlb,x) tlb_remove_page((tlb),virt_to_page(x)) +#define __pmd_free_tlb(tlb,x, address) tlb_remove_page((tlb),virt_to_page(x)) #endif #define check_pgt_cache() do { } while (0) diff --git a/arch/um/include/asm/tlb.h b/arch/um/include/asm/tlb.h index 5240fa1c5e08..660caedac9eb 100644 --- a/arch/um/include/asm/tlb.h +++ b/arch/um/include/asm/tlb.h @@ -116,11 +116,11 @@ static inline void tlb_remove_page(struct mmu_gather *tlb, struct page *page) __tlb_remove_tlb_entry(tlb, ptep, address); \ } while (0) -#define pte_free_tlb(tlb, ptep) __pte_free_tlb(tlb, ptep) +#define pte_free_tlb(tlb, ptep, addr) __pte_free_tlb(tlb, ptep, addr) -#define pud_free_tlb(tlb, pudp) __pud_free_tlb(tlb, pudp) +#define pud_free_tlb(tlb, pudp, addr) __pud_free_tlb(tlb, pudp, addr) -#define pmd_free_tlb(tlb, pmdp) __pmd_free_tlb(tlb, pmdp) +#define pmd_free_tlb(tlb, pmdp, addr) __pmd_free_tlb(tlb, pmdp, addr) #define tlb_migrate_finish(mm) do {} while (0) diff --git a/arch/x86/include/asm/pgalloc.h b/arch/x86/include/asm/pgalloc.h index dd14c54ac718..0e8c2a0fd922 100644 --- a/arch/x86/include/asm/pgalloc.h +++ b/arch/x86/include/asm/pgalloc.h @@ -46,7 +46,13 @@ static inline void pte_free(struct mm_struct *mm, struct page *pte) __free_page(pte); } -extern void __pte_free_tlb(struct mmu_gather *tlb, struct page *pte); +extern void ___pte_free_tlb(struct mmu_gather *tlb, struct page *pte); + +static inline void __pte_free_tlb(struct mmu_gather *tlb, struct page *pte, + unsigned long address) +{ + ___pte_free_tlb(tlb, pte); +} static inline void pmd_populate_kernel(struct mm_struct *mm, pmd_t *pmd, pte_t *pte) @@ -78,7 +84,13 @@ static inline void pmd_free(struct mm_struct *mm, pmd_t *pmd) free_page((unsigned long)pmd); } -extern void __pmd_free_tlb(struct mmu_gather *tlb, pmd_t *pmd); +extern void ___pmd_free_tlb(struct mmu_gather *tlb, pmd_t *pmd); + +static inline void __pmd_free_tlb(struct mmu_gather *tlb, pmd_t *pmd, + unsigned long adddress) +{ + ___pmd_free_tlb(tlb, pmd); +} #ifdef CONFIG_X86_PAE extern void pud_populate(struct mm_struct *mm, pud_t *pudp, pmd_t *pmd); @@ -108,7 +120,14 @@ static inline void pud_free(struct mm_struct *mm, pud_t *pud) free_page((unsigned long)pud); } -extern void __pud_free_tlb(struct mmu_gather *tlb, pud_t *pud); +extern void ___pud_free_tlb(struct mmu_gather *tlb, pud_t *pud); + +static inline void __pud_free_tlb(struct mmu_gather *tlb, pud_t *pud, + unsigned long address) +{ + ___pud_free_tlb(tlb, pud); +} + #endif /* PAGETABLE_LEVELS > 3 */ #endif /* PAGETABLE_LEVELS > 2 */ diff --git a/arch/x86/mm/pgtable.c b/arch/x86/mm/pgtable.c index 8e43bdd45456..af8f9650058c 100644 --- a/arch/x86/mm/pgtable.c +++ b/arch/x86/mm/pgtable.c @@ -25,7 +25,7 @@ pgtable_t pte_alloc_one(struct mm_struct *mm, unsigned long address) return pte; } -void __pte_free_tlb(struct mmu_gather *tlb, struct page *pte) +void ___pte_free_tlb(struct mmu_gather *tlb, struct page *pte) { pgtable_page_dtor(pte); paravirt_release_pte(page_to_pfn(pte)); @@ -33,14 +33,14 @@ void __pte_free_tlb(struct mmu_gather *tlb, struct page *pte) } #if PAGETABLE_LEVELS > 2 -void __pmd_free_tlb(struct mmu_gather *tlb, pmd_t *pmd) +void ___pmd_free_tlb(struct mmu_gather *tlb, pmd_t *pmd) { paravirt_release_pmd(__pa(pmd) >> PAGE_SHIFT); tlb_remove_page(tlb, virt_to_page(pmd)); } #if PAGETABLE_LEVELS > 3 -void __pud_free_tlb(struct mmu_gather *tlb, pud_t *pud) +void ___pud_free_tlb(struct mmu_gather *tlb, pud_t *pud) { paravirt_release_pud(__pa(pud) >> PAGE_SHIFT); tlb_remove_page(tlb, virt_to_page(pud)); diff --git a/arch/xtensa/include/asm/tlb.h b/arch/xtensa/include/asm/tlb.h index 31c220faca02..0d766f9c1083 100644 --- a/arch/xtensa/include/asm/tlb.h +++ b/arch/xtensa/include/asm/tlb.h @@ -42,6 +42,6 @@ #include -#define __pte_free_tlb(tlb, pte) pte_free((tlb)->mm, pte) +#define __pte_free_tlb(tlb, pte, address) pte_free((tlb)->mm, pte) #endif /* _XTENSA_TLB_H */ diff --git a/include/asm-generic/4level-fixup.h b/include/asm-generic/4level-fixup.h index 9d40e879f99e..77ff547730af 100644 --- a/include/asm-generic/4level-fixup.h +++ b/include/asm-generic/4level-fixup.h @@ -27,9 +27,9 @@ #define pud_page_vaddr(pud) pgd_page_vaddr(pud) #undef pud_free_tlb -#define pud_free_tlb(tlb, x) do { } while (0) +#define pud_free_tlb(tlb, x, addr) do { } while (0) #define pud_free(mm, x) do { } while (0) -#define __pud_free_tlb(tlb, x) do { } while (0) +#define __pud_free_tlb(tlb, x, addr) do { } while (0) #undef pud_addr_end #define pud_addr_end(addr, end) (end) diff --git a/include/asm-generic/pgtable-nopmd.h b/include/asm-generic/pgtable-nopmd.h index a7cdc48e8b78..725612b793ce 100644 --- a/include/asm-generic/pgtable-nopmd.h +++ b/include/asm-generic/pgtable-nopmd.h @@ -59,7 +59,7 @@ static inline pmd_t * pmd_offset(pud_t * pud, unsigned long address) static inline void pmd_free(struct mm_struct *mm, pmd_t *pmd) { } -#define __pmd_free_tlb(tlb, x) do { } while (0) +#define __pmd_free_tlb(tlb, x, a) do { } while (0) #undef pmd_addr_end #define pmd_addr_end(addr, end) (end) diff --git a/include/asm-generic/pgtable-nopud.h b/include/asm-generic/pgtable-nopud.h index 87cf449a6df3..810431d8351b 100644 --- a/include/asm-generic/pgtable-nopud.h +++ b/include/asm-generic/pgtable-nopud.h @@ -52,7 +52,7 @@ static inline pud_t * pud_offset(pgd_t * pgd, unsigned long address) */ #define pud_alloc_one(mm, address) NULL #define pud_free(mm, x) do { } while (0) -#define __pud_free_tlb(tlb, x) do { } while (0) +#define __pud_free_tlb(tlb, x, a) do { } while (0) #undef pud_addr_end #define pud_addr_end(addr, end) (end) diff --git a/include/asm-generic/tlb.h b/include/asm-generic/tlb.h index f490e43a90b9..e43f9766259f 100644 --- a/include/asm-generic/tlb.h +++ b/include/asm-generic/tlb.h @@ -123,24 +123,24 @@ static inline void tlb_remove_page(struct mmu_gather *tlb, struct page *page) __tlb_remove_tlb_entry(tlb, ptep, address); \ } while (0) -#define pte_free_tlb(tlb, ptep) \ +#define pte_free_tlb(tlb, ptep, address) \ do { \ tlb->need_flush = 1; \ - __pte_free_tlb(tlb, ptep); \ + __pte_free_tlb(tlb, ptep, address); \ } while (0) #ifndef __ARCH_HAS_4LEVEL_HACK -#define pud_free_tlb(tlb, pudp) \ +#define pud_free_tlb(tlb, pudp, address) \ do { \ tlb->need_flush = 1; \ - __pud_free_tlb(tlb, pudp); \ + __pud_free_tlb(tlb, pudp, address); \ } while (0) #endif -#define pmd_free_tlb(tlb, pmdp) \ +#define pmd_free_tlb(tlb, pmdp, address) \ do { \ tlb->need_flush = 1; \ - __pmd_free_tlb(tlb, pmdp); \ + __pmd_free_tlb(tlb, pmdp, address); \ } while (0) #define tlb_migrate_finish(mm) do {} while (0) diff --git a/mm/memory.c b/mm/memory.c index 65216194eb8d..aede2ce3aba4 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -135,11 +135,12 @@ void pmd_clear_bad(pmd_t *pmd) * Note: this doesn't free the actual pages themselves. That * has been handled earlier when unmapping all the memory regions. */ -static void free_pte_range(struct mmu_gather *tlb, pmd_t *pmd) +static void free_pte_range(struct mmu_gather *tlb, pmd_t *pmd, + unsigned long addr) { pgtable_t token = pmd_pgtable(*pmd); pmd_clear(pmd); - pte_free_tlb(tlb, token); + pte_free_tlb(tlb, token, addr); tlb->mm->nr_ptes--; } @@ -157,7 +158,7 @@ static inline void free_pmd_range(struct mmu_gather *tlb, pud_t *pud, next = pmd_addr_end(addr, end); if (pmd_none_or_clear_bad(pmd)) continue; - free_pte_range(tlb, pmd); + free_pte_range(tlb, pmd, addr); } while (pmd++, addr = next, addr != end); start &= PUD_MASK; @@ -173,7 +174,7 @@ static inline void free_pmd_range(struct mmu_gather *tlb, pud_t *pud, pmd = pmd_offset(pud, start); pud_clear(pud); - pmd_free_tlb(tlb, pmd); + pmd_free_tlb(tlb, pmd, start); } static inline void free_pud_range(struct mmu_gather *tlb, pgd_t *pgd, @@ -206,7 +207,7 @@ static inline void free_pud_range(struct mmu_gather *tlb, pgd_t *pgd, pud = pud_offset(pgd, start); pgd_clear(pgd); - pud_free_tlb(tlb, pud); + pud_free_tlb(tlb, pud, start); } /* -- cgit v1.2.3 From 463d018323851a608eef52a9427b0585005c647f Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Tue, 14 Jul 2009 00:33:35 +0200 Subject: cfg80211: make aware of net namespaces In order to make cfg80211/nl80211 aware of network namespaces, we have to do the following things: * del_virtual_intf method takes an interface index rather than a netdev pointer - simply change this * nl80211 uses init_net a lot, it changes to use the sender's network namespace * scan requests use the interface index, hold a netdev pointer and reference instead * we want a wiphy and its associated virtual interfaces to be in one netns together, so - we need to be able to change ns for a given interface, so export dev_change_net_namespace() - for each virtual interface set the NETIF_F_NETNS_LOCAL flag, and clear that flag only when the wiphy changes ns, to disallow breaking this invariant * when a network namespace goes away, we need to reparent the wiphy to init_net * cfg80211 users that support creating virtual interfaces must create them in the wiphy's namespace, currently this affects only mac80211 The end result is that you can now switch an entire wiphy into a different network namespace with the new command iw phy# set netns and all virtual interfaces will follow (or the operation fails). Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- include/linux/nl80211.h | 9 +++ include/net/cfg80211.h | 40 +++++++++- net/core/dev.c | 1 + net/mac80211/cfg.c | 14 +--- net/wireless/core.c | 75 ++++++++++++++++-- net/wireless/core.h | 5 +- net/wireless/nl80211.c | 202 +++++++++++++++++++++++++++++++++--------------- net/wireless/scan.c | 22 ++---- net/wireless/sme.c | 3 +- 9 files changed, 272 insertions(+), 99 deletions(-) (limited to 'include') diff --git a/include/linux/nl80211.h b/include/linux/nl80211.h index 962e2232a074..cb3dc6027fd9 100644 --- a/include/linux/nl80211.h +++ b/include/linux/nl80211.h @@ -262,6 +262,9 @@ * reasons, for this the %NL80211_ATTR_DISCONNECTED_BY_AP and * %NL80211_ATTR_REASON_CODE attributes are used. * + * @NL80211_CMD_SET_WIPHY_NETNS: Set a wiphy's netns. Note that all devices + * associated with this wiphy must be down and will follow. + * * @NL80211_CMD_MAX: highest used command number * @__NL80211_CMD_AFTER_LAST: internal use */ @@ -336,6 +339,8 @@ enum nl80211_commands { NL80211_CMD_ROAM, NL80211_CMD_DISCONNECT, + NL80211_CMD_SET_WIPHY_NETNS, + /* add new commands above here */ /* used to define NL80211_CMD_MAX below */ @@ -573,6 +578,8 @@ enum nl80211_commands { * and join_ibss(), key information is in a nested attribute each * with %NL80211_KEY_* sub-attributes * + * @NL80211_ATTR_PID: Process ID of a network namespace. + * * @NL80211_ATTR_MAX: highest attribute number currently defined * @__NL80211_ATTR_AFTER_LAST: internal use */ @@ -701,6 +708,8 @@ enum nl80211_attrs { NL80211_ATTR_KEY, NL80211_ATTR_KEYS, + NL80211_ATTR_PID, + /* add attributes here, update the policy in nl80211.c */ __NL80211_ATTR_AFTER_LAST, diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index a981ca8a5701..0d278777e39c 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -542,7 +542,7 @@ struct cfg80211_ssid { * @ie: optional information element(s) to add into Probe Request or %NULL * @ie_len: length of ie in octets * @wiphy: the wiphy this was for - * @ifidx: the interface index + * @dev: the interface */ struct cfg80211_scan_request { struct cfg80211_ssid *ssids; @@ -554,7 +554,7 @@ struct cfg80211_scan_request { /* internal */ struct wiphy *wiphy; - int ifidx; + struct net_device *dev; bool aborted; }; @@ -845,7 +845,8 @@ struct cfg80211_bitrate_mask { * @resume: wiphy device needs to be resumed * * @add_virtual_intf: create a new virtual interface with the given name, - * must set the struct wireless_dev's iftype. + * must set the struct wireless_dev's iftype. Beware: You must create + * the new netdev in the wiphy's network namespace! * * @del_virtual_intf: remove the virtual interface determined by ifindex. * @@ -937,7 +938,7 @@ struct cfg80211_ops { int (*add_virtual_intf)(struct wiphy *wiphy, char *name, enum nl80211_iftype type, u32 *flags, struct vif_params *params); - int (*del_virtual_intf)(struct wiphy *wiphy, int ifindex); + int (*del_virtual_intf)(struct wiphy *wiphy, struct net_device *dev); int (*change_virtual_intf)(struct wiphy *wiphy, struct net_device *dev, enum nl80211_iftype type, u32 *flags, @@ -1088,6 +1089,9 @@ struct cfg80211_ops { * @frag_threshold: Fragmentation threshold (dot11FragmentationThreshold); * -1 = fragmentation disabled, only odd values >= 256 used * @rts_threshold: RTS threshold (dot11RTSThreshold); -1 = RTS/CTS disabled + * @net: the network namespace this wiphy currently lives in + * @netnsok: if set to false, do not allow changing the netns of this + * wiphy at all */ struct wiphy { /* assign these fields before you register the wiphy */ @@ -1101,6 +1105,8 @@ struct wiphy { bool custom_regulatory; bool strict_regulatory; + bool netnsok; + enum cfg80211_signal_type signal_type; int bss_priv_size; @@ -1139,9 +1145,35 @@ struct wiphy { /* dir in debugfs: ieee80211/ */ struct dentry *debugfsdir; +#ifdef CONFIG_NET_NS + /* the network namespace this phy lives in currently */ + struct net *_net; +#endif + char priv[0] __attribute__((__aligned__(NETDEV_ALIGN))); }; +#ifdef CONFIG_NET_NS +static inline struct net *wiphy_net(struct wiphy *wiphy) +{ + return wiphy->_net; +} + +static inline void wiphy_net_set(struct wiphy *wiphy, struct net *net) +{ + wiphy->_net = net; +} +#else +static inline struct net *wiphy_net(struct wiphy *wiphy) +{ + return &init_net; +} + +static inline void wiphy_net_set(struct wiphy *wiphy, struct net *net) +{ +} +#endif + /** * wiphy_priv - return priv from wiphy * diff --git a/net/core/dev.c b/net/core/dev.c index d6c657ee413d..71347668c506 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -5344,6 +5344,7 @@ int dev_change_net_namespace(struct net_device *dev, struct net *net, const char out: return err; } +EXPORT_SYMBOL_GPL(dev_change_net_namespace); static int dev_cpu_callback(struct notifier_block *nfb, unsigned long action, diff --git a/net/mac80211/cfg.c b/net/mac80211/cfg.c index 52928ad90570..4bbf5007799b 100644 --- a/net/mac80211/cfg.c +++ b/net/mac80211/cfg.c @@ -57,19 +57,9 @@ static int ieee80211_add_iface(struct wiphy *wiphy, char *name, return 0; } -static int ieee80211_del_iface(struct wiphy *wiphy, int ifindex) +static int ieee80211_del_iface(struct wiphy *wiphy, struct net_device *dev) { - struct net_device *dev; - struct ieee80211_sub_if_data *sdata; - - /* we're under RTNL */ - dev = __dev_get_by_index(&init_net, ifindex); - if (!dev) - return -ENODEV; - - sdata = IEEE80211_DEV_TO_SUB_IF(dev); - - ieee80211_if_remove(sdata); + ieee80211_if_remove(IEEE80211_DEV_TO_SUB_IF(dev)); return 0; } diff --git a/net/wireless/core.c b/net/wireless/core.c index 6891cd0e38d5..442c9f389799 100644 --- a/net/wireless/core.c +++ b/net/wireless/core.c @@ -106,7 +106,7 @@ __cfg80211_rdev_from_info(struct genl_info *info) if (info->attrs[NL80211_ATTR_IFINDEX]) { ifindex = nla_get_u32(info->attrs[NL80211_ATTR_IFINDEX]); - dev = dev_get_by_index(&init_net, ifindex); + dev = dev_get_by_index(genl_info_net(info), ifindex); if (dev) { if (dev->ieee80211_ptr) byifidx = @@ -151,13 +151,13 @@ cfg80211_get_dev_from_info(struct genl_info *info) } struct cfg80211_registered_device * -cfg80211_get_dev_from_ifindex(int ifindex) +cfg80211_get_dev_from_ifindex(struct net *net, int ifindex) { struct cfg80211_registered_device *rdev = ERR_PTR(-ENODEV); struct net_device *dev; mutex_lock(&cfg80211_mutex); - dev = dev_get_by_index(&init_net, ifindex); + dev = dev_get_by_index(net, ifindex); if (!dev) goto out; if (dev->ieee80211_ptr) { @@ -222,6 +222,42 @@ int cfg80211_dev_rename(struct cfg80211_registered_device *rdev, return 0; } +int cfg80211_switch_netns(struct cfg80211_registered_device *rdev, + struct net *net) +{ + struct wireless_dev *wdev; + int err = 0; + + if (!rdev->wiphy.netnsok) + return -EOPNOTSUPP; + + list_for_each_entry(wdev, &rdev->netdev_list, list) { + wdev->netdev->features &= ~NETIF_F_NETNS_LOCAL; + err = dev_change_net_namespace(wdev->netdev, net, "wlan%d"); + if (err) + break; + wdev->netdev->features |= NETIF_F_NETNS_LOCAL; + } + + if (err) { + /* failed -- clean up to old netns */ + net = wiphy_net(&rdev->wiphy); + + list_for_each_entry_continue_reverse(wdev, &rdev->netdev_list, + list) { + wdev->netdev->features &= ~NETIF_F_NETNS_LOCAL; + err = dev_change_net_namespace(wdev->netdev, net, + "wlan%d"); + WARN_ON(err); + wdev->netdev->features |= NETIF_F_NETNS_LOCAL; + } + } + + wiphy_net_set(&rdev->wiphy, net); + + return err; +} + static void cfg80211_rfkill_poll(struct rfkill *rfkill, void *data) { struct cfg80211_registered_device *rdev = data; @@ -375,6 +411,8 @@ struct wiphy *wiphy_new(const struct cfg80211_ops *ops, int sizeof_priv) rdev->wiphy.dev.class = &ieee80211_class; rdev->wiphy.dev.platform_data = rdev; + wiphy_net_set(&rdev->wiphy, &init_net); + rdev->rfkill_ops.set_block = cfg80211_rfkill_set_block; rdev->rfkill = rfkill_alloc(dev_name(&rdev->wiphy.dev), &rdev->wiphy.dev, RFKILL_TYPE_WLAN, @@ -615,6 +653,9 @@ static int cfg80211_netdev_notifier_call(struct notifier_block * nb, spin_lock_init(&wdev->event_lock); mutex_lock(&rdev->devlist_mtx); list_add(&wdev->list, &rdev->netdev_list); + /* can only change netns with wiphy */ + dev->features |= NETIF_F_NETNS_LOCAL; + if (sysfs_create_link(&dev->dev.kobj, &rdev->wiphy.dev.kobj, "phy80211")) { printk(KERN_ERR "wireless: failed to add phy80211 " @@ -705,10 +746,32 @@ static struct notifier_block cfg80211_netdev_notifier = { .notifier_call = cfg80211_netdev_notifier_call, }; -static int cfg80211_init(void) +static void __net_exit cfg80211_pernet_exit(struct net *net) +{ + struct cfg80211_registered_device *rdev; + + rtnl_lock(); + mutex_lock(&cfg80211_mutex); + list_for_each_entry(rdev, &cfg80211_rdev_list, list) { + if (net_eq(wiphy_net(&rdev->wiphy), net)) + WARN_ON(cfg80211_switch_netns(rdev, &init_net)); + } + mutex_unlock(&cfg80211_mutex); + rtnl_unlock(); +} + +static struct pernet_operations cfg80211_pernet_ops = { + .exit = cfg80211_pernet_exit, +}; + +static int __init cfg80211_init(void) { int err; + err = register_pernet_device(&cfg80211_pernet_ops); + if (err) + goto out_fail_pernet; + err = wiphy_sysfs_init(); if (err) goto out_fail_sysfs; @@ -736,9 +799,10 @@ out_fail_nl80211: out_fail_notifier: wiphy_sysfs_exit(); out_fail_sysfs: + unregister_pernet_device(&cfg80211_pernet_ops); +out_fail_pernet: return err; } - subsys_initcall(cfg80211_init); static void cfg80211_exit(void) @@ -748,5 +812,6 @@ static void cfg80211_exit(void) unregister_netdevice_notifier(&cfg80211_netdev_notifier); wiphy_sysfs_exit(); regulatory_exit(); + unregister_pernet_device(&cfg80211_pernet_ops); } module_exit(cfg80211_exit); diff --git a/net/wireless/core.h b/net/wireless/core.h index 2ec8ddbe57de..4276b70cd975 100644 --- a/net/wireless/core.h +++ b/net/wireless/core.h @@ -170,7 +170,10 @@ struct wiphy *wiphy_idx_to_wiphy(int wiphy_idx); /* identical to cfg80211_get_dev_from_info but only operate on ifindex */ extern struct cfg80211_registered_device * -cfg80211_get_dev_from_ifindex(int ifindex); +cfg80211_get_dev_from_ifindex(struct net *net, int ifindex); + +int cfg80211_switch_netns(struct cfg80211_registered_device *rdev, + struct net *net); static inline void cfg80211_lock_rdev(struct cfg80211_registered_device *rdev) { diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index da450ef1fc7e..7880a9c4cdda 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -14,8 +14,10 @@ #include #include #include +#include #include #include +#include #include "core.h" #include "nl80211.h" #include "reg.h" @@ -27,24 +29,26 @@ static struct genl_family nl80211_fam = { .hdrsize = 0, /* no private header */ .version = 1, /* no particular meaning now */ .maxattr = NL80211_ATTR_MAX, + .netnsok = true, }; /* internal helper: get rdev and dev */ -static int get_rdev_dev_by_info_ifindex(struct nlattr **attrs, +static int get_rdev_dev_by_info_ifindex(struct genl_info *info, struct cfg80211_registered_device **rdev, struct net_device **dev) { + struct nlattr **attrs = info->attrs; int ifindex; if (!attrs[NL80211_ATTR_IFINDEX]) return -EINVAL; ifindex = nla_get_u32(attrs[NL80211_ATTR_IFINDEX]); - *dev = dev_get_by_index(&init_net, ifindex); + *dev = dev_get_by_index(genl_info_net(info), ifindex); if (!*dev) return -ENODEV; - *rdev = cfg80211_get_dev_from_ifindex(ifindex); + *rdev = cfg80211_get_dev_from_ifindex(genl_info_net(info), ifindex); if (IS_ERR(*rdev)) { dev_put(*dev); return PTR_ERR(*rdev); @@ -133,6 +137,7 @@ static struct nla_policy nl80211_policy[NL80211_ATTR_MAX+1] __read_mostly = { [NL80211_ATTR_PRIVACY] = { .type = NLA_FLAG }, [NL80211_ATTR_CIPHER_SUITE_GROUP] = { .type = NLA_U32 }, [NL80211_ATTR_WPA_VERSIONS] = { .type = NLA_U32 }, + [NL80211_ATTR_PID] = { .type = NLA_U32 }, }; /* policy for the attributes */ @@ -532,6 +537,10 @@ static int nl80211_send_wiphy(struct sk_buff *msg, u32 pid, u32 seq, int flags, CMD(deauth, DEAUTHENTICATE); CMD(disassoc, DISASSOCIATE); CMD(join_ibss, JOIN_IBSS); + if (dev->wiphy.netnsok) { + i++; + NLA_PUT_U32(msg, i, NL80211_CMD_SET_WIPHY_NETNS); + } #undef CMD @@ -562,6 +571,8 @@ static int nl80211_dump_wiphy(struct sk_buff *skb, struct netlink_callback *cb) mutex_lock(&cfg80211_mutex); list_for_each_entry(dev, &cfg80211_rdev_list, list) { + if (!net_eq(wiphy_net(&dev->wiphy), sock_net(skb->sk))) + continue; if (++idx <= start) continue; if (nl80211_send_wiphy(skb, NETLINK_CB(cb->skb).pid, @@ -867,6 +878,8 @@ static int nl80211_dump_interface(struct sk_buff *skb, struct netlink_callback * mutex_lock(&cfg80211_mutex); list_for_each_entry(dev, &cfg80211_rdev_list, list) { + if (!net_eq(wiphy_net(&dev->wiphy), sock_net(skb->sk))) + continue; if (wp_idx < wp_start) { wp_idx++; continue; @@ -907,7 +920,7 @@ static int nl80211_get_interface(struct sk_buff *skb, struct genl_info *info) struct net_device *netdev; int err; - err = get_rdev_dev_by_info_ifindex(info->attrs, &dev, &netdev); + err = get_rdev_dev_by_info_ifindex(info, &dev, &netdev); if (err) return err; @@ -975,7 +988,7 @@ static int nl80211_set_interface(struct sk_buff *skb, struct genl_info *info) rtnl_lock(); - err = get_rdev_dev_by_info_ifindex(info->attrs, &rdev, &dev); + err = get_rdev_dev_by_info_ifindex(info, &rdev, &dev); if (err) goto unlock_rtnl; @@ -1098,26 +1111,25 @@ static int nl80211_new_interface(struct sk_buff *skb, struct genl_info *info) static int nl80211_del_interface(struct sk_buff *skb, struct genl_info *info) { struct cfg80211_registered_device *rdev; - int ifindex, err; + int err; struct net_device *dev; rtnl_lock(); - err = get_rdev_dev_by_info_ifindex(info->attrs, &rdev, &dev); + err = get_rdev_dev_by_info_ifindex(info, &rdev, &dev); if (err) goto unlock_rtnl; - ifindex = dev->ifindex; - dev_put(dev); if (!rdev->ops->del_virtual_intf) { err = -EOPNOTSUPP; goto out; } - err = rdev->ops->del_virtual_intf(&rdev->wiphy, ifindex); + err = rdev->ops->del_virtual_intf(&rdev->wiphy, dev); out: cfg80211_unlock_rdev(rdev); + dev_put(dev); unlock_rtnl: rtnl_unlock(); return err; @@ -1195,7 +1207,7 @@ static int nl80211_get_key(struct sk_buff *skb, struct genl_info *info) rtnl_lock(); - err = get_rdev_dev_by_info_ifindex(info->attrs, &rdev, &dev); + err = get_rdev_dev_by_info_ifindex(info, &rdev, &dev); if (err) goto unlock_rtnl; @@ -1274,7 +1286,7 @@ static int nl80211_set_key(struct sk_buff *skb, struct genl_info *info) rtnl_lock(); - err = get_rdev_dev_by_info_ifindex(info->attrs, &rdev, &dev); + err = get_rdev_dev_by_info_ifindex(info, &rdev, &dev); if (err) goto unlock_rtnl; @@ -1333,7 +1345,7 @@ static int nl80211_new_key(struct sk_buff *skb, struct genl_info *info) rtnl_lock(); - err = get_rdev_dev_by_info_ifindex(info->attrs, &rdev, &dev); + err = get_rdev_dev_by_info_ifindex(info, &rdev, &dev); if (err) goto unlock_rtnl; @@ -1380,7 +1392,7 @@ static int nl80211_del_key(struct sk_buff *skb, struct genl_info *info) rtnl_lock(); - err = get_rdev_dev_by_info_ifindex(info->attrs, &rdev, &dev); + err = get_rdev_dev_by_info_ifindex(info, &rdev, &dev); if (err) goto unlock_rtnl; @@ -1429,7 +1441,7 @@ static int nl80211_addset_beacon(struct sk_buff *skb, struct genl_info *info) rtnl_lock(); - err = get_rdev_dev_by_info_ifindex(info->attrs, &rdev, &dev); + err = get_rdev_dev_by_info_ifindex(info, &rdev, &dev); if (err) goto unlock_rtnl; @@ -1516,7 +1528,7 @@ static int nl80211_del_beacon(struct sk_buff *skb, struct genl_info *info) rtnl_lock(); - err = get_rdev_dev_by_info_ifindex(info->attrs, &rdev, &dev); + err = get_rdev_dev_by_info_ifindex(info, &rdev, &dev); if (err) goto unlock_rtnl; @@ -1726,13 +1738,13 @@ static int nl80211_dump_station(struct sk_buff *skb, rtnl_lock(); - netdev = __dev_get_by_index(&init_net, ifidx); + netdev = __dev_get_by_index(sock_net(skb->sk), ifidx); if (!netdev) { err = -ENODEV; goto out_rtnl; } - dev = cfg80211_get_dev_from_ifindex(ifidx); + dev = cfg80211_get_dev_from_ifindex(sock_net(skb->sk), ifidx); if (IS_ERR(dev)) { err = PTR_ERR(dev); goto out_rtnl; @@ -1791,7 +1803,7 @@ static int nl80211_get_station(struct sk_buff *skb, struct genl_info *info) rtnl_lock(); - err = get_rdev_dev_by_info_ifindex(info->attrs, &rdev, &dev); + err = get_rdev_dev_by_info_ifindex(info, &rdev, &dev); if (err) goto out_rtnl; @@ -1829,14 +1841,16 @@ static int nl80211_get_station(struct sk_buff *skb, struct genl_info *info) /* * Get vlan interface making sure it is on the right wiphy. */ -static int get_vlan(struct nlattr *vlanattr, +static int get_vlan(struct genl_info *info, struct cfg80211_registered_device *rdev, struct net_device **vlan) { + struct nlattr *vlanattr = info->attrs[NL80211_ATTR_STA_VLAN]; *vlan = NULL; if (vlanattr) { - *vlan = dev_get_by_index(&init_net, nla_get_u32(vlanattr)); + *vlan = dev_get_by_index(genl_info_net(info), + nla_get_u32(vlanattr)); if (!*vlan) return -ENODEV; if (!(*vlan)->ieee80211_ptr) @@ -1891,11 +1905,11 @@ static int nl80211_set_station(struct sk_buff *skb, struct genl_info *info) rtnl_lock(); - err = get_rdev_dev_by_info_ifindex(info->attrs, &rdev, &dev); + err = get_rdev_dev_by_info_ifindex(info, &rdev, &dev); if (err) goto out_rtnl; - err = get_vlan(info->attrs[NL80211_ATTR_STA_VLAN], rdev, ¶ms.vlan); + err = get_vlan(info, rdev, ¶ms.vlan); if (err) goto out; @@ -2004,11 +2018,11 @@ static int nl80211_new_station(struct sk_buff *skb, struct genl_info *info) rtnl_lock(); - err = get_rdev_dev_by_info_ifindex(info->attrs, &rdev, &dev); + err = get_rdev_dev_by_info_ifindex(info, &rdev, &dev); if (err) goto out_rtnl; - err = get_vlan(info->attrs[NL80211_ATTR_STA_VLAN], rdev, ¶ms.vlan); + err = get_vlan(info, rdev, ¶ms.vlan); if (err) goto out; @@ -2079,7 +2093,7 @@ static int nl80211_del_station(struct sk_buff *skb, struct genl_info *info) rtnl_lock(); - err = get_rdev_dev_by_info_ifindex(info->attrs, &rdev, &dev); + err = get_rdev_dev_by_info_ifindex(info, &rdev, &dev); if (err) goto out_rtnl; @@ -2185,13 +2199,13 @@ static int nl80211_dump_mpath(struct sk_buff *skb, rtnl_lock(); - netdev = __dev_get_by_index(&init_net, ifidx); + netdev = __dev_get_by_index(sock_net(skb->sk), ifidx); if (!netdev) { err = -ENODEV; goto out_rtnl; } - dev = cfg80211_get_dev_from_ifindex(ifidx); + dev = cfg80211_get_dev_from_ifindex(sock_net(skb->sk), ifidx); if (IS_ERR(dev)) { err = PTR_ERR(dev); goto out_rtnl; @@ -2255,7 +2269,7 @@ static int nl80211_get_mpath(struct sk_buff *skb, struct genl_info *info) rtnl_lock(); - err = get_rdev_dev_by_info_ifindex(info->attrs, &rdev, &dev); + err = get_rdev_dev_by_info_ifindex(info, &rdev, &dev); if (err) goto out_rtnl; @@ -2314,7 +2328,7 @@ static int nl80211_set_mpath(struct sk_buff *skb, struct genl_info *info) rtnl_lock(); - err = get_rdev_dev_by_info_ifindex(info->attrs, &rdev, &dev); + err = get_rdev_dev_by_info_ifindex(info, &rdev, &dev); if (err) goto out_rtnl; @@ -2362,7 +2376,7 @@ static int nl80211_new_mpath(struct sk_buff *skb, struct genl_info *info) rtnl_lock(); - err = get_rdev_dev_by_info_ifindex(info->attrs, &rdev, &dev); + err = get_rdev_dev_by_info_ifindex(info, &rdev, &dev); if (err) goto out_rtnl; @@ -2404,7 +2418,7 @@ static int nl80211_del_mpath(struct sk_buff *skb, struct genl_info *info) rtnl_lock(); - err = get_rdev_dev_by_info_ifindex(info->attrs, &rdev, &dev); + err = get_rdev_dev_by_info_ifindex(info, &rdev, &dev); if (err) goto out_rtnl; @@ -2455,7 +2469,7 @@ static int nl80211_set_bss(struct sk_buff *skb, struct genl_info *info) rtnl_lock(); - err = get_rdev_dev_by_info_ifindex(info->attrs, &rdev, &dev); + err = get_rdev_dev_by_info_ifindex(info, &rdev, &dev); if (err) goto out_rtnl; @@ -2574,7 +2588,7 @@ static int nl80211_get_mesh_params(struct sk_buff *skb, rtnl_lock(); /* Look up our device */ - err = get_rdev_dev_by_info_ifindex(info->attrs, &rdev, &dev); + err = get_rdev_dev_by_info_ifindex(info, &rdev, &dev); if (err) goto out_rtnl; @@ -2691,7 +2705,7 @@ static int nl80211_set_mesh_params(struct sk_buff *skb, struct genl_info *info) rtnl_lock(); - err = get_rdev_dev_by_info_ifindex(info->attrs, &rdev, &dev); + err = get_rdev_dev_by_info_ifindex(info, &rdev, &dev); if (err) goto out_rtnl; @@ -2947,7 +2961,7 @@ static int nl80211_trigger_scan(struct sk_buff *skb, struct genl_info *info) rtnl_lock(); - err = get_rdev_dev_by_info_ifindex(info->attrs, &rdev, &dev); + err = get_rdev_dev_by_info_ifindex(info, &rdev, &dev); if (err) goto out_rtnl; @@ -3069,14 +3083,16 @@ static int nl80211_trigger_scan(struct sk_buff *skb, struct genl_info *info) request->ie_len); } - request->ifidx = dev->ifindex; + request->dev = dev; request->wiphy = &rdev->wiphy; rdev->scan_req = request; err = rdev->ops->scan(&rdev->wiphy, dev, request); - if (!err) + if (!err) { nl80211_send_scan_start(rdev, dev); + dev_hold(dev); + } out_free: if (err) { @@ -3198,11 +3214,11 @@ static int nl80211_dump_scan(struct sk_buff *skb, cb->args[0] = ifidx; } - dev = dev_get_by_index(&init_net, ifidx); + dev = dev_get_by_index(sock_net(skb->sk), ifidx); if (!dev) return -ENODEV; - rdev = cfg80211_get_dev_from_ifindex(ifidx); + rdev = cfg80211_get_dev_from_ifindex(sock_net(skb->sk), ifidx); if (IS_ERR(rdev)) { err = PTR_ERR(rdev); goto out_put_netdev; @@ -3312,7 +3328,7 @@ static int nl80211_authenticate(struct sk_buff *skb, struct genl_info *info) rtnl_lock(); - err = get_rdev_dev_by_info_ifindex(info->attrs, &rdev, &dev); + err = get_rdev_dev_by_info_ifindex(info, &rdev, &dev); if (err) goto unlock_rtnl; @@ -3448,7 +3464,7 @@ static int nl80211_associate(struct sk_buff *skb, struct genl_info *info) rtnl_lock(); - err = get_rdev_dev_by_info_ifindex(info->attrs, &rdev, &dev); + err = get_rdev_dev_by_info_ifindex(info, &rdev, &dev); if (err) goto unlock_rtnl; @@ -3531,7 +3547,7 @@ static int nl80211_deauthenticate(struct sk_buff *skb, struct genl_info *info) rtnl_lock(); - err = get_rdev_dev_by_info_ifindex(info->attrs, &rdev, &dev); + err = get_rdev_dev_by_info_ifindex(info, &rdev, &dev); if (err) goto unlock_rtnl; @@ -3593,7 +3609,7 @@ static int nl80211_disassociate(struct sk_buff *skb, struct genl_info *info) rtnl_lock(); - err = get_rdev_dev_by_info_ifindex(info->attrs, &rdev, &dev); + err = get_rdev_dev_by_info_ifindex(info, &rdev, &dev); if (err) goto unlock_rtnl; @@ -3666,7 +3682,7 @@ static int nl80211_join_ibss(struct sk_buff *skb, struct genl_info *info) rtnl_lock(); - err = get_rdev_dev_by_info_ifindex(info->attrs, &rdev, &dev); + err = get_rdev_dev_by_info_ifindex(info, &rdev, &dev); if (err) goto unlock_rtnl; @@ -3739,7 +3755,7 @@ static int nl80211_leave_ibss(struct sk_buff *skb, struct genl_info *info) rtnl_lock(); - err = get_rdev_dev_by_info_ifindex(info->attrs, &rdev, &dev); + err = get_rdev_dev_by_info_ifindex(info, &rdev, &dev); if (err) goto unlock_rtnl; @@ -3924,7 +3940,7 @@ static int nl80211_connect(struct sk_buff *skb, struct genl_info *info) return err; rtnl_lock(); - err = get_rdev_dev_by_info_ifindex(info->attrs, &rdev, &dev); + err = get_rdev_dev_by_info_ifindex(info, &rdev, &dev); if (err) goto unlock_rtnl; @@ -4000,7 +4016,7 @@ static int nl80211_disconnect(struct sk_buff *skb, struct genl_info *info) rtnl_lock(); - err = get_rdev_dev_by_info_ifindex(info->attrs, &rdev, &dev); + err = get_rdev_dev_by_info_ifindex(info, &rdev, &dev); if (err) goto unlock_rtnl; @@ -4024,6 +4040,47 @@ unlock_rtnl: return err; } +static int nl80211_wiphy_netns(struct sk_buff *skb, struct genl_info *info) +{ + struct cfg80211_registered_device *rdev; + struct net *net; + int err; + u32 pid; + + if (!info->attrs[NL80211_ATTR_PID]) + return -EINVAL; + + pid = nla_get_u32(info->attrs[NL80211_ATTR_PID]); + + rtnl_lock(); + + rdev = cfg80211_get_dev_from_info(info); + if (IS_ERR(rdev)) { + err = PTR_ERR(rdev); + goto out; + } + + net = get_net_ns_by_pid(pid); + if (IS_ERR(net)) { + err = PTR_ERR(net); + goto out; + } + + err = 0; + + /* check if anything to do */ + if (net_eq(wiphy_net(&rdev->wiphy), net)) + goto out_put_net; + + err = cfg80211_switch_netns(rdev, net); + out_put_net: + put_net(net); + out: + cfg80211_unlock_rdev(rdev); + rtnl_unlock(); + return err; +} + static struct genl_ops nl80211_ops[] = { { .cmd = NL80211_CMD_GET_WIPHY, @@ -4257,6 +4314,12 @@ static struct genl_ops nl80211_ops[] = { .policy = nl80211_policy, .flags = GENL_ADMIN_PERM, }, + { + .cmd = NL80211_CMD_SET_WIPHY_NETNS, + .doit = nl80211_wiphy_netns, + .policy = nl80211_policy, + .flags = GENL_ADMIN_PERM, + }, }; static struct genl_multicast_group nl80211_mlme_mcgrp = { .name = "mlme", @@ -4288,7 +4351,8 @@ void nl80211_notify_dev_rename(struct cfg80211_registered_device *rdev) return; } - genlmsg_multicast(msg, 0, nl80211_config_mcgrp.id, GFP_KERNEL); + genlmsg_multicast_netns(wiphy_net(&rdev->wiphy), msg, 0, + nl80211_config_mcgrp.id, GFP_KERNEL); } static int nl80211_add_scan_req(struct sk_buff *msg, @@ -4365,7 +4429,8 @@ void nl80211_send_scan_start(struct cfg80211_registered_device *rdev, return; } - genlmsg_multicast(msg, 0, nl80211_scan_mcgrp.id, GFP_KERNEL); + genlmsg_multicast_netns(wiphy_net(&rdev->wiphy), msg, 0, + nl80211_scan_mcgrp.id, GFP_KERNEL); } void nl80211_send_scan_done(struct cfg80211_registered_device *rdev, @@ -4383,7 +4448,8 @@ void nl80211_send_scan_done(struct cfg80211_registered_device *rdev, return; } - genlmsg_multicast(msg, 0, nl80211_scan_mcgrp.id, GFP_KERNEL); + genlmsg_multicast_netns(wiphy_net(&rdev->wiphy), msg, 0, + nl80211_scan_mcgrp.id, GFP_KERNEL); } void nl80211_send_scan_aborted(struct cfg80211_registered_device *rdev, @@ -4401,7 +4467,8 @@ void nl80211_send_scan_aborted(struct cfg80211_registered_device *rdev, return; } - genlmsg_multicast(msg, 0, nl80211_scan_mcgrp.id, GFP_KERNEL); + genlmsg_multicast_netns(wiphy_net(&rdev->wiphy), msg, 0, + nl80211_scan_mcgrp.id, GFP_KERNEL); } /* @@ -4450,7 +4517,10 @@ void nl80211_send_reg_change_event(struct regulatory_request *request) return; } - genlmsg_multicast(msg, 0, nl80211_regulatory_mcgrp.id, GFP_KERNEL); + rtnl_lock(); + genlmsg_multicast_allns(msg, 0, nl80211_regulatory_mcgrp.id, + GFP_KERNEL); + rtnl_unlock(); return; @@ -4486,7 +4556,8 @@ static void nl80211_send_mlme_event(struct cfg80211_registered_device *rdev, return; } - genlmsg_multicast(msg, 0, nl80211_mlme_mcgrp.id, gfp); + genlmsg_multicast_netns(wiphy_net(&rdev->wiphy), msg, 0, + nl80211_mlme_mcgrp.id, gfp); return; nla_put_failure: @@ -4553,7 +4624,8 @@ static void nl80211_send_mlme_timeout(struct cfg80211_registered_device *rdev, return; } - genlmsg_multicast(msg, 0, nl80211_mlme_mcgrp.id, gfp); + genlmsg_multicast_netns(wiphy_net(&rdev->wiphy), msg, 0, + nl80211_mlme_mcgrp.id, gfp); return; nla_put_failure: @@ -4611,7 +4683,8 @@ void nl80211_send_connect_result(struct cfg80211_registered_device *rdev, return; } - genlmsg_multicast(msg, 0, nl80211_mlme_mcgrp.id, gfp); + genlmsg_multicast_netns(wiphy_net(&rdev->wiphy), msg, 0, + nl80211_mlme_mcgrp.id, gfp); return; nla_put_failure: @@ -4651,7 +4724,8 @@ void nl80211_send_roamed(struct cfg80211_registered_device *rdev, return; } - genlmsg_multicast(msg, 0, nl80211_mlme_mcgrp.id, gfp); + genlmsg_multicast_netns(wiphy_net(&rdev->wiphy), msg, 0, + nl80211_mlme_mcgrp.id, gfp); return; nla_put_failure: @@ -4691,7 +4765,8 @@ void nl80211_send_disconnected(struct cfg80211_registered_device *rdev, return; } - genlmsg_multicast(msg, 0, nl80211_mlme_mcgrp.id, GFP_KERNEL); + genlmsg_multicast_netns(wiphy_net(&rdev->wiphy), msg, 0, + nl80211_mlme_mcgrp.id, GFP_KERNEL); return; nla_put_failure: @@ -4726,7 +4801,8 @@ void nl80211_send_ibss_bssid(struct cfg80211_registered_device *rdev, return; } - genlmsg_multicast(msg, 0, nl80211_mlme_mcgrp.id, gfp); + genlmsg_multicast_netns(wiphy_net(&rdev->wiphy), msg, 0, + nl80211_mlme_mcgrp.id, gfp); return; nla_put_failure: @@ -4766,7 +4842,8 @@ void nl80211_michael_mic_failure(struct cfg80211_registered_device *rdev, return; } - genlmsg_multicast(msg, 0, nl80211_mlme_mcgrp.id, gfp); + genlmsg_multicast_netns(wiphy_net(&rdev->wiphy), msg, 0, + nl80211_mlme_mcgrp.id, gfp); return; nla_put_failure: @@ -4819,7 +4896,10 @@ void nl80211_send_beacon_hint_event(struct wiphy *wiphy, return; } - genlmsg_multicast(msg, 0, nl80211_regulatory_mcgrp.id, GFP_ATOMIC); + rcu_read_lock(); + genlmsg_multicast_allns(msg, 0, nl80211_regulatory_mcgrp.id, + GFP_ATOMIC); + rcu_read_unlock(); return; diff --git a/net/wireless/scan.c b/net/wireless/scan.c index decc59fe0ee8..1b578b8cb1c9 100644 --- a/net/wireless/scan.c +++ b/net/wireless/scan.c @@ -32,9 +32,7 @@ void __cfg80211_scan_done(struct work_struct *wk) mutex_lock(&rdev->mtx); request = rdev->scan_req; - dev = dev_get_by_index(&init_net, request->ifidx); - if (!dev) - goto out; + dev = request->dev; /* * This must be before sending the other events! @@ -58,7 +56,6 @@ void __cfg80211_scan_done(struct work_struct *wk) dev_put(dev); - out: cfg80211_unlock_rdev(rdev); wiphy_to_dev(request->wiphy)->scan_req = NULL; kfree(request); @@ -66,17 +63,10 @@ void __cfg80211_scan_done(struct work_struct *wk) void cfg80211_scan_done(struct cfg80211_scan_request *request, bool aborted) { - struct net_device *dev = dev_get_by_index(&init_net, request->ifidx); - if (WARN_ON(!dev)) { - kfree(request); - return; - } - WARN_ON(request != wiphy_to_dev(request->wiphy)->scan_req); request->aborted = aborted; schedule_work(&wiphy_to_dev(request->wiphy)->scan_done_wk); - dev_put(dev); } EXPORT_SYMBOL(cfg80211_scan_done); @@ -592,7 +582,7 @@ int cfg80211_wext_siwscan(struct net_device *dev, if (!netif_running(dev)) return -ENETDOWN; - rdev = cfg80211_get_dev_from_ifindex(dev->ifindex); + rdev = cfg80211_get_dev_from_ifindex(dev_net(dev), dev->ifindex); if (IS_ERR(rdev)) return PTR_ERR(rdev); @@ -617,7 +607,7 @@ int cfg80211_wext_siwscan(struct net_device *dev, } creq->wiphy = wiphy; - creq->ifidx = dev->ifindex; + creq->dev = dev; creq->ssids = (void *)(creq + 1); creq->channels = (void *)(creq->ssids + 1); creq->n_channels = n_channels; @@ -654,8 +644,10 @@ int cfg80211_wext_siwscan(struct net_device *dev, if (err) { rdev->scan_req = NULL; kfree(creq); - } else + } else { nl80211_send_scan_start(rdev, dev); + dev_hold(dev); + } out: cfg80211_unlock_rdev(rdev); return err; @@ -948,7 +940,7 @@ int cfg80211_wext_giwscan(struct net_device *dev, if (!netif_running(dev)) return -ENETDOWN; - rdev = cfg80211_get_dev_from_ifindex(dev->ifindex); + rdev = cfg80211_get_dev_from_ifindex(dev_net(dev), dev->ifindex); if (IS_ERR(rdev)) return PTR_ERR(rdev); diff --git a/net/wireless/sme.c b/net/wireless/sme.c index 82de2d9795f4..a19741097989 100644 --- a/net/wireless/sme.c +++ b/net/wireless/sme.c @@ -86,7 +86,7 @@ static int cfg80211_conn_scan(struct wireless_dev *wdev) wdev->conn->params.ssid_len); request->ssids[0].ssid_len = wdev->conn->params.ssid_len; - request->ifidx = wdev->netdev->ifindex; + request->dev = wdev->netdev; request->wiphy = &rdev->wiphy; rdev->scan_req = request; @@ -95,6 +95,7 @@ static int cfg80211_conn_scan(struct wireless_dev *wdev) if (!err) { wdev->conn->state = CFG80211_CONN_SCANNING; nl80211_send_scan_start(rdev, wdev->netdev); + dev_hold(wdev->netdev); } else { rdev->scan_req = NULL; kfree(request); -- cgit v1.2.3 From d1eba248469272ae0618288bccf65b24d017f1d2 Mon Sep 17 00:00:00 2001 From: Sujith Date: Thu, 23 Jul 2009 15:31:31 +0530 Subject: mac80211: Add a few 802.11n defines for AMPDU parameters Signed-off-by: Sujith Acked-by: Johannes Berg Signed-off-by: John W. Linville --- include/linux/ieee80211.h | 25 +++++++++++++++++++++++++ 1 file changed, 25 insertions(+) (limited to 'include') diff --git a/include/linux/ieee80211.h b/include/linux/ieee80211.h index 23343ab0bb03..21556a2d9e7e 100644 --- a/include/linux/ieee80211.h +++ b/include/linux/ieee80211.h @@ -802,6 +802,31 @@ struct ieee80211_ht_cap { #define IEEE80211_HT_AMPDU_PARM_FACTOR 0x03 #define IEEE80211_HT_AMPDU_PARM_DENSITY 0x1C +/* + * Maximum length of AMPDU that the STA can receive. + * Length = 2 ^ (13 + max_ampdu_length_exp) - 1 (octets) + */ +enum ieee80211_max_ampdu_length_exp { + IEEE80211_HT_MAX_AMPDU_8K = 0, + IEEE80211_HT_MAX_AMPDU_16K = 1, + IEEE80211_HT_MAX_AMPDU_32K = 2, + IEEE80211_HT_MAX_AMPDU_64K = 3 +}; + +#define IEEE80211_HT_MAX_AMPDU_FACTOR 13 + +/* Minimum MPDU start spacing */ +enum ieee80211_min_mpdu_spacing { + IEEE80211_HT_MPDU_DENSITY_NONE = 0, /* No restriction */ + IEEE80211_HT_MPDU_DENSITY_0_25 = 1, /* 1/4 usec */ + IEEE80211_HT_MPDU_DENSITY_0_5 = 2, /* 1/2 usec */ + IEEE80211_HT_MPDU_DENSITY_1 = 3, /* 1 usec */ + IEEE80211_HT_MPDU_DENSITY_2 = 4, /* 2 usec */ + IEEE80211_HT_MPDU_DENSITY_4 = 5, /* 4 usec */ + IEEE80211_HT_MPDU_DENSITY_8 = 6, /* 8 usec */ + IEEE80211_HT_MPDU_DENSITY_16 = 7 /* 16 usec */ +}; + /** * struct ieee80211_ht_info - HT information * -- cgit v1.2.3 From 3fa52056f3a8e755708241d5795e6d3e6f55ad85 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Fri, 24 Jul 2009 13:23:09 +0200 Subject: mac80211: fix PS-poll response, race When a station queries us for a PS-poll response, we wrongly queue the frame on the virtual interface's queue rather than the pending queue. Additionally, fix a race condition where we could potentially send multiple frames to the sleeping station due to using a station flag rather than a packet flag. When converting to a packet flag, we can also convert p54 and remove the filter clearing we added for it. (Also remove a now dead function) Signed-off-by: Johannes Berg Reported-by: Bob Copeland Tested-by: Bob Copeland Cc: Christian Lamparter Signed-off-by: John W. Linville --- drivers/net/wireless/p54/txrx.c | 2 +- include/net/mac80211.h | 4 ++++ net/mac80211/rx.c | 11 ++++++----- net/mac80211/sta_info.h | 13 ------------- net/mac80211/tx.c | 19 +------------------ 5 files changed, 12 insertions(+), 37 deletions(-) (limited to 'include') diff --git a/drivers/net/wireless/p54/txrx.c b/drivers/net/wireless/p54/txrx.c index 0d589d68e547..c32a0d2fa1f7 100644 --- a/drivers/net/wireless/p54/txrx.c +++ b/drivers/net/wireless/p54/txrx.c @@ -614,7 +614,7 @@ static void p54_tx_80211_header(struct p54_common *priv, struct sk_buff *skb, if (info->flags & IEEE80211_TX_CTL_ASSIGN_SEQ) *flags |= P54_HDR_FLAG_DATA_OUT_SEQNR; - if (info->flags & IEEE80211_TX_CTL_CLEAR_PS_FILT) + if (info->flags & IEEE80211_TX_CTL_PSPOLL_RESPONSE) *flags |= P54_HDR_FLAG_DATA_OUT_NOCANCEL; *queue = skb_get_queue_mapping(skb) + P54_QUEUE_DATA; diff --git a/include/net/mac80211.h b/include/net/mac80211.h index 7dd67a1ff4d5..d4e09a06b4a2 100644 --- a/include/net/mac80211.h +++ b/include/net/mac80211.h @@ -243,6 +243,9 @@ struct ieee80211_bss_conf { * used to indicate that a frame was already retried due to PS * @IEEE80211_TX_INTFL_DONT_ENCRYPT: completely internal to mac80211, * used to indicate frame should not be encrypted + * @IEEE80211_TX_CTL_PSPOLL_RESPONSE: (internal?) + * This frame is a response to a PS-poll frame and should be sent + * although the station is in powersave mode. */ enum mac80211_tx_control_flags { IEEE80211_TX_CTL_REQ_TX_STATUS = BIT(0), @@ -262,6 +265,7 @@ enum mac80211_tx_control_flags { IEEE80211_TX_INTFL_NEED_TXPROCESSING = BIT(14), IEEE80211_TX_INTFL_RETRIED = BIT(15), IEEE80211_TX_INTFL_DONT_ENCRYPT = BIT(16), + IEEE80211_TX_CTL_PSPOLL_RESPONSE = BIT(17), }; /** diff --git a/net/mac80211/rx.c b/net/mac80211/rx.c index cb95a3116034..f195705146bd 100644 --- a/net/mac80211/rx.c +++ b/net/mac80211/rx.c @@ -783,7 +783,7 @@ static void ap_sta_ps_start(struct sta_info *sta) struct ieee80211_local *local = sdata->local; atomic_inc(&sdata->bss->num_sta_ps); - set_and_clear_sta_flags(sta, WLAN_STA_PS, WLAN_STA_PSPOLL); + set_sta_flags(sta, WLAN_STA_PS); drv_sta_notify(local, &sdata->vif, STA_NOTIFY_SLEEP, &sta->sta); #ifdef CONFIG_MAC80211_VERBOSE_PS_DEBUG printk(KERN_DEBUG "%s: STA %pM aid %d enters power save mode\n", @@ -799,7 +799,7 @@ static int ap_sta_ps_end(struct sta_info *sta) atomic_dec(&sdata->bss->num_sta_ps); - clear_sta_flags(sta, WLAN_STA_PS | WLAN_STA_PSPOLL); + clear_sta_flags(sta, WLAN_STA_PS); drv_sta_notify(local, &sdata->vif, STA_NOTIFY_AWAKE, &sta->sta); if (!skb_queue_empty(&sta->ps_tx_buf)) @@ -1117,14 +1117,15 @@ ieee80211_rx_h_ps_poll(struct ieee80211_rx_data *rx) skb_queue_empty(&rx->sta->ps_tx_buf); if (skb) { + struct ieee80211_tx_info *info = IEEE80211_SKB_CB(skb); struct ieee80211_hdr *hdr = (struct ieee80211_hdr *) skb->data; /* - * Tell TX path to send one frame even though the STA may + * Tell TX path to send this frame even though the STA may * still remain is PS mode after this frame exchange. */ - set_sta_flags(rx->sta, WLAN_STA_PSPOLL); + info->flags |= IEEE80211_TX_CTL_PSPOLL_RESPONSE; #ifdef CONFIG_MAC80211_VERBOSE_PS_DEBUG printk(KERN_DEBUG "STA %pM aid %d: PS Poll (entries after %d)\n", @@ -1139,7 +1140,7 @@ ieee80211_rx_h_ps_poll(struct ieee80211_rx_data *rx) else hdr->frame_control |= cpu_to_le16(IEEE80211_FCTL_MOREDATA); - dev_queue_xmit(skb); + ieee80211_add_pending_skb(rx->local, skb); if (no_pending_pkts) sta_info_clear_tim_bit(rx->sta); diff --git a/net/mac80211/sta_info.h b/net/mac80211/sta_info.h index 4ecf10a9bd00..ccc3adf962c7 100644 --- a/net/mac80211/sta_info.h +++ b/net/mac80211/sta_info.h @@ -30,7 +30,6 @@ * @WLAN_STA_ASSOC_AP: We're associated to that station, it is an AP. * @WLAN_STA_WME: Station is a QoS-STA. * @WLAN_STA_WDS: Station is one of our WDS peers. - * @WLAN_STA_PSPOLL: Station has just PS-polled us. * @WLAN_STA_CLEAR_PS_FILT: Clear PS filter in hardware (using the * IEEE80211_TX_CTL_CLEAR_PS_FILT control flag) when the next * frame to this station is transmitted. @@ -47,7 +46,6 @@ enum ieee80211_sta_info_flags { WLAN_STA_ASSOC_AP = 1<<5, WLAN_STA_WME = 1<<6, WLAN_STA_WDS = 1<<7, - WLAN_STA_PSPOLL = 1<<8, WLAN_STA_CLEAR_PS_FILT = 1<<9, WLAN_STA_MFP = 1<<10, WLAN_STA_SUSPEND = 1<<11 @@ -359,17 +357,6 @@ static inline void clear_sta_flags(struct sta_info *sta, const u32 flags) spin_unlock_irqrestore(&sta->flaglock, irqfl); } -static inline void set_and_clear_sta_flags(struct sta_info *sta, - const u32 set, const u32 clear) -{ - unsigned long irqfl; - - spin_lock_irqsave(&sta->flaglock, irqfl); - sta->flags |= set; - sta->flags &= ~clear; - spin_unlock_irqrestore(&sta->flaglock, irqfl); -} - static inline u32 test_sta_flags(struct sta_info *sta, const u32 flags) { u32 ret; diff --git a/net/mac80211/tx.c b/net/mac80211/tx.c index 70ff4f065665..edacad1fb1dc 100644 --- a/net/mac80211/tx.c +++ b/net/mac80211/tx.c @@ -373,7 +373,7 @@ ieee80211_tx_h_unicast_ps_buf(struct ieee80211_tx_data *tx) staflags = get_sta_flags(sta); if (unlikely((staflags & WLAN_STA_PS) && - !(staflags & WLAN_STA_PSPOLL))) { + !(info->flags & IEEE80211_TX_CTL_PSPOLL_RESPONSE))) { #ifdef CONFIG_MAC80211_VERBOSE_PS_DEBUG printk(KERN_DEBUG "STA %pM aid %d: PS buffer (entries " "before %d)\n", @@ -412,24 +412,7 @@ ieee80211_tx_h_unicast_ps_buf(struct ieee80211_tx_data *tx) sta->sta.addr); } #endif /* CONFIG_MAC80211_VERBOSE_PS_DEBUG */ - if (test_and_clear_sta_flags(sta, WLAN_STA_PSPOLL)) { - /* - * The sleeping station with pending data is now snoozing. - * It queried us for its buffered frames and will go back - * to deep sleep once it got everything. - * - * inform the driver, in case the hardware does powersave - * frame filtering and keeps a station blacklist on its own - * (e.g: p54), so that frames can be delivered unimpeded. - * - * Note: It should be safe to disable the filter now. - * As, it is really unlikely that we still have any pending - * frame for this station in the hw's buffers/fifos left, - * that is not rejected with a unsuccessful tx_status yet. - */ - info->flags |= IEEE80211_TX_CTL_CLEAR_PS_FILT; - } return TX_CONTINUE; } -- cgit v1.2.3 From 7cb7f45c7feef43c8f71f5cfedfc0b19be2142f7 Mon Sep 17 00:00:00 2001 From: Len Brown Date: Mon, 27 Jul 2009 18:42:38 -0400 Subject: Revert "ACPICA: Remove obsolete acpi_os_validate_address interface" This reverts commit f9ca058430333c9a24c5ca926aa445125f88df18. which caused a regression: http://bugzilla.kernel.org/show_bug.cgi?id=13620 Signed-off-by: Lin Ming Signed-off-by: Len Brown --- drivers/acpi/acpica/acobject.h | 1 + drivers/acpi/acpica/dsopcode.c | 24 ++++++++++++++++++++++++ drivers/acpi/acpica/exfldio.c | 6 ++++++ include/acpi/acpiosxf.h | 4 ++++ 4 files changed, 35 insertions(+) (limited to 'include') diff --git a/drivers/acpi/acpica/acobject.h b/drivers/acpi/acpica/acobject.h index 544dcf834922..eb6f038b03d9 100644 --- a/drivers/acpi/acpica/acobject.h +++ b/drivers/acpi/acpica/acobject.h @@ -97,6 +97,7 @@ #define AOPOBJ_OBJECT_INITIALIZED 0x08 #define AOPOBJ_SETUP_COMPLETE 0x10 #define AOPOBJ_SINGLE_DATUM 0x20 +#define AOPOBJ_INVALID 0x40 /* Used if host OS won't allow an op_region address */ /****************************************************************************** * diff --git a/drivers/acpi/acpica/dsopcode.c b/drivers/acpi/acpica/dsopcode.c index 584d766e6f12..b79978f7bc71 100644 --- a/drivers/acpi/acpica/dsopcode.c +++ b/drivers/acpi/acpica/dsopcode.c @@ -397,6 +397,30 @@ acpi_status acpi_ds_get_region_arguments(union acpi_operand_object *obj_desc) status = acpi_ds_execute_arguments(node, acpi_ns_get_parent_node(node), extra_desc->extra.aml_length, extra_desc->extra.aml_start); + if (ACPI_FAILURE(status)) { + return_ACPI_STATUS(status); + } + + /* Validate the region address/length via the host OS */ + + status = acpi_os_validate_address(obj_desc->region.space_id, + obj_desc->region.address, + (acpi_size) obj_desc->region.length, + acpi_ut_get_node_name(node)); + + if (ACPI_FAILURE(status)) { + /* + * Invalid address/length. We will emit an error message and mark + * the region as invalid, so that it will cause an additional error if + * it is ever used. Then return AE_OK. + */ + ACPI_EXCEPTION((AE_INFO, status, + "During address validation of OpRegion [%4.4s]", + node->name.ascii)); + obj_desc->common.flags |= AOPOBJ_INVALID; + status = AE_OK; + } + return_ACPI_STATUS(status); } diff --git a/drivers/acpi/acpica/exfldio.c b/drivers/acpi/acpica/exfldio.c index d4075b821021..6687be167f5f 100644 --- a/drivers/acpi/acpica/exfldio.c +++ b/drivers/acpi/acpica/exfldio.c @@ -113,6 +113,12 @@ acpi_ex_setup_region(union acpi_operand_object *obj_desc, } } + /* Exit if Address/Length have been disallowed by the host OS */ + + if (rgn_desc->common.flags & AOPOBJ_INVALID) { + return_ACPI_STATUS(AE_AML_ILLEGAL_ADDRESS); + } + /* * Exit now for SMBus address space, it has a non-linear address space * and the request cannot be directly validated diff --git a/include/acpi/acpiosxf.h b/include/acpi/acpiosxf.h index 3e798593b17b..ab0b85cf21f3 100644 --- a/include/acpi/acpiosxf.h +++ b/include/acpi/acpiosxf.h @@ -242,6 +242,10 @@ acpi_os_derive_pci_id(acpi_handle rhandle, acpi_status acpi_os_validate_interface(char *interface); acpi_status acpi_osi_invalidate(char* interface); +acpi_status +acpi_os_validate_address(u8 space_id, acpi_physical_address address, + acpi_size length, char *name); + u64 acpi_os_get_timer(void); acpi_status acpi_os_signal(u32 function, void *info); -- cgit v1.2.3 From 27fed4175acf81ddd91d9a4ee2fd298981f60295 Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Mon, 27 Jul 2009 18:39:45 -0700 Subject: ip: fix logic of reverse path filter sysctl Even though reverse path filter was changed from simple boolean to trinary control, the loose mode only works if both all and device are configured because of this logic error. Signed-off-by: Stephen Hemminger Signed-off-by: David S. Miller --- include/linux/inetdevice.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/inetdevice.h b/include/linux/inetdevice.h index acef2a770b6b..ad27c7da8798 100644 --- a/include/linux/inetdevice.h +++ b/include/linux/inetdevice.h @@ -82,7 +82,7 @@ static inline void ipv4_devconf_setall(struct in_device *in_dev) #define IN_DEV_FORWARD(in_dev) IN_DEV_CONF_GET((in_dev), FORWARDING) #define IN_DEV_MFORWARD(in_dev) IN_DEV_ANDCONF((in_dev), MC_FORWARDING) -#define IN_DEV_RPFILTER(in_dev) IN_DEV_ANDCONF((in_dev), RP_FILTER) +#define IN_DEV_RPFILTER(in_dev) IN_DEV_MAXCONF((in_dev), RP_FILTER) #define IN_DEV_SOURCE_ROUTE(in_dev) IN_DEV_ANDCONF((in_dev), \ ACCEPT_SOURCE_ROUTE) #define IN_DEV_BOOTP_RELAY(in_dev) IN_DEV_ANDCONF((in_dev), BOOTP_RELAY) -- cgit v1.2.3 From 3885123da8335dc6b67387e5e626acbffc56f664 Mon Sep 17 00:00:00 2001 From: FUJITA Tomonori Date: Fri, 10 Jul 2009 10:04:50 +0900 Subject: swiotlb: remove unused swiotlb_alloc_boot() Nobody uses swiotlb_alloc_boot(). Signed-off-by: FUJITA Tomonori Acked-by: Becky Bruce --- arch/x86/kernel/pci-swiotlb.c | 5 ----- include/linux/swiotlb.h | 2 -- lib/swiotlb.c | 7 +------ 3 files changed, 1 insertion(+), 13 deletions(-) (limited to 'include') diff --git a/arch/x86/kernel/pci-swiotlb.c b/arch/x86/kernel/pci-swiotlb.c index 6af96ee44200..0ac7cd524788 100644 --- a/arch/x86/kernel/pci-swiotlb.c +++ b/arch/x86/kernel/pci-swiotlb.c @@ -13,11 +13,6 @@ int swiotlb __read_mostly; -void * __init swiotlb_alloc_boot(size_t size, unsigned long nslabs) -{ - return alloc_bootmem_low_pages(size); -} - void *swiotlb_alloc(unsigned order, unsigned long nslabs) { return (void *)__get_free_pages(GFP_DMA | __GFP_NOWARN, order); diff --git a/include/linux/swiotlb.h b/include/linux/swiotlb.h index cb1a6631b8f4..94db70444c17 100644 --- a/include/linux/swiotlb.h +++ b/include/linux/swiotlb.h @@ -14,7 +14,6 @@ struct scatterlist; */ #define IO_TLB_SEGSIZE 128 - /* * log of the size of each IO TLB slab. The number of slabs is command line * controllable. @@ -24,7 +23,6 @@ struct scatterlist; extern void swiotlb_init(void); -extern void *swiotlb_alloc_boot(size_t bytes, unsigned long nslabs); extern void *swiotlb_alloc(unsigned order, unsigned long nslabs); extern dma_addr_t swiotlb_phys_to_bus(struct device *hwdev, diff --git a/lib/swiotlb.c b/lib/swiotlb.c index bffe6d7ef9d9..9edfdd442edc 100644 --- a/lib/swiotlb.c +++ b/lib/swiotlb.c @@ -114,11 +114,6 @@ setup_io_tlb_npages(char *str) __setup("swiotlb=", setup_io_tlb_npages); /* make io_tlb_overflow tunable too? */ -void * __weak __init swiotlb_alloc_boot(size_t size, unsigned long nslabs) -{ - return alloc_bootmem_low_pages(size); -} - void * __weak swiotlb_alloc(unsigned order, unsigned long nslabs) { return (void *)__get_free_pages(GFP_DMA | __GFP_NOWARN, order); @@ -189,7 +184,7 @@ swiotlb_init_with_default_size(size_t default_size) /* * Get IO TLB memory from the low pages */ - io_tlb_start = swiotlb_alloc_boot(bytes, io_tlb_nslabs); + io_tlb_start = alloc_bootmem_low_pages(bytes); if (!io_tlb_start) panic("Cannot allocate SWIOTLB buffer"); io_tlb_end = io_tlb_start + bytes; -- cgit v1.2.3 From bb52196be37ce154ddc50b1f39496146d181cbe7 Mon Sep 17 00:00:00 2001 From: FUJITA Tomonori Date: Fri, 10 Jul 2009 10:04:51 +0900 Subject: swiotlb: remove unused swiotlb_alloc() Nobody uses swiotlb_alloc(). Signed-off-by: FUJITA Tomonori Acked-by: Becky Bruce --- arch/x86/kernel/pci-swiotlb.c | 5 ----- include/linux/swiotlb.h | 2 -- lib/swiotlb.c | 8 ++------ 3 files changed, 2 insertions(+), 13 deletions(-) (limited to 'include') diff --git a/arch/x86/kernel/pci-swiotlb.c b/arch/x86/kernel/pci-swiotlb.c index 0ac7cd524788..ea675cfe76fe 100644 --- a/arch/x86/kernel/pci-swiotlb.c +++ b/arch/x86/kernel/pci-swiotlb.c @@ -13,11 +13,6 @@ int swiotlb __read_mostly; -void *swiotlb_alloc(unsigned order, unsigned long nslabs) -{ - return (void *)__get_free_pages(GFP_DMA | __GFP_NOWARN, order); -} - dma_addr_t swiotlb_phys_to_bus(struct device *hwdev, phys_addr_t paddr) { return paddr; diff --git a/include/linux/swiotlb.h b/include/linux/swiotlb.h index 94db70444c17..6bc50944040f 100644 --- a/include/linux/swiotlb.h +++ b/include/linux/swiotlb.h @@ -23,8 +23,6 @@ struct scatterlist; extern void swiotlb_init(void); -extern void *swiotlb_alloc(unsigned order, unsigned long nslabs); - extern dma_addr_t swiotlb_phys_to_bus(struct device *hwdev, phys_addr_t address); extern phys_addr_t swiotlb_bus_to_phys(struct device *hwdev, diff --git a/lib/swiotlb.c b/lib/swiotlb.c index 9edfdd442edc..3c4c21cdf43d 100644 --- a/lib/swiotlb.c +++ b/lib/swiotlb.c @@ -114,11 +114,6 @@ setup_io_tlb_npages(char *str) __setup("swiotlb=", setup_io_tlb_npages); /* make io_tlb_overflow tunable too? */ -void * __weak swiotlb_alloc(unsigned order, unsigned long nslabs) -{ - return (void *)__get_free_pages(GFP_DMA | __GFP_NOWARN, order); -} - dma_addr_t __weak swiotlb_phys_to_bus(struct device *hwdev, phys_addr_t paddr) { return paddr; @@ -240,7 +235,8 @@ swiotlb_late_init_with_default_size(size_t default_size) bytes = io_tlb_nslabs << IO_TLB_SHIFT; while ((SLABS_PER_PAGE << order) > IO_TLB_MIN_SLABS) { - io_tlb_start = swiotlb_alloc(order, io_tlb_nslabs); + io_tlb_start = (void *)__get_free_pages(GFP_DMA | __GFP_NOWARN, + order); if (io_tlb_start) break; order--; -- cgit v1.2.3 From cf56e3f2e8a8d5b7bc719980869b0e7985c256f3 Mon Sep 17 00:00:00 2001 From: FUJITA Tomonori Date: Fri, 10 Jul 2009 10:04:52 +0900 Subject: swiotlb: remove swiotlb_arch_range_needs_mapping Nobody uses swiotlb_arch_range_needs_mapping(). Signed-off-by: FUJITA Tomonori Acked-by: Becky Bruce --- arch/x86/kernel/pci-swiotlb.c | 5 ----- include/linux/swiotlb.h | 2 -- lib/swiotlb.c | 15 ++------------- 3 files changed, 2 insertions(+), 20 deletions(-) (limited to 'include') diff --git a/arch/x86/kernel/pci-swiotlb.c b/arch/x86/kernel/pci-swiotlb.c index ea675cfe76fe..165bd7f93bb1 100644 --- a/arch/x86/kernel/pci-swiotlb.c +++ b/arch/x86/kernel/pci-swiotlb.c @@ -23,11 +23,6 @@ phys_addr_t swiotlb_bus_to_phys(struct device *hwdev, dma_addr_t baddr) return baddr; } -int __weak swiotlb_arch_range_needs_mapping(phys_addr_t paddr, size_t size) -{ - return 0; -} - static void *x86_swiotlb_alloc_coherent(struct device *hwdev, size_t size, dma_addr_t *dma_handle, gfp_t flags) { diff --git a/include/linux/swiotlb.h b/include/linux/swiotlb.h index 6bc50944040f..a977da24f17c 100644 --- a/include/linux/swiotlb.h +++ b/include/linux/swiotlb.h @@ -28,8 +28,6 @@ extern dma_addr_t swiotlb_phys_to_bus(struct device *hwdev, extern phys_addr_t swiotlb_bus_to_phys(struct device *hwdev, dma_addr_t address); -extern int swiotlb_arch_range_needs_mapping(phys_addr_t paddr, size_t size); - extern void *swiotlb_alloc_coherent(struct device *hwdev, size_t size, dma_addr_t *dma_handle, gfp_t flags); diff --git a/lib/swiotlb.c b/lib/swiotlb.c index 3c4c21cdf43d..dc1cd1f5369e 100644 --- a/lib/swiotlb.c +++ b/lib/swiotlb.c @@ -141,11 +141,6 @@ int __weak swiotlb_arch_address_needs_mapping(struct device *hwdev, return !is_buffer_dma_capable(dma_get_mask(hwdev), addr, size); } -int __weak swiotlb_arch_range_needs_mapping(phys_addr_t paddr, size_t size) -{ - return 0; -} - static void swiotlb_print_info(unsigned long bytes) { phys_addr_t pstart, pend; @@ -312,11 +307,6 @@ address_needs_mapping(struct device *hwdev, dma_addr_t addr, size_t size) return swiotlb_arch_address_needs_mapping(hwdev, addr, size); } -static inline int range_needs_mapping(phys_addr_t paddr, size_t size) -{ - return swiotlb_force || swiotlb_arch_range_needs_mapping(paddr, size); -} - static int is_swiotlb_buffer(char *addr) { return addr >= io_tlb_start && addr < io_tlb_end; @@ -646,8 +636,7 @@ dma_addr_t swiotlb_map_page(struct device *dev, struct page *page, * we can safely return the device addr and not worry about bounce * buffering it. */ - if (!address_needs_mapping(dev, dev_addr, size) && - !range_needs_mapping(phys, size)) + if (!address_needs_mapping(dev, dev_addr, size) && !swiotlb_force) return dev_addr; /* @@ -810,7 +799,7 @@ swiotlb_map_sg_attrs(struct device *hwdev, struct scatterlist *sgl, int nelems, phys_addr_t paddr = sg_phys(sg); dma_addr_t dev_addr = swiotlb_phys_to_bus(hwdev, paddr); - if (range_needs_mapping(paddr, sg->length) || + if (swiotlb_force || address_needs_mapping(hwdev, dev_addr, sg->length)) { void *map = map_single(hwdev, sg_phys(sg), sg->length, dir); -- cgit v1.2.3 From 8f2502fd8157632909ff335a3c628e7caeec5e03 Mon Sep 17 00:00:00 2001 From: FUJITA Tomonori Date: Fri, 10 Jul 2009 10:05:00 +0900 Subject: remove is_buffer_dma_capable() is_buffer_dma_capable() was replaced with dma_capable(). is_buffer_dma_capable() tells if a buffer is dma-capable or not. However, it doesn't take a pointer to struct device so it doesn't work for POWERPC. Signed-off-by: FUJITA Tomonori Acked-by: Becky Bruce --- include/linux/dma-mapping.h | 5 ----- 1 file changed, 5 deletions(-) (limited to 'include') diff --git a/include/linux/dma-mapping.h b/include/linux/dma-mapping.h index 07dfd460d286..c0f6c3cd788c 100644 --- a/include/linux/dma-mapping.h +++ b/include/linux/dma-mapping.h @@ -98,11 +98,6 @@ static inline int is_device_dma_capable(struct device *dev) return dev->dma_mask != NULL && *dev->dma_mask != DMA_MASK_NONE; } -static inline int is_buffer_dma_capable(u64 mask, dma_addr_t addr, size_t size) -{ - return addr + size <= mask; -} - #ifdef CONFIG_HAS_DMA #include #else -- cgit v1.2.3 From 862d196b27bd30a5ab8109d5cdb37980d81b1fe9 Mon Sep 17 00:00:00 2001 From: FUJITA Tomonori Date: Fri, 10 Jul 2009 10:05:02 +0900 Subject: swiotlb: use phys_to_dma and dma_to_phys This converts swiotlb to use phys_to_dma and dma_to_phys instead of swiotlb_phys_to_bus() and swiotlb_bus_to_phys(). swiotlb_phys_to_bus() and swiotlb_bus_to_phys() are not necessary so this patch also removes them. Signed-off-by: FUJITA Tomonori Acked-by: Becky Bruce --- include/linux/swiotlb.h | 5 ----- lib/swiotlb.c | 22 ++++++---------------- 2 files changed, 6 insertions(+), 21 deletions(-) (limited to 'include') diff --git a/include/linux/swiotlb.h b/include/linux/swiotlb.h index a977da24f17c..73b1f1cec423 100644 --- a/include/linux/swiotlb.h +++ b/include/linux/swiotlb.h @@ -23,11 +23,6 @@ struct scatterlist; extern void swiotlb_init(void); -extern dma_addr_t swiotlb_phys_to_bus(struct device *hwdev, - phys_addr_t address); -extern phys_addr_t swiotlb_bus_to_phys(struct device *hwdev, - dma_addr_t address); - extern void *swiotlb_alloc_coherent(struct device *hwdev, size_t size, dma_addr_t *dma_handle, gfp_t flags); diff --git a/lib/swiotlb.c b/lib/swiotlb.c index a012d93792b7..9e2fe3e1804c 100644 --- a/lib/swiotlb.c +++ b/lib/swiotlb.c @@ -114,21 +114,11 @@ setup_io_tlb_npages(char *str) __setup("swiotlb=", setup_io_tlb_npages); /* make io_tlb_overflow tunable too? */ -dma_addr_t __weak swiotlb_phys_to_bus(struct device *hwdev, phys_addr_t paddr) -{ - return paddr; -} - -phys_addr_t __weak swiotlb_bus_to_phys(struct device *hwdev, dma_addr_t baddr) -{ - return baddr; -} - /* Note that this doesn't work with highmem page */ static dma_addr_t swiotlb_virt_to_bus(struct device *hwdev, volatile void *address) { - return swiotlb_phys_to_bus(hwdev, virt_to_phys(address)); + return phys_to_dma(hwdev, virt_to_phys(address)); } static void swiotlb_print_info(unsigned long bytes) @@ -567,7 +557,7 @@ void swiotlb_free_coherent(struct device *hwdev, size_t size, void *vaddr, dma_addr_t dev_addr) { - phys_addr_t paddr = swiotlb_bus_to_phys(hwdev, dev_addr); + phys_addr_t paddr = dma_to_phys(hwdev, dev_addr); WARN_ON(irqs_disabled()); if (!is_swiotlb_buffer(paddr)) @@ -612,7 +602,7 @@ dma_addr_t swiotlb_map_page(struct device *dev, struct page *page, struct dma_attrs *attrs) { phys_addr_t phys = page_to_phys(page) + offset; - dma_addr_t dev_addr = swiotlb_phys_to_bus(dev, phys); + dma_addr_t dev_addr = phys_to_dma(dev, phys); void *map; BUG_ON(dir == DMA_NONE); @@ -656,7 +646,7 @@ EXPORT_SYMBOL_GPL(swiotlb_map_page); static void unmap_single(struct device *hwdev, dma_addr_t dev_addr, size_t size, int dir) { - phys_addr_t paddr = swiotlb_bus_to_phys(hwdev, dev_addr); + phys_addr_t paddr = dma_to_phys(hwdev, dev_addr); BUG_ON(dir == DMA_NONE); @@ -699,7 +689,7 @@ static void swiotlb_sync_single(struct device *hwdev, dma_addr_t dev_addr, size_t size, int dir, int target) { - phys_addr_t paddr = swiotlb_bus_to_phys(hwdev, dev_addr); + phys_addr_t paddr = dma_to_phys(hwdev, dev_addr); BUG_ON(dir == DMA_NONE); @@ -788,7 +778,7 @@ swiotlb_map_sg_attrs(struct device *hwdev, struct scatterlist *sgl, int nelems, for_each_sg(sgl, sg, nelems, i) { phys_addr_t paddr = sg_phys(sg); - dma_addr_t dev_addr = swiotlb_phys_to_bus(hwdev, paddr); + dma_addr_t dev_addr = phys_to_dma(hwdev, paddr); if (swiotlb_force || !dma_capable(hwdev, dev_addr, sg->length)) { -- cgit v1.2.3 From 0c193054a4c1cf190d2f23e5e91bd14402e43912 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Mon, 27 Jul 2009 19:09:19 -0400 Subject: nfsd41: hange from page to memory based drc limits NFSD_SLOT_CACHE_SIZE is the size of all encoded operation responses (excluding the sequence operation) that we want to cache. For now, keep NFSD_SLOT_CACHE_SIZE at PAGE_SIZE. It will be reduced when the DRC is changed from page based to memory based. Signed-off-by: Andy Adamson Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 28 +++++++++++++--------------- fs/nfsd/nfssvc.c | 13 ++++++------- include/linux/nfsd/nfsd.h | 4 ++-- include/linux/nfsd/state.h | 1 + 4 files changed, 22 insertions(+), 24 deletions(-) (limited to 'include') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 70cba3fbfa6d..e2b11b1b515c 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -414,31 +414,31 @@ gen_sessionid(struct nfsd4_session *ses) /* * Give the client the number of slots it requests bound by - * NFSD_MAX_SLOTS_PER_SESSION and by sv_drc_max_pages. + * NFSD_MAX_SLOTS_PER_SESSION and by nfsd_drc_max_mem. * - * If we run out of pages (sv_drc_pages_used == sv_drc_max_pages) we - * should (up to a point) re-negotiate active sessions and reduce their - * slot usage to make rooom for new connections. For now we just fail the - * create session. + * If we run out of reserved DRC memory we should (up to a point) re-negotiate + * active sessions and reduce their slot usage to make rooom for new + * connections. For now we just fail the create session. */ static int set_forechannel_maxreqs(struct nfsd4_channel_attrs *fchan) { - int np; + int mem; if (fchan->maxreqs < 1) return nfserr_inval; else if (fchan->maxreqs > NFSD_MAX_SLOTS_PER_SESSION) fchan->maxreqs = NFSD_MAX_SLOTS_PER_SESSION; - np = fchan->maxreqs * NFSD_PAGES_PER_SLOT; + mem = fchan->maxreqs * NFSD_SLOT_CACHE_SIZE; spin_lock(&nfsd_drc_lock); - if (np + nfsd_drc_pages_used > nfsd_drc_max_pages) - np = nfsd_drc_max_pages - nfsd_drc_pages_used; - nfsd_drc_pages_used += np; + if (mem + nfsd_drc_mem_used > nfsd_drc_max_mem) + mem = ((nfsd_drc_max_mem - nfsd_drc_mem_used) / + NFSD_SLOT_CACHE_SIZE) * NFSD_SLOT_CACHE_SIZE; + nfsd_drc_mem_used += mem; spin_unlock(&nfsd_drc_lock); - fchan->maxreqs = np / NFSD_PAGES_PER_SLOT; + fchan->maxreqs = mem / NFSD_SLOT_CACHE_SIZE; if (fchan->maxreqs == 0) return nfserr_resource; return 0; @@ -465,9 +465,7 @@ static int init_forechannel_attrs(struct svc_rqst *rqstp, fchan->maxresp_sz = maxcount; session_fchan->maxresp_sz = fchan->maxresp_sz; - /* Set the max response cached size our default which is - * a multiple of PAGE_SIZE and small */ - session_fchan->maxresp_cached = NFSD_PAGES_PER_SLOT * PAGE_SIZE; + session_fchan->maxresp_cached = NFSD_SLOT_CACHE_SIZE; fchan->maxresp_cached = session_fchan->maxresp_cached; /* Use the client's maxops if possible */ @@ -585,7 +583,7 @@ free_session(struct kref *kref) nfsd4_release_respages(e->ce_respages, e->ce_resused); } spin_lock(&nfsd_drc_lock); - nfsd_drc_pages_used -= ses->se_fchannel.maxreqs * NFSD_PAGES_PER_SLOT; + nfsd_drc_mem_used -= ses->se_fchannel.maxreqs * NFSD_SLOT_CACHE_SIZE; spin_unlock(&nfsd_drc_lock); kfree(ses); } diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c index 9be2a1932f8a..5a280a9cb540 100644 --- a/fs/nfsd/nfssvc.c +++ b/fs/nfsd/nfssvc.c @@ -74,8 +74,8 @@ struct svc_serv *nfsd_serv; * nfsd_drc_pages_used tracks the current version 4.1 DRC memory usage. */ spinlock_t nfsd_drc_lock; -unsigned int nfsd_drc_max_pages; -unsigned int nfsd_drc_pages_used; +unsigned int nfsd_drc_max_mem; +unsigned int nfsd_drc_mem_used; #if defined(CONFIG_NFSD_V2_ACL) || defined(CONFIG_NFSD_V3_ACL) static struct svc_stat nfsd_acl_svcstats; @@ -247,12 +247,11 @@ void nfsd_reset_versions(void) static void set_max_drc(void) { #define NFSD_DRC_SIZE_SHIFT 10 - nfsd_drc_max_pages = nr_free_buffer_pages() - >> NFSD_DRC_SIZE_SHIFT; - nfsd_drc_pages_used = 0; + nfsd_drc_max_mem = (nr_free_buffer_pages() + >> NFSD_DRC_SIZE_SHIFT) * PAGE_SIZE; + nfsd_drc_mem_used = 0; spin_lock_init(&nfsd_drc_lock); - dprintk("%s nfsd_drc_max_pages %u\n", __func__, - nfsd_drc_max_pages); + dprintk("%s nfsd_drc_max_mem %u \n", __func__, nfsd_drc_max_mem); } int nfsd_create_serv(void) diff --git a/include/linux/nfsd/nfsd.h b/include/linux/nfsd/nfsd.h index 2571f856908f..2812ed52669d 100644 --- a/include/linux/nfsd/nfsd.h +++ b/include/linux/nfsd/nfsd.h @@ -57,8 +57,8 @@ extern u32 nfsd_supported_minorversion; extern struct mutex nfsd_mutex; extern struct svc_serv *nfsd_serv; extern spinlock_t nfsd_drc_lock; -extern unsigned int nfsd_drc_max_pages; -extern unsigned int nfsd_drc_pages_used; +extern unsigned int nfsd_drc_max_mem; +extern unsigned int nfsd_drc_mem_used; extern struct seq_operations nfs_exports_op; diff --git a/include/linux/nfsd/state.h b/include/linux/nfsd/state.h index 57ab2ed08459..a6c87d623891 100644 --- a/include/linux/nfsd/state.h +++ b/include/linux/nfsd/state.h @@ -96,6 +96,7 @@ struct nfs4_cb_conn { #define NFSD_MAX_SLOTS_PER_SESSION 128 /* Maximum number of pages per slot cache entry */ #define NFSD_PAGES_PER_SLOT 1 +#define NFSD_SLOT_CACHE_SIZE PAGE_SIZE /* Maximum number of operations per session compound */ #define NFSD_MAX_OPS_PER_COMPOUND 16 -- cgit v1.2.3 From 49557cc74c7bdf6a984be227ead9a84b3a26f053 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Thu, 23 Jul 2009 19:02:16 -0400 Subject: nfsd41: Use separate DRC for setclientid Instead of trying to share the generic 4.1 reply cache code for the CREATE_SESSION reply cache, it's simpler to handle CREATE_SESSION separately. The nfs41 single slot clientid DRC holds the results of create session processing. CREATE_SESSION can be preceeded by a SEQUENCE operation (an embedded CREATE_SESSION) and the create session single slot cache must be maintained. nfsd4_replay_cache_entry() and nfsd4_store_cache_entry() do not implement the replay of an embedded CREATE_SESSION. The clientid DRC slot does not need the inuse, cachethis or other fields that the multiple slot session cache uses. Replace the clientid DRC cache struct nfs4_slot cache with a new nfsd4_clid_slot cache. Save the xdr struct nfsd4_create_session into the cache at the end of processing, and on a replay, replace the struct for the replay request with the cached version all while under the state lock. nfsd4_proc_compound will handle both the solo and embedded CREATE_SESSION case via the normal use of encode_operation. Errors that do not change the create session cache: A create session NFS4ERR_STALE_CLIENTID error means that a client record (and associated create session slot) could not be found and therefore can't be changed. NFSERR_SEQ_MISORDERED errors do not change the slot cache. All other errors get cached. Remove the clientid DRC specific check in nfs4svc_encode_compoundres to put the session only if cstate.session is set which will now always be true. Signed-off-by: Andy Adamson Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4proc.c | 2 +- fs/nfsd/nfs4state.c | 64 +++++++++++++++++++++++++++------------------- fs/nfsd/nfs4xdr.c | 3 +-- include/linux/nfsd/state.h | 21 ++++++++++++++- include/linux/nfsd/xdr4.h | 12 --------- 5 files changed, 60 insertions(+), 42 deletions(-) (limited to 'include') diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index d781658e8084..d606c6a427de 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -1120,7 +1120,7 @@ nfsd4_proc_compound(struct svc_rqst *rqstp, BUG_ON(op->status == nfs_ok); encode_op: - /* Only from SEQUENCE or CREATE_SESSION */ + /* Only from SEQUENCE */ if (resp->cstate.status == nfserr_replay_cache) { dprintk("%s NFS4.1 replay from cache\n", __func__); if (nfsd4_not_cached(resp)) diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 99df8e7a687b..7729d092c8a5 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -653,8 +653,6 @@ static inline void free_client(struct nfs4_client *clp) { shutdown_callback_client(clp); - nfsd4_release_respages(clp->cl_slot.sl_cache_entry.ce_respages, - clp->cl_slot.sl_cache_entry.ce_resused); if (clp->cl_cred.cr_group_info) put_group_info(clp->cl_cred.cr_group_info); kfree(clp->cl_principal); @@ -1293,12 +1291,11 @@ out_copy: exid->clientid.cl_boot = new->cl_clientid.cl_boot; exid->clientid.cl_id = new->cl_clientid.cl_id; - new->cl_slot.sl_seqid = 0; exid->seqid = 1; nfsd4_set_ex_flags(new, exid); dprintk("nfsd4_exchange_id seqid %d flags %x\n", - new->cl_slot.sl_seqid, new->cl_exchange_flags); + new->cl_cs_slot.sl_seqid, new->cl_exchange_flags); status = nfs_ok; out: @@ -1334,15 +1331,35 @@ check_slot_seqid(u32 seqid, u32 slot_seqid, int slot_inuse) return nfserr_seq_misordered; } +/* + * Cache the create session result into the create session single DRC + * slot cache by saving the xdr structure. sl_seqid has been set. + * Do this for solo or embedded create session operations. + */ +static void +nfsd4_cache_create_session(struct nfsd4_create_session *cr_ses, + struct nfsd4_clid_slot *slot, int nfserr) +{ + slot->sl_status = nfserr; + memcpy(&slot->sl_cr_ses, cr_ses, sizeof(*cr_ses)); +} + +static __be32 +nfsd4_replay_create_session(struct nfsd4_create_session *cr_ses, + struct nfsd4_clid_slot *slot) +{ + memcpy(cr_ses, &slot->sl_cr_ses, sizeof(*cr_ses)); + return slot->sl_status; +} + __be32 nfsd4_create_session(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, struct nfsd4_create_session *cr_ses) { u32 ip_addr = svc_addr_in(rqstp)->sin_addr.s_addr; - struct nfsd4_compoundres *resp = rqstp->rq_resp; struct nfs4_client *conf, *unconf; - struct nfsd4_slot *slot = NULL; + struct nfsd4_clid_slot *cs_slot = NULL; int status = 0; nfs4_lock_state(); @@ -1350,25 +1367,22 @@ nfsd4_create_session(struct svc_rqst *rqstp, conf = find_confirmed_client(&cr_ses->clientid); if (conf) { - slot = &conf->cl_slot; - status = check_slot_seqid(cr_ses->seqid, slot->sl_seqid, - slot->sl_inuse); + cs_slot = &conf->cl_cs_slot; + status = check_slot_seqid(cr_ses->seqid, cs_slot->sl_seqid, 0); if (status == nfserr_replay_cache) { dprintk("Got a create_session replay! seqid= %d\n", - slot->sl_seqid); - cstate->slot = slot; - cstate->status = status; + cs_slot->sl_seqid); /* Return the cached reply status */ - status = nfsd4_replay_cache_entry(resp, NULL); + status = nfsd4_replay_create_session(cr_ses, cs_slot); goto out; - } else if (cr_ses->seqid != conf->cl_slot.sl_seqid + 1) { + } else if (cr_ses->seqid != cs_slot->sl_seqid + 1) { status = nfserr_seq_misordered; dprintk("Sequence misordered!\n"); dprintk("Expected seqid= %d but got seqid= %d\n", - slot->sl_seqid, cr_ses->seqid); + cs_slot->sl_seqid, cr_ses->seqid); goto out; } - conf->cl_slot.sl_seqid++; + cs_slot->sl_seqid++; } else if (unconf) { if (!same_creds(&unconf->cl_cred, &rqstp->rq_cred) || (ip_addr != unconf->cl_addr)) { @@ -1376,16 +1390,15 @@ nfsd4_create_session(struct svc_rqst *rqstp, goto out; } - slot = &unconf->cl_slot; - status = check_slot_seqid(cr_ses->seqid, slot->sl_seqid, - slot->sl_inuse); + cs_slot = &unconf->cl_cs_slot; + status = check_slot_seqid(cr_ses->seqid, cs_slot->sl_seqid, 0); if (status) { /* an unconfirmed replay returns misordered */ status = nfserr_seq_misordered; - goto out; + goto out_cache; } - slot->sl_seqid++; /* from 0 to 1 */ + cs_slot->sl_seqid++; /* from 0 to 1 */ move_to_confirmed(unconf); /* @@ -1406,12 +1419,11 @@ nfsd4_create_session(struct svc_rqst *rqstp, memcpy(cr_ses->sessionid.data, conf->cl_sessionid.data, NFS4_MAX_SESSIONID_LEN); - cr_ses->seqid = slot->sl_seqid; + cr_ses->seqid = cs_slot->sl_seqid; - slot->sl_inuse = true; - cstate->slot = slot; - /* Ensure a page is used for the cache */ - slot->sl_cache_entry.ce_cachethis = 1; +out_cache: + /* cache solo and embedded create sessions under the state lock */ + nfsd4_cache_create_session(cr_ses, cs_slot, status); out: nfs4_unlock_state(); dprintk("%s returns %d\n", __func__, ntohl(status)); diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index 2dcc7feaa6ff..fdf632bf1cfe 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -3313,8 +3313,7 @@ nfs4svc_encode_compoundres(struct svc_rqst *rqstp, __be32 *p, struct nfsd4_compo dprintk("%s: SET SLOT STATE TO AVAILABLE\n", __func__); resp->cstate.slot->sl_inuse = 0; } - if (resp->cstate.session) - nfsd4_put_session(resp->cstate.session); + nfsd4_put_session(resp->cstate.session); } return 1; } diff --git a/include/linux/nfsd/state.h b/include/linux/nfsd/state.h index a6c87d623891..58bb19784e12 100644 --- a/include/linux/nfsd/state.h +++ b/include/linux/nfsd/state.h @@ -127,6 +127,25 @@ struct nfsd4_channel_attrs { u32 rdma_attrs; }; +struct nfsd4_create_session { + clientid_t clientid; + struct nfs4_sessionid sessionid; + u32 seqid; + u32 flags; + struct nfsd4_channel_attrs fore_channel; + struct nfsd4_channel_attrs back_channel; + u32 callback_prog; + u32 uid; + u32 gid; +}; + +/* The single slot clientid cache structure */ +struct nfsd4_clid_slot { + u32 sl_seqid; + __be32 sl_status; + struct nfsd4_create_session sl_cr_ses; +}; + struct nfsd4_session { struct kref se_ref; struct list_head se_hash; /* hash by sessionid */ @@ -193,7 +212,7 @@ struct nfs4_client { /* for nfs41 */ struct list_head cl_sessions; - struct nfsd4_slot cl_slot; /* create_session slot */ + struct nfsd4_clid_slot cl_cs_slot; /* create_session slot */ u32 cl_exchange_flags; struct nfs4_sessionid cl_sessionid; }; diff --git a/include/linux/nfsd/xdr4.h b/include/linux/nfsd/xdr4.h index 2bacf7535069..5e4beb0deb80 100644 --- a/include/linux/nfsd/xdr4.h +++ b/include/linux/nfsd/xdr4.h @@ -366,18 +366,6 @@ struct nfsd4_exchange_id { int spa_how; }; -struct nfsd4_create_session { - clientid_t clientid; - struct nfs4_sessionid sessionid; - u32 seqid; - u32 flags; - struct nfsd4_channel_attrs fore_channel; - struct nfsd4_channel_attrs back_channel; - u32 callback_prog; - u32 uid; - u32 gid; -}; - struct nfsd4_sequence { struct nfs4_sessionid sessionid; /* request/response */ u32 seqid; /* request/response */ -- cgit v1.2.3 From 5920dadfb4aec6c1372c5570e71bcd3b4837e63c Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Wed, 15 Jul 2009 17:11:41 +0900 Subject: libata: accept late unlocking of HPA On certain configurations, HPA isn't or can't be unlocked during probing but it somehow ends up unlocked afterwards. In the following thread, the problem can be reliably reproduced after resuming from STR. The BIOS turns on HPA during boot but forgets to do it during resume. http://thread.gmane.org/gmane.linux.kernel/858310 This patch updates libata revalidation such that it considers native n_sectors. If the device size has increased to match native n_sectors, it's assumed that HPA has been unlocked involuntarily and the device is recognized as the same one. This should be fairly safe while nicely working around the problem. Signed-off-by: Tejun Heo Reported-by: Christof Warlich Signed-off-by: Jeff Garzik --- drivers/ata/libata-core.c | 30 +++++++++++++++++++++++------- include/linux/libata.h | 1 + 2 files changed, 24 insertions(+), 7 deletions(-) (limited to 'include') diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c index 2c6aedaef718..8ac98ff16d7d 100644 --- a/drivers/ata/libata-core.c +++ b/drivers/ata/libata-core.c @@ -1515,6 +1515,7 @@ static int ata_hpa_resize(struct ata_device *dev) return rc; } + dev->n_native_sectors = native_sectors; /* nothing to do? */ if (native_sectors <= sectors || !ata_ignore_hpa) { @@ -4099,6 +4100,7 @@ int ata_dev_revalidate(struct ata_device *dev, unsigned int new_class, unsigned int readid_flags) { u64 n_sectors = dev->n_sectors; + u64 n_native_sectors = dev->n_native_sectors; int rc; if (!ata_dev_enabled(dev)) @@ -4128,16 +4130,30 @@ int ata_dev_revalidate(struct ata_device *dev, unsigned int new_class, /* verify n_sectors hasn't changed */ if (dev->class == ATA_DEV_ATA && n_sectors && dev->n_sectors != n_sectors) { - ata_dev_printk(dev, KERN_INFO, "n_sectors mismatch " + ata_dev_printk(dev, KERN_WARNING, "n_sectors mismatch " "%llu != %llu\n", (unsigned long long)n_sectors, (unsigned long long)dev->n_sectors); - - /* restore original n_sectors */ - dev->n_sectors = n_sectors; - - rc = -ENODEV; - goto fail; + /* + * Something could have caused HPA to be unlocked + * involuntarily. If n_native_sectors hasn't changed + * and the new size matches it, keep the device. + */ + if (dev->n_native_sectors == n_native_sectors && + dev->n_sectors > n_sectors && + dev->n_sectors == n_native_sectors) { + ata_dev_printk(dev, KERN_WARNING, + "new n_sectors matches native, probably " + "late HPA unlock, continuing\n"); + /* keep using the old n_sectors */ + dev->n_sectors = n_sectors; + } else { + /* restore original n_[native]_sectors and fail */ + dev->n_native_sectors = n_native_sectors; + dev->n_sectors = n_sectors; + rc = -ENODEV; + goto fail; + } } return 0; diff --git a/include/linux/libata.h b/include/linux/libata.h index 79b6d7fd4ac2..e5b6e33c6571 100644 --- a/include/linux/libata.h +++ b/include/linux/libata.h @@ -589,6 +589,7 @@ struct ata_device { #endif /* n_sector is CLEAR_BEGIN, read comment above CLEAR_BEGIN */ u64 n_sectors; /* size of device, if ATA */ + u64 n_native_sectors; /* native size, if ATA */ unsigned int class; /* ATA_DEV_xxx */ unsigned long unpark_deadline; -- cgit v1.2.3 From e024e11070a0a0dc7163ce1ec2da354a638bdbed Mon Sep 17 00:00:00 2001 From: Dave Airlie Date: Wed, 24 Jun 2009 09:48:08 +1000 Subject: drm/radeon/kms: add initial colortiling support. MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This adds new set/get tiling interfaces where the pitch and macro/micro tiling enables can be set. Along with a flag to decide if this object should have a surface when mapped. The only thing we need to allocate with a mapped surface should be the frontbuffer. Note rotate scanout shouldn't require one, and back/depth shouldn't either, though mesa needs some fixes. It fixes the TTM interfaces along Thomas's suggestions, and I've tested the surface stealing code with two X servers and not seen any lockdep issues. I've stopped tiling the fbcon frontbuffer, as I don't see there being any advantage other than testing, I've left the testing commands in there, just flip the fb_tiled to true in radeon_fb.c Open: Can we integrate endian swapping in with this? Future features: texture tiling - need to relocate texture registers TXOFFSET* with tiling info. This also merges Michel's cleanup surfaces regs at init time patch even though it makes sense on its own, this patch really relies on it. Some PowerMac firmwares set up a tiling surface at the beginning of VRAM which messes us up otherwise. that patch is: Signed-off-by: Michel Dänzer Signed-off-by: Dave Airlie --- drivers/gpu/drm/radeon/atombios_crtc.c | 11 ++- drivers/gpu/drm/radeon/r100.c | 79 ++++++++++++++++- drivers/gpu/drm/radeon/r300.c | 51 ++++++++++- drivers/gpu/drm/radeon/r300_reg.h | 4 +- drivers/gpu/drm/radeon/radeon.h | 32 ++++++- drivers/gpu/drm/radeon/radeon_asic.h | 19 ++++ drivers/gpu/drm/radeon/radeon_device.c | 2 + drivers/gpu/drm/radeon/radeon_fb.c | 12 ++- drivers/gpu/drm/radeon/radeon_gem.c | 41 +++++++++ drivers/gpu/drm/radeon/radeon_kms.c | 2 + drivers/gpu/drm/radeon/radeon_legacy_crtc.c | 15 ++-- drivers/gpu/drm/radeon/radeon_object.c | 130 ++++++++++++++++++++++++++++ drivers/gpu/drm/radeon/radeon_ttm.c | 2 + drivers/gpu/drm/ttm/ttm_bo.c | 5 +- drivers/gpu/drm/ttm/ttm_bo_vm.c | 3 + include/drm/radeon_drm.h | 23 ++++- include/drm/ttm/ttm_bo_driver.h | 15 ++++ 17 files changed, 427 insertions(+), 19 deletions(-) (limited to 'include') diff --git a/drivers/gpu/drm/radeon/atombios_crtc.c b/drivers/gpu/drm/radeon/atombios_crtc.c index e64a199b5ee1..eac26cdb5dae 100644 --- a/drivers/gpu/drm/radeon/atombios_crtc.c +++ b/drivers/gpu/drm/radeon/atombios_crtc.c @@ -327,7 +327,7 @@ int atombios_crtc_set_base(struct drm_crtc *crtc, int x, int y, struct drm_gem_object *obj; struct drm_radeon_gem_object *obj_priv; uint64_t fb_location; - uint32_t fb_format, fb_pitch_pixels; + uint32_t fb_format, fb_pitch_pixels, tiling_flags; if (!crtc->fb) return -EINVAL; @@ -364,7 +364,14 @@ int atombios_crtc_set_base(struct drm_crtc *crtc, int x, int y, return -EINVAL; } - /* TODO tiling */ + radeon_object_get_tiling_flags(obj->driver_private, + &tiling_flags, NULL); + if (tiling_flags & RADEON_TILING_MACRO) + fb_format |= AVIVO_D1GRPH_MACRO_ADDRESS_MODE; + + if (tiling_flags & RADEON_TILING_MICRO) + fb_format |= AVIVO_D1GRPH_TILED; + if (radeon_crtc->crtc_id == 0) WREG32(AVIVO_D1VGA_CONTROL, 0); else diff --git a/drivers/gpu/drm/radeon/r100.c b/drivers/gpu/drm/radeon/r100.c index 0d05909f03f6..69bd7cb59972 100644 --- a/drivers/gpu/drm/radeon/r100.c +++ b/drivers/gpu/drm/radeon/r100.c @@ -909,6 +909,7 @@ static int r100_packet0_check(struct radeon_cs_parser *p, unsigned idx; bool onereg; int r; + u32 tile_flags = 0; ib = p->ib->ptr; ib_chunk = &p->chunks[p->chunk_ib_idx]; @@ -942,7 +943,20 @@ static int r100_packet0_check(struct radeon_cs_parser *p, } tmp = ib_chunk->kdata[idx] & 0x003fffff; tmp += (((u32)reloc->lobj.gpu_offset) >> 10); - ib[idx] = (ib_chunk->kdata[idx] & 0xffc00000) | tmp; + + if (reloc->lobj.tiling_flags & RADEON_TILING_MACRO) + tile_flags |= RADEON_DST_TILE_MACRO; + if (reloc->lobj.tiling_flags & RADEON_TILING_MICRO) { + if (reg == RADEON_SRC_PITCH_OFFSET) { + DRM_ERROR("Cannot src blit from microtiled surface\n"); + r100_cs_dump_packet(p, pkt); + return -EINVAL; + } + tile_flags |= RADEON_DST_TILE_MICRO; + } + + tmp |= tile_flags; + ib[idx] = (ib_chunk->kdata[idx] & 0x3fc00000) | tmp; break; case RADEON_RB3D_DEPTHOFFSET: case RADEON_RB3D_COLOROFFSET: @@ -987,6 +1001,25 @@ static int r100_packet0_check(struct radeon_cs_parser *p, } ib[idx] = ib_chunk->kdata[idx] + ((u32)reloc->lobj.gpu_offset); break; + case R300_RB3D_COLORPITCH0: + case RADEON_RB3D_COLORPITCH: + r = r100_cs_packet_next_reloc(p, &reloc); + if (r) { + DRM_ERROR("No reloc for ib[%d]=0x%04X\n", + idx, reg); + r100_cs_dump_packet(p, pkt); + return r; + } + + if (reloc->lobj.tiling_flags & RADEON_TILING_MACRO) + tile_flags |= RADEON_COLOR_TILE_ENABLE; + if (reloc->lobj.tiling_flags & RADEON_TILING_MICRO) + tile_flags |= RADEON_COLOR_MICROTILE_ENABLE; + + tmp = ib_chunk->kdata[idx] & ~(0x7 << 16); + tmp |= tile_flags; + ib[idx] = tmp; + break; default: /* FIXME: we don't want to allow anyothers packet */ break; @@ -1707,3 +1740,47 @@ int r100_debugfs_mc_info_init(struct radeon_device *rdev) return 0; #endif } + +int r100_set_surface_reg(struct radeon_device *rdev, int reg, + uint32_t tiling_flags, uint32_t pitch, + uint32_t offset, uint32_t obj_size) +{ + int surf_index = reg * 16; + int flags = 0; + + /* r100/r200 divide by 16 */ + if (rdev->family < CHIP_R300) + flags = pitch / 16; + else + flags = pitch / 8; + + if (rdev->family <= CHIP_RS200) { + if ((tiling_flags & (RADEON_TILING_MACRO|RADEON_TILING_MICRO)) + == (RADEON_TILING_MACRO|RADEON_TILING_MICRO)) + flags |= RADEON_SURF_TILE_COLOR_BOTH; + if (tiling_flags & RADEON_TILING_MACRO) + flags |= RADEON_SURF_TILE_COLOR_MACRO; + } else if (rdev->family <= CHIP_RV280) { + if (tiling_flags & (RADEON_TILING_MACRO)) + flags |= R200_SURF_TILE_COLOR_MACRO; + if (tiling_flags & RADEON_TILING_MICRO) + flags |= R200_SURF_TILE_COLOR_MICRO; + } else { + if (tiling_flags & RADEON_TILING_MACRO) + flags |= R300_SURF_TILE_MACRO; + if (tiling_flags & RADEON_TILING_MICRO) + flags |= R300_SURF_TILE_MICRO; + } + + DRM_DEBUG("writing surface %d %d %x %x\n", reg, flags, offset, offset+obj_size-1); + WREG32(RADEON_SURFACE0_INFO + surf_index, flags); + WREG32(RADEON_SURFACE0_LOWER_BOUND + surf_index, offset); + WREG32(RADEON_SURFACE0_UPPER_BOUND + surf_index, offset + obj_size - 1); + return 0; +} + +void r100_clear_surface_reg(struct radeon_device *rdev, int reg) +{ + int surf_index = reg * 16; + WREG32(RADEON_SURFACE0_INFO + surf_index, 0); +} diff --git a/drivers/gpu/drm/radeon/r300.c b/drivers/gpu/drm/radeon/r300.c index 0e0e094da503..28e5777658be 100644 --- a/drivers/gpu/drm/radeon/r300.c +++ b/drivers/gpu/drm/radeon/r300.c @@ -30,6 +30,7 @@ #include "drm.h" #include "radeon_reg.h" #include "radeon.h" +#include "radeon_drm.h" /* r300,r350,rv350,rv370,rv380 depends on : */ void r100_hdp_reset(struct radeon_device *rdev); @@ -1023,7 +1024,7 @@ static int r300_packet0_check(struct radeon_cs_parser *p, struct radeon_cs_reloc *reloc; struct r300_cs_track *track; volatile uint32_t *ib; - uint32_t tmp; + uint32_t tmp, tile_flags = 0; unsigned i; int r; @@ -1052,7 +1053,19 @@ static int r300_packet0_check(struct radeon_cs_parser *p, } tmp = ib_chunk->kdata[idx] & 0x003fffff; tmp += (((u32)reloc->lobj.gpu_offset) >> 10); - ib[idx] = (ib_chunk->kdata[idx] & 0xffc00000) | tmp; + + if (reloc->lobj.tiling_flags & RADEON_TILING_MACRO) + tile_flags |= RADEON_DST_TILE_MACRO; + if (reloc->lobj.tiling_flags & RADEON_TILING_MICRO) { + if (reg == RADEON_SRC_PITCH_OFFSET) { + DRM_ERROR("Cannot src blit from microtiled surface\n"); + r100_cs_dump_packet(p, pkt); + return -EINVAL; + } + tile_flags |= RADEON_DST_TILE_MICRO; + } + tmp |= tile_flags; + ib[idx] = (ib_chunk->kdata[idx] & 0x3fc00000) | tmp; break; case R300_RB3D_COLOROFFSET0: case R300_RB3D_COLOROFFSET1: @@ -1141,6 +1154,23 @@ static int r300_packet0_check(struct radeon_cs_parser *p, /* RB3D_COLORPITCH1 */ /* RB3D_COLORPITCH2 */ /* RB3D_COLORPITCH3 */ + r = r100_cs_packet_next_reloc(p, &reloc); + if (r) { + DRM_ERROR("No reloc for ib[%d]=0x%04X\n", + idx, reg); + r100_cs_dump_packet(p, pkt); + return r; + } + + if (reloc->lobj.tiling_flags & RADEON_TILING_MACRO) + tile_flags |= R300_COLOR_TILE_ENABLE; + if (reloc->lobj.tiling_flags & RADEON_TILING_MICRO) + tile_flags |= R300_COLOR_MICROTILE_ENABLE; + + tmp = ib_chunk->kdata[idx] & ~(0x7 << 16); + tmp |= tile_flags; + ib[idx] = tmp; + i = (reg - 0x4E38) >> 2; track->cb[i].pitch = ib_chunk->kdata[idx] & 0x3FFE; switch (((ib_chunk->kdata[idx] >> 21) & 0xF)) { @@ -1196,6 +1226,23 @@ static int r300_packet0_check(struct radeon_cs_parser *p, break; case 0x4F24: /* ZB_DEPTHPITCH */ + r = r100_cs_packet_next_reloc(p, &reloc); + if (r) { + DRM_ERROR("No reloc for ib[%d]=0x%04X\n", + idx, reg); + r100_cs_dump_packet(p, pkt); + return r; + } + + if (reloc->lobj.tiling_flags & RADEON_TILING_MACRO) + tile_flags |= R300_DEPTHMACROTILE_ENABLE; + if (reloc->lobj.tiling_flags & RADEON_TILING_MICRO) + tile_flags |= R300_DEPTHMICROTILE_TILED;; + + tmp = ib_chunk->kdata[idx] & ~(0x7 << 16); + tmp |= tile_flags; + ib[idx] = tmp; + track->zb.pitch = ib_chunk->kdata[idx] & 0x3FFC; break; case 0x4104: diff --git a/drivers/gpu/drm/radeon/r300_reg.h b/drivers/gpu/drm/radeon/r300_reg.h index 70f48609515e..4b7afef35a65 100644 --- a/drivers/gpu/drm/radeon/r300_reg.h +++ b/drivers/gpu/drm/radeon/r300_reg.h @@ -27,7 +27,9 @@ #ifndef _R300_REG_H_ #define _R300_REG_H_ - +#define R300_SURF_TILE_MACRO (1<<16) +#define R300_SURF_TILE_MICRO (2<<16) +#define R300_SURF_TILE_BOTH (3<<16) #define R300_MC_INIT_MISC_LAT_TIMER 0x180 diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h index 7f007185e7f7..af12a2fe3221 100644 --- a/drivers/gpu/drm/radeon/radeon.h +++ b/drivers/gpu/drm/radeon/radeon.h @@ -201,6 +201,14 @@ int radeon_fence_wait_last(struct radeon_device *rdev); struct radeon_fence *radeon_fence_ref(struct radeon_fence *fence); void radeon_fence_unref(struct radeon_fence **fence); +/* + * Tiling registers + */ +struct radeon_surface_reg { + struct radeon_object *robj; +}; + +#define RADEON_GEM_MAX_SURFACES 8 /* * Radeon buffer. @@ -213,6 +221,7 @@ struct radeon_object_list { uint64_t gpu_offset; unsigned rdomain; unsigned wdomain; + uint32_t tiling_flags; }; int radeon_object_init(struct radeon_device *rdev); @@ -242,8 +251,15 @@ void radeon_object_list_clean(struct list_head *head); int radeon_object_fbdev_mmap(struct radeon_object *robj, struct vm_area_struct *vma); unsigned long radeon_object_size(struct radeon_object *robj); - - +void radeon_object_clear_surface_reg(struct radeon_object *robj); +int radeon_object_check_tiling(struct radeon_object *robj, bool has_moved, + bool force_drop); +void radeon_object_set_tiling_flags(struct radeon_object *robj, + uint32_t tiling_flags, uint32_t pitch); +void radeon_object_get_tiling_flags(struct radeon_object *robj, uint32_t *tiling_flags, uint32_t *pitch); +void radeon_bo_move_notify(struct ttm_buffer_object *bo, + struct ttm_mem_reg *mem); +void radeon_bo_fault_reserve_notify(struct ttm_buffer_object *bo); /* * GEM objects. */ @@ -535,6 +551,11 @@ struct radeon_asic { void (*set_memory_clock)(struct radeon_device *rdev, uint32_t mem_clock); void (*set_pcie_lanes)(struct radeon_device *rdev, int lanes); void (*set_clock_gating)(struct radeon_device *rdev, int enable); + + int (*set_surface_reg)(struct radeon_device *rdev, int reg, + uint32_t tiling_flags, uint32_t pitch, + uint32_t offset, uint32_t obj_size); + int (*clear_surface_reg)(struct radeon_device *rdev, int reg); }; union radeon_asic_config { @@ -568,6 +589,10 @@ int radeon_gem_busy_ioctl(struct drm_device *dev, void *data, int radeon_gem_wait_idle_ioctl(struct drm_device *dev, void *data, struct drm_file *filp); int radeon_cs_ioctl(struct drm_device *dev, void *data, struct drm_file *filp); +int radeon_gem_set_tiling_ioctl(struct drm_device *dev, void *data, + struct drm_file *filp); +int radeon_gem_get_tiling_ioctl(struct drm_device *dev, void *data, + struct drm_file *filp); /* @@ -627,6 +652,7 @@ struct radeon_device { bool shutdown; bool suspend; bool need_dma32; + struct radeon_surface_reg surface_regs[RADEON_GEM_MAX_SURFACES]; }; int radeon_device_init(struct radeon_device *rdev, @@ -801,5 +827,7 @@ static inline void radeon_ring_write(struct radeon_device *rdev, uint32_t v) #define radeon_set_memory_clock(rdev, e) (rdev)->asic->set_engine_clock((rdev), (e)) #define radeon_set_pcie_lanes(rdev, l) (rdev)->asic->set_pcie_lanes((rdev), (l)) #define radeon_set_clock_gating(rdev, e) (rdev)->asic->set_clock_gating((rdev), (e)) +#define radeon_set_surface_reg(rdev, r, f, p, o, s) ((rdev)->asic->set_surface_reg((rdev), (r), (f), (p), (o), (s))) +#define radeon_clear_surface_reg(rdev, r) ((rdev)->asic->clear_surface_reg((rdev), (r))) #endif diff --git a/drivers/gpu/drm/radeon/radeon_asic.h b/drivers/gpu/drm/radeon/radeon_asic.h index e2e567395df8..dd903d329406 100644 --- a/drivers/gpu/drm/radeon/radeon_asic.h +++ b/drivers/gpu/drm/radeon/radeon_asic.h @@ -71,6 +71,10 @@ int r100_copy_blit(struct radeon_device *rdev, uint64_t dst_offset, unsigned num_pages, struct radeon_fence *fence); +int r100_set_surface_reg(struct radeon_device *rdev, int reg, + uint32_t tiling_flags, uint32_t pitch, + uint32_t offset, uint32_t obj_size); +int r100_clear_surface_reg(struct radeon_device *rdev, int reg); static struct radeon_asic r100_asic = { .init = &r100_init, @@ -100,6 +104,8 @@ static struct radeon_asic r100_asic = { .set_memory_clock = NULL, .set_pcie_lanes = NULL, .set_clock_gating = &radeon_legacy_set_clock_gating, + .set_surface_reg = r100_set_surface_reg, + .clear_surface_reg = r100_clear_surface_reg, }; @@ -128,6 +134,7 @@ int r300_copy_dma(struct radeon_device *rdev, uint64_t dst_offset, unsigned num_pages, struct radeon_fence *fence); + static struct radeon_asic r300_asic = { .init = &r300_init, .errata = &r300_errata, @@ -156,6 +163,8 @@ static struct radeon_asic r300_asic = { .set_memory_clock = NULL, .set_pcie_lanes = &rv370_set_pcie_lanes, .set_clock_gating = &radeon_legacy_set_clock_gating, + .set_surface_reg = r100_set_surface_reg, + .clear_surface_reg = r100_clear_surface_reg, }; /* @@ -193,6 +202,8 @@ static struct radeon_asic r420_asic = { .set_memory_clock = &radeon_atom_set_memory_clock, .set_pcie_lanes = &rv370_set_pcie_lanes, .set_clock_gating = &radeon_atom_set_clock_gating, + .set_surface_reg = r100_set_surface_reg, + .clear_surface_reg = r100_clear_surface_reg, }; @@ -237,6 +248,8 @@ static struct radeon_asic rs400_asic = { .set_memory_clock = NULL, .set_pcie_lanes = NULL, .set_clock_gating = &radeon_legacy_set_clock_gating, + .set_surface_reg = r100_set_surface_reg, + .clear_surface_reg = r100_clear_surface_reg, }; @@ -322,6 +335,8 @@ static struct radeon_asic rs690_asic = { .set_memory_clock = &radeon_atom_set_memory_clock, .set_pcie_lanes = NULL, .set_clock_gating = &radeon_atom_set_clock_gating, + .set_surface_reg = r100_set_surface_reg, + .clear_surface_reg = r100_clear_surface_reg, }; @@ -367,6 +382,8 @@ static struct radeon_asic rv515_asic = { .set_memory_clock = &radeon_atom_set_memory_clock, .set_pcie_lanes = &rv370_set_pcie_lanes, .set_clock_gating = &radeon_atom_set_clock_gating, + .set_surface_reg = r100_set_surface_reg, + .clear_surface_reg = r100_clear_surface_reg, }; @@ -405,6 +422,8 @@ static struct radeon_asic r520_asic = { .set_memory_clock = &radeon_atom_set_memory_clock, .set_pcie_lanes = &rv370_set_pcie_lanes, .set_clock_gating = &radeon_atom_set_clock_gating, + .set_surface_reg = r100_set_surface_reg, + .clear_surface_reg = r100_clear_surface_reg, }; /* diff --git a/drivers/gpu/drm/radeon/radeon_device.c b/drivers/gpu/drm/radeon/radeon_device.c index cdef6eb01baf..f23083bbba3f 100644 --- a/drivers/gpu/drm/radeon/radeon_device.c +++ b/drivers/gpu/drm/radeon/radeon_device.c @@ -48,6 +48,8 @@ static void radeon_surface_init(struct radeon_device *rdev) i * (RADEON_SURFACE1_INFO - RADEON_SURFACE0_INFO), 0); } + /* enable surfaces */ + WREG32(RADEON_SURFACE_CNTL, 0); } } diff --git a/drivers/gpu/drm/radeon/radeon_fb.c b/drivers/gpu/drm/radeon/radeon_fb.c index 260870a29d83..36d2f5588f20 100644 --- a/drivers/gpu/drm/radeon/radeon_fb.c +++ b/drivers/gpu/drm/radeon/radeon_fb.c @@ -471,10 +471,10 @@ static struct notifier_block paniced = { .notifier_call = radeonfb_panic, }; -static int radeon_align_pitch(struct radeon_device *rdev, int width, int bpp) +static int radeon_align_pitch(struct radeon_device *rdev, int width, int bpp, bool tiled) { int aligned = width; - int align_large = (ASIC_IS_AVIVO(rdev)); + int align_large = (ASIC_IS_AVIVO(rdev)) || tiled; int pitch_mask = 0; switch (bpp / 8) { @@ -512,12 +512,13 @@ int radeonfb_create(struct radeon_device *rdev, u64 fb_gpuaddr; void *fbptr = NULL; unsigned long tmp; + bool fb_tiled = false; /* useful for testing */ mode_cmd.width = surface_width; mode_cmd.height = surface_height; mode_cmd.bpp = 32; /* need to align pitch with crtc limits */ - mode_cmd.pitch = radeon_align_pitch(rdev, mode_cmd.width, mode_cmd.bpp) * ((mode_cmd.bpp + 1) / 8); + mode_cmd.pitch = radeon_align_pitch(rdev, mode_cmd.width, mode_cmd.bpp, fb_tiled) * ((mode_cmd.bpp + 1) / 8); mode_cmd.depth = 24; size = mode_cmd.pitch * mode_cmd.height; @@ -535,6 +536,8 @@ int radeonfb_create(struct radeon_device *rdev, } robj = gobj->driver_private; + if (fb_tiled) + radeon_object_set_tiling_flags(robj, RADEON_TILING_MACRO|RADEON_TILING_SURFACE, mode_cmd.pitch); mutex_lock(&rdev->ddev->struct_mutex); fb = radeon_framebuffer_create(rdev->ddev, &mode_cmd, gobj); if (fb == NULL) { @@ -563,6 +566,9 @@ int radeonfb_create(struct radeon_device *rdev, } rfbdev = info->par; + if (fb_tiled) + radeon_object_check_tiling(robj, 0, 0); + ret = radeon_object_kmap(robj, &fbptr); if (ret) { goto out_unref; diff --git a/drivers/gpu/drm/radeon/radeon_gem.c b/drivers/gpu/drm/radeon/radeon_gem.c index eb516034235d..12542087b298 100644 --- a/drivers/gpu/drm/radeon/radeon_gem.c +++ b/drivers/gpu/drm/radeon/radeon_gem.c @@ -285,3 +285,44 @@ int radeon_gem_wait_idle_ioctl(struct drm_device *dev, void *data, mutex_unlock(&dev->struct_mutex); return r; } + +int radeon_gem_set_tiling_ioctl(struct drm_device *dev, void *data, + struct drm_file *filp) +{ + struct drm_radeon_gem_set_tiling *args = data; + struct drm_gem_object *gobj; + struct radeon_object *robj; + int r = 0; + + DRM_DEBUG("%d \n", args->handle); + gobj = drm_gem_object_lookup(dev, filp, args->handle); + if (gobj == NULL) + return -EINVAL; + robj = gobj->driver_private; + radeon_object_set_tiling_flags(robj, args->tiling_flags, args->pitch); + mutex_lock(&dev->struct_mutex); + drm_gem_object_unreference(gobj); + mutex_unlock(&dev->struct_mutex); + return r; +} + +int radeon_gem_get_tiling_ioctl(struct drm_device *dev, void *data, + struct drm_file *filp) +{ + struct drm_radeon_gem_get_tiling *args = data; + struct drm_gem_object *gobj; + struct radeon_object *robj; + int r = 0; + + DRM_DEBUG("\n"); + gobj = drm_gem_object_lookup(dev, filp, args->handle); + if (gobj == NULL) + return -EINVAL; + robj = gobj->driver_private; + radeon_object_get_tiling_flags(robj, &args->tiling_flags, + &args->pitch); + mutex_lock(&dev->struct_mutex); + drm_gem_object_unreference(gobj); + mutex_unlock(&dev->struct_mutex); + return r; +} diff --git a/drivers/gpu/drm/radeon/radeon_kms.c b/drivers/gpu/drm/radeon/radeon_kms.c index 4612a7c146d1..937a2f1cdb46 100644 --- a/drivers/gpu/drm/radeon/radeon_kms.c +++ b/drivers/gpu/drm/radeon/radeon_kms.c @@ -291,5 +291,7 @@ struct drm_ioctl_desc radeon_ioctls_kms[] = { DRM_IOCTL_DEF(DRM_RADEON_GEM_WAIT_IDLE, radeon_gem_wait_idle_ioctl, DRM_AUTH), DRM_IOCTL_DEF(DRM_RADEON_CS, radeon_cs_ioctl, DRM_AUTH), DRM_IOCTL_DEF(DRM_RADEON_INFO, radeon_info_ioctl, DRM_AUTH), + DRM_IOCTL_DEF(DRM_RADEON_GEM_SET_TILING, radeon_gem_set_tiling_ioctl, DRM_AUTH), + DRM_IOCTL_DEF(DRM_RADEON_GEM_GET_TILING, radeon_gem_get_tiling_ioctl, DRM_AUTH), }; int radeon_max_kms_ioctl = DRM_ARRAY_SIZE(radeon_ioctls_kms); diff --git a/drivers/gpu/drm/radeon/radeon_legacy_crtc.c b/drivers/gpu/drm/radeon/radeon_legacy_crtc.c index 14c1a5107fc9..0613790e2a5b 100644 --- a/drivers/gpu/drm/radeon/radeon_legacy_crtc.c +++ b/drivers/gpu/drm/radeon/radeon_legacy_crtc.c @@ -235,6 +235,7 @@ int radeon_crtc_set_base(struct drm_crtc *crtc, int x, int y, uint64_t base; uint32_t crtc_offset, crtc_offset_cntl, crtc_tile_x0_y0 = 0; uint32_t crtc_pitch, pitch_pixels; + uint32_t tiling_flags; DRM_DEBUG("\n"); @@ -258,8 +259,12 @@ int radeon_crtc_set_base(struct drm_crtc *crtc, int x, int y, (crtc->fb->bits_per_pixel * 8)); crtc_pitch |= crtc_pitch << 16; - /* TODO tiling */ - if (0) { + radeon_object_get_tiling_flags(obj->driver_private, + &tiling_flags, NULL); + if (tiling_flags & RADEON_TILING_MICRO) + DRM_ERROR("trying to scanout microtiled buffer\n"); + + if (tiling_flags & RADEON_TILING_MACRO) { if (ASIC_IS_R300(rdev)) crtc_offset_cntl |= (R300_CRTC_X_Y_MODE_EN | R300_CRTC_MICRO_TILE_BUFFER_DIS | @@ -275,15 +280,13 @@ int radeon_crtc_set_base(struct drm_crtc *crtc, int x, int y, crtc_offset_cntl &= ~RADEON_CRTC_TILE_EN; } - - /* TODO more tiling */ - if (0) { + if (tiling_flags & RADEON_TILING_MACRO) { if (ASIC_IS_R300(rdev)) { crtc_tile_x0_y0 = x | (y << 16); base &= ~0x7ff; } else { int byteshift = crtc->fb->bits_per_pixel >> 4; - int tile_addr = (((y >> 3) * crtc->fb->width + x) >> (8 - byteshift)) << 11; + int tile_addr = (((y >> 3) * pitch_pixels + x) >> (8 - byteshift)) << 11; base += tile_addr + ((x << byteshift) % 256) + ((y % 8) << 8); crtc_offset_cntl |= (y % 16); } diff --git a/drivers/gpu/drm/radeon/radeon_object.c b/drivers/gpu/drm/radeon/radeon_object.c index bac0d06c52ac..d5b1fd562d88 100644 --- a/drivers/gpu/drm/radeon/radeon_object.c +++ b/drivers/gpu/drm/radeon/radeon_object.c @@ -44,6 +44,9 @@ struct radeon_object { uint64_t gpu_addr; void *kptr; bool is_iomem; + uint32_t tiling_flags; + uint32_t pitch; + int surface_reg; }; int radeon_ttm_init(struct radeon_device *rdev); @@ -70,6 +73,7 @@ static void radeon_ttm_object_object_destroy(struct ttm_buffer_object *tobj) robj = container_of(tobj, struct radeon_object, tobj); list_del_init(&robj->list); + radeon_object_clear_surface_reg(robj); kfree(robj); } @@ -141,6 +145,7 @@ int radeon_object_create(struct radeon_device *rdev, } robj->rdev = rdev; robj->gobj = gobj; + robj->surface_reg = -1; INIT_LIST_HEAD(&robj->list); flags = radeon_object_flags_from_domain(domain); @@ -435,6 +440,7 @@ int radeon_object_list_validate(struct list_head *head, void *fence) radeon_object_gpu_addr(robj); } lobj->gpu_offset = robj->gpu_addr; + lobj->tiling_flags = robj->tiling_flags; if (fence) { old_fence = (struct radeon_fence *)robj->tobj.sync_obj; robj->tobj.sync_obj = radeon_fence_ref(fence); @@ -479,3 +485,127 @@ unsigned long radeon_object_size(struct radeon_object *robj) { return robj->tobj.num_pages << PAGE_SHIFT; } + +int radeon_object_get_surface_reg(struct radeon_object *robj) +{ + struct radeon_device *rdev = robj->rdev; + struct radeon_surface_reg *reg; + struct radeon_object *old_object; + int steal; + int i; + + if (!robj->tiling_flags) + return 0; + + if (robj->surface_reg >= 0) { + reg = &rdev->surface_regs[robj->surface_reg]; + i = robj->surface_reg; + goto out; + } + + steal = -1; + for (i = 0; i < RADEON_GEM_MAX_SURFACES; i++) { + + reg = &rdev->surface_regs[i]; + if (!reg->robj) + break; + + old_object = reg->robj; + if (old_object->pin_count == 0) + steal = i; + } + + /* if we are all out */ + if (i == RADEON_GEM_MAX_SURFACES) { + if (steal == -1) + return -ENOMEM; + /* find someone with a surface reg and nuke their BO */ + reg = &rdev->surface_regs[steal]; + old_object = reg->robj; + /* blow away the mapping */ + DRM_DEBUG("stealing surface reg %d from %p\n", steal, old_object); + ttm_bo_unmap_virtual(&old_object->tobj); + old_object->surface_reg = -1; + i = steal; + } + + robj->surface_reg = i; + reg->robj = robj; + +out: + radeon_set_surface_reg(rdev, i, robj->tiling_flags, robj->pitch, + robj->tobj.mem.mm_node->start << PAGE_SHIFT, + robj->tobj.num_pages << PAGE_SHIFT); + return 0; +} + +void radeon_object_clear_surface_reg(struct radeon_object *robj) +{ + struct radeon_device *rdev = robj->rdev; + struct radeon_surface_reg *reg; + + if (robj->surface_reg == -1) + return; + + reg = &rdev->surface_regs[robj->surface_reg]; + radeon_clear_surface_reg(rdev, robj->surface_reg); + + reg->robj = NULL; + robj->surface_reg = -1; +} + +void radeon_object_set_tiling_flags(struct radeon_object *robj, + uint32_t tiling_flags, uint32_t pitch) +{ + robj->tiling_flags = tiling_flags; + robj->pitch = pitch; +} + +void radeon_object_get_tiling_flags(struct radeon_object *robj, + uint32_t *tiling_flags, + uint32_t *pitch) +{ + if (tiling_flags) + *tiling_flags = robj->tiling_flags; + if (pitch) + *pitch = robj->pitch; +} + +int radeon_object_check_tiling(struct radeon_object *robj, bool has_moved, + bool force_drop) +{ + if (!(robj->tiling_flags & RADEON_TILING_SURFACE)) + return 0; + + if (force_drop) { + radeon_object_clear_surface_reg(robj); + return 0; + } + + if (robj->tobj.mem.mem_type != TTM_PL_VRAM) { + if (!has_moved) + return 0; + + if (robj->surface_reg >= 0) + radeon_object_clear_surface_reg(robj); + return 0; + } + + if ((robj->surface_reg >= 0) && !has_moved) + return 0; + + return radeon_object_get_surface_reg(robj); +} + +void radeon_bo_move_notify(struct ttm_buffer_object *bo, + struct ttm_mem_reg *mem) +{ + struct radeon_object *robj = container_of(bo, struct radeon_object, tobj); + radeon_object_check_tiling(robj, 0, 1); +} + +void radeon_bo_fault_reserve_notify(struct ttm_buffer_object *bo) +{ + struct radeon_object *robj = container_of(bo, struct radeon_object, tobj); + radeon_object_check_tiling(robj, 0, 0); +} diff --git a/drivers/gpu/drm/radeon/radeon_ttm.c b/drivers/gpu/drm/radeon/radeon_ttm.c index 4ca9aa9203d0..37e1cbcce3a9 100644 --- a/drivers/gpu/drm/radeon/radeon_ttm.c +++ b/drivers/gpu/drm/radeon/radeon_ttm.c @@ -429,6 +429,8 @@ static struct ttm_bo_driver radeon_bo_driver = { .sync_obj_flush = &radeon_sync_obj_flush, .sync_obj_unref = &radeon_sync_obj_unref, .sync_obj_ref = &radeon_sync_obj_ref, + .move_notify = &radeon_bo_move_notify, + .fault_reserve_notify = &radeon_bo_fault_reserve_notify, }; int radeon_ttm_init(struct radeon_device *rdev) diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c index e55e7972c897..6538d4236989 100644 --- a/drivers/gpu/drm/ttm/ttm_bo.c +++ b/drivers/gpu/drm/ttm/ttm_bo.c @@ -43,7 +43,6 @@ #define TTM_BO_HASH_ORDER 13 static int ttm_bo_setup_vm(struct ttm_buffer_object *bo); -static void ttm_bo_unmap_virtual(struct ttm_buffer_object *bo); static int ttm_bo_swapout(struct ttm_mem_shrink *shrink); static inline uint32_t ttm_bo_type_flags(unsigned type) @@ -307,6 +306,9 @@ static int ttm_bo_handle_move_mem(struct ttm_buffer_object *bo, } + if (bdev->driver->move_notify) + bdev->driver->move_notify(bo, mem); + if (!(old_man->flags & TTM_MEMTYPE_FLAG_FIXED) && !(new_man->flags & TTM_MEMTYPE_FLAG_FIXED)) ret = ttm_bo_move_ttm(bo, evict, no_wait, mem); @@ -1451,6 +1453,7 @@ void ttm_bo_unmap_virtual(struct ttm_buffer_object *bo) unmap_mapping_range(bdev->dev_mapping, offset, holelen, 1); } +EXPORT_SYMBOL(ttm_bo_unmap_virtual); static void ttm_bo_vm_insert_rb(struct ttm_buffer_object *bo) { diff --git a/drivers/gpu/drm/ttm/ttm_bo_vm.c b/drivers/gpu/drm/ttm/ttm_bo_vm.c index 40b75032ea47..41c907f6c560 100644 --- a/drivers/gpu/drm/ttm/ttm_bo_vm.c +++ b/drivers/gpu/drm/ttm/ttm_bo_vm.c @@ -101,6 +101,9 @@ static int ttm_bo_vm_fault(struct vm_area_struct *vma, struct vm_fault *vmf) return VM_FAULT_NOPAGE; } + if (bdev->driver->fault_reserve_notify) + bdev->driver->fault_reserve_notify(bo); + /* * Wait for buffer data in transit, due to a pipelined * move. diff --git a/include/drm/radeon_drm.h b/include/drm/radeon_drm.h index 41862e9a4c20..af4b4826997e 100644 --- a/include/drm/radeon_drm.h +++ b/include/drm/radeon_drm.h @@ -506,6 +506,8 @@ typedef struct { #define DRM_RADEON_GEM_WAIT_IDLE 0x24 #define DRM_RADEON_CS 0x26 #define DRM_RADEON_INFO 0x27 +#define DRM_RADEON_GEM_SET_TILING 0x28 +#define DRM_RADEON_GEM_GET_TILING 0x29 #define DRM_IOCTL_RADEON_CP_INIT DRM_IOW( DRM_COMMAND_BASE + DRM_RADEON_CP_INIT, drm_radeon_init_t) #define DRM_IOCTL_RADEON_CP_START DRM_IO( DRM_COMMAND_BASE + DRM_RADEON_CP_START) @@ -544,7 +546,8 @@ typedef struct { #define DRM_IOCTL_RADEON_GEM_WAIT_IDLE DRM_IOW(DRM_COMMAND_BASE + DRM_RADEON_GEM_WAIT_IDLE, struct drm_radeon_gem_wait_idle) #define DRM_IOCTL_RADEON_CS DRM_IOWR(DRM_COMMAND_BASE + DRM_RADEON_CS, struct drm_radeon_cs) #define DRM_IOCTL_RADEON_INFO DRM_IOWR(DRM_COMMAND_BASE + DRM_RADEON_INFO, struct drm_radeon_info) - +#define DRM_IOCTL_RADEON_SET_TILING DRM_IOWR(DRM_COMMAND_BASE + DRM_RADEON_GEM_SET_TILING, struct drm_radeon_gem_set_tiling) +#define DRM_IOCTL_RADEON_GET_TILING DRM_IOWR(DRM_COMMAND_BASE + DRM_RADEON_GEM_GET_TILING, struct drm_radeon_gem_get_tiling) typedef struct drm_radeon_init { enum { @@ -796,6 +799,24 @@ struct drm_radeon_gem_create { uint32_t flags; }; +#define RADEON_TILING_MACRO 0x1 +#define RADEON_TILING_MICRO 0x2 +#define RADEON_TILING_SWAP 0x4 +#define RADEON_TILING_SURFACE 0x8 /* this object requires a surface + * when mapped - i.e. front buffer */ + +struct drm_radeon_gem_set_tiling { + uint32_t handle; + uint32_t tiling_flags; + uint32_t pitch; +}; + +struct drm_radeon_gem_get_tiling { + uint32_t handle; + uint32_t tiling_flags; + uint32_t pitch; +}; + struct drm_radeon_gem_mmap { uint32_t handle; uint32_t pad; diff --git a/include/drm/ttm/ttm_bo_driver.h b/include/drm/ttm/ttm_bo_driver.h index ea83dd23a4d7..a68829db381a 100644 --- a/include/drm/ttm/ttm_bo_driver.h +++ b/include/drm/ttm/ttm_bo_driver.h @@ -354,6 +354,14 @@ struct ttm_bo_driver { int (*sync_obj_flush) (void *sync_obj, void *sync_arg); void (*sync_obj_unref) (void **sync_obj); void *(*sync_obj_ref) (void *sync_obj); + + /* hook to notify driver about a driver move so it + * can do tiling things */ + void (*move_notify)(struct ttm_buffer_object *bo, + struct ttm_mem_reg *new_mem); + /* notify the driver we are taking a fault on this BO + * and have reserved it */ + void (*fault_reserve_notify)(struct ttm_buffer_object *bo); }; #define TTM_NUM_MEM_TYPES 8 @@ -653,6 +661,13 @@ extern int ttm_bo_device_init(struct ttm_bo_device *bdev, struct ttm_bo_driver *driver, uint64_t file_page_offset, bool need_dma32); +/** + * ttm_bo_unmap_virtual + * + * @bo: tear down the virtual mappings for this BO + */ +extern void ttm_bo_unmap_virtual(struct ttm_buffer_object *bo); + /** * ttm_bo_reserve: * -- cgit v1.2.3 From dddac6a7b445de95515f64fdf82fe5dc36c02f26 Mon Sep 17 00:00:00 2001 From: Alan Jenkins Date: Wed, 29 Jul 2009 21:07:55 +0200 Subject: PM / Hibernate: Replace bdget call with simple atomic_inc of i_count Create bdgrab(). This function copies an existing reference to a block_device. It is safe to call from any context. Hibernation code wishes to copy a reference to the active swap device. Right now it calls bdget() under a spinlock, but this is wrong because bdget() can sleep. It doesn't need a full bdget() because we already hold a reference to active swap devices (and the spinlock protects against swapoff). Fixes http://bugzilla.kernel.org/show_bug.cgi?id=13827 Signed-off-by: Alan Jenkins Signed-off-by: Rafael J. Wysocki --- fs/block_dev.c | 10 ++++++++++ include/linux/fs.h | 1 + mm/swapfile.c | 4 ++-- 3 files changed, 13 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/fs/block_dev.c b/fs/block_dev.c index 3a6d4fb2a329..94dfda24c06e 100644 --- a/fs/block_dev.c +++ b/fs/block_dev.c @@ -564,6 +564,16 @@ struct block_device *bdget(dev_t dev) EXPORT_SYMBOL(bdget); +/** + * bdgrab -- Grab a reference to an already referenced block device + * @bdev: Block device to grab a reference to. + */ +struct block_device *bdgrab(struct block_device *bdev) +{ + atomic_inc(&bdev->bd_inode->i_count); + return bdev; +} + long nr_blockdev_pages(void) { struct block_device *bdev; diff --git a/include/linux/fs.h b/include/linux/fs.h index 0872372184fe..a36ffa5a77a4 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -1946,6 +1946,7 @@ extern void putname(const char *name); extern int register_blkdev(unsigned int, const char *); extern void unregister_blkdev(unsigned int, const char *); extern struct block_device *bdget(dev_t); +extern struct block_device *bdgrab(struct block_device *bdev); extern void bd_set_size(struct block_device *, loff_t size); extern void bd_forget(struct inode *inode); extern void bdput(struct block_device *); diff --git a/mm/swapfile.c b/mm/swapfile.c index d1ade1a48ee7..8ffdc0d23c53 100644 --- a/mm/swapfile.c +++ b/mm/swapfile.c @@ -753,7 +753,7 @@ int swap_type_of(dev_t device, sector_t offset, struct block_device **bdev_p) if (!bdev) { if (bdev_p) - *bdev_p = bdget(sis->bdev->bd_dev); + *bdev_p = bdgrab(sis->bdev); spin_unlock(&swap_lock); return i; @@ -765,7 +765,7 @@ int swap_type_of(dev_t device, sector_t offset, struct block_device **bdev_p) struct swap_extent, list); if (se->start_block == offset) { if (bdev_p) - *bdev_p = bdget(sis->bdev->bd_dev); + *bdev_p = bdgrab(sis->bdev); spin_unlock(&swap_lock); bdput(bdev); -- cgit v1.2.3 From e043e42bdb66885b3ac10d27a01ccb9972e2b0a3 Mon Sep 17 00:00:00 2001 From: OGAWA Hirofumi Date: Wed, 29 Jul 2009 12:15:56 -0700 Subject: pty: avoid forcing 'low_latency' tty flag We really don't want to mark the pty as a low-latency device, because as Alan points out, the ->write method can be called from an IRQ (ppp?), and that means we can't use ->low_latency=1 as we take mutexes in the low_latency case. So rather than using low_latency to force the written data to be pushed to the ldisc handling at 'write()' time, just make the reader side (or the poll function) do the flush when it checks whether there is data to be had. This also fixes the problem with lost data in an emacs compile buffer (bugzilla 13815), and we can thus revert the low_latency pty hack (commit 3a54297478e6578f96fd54bf4daa1751130aca86: "pty: quickfix for the pty ENXIO timing problems"). Signed-off-by: OGAWA Hirofumi Tested-by: Aneesh Kumar K.V [ Modified to do the tty_flush_to_ldisc() inside input_available_p() so that it triggers for both read and poll() - Linus] Signed-off-by: Linus Torvalds --- drivers/char/n_tty.c | 1 + drivers/char/pty.c | 2 -- drivers/char/tty_buffer.c | 13 +++++++++++++ include/linux/tty.h | 1 + 4 files changed, 15 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/drivers/char/n_tty.c b/drivers/char/n_tty.c index ff47907ff1bf..973be2f44195 100644 --- a/drivers/char/n_tty.c +++ b/drivers/char/n_tty.c @@ -1583,6 +1583,7 @@ static int n_tty_open(struct tty_struct *tty) static inline int input_available_p(struct tty_struct *tty, int amt) { + tty_flush_to_ldisc(tty); if (tty->icanon) { if (tty->canon_data) return 1; diff --git a/drivers/char/pty.c b/drivers/char/pty.c index 3850a68f265a..6e6942c45f5b 100644 --- a/drivers/char/pty.c +++ b/drivers/char/pty.c @@ -52,7 +52,6 @@ static void pty_close(struct tty_struct *tty, struct file *filp) return; tty->link->packet = 0; set_bit(TTY_OTHER_CLOSED, &tty->link->flags); - tty_flip_buffer_push(tty->link); wake_up_interruptible(&tty->link->read_wait); wake_up_interruptible(&tty->link->write_wait); if (tty->driver->subtype == PTY_TYPE_MASTER) { @@ -208,7 +207,6 @@ static int pty_open(struct tty_struct *tty, struct file *filp) clear_bit(TTY_OTHER_CLOSED, &tty->link->flags); set_bit(TTY_THROTTLED, &tty->flags); retval = 0; - tty->low_latency = 1; out: return retval; } diff --git a/drivers/char/tty_buffer.c b/drivers/char/tty_buffer.c index 810ee25d66a4..3108991c5c8b 100644 --- a/drivers/char/tty_buffer.c +++ b/drivers/char/tty_buffer.c @@ -461,6 +461,19 @@ static void flush_to_ldisc(struct work_struct *work) tty_ldisc_deref(disc); } +/** + * tty_flush_to_ldisc + * @tty: tty to push + * + * Push the terminal flip buffers to the line discipline. + * + * Must not be called from IRQ context. + */ +void tty_flush_to_ldisc(struct tty_struct *tty) +{ + flush_to_ldisc(&tty->buf.work.work); +} + /** * tty_flip_buffer_push - terminal * @tty: tty to push diff --git a/include/linux/tty.h b/include/linux/tty.h index 1488d8c81aac..e8c6c9136c97 100644 --- a/include/linux/tty.h +++ b/include/linux/tty.h @@ -394,6 +394,7 @@ extern void __do_SAK(struct tty_struct *tty); extern void disassociate_ctty(int priv); extern void no_tty(void); extern void tty_flip_buffer_push(struct tty_struct *tty); +extern void tty_flush_to_ldisc(struct tty_struct *tty); extern void tty_buffer_free_all(struct tty_struct *tty); extern void tty_buffer_flush(struct tty_struct *tty); extern void tty_buffer_init(struct tty_struct *tty); -- cgit v1.2.3 From 0e82ffe3b90bcad72cfe80e4379946b8fb0691ca Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Mon, 27 Jul 2009 12:01:50 +0200 Subject: cfg80211: combine iwfreq implementations Until now we implemented iwfreq for managed mode, we needed to keep the implementations separate, but now that we have all versions implemented we can combine them and export just one handler. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- drivers/net/wireless/iwmc3200wifi/wext.c | 34 +-------------- include/net/cfg80211.h | 20 +++------ net/mac80211/wext.c | 73 +------------------------------- net/wireless/core.h | 3 ++ net/wireless/ibss.c | 5 +-- net/wireless/nl80211.c | 2 + net/wireless/wext-compat.c | 54 ++++++++++++++++++++++- net/wireless/wext-compat.h | 21 +++++++++ net/wireless/wext-sme.c | 5 +-- 9 files changed, 91 insertions(+), 126 deletions(-) create mode 100644 net/wireless/wext-compat.h (limited to 'include') diff --git a/drivers/net/wireless/iwmc3200wifi/wext.c b/drivers/net/wireless/iwmc3200wifi/wext.c index c3c90d5963bf..8058e9991c30 100644 --- a/drivers/net/wireless/iwmc3200wifi/wext.c +++ b/drivers/net/wireless/iwmc3200wifi/wext.c @@ -27,36 +27,6 @@ #include "iwm.h" #include "commands.h" -static int iwm_wext_siwfreq(struct net_device *dev, - struct iw_request_info *info, - struct iw_freq *freq, char *extra) -{ - struct iwm_priv *iwm = ndev_to_iwm(dev); - - switch (iwm->conf.mode) { - case UMAC_MODE_IBSS: - return cfg80211_ibss_wext_siwfreq(dev, info, freq, extra); - default: - return -EOPNOTSUPP; - } -} - -static int iwm_wext_giwfreq(struct net_device *dev, - struct iw_request_info *info, - struct iw_freq *freq, char *extra) -{ - struct iwm_priv *iwm = ndev_to_iwm(dev); - - switch (iwm->conf.mode) { - case UMAC_MODE_IBSS: - return cfg80211_ibss_wext_giwfreq(dev, info, freq, extra); - case UMAC_MODE_BSS: - return cfg80211_mgd_wext_giwfreq(dev, info, freq, extra); - default: - return -EOPNOTSUPP; - } -} - static int iwm_wext_siwap(struct net_device *dev, struct iw_request_info *info, struct sockaddr *ap_addr, char *extra) { @@ -125,8 +95,8 @@ static const iw_handler iwm_handlers[] = (iw_handler) cfg80211_wext_giwname, /* SIOCGIWNAME */ (iw_handler) NULL, /* SIOCSIWNWID */ (iw_handler) NULL, /* SIOCGIWNWID */ - (iw_handler) iwm_wext_siwfreq, /* SIOCSIWFREQ */ - (iw_handler) iwm_wext_giwfreq, /* SIOCGIWFREQ */ + (iw_handler) cfg80211_wext_siwfreq, /* SIOCSIWFREQ */ + (iw_handler) cfg80211_wext_giwfreq, /* SIOCGIWFREQ */ (iw_handler) cfg80211_wext_siwmode, /* SIOCSIWMODE */ (iw_handler) cfg80211_wext_giwmode, /* SIOCGIWMODE */ (iw_handler) NULL, /* SIOCSIWSENS */ diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index 0d278777e39c..5d249c4bf225 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -1595,12 +1595,6 @@ int cfg80211_wext_siwmlme(struct net_device *dev, int cfg80211_wext_giwrange(struct net_device *dev, struct iw_request_info *info, struct iw_point *data, char *extra); -int cfg80211_ibss_wext_siwfreq(struct net_device *dev, - struct iw_request_info *info, - struct iw_freq *freq, char *extra); -int cfg80211_ibss_wext_giwfreq(struct net_device *dev, - struct iw_request_info *info, - struct iw_freq *freq, char *extra); int cfg80211_ibss_wext_siwessid(struct net_device *dev, struct iw_request_info *info, struct iw_point *data, char *ssid); @@ -1614,12 +1608,6 @@ int cfg80211_ibss_wext_giwap(struct net_device *dev, struct iw_request_info *info, struct sockaddr *ap_addr, char *extra); -int cfg80211_mgd_wext_siwfreq(struct net_device *dev, - struct iw_request_info *info, - struct iw_freq *freq, char *extra); -int cfg80211_mgd_wext_giwfreq(struct net_device *dev, - struct iw_request_info *info, - struct iw_freq *freq, char *extra); int cfg80211_mgd_wext_siwessid(struct net_device *dev, struct iw_request_info *info, struct iw_point *data, char *ssid); @@ -1642,8 +1630,12 @@ int cfg80211_wext_giwauth(struct net_device *dev, struct iw_request_info *info, struct iw_param *data, char *extra); -struct ieee80211_channel *cfg80211_wext_freq(struct wiphy *wiphy, - struct iw_freq *freq); +int cfg80211_wext_siwfreq(struct net_device *dev, + struct iw_request_info *info, + struct iw_freq *freq, char *extra); +int cfg80211_wext_giwfreq(struct net_device *dev, + struct iw_request_info *info, + struct iw_freq *freq, char *extra); int cfg80211_wext_siwrate(struct net_device *dev, struct iw_request_info *info, diff --git a/net/mac80211/wext.c b/net/mac80211/wext.c index 5acb8140ee58..7cd9aa79ef52 100644 --- a/net/mac80211/wext.c +++ b/net/mac80211/wext.c @@ -27,75 +27,6 @@ #include "aes_ccm.h" -static int ieee80211_ioctl_siwfreq(struct net_device *dev, - struct iw_request_info *info, - struct iw_freq *freq, char *extra) -{ - struct ieee80211_sub_if_data *sdata = IEEE80211_DEV_TO_SUB_IF(dev); - struct ieee80211_local *local = sdata->local; - struct ieee80211_channel *chan; - - if (sdata->vif.type == NL80211_IFTYPE_ADHOC) - return cfg80211_ibss_wext_siwfreq(dev, info, freq, extra); - else if (sdata->vif.type == NL80211_IFTYPE_STATION) - return cfg80211_mgd_wext_siwfreq(dev, info, freq, extra); - - /* freq->e == 0: freq->m = channel; otherwise freq = m * 10^e */ - if (freq->e == 0) { - if (freq->m < 0) - return -EINVAL; - else - chan = ieee80211_get_channel(local->hw.wiphy, - ieee80211_channel_to_frequency(freq->m)); - } else { - int i, div = 1000000; - for (i = 0; i < freq->e; i++) - div /= 10; - if (div <= 0) - return -EINVAL; - chan = ieee80211_get_channel(local->hw.wiphy, freq->m / div); - } - - if (!chan) - return -EINVAL; - - if (chan->flags & IEEE80211_CHAN_DISABLED) - return -EINVAL; - - /* - * no change except maybe auto -> fixed, ignore the HT - * setting so you can fix a channel you're on already - */ - if (local->oper_channel == chan) - return 0; - - local->oper_channel = chan; - local->oper_channel_type = NL80211_CHAN_NO_HT; - ieee80211_hw_config(local, 0); - - return 0; -} - - -static int ieee80211_ioctl_giwfreq(struct net_device *dev, - struct iw_request_info *info, - struct iw_freq *freq, char *extra) -{ - struct ieee80211_local *local = wdev_priv(dev->ieee80211_ptr); - struct ieee80211_sub_if_data *sdata = IEEE80211_DEV_TO_SUB_IF(dev); - - if (sdata->vif.type == NL80211_IFTYPE_ADHOC) - return cfg80211_ibss_wext_giwfreq(dev, info, freq, extra); - else if (sdata->vif.type == NL80211_IFTYPE_STATION) - return cfg80211_mgd_wext_giwfreq(dev, info, freq, extra); - - freq->m = local->oper_channel->center_freq; - freq->e = 6; - - return 0; -} - - static int ieee80211_ioctl_siwessid(struct net_device *dev, struct iw_request_info *info, struct iw_point *data, char *ssid) @@ -173,8 +104,8 @@ static const iw_handler ieee80211_handler[] = (iw_handler) cfg80211_wext_giwname, /* SIOCGIWNAME */ (iw_handler) NULL, /* SIOCSIWNWID */ (iw_handler) NULL, /* SIOCGIWNWID */ - (iw_handler) ieee80211_ioctl_siwfreq, /* SIOCSIWFREQ */ - (iw_handler) ieee80211_ioctl_giwfreq, /* SIOCGIWFREQ */ + (iw_handler) cfg80211_wext_siwfreq, /* SIOCSIWFREQ */ + (iw_handler) cfg80211_wext_giwfreq, /* SIOCGIWFREQ */ (iw_handler) cfg80211_wext_siwmode, /* SIOCSIWMODE */ (iw_handler) cfg80211_wext_giwmode, /* SIOCGIWMODE */ (iw_handler) NULL, /* SIOCSIWSENS */ diff --git a/net/wireless/core.h b/net/wireless/core.h index 4276b70cd975..6d903c1d721d 100644 --- a/net/wireless/core.h +++ b/net/wireless/core.h @@ -66,6 +66,9 @@ struct cfg80211_registered_device { struct work_struct conn_work; struct work_struct event_work; + /* current channel */ + struct ieee80211_channel *channel; + #ifdef CONFIG_CFG80211_DEBUGFS /* Debugfs entries */ struct wiphy_debugfsdentries { diff --git a/net/wireless/ibss.c b/net/wireless/ibss.c index 8b65e212ae49..de9ac49cd907 100644 --- a/net/wireless/ibss.c +++ b/net/wireless/ibss.c @@ -7,6 +7,7 @@ #include #include #include +#include "wext-compat.h" #include "nl80211.h" @@ -312,8 +313,6 @@ int cfg80211_ibss_wext_siwfreq(struct net_device *dev, return err; } -/* temporary symbol - mark GPL - in the future the handler won't be */ -EXPORT_SYMBOL_GPL(cfg80211_ibss_wext_siwfreq); int cfg80211_ibss_wext_giwfreq(struct net_device *dev, struct iw_request_info *info, @@ -342,8 +341,6 @@ int cfg80211_ibss_wext_giwfreq(struct net_device *dev, /* no channel if not joining */ return -EINVAL; } -/* temporary symbol - mark GPL - in the future the handler won't be */ -EXPORT_SYMBOL_GPL(cfg80211_ibss_wext_giwfreq); int cfg80211_ibss_wext_siwessid(struct net_device *dev, struct iw_request_info *info, diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index c951eb2b07d5..0cd548267d4a 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -757,6 +757,8 @@ static int nl80211_set_wiphy(struct sk_buff *skb, struct genl_info *info) channel_type); if (result) goto bad_res; + + rdev->channel = chan; } changed = 0; diff --git a/net/wireless/wext-compat.c b/net/wireless/wext-compat.c index c7351a98e660..fc2e7768967d 100644 --- a/net/wireless/wext-compat.c +++ b/net/wireless/wext-compat.c @@ -14,6 +14,7 @@ #include #include #include +#include "wext-compat.h" #include "core.h" int cfg80211_wext_giwname(struct net_device *dev, @@ -300,7 +301,6 @@ struct ieee80211_channel *cfg80211_wext_freq(struct wiphy *wiphy, return ERR_PTR(-EINVAL); return chan; } -EXPORT_SYMBOL_GPL(cfg80211_wext_freq); int cfg80211_wext_siwrts(struct net_device *dev, struct iw_request_info *info, @@ -759,6 +759,58 @@ int cfg80211_wext_giwencode(struct net_device *dev, } EXPORT_SYMBOL_GPL(cfg80211_wext_giwencode); +int cfg80211_wext_siwfreq(struct net_device *dev, + struct iw_request_info *info, + struct iw_freq *freq, char *extra) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + struct cfg80211_registered_device *rdev = wiphy_to_dev(wdev->wiphy); + struct ieee80211_channel *chan; + int err; + + switch (wdev->iftype) { + case NL80211_IFTYPE_STATION: + return cfg80211_mgd_wext_siwfreq(dev, info, freq, extra); + case NL80211_IFTYPE_ADHOC: + return cfg80211_ibss_wext_siwfreq(dev, info, freq, extra); + default: + chan = cfg80211_wext_freq(wdev->wiphy, freq); + if (!chan) + return -EINVAL; + if (IS_ERR(chan)) + return PTR_ERR(chan); + err = rdev->ops->set_channel(wdev->wiphy, chan, + NL80211_CHAN_NO_HT); + if (err) + return err; + rdev->channel = chan; + return 0; + } +} +EXPORT_SYMBOL_GPL(cfg80211_wext_siwfreq); + +int cfg80211_wext_giwfreq(struct net_device *dev, + struct iw_request_info *info, + struct iw_freq *freq, char *extra) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + struct cfg80211_registered_device *rdev = wiphy_to_dev(wdev->wiphy); + + switch (wdev->iftype) { + case NL80211_IFTYPE_STATION: + return cfg80211_mgd_wext_giwfreq(dev, info, freq, extra); + case NL80211_IFTYPE_ADHOC: + return cfg80211_ibss_wext_giwfreq(dev, info, freq, extra); + default: + if (!rdev->channel) + return -EINVAL; + freq->m = rdev->channel->center_freq; + freq->e = 6; + return 0; + } +} +EXPORT_SYMBOL_GPL(cfg80211_wext_giwfreq); + int cfg80211_wext_siwtxpower(struct net_device *dev, struct iw_request_info *info, union iwreq_data *data, char *extra) diff --git a/net/wireless/wext-compat.h b/net/wireless/wext-compat.h new file mode 100644 index 000000000000..23a6b5a83f2d --- /dev/null +++ b/net/wireless/wext-compat.h @@ -0,0 +1,21 @@ +#ifndef __WEXT_COMPAT +#define __WEXT_COMPAT + +int cfg80211_ibss_wext_siwfreq(struct net_device *dev, + struct iw_request_info *info, + struct iw_freq *freq, char *extra); +int cfg80211_ibss_wext_giwfreq(struct net_device *dev, + struct iw_request_info *info, + struct iw_freq *freq, char *extra); + +int cfg80211_mgd_wext_siwfreq(struct net_device *dev, + struct iw_request_info *info, + struct iw_freq *freq, char *extra); +int cfg80211_mgd_wext_giwfreq(struct net_device *dev, + struct iw_request_info *info, + struct iw_freq *freq, char *extra); + +struct ieee80211_channel *cfg80211_wext_freq(struct wiphy *wiphy, + struct iw_freq *freq); + +#endif /* __WEXT_COMPAT */ diff --git a/net/wireless/wext-sme.c b/net/wireless/wext-sme.c index 4c689fd865b0..509279a1cfb2 100644 --- a/net/wireless/wext-sme.c +++ b/net/wireless/wext-sme.c @@ -8,6 +8,7 @@ #include #include #include +#include "wext-compat.h" #include "nl80211.h" int cfg80211_mgd_wext_connect(struct cfg80211_registered_device *rdev, @@ -108,8 +109,6 @@ int cfg80211_mgd_wext_siwfreq(struct net_device *dev, cfg80211_unlock_rdev(rdev); return err; } -/* temporary symbol - mark GPL - in the future the handler won't be */ -EXPORT_SYMBOL_GPL(cfg80211_mgd_wext_siwfreq); int cfg80211_mgd_wext_giwfreq(struct net_device *dev, struct iw_request_info *info, @@ -138,8 +137,6 @@ int cfg80211_mgd_wext_giwfreq(struct net_device *dev, /* no channel if not joining */ return -EINVAL; } -/* temporary symbol - mark GPL - in the future the handler won't be */ -EXPORT_SYMBOL_GPL(cfg80211_mgd_wext_giwfreq); int cfg80211_mgd_wext_siwessid(struct net_device *dev, struct iw_request_info *info, -- cgit v1.2.3 From 562e482265ac4d660d9f0114419591d62f44361d Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Mon, 27 Jul 2009 12:01:51 +0200 Subject: cfg80211: combine IWAP handlers Since we now have IWAP handlers for all modes, we can combine them into one. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- drivers/net/wireless/iwmc3200wifi/wext.c | 34 ++------------------- include/net/cfg80211.h | 24 ++++----------- net/mac80211/wext.c | 41 ++----------------------- net/wireless/ibss.c | 4 --- net/wireless/wext-compat.c | 52 +++++++++++++++++++++++++++----- net/wireless/wext-compat.h | 12 ++++++++ net/wireless/wext-sme.c | 4 --- 7 files changed, 66 insertions(+), 105 deletions(-) (limited to 'include') diff --git a/drivers/net/wireless/iwmc3200wifi/wext.c b/drivers/net/wireless/iwmc3200wifi/wext.c index 8058e9991c30..5319b16474e3 100644 --- a/drivers/net/wireless/iwmc3200wifi/wext.c +++ b/drivers/net/wireless/iwmc3200wifi/wext.c @@ -27,36 +27,6 @@ #include "iwm.h" #include "commands.h" -static int iwm_wext_siwap(struct net_device *dev, struct iw_request_info *info, - struct sockaddr *ap_addr, char *extra) -{ - struct iwm_priv *iwm = ndev_to_iwm(dev); - - switch (iwm->conf.mode) { - case UMAC_MODE_IBSS: - return cfg80211_ibss_wext_siwap(dev, info, ap_addr, extra); - case UMAC_MODE_BSS: - return cfg80211_mgd_wext_siwap(dev, info, ap_addr, extra); - default: - return -EOPNOTSUPP; - } -} - -static int iwm_wext_giwap(struct net_device *dev, struct iw_request_info *info, - struct sockaddr *ap_addr, char *extra) -{ - struct iwm_priv *iwm = ndev_to_iwm(dev); - - switch (iwm->conf.mode) { - case UMAC_MODE_IBSS: - return cfg80211_ibss_wext_giwap(dev, info, ap_addr, extra); - case UMAC_MODE_BSS: - return cfg80211_mgd_wext_giwap(dev, info, ap_addr, extra); - default: - return -EOPNOTSUPP; - } -} - static int iwm_wext_siwessid(struct net_device *dev, struct iw_request_info *info, struct iw_point *data, char *ssid) @@ -111,8 +81,8 @@ static const iw_handler iwm_handlers[] = (iw_handler) NULL, /* SIOCGIWSPY */ (iw_handler) NULL, /* SIOCSIWTHRSPY */ (iw_handler) NULL, /* SIOCGIWTHRSPY */ - (iw_handler) iwm_wext_siwap, /* SIOCSIWAP */ - (iw_handler) iwm_wext_giwap, /* SIOCGIWAP */ + (iw_handler) cfg80211_wext_siwap, /* SIOCSIWAP */ + (iw_handler) cfg80211_wext_giwap, /* SIOCGIWAP */ (iw_handler) NULL, /* SIOCSIWMLME */ (iw_handler) NULL, /* SIOCGIWAPLIST */ (iw_handler) cfg80211_wext_siwscan, /* SIOCSIWSCAN */ diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index 5d249c4bf225..3348c16e1f35 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -1601,12 +1601,6 @@ int cfg80211_ibss_wext_siwessid(struct net_device *dev, int cfg80211_ibss_wext_giwessid(struct net_device *dev, struct iw_request_info *info, struct iw_point *data, char *ssid); -int cfg80211_ibss_wext_siwap(struct net_device *dev, - struct iw_request_info *info, - struct sockaddr *ap_addr, char *extra); -int cfg80211_ibss_wext_giwap(struct net_device *dev, - struct iw_request_info *info, - struct sockaddr *ap_addr, char *extra); int cfg80211_mgd_wext_siwessid(struct net_device *dev, struct iw_request_info *info, @@ -1614,12 +1608,6 @@ int cfg80211_mgd_wext_siwessid(struct net_device *dev, int cfg80211_mgd_wext_giwessid(struct net_device *dev, struct iw_request_info *info, struct iw_point *data, char *ssid); -int cfg80211_mgd_wext_siwap(struct net_device *dev, - struct iw_request_info *info, - struct sockaddr *ap_addr, char *extra); -int cfg80211_mgd_wext_giwap(struct net_device *dev, - struct iw_request_info *info, - struct sockaddr *ap_addr, char *extra); int cfg80211_wext_siwgenie(struct net_device *dev, struct iw_request_info *info, struct iw_point *data, char *extra); @@ -1686,12 +1674,12 @@ int cfg80211_wext_giwpower(struct net_device *dev, struct iw_request_info *info, struct iw_param *wrq, char *extra); -int cfg80211_wds_wext_siwap(struct net_device *dev, - struct iw_request_info *info, - struct sockaddr *addr, char *extra); -int cfg80211_wds_wext_giwap(struct net_device *dev, - struct iw_request_info *info, - struct sockaddr *addr, char *extra); +int cfg80211_wext_siwap(struct net_device *dev, + struct iw_request_info *info, + struct sockaddr *ap_addr, char *extra); +int cfg80211_wext_giwap(struct net_device *dev, + struct iw_request_info *info, + struct sockaddr *ap_addr, char *extra); /* * callbacks for asynchronous cfg80211 methods, notification diff --git a/net/mac80211/wext.c b/net/mac80211/wext.c index 7cd9aa79ef52..72866c8b8c3d 100644 --- a/net/mac80211/wext.c +++ b/net/mac80211/wext.c @@ -59,43 +59,6 @@ static int ieee80211_ioctl_giwessid(struct net_device *dev, } -static int ieee80211_ioctl_siwap(struct net_device *dev, - struct iw_request_info *info, - struct sockaddr *ap_addr, char *extra) -{ - struct ieee80211_sub_if_data *sdata = IEEE80211_DEV_TO_SUB_IF(dev); - - if (sdata->vif.type == NL80211_IFTYPE_ADHOC) - return cfg80211_ibss_wext_siwap(dev, info, ap_addr, extra); - - if (sdata->vif.type == NL80211_IFTYPE_STATION) - return cfg80211_mgd_wext_siwap(dev, info, ap_addr, extra); - - if (sdata->vif.type == NL80211_IFTYPE_WDS) - return cfg80211_wds_wext_siwap(dev, info, ap_addr, extra); - return -EOPNOTSUPP; -} - - -static int ieee80211_ioctl_giwap(struct net_device *dev, - struct iw_request_info *info, - struct sockaddr *ap_addr, char *extra) -{ - struct ieee80211_sub_if_data *sdata = IEEE80211_DEV_TO_SUB_IF(dev); - - if (sdata->vif.type == NL80211_IFTYPE_ADHOC) - return cfg80211_ibss_wext_giwap(dev, info, ap_addr, extra); - - if (sdata->vif.type == NL80211_IFTYPE_STATION) - return cfg80211_mgd_wext_giwap(dev, info, ap_addr, extra); - - if (sdata->vif.type == NL80211_IFTYPE_WDS) - return cfg80211_wds_wext_giwap(dev, info, ap_addr, extra); - - return -EOPNOTSUPP; -} - - /* Structures to export the Wireless Handlers */ static const iw_handler ieee80211_handler[] = @@ -120,8 +83,8 @@ static const iw_handler ieee80211_handler[] = (iw_handler) NULL, /* SIOCGIWSPY */ (iw_handler) NULL, /* SIOCSIWTHRSPY */ (iw_handler) NULL, /* SIOCGIWTHRSPY */ - (iw_handler) ieee80211_ioctl_siwap, /* SIOCSIWAP */ - (iw_handler) ieee80211_ioctl_giwap, /* SIOCGIWAP */ + (iw_handler) cfg80211_wext_siwap, /* SIOCSIWAP */ + (iw_handler) cfg80211_wext_giwap, /* SIOCGIWAP */ (iw_handler) cfg80211_wext_siwmlme, /* SIOCSIWMLME */ (iw_handler) NULL, /* SIOCGIWAPLIST */ (iw_handler) cfg80211_wext_siwscan, /* SIOCSIWSCAN */ diff --git a/net/wireless/ibss.c b/net/wireless/ibss.c index de9ac49cd907..f955225ed911 100644 --- a/net/wireless/ibss.c +++ b/net/wireless/ibss.c @@ -466,8 +466,6 @@ int cfg80211_ibss_wext_siwap(struct net_device *dev, return err; } -/* temporary symbol - mark GPL - in the future the handler won't be */ -EXPORT_SYMBOL_GPL(cfg80211_ibss_wext_siwap); int cfg80211_ibss_wext_giwap(struct net_device *dev, struct iw_request_info *info, @@ -493,6 +491,4 @@ int cfg80211_ibss_wext_giwap(struct net_device *dev, return 0; } -/* temporary symbol - mark GPL - in the future the handler won't be */ -EXPORT_SYMBOL_GPL(cfg80211_ibss_wext_giwap); #endif diff --git a/net/wireless/wext-compat.c b/net/wireless/wext-compat.c index fc2e7768967d..c27774bd0107 100644 --- a/net/wireless/wext-compat.c +++ b/net/wireless/wext-compat.c @@ -1149,9 +1149,9 @@ int cfg80211_wext_giwpower(struct net_device *dev, } EXPORT_SYMBOL_GPL(cfg80211_wext_giwpower); -int cfg80211_wds_wext_siwap(struct net_device *dev, - struct iw_request_info *info, - struct sockaddr *addr, char *extra) +static int cfg80211_wds_wext_siwap(struct net_device *dev, + struct iw_request_info *info, + struct sockaddr *addr, char *extra) { struct wireless_dev *wdev = dev->ieee80211_ptr; struct cfg80211_registered_device *rdev = wiphy_to_dev(wdev->wiphy); @@ -1177,11 +1177,10 @@ int cfg80211_wds_wext_siwap(struct net_device *dev, return 0; } -EXPORT_SYMBOL_GPL(cfg80211_wds_wext_siwap); -int cfg80211_wds_wext_giwap(struct net_device *dev, - struct iw_request_info *info, - struct sockaddr *addr, char *extra) +static int cfg80211_wds_wext_giwap(struct net_device *dev, + struct iw_request_info *info, + struct sockaddr *addr, char *extra) { struct wireless_dev *wdev = dev->ieee80211_ptr; @@ -1193,7 +1192,6 @@ int cfg80211_wds_wext_giwap(struct net_device *dev, return 0; } -EXPORT_SYMBOL_GPL(cfg80211_wds_wext_giwap); int cfg80211_wext_siwrate(struct net_device *dev, struct iw_request_info *info, @@ -1327,3 +1325,41 @@ struct iw_statistics *cfg80211_wireless_stats(struct net_device *dev) return &wstats; } EXPORT_SYMBOL_GPL(cfg80211_wireless_stats); + +int cfg80211_wext_siwap(struct net_device *dev, + struct iw_request_info *info, + struct sockaddr *ap_addr, char *extra) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + + switch (wdev->iftype) { + case NL80211_IFTYPE_ADHOC: + return cfg80211_ibss_wext_siwap(dev, info, ap_addr, extra); + case NL80211_IFTYPE_STATION: + return cfg80211_mgd_wext_siwap(dev, info, ap_addr, extra); + case NL80211_IFTYPE_WDS: + return cfg80211_wds_wext_siwap(dev, info, ap_addr, extra); + default: + return -EOPNOTSUPP; + } +} +EXPORT_SYMBOL_GPL(cfg80211_wext_siwap); + +int cfg80211_wext_giwap(struct net_device *dev, + struct iw_request_info *info, + struct sockaddr *ap_addr, char *extra) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + + switch (wdev->iftype) { + case NL80211_IFTYPE_ADHOC: + return cfg80211_ibss_wext_giwap(dev, info, ap_addr, extra); + case NL80211_IFTYPE_STATION: + return cfg80211_mgd_wext_giwap(dev, info, ap_addr, extra); + case NL80211_IFTYPE_WDS: + return cfg80211_wds_wext_giwap(dev, info, ap_addr, extra); + default: + return -EOPNOTSUPP; + } +} +EXPORT_SYMBOL_GPL(cfg80211_wext_giwap); diff --git a/net/wireless/wext-compat.h b/net/wireless/wext-compat.h index 23a6b5a83f2d..51028ebf19ae 100644 --- a/net/wireless/wext-compat.h +++ b/net/wireless/wext-compat.h @@ -7,6 +7,12 @@ int cfg80211_ibss_wext_siwfreq(struct net_device *dev, int cfg80211_ibss_wext_giwfreq(struct net_device *dev, struct iw_request_info *info, struct iw_freq *freq, char *extra); +int cfg80211_ibss_wext_siwap(struct net_device *dev, + struct iw_request_info *info, + struct sockaddr *ap_addr, char *extra); +int cfg80211_ibss_wext_giwap(struct net_device *dev, + struct iw_request_info *info, + struct sockaddr *ap_addr, char *extra); int cfg80211_mgd_wext_siwfreq(struct net_device *dev, struct iw_request_info *info, @@ -14,6 +20,12 @@ int cfg80211_mgd_wext_siwfreq(struct net_device *dev, int cfg80211_mgd_wext_giwfreq(struct net_device *dev, struct iw_request_info *info, struct iw_freq *freq, char *extra); +int cfg80211_mgd_wext_siwap(struct net_device *dev, + struct iw_request_info *info, + struct sockaddr *ap_addr, char *extra); +int cfg80211_mgd_wext_giwap(struct net_device *dev, + struct iw_request_info *info, + struct sockaddr *ap_addr, char *extra); struct ieee80211_channel *cfg80211_wext_freq(struct wiphy *wiphy, struct iw_freq *freq); diff --git a/net/wireless/wext-sme.c b/net/wireless/wext-sme.c index 509279a1cfb2..1aa31cc55113 100644 --- a/net/wireless/wext-sme.c +++ b/net/wireless/wext-sme.c @@ -273,8 +273,6 @@ int cfg80211_mgd_wext_siwap(struct net_device *dev, cfg80211_unlock_rdev(wiphy_to_dev(wdev->wiphy)); return err; } -/* temporary symbol - mark GPL - in the future the handler won't be */ -EXPORT_SYMBOL_GPL(cfg80211_mgd_wext_siwap); int cfg80211_mgd_wext_giwap(struct net_device *dev, struct iw_request_info *info, @@ -299,8 +297,6 @@ int cfg80211_mgd_wext_giwap(struct net_device *dev, return 0; } -/* temporary symbol - mark GPL - in the future the handler won't be */ -EXPORT_SYMBOL_GPL(cfg80211_mgd_wext_giwap); int cfg80211_wext_siwgenie(struct net_device *dev, struct iw_request_info *info, -- cgit v1.2.3 From 1f9298f96082692bdfe73af6fc2167f627f21647 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Mon, 27 Jul 2009 12:01:52 +0200 Subject: cfg80211: combine IWESSID handlers Since we now have handlers IWESSID for all modes, we can combine them into one. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- drivers/net/wireless/iwmc3200wifi/wext.c | 36 ++------------------------------ include/net/cfg80211.h | 20 ++++++------------ net/mac80211/wext.c | 35 ++----------------------------- net/wireless/ibss.c | 4 ---- net/wireless/wext-compat.c | 34 ++++++++++++++++++++++++++++++ net/wireless/wext-compat.h | 12 +++++++++++ net/wireless/wext-sme.c | 4 ---- 7 files changed, 56 insertions(+), 89 deletions(-) (limited to 'include') diff --git a/drivers/net/wireless/iwmc3200wifi/wext.c b/drivers/net/wireless/iwmc3200wifi/wext.c index 5319b16474e3..9196024a2890 100644 --- a/drivers/net/wireless/iwmc3200wifi/wext.c +++ b/drivers/net/wireless/iwmc3200wifi/wext.c @@ -27,38 +27,6 @@ #include "iwm.h" #include "commands.h" -static int iwm_wext_siwessid(struct net_device *dev, - struct iw_request_info *info, - struct iw_point *data, char *ssid) -{ - struct iwm_priv *iwm = ndev_to_iwm(dev); - - switch (iwm->conf.mode) { - case UMAC_MODE_IBSS: - return cfg80211_ibss_wext_siwessid(dev, info, data, ssid); - case UMAC_MODE_BSS: - return cfg80211_mgd_wext_siwessid(dev, info, data, ssid); - default: - return -EOPNOTSUPP; - } -} - -static int iwm_wext_giwessid(struct net_device *dev, - struct iw_request_info *info, - struct iw_point *data, char *ssid) -{ - struct iwm_priv *iwm = ndev_to_iwm(dev); - - switch (iwm->conf.mode) { - case UMAC_MODE_IBSS: - return cfg80211_ibss_wext_giwessid(dev, info, data, ssid); - case UMAC_MODE_BSS: - return cfg80211_mgd_wext_giwessid(dev, info, data, ssid); - default: - return -EOPNOTSUPP; - } -} - static const iw_handler iwm_handlers[] = { (iw_handler) NULL, /* SIOCSIWCOMMIT */ @@ -87,8 +55,8 @@ static const iw_handler iwm_handlers[] = (iw_handler) NULL, /* SIOCGIWAPLIST */ (iw_handler) cfg80211_wext_siwscan, /* SIOCSIWSCAN */ (iw_handler) cfg80211_wext_giwscan, /* SIOCGIWSCAN */ - (iw_handler) iwm_wext_siwessid, /* SIOCSIWESSID */ - (iw_handler) iwm_wext_giwessid, /* SIOCGIWESSID */ + (iw_handler) cfg80211_wext_siwessid, /* SIOCSIWESSID */ + (iw_handler) cfg80211_wext_giwessid, /* SIOCGIWESSID */ (iw_handler) NULL, /* SIOCSIWNICKN */ (iw_handler) NULL, /* SIOCGIWNICKN */ (iw_handler) NULL, /* -- hole -- */ diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index 3348c16e1f35..e1b92358242b 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -1595,19 +1595,6 @@ int cfg80211_wext_siwmlme(struct net_device *dev, int cfg80211_wext_giwrange(struct net_device *dev, struct iw_request_info *info, struct iw_point *data, char *extra); -int cfg80211_ibss_wext_siwessid(struct net_device *dev, - struct iw_request_info *info, - struct iw_point *data, char *ssid); -int cfg80211_ibss_wext_giwessid(struct net_device *dev, - struct iw_request_info *info, - struct iw_point *data, char *ssid); - -int cfg80211_mgd_wext_siwessid(struct net_device *dev, - struct iw_request_info *info, - struct iw_point *data, char *ssid); -int cfg80211_mgd_wext_giwessid(struct net_device *dev, - struct iw_request_info *info, - struct iw_point *data, char *ssid); int cfg80211_wext_siwgenie(struct net_device *dev, struct iw_request_info *info, struct iw_point *data, char *extra); @@ -1624,7 +1611,12 @@ int cfg80211_wext_siwfreq(struct net_device *dev, int cfg80211_wext_giwfreq(struct net_device *dev, struct iw_request_info *info, struct iw_freq *freq, char *extra); - +int cfg80211_wext_siwessid(struct net_device *dev, + struct iw_request_info *info, + struct iw_point *data, char *ssid); +int cfg80211_wext_giwessid(struct net_device *dev, + struct iw_request_info *info, + struct iw_point *data, char *ssid); int cfg80211_wext_siwrate(struct net_device *dev, struct iw_request_info *info, struct iw_param *rate, char *extra); diff --git a/net/mac80211/wext.c b/net/mac80211/wext.c index 72866c8b8c3d..aa250c3e8fda 100644 --- a/net/mac80211/wext.c +++ b/net/mac80211/wext.c @@ -27,37 +27,6 @@ #include "aes_ccm.h" -static int ieee80211_ioctl_siwessid(struct net_device *dev, - struct iw_request_info *info, - struct iw_point *data, char *ssid) -{ - struct ieee80211_sub_if_data *sdata = IEEE80211_DEV_TO_SUB_IF(dev); - - if (sdata->vif.type == NL80211_IFTYPE_ADHOC) - return cfg80211_ibss_wext_siwessid(dev, info, data, ssid); - else if (sdata->vif.type == NL80211_IFTYPE_STATION) - return cfg80211_mgd_wext_siwessid(dev, info, data, ssid); - - return -EOPNOTSUPP; -} - - -static int ieee80211_ioctl_giwessid(struct net_device *dev, - struct iw_request_info *info, - struct iw_point *data, char *ssid) -{ - struct ieee80211_sub_if_data *sdata; - - sdata = IEEE80211_DEV_TO_SUB_IF(dev); - - if (sdata->vif.type == NL80211_IFTYPE_ADHOC) - return cfg80211_ibss_wext_giwessid(dev, info, data, ssid); - else if (sdata->vif.type == NL80211_IFTYPE_STATION) - return cfg80211_mgd_wext_giwessid(dev, info, data, ssid); - - return -EOPNOTSUPP; -} - /* Structures to export the Wireless Handlers */ @@ -89,8 +58,8 @@ static const iw_handler ieee80211_handler[] = (iw_handler) NULL, /* SIOCGIWAPLIST */ (iw_handler) cfg80211_wext_siwscan, /* SIOCSIWSCAN */ (iw_handler) cfg80211_wext_giwscan, /* SIOCGIWSCAN */ - (iw_handler) ieee80211_ioctl_siwessid, /* SIOCSIWESSID */ - (iw_handler) ieee80211_ioctl_giwessid, /* SIOCGIWESSID */ + (iw_handler) cfg80211_wext_siwessid, /* SIOCSIWESSID */ + (iw_handler) cfg80211_wext_giwessid, /* SIOCGIWESSID */ (iw_handler) NULL, /* SIOCSIWNICKN */ (iw_handler) NULL, /* SIOCGIWNICKN */ (iw_handler) NULL, /* -- hole -- */ diff --git a/net/wireless/ibss.c b/net/wireless/ibss.c index f955225ed911..4d7a084b35e2 100644 --- a/net/wireless/ibss.c +++ b/net/wireless/ibss.c @@ -381,8 +381,6 @@ int cfg80211_ibss_wext_siwessid(struct net_device *dev, return err; } -/* temporary symbol - mark GPL - in the future the handler won't be */ -EXPORT_SYMBOL_GPL(cfg80211_ibss_wext_siwessid); int cfg80211_ibss_wext_giwessid(struct net_device *dev, struct iw_request_info *info, @@ -410,8 +408,6 @@ int cfg80211_ibss_wext_giwessid(struct net_device *dev, return 0; } -/* temporary symbol - mark GPL - in the future the handler won't be */ -EXPORT_SYMBOL_GPL(cfg80211_ibss_wext_giwessid); int cfg80211_ibss_wext_siwap(struct net_device *dev, struct iw_request_info *info, diff --git a/net/wireless/wext-compat.c b/net/wireless/wext-compat.c index c27774bd0107..083e4c33d95d 100644 --- a/net/wireless/wext-compat.c +++ b/net/wireless/wext-compat.c @@ -1363,3 +1363,37 @@ int cfg80211_wext_giwap(struct net_device *dev, } } EXPORT_SYMBOL_GPL(cfg80211_wext_giwap); + +int cfg80211_wext_siwessid(struct net_device *dev, + struct iw_request_info *info, + struct iw_point *data, char *ssid) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + + switch (wdev->iftype) { + case NL80211_IFTYPE_ADHOC: + return cfg80211_ibss_wext_siwessid(dev, info, data, ssid); + case NL80211_IFTYPE_STATION: + return cfg80211_mgd_wext_siwessid(dev, info, data, ssid); + default: + return -EOPNOTSUPP; + } +} +EXPORT_SYMBOL_GPL(cfg80211_wext_siwessid); + +int cfg80211_wext_giwessid(struct net_device *dev, + struct iw_request_info *info, + struct iw_point *data, char *ssid) +{ + struct wireless_dev *wdev = dev->ieee80211_ptr; + + switch (wdev->iftype) { + case NL80211_IFTYPE_ADHOC: + return cfg80211_ibss_wext_giwessid(dev, info, data, ssid); + case NL80211_IFTYPE_STATION: + return cfg80211_mgd_wext_giwessid(dev, info, data, ssid); + default: + return -EOPNOTSUPP; + } +} +EXPORT_SYMBOL_GPL(cfg80211_wext_giwessid); diff --git a/net/wireless/wext-compat.h b/net/wireless/wext-compat.h index 51028ebf19ae..c0310d93c2e5 100644 --- a/net/wireless/wext-compat.h +++ b/net/wireless/wext-compat.h @@ -13,6 +13,12 @@ int cfg80211_ibss_wext_siwap(struct net_device *dev, int cfg80211_ibss_wext_giwap(struct net_device *dev, struct iw_request_info *info, struct sockaddr *ap_addr, char *extra); +int cfg80211_ibss_wext_siwessid(struct net_device *dev, + struct iw_request_info *info, + struct iw_point *data, char *ssid); +int cfg80211_ibss_wext_giwessid(struct net_device *dev, + struct iw_request_info *info, + struct iw_point *data, char *ssid); int cfg80211_mgd_wext_siwfreq(struct net_device *dev, struct iw_request_info *info, @@ -26,6 +32,12 @@ int cfg80211_mgd_wext_siwap(struct net_device *dev, int cfg80211_mgd_wext_giwap(struct net_device *dev, struct iw_request_info *info, struct sockaddr *ap_addr, char *extra); +int cfg80211_mgd_wext_siwessid(struct net_device *dev, + struct iw_request_info *info, + struct iw_point *data, char *ssid); +int cfg80211_mgd_wext_giwessid(struct net_device *dev, + struct iw_request_info *info, + struct iw_point *data, char *ssid); struct ieee80211_channel *cfg80211_wext_freq(struct wiphy *wiphy, struct iw_freq *freq); diff --git a/net/wireless/wext-sme.c b/net/wireless/wext-sme.c index 1aa31cc55113..7bacbd1c2af6 100644 --- a/net/wireless/wext-sme.c +++ b/net/wireless/wext-sme.c @@ -192,8 +192,6 @@ int cfg80211_mgd_wext_siwessid(struct net_device *dev, cfg80211_unlock_rdev(wiphy_to_dev(wdev->wiphy)); return err; } -/* temporary symbol - mark GPL - in the future the handler won't be */ -EXPORT_SYMBOL_GPL(cfg80211_mgd_wext_siwessid); int cfg80211_mgd_wext_giwessid(struct net_device *dev, struct iw_request_info *info, @@ -218,8 +216,6 @@ int cfg80211_mgd_wext_giwessid(struct net_device *dev, return 0; } -/* temporary symbol - mark GPL - in the future the handler won't be */ -EXPORT_SYMBOL_GPL(cfg80211_mgd_wext_giwessid); int cfg80211_mgd_wext_siwap(struct net_device *dev, struct iw_request_info *info, -- cgit v1.2.3 From 096b7fe012d66ed55e98bc8022405ede0cc80e96 Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Wed, 29 Jul 2009 15:04:04 -0700 Subject: cgroups: fix pid namespace bug The bug was introduced by commit cc31edceee04a7b87f2be48f9489ebb72d264844 ("cgroups: convert tasks file to use a seq_file with shared pid array"). We cache a pid array for all threads that are opening the same "tasks" file, but the pids in the array are always from the namespace of the last process that opened the file, so all other threads will read pids from that namespace instead of their own namespaces. To fix it, we maintain a list of pid arrays, which is keyed by pid_ns. The list will be of length 1 at most time. Reported-by: Paul Menage Idea-by: Paul Menage Signed-off-by: Li Zefan Reviewed-by: Serge Hallyn Cc: Balbir Singh Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/cgroup.h | 11 +++--- kernel/cgroup.c | 96 +++++++++++++++++++++++++++++++++++++------------- 2 files changed, 76 insertions(+), 31 deletions(-) (limited to 'include') diff --git a/include/linux/cgroup.h b/include/linux/cgroup.h index 665fa70e4094..20411d2876f8 100644 --- a/include/linux/cgroup.h +++ b/include/linux/cgroup.h @@ -179,14 +179,11 @@ struct cgroup { */ struct list_head release_list; - /* pids_mutex protects the fields below */ + /* pids_mutex protects pids_list and cached pid arrays. */ struct rw_semaphore pids_mutex; - /* Array of process ids in the cgroup */ - pid_t *tasks_pids; - /* How many files are using the current tasks_pids array */ - int pids_use_count; - /* Length of the current tasks_pids array */ - int pids_length; + + /* Linked list of struct cgroup_pids */ + struct list_head pids_list; /* For RCU-protected deletion */ struct rcu_head rcu_head; diff --git a/kernel/cgroup.c b/kernel/cgroup.c index 3737a682cdf5..250dac05680f 100644 --- a/kernel/cgroup.c +++ b/kernel/cgroup.c @@ -47,6 +47,7 @@ #include #include #include +#include #include @@ -960,6 +961,7 @@ static void init_cgroup_housekeeping(struct cgroup *cgrp) INIT_LIST_HEAD(&cgrp->children); INIT_LIST_HEAD(&cgrp->css_sets); INIT_LIST_HEAD(&cgrp->release_list); + INIT_LIST_HEAD(&cgrp->pids_list); init_rwsem(&cgrp->pids_mutex); } static void init_cgroup_root(struct cgroupfs_root *root) @@ -2201,12 +2203,30 @@ err: return ret; } +/* + * Cache pids for all threads in the same pid namespace that are + * opening the same "tasks" file. + */ +struct cgroup_pids { + /* The node in cgrp->pids_list */ + struct list_head list; + /* The cgroup those pids belong to */ + struct cgroup *cgrp; + /* The namepsace those pids belong to */ + struct pid_namespace *ns; + /* Array of process ids in the cgroup */ + pid_t *tasks_pids; + /* How many files are using the this tasks_pids array */ + int use_count; + /* Length of the current tasks_pids array */ + int length; +}; + static int cmppid(const void *a, const void *b) { return *(pid_t *)a - *(pid_t *)b; } - /* * seq_file methods for the "tasks" file. The seq_file position is the * next pid to display; the seq_file iterator is a pointer to the pid @@ -2221,45 +2241,47 @@ static void *cgroup_tasks_start(struct seq_file *s, loff_t *pos) * after a seek to the start). Use a binary-search to find the * next pid to display, if any */ - struct cgroup *cgrp = s->private; + struct cgroup_pids *cp = s->private; + struct cgroup *cgrp = cp->cgrp; int index = 0, pid = *pos; int *iter; down_read(&cgrp->pids_mutex); if (pid) { - int end = cgrp->pids_length; + int end = cp->length; while (index < end) { int mid = (index + end) / 2; - if (cgrp->tasks_pids[mid] == pid) { + if (cp->tasks_pids[mid] == pid) { index = mid; break; - } else if (cgrp->tasks_pids[mid] <= pid) + } else if (cp->tasks_pids[mid] <= pid) index = mid + 1; else end = mid; } } /* If we're off the end of the array, we're done */ - if (index >= cgrp->pids_length) + if (index >= cp->length) return NULL; /* Update the abstract position to be the actual pid that we found */ - iter = cgrp->tasks_pids + index; + iter = cp->tasks_pids + index; *pos = *iter; return iter; } static void cgroup_tasks_stop(struct seq_file *s, void *v) { - struct cgroup *cgrp = s->private; + struct cgroup_pids *cp = s->private; + struct cgroup *cgrp = cp->cgrp; up_read(&cgrp->pids_mutex); } static void *cgroup_tasks_next(struct seq_file *s, void *v, loff_t *pos) { - struct cgroup *cgrp = s->private; + struct cgroup_pids *cp = s->private; int *p = v; - int *end = cgrp->tasks_pids + cgrp->pids_length; + int *end = cp->tasks_pids + cp->length; /* * Advance to the next pid in the array. If this goes off the @@ -2286,26 +2308,33 @@ static struct seq_operations cgroup_tasks_seq_operations = { .show = cgroup_tasks_show, }; -static void release_cgroup_pid_array(struct cgroup *cgrp) +static void release_cgroup_pid_array(struct cgroup_pids *cp) { + struct cgroup *cgrp = cp->cgrp; + down_write(&cgrp->pids_mutex); - BUG_ON(!cgrp->pids_use_count); - if (!--cgrp->pids_use_count) { - kfree(cgrp->tasks_pids); - cgrp->tasks_pids = NULL; - cgrp->pids_length = 0; + BUG_ON(!cp->use_count); + if (!--cp->use_count) { + list_del(&cp->list); + put_pid_ns(cp->ns); + kfree(cp->tasks_pids); + kfree(cp); } up_write(&cgrp->pids_mutex); } static int cgroup_tasks_release(struct inode *inode, struct file *file) { - struct cgroup *cgrp = __d_cgrp(file->f_dentry->d_parent); + struct seq_file *seq; + struct cgroup_pids *cp; if (!(file->f_mode & FMODE_READ)) return 0; - release_cgroup_pid_array(cgrp); + seq = file->private_data; + cp = seq->private; + + release_cgroup_pid_array(cp); return seq_release(inode, file); } @@ -2324,6 +2353,8 @@ static struct file_operations cgroup_tasks_operations = { static int cgroup_tasks_open(struct inode *unused, struct file *file) { struct cgroup *cgrp = __d_cgrp(file->f_dentry->d_parent); + struct pid_namespace *ns = current->nsproxy->pid_ns; + struct cgroup_pids *cp; pid_t *pidarray; int npids; int retval; @@ -2350,20 +2381,37 @@ static int cgroup_tasks_open(struct inode *unused, struct file *file) * array if necessary */ down_write(&cgrp->pids_mutex); - kfree(cgrp->tasks_pids); - cgrp->tasks_pids = pidarray; - cgrp->pids_length = npids; - cgrp->pids_use_count++; + + list_for_each_entry(cp, &cgrp->pids_list, list) { + if (ns == cp->ns) + goto found; + } + + cp = kzalloc(sizeof(*cp), GFP_KERNEL); + if (!cp) { + up_write(&cgrp->pids_mutex); + kfree(pidarray); + return -ENOMEM; + } + cp->cgrp = cgrp; + cp->ns = ns; + get_pid_ns(ns); + list_add(&cp->list, &cgrp->pids_list); +found: + kfree(cp->tasks_pids); + cp->tasks_pids = pidarray; + cp->length = npids; + cp->use_count++; up_write(&cgrp->pids_mutex); file->f_op = &cgroup_tasks_operations; retval = seq_open(file, &cgroup_tasks_seq_operations); if (retval) { - release_cgroup_pid_array(cgrp); + release_cgroup_pid_array(cp); return retval; } - ((struct seq_file *)file->private_data)->private = cgrp; + ((struct seq_file *)file->private_data)->private = cp; return 0; } -- cgit v1.2.3 From 887032670d47366a8c8f25396ea7c14b7b2cc620 Mon Sep 17 00:00:00 2001 From: KAMEZAWA Hiroyuki Date: Wed, 29 Jul 2009 15:04:06 -0700 Subject: cgroup avoid permanent sleep at rmdir After commit ec64f51545fffbc4cb968f0cea56341a4b07e85a ("cgroup: fix frequent -EBUSY at rmdir"), cgroup's rmdir (especially against memcg) doesn't return -EBUSY by temporary ref counts. That commit expects all refs after pre_destroy() is temporary but...it wasn't. Then, rmdir can wait permanently. This patch tries to fix that and change followings. - set CGRP_WAIT_ON_RMDIR flag before pre_destroy(). - clear CGRP_WAIT_ON_RMDIR flag when the subsys finds racy case. if there are sleeping ones, wakes them up. - rmdir() sleeps only when CGRP_WAIT_ON_RMDIR flag is set. Tested-by: Daisuke Nishimura Reported-by: Daisuke Nishimura Reviewed-by: Paul Menage Acked-by: Balbir Sigh Signed-off-by: KAMEZAWA Hiroyuki Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/cgroup.h | 17 ++++++++++++++++ kernel/cgroup.c | 55 ++++++++++++++++++++++++++++++++++---------------- mm/memcontrol.c | 23 ++++++++++++++++++--- 3 files changed, 75 insertions(+), 20 deletions(-) (limited to 'include') diff --git a/include/linux/cgroup.h b/include/linux/cgroup.h index 20411d2876f8..90bba9e62286 100644 --- a/include/linux/cgroup.h +++ b/include/linux/cgroup.h @@ -362,6 +362,23 @@ int cgroup_task_count(const struct cgroup *cgrp); /* Return true if cgrp is a descendant of the task's cgroup */ int cgroup_is_descendant(const struct cgroup *cgrp, struct task_struct *task); +/* + * When the subsys has to access css and may add permanent refcnt to css, + * it should take care of racy conditions with rmdir(). Following set of + * functions, is for stop/restart rmdir if necessary. + * Because these will call css_get/put, "css" should be alive css. + * + * cgroup_exclude_rmdir(); + * ...do some jobs which may access arbitrary empty cgroup + * cgroup_release_and_wakeup_rmdir(); + * + * When someone removes a cgroup while cgroup_exclude_rmdir() holds it, + * it sleeps and cgroup_release_and_wakeup_rmdir() will wake him up. + */ + +void cgroup_exclude_rmdir(struct cgroup_subsys_state *css); +void cgroup_release_and_wakeup_rmdir(struct cgroup_subsys_state *css); + /* * Control Group subsystem type. * See Documentation/cgroups/cgroups.txt for details diff --git a/kernel/cgroup.c b/kernel/cgroup.c index 250dac05680f..b6eadfe30e7b 100644 --- a/kernel/cgroup.c +++ b/kernel/cgroup.c @@ -735,16 +735,28 @@ static void cgroup_d_remove_dir(struct dentry *dentry) * reference to css->refcnt. In general, this refcnt is expected to goes down * to zero, soon. * - * CGRP_WAIT_ON_RMDIR flag is modified under cgroup's inode->i_mutex; + * CGRP_WAIT_ON_RMDIR flag is set under cgroup's inode->i_mutex; */ DECLARE_WAIT_QUEUE_HEAD(cgroup_rmdir_waitq); -static void cgroup_wakeup_rmdir_waiters(const struct cgroup *cgrp) +static void cgroup_wakeup_rmdir_waiter(struct cgroup *cgrp) { - if (unlikely(test_bit(CGRP_WAIT_ON_RMDIR, &cgrp->flags))) + if (unlikely(test_and_clear_bit(CGRP_WAIT_ON_RMDIR, &cgrp->flags))) wake_up_all(&cgroup_rmdir_waitq); } +void cgroup_exclude_rmdir(struct cgroup_subsys_state *css) +{ + css_get(css); +} + +void cgroup_release_and_wakeup_rmdir(struct cgroup_subsys_state *css) +{ + cgroup_wakeup_rmdir_waiter(css->cgroup); + css_put(css); +} + + static int rebind_subsystems(struct cgroupfs_root *root, unsigned long final_bits) { @@ -1359,7 +1371,7 @@ int cgroup_attach_task(struct cgroup *cgrp, struct task_struct *tsk) * wake up rmdir() waiter. the rmdir should fail since the cgroup * is no longer empty. */ - cgroup_wakeup_rmdir_waiters(cgrp); + cgroup_wakeup_rmdir_waiter(cgrp); return 0; } @@ -2743,34 +2755,43 @@ again: } mutex_unlock(&cgroup_mutex); + /* + * In general, subsystem has no css->refcnt after pre_destroy(). But + * in racy cases, subsystem may have to get css->refcnt after + * pre_destroy() and it makes rmdir return with -EBUSY. This sometimes + * make rmdir return -EBUSY too often. To avoid that, we use waitqueue + * for cgroup's rmdir. CGRP_WAIT_ON_RMDIR is for synchronizing rmdir + * and subsystem's reference count handling. Please see css_get/put + * and css_tryget() and cgroup_wakeup_rmdir_waiter() implementation. + */ + set_bit(CGRP_WAIT_ON_RMDIR, &cgrp->flags); + /* * Call pre_destroy handlers of subsys. Notify subsystems * that rmdir() request comes. */ ret = cgroup_call_pre_destroy(cgrp); - if (ret) + if (ret) { + clear_bit(CGRP_WAIT_ON_RMDIR, &cgrp->flags); return ret; + } mutex_lock(&cgroup_mutex); parent = cgrp->parent; if (atomic_read(&cgrp->count) || !list_empty(&cgrp->children)) { + clear_bit(CGRP_WAIT_ON_RMDIR, &cgrp->flags); mutex_unlock(&cgroup_mutex); return -EBUSY; } - /* - * css_put/get is provided for subsys to grab refcnt to css. In typical - * case, subsystem has no reference after pre_destroy(). But, under - * hierarchy management, some *temporal* refcnt can be hold. - * To avoid returning -EBUSY to a user, waitqueue is used. If subsys - * is really busy, it should return -EBUSY at pre_destroy(). wake_up - * is called when css_put() is called and refcnt goes down to 0. - */ - set_bit(CGRP_WAIT_ON_RMDIR, &cgrp->flags); prepare_to_wait(&cgroup_rmdir_waitq, &wait, TASK_INTERRUPTIBLE); - if (!cgroup_clear_css_refs(cgrp)) { mutex_unlock(&cgroup_mutex); - schedule(); + /* + * Because someone may call cgroup_wakeup_rmdir_waiter() before + * prepare_to_wait(), we need to check this flag. + */ + if (test_bit(CGRP_WAIT_ON_RMDIR, &cgrp->flags)) + schedule(); finish_wait(&cgroup_rmdir_waitq, &wait); clear_bit(CGRP_WAIT_ON_RMDIR, &cgrp->flags); if (signal_pending(current)) @@ -3342,7 +3363,7 @@ void __css_put(struct cgroup_subsys_state *css) set_bit(CGRP_RELEASABLE, &cgrp->flags); check_for_release(cgrp); } - cgroup_wakeup_rmdir_waiters(cgrp); + cgroup_wakeup_rmdir_waiter(cgrp); } rcu_read_unlock(); } diff --git a/mm/memcontrol.c b/mm/memcontrol.c index e717964cb5a0..fd4529d86de5 100644 --- a/mm/memcontrol.c +++ b/mm/memcontrol.c @@ -1207,6 +1207,12 @@ static int mem_cgroup_move_account(struct page_cgroup *pc, ret = 0; out: unlock_page_cgroup(pc); + /* + * We charges against "to" which may not have any tasks. Then, "to" + * can be under rmdir(). But in current implementation, caller of + * this function is just force_empty() and it's garanteed that + * "to" is never removed. So, we don't check rmdir status here. + */ return ret; } @@ -1428,6 +1434,7 @@ __mem_cgroup_commit_charge_swapin(struct page *page, struct mem_cgroup *ptr, return; if (!ptr) return; + cgroup_exclude_rmdir(&ptr->css); pc = lookup_page_cgroup(page); mem_cgroup_lru_del_before_commit_swapcache(page); __mem_cgroup_commit_charge(ptr, pc, ctype); @@ -1457,8 +1464,12 @@ __mem_cgroup_commit_charge_swapin(struct page *page, struct mem_cgroup *ptr, } rcu_read_unlock(); } - /* add this page(page_cgroup) to the LRU we want. */ - + /* + * At swapin, we may charge account against cgroup which has no tasks. + * So, rmdir()->pre_destroy() can be called while we do this charge. + * In that case, we need to call pre_destroy() again. check it here. + */ + cgroup_release_and_wakeup_rmdir(&ptr->css); } void mem_cgroup_commit_charge_swapin(struct page *page, struct mem_cgroup *ptr) @@ -1664,7 +1675,7 @@ void mem_cgroup_end_migration(struct mem_cgroup *mem, if (!mem) return; - + cgroup_exclude_rmdir(&mem->css); /* at migration success, oldpage->mapping is NULL. */ if (oldpage->mapping) { target = oldpage; @@ -1704,6 +1715,12 @@ void mem_cgroup_end_migration(struct mem_cgroup *mem, */ if (ctype == MEM_CGROUP_CHARGE_TYPE_MAPPED) mem_cgroup_uncharge_page(target); + /* + * At migration, we may charge account against cgroup which has no tasks + * So, rmdir()->pre_destroy() can be called while we do this charge. + * In that case, we need to call pre_destroy() again. check it here. + */ + cgroup_release_and_wakeup_rmdir(&mem->css); } /* -- cgit v1.2.3 From f5a55efa140f5e9c9dd0f398fef54f20cdb74ec9 Mon Sep 17 00:00:00 2001 From: Dave Jones Date: Wed, 29 Jul 2009 15:04:12 -0700 Subject: pps.h needs Found with make headers_check /usr/include/linux/pps.h:52: found __[us]{8,16,32,64} type without #include Signed-off-by: Dave Jones Cc: Rodolfo Giometti Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/pps.h | 2 ++ 1 file changed, 2 insertions(+) (limited to 'include') diff --git a/include/linux/pps.h b/include/linux/pps.h index cfe5c7214ec6..0194ab06177b 100644 --- a/include/linux/pps.h +++ b/include/linux/pps.h @@ -22,6 +22,8 @@ #ifndef _PPS_H_ #define _PPS_H_ +#include + #define PPS_VERSION "5.3.6" #define PPS_MAX_SOURCES 16 /* should be enough... */ -- cgit v1.2.3 From 534acc057b5a08ec33fa57cdd2f5a09ef124e7f2 Mon Sep 17 00:00:00 2001 From: Dave Hansen Date: Wed, 29 Jul 2009 15:04:18 -0700 Subject: lib: flexible array implementation Once a structure goes over PAGE_SIZE*2, we see occasional allocation failures. Some people have chosen to switch over to things like vmalloc() that will let them keep array-like access to such a large structures. But, vmalloc() has plenty of downsides. Here's an alternative. I think it's what Andrew was suggesting here: http://lkml.org/lkml/2009/7/2/518 I call it a flexible array. It does all of its work in PAGE_SIZE bits, so never does an order>0 allocation. The base level has PAGE_SIZE-2*sizeof(int) bytes of storage for pointers to the second level. So, with a 32-bit arch, you get about 4MB (4183112 bytes) of total storage when the objects pack nicely into a page. It is half that on 64-bit because the pointers are twice the size. There's a table detailing this in the code. There are kerneldocs for the functions, but here's an overview: flex_array_alloc() - dynamically allocate a base structure flex_array_free() - free the array and all of the second-level pages flex_array_free_parts() - free the second-level pages, but not the base (for static bases) flex_array_put() - copy into the array at the given index flex_array_get() - copy out of the array at the given index flex_array_prealloc() - preallocate the second-level pages between the given indexes to guarantee no allocs will occur at put() time. We could also potentially just pass the "element_size" into each of the API functions instead of storing it internally. That would get us one more base pointer on 32-bit. I've been testing this by running it in userspace. The header and patch that I've been using are here, as well as the little script I'm using to generate the size table which goes in the kerneldocs. http://sr71.net/~dave/linux/flexarray/ [akpm@linux-foundation.org: coding-style fixes] Signed-off-by: Dave Hansen Reviewed-by: KAMEZAWA Hiroyuki Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/flex_array.h | 47 ++++++++ lib/Makefile | 2 +- lib/flex_array.c | 269 +++++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 317 insertions(+), 1 deletion(-) create mode 100644 include/linux/flex_array.h create mode 100644 lib/flex_array.c (limited to 'include') diff --git a/include/linux/flex_array.h b/include/linux/flex_array.h new file mode 100644 index 000000000000..23c1ec79a31b --- /dev/null +++ b/include/linux/flex_array.h @@ -0,0 +1,47 @@ +#ifndef _FLEX_ARRAY_H +#define _FLEX_ARRAY_H + +#include +#include + +#define FLEX_ARRAY_PART_SIZE PAGE_SIZE +#define FLEX_ARRAY_BASE_SIZE PAGE_SIZE + +struct flex_array_part; + +/* + * This is meant to replace cases where an array-like + * structure has gotten too big to fit into kmalloc() + * and the developer is getting tempted to use + * vmalloc(). + */ + +struct flex_array { + union { + struct { + int element_size; + int total_nr_elements; + struct flex_array_part *parts[0]; + }; + /* + * This little trick makes sure that + * sizeof(flex_array) == PAGE_SIZE + */ + char padding[FLEX_ARRAY_BASE_SIZE]; + }; +}; + +#define FLEX_ARRAY_INIT(size, total) { { {\ + .element_size = (size), \ + .total_nr_elements = (total), \ +} } } + +struct flex_array *flex_array_alloc(int element_size, int total, gfp_t flags); +int flex_array_prealloc(struct flex_array *fa, int start, int end, gfp_t flags); +void flex_array_free(struct flex_array *fa); +void flex_array_free_parts(struct flex_array *fa); +int flex_array_put(struct flex_array *fa, int element_nr, void *src, + gfp_t flags); +void *flex_array_get(struct flex_array *fa, int element_nr); + +#endif /* _FLEX_ARRAY_H */ diff --git a/lib/Makefile b/lib/Makefile index b6d1857bbf08..2e78277eff9d 100644 --- a/lib/Makefile +++ b/lib/Makefile @@ -12,7 +12,7 @@ lib-y := ctype.o string.o vsprintf.o cmdline.o \ idr.o int_sqrt.o extable.o prio_tree.o \ sha1.o irq_regs.o reciprocal_div.o argv_split.o \ proportions.o prio_heap.o ratelimit.o show_mem.o \ - is_single_threaded.o plist.o decompress.o + is_single_threaded.o plist.o decompress.o flex_array.o lib-$(CONFIG_MMU) += ioremap.o lib-$(CONFIG_SMP) += cpumask.o diff --git a/lib/flex_array.c b/lib/flex_array.c new file mode 100644 index 000000000000..0e7894ce8882 --- /dev/null +++ b/lib/flex_array.c @@ -0,0 +1,269 @@ +/* + * Flexible array managed in PAGE_SIZE parts + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. + * + * Copyright IBM Corporation, 2009 + * + * Author: Dave Hansen + */ + +#include +#include +#include + +struct flex_array_part { + char elements[FLEX_ARRAY_PART_SIZE]; +}; + +static inline int __elements_per_part(int element_size) +{ + return FLEX_ARRAY_PART_SIZE / element_size; +} + +static inline int bytes_left_in_base(void) +{ + int element_offset = offsetof(struct flex_array, parts); + int bytes_left = FLEX_ARRAY_BASE_SIZE - element_offset; + return bytes_left; +} + +static inline int nr_base_part_ptrs(void) +{ + return bytes_left_in_base() / sizeof(struct flex_array_part *); +} + +/* + * If a user requests an allocation which is small + * enough, we may simply use the space in the + * flex_array->parts[] array to store the user + * data. + */ +static inline int elements_fit_in_base(struct flex_array *fa) +{ + int data_size = fa->element_size * fa->total_nr_elements; + if (data_size <= bytes_left_in_base()) + return 1; + return 0; +} + +/** + * flex_array_alloc - allocate a new flexible array + * @element_size: the size of individual elements in the array + * @total: total number of elements that this should hold + * + * Note: all locking must be provided by the caller. + * + * @total is used to size internal structures. If the user ever + * accesses any array indexes >=@total, it will produce errors. + * + * The maximum number of elements is defined as: the number of + * elements that can be stored in a page times the number of + * page pointers that we can fit in the base structure or (using + * integer math): + * + * (PAGE_SIZE/element_size) * (PAGE_SIZE-8)/sizeof(void *) + * + * Here's a table showing example capacities. Note that the maximum + * index that the get/put() functions is just nr_objects-1. This + * basically means that you get 4MB of storage on 32-bit and 2MB on + * 64-bit. + * + * + * Element size | Objects | Objects | + * PAGE_SIZE=4k | 32-bit | 64-bit | + * ---------------------------------| + * 1 bytes | 4186112 | 2093056 | + * 2 bytes | 2093056 | 1046528 | + * 3 bytes | 1395030 | 697515 | + * 4 bytes | 1046528 | 523264 | + * 32 bytes | 130816 | 65408 | + * 33 bytes | 126728 | 63364 | + * 2048 bytes | 2044 | 1022 | + * 2049 bytes | 1022 | 511 | + * void * | 1046528 | 261632 | + * + * Since 64-bit pointers are twice the size, we lose half the + * capacity in the base structure. Also note that no effort is made + * to efficiently pack objects across page boundaries. + */ +struct flex_array *flex_array_alloc(int element_size, int total, gfp_t flags) +{ + struct flex_array *ret; + int max_size = nr_base_part_ptrs() * __elements_per_part(element_size); + + /* max_size will end up 0 if element_size > PAGE_SIZE */ + if (total > max_size) + return NULL; + ret = kzalloc(sizeof(struct flex_array), flags); + if (!ret) + return NULL; + ret->element_size = element_size; + ret->total_nr_elements = total; + return ret; +} + +static int fa_element_to_part_nr(struct flex_array *fa, int element_nr) +{ + return element_nr / __elements_per_part(fa->element_size); +} + +/** + * flex_array_free_parts - just free the second-level pages + * @src: address of data to copy into the array + * @element_nr: index of the position in which to insert + * the new element. + * + * This is to be used in cases where the base 'struct flex_array' + * has been statically allocated and should not be free. + */ +void flex_array_free_parts(struct flex_array *fa) +{ + int part_nr; + int max_part = nr_base_part_ptrs(); + + if (elements_fit_in_base(fa)) + return; + for (part_nr = 0; part_nr < max_part; part_nr++) + kfree(fa->parts[part_nr]); +} + +void flex_array_free(struct flex_array *fa) +{ + flex_array_free_parts(fa); + kfree(fa); +} + +static int fa_index_inside_part(struct flex_array *fa, int element_nr) +{ + return element_nr % __elements_per_part(fa->element_size); +} + +static int index_inside_part(struct flex_array *fa, int element_nr) +{ + int part_offset = fa_index_inside_part(fa, element_nr); + return part_offset * fa->element_size; +} + +static struct flex_array_part * +__fa_get_part(struct flex_array *fa, int part_nr, gfp_t flags) +{ + struct flex_array_part *part = fa->parts[part_nr]; + if (!part) { + /* + * This leaves the part pages uninitialized + * and with potentially random data, just + * as if the user had kmalloc()'d the whole. + * __GFP_ZERO can be used to zero it. + */ + part = kmalloc(FLEX_ARRAY_PART_SIZE, flags); + if (!part) + return NULL; + fa->parts[part_nr] = part; + } + return part; +} + +/** + * flex_array_put - copy data into the array at @element_nr + * @src: address of data to copy into the array + * @element_nr: index of the position in which to insert + * the new element. + * + * Note that this *copies* the contents of @src into + * the array. If you are trying to store an array of + * pointers, make sure to pass in &ptr instead of ptr. + * + * Locking must be provided by the caller. + */ +int flex_array_put(struct flex_array *fa, int element_nr, void *src, gfp_t flags) +{ + int part_nr = fa_element_to_part_nr(fa, element_nr); + struct flex_array_part *part; + void *dst; + + if (element_nr >= fa->total_nr_elements) + return -ENOSPC; + if (elements_fit_in_base(fa)) + part = (struct flex_array_part *)&fa->parts[0]; + else + part = __fa_get_part(fa, part_nr, flags); + if (!part) + return -ENOMEM; + dst = &part->elements[index_inside_part(fa, element_nr)]; + memcpy(dst, src, fa->element_size); + return 0; +} + +/** + * flex_array_prealloc - guarantee that array space exists + * @start: index of first array element for which space is allocated + * @end: index of last (inclusive) element for which space is allocated + * + * This will guarantee that no future calls to flex_array_put() + * will allocate memory. It can be used if you are expecting to + * be holding a lock or in some atomic context while writing + * data into the array. + * + * Locking must be provided by the caller. + */ +int flex_array_prealloc(struct flex_array *fa, int start, int end, gfp_t flags) +{ + int start_part; + int end_part; + int part_nr; + struct flex_array_part *part; + + if (start >= fa->total_nr_elements || end >= fa->total_nr_elements) + return -ENOSPC; + if (elements_fit_in_base(fa)) + return 0; + start_part = fa_element_to_part_nr(fa, start); + end_part = fa_element_to_part_nr(fa, end); + for (part_nr = start_part; part_nr <= end_part; part_nr++) { + part = __fa_get_part(fa, part_nr, flags); + if (!part) + return -ENOMEM; + } + return 0; +} + +/** + * flex_array_get - pull data back out of the array + * @element_nr: index of the element to fetch from the array + * + * Returns a pointer to the data at index @element_nr. Note + * that this is a copy of the data that was passed in. If you + * are using this to store pointers, you'll get back &ptr. + * + * Locking must be provided by the caller. + */ +void *flex_array_get(struct flex_array *fa, int element_nr) +{ + int part_nr = fa_element_to_part_nr(fa, element_nr); + struct flex_array_part *part; + int index; + + if (element_nr >= fa->total_nr_elements) + return NULL; + if (!fa->parts[part_nr]) + return NULL; + if (elements_fit_in_base(fa)) + part = (struct flex_array_part *)&fa->parts[0]; + else + part = fa->parts[part_nr]; + index = index_inside_part(fa, element_nr); + return &part->elements[index_inside_part(fa, element_nr)]; +} -- cgit v1.2.3 From 812ed032cdc8138b7546eecc996879756b92d801 Mon Sep 17 00:00:00 2001 From: Jiri Slaby Date: Wed, 29 Jul 2009 15:04:19 -0700 Subject: uio: mark uio.h functions __KERNEL__ only To avoid userspace build failures such as: .../linux/uio.h:37: error: expected `=', `,', `;', `asm' or `__attribute__' before `iov_length' .../linux/uio.h:47: error: expected declaration specifiers or `...' before `size_t' move uio functions inside a __KERNEL__ block. Signed-off-by: Jiri Slaby Acked-by: Sam Ravnborg Cc: Alexander Viro Cc: Christoph Hellwig Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/uio.h | 17 ++++++++--------- 1 file changed, 8 insertions(+), 9 deletions(-) (limited to 'include') diff --git a/include/linux/uio.h b/include/linux/uio.h index b7fe13883bdb..98c114323a8b 100644 --- a/include/linux/uio.h +++ b/include/linux/uio.h @@ -19,15 +19,6 @@ struct iovec __kernel_size_t iov_len; /* Must be size_t (1003.1g) */ }; -#ifdef __KERNEL__ - -struct kvec { - void *iov_base; /* and that should *never* hold a userland pointer */ - size_t iov_len; -}; - -#endif - /* * UIO_MAXIOV shall be at least 16 1003.1g (5.4.1.1) */ @@ -35,6 +26,13 @@ struct kvec { #define UIO_FASTIOV 8 #define UIO_MAXIOV 1024 +#ifdef __KERNEL__ + +struct kvec { + void *iov_base; /* and that should *never* hold a userland pointer */ + size_t iov_len; +}; + /* * Total number of bytes covered by an iovec. * @@ -53,5 +51,6 @@ static inline size_t iov_length(const struct iovec *iov, unsigned long nr_segs) } unsigned long iov_shorten(struct iovec *iov, unsigned long nr_segs, size_t to); +#endif #endif -- cgit v1.2.3 From 2e04ef76916d1e29a077ea9d0f2003c8fd86724d Mon Sep 17 00:00:00 2001 From: Rusty Russell Date: Thu, 30 Jul 2009 16:03:45 -0600 Subject: lguest: fix comment style I don't really notice it (except to begrudge the extra vertical space), but Ingo does. And he pointed out that one excuse of lguest is as a teaching tool, it should set a good example. Signed-off-by: Rusty Russell Cc: Ingo Molnar --- Documentation/lguest/lguest.c | 540 ++++++++++++++++++++++------------ arch/x86/include/asm/lguest.h | 3 +- arch/x86/include/asm/lguest_hcall.h | 10 +- arch/x86/lguest/boot.c | 428 +++++++++++++++++---------- arch/x86/lguest/i386_head.S | 110 ++++--- drivers/lguest/core.c | 114 ++++--- drivers/lguest/hypercalls.c | 141 ++++++--- drivers/lguest/interrupts_and_traps.c | 288 ++++++++++++------ drivers/lguest/lg.h | 23 +- drivers/lguest/lguest_device.c | 150 ++++++---- drivers/lguest/lguest_user.c | 137 ++++++--- drivers/lguest/page_tables.c | 427 ++++++++++++++++++--------- drivers/lguest/segments.c | 106 ++++--- drivers/lguest/x86/core.c | 372 +++++++++++++++-------- drivers/lguest/x86/switcher_32.S | 18 +- include/linux/lguest.h | 36 ++- include/linux/lguest_launcher.h | 18 +- 17 files changed, 1906 insertions(+), 1015 deletions(-) (limited to 'include') diff --git a/Documentation/lguest/lguest.c b/Documentation/lguest/lguest.c index 45d7d6dcae7a..aa66a52b73e9 100644 --- a/Documentation/lguest/lguest.c +++ b/Documentation/lguest/lguest.c @@ -1,7 +1,9 @@ -/*P:100 This is the Launcher code, a simple program which lays out the - * "physical" memory for the new Guest by mapping the kernel image and - * the virtual devices, then opens /dev/lguest to tell the kernel - * about the Guest and control it. :*/ +/*P:100 + * This is the Launcher code, a simple program which lays out the "physical" + * memory for the new Guest by mapping the kernel image and the virtual + * devices, then opens /dev/lguest to tell the kernel about the Guest and + * control it. +:*/ #define _LARGEFILE64_SOURCE #define _GNU_SOURCE #include @@ -46,13 +48,15 @@ #include "linux/virtio_rng.h" #include "linux/virtio_ring.h" #include "asm/bootparam.h" -/*L:110 We can ignore the 39 include files we need for this program, but I do - * want to draw attention to the use of kernel-style types. +/*L:110 + * We can ignore the 39 include files we need for this program, but I do want + * to draw attention to the use of kernel-style types. * * As Linus said, "C is a Spartan language, and so should your naming be." I * like these abbreviations, so we define them here. Note that u64 is always * unsigned long long, which works on all Linux systems: this means that we can - * use %llu in printf for any u64. */ + * use %llu in printf for any u64. + */ typedef unsigned long long u64; typedef uint32_t u32; typedef uint16_t u16; @@ -69,8 +73,10 @@ typedef uint8_t u8; /* This will occupy 3 pages: it must be a power of 2. */ #define VIRTQUEUE_NUM 256 -/*L:120 verbose is both a global flag and a macro. The C preprocessor allows - * this, and although I wouldn't recommend it, it works quite nicely here. */ +/*L:120 + * verbose is both a global flag and a macro. The C preprocessor allows + * this, and although I wouldn't recommend it, it works quite nicely here. + */ static bool verbose; #define verbose(args...) \ do { if (verbose) printf(args); } while(0) @@ -100,8 +106,7 @@ struct device_list /* A single linked list of devices. */ struct device *dev; - /* And a pointer to the last device for easy append and also for - * configuration appending. */ + /* And a pointer to the last device for easy append. */ struct device *lastdev; }; @@ -168,20 +173,24 @@ static char **main_args; /* The original tty settings to restore on exit. */ static struct termios orig_term; -/* We have to be careful with barriers: our devices are all run in separate +/* + * We have to be careful with barriers: our devices are all run in separate * threads and so we need to make sure that changes visible to the Guest happen - * in precise order. */ + * in precise order. + */ #define wmb() __asm__ __volatile__("" : : : "memory") #define mb() __asm__ __volatile__("" : : : "memory") -/* Convert an iovec element to the given type. +/* + * Convert an iovec element to the given type. * * This is a fairly ugly trick: we need to know the size of the type and * alignment requirement to check the pointer is kosher. It's also nice to * have the name of the type in case we report failure. * * Typing those three things all the time is cumbersome and error prone, so we - * have a macro which sets them all up and passes to the real function. */ + * have a macro which sets them all up and passes to the real function. + */ #define convert(iov, type) \ ((type *)_convert((iov), sizeof(type), __alignof__(type), #type)) @@ -198,8 +207,10 @@ static void *_convert(struct iovec *iov, size_t size, size_t align, /* Wrapper for the last available index. Makes it easier to change. */ #define lg_last_avail(vq) ((vq)->last_avail_idx) -/* The virtio configuration space is defined to be little-endian. x86 is - * little-endian too, but it's nice to be explicit so we have these helpers. */ +/* + * The virtio configuration space is defined to be little-endian. x86 is + * little-endian too, but it's nice to be explicit so we have these helpers. + */ #define cpu_to_le16(v16) (v16) #define cpu_to_le32(v32) (v32) #define cpu_to_le64(v64) (v64) @@ -241,11 +252,12 @@ static u8 *get_feature_bits(struct device *dev) + dev->num_vq * sizeof(struct lguest_vqconfig); } -/*L:100 The Launcher code itself takes us out into userspace, that scary place - * where pointers run wild and free! Unfortunately, like most userspace - * programs, it's quite boring (which is why everyone likes to hack on the - * kernel!). Perhaps if you make up an Lguest Drinking Game at this point, it - * will get you through this section. Or, maybe not. +/*L:100 + * The Launcher code itself takes us out into userspace, that scary place where + * pointers run wild and free! Unfortunately, like most userspace programs, + * it's quite boring (which is why everyone likes to hack on the kernel!). + * Perhaps if you make up an Lguest Drinking Game at this point, it will get + * you through this section. Or, maybe not. * * The Launcher sets up a big chunk of memory to be the Guest's "physical" * memory and stores it in "guest_base". In other words, Guest physical == @@ -253,7 +265,8 @@ static u8 *get_feature_bits(struct device *dev) * * This can be tough to get your head around, but usually it just means that we * use these trivial conversion functions when the Guest gives us it's - * "physical" addresses: */ + * "physical" addresses: + */ static void *from_guest_phys(unsigned long addr) { return guest_base + addr; @@ -268,7 +281,8 @@ static unsigned long to_guest_phys(const void *addr) * Loading the Kernel. * * We start with couple of simple helper routines. open_or_die() avoids - * error-checking code cluttering the callers: */ + * error-checking code cluttering the callers: + */ static int open_or_die(const char *name, int flags) { int fd = open(name, flags); @@ -283,8 +297,10 @@ static void *map_zeroed_pages(unsigned int num) int fd = open_or_die("/dev/zero", O_RDONLY); void *addr; - /* We use a private mapping (ie. if we write to the page, it will be - * copied). */ + /* + * We use a private mapping (ie. if we write to the page, it will be + * copied). + */ addr = mmap(NULL, getpagesize() * num, PROT_READ|PROT_WRITE|PROT_EXEC, MAP_PRIVATE, fd, 0); if (addr == MAP_FAILED) @@ -305,20 +321,24 @@ static void *get_pages(unsigned int num) return addr; } -/* This routine is used to load the kernel or initrd. It tries mmap, but if +/* + * This routine is used to load the kernel or initrd. It tries mmap, but if * that fails (Plan 9's kernel file isn't nicely aligned on page boundaries), - * it falls back to reading the memory in. */ + * it falls back to reading the memory in. + */ static void map_at(int fd, void *addr, unsigned long offset, unsigned long len) { ssize_t r; - /* We map writable even though for some segments are marked read-only. + /* + * We map writable even though for some segments are marked read-only. * The kernel really wants to be writable: it patches its own * instructions. * * MAP_PRIVATE means that the page won't be copied until a write is * done to it. This allows us to share untouched memory between - * Guests. */ + * Guests. + */ if (mmap(addr, len, PROT_READ|PROT_WRITE|PROT_EXEC, MAP_FIXED|MAP_PRIVATE, fd, offset) != MAP_FAILED) return; @@ -329,7 +349,8 @@ static void map_at(int fd, void *addr, unsigned long offset, unsigned long len) err(1, "Reading offset %lu len %lu gave %zi", offset, len, r); } -/* This routine takes an open vmlinux image, which is in ELF, and maps it into +/* + * This routine takes an open vmlinux image, which is in ELF, and maps it into * the Guest memory. ELF = Embedded Linking Format, which is the format used * by all modern binaries on Linux including the kernel. * @@ -337,23 +358,28 @@ static void map_at(int fd, void *addr, unsigned long offset, unsigned long len) * address. We use the physical address; the Guest will map itself to the * virtual address. * - * We return the starting address. */ + * We return the starting address. + */ static unsigned long map_elf(int elf_fd, const Elf32_Ehdr *ehdr) { Elf32_Phdr phdr[ehdr->e_phnum]; unsigned int i; - /* Sanity checks on the main ELF header: an x86 executable with a - * reasonable number of correctly-sized program headers. */ + /* + * Sanity checks on the main ELF header: an x86 executable with a + * reasonable number of correctly-sized program headers. + */ if (ehdr->e_type != ET_EXEC || ehdr->e_machine != EM_386 || ehdr->e_phentsize != sizeof(Elf32_Phdr) || ehdr->e_phnum < 1 || ehdr->e_phnum > 65536U/sizeof(Elf32_Phdr)) errx(1, "Malformed elf header"); - /* An ELF executable contains an ELF header and a number of "program" + /* + * An ELF executable contains an ELF header and a number of "program" * headers which indicate which parts ("segments") of the program to - * load where. */ + * load where. + */ /* We read in all the program headers at once: */ if (lseek(elf_fd, ehdr->e_phoff, SEEK_SET) < 0) @@ -361,8 +387,10 @@ static unsigned long map_elf(int elf_fd, const Elf32_Ehdr *ehdr) if (read(elf_fd, phdr, sizeof(phdr)) != sizeof(phdr)) err(1, "Reading program headers"); - /* Try all the headers: there are usually only three. A read-only one, - * a read-write one, and a "note" section which we don't load. */ + /* + * Try all the headers: there are usually only three. A read-only one, + * a read-write one, and a "note" section which we don't load. + */ for (i = 0; i < ehdr->e_phnum; i++) { /* If this isn't a loadable segment, we ignore it */ if (phdr[i].p_type != PT_LOAD) @@ -380,13 +408,15 @@ static unsigned long map_elf(int elf_fd, const Elf32_Ehdr *ehdr) return ehdr->e_entry; } -/*L:150 A bzImage, unlike an ELF file, is not meant to be loaded. You're - * supposed to jump into it and it will unpack itself. We used to have to - * perform some hairy magic because the unpacking code scared me. +/*L:150 + * A bzImage, unlike an ELF file, is not meant to be loaded. You're supposed + * to jump into it and it will unpack itself. We used to have to perform some + * hairy magic because the unpacking code scared me. * * Fortunately, Jeremy Fitzhardinge convinced me it wasn't that hard and wrote * a small patch to jump over the tricky bits in the Guest, so now we just read - * the funky header so we know where in the file to load, and away we go! */ + * the funky header so we know where in the file to load, and away we go! + */ static unsigned long load_bzimage(int fd) { struct boot_params boot; @@ -394,8 +424,10 @@ static unsigned long load_bzimage(int fd) /* Modern bzImages get loaded at 1M. */ void *p = from_guest_phys(0x100000); - /* Go back to the start of the file and read the header. It should be - * a Linux boot header (see Documentation/x86/i386/boot.txt) */ + /* + * Go back to the start of the file and read the header. It should be + * a Linux boot header (see Documentation/x86/i386/boot.txt) + */ lseek(fd, 0, SEEK_SET); read(fd, &boot, sizeof(boot)); @@ -414,9 +446,11 @@ static unsigned long load_bzimage(int fd) return boot.hdr.code32_start; } -/*L:140 Loading the kernel is easy when it's a "vmlinux", but most kernels +/*L:140 + * Loading the kernel is easy when it's a "vmlinux", but most kernels * come wrapped up in the self-decompressing "bzImage" format. With a little - * work, we can load those, too. */ + * work, we can load those, too. + */ static unsigned long load_kernel(int fd) { Elf32_Ehdr hdr; @@ -433,24 +467,28 @@ static unsigned long load_kernel(int fd) return load_bzimage(fd); } -/* This is a trivial little helper to align pages. Andi Kleen hated it because +/* + * This is a trivial little helper to align pages. Andi Kleen hated it because * it calls getpagesize() twice: "it's dumb code." * * Kernel guys get really het up about optimization, even when it's not - * necessary. I leave this code as a reaction against that. */ + * necessary. I leave this code as a reaction against that. + */ static inline unsigned long page_align(unsigned long addr) { /* Add upwards and truncate downwards. */ return ((addr + getpagesize()-1) & ~(getpagesize()-1)); } -/*L:180 An "initial ram disk" is a disk image loaded into memory along with - * the kernel which the kernel can use to boot from without needing any - * drivers. Most distributions now use this as standard: the initrd contains - * the code to load the appropriate driver modules for the current machine. +/*L:180 + * An "initial ram disk" is a disk image loaded into memory along with the + * kernel which the kernel can use to boot from without needing any drivers. + * Most distributions now use this as standard: the initrd contains the code to + * load the appropriate driver modules for the current machine. * * Importantly, James Morris works for RedHat, and Fedora uses initrds for its - * kernels. He sent me this (and tells me when I break it). */ + * kernels. He sent me this (and tells me when I break it). + */ static unsigned long load_initrd(const char *name, unsigned long mem) { int ifd; @@ -462,12 +500,16 @@ static unsigned long load_initrd(const char *name, unsigned long mem) if (fstat(ifd, &st) < 0) err(1, "fstat() on initrd '%s'", name); - /* We map the initrd at the top of memory, but mmap wants it to be - * page-aligned, so we round the size up for that. */ + /* + * We map the initrd at the top of memory, but mmap wants it to be + * page-aligned, so we round the size up for that. + */ len = page_align(st.st_size); map_at(ifd, from_guest_phys(mem - len), 0, st.st_size); - /* Once a file is mapped, you can close the file descriptor. It's a - * little odd, but quite useful. */ + /* + * Once a file is mapped, you can close the file descriptor. It's a + * little odd, but quite useful. + */ close(ifd); verbose("mapped initrd %s size=%lu @ %p\n", name, len, (void*)mem-len); @@ -476,8 +518,10 @@ static unsigned long load_initrd(const char *name, unsigned long mem) } /*:*/ -/* Simple routine to roll all the commandline arguments together with spaces - * between them. */ +/* + * Simple routine to roll all the commandline arguments together with spaces + * between them. + */ static void concat(char *dst, char *args[]) { unsigned int i, len = 0; @@ -494,10 +538,12 @@ static void concat(char *dst, char *args[]) dst[len] = '\0'; } -/*L:185 This is where we actually tell the kernel to initialize the Guest. We +/*L:185 + * This is where we actually tell the kernel to initialize the Guest. We * saw the arguments it expects when we looked at initialize() in lguest_user.c: * the base of Guest "physical" memory, the top physical page to allow and the - * entry point for the Guest. */ + * entry point for the Guest. + */ static void tell_kernel(unsigned long start) { unsigned long args[] = { LHREQ_INITIALIZE, @@ -522,20 +568,26 @@ static void tell_kernel(unsigned long start) static void *_check_pointer(unsigned long addr, unsigned int size, unsigned int line) { - /* We have to separately check addr and addr+size, because size could - * be huge and addr + size might wrap around. */ + /* + * We have to separately check addr and addr+size, because size could + * be huge and addr + size might wrap around. + */ if (addr >= guest_limit || addr + size >= guest_limit) errx(1, "%s:%i: Invalid address %#lx", __FILE__, line, addr); - /* We return a pointer for the caller's convenience, now we know it's - * safe to use. */ + /* + * We return a pointer for the caller's convenience, now we know it's + * safe to use. + */ return from_guest_phys(addr); } /* A macro which transparently hands the line number to the real function. */ #define check_pointer(addr,size) _check_pointer(addr, size, __LINE__) -/* Each buffer in the virtqueues is actually a chain of descriptors. This +/* + * Each buffer in the virtqueues is actually a chain of descriptors. This * function returns the next descriptor in the chain, or vq->vring.num if we're - * at the end. */ + * at the end. + */ static unsigned next_desc(struct vring_desc *desc, unsigned int i, unsigned int max) { @@ -576,12 +628,14 @@ static void trigger_irq(struct virtqueue *vq) err(1, "Triggering irq %i", vq->config.irq); } -/* This looks in the virtqueue and for the first available buffer, and converts +/* + * This looks in the virtqueue and for the first available buffer, and converts * it to an iovec for convenient access. Since descriptors consist of some * number of output then some number of input descriptors, it's actually two * iovecs, but we pack them into one and note how many of each there were. * - * This function returns the descriptor number found. */ + * This function returns the descriptor number found. + */ static unsigned wait_for_vq_desc(struct virtqueue *vq, struct iovec iov[], unsigned int *out_num, unsigned int *in_num) @@ -599,8 +653,10 @@ static unsigned wait_for_vq_desc(struct virtqueue *vq, /* OK, now we need to know about added descriptors. */ vq->vring.used->flags &= ~VRING_USED_F_NO_NOTIFY; - /* They could have slipped one in as we were doing that: make - * sure it's written, then check again. */ + /* + * They could have slipped one in as we were doing that: make + * sure it's written, then check again. + */ mb(); if (last_avail != vq->vring.avail->idx) { vq->vring.used->flags |= VRING_USED_F_NO_NOTIFY; @@ -620,8 +676,10 @@ static unsigned wait_for_vq_desc(struct virtqueue *vq, errx(1, "Guest moved used index from %u to %u", last_avail, vq->vring.avail->idx); - /* Grab the next descriptor number they're advertising, and increment - * the index we've seen. */ + /* + * Grab the next descriptor number they're advertising, and increment + * the index we've seen. + */ head = vq->vring.avail->ring[last_avail % vq->vring.num]; lg_last_avail(vq)++; @@ -636,8 +694,10 @@ static unsigned wait_for_vq_desc(struct virtqueue *vq, desc = vq->vring.desc; i = head; - /* If this is an indirect entry, then this buffer contains a descriptor - * table which we handle as if it's any normal descriptor chain. */ + /* + * If this is an indirect entry, then this buffer contains a descriptor + * table which we handle as if it's any normal descriptor chain. + */ if (desc[i].flags & VRING_DESC_F_INDIRECT) { if (desc[i].len % sizeof(struct vring_desc)) errx(1, "Invalid size for indirect buffer table"); @@ -656,8 +716,10 @@ static unsigned wait_for_vq_desc(struct virtqueue *vq, if (desc[i].flags & VRING_DESC_F_WRITE) (*in_num)++; else { - /* If it's an output descriptor, they're all supposed - * to come before any input descriptors. */ + /* + * If it's an output descriptor, they're all supposed + * to come before any input descriptors. + */ if (*in_num) errx(1, "Descriptor has out after in"); (*out_num)++; @@ -671,14 +733,18 @@ static unsigned wait_for_vq_desc(struct virtqueue *vq, return head; } -/* After we've used one of their buffers, we tell them about it. We'll then - * want to send them an interrupt, using trigger_irq(). */ +/* + * After we've used one of their buffers, we tell them about it. We'll then + * want to send them an interrupt, using trigger_irq(). + */ static void add_used(struct virtqueue *vq, unsigned int head, int len) { struct vring_used_elem *used; - /* The virtqueue contains a ring of used buffers. Get a pointer to the - * next entry in that used ring. */ + /* + * The virtqueue contains a ring of used buffers. Get a pointer to the + * next entry in that used ring. + */ used = &vq->vring.used->ring[vq->vring.used->idx % vq->vring.num]; used->id = head; used->len = len; @@ -698,7 +764,8 @@ static void add_used_and_trigger(struct virtqueue *vq, unsigned head, int len) /* * The Console * - * We associate some data with the console for our exit hack. */ + * We associate some data with the console for our exit hack. + */ struct console_abort { /* How many times have they hit ^C? */ @@ -725,20 +792,24 @@ static void console_input(struct virtqueue *vq) if (len <= 0) { /* Ran out of input? */ warnx("Failed to get console input, ignoring console."); - /* For simplicity, dying threads kill the whole Launcher. So - * just nap here. */ + /* + * For simplicity, dying threads kill the whole Launcher. So + * just nap here. + */ for (;;) pause(); } add_used_and_trigger(vq, head, len); - /* Three ^C within one second? Exit. + /* + * Three ^C within one second? Exit. * * This is such a hack, but works surprisingly well. Each ^C has to * be in a buffer by itself, so they can't be too fast. But we check * that we get three within about a second, so they can't be too - * slow. */ + * slow. + */ if (len != 1 || ((char *)iov[0].iov_base)[0] != 3) { abort->count = 0; return; @@ -809,8 +880,7 @@ static bool will_block(int fd) return select(fd+1, &fdset, NULL, NULL, &zero) != 1; } -/* This is where we handle packets coming in from the tun device to our - * Guest. */ +/* This handles packets coming in from the tun device to our Guest. */ static void net_input(struct virtqueue *vq) { int len; @@ -842,8 +912,10 @@ static int do_thread(void *_vq) return 0; } -/* When a child dies, we kill our entire process group with SIGTERM. This - * also has the side effect that the shell restores the console for us! */ +/* + * When a child dies, we kill our entire process group with SIGTERM. This + * also has the side effect that the shell restores the console for us! + */ static void kill_launcher(int signal) { kill(0, SIGTERM); @@ -880,9 +952,10 @@ static void reset_device(struct device *dev) static void create_thread(struct virtqueue *vq) { - /* Create stack for thread and run it. Since stack grows - * upwards, we point the stack pointer to the end of this - * region. */ + /* + * Create stack for thread and run it. Since the stack grows upwards, + * we point the stack pointer to the end of this region. + */ char *stack = malloc(32768); unsigned long args[] = { LHREQ_EVENTFD, vq->config.pfn*getpagesize(), 0 }; @@ -981,8 +1054,11 @@ static void handle_output(unsigned long addr) } } - /* Early console write is done using notify on a nul-terminated string - * in Guest memory. */ + /* + * Early console write is done using notify on a nul-terminated string + * in Guest memory. It's also great for hacking debugging messages + * into a Guest. + */ if (addr >= guest_limit) errx(1, "Bad NOTIFY %#lx", addr); @@ -998,10 +1074,12 @@ static void handle_output(unsigned long addr) * routines to allocate and manage them. */ -/* The layout of the device page is a "struct lguest_device_desc" followed by a +/* + * The layout of the device page is a "struct lguest_device_desc" followed by a * number of virtqueue descriptors, then two sets of feature bits, then an * array of configuration bytes. This routine returns the configuration - * pointer. */ + * pointer. + */ static u8 *device_config(const struct device *dev) { return (void *)(dev->desc + 1) @@ -1009,9 +1087,11 @@ static u8 *device_config(const struct device *dev) + dev->feature_len * 2; } -/* This routine allocates a new "struct lguest_device_desc" from descriptor +/* + * This routine allocates a new "struct lguest_device_desc" from descriptor * table page just above the Guest's normal memory. It returns a pointer to - * that descriptor. */ + * that descriptor. + */ static struct lguest_device_desc *new_dev_desc(u16 type) { struct lguest_device_desc d = { .type = type }; @@ -1032,8 +1112,10 @@ static struct lguest_device_desc *new_dev_desc(u16 type) return memcpy(p, &d, sizeof(d)); } -/* Each device descriptor is followed by the description of its virtqueues. We - * specify how many descriptors the virtqueue is to have. */ +/* + * Each device descriptor is followed by the description of its virtqueues. We + * specify how many descriptors the virtqueue is to have. + */ static void add_virtqueue(struct device *dev, unsigned int num_descs, void (*service)(struct virtqueue *)) { @@ -1061,10 +1143,12 @@ static void add_virtqueue(struct device *dev, unsigned int num_descs, /* Initialize the vring. */ vring_init(&vq->vring, num_descs, p, LGUEST_VRING_ALIGN); - /* Append virtqueue to this device's descriptor. We use + /* + * Append virtqueue to this device's descriptor. We use * device_config() to get the end of the device's current virtqueues; * we check that we haven't added any config or feature information - * yet, otherwise we'd be overwriting them. */ + * yet, otherwise we'd be overwriting them. + */ assert(dev->desc->config_len == 0 && dev->desc->feature_len == 0); memcpy(device_config(dev), &vq->config, sizeof(vq->config)); dev->num_vq++; @@ -1072,14 +1156,18 @@ static void add_virtqueue(struct device *dev, unsigned int num_descs, verbose("Virtqueue page %#lx\n", to_guest_phys(p)); - /* Add to tail of list, so dev->vq is first vq, dev->vq->next is - * second. */ + /* + * Add to tail of list, so dev->vq is first vq, dev->vq->next is + * second. + */ for (i = &dev->vq; *i; i = &(*i)->next); *i = vq; } -/* The first half of the feature bitmask is for us to advertise features. The - * second half is for the Guest to accept features. */ +/* + * The first half of the feature bitmask is for us to advertise features. The + * second half is for the Guest to accept features. + */ static void add_feature(struct device *dev, unsigned bit) { u8 *features = get_feature_bits(dev); @@ -1093,9 +1181,11 @@ static void add_feature(struct device *dev, unsigned bit) features[bit / CHAR_BIT] |= (1 << (bit % CHAR_BIT)); } -/* This routine sets the configuration fields for an existing device's +/* + * This routine sets the configuration fields for an existing device's * descriptor. It only works for the last device, but that's OK because that's - * how we use it. */ + * how we use it. + */ static void set_config(struct device *dev, unsigned len, const void *conf) { /* Check we haven't overflowed our single page. */ @@ -1110,10 +1200,12 @@ static void set_config(struct device *dev, unsigned len, const void *conf) assert(dev->desc->config_len == len); } -/* This routine does all the creation and setup of a new device, including +/* + * This routine does all the creation and setup of a new device, including * calling new_dev_desc() to allocate the descriptor and device memory. * - * See what I mean about userspace being boring? */ + * See what I mean about userspace being boring? + */ static struct device *new_device(const char *name, u16 type) { struct device *dev = malloc(sizeof(*dev)); @@ -1126,10 +1218,12 @@ static struct device *new_device(const char *name, u16 type) dev->num_vq = 0; dev->running = false; - /* Append to device list. Prepending to a single-linked list is + /* + * Append to device list. Prepending to a single-linked list is * easier, but the user expects the devices to be arranged on the bus * in command-line order. The first network device on the command line - * is eth0, the first block device /dev/vda, etc. */ + * is eth0, the first block device /dev/vda, etc. + */ if (devices.lastdev) devices.lastdev->next = dev; else @@ -1139,8 +1233,10 @@ static struct device *new_device(const char *name, u16 type) return dev; } -/* Our first setup routine is the console. It's a fairly simple device, but - * UNIX tty handling makes it uglier than it could be. */ +/* + * Our first setup routine is the console. It's a fairly simple device, but + * UNIX tty handling makes it uglier than it could be. + */ static void setup_console(void) { struct device *dev; @@ -1148,8 +1244,10 @@ static void setup_console(void) /* If we can save the initial standard input settings... */ if (tcgetattr(STDIN_FILENO, &orig_term) == 0) { struct termios term = orig_term; - /* Then we turn off echo, line buffering and ^C etc. We want a - * raw input stream to the Guest. */ + /* + * Then we turn off echo, line buffering and ^C etc: We want a + * raw input stream to the Guest. + */ term.c_lflag &= ~(ISIG|ICANON|ECHO); tcsetattr(STDIN_FILENO, TCSANOW, &term); } @@ -1160,10 +1258,12 @@ static void setup_console(void) dev->priv = malloc(sizeof(struct console_abort)); ((struct console_abort *)dev->priv)->count = 0; - /* The console needs two virtqueues: the input then the output. When + /* + * The console needs two virtqueues: the input then the output. When * they put something the input queue, we make sure we're listening to * stdin. When they put something in the output queue, we write it to - * stdout. */ + * stdout. + */ add_virtqueue(dev, VIRTQUEUE_NUM, console_input); add_virtqueue(dev, VIRTQUEUE_NUM, console_output); @@ -1171,7 +1271,8 @@ static void setup_console(void) } /*:*/ -/*M:010 Inter-guest networking is an interesting area. Simplest is to have a +/*M:010 + * Inter-guest networking is an interesting area. Simplest is to have a * --sharenet= option which opens or creates a named pipe. This can be * used to send packets to another guest in a 1:1 manner. * @@ -1185,7 +1286,8 @@ static void setup_console(void) * multiple inter-guest channels behind one interface, although it would * require some manner of hotplugging new virtio channels. * - * Finally, we could implement a virtio network switch in the kernel. :*/ + * Finally, we could implement a virtio network switch in the kernel. +:*/ static u32 str2ip(const char *ipaddr) { @@ -1210,11 +1312,13 @@ static void str2mac(const char *macaddr, unsigned char mac[6]) mac[5] = m[5]; } -/* This code is "adapted" from libbridge: it attaches the Host end of the +/* + * This code is "adapted" from libbridge: it attaches the Host end of the * network device to the bridge device specified by the command line. * * This is yet another James Morris contribution (I'm an IP-level guy, so I - * dislike bridging), and I just try not to break it. */ + * dislike bridging), and I just try not to break it. + */ static void add_to_bridge(int fd, const char *if_name, const char *br_name) { int ifidx; @@ -1234,9 +1338,11 @@ static void add_to_bridge(int fd, const char *if_name, const char *br_name) err(1, "can't add %s to bridge %s", if_name, br_name); } -/* This sets up the Host end of the network device with an IP address, brings +/* + * This sets up the Host end of the network device with an IP address, brings * it up so packets will flow, the copies the MAC address into the hwaddr - * pointer. */ + * pointer. + */ static void configure_device(int fd, const char *tapif, u32 ipaddr) { struct ifreq ifr; @@ -1263,10 +1369,12 @@ static int get_tun_device(char tapif[IFNAMSIZ]) /* Start with this zeroed. Messy but sure. */ memset(&ifr, 0, sizeof(ifr)); - /* We open the /dev/net/tun device and tell it we want a tap device. A + /* + * We open the /dev/net/tun device and tell it we want a tap device. A * tap device is like a tun device, only somehow different. To tell * the truth, I completely blundered my way through this code, but it - * works now! */ + * works now! + */ netfd = open_or_die("/dev/net/tun", O_RDWR); ifr.ifr_flags = IFF_TAP | IFF_NO_PI | IFF_VNET_HDR; strcpy(ifr.ifr_name, "tap%d"); @@ -1277,18 +1385,22 @@ static int get_tun_device(char tapif[IFNAMSIZ]) TUN_F_CSUM|TUN_F_TSO4|TUN_F_TSO6|TUN_F_TSO_ECN) != 0) err(1, "Could not set features for tun device"); - /* We don't need checksums calculated for packets coming in this - * device: trust us! */ + /* + * We don't need checksums calculated for packets coming in this + * device: trust us! + */ ioctl(netfd, TUNSETNOCSUM, 1); memcpy(tapif, ifr.ifr_name, IFNAMSIZ); return netfd; } -/*L:195 Our network is a Host<->Guest network. This can either use bridging or +/*L:195 + * Our network is a Host<->Guest network. This can either use bridging or * routing, but the principle is the same: it uses the "tun" device to inject * packets into the Host as if they came in from a normal network card. We - * just shunt packets between the Guest and the tun device. */ + * just shunt packets between the Guest and the tun device. + */ static void setup_tun_net(char *arg) { struct device *dev; @@ -1305,13 +1417,14 @@ static void setup_tun_net(char *arg) dev = new_device("net", VIRTIO_ID_NET); dev->priv = net_info; - /* Network devices need a receive and a send queue, just like - * console. */ + /* Network devices need a recv and a send queue, just like console. */ add_virtqueue(dev, VIRTQUEUE_NUM, net_input); add_virtqueue(dev, VIRTQUEUE_NUM, net_output); - /* We need a socket to perform the magic network ioctls to bring up the - * tap interface, connect to the bridge etc. Any socket will do! */ + /* + * We need a socket to perform the magic network ioctls to bring up the + * tap interface, connect to the bridge etc. Any socket will do! + */ ipfd = socket(PF_INET, SOCK_DGRAM, IPPROTO_IP); if (ipfd < 0) err(1, "opening IP socket"); @@ -1366,7 +1479,8 @@ static void setup_tun_net(char *arg) devices.device_num, tapif, arg); } -/* Our block (disk) device should be really simple: the Guest asks for a block +/* + * Our block (disk) device should be really simple: the Guest asks for a block * number and we read or write that position in the file. Unfortunately, that * was amazingly slow: the Guest waits until the read is finished before * running anything else, even if it could have been doing useful work. @@ -1374,7 +1488,9 @@ static void setup_tun_net(char *arg) * We could use async I/O, except it's reputed to suck so hard that characters * actually go missing from your code when you try to use it. * - * So we farm the I/O out to thread, and communicate with it via a pipe. */ + * So this was one reason why lguest now does all virtqueue servicing in + * separate threads: it's more efficient and more like a real device. + */ /* This hangs off device->priv. */ struct vblk_info @@ -1412,9 +1528,11 @@ static void blk_request(struct virtqueue *vq) /* Get the next request. */ head = wait_for_vq_desc(vq, iov, &out_num, &in_num); - /* Every block request should contain at least one output buffer + /* + * Every block request should contain at least one output buffer * (detailing the location on disk and the type of request) and one - * input buffer (to hold the result). */ + * input buffer (to hold the result). + */ if (out_num == 0 || in_num == 0) errx(1, "Bad virtblk cmd %u out=%u in=%u", head, out_num, in_num); @@ -1423,33 +1541,41 @@ static void blk_request(struct virtqueue *vq) in = convert(&iov[out_num+in_num-1], u8); off = out->sector * 512; - /* The block device implements "barriers", where the Guest indicates + /* + * The block device implements "barriers", where the Guest indicates * that it wants all previous writes to occur before this write. We * don't have a way of asking our kernel to do a barrier, so we just - * synchronize all the data in the file. Pretty poor, no? */ + * synchronize all the data in the file. Pretty poor, no? + */ if (out->type & VIRTIO_BLK_T_BARRIER) fdatasync(vblk->fd); - /* In general the virtio block driver is allowed to try SCSI commands. - * It'd be nice if we supported eject, for example, but we don't. */ + /* + * In general the virtio block driver is allowed to try SCSI commands. + * It'd be nice if we supported eject, for example, but we don't. + */ if (out->type & VIRTIO_BLK_T_SCSI_CMD) { fprintf(stderr, "Scsi commands unsupported\n"); *in = VIRTIO_BLK_S_UNSUPP; wlen = sizeof(*in); } else if (out->type & VIRTIO_BLK_T_OUT) { - /* Write */ - - /* Move to the right location in the block file. This can fail - * if they try to write past end. */ + /* + * Write + * + * Move to the right location in the block file. This can fail + * if they try to write past end. + */ if (lseek64(vblk->fd, off, SEEK_SET) != off) err(1, "Bad seek to sector %llu", out->sector); ret = writev(vblk->fd, iov+1, out_num-1); verbose("WRITE to sector %llu: %i\n", out->sector, ret); - /* Grr... Now we know how long the descriptor they sent was, we + /* + * Grr... Now we know how long the descriptor they sent was, we * make sure they didn't try to write over the end of the block - * file (possibly extending it). */ + * file (possibly extending it). + */ if (ret > 0 && off + ret > vblk->len) { /* Trim it back to the correct length */ ftruncate64(vblk->fd, vblk->len); @@ -1459,10 +1585,12 @@ static void blk_request(struct virtqueue *vq) wlen = sizeof(*in); *in = (ret >= 0 ? VIRTIO_BLK_S_OK : VIRTIO_BLK_S_IOERR); } else { - /* Read */ - - /* Move to the right location in the block file. This can fail - * if they try to read past end. */ + /* + * Read + * + * Move to the right location in the block file. This can fail + * if they try to read past end. + */ if (lseek64(vblk->fd, off, SEEK_SET) != off) err(1, "Bad seek to sector %llu", out->sector); @@ -1477,10 +1605,12 @@ static void blk_request(struct virtqueue *vq) } } - /* OK, so we noted that it was pretty poor to use an fdatasync as a + /* + * OK, so we noted that it was pretty poor to use an fdatasync as a * barrier. But Christoph Hellwig points out that we need a sync * *afterwards* as well: "Barriers specify no reordering to the front - * or the back." And Jens Axboe confirmed it, so here we are: */ + * or the back." And Jens Axboe confirmed it, so here we are: + */ if (out->type & VIRTIO_BLK_T_BARRIER) fdatasync(vblk->fd); @@ -1494,7 +1624,7 @@ static void setup_block_file(const char *filename) struct vblk_info *vblk; struct virtio_blk_config conf; - /* The device responds to return from I/O thread. */ + /* Creat the device. */ dev = new_device("block", VIRTIO_ID_BLOCK); /* The device has one virtqueue, where the Guest places requests. */ @@ -1513,8 +1643,10 @@ static void setup_block_file(const char *filename) /* Tell Guest how many sectors this device has. */ conf.capacity = cpu_to_le64(vblk->len / 512); - /* Tell Guest not to put in too many descriptors at once: two are used - * for the in and out elements. */ + /* + * Tell Guest not to put in too many descriptors at once: two are used + * for the in and out elements. + */ add_feature(dev, VIRTIO_BLK_F_SEG_MAX); conf.seg_max = cpu_to_le32(VIRTQUEUE_NUM - 2); @@ -1525,16 +1657,18 @@ static void setup_block_file(const char *filename) ++devices.device_num, le64_to_cpu(conf.capacity)); } -struct rng_info { - int rfd; -}; - -/* Our random number generator device reads from /dev/random into the Guest's +/*L:211 + * Our random number generator device reads from /dev/random into the Guest's * input buffers. The usual case is that the Guest doesn't want random numbers * and so has no buffers although /dev/random is still readable, whereas * console is the reverse. * - * The same logic applies, however. */ + * The same logic applies, however. + */ +struct rng_info { + int rfd; +}; + static void rng_input(struct virtqueue *vq) { int len; @@ -1547,9 +1681,11 @@ static void rng_input(struct virtqueue *vq) if (out_num) errx(1, "Output buffers in rng?"); - /* This is why we convert to iovecs: the readv() call uses them, and so + /* + * This is why we convert to iovecs: the readv() call uses them, and so * it reads straight into the Guest's buffer. We loop to make sure we - * fill it. */ + * fill it. + */ while (!iov_empty(iov, in_num)) { len = readv(rng_info->rfd, iov, in_num); if (len <= 0) @@ -1562,15 +1698,18 @@ static void rng_input(struct virtqueue *vq) add_used(vq, head, totlen); } -/* And this creates a "hardware" random number device for the Guest. */ +/*L:199 + * This creates a "hardware" random number device for the Guest. + */ static void setup_rng(void) { struct device *dev; struct rng_info *rng_info = malloc(sizeof(*rng_info)); + /* Our device's privat info simply contains the /dev/random fd. */ rng_info->rfd = open_or_die("/dev/random", O_RDONLY); - /* The device responds to return from I/O thread. */ + /* Create the new device. */ dev = new_device("rng", VIRTIO_ID_RNG); dev->priv = rng_info; @@ -1586,8 +1725,10 @@ static void __attribute__((noreturn)) restart_guest(void) { unsigned int i; - /* Since we don't track all open fds, we simply close everything beyond - * stderr. */ + /* + * Since we don't track all open fds, we simply close everything beyond + * stderr. + */ for (i = 3; i < FD_SETSIZE; i++) close(i); @@ -1598,8 +1739,10 @@ static void __attribute__((noreturn)) restart_guest(void) err(1, "Could not exec %s", main_args[0]); } -/*L:220 Finally we reach the core of the Launcher which runs the Guest, serves - * its input and output, and finally, lays it to rest. */ +/*L:220 + * Finally we reach the core of the Launcher which runs the Guest, serves + * its input and output, and finally, lays it to rest. + */ static void __attribute__((noreturn)) run_guest(void) { for (;;) { @@ -1634,7 +1777,7 @@ static void __attribute__((noreturn)) run_guest(void) * * Are you ready? Take a deep breath and join me in the core of the Host, in * "make Host". - :*/ +:*/ static struct option opts[] = { { "verbose", 0, NULL, 'v' }, @@ -1655,8 +1798,7 @@ static void usage(void) /*L:105 The main routine is where the real work begins: */ int main(int argc, char *argv[]) { - /* Memory, top-level pagetable, code startpoint and size of the - * (optional) initrd. */ + /* Memory, code startpoint and size of the (optional) initrd. */ unsigned long mem = 0, start, initrd_size = 0; /* Two temporaries. */ int i, c; @@ -1668,24 +1810,30 @@ int main(int argc, char *argv[]) /* Save the args: we "reboot" by execing ourselves again. */ main_args = argv; - /* First we initialize the device list. We keep a pointer to the last + /* + * First we initialize the device list. We keep a pointer to the last * device, and the next interrupt number to use for devices (1: - * remember that 0 is used by the timer). */ + * remember that 0 is used by the timer). + */ devices.lastdev = NULL; devices.next_irq = 1; cpu_id = 0; - /* We need to know how much memory so we can set up the device + /* + * We need to know how much memory so we can set up the device * descriptor and memory pages for the devices as we parse the command * line. So we quickly look through the arguments to find the amount - * of memory now. */ + * of memory now. + */ for (i = 1; i < argc; i++) { if (argv[i][0] != '-') { mem = atoi(argv[i]) * 1024 * 1024; - /* We start by mapping anonymous pages over all of + /* + * We start by mapping anonymous pages over all of * guest-physical memory range. This fills it with 0, * and ensures that the Guest won't be killed when it - * tries to access it. */ + * tries to access it. + */ guest_base = map_zeroed_pages(mem / getpagesize() + DEVICE_PAGES); guest_limit = mem; @@ -1718,8 +1866,10 @@ int main(int argc, char *argv[]) usage(); } } - /* After the other arguments we expect memory and kernel image name, - * followed by command line arguments for the kernel. */ + /* + * After the other arguments we expect memory and kernel image name, + * followed by command line arguments for the kernel. + */ if (optind + 2 > argc) usage(); @@ -1737,20 +1887,26 @@ int main(int argc, char *argv[]) /* Map the initrd image if requested (at top of physical memory) */ if (initrd_name) { initrd_size = load_initrd(initrd_name, mem); - /* These are the location in the Linux boot header where the - * start and size of the initrd are expected to be found. */ + /* + * These are the location in the Linux boot header where the + * start and size of the initrd are expected to be found. + */ boot->hdr.ramdisk_image = mem - initrd_size; boot->hdr.ramdisk_size = initrd_size; /* The bootloader type 0xFF means "unknown"; that's OK. */ boot->hdr.type_of_loader = 0xFF; } - /* The Linux boot header contains an "E820" memory map: ours is a - * simple, single region. */ + /* + * The Linux boot header contains an "E820" memory map: ours is a + * simple, single region. + */ boot->e820_entries = 1; boot->e820_map[0] = ((struct e820entry) { 0, mem, E820_RAM }); - /* The boot header contains a command line pointer: we put the command - * line after the boot header. */ + /* + * The boot header contains a command line pointer: we put the command + * line after the boot header. + */ boot->hdr.cmd_line_ptr = to_guest_phys(boot + 1); /* We use a simple helper to copy the arguments separated by spaces. */ concat((char *)(boot + 1), argv+optind+2); @@ -1764,8 +1920,10 @@ int main(int argc, char *argv[]) /* Tell the entry path not to try to reload segment registers. */ boot->hdr.loadflags |= KEEP_SEGMENTS; - /* We tell the kernel to initialize the Guest: this returns the open - * /dev/lguest file descriptor. */ + /* + * We tell the kernel to initialize the Guest: this returns the open + * /dev/lguest file descriptor. + */ tell_kernel(start); /* Ensure that we terminate if a child dies. */ diff --git a/arch/x86/include/asm/lguest.h b/arch/x86/include/asm/lguest.h index 313389cd50d2..5136dad57cbb 100644 --- a/arch/x86/include/asm/lguest.h +++ b/arch/x86/include/asm/lguest.h @@ -17,8 +17,7 @@ /* Pages for switcher itself, then two pages per cpu */ #define TOTAL_SWITCHER_PAGES (SHARED_SWITCHER_PAGES + 2 * nr_cpu_ids) -/* We map at -4M (-2M when PAE is activated) for ease of mapping - * into the guest (one PTE page). */ +/* We map at -4M (-2M for PAE) for ease of mapping (one PTE page). */ #ifdef CONFIG_X86_PAE #define SWITCHER_ADDR 0xFFE00000 #else diff --git a/arch/x86/include/asm/lguest_hcall.h b/arch/x86/include/asm/lguest_hcall.h index 33600a66755f..cceb73e12e50 100644 --- a/arch/x86/include/asm/lguest_hcall.h +++ b/arch/x86/include/asm/lguest_hcall.h @@ -30,7 +30,8 @@ #include #include -/*G:030 But first, how does our Guest contact the Host to ask for privileged +/*G:030 + * But first, how does our Guest contact the Host to ask for privileged * operations? There are two ways: the direct way is to make a "hypercall", * to make requests of the Host Itself. * @@ -41,16 +42,15 @@ * * Grossly invalid calls result in Sudden Death at the hands of the vengeful * Host, rather than returning failure. This reflects Winston Churchill's - * definition of a gentleman: "someone who is only rude intentionally". */ -/*:*/ + * definition of a gentleman: "someone who is only rude intentionally". +:*/ /* Can't use our min() macro here: needs to be a constant */ #define LGUEST_IRQS (NR_IRQS < 32 ? NR_IRQS: 32) #define LHCALL_RING_SIZE 64 struct hcall_args { - /* These map directly onto eax, ebx, ecx, edx and esi - * in struct lguest_regs */ + /* These map directly onto eax/ebx/ecx/edx/esi in struct lguest_regs */ unsigned long arg0, arg1, arg2, arg3, arg4; }; diff --git a/arch/x86/lguest/boot.c b/arch/x86/lguest/boot.c index f2bf1f73d468..025c04d18f2b 100644 --- a/arch/x86/lguest/boot.c +++ b/arch/x86/lguest/boot.c @@ -22,7 +22,8 @@ * * So how does the kernel know it's a Guest? We'll see that later, but let's * just say that we end up here where we replace the native functions various - * "paravirt" structures with our Guest versions, then boot like normal. :*/ + * "paravirt" structures with our Guest versions, then boot like normal. +:*/ /* * Copyright (C) 2006, Rusty Russell IBM Corporation. @@ -74,7 +75,8 @@ * * The Guest in our tale is a simple creature: identical to the Host but * behaving in simplified but equivalent ways. In particular, the Guest is the - * same kernel as the Host (or at least, built from the same source code). :*/ + * same kernel as the Host (or at least, built from the same source code). +:*/ struct lguest_data lguest_data = { .hcall_status = { [0 ... LHCALL_RING_SIZE-1] = 0xFF }, @@ -85,7 +87,8 @@ struct lguest_data lguest_data = { .syscall_vec = SYSCALL_VECTOR, }; -/*G:037 async_hcall() is pretty simple: I'm quite proud of it really. We have a +/*G:037 + * async_hcall() is pretty simple: I'm quite proud of it really. We have a * ring buffer of stored hypercalls which the Host will run though next time we * do a normal hypercall. Each entry in the ring has 5 slots for the hypercall * arguments, and a "hcall_status" word which is 0 if the call is ready to go, @@ -94,7 +97,8 @@ struct lguest_data lguest_data = { * If we come around to a slot which hasn't been finished, then the table is * full and we just make the hypercall directly. This has the nice side * effect of causing the Host to run all the stored calls in the ring buffer - * which empties it for next time! */ + * which empties it for next time! + */ static void async_hcall(unsigned long call, unsigned long arg1, unsigned long arg2, unsigned long arg3, unsigned long arg4) @@ -103,9 +107,11 @@ static void async_hcall(unsigned long call, unsigned long arg1, static unsigned int next_call; unsigned long flags; - /* Disable interrupts if not already disabled: we don't want an + /* + * Disable interrupts if not already disabled: we don't want an * interrupt handler making a hypercall while we're already doing - * one! */ + * one! + */ local_irq_save(flags); if (lguest_data.hcall_status[next_call] != 0xFF) { /* Table full, so do normal hcall which will flush table. */ @@ -125,8 +131,9 @@ static void async_hcall(unsigned long call, unsigned long arg1, local_irq_restore(flags); } -/*G:035 Notice the lazy_hcall() above, rather than hcall(). This is our first - * real optimization trick! +/*G:035 + * Notice the lazy_hcall() above, rather than hcall(). This is our first real + * optimization trick! * * When lazy_mode is set, it means we're allowed to defer all hypercalls and do * them as a batch when lazy_mode is eventually turned off. Because hypercalls @@ -136,7 +143,8 @@ static void async_hcall(unsigned long call, unsigned long arg1, * lguest_leave_lazy_mode(). * * So, when we're in lazy mode, we call async_hcall() to store the call for - * future processing: */ + * future processing: + */ static void lazy_hcall1(unsigned long call, unsigned long arg1) { @@ -208,9 +216,11 @@ static void lguest_end_context_switch(struct task_struct *next) * check there before it tries to deliver an interrupt. */ -/* save_flags() is expected to return the processor state (ie. "flags"). The +/* + * save_flags() is expected to return the processor state (ie. "flags"). The * flags word contains all kind of stuff, but in practice Linux only cares - * about the interrupt flag. Our "save_flags()" just returns that. */ + * about the interrupt flag. Our "save_flags()" just returns that. + */ static unsigned long save_fl(void) { return lguest_data.irq_enabled; @@ -222,13 +232,15 @@ static void irq_disable(void) lguest_data.irq_enabled = 0; } -/* Let's pause a moment. Remember how I said these are called so often? +/* + * Let's pause a moment. Remember how I said these are called so often? * Jeremy Fitzhardinge optimized them so hard early in 2009 that he had to * break some rules. In particular, these functions are assumed to save their * own registers if they need to: normal C functions assume they can trash the * eax register. To use normal C functions, we use * PV_CALLEE_SAVE_REGS_THUNK(), which pushes %eax onto the stack, calls the - * C function, then restores it. */ + * C function, then restores it. + */ PV_CALLEE_SAVE_REGS_THUNK(save_fl); PV_CALLEE_SAVE_REGS_THUNK(irq_disable); /*:*/ @@ -237,18 +249,20 @@ PV_CALLEE_SAVE_REGS_THUNK(irq_disable); extern void lg_irq_enable(void); extern void lg_restore_fl(unsigned long flags); -/*M:003 Note that we don't check for outstanding interrupts when we re-enable - * them (or when we unmask an interrupt). This seems to work for the moment, - * since interrupts are rare and we'll just get the interrupt on the next timer - * tick, but now we can run with CONFIG_NO_HZ, we should revisit this. One way - * would be to put the "irq_enabled" field in a page by itself, and have the - * Host write-protect it when an interrupt comes in when irqs are disabled. - * There will then be a page fault as soon as interrupts are re-enabled. +/*M:003 + * Note that we don't check for outstanding interrupts when we re-enable them + * (or when we unmask an interrupt). This seems to work for the moment, since + * interrupts are rare and we'll just get the interrupt on the next timer tick, + * but now we can run with CONFIG_NO_HZ, we should revisit this. One way would + * be to put the "irq_enabled" field in a page by itself, and have the Host + * write-protect it when an interrupt comes in when irqs are disabled. There + * will then be a page fault as soon as interrupts are re-enabled. * * A better method is to implement soft interrupt disable generally for x86: * instead of disabling interrupts, we set a flag. If an interrupt does come * in, we then disable them for real. This is uncommon, so we could simply use - * a hypercall for interrupt control and not worry about efficiency. :*/ + * a hypercall for interrupt control and not worry about efficiency. +:*/ /*G:034 * The Interrupt Descriptor Table (IDT). @@ -261,10 +275,12 @@ extern void lg_restore_fl(unsigned long flags); static void lguest_write_idt_entry(gate_desc *dt, int entrynum, const gate_desc *g) { - /* The gate_desc structure is 8 bytes long: we hand it to the Host in + /* + * The gate_desc structure is 8 bytes long: we hand it to the Host in * two 32-bit chunks. The whole 32-bit kernel used to hand descriptors * around like this; typesafety wasn't a big concern in Linux's early - * years. */ + * years. + */ u32 *desc = (u32 *)g; /* Keep the local copy up to date. */ native_write_idt_entry(dt, entrynum, g); @@ -272,9 +288,11 @@ static void lguest_write_idt_entry(gate_desc *dt, kvm_hypercall3(LHCALL_LOAD_IDT_ENTRY, entrynum, desc[0], desc[1]); } -/* Changing to a different IDT is very rare: we keep the IDT up-to-date every +/* + * Changing to a different IDT is very rare: we keep the IDT up-to-date every * time it is written, so we can simply loop through all entries and tell the - * Host about them. */ + * Host about them. + */ static void lguest_load_idt(const struct desc_ptr *desc) { unsigned int i; @@ -305,9 +323,11 @@ static void lguest_load_gdt(const struct desc_ptr *desc) kvm_hypercall3(LHCALL_LOAD_GDT_ENTRY, i, gdt[i].a, gdt[i].b); } -/* For a single GDT entry which changes, we do the lazy thing: alter our GDT, +/* + * For a single GDT entry which changes, we do the lazy thing: alter our GDT, * then tell the Host to reload the entire thing. This operation is so rare - * that this naive implementation is reasonable. */ + * that this naive implementation is reasonable. + */ static void lguest_write_gdt_entry(struct desc_struct *dt, int entrynum, const void *desc, int type) { @@ -317,29 +337,36 @@ static void lguest_write_gdt_entry(struct desc_struct *dt, int entrynum, dt[entrynum].a, dt[entrynum].b); } -/* OK, I lied. There are three "thread local storage" GDT entries which change +/* + * OK, I lied. There are three "thread local storage" GDT entries which change * on every context switch (these three entries are how glibc implements - * __thread variables). So we have a hypercall specifically for this case. */ + * __thread variables). So we have a hypercall specifically for this case. + */ static void lguest_load_tls(struct thread_struct *t, unsigned int cpu) { - /* There's one problem which normal hardware doesn't have: the Host + /* + * There's one problem which normal hardware doesn't have: the Host * can't handle us removing entries we're currently using. So we clear - * the GS register here: if it's needed it'll be reloaded anyway. */ + * the GS register here: if it's needed it'll be reloaded anyway. + */ lazy_load_gs(0); lazy_hcall2(LHCALL_LOAD_TLS, __pa(&t->tls_array), cpu); } -/*G:038 That's enough excitement for now, back to ploughing through each of - * the different pv_ops structures (we're about 1/3 of the way through). +/*G:038 + * That's enough excitement for now, back to ploughing through each of the + * different pv_ops structures (we're about 1/3 of the way through). * * This is the Local Descriptor Table, another weird Intel thingy. Linux only * uses this for some strange applications like Wine. We don't do anything - * here, so they'll get an informative and friendly Segmentation Fault. */ + * here, so they'll get an informative and friendly Segmentation Fault. + */ static void lguest_set_ldt(const void *addr, unsigned entries) { } -/* This loads a GDT entry into the "Task Register": that entry points to a +/* + * This loads a GDT entry into the "Task Register": that entry points to a * structure called the Task State Segment. Some comments scattered though the * kernel code indicate that this used for task switching in ages past, along * with blood sacrifice and astrology. @@ -347,19 +374,21 @@ static void lguest_set_ldt(const void *addr, unsigned entries) * Now there's nothing interesting in here that we don't get told elsewhere. * But the native version uses the "ltr" instruction, which makes the Host * complain to the Guest about a Segmentation Fault and it'll oops. So we - * override the native version with a do-nothing version. */ + * override the native version with a do-nothing version. + */ static void lguest_load_tr_desc(void) { } -/* The "cpuid" instruction is a way of querying both the CPU identity +/* + * The "cpuid" instruction is a way of querying both the CPU identity * (manufacturer, model, etc) and its features. It was introduced before the * Pentium in 1993 and keeps getting extended by both Intel, AMD and others. * As you might imagine, after a decade and a half this treatment, it is now a * giant ball of hair. Its entry in the current Intel manual runs to 28 pages. * * This instruction even it has its own Wikipedia entry. The Wikipedia entry - * has been translated into 4 languages. I am not making this up! + * has been translated into 5 languages. I am not making this up! * * We could get funky here and identify ourselves as "GenuineLguest", but * instead we just use the real "cpuid" instruction. Then I pretty much turned @@ -371,7 +400,8 @@ static void lguest_load_tr_desc(void) * Replacing the cpuid so we can turn features off is great for the kernel, but * anyone (including userspace) can just use the raw "cpuid" instruction and * the Host won't even notice since it isn't privileged. So we try not to get - * too worked up about it. */ + * too worked up about it. + */ static void lguest_cpuid(unsigned int *ax, unsigned int *bx, unsigned int *cx, unsigned int *dx) { @@ -379,43 +409,63 @@ static void lguest_cpuid(unsigned int *ax, unsigned int *bx, native_cpuid(ax, bx, cx, dx); switch (function) { - case 0: /* ID and highest CPUID. Futureproof a little by sticking to - * older ones. */ + /* + * CPUID 0 gives the highest legal CPUID number (and the ID string). + * We futureproof our code a little by sticking to known CPUID values. + */ + case 0: if (*ax > 5) *ax = 5; break; - case 1: /* Basic feature request. */ - /* We only allow kernel to see SSE3, CMPXCHG16B and SSSE3 */ + + /* + * CPUID 1 is a basic feature request. + * + * CX: we only allow kernel to see SSE3, CMPXCHG16B and SSSE3 + * DX: SSE, SSE2, FXSR, MMX, CMOV, CMPXCHG8B, TSC, FPU and PAE. + */ + case 1: *cx &= 0x00002201; - /* SSE, SSE2, FXSR, MMX, CMOV, CMPXCHG8B, TSC, FPU, PAE. */ *dx &= 0x07808151; - /* The Host can do a nice optimization if it knows that the + /* + * The Host can do a nice optimization if it knows that the * kernel mappings (addresses above 0xC0000000 or whatever * PAGE_OFFSET is set to) haven't changed. But Linux calls * flush_tlb_user() for both user and kernel mappings unless - * the Page Global Enable (PGE) feature bit is set. */ + * the Page Global Enable (PGE) feature bit is set. + */ *dx |= 0x00002000; - /* We also lie, and say we're family id 5. 6 or greater + /* + * We also lie, and say we're family id 5. 6 or greater * leads to a rdmsr in early_init_intel which we can't handle. - * Family ID is returned as bits 8-12 in ax. */ + * Family ID is returned as bits 8-12 in ax. + */ *ax &= 0xFFFFF0FF; *ax |= 0x00000500; break; + /* + * 0x80000000 returns the highest Extended Function, so we futureproof + * like we do above by limiting it to known fields. + */ case 0x80000000: - /* Futureproof this a little: if they ask how much extended - * processor information there is, limit it to known fields. */ if (*ax > 0x80000008) *ax = 0x80000008; break; + + /* + * PAE systems can mark pages as non-executable. Linux calls this the + * NX bit. Intel calls it XD (eXecute Disable), AMD EVP (Enhanced + * Virus Protection). We just switch turn if off here, since we don't + * support it. + */ case 0x80000001: - /* Here we should fix nx cap depending on host. */ - /* For this version of PAE, we just clear NX bit. */ *dx &= ~(1 << 20); break; } } -/* Intel has four control registers, imaginatively named cr0, cr2, cr3 and cr4. +/* + * Intel has four control registers, imaginatively named cr0, cr2, cr3 and cr4. * I assume there's a cr1, but it hasn't bothered us yet, so we'll not bother * it. The Host needs to know when the Guest wants to change them, so we have * a whole series of functions like read_cr0() and write_cr0(). @@ -430,7 +480,8 @@ static void lguest_cpuid(unsigned int *ax, unsigned int *bx, * name like "FPUTRAP bit" be a little less cryptic? * * We store cr0 locally because the Host never changes it. The Guest sometimes - * wants to read it and we'd prefer not to bother the Host unnecessarily. */ + * wants to read it and we'd prefer not to bother the Host unnecessarily. + */ static unsigned long current_cr0; static void lguest_write_cr0(unsigned long val) { @@ -443,18 +494,22 @@ static unsigned long lguest_read_cr0(void) return current_cr0; } -/* Intel provided a special instruction to clear the TS bit for people too cool +/* + * Intel provided a special instruction to clear the TS bit for people too cool * to use write_cr0() to do it. This "clts" instruction is faster, because all - * the vowels have been optimized out. */ + * the vowels have been optimized out. + */ static void lguest_clts(void) { lazy_hcall1(LHCALL_TS, 0); current_cr0 &= ~X86_CR0_TS; } -/* cr2 is the virtual address of the last page fault, which the Guest only ever +/* + * cr2 is the virtual address of the last page fault, which the Guest only ever * reads. The Host kindly writes this into our "struct lguest_data", so we - * just read it out of there. */ + * just read it out of there. + */ static unsigned long lguest_read_cr2(void) { return lguest_data.cr2; @@ -463,10 +518,12 @@ static unsigned long lguest_read_cr2(void) /* See lguest_set_pte() below. */ static bool cr3_changed = false; -/* cr3 is the current toplevel pagetable page: the principle is the same as +/* + * cr3 is the current toplevel pagetable page: the principle is the same as * cr0. Keep a local copy, and tell the Host when it changes. The only * difference is that our local copy is in lguest_data because the Host needs - * to set it upon our initial hypercall. */ + * to set it upon our initial hypercall. + */ static void lguest_write_cr3(unsigned long cr3) { lguest_data.pgdir = cr3; @@ -538,10 +595,12 @@ static void lguest_write_cr4(unsigned long val) * the real page tables based on the Guests'. */ -/* The Guest calls this to set a second-level entry (pte), ie. to map a page +/* + * The Guest calls this to set a second-level entry (pte), ie. to map a page * into a process' address space. We set the entry then tell the Host the * toplevel and address this corresponds to. The Guest uses one pagetable per - * process, so we need to tell the Host which one we're changing (mm->pgd). */ + * process, so we need to tell the Host which one we're changing (mm->pgd). + */ static void lguest_pte_update(struct mm_struct *mm, unsigned long addr, pte_t *ptep) { @@ -560,10 +619,13 @@ static void lguest_set_pte_at(struct mm_struct *mm, unsigned long addr, lguest_pte_update(mm, addr, ptep); } -/* The Guest calls lguest_set_pud to set a top-level entry and lguest_set_pmd +/* + * The Guest calls lguest_set_pud to set a top-level entry and lguest_set_pmd * to set a middle-level entry when PAE is activated. + * * Again, we set the entry then tell the Host which page we changed, - * and the index of the entry we changed. */ + * and the index of the entry we changed. + */ #ifdef CONFIG_X86_PAE static void lguest_set_pud(pud_t *pudp, pud_t pudval) { @@ -582,8 +644,7 @@ static void lguest_set_pmd(pmd_t *pmdp, pmd_t pmdval) } #else -/* The Guest calls lguest_set_pmd to set a top-level entry when PAE is not - * activated. */ +/* The Guest calls lguest_set_pmd to set a top-level entry when !PAE. */ static void lguest_set_pmd(pmd_t *pmdp, pmd_t pmdval) { native_set_pmd(pmdp, pmdval); @@ -592,7 +653,8 @@ static void lguest_set_pmd(pmd_t *pmdp, pmd_t pmdval) } #endif -/* There are a couple of legacy places where the kernel sets a PTE, but we +/* + * There are a couple of legacy places where the kernel sets a PTE, but we * don't know the top level any more. This is useless for us, since we don't * know which pagetable is changing or what address, so we just tell the Host * to forget all of them. Fortunately, this is very rare. @@ -600,7 +662,8 @@ static void lguest_set_pmd(pmd_t *pmdp, pmd_t pmdval) * ... except in early boot when the kernel sets up the initial pagetables, * which makes booting astonishingly slow: 1.83 seconds! So we don't even tell * the Host anything changed until we've done the first page table switch, - * which brings boot back to 0.25 seconds. */ + * which brings boot back to 0.25 seconds. + */ static void lguest_set_pte(pte_t *ptep, pte_t pteval) { native_set_pte(ptep, pteval); @@ -628,7 +691,8 @@ void lguest_pmd_clear(pmd_t *pmdp) } #endif -/* Unfortunately for Lguest, the pv_mmu_ops for page tables were based on +/* + * Unfortunately for Lguest, the pv_mmu_ops for page tables were based on * native page table operations. On native hardware you can set a new page * table entry whenever you want, but if you want to remove one you have to do * a TLB flush (a TLB is a little cache of page table entries kept by the CPU). @@ -637,24 +701,29 @@ void lguest_pmd_clear(pmd_t *pmdp) * called when a valid entry is written, not when it's removed (ie. marked not * present). Instead, this is where we come when the Guest wants to remove a * page table entry: we tell the Host to set that entry to 0 (ie. the present - * bit is zero). */ + * bit is zero). + */ static void lguest_flush_tlb_single(unsigned long addr) { /* Simply set it to zero: if it was not, it will fault back in. */ lazy_hcall3(LHCALL_SET_PTE, lguest_data.pgdir, addr, 0); } -/* This is what happens after the Guest has removed a large number of entries. +/* + * This is what happens after the Guest has removed a large number of entries. * This tells the Host that any of the page table entries for userspace might - * have changed, ie. virtual addresses below PAGE_OFFSET. */ + * have changed, ie. virtual addresses below PAGE_OFFSET. + */ static void lguest_flush_tlb_user(void) { lazy_hcall1(LHCALL_FLUSH_TLB, 0); } -/* This is called when the kernel page tables have changed. That's not very +/* + * This is called when the kernel page tables have changed. That's not very * common (unless the Guest is using highmem, which makes the Guest extremely - * slow), so it's worth separating this from the user flushing above. */ + * slow), so it's worth separating this from the user flushing above. + */ static void lguest_flush_tlb_kernel(void) { lazy_hcall1(LHCALL_FLUSH_TLB, 1); @@ -691,23 +760,27 @@ static struct irq_chip lguest_irq_controller = { .unmask = enable_lguest_irq, }; -/* This sets up the Interrupt Descriptor Table (IDT) entry for each hardware +/* + * This sets up the Interrupt Descriptor Table (IDT) entry for each hardware * interrupt (except 128, which is used for system calls), and then tells the * Linux infrastructure that each interrupt is controlled by our level-based - * lguest interrupt controller. */ + * lguest interrupt controller. + */ static void __init lguest_init_IRQ(void) { unsigned int i; for (i = FIRST_EXTERNAL_VECTOR; i < NR_VECTORS; i++) { - /* Some systems map "vectors" to interrupts weirdly. Lguest has - * a straightforward 1 to 1 mapping, so force that here. */ + /* Some systems map "vectors" to interrupts weirdly. Not us! */ __get_cpu_var(vector_irq)[i] = i - FIRST_EXTERNAL_VECTOR; if (i != SYSCALL_VECTOR) set_intr_gate(i, interrupt[i - FIRST_EXTERNAL_VECTOR]); } - /* This call is required to set up for 4k stacks, where we have - * separate stacks for hard and soft interrupts. */ + + /* + * This call is required to set up for 4k stacks, where we have + * separate stacks for hard and soft interrupts. + */ irq_ctx_init(smp_processor_id()); } @@ -729,31 +802,39 @@ static unsigned long lguest_get_wallclock(void) return lguest_data.time.tv_sec; } -/* The TSC is an Intel thing called the Time Stamp Counter. The Host tells us +/* + * The TSC is an Intel thing called the Time Stamp Counter. The Host tells us * what speed it runs at, or 0 if it's unusable as a reliable clock source. * This matches what we want here: if we return 0 from this function, the x86 - * TSC clock will give up and not register itself. */ + * TSC clock will give up and not register itself. + */ static unsigned long lguest_tsc_khz(void) { return lguest_data.tsc_khz; } -/* If we can't use the TSC, the kernel falls back to our lower-priority - * "lguest_clock", where we read the time value given to us by the Host. */ +/* + * If we can't use the TSC, the kernel falls back to our lower-priority + * "lguest_clock", where we read the time value given to us by the Host. + */ static cycle_t lguest_clock_read(struct clocksource *cs) { unsigned long sec, nsec; - /* Since the time is in two parts (seconds and nanoseconds), we risk + /* + * Since the time is in two parts (seconds and nanoseconds), we risk * reading it just as it's changing from 99 & 0.999999999 to 100 and 0, * and getting 99 and 0. As Linux tends to come apart under the stress - * of time travel, we must be careful: */ + * of time travel, we must be careful: + */ do { /* First we read the seconds part. */ sec = lguest_data.time.tv_sec; - /* This read memory barrier tells the compiler and the CPU that + /* + * This read memory barrier tells the compiler and the CPU that * this can't be reordered: we have to complete the above - * before going on. */ + * before going on. + */ rmb(); /* Now we read the nanoseconds part. */ nsec = lguest_data.time.tv_nsec; @@ -777,9 +858,11 @@ static struct clocksource lguest_clock = { .flags = CLOCK_SOURCE_IS_CONTINUOUS, }; -/* We also need a "struct clock_event_device": Linux asks us to set it to go +/* + * We also need a "struct clock_event_device": Linux asks us to set it to go * off some time in the future. Actually, James Morris figured all this out, I - * just applied the patch. */ + * just applied the patch. + */ static int lguest_clockevent_set_next_event(unsigned long delta, struct clock_event_device *evt) { @@ -829,8 +912,10 @@ static struct clock_event_device lguest_clockevent = { .max_delta_ns = LG_CLOCK_MAX_DELTA, }; -/* This is the Guest timer interrupt handler (hardware interrupt 0). We just - * call the clockevent infrastructure and it does whatever needs doing. */ +/* + * This is the Guest timer interrupt handler (hardware interrupt 0). We just + * call the clockevent infrastructure and it does whatever needs doing. + */ static void lguest_time_irq(unsigned int irq, struct irq_desc *desc) { unsigned long flags; @@ -841,10 +926,12 @@ static void lguest_time_irq(unsigned int irq, struct irq_desc *desc) local_irq_restore(flags); } -/* At some point in the boot process, we get asked to set up our timing +/* + * At some point in the boot process, we get asked to set up our timing * infrastructure. The kernel doesn't expect timer interrupts before this, but * we cleverly initialized the "blocked_interrupts" field of "struct - * lguest_data" so that timer interrupts were blocked until now. */ + * lguest_data" so that timer interrupts were blocked until now. + */ static void lguest_time_init(void) { /* Set up the timer interrupt (0) to go to our simple timer routine */ @@ -868,14 +955,16 @@ static void lguest_time_init(void) * to work. They're pretty simple. */ -/* The Guest needs to tell the Host what stack it expects traps to use. For +/* + * The Guest needs to tell the Host what stack it expects traps to use. For * native hardware, this is part of the Task State Segment mentioned above in * lguest_load_tr_desc(), but to help hypervisors there's this special call. * * We tell the Host the segment we want to use (__KERNEL_DS is the kernel data * segment), the privilege level (we're privilege level 1, the Host is 0 and * will not tolerate us trying to use that), the stack pointer, and the number - * of pages in the stack. */ + * of pages in the stack. + */ static void lguest_load_sp0(struct tss_struct *tss, struct thread_struct *thread) { @@ -889,7 +978,8 @@ static void lguest_set_debugreg(int regno, unsigned long value) /* FIXME: Implement */ } -/* There are times when the kernel wants to make sure that no memory writes are +/* + * There are times when the kernel wants to make sure that no memory writes are * caught in the cache (that they've all reached real hardware devices). This * doesn't matter for the Guest which has virtual hardware. * @@ -903,11 +993,13 @@ static void lguest_wbinvd(void) { } -/* If the Guest expects to have an Advanced Programmable Interrupt Controller, +/* + * If the Guest expects to have an Advanced Programmable Interrupt Controller, * we play dumb by ignoring writes and returning 0 for reads. So it's no * longer Programmable nor Controlling anything, and I don't think 8 lines of * code qualifies for Advanced. It will also never interrupt anything. It - * does, however, allow us to get through the Linux boot code. */ + * does, however, allow us to get through the Linux boot code. + */ #ifdef CONFIG_X86_LOCAL_APIC static void lguest_apic_write(u32 reg, u32 v) { @@ -956,11 +1048,13 @@ static void lguest_safe_halt(void) kvm_hypercall0(LHCALL_HALT); } -/* The SHUTDOWN hypercall takes a string to describe what's happening, and +/* + * The SHUTDOWN hypercall takes a string to describe what's happening, and * an argument which says whether this to restart (reboot) the Guest or not. * * Note that the Host always prefers that the Guest speak in physical addresses - * rather than virtual addresses, so we use __pa() here. */ + * rather than virtual addresses, so we use __pa() here. + */ static void lguest_power_off(void) { kvm_hypercall2(LHCALL_SHUTDOWN, __pa("Power down"), @@ -991,8 +1085,10 @@ static __init char *lguest_memory_setup(void) * nice to move it back to lguest_init. Patch welcome... */ atomic_notifier_chain_register(&panic_notifier_list, &paniced); - /* The Linux bootloader header contains an "e820" memory map: the - * Launcher populated the first entry with our memory limit. */ + /* + *The Linux bootloader header contains an "e820" memory map: the + * Launcher populated the first entry with our memory limit. + */ e820_add_region(boot_params.e820_map[0].addr, boot_params.e820_map[0].size, boot_params.e820_map[0].type); @@ -1001,16 +1097,17 @@ static __init char *lguest_memory_setup(void) return "LGUEST"; } -/* We will eventually use the virtio console device to produce console output, +/* + * We will eventually use the virtio console device to produce console output, * but before that is set up we use LHCALL_NOTIFY on normal memory to produce - * console output. */ + * console output. + */ static __init int early_put_chars(u32 vtermno, const char *buf, int count) { char scratch[17]; unsigned int len = count; - /* We use a nul-terminated string, so we have to make a copy. Icky, - * huh? */ + /* We use a nul-terminated string, so we make a copy. Icky, huh? */ if (len > sizeof(scratch) - 1) len = sizeof(scratch) - 1; scratch[len] = '\0'; @@ -1021,8 +1118,10 @@ static __init int early_put_chars(u32 vtermno, const char *buf, int count) return len; } -/* Rebooting also tells the Host we're finished, but the RESTART flag tells the - * Launcher to reboot us. */ +/* + * Rebooting also tells the Host we're finished, but the RESTART flag tells the + * Launcher to reboot us. + */ static void lguest_restart(char *reason) { kvm_hypercall2(LHCALL_SHUTDOWN, __pa(reason), LGUEST_SHUTDOWN_RESTART); @@ -1049,7 +1148,8 @@ static void lguest_restart(char *reason) * fit comfortably. * * First we need assembly templates of each of the patchable Guest operations, - * and these are in i386_head.S. */ + * and these are in i386_head.S. + */ /*G:060 We construct a table from the assembler templates: */ static const struct lguest_insns @@ -1060,9 +1160,11 @@ static const struct lguest_insns [PARAVIRT_PATCH(pv_irq_ops.save_fl)] = { lgstart_pushf, lgend_pushf }, }; -/* Now our patch routine is fairly simple (based on the native one in +/* + * Now our patch routine is fairly simple (based on the native one in * paravirt.c). If we have a replacement, we copy it in and return how much of - * the available space we used. */ + * the available space we used. + */ static unsigned lguest_patch(u8 type, u16 clobber, void *ibuf, unsigned long addr, unsigned len) { @@ -1074,8 +1176,7 @@ static unsigned lguest_patch(u8 type, u16 clobber, void *ibuf, insn_len = lguest_insns[type].end - lguest_insns[type].start; - /* Similarly if we can't fit replacement (shouldn't happen, but let's - * be thorough). */ + /* Similarly if it can't fit (doesn't happen, but let's be thorough). */ if (len < insn_len) return paravirt_patch_default(type, clobber, ibuf, addr, len); @@ -1084,22 +1185,28 @@ static unsigned lguest_patch(u8 type, u16 clobber, void *ibuf, return insn_len; } -/*G:029 Once we get to lguest_init(), we know we're a Guest. The various +/*G:029 + * Once we get to lguest_init(), we know we're a Guest. The various * pv_ops structures in the kernel provide points for (almost) every routine we - * have to override to avoid privileged instructions. */ + * have to override to avoid privileged instructions. + */ __init void lguest_init(void) { - /* We're under lguest, paravirt is enabled, and we're running at - * privilege level 1, not 0 as normal. */ + /* We're under lguest. */ pv_info.name = "lguest"; + /* Paravirt is enabled. */ pv_info.paravirt_enabled = 1; + /* We're running at privilege level 1, not 0 as normal. */ pv_info.kernel_rpl = 1; + /* Everyone except Xen runs with this set. */ pv_info.shared_kernel_pmd = 1; - /* We set up all the lguest overrides for sensitive operations. These - * are detailed with the operations themselves. */ + /* + * We set up all the lguest overrides for sensitive operations. These + * are detailed with the operations themselves. + */ - /* interrupt-related operations */ + /* Interrupt-related operations */ pv_irq_ops.init_IRQ = lguest_init_IRQ; pv_irq_ops.save_fl = PV_CALLEE_SAVE(save_fl); pv_irq_ops.restore_fl = __PV_IS_CALLEE_SAVE(lg_restore_fl); @@ -1107,11 +1214,11 @@ __init void lguest_init(void) pv_irq_ops.irq_enable = __PV_IS_CALLEE_SAVE(lg_irq_enable); pv_irq_ops.safe_halt = lguest_safe_halt; - /* init-time operations */ + /* Setup operations */ pv_init_ops.memory_setup = lguest_memory_setup; pv_init_ops.patch = lguest_patch; - /* Intercepts of various cpu instructions */ + /* Intercepts of various CPU instructions */ pv_cpu_ops.load_gdt = lguest_load_gdt; pv_cpu_ops.cpuid = lguest_cpuid; pv_cpu_ops.load_idt = lguest_load_idt; @@ -1132,7 +1239,7 @@ __init void lguest_init(void) pv_cpu_ops.start_context_switch = paravirt_start_context_switch; pv_cpu_ops.end_context_switch = lguest_end_context_switch; - /* pagetable management */ + /* Pagetable management */ pv_mmu_ops.write_cr3 = lguest_write_cr3; pv_mmu_ops.flush_tlb_user = lguest_flush_tlb_user; pv_mmu_ops.flush_tlb_single = lguest_flush_tlb_single; @@ -1154,54 +1261,71 @@ __init void lguest_init(void) pv_mmu_ops.pte_update_defer = lguest_pte_update; #ifdef CONFIG_X86_LOCAL_APIC - /* apic read/write intercepts */ + /* APIC read/write intercepts */ set_lguest_basic_apic_ops(); #endif - /* time operations */ + /* Time operations */ pv_time_ops.get_wallclock = lguest_get_wallclock; pv_time_ops.time_init = lguest_time_init; pv_time_ops.get_tsc_khz = lguest_tsc_khz; - /* Now is a good time to look at the implementations of these functions - * before returning to the rest of lguest_init(). */ + /* + * Now is a good time to look at the implementations of these functions + * before returning to the rest of lguest_init(). + */ - /*G:070 Now we've seen all the paravirt_ops, we return to + /*G:070 + * Now we've seen all the paravirt_ops, we return to * lguest_init() where the rest of the fairly chaotic boot setup - * occurs. */ + * occurs. + */ - /* The stack protector is a weird thing where gcc places a canary + /* + * The stack protector is a weird thing where gcc places a canary * value on the stack and then checks it on return. This file is * compiled with -fno-stack-protector it, so we got this far without * problems. The value of the canary is kept at offset 20 from the * %gs register, so we need to set that up before calling C functions - * in other files. */ + * in other files. + */ setup_stack_canary_segment(0); - /* We could just call load_stack_canary_segment(), but we might as - * call switch_to_new_gdt() which loads the whole table and sets up - * the per-cpu segment descriptor register %fs as well. */ + + /* + * We could just call load_stack_canary_segment(), but we might as well + * call switch_to_new_gdt() which loads the whole table and sets up the + * per-cpu segment descriptor register %fs as well. + */ switch_to_new_gdt(0); /* As described in head_32.S, we map the first 128M of memory. */ max_pfn_mapped = (128*1024*1024) >> PAGE_SHIFT; - /* The Host<->Guest Switcher lives at the top of our address space, and + /* + * The Host<->Guest Switcher lives at the top of our address space, and * the Host told us how big it is when we made LGUEST_INIT hypercall: - * it put the answer in lguest_data.reserve_mem */ + * it put the answer in lguest_data.reserve_mem + */ reserve_top_address(lguest_data.reserve_mem); - /* If we don't initialize the lock dependency checker now, it crashes - * paravirt_disable_iospace. */ + /* + * If we don't initialize the lock dependency checker now, it crashes + * paravirt_disable_iospace. + */ lockdep_init(); - /* The IDE code spends about 3 seconds probing for disks: if we reserve + /* + * The IDE code spends about 3 seconds probing for disks: if we reserve * all the I/O ports up front it can't get them and so doesn't probe. * Other device drivers are similar (but less severe). This cuts the - * kernel boot time on my machine from 4.1 seconds to 0.45 seconds. */ + * kernel boot time on my machine from 4.1 seconds to 0.45 seconds. + */ paravirt_disable_iospace(); - /* This is messy CPU setup stuff which the native boot code does before - * start_kernel, so we have to do, too: */ + /* + * This is messy CPU setup stuff which the native boot code does before + * start_kernel, so we have to do, too: + */ cpu_detect(&new_cpu_data); /* head.S usually sets up the first capability word, so do it here. */ new_cpu_data.x86_capability[0] = cpuid_edx(1); @@ -1218,22 +1342,28 @@ __init void lguest_init(void) acpi_ht = 0; #endif - /* We set the preferred console to "hvc". This is the "hypervisor + /* + * We set the preferred console to "hvc". This is the "hypervisor * virtual console" driver written by the PowerPC people, which we also - * adapted for lguest's use. */ + * adapted for lguest's use. + */ add_preferred_console("hvc", 0, NULL); /* Register our very early console. */ virtio_cons_early_init(early_put_chars); - /* Last of all, we set the power management poweroff hook to point to + /* + * Last of all, we set the power management poweroff hook to point to * the Guest routine to power off, and the reboot hook to our restart - * routine. */ + * routine. + */ pm_power_off = lguest_power_off; machine_ops.restart = lguest_restart; - /* Now we're set up, call i386_start_kernel() in head32.c and we proceed - * to boot as normal. It never returns. */ + /* + * Now we're set up, call i386_start_kernel() in head32.c and we proceed + * to boot as normal. It never returns. + */ i386_start_kernel(); } /* diff --git a/arch/x86/lguest/i386_head.S b/arch/x86/lguest/i386_head.S index a9c8cfe61cd4..db6aa95eb054 100644 --- a/arch/x86/lguest/i386_head.S +++ b/arch/x86/lguest/i386_head.S @@ -5,7 +5,8 @@ #include #include -/*G:020 Our story starts with the kernel booting into startup_32 in +/*G:020 + * Our story starts with the kernel booting into startup_32 in * arch/x86/kernel/head_32.S. It expects a boot header, which is created by * the bootloader (the Launcher in our case). * @@ -21,11 +22,14 @@ * data without remembering to subtract __PAGE_OFFSET! * * The .section line puts this code in .init.text so it will be discarded after - * boot. */ + * boot. + */ .section .init.text, "ax", @progbits ENTRY(lguest_entry) - /* We make the "initialization" hypercall now to tell the Host about - * us, and also find out where it put our page tables. */ + /* + * We make the "initialization" hypercall now to tell the Host about + * us, and also find out where it put our page tables. + */ movl $LHCALL_LGUEST_INIT, %eax movl $lguest_data - __PAGE_OFFSET, %ebx .byte 0x0f,0x01,0xc1 /* KVM_HYPERCALL */ @@ -33,13 +37,14 @@ ENTRY(lguest_entry) /* Set up the initial stack so we can run C code. */ movl $(init_thread_union+THREAD_SIZE),%esp - /* Jumps are relative, and we're running __PAGE_OFFSET too low at the - * moment. */ + /* Jumps are relative: we're running __PAGE_OFFSET too low. */ jmp lguest_init+__PAGE_OFFSET -/*G:055 We create a macro which puts the assembler code between lgstart_ and - * lgend_ markers. These templates are put in the .text section: they can't be - * discarded after boot as we may need to patch modules, too. */ +/*G:055 + * We create a macro which puts the assembler code between lgstart_ and lgend_ + * markers. These templates are put in the .text section: they can't be + * discarded after boot as we may need to patch modules, too. + */ .text #define LGUEST_PATCH(name, insns...) \ lgstart_##name: insns; lgend_##name:; \ @@ -48,58 +53,74 @@ ENTRY(lguest_entry) LGUEST_PATCH(cli, movl $0, lguest_data+LGUEST_DATA_irq_enabled) LGUEST_PATCH(pushf, movl lguest_data+LGUEST_DATA_irq_enabled, %eax) -/*G:033 But using those wrappers is inefficient (we'll see why that doesn't - * matter for save_fl and irq_disable later). If we write our routines - * carefully in assembler, we can avoid clobbering any registers and avoid - * jumping through the wrapper functions. +/*G:033 + * But using those wrappers is inefficient (we'll see why that doesn't matter + * for save_fl and irq_disable later). If we write our routines carefully in + * assembler, we can avoid clobbering any registers and avoid jumping through + * the wrapper functions. * * I skipped over our first piece of assembler, but this one is worth studying - * in a bit more detail so I'll describe in easy stages. First, the routine - * to enable interrupts: */ + * in a bit more detail so I'll describe in easy stages. First, the routine to + * enable interrupts: + */ ENTRY(lg_irq_enable) - /* The reverse of irq_disable, this sets lguest_data.irq_enabled to - * X86_EFLAGS_IF (ie. "Interrupts enabled"). */ + /* + * The reverse of irq_disable, this sets lguest_data.irq_enabled to + * X86_EFLAGS_IF (ie. "Interrupts enabled"). + */ movl $X86_EFLAGS_IF, lguest_data+LGUEST_DATA_irq_enabled - /* But now we need to check if the Host wants to know: there might have + /* + * But now we need to check if the Host wants to know: there might have * been interrupts waiting to be delivered, in which case it will have * set lguest_data.irq_pending to X86_EFLAGS_IF. If it's not zero, we - * jump to send_interrupts, otherwise we're done. */ + * jump to send_interrupts, otherwise we're done. + */ testl $0, lguest_data+LGUEST_DATA_irq_pending jnz send_interrupts - /* One cool thing about x86 is that you can do many things without using + /* + * One cool thing about x86 is that you can do many things without using * a register. In this case, the normal path hasn't needed to save or - * restore any registers at all! */ + * restore any registers at all! + */ ret send_interrupts: - /* OK, now we need a register: eax is used for the hypercall number, + /* + * OK, now we need a register: eax is used for the hypercall number, * which is LHCALL_SEND_INTERRUPTS. * * We used not to bother with this pending detection at all, which was * much simpler. Sooner or later the Host would realize it had to * send us an interrupt. But that turns out to make performance 7 * times worse on a simple tcp benchmark. So now we do this the hard - * way. */ + * way. + */ pushl %eax movl $LHCALL_SEND_INTERRUPTS, %eax - /* This is a vmcall instruction (same thing that KVM uses). Older + /* + * This is a vmcall instruction (same thing that KVM uses). Older * assembler versions might not know the "vmcall" instruction, so we - * create one manually here. */ + * create one manually here. + */ .byte 0x0f,0x01,0xc1 /* KVM_HYPERCALL */ popl %eax ret -/* Finally, the "popf" or "restore flags" routine. The %eax register holds the +/* + * Finally, the "popf" or "restore flags" routine. The %eax register holds the * flags (in practice, either X86_EFLAGS_IF or 0): if it's X86_EFLAGS_IF we're - * enabling interrupts again, if it's 0 we're leaving them off. */ + * enabling interrupts again, if it's 0 we're leaving them off. + */ ENTRY(lg_restore_fl) /* This is just "lguest_data.irq_enabled = flags;" */ movl %eax, lguest_data+LGUEST_DATA_irq_enabled - /* Now, if the %eax value has enabled interrupts and + /* + * Now, if the %eax value has enabled interrupts and * lguest_data.irq_pending is set, we want to tell the Host so it can * deliver any outstanding interrupts. Fortunately, both values will * be X86_EFLAGS_IF (ie. 512) in that case, and the "testl" * instruction will AND them together for us. If both are set, we - * jump to send_interrupts. */ + * jump to send_interrupts. + */ testl lguest_data+LGUEST_DATA_irq_pending, %eax jnz send_interrupts /* Again, the normal path has used no extra registers. Clever, huh? */ @@ -109,22 +130,24 @@ ENTRY(lg_restore_fl) .global lguest_noirq_start .global lguest_noirq_end -/*M:004 When the Host reflects a trap or injects an interrupt into the Guest, - * it sets the eflags interrupt bit on the stack based on - * lguest_data.irq_enabled, so the Guest iret logic does the right thing when - * restoring it. However, when the Host sets the Guest up for direct traps, - * such as system calls, the processor is the one to push eflags onto the - * stack, and the interrupt bit will be 1 (in reality, interrupts are always - * enabled in the Guest). +/*M:004 + * When the Host reflects a trap or injects an interrupt into the Guest, it + * sets the eflags interrupt bit on the stack based on lguest_data.irq_enabled, + * so the Guest iret logic does the right thing when restoring it. However, + * when the Host sets the Guest up for direct traps, such as system calls, the + * processor is the one to push eflags onto the stack, and the interrupt bit + * will be 1 (in reality, interrupts are always enabled in the Guest). * * This turns out to be harmless: the only trap which should happen under Linux * with interrupts disabled is Page Fault (due to our lazy mapping of vmalloc * regions), which has to be reflected through the Host anyway. If another * trap *does* go off when interrupts are disabled, the Guest will panic, and - * we'll never get to this iret! :*/ + * we'll never get to this iret! +:*/ -/*G:045 There is one final paravirt_op that the Guest implements, and glancing - * at it you can see why I left it to last. It's *cool*! It's in *assembler*! +/*G:045 + * There is one final paravirt_op that the Guest implements, and glancing at it + * you can see why I left it to last. It's *cool*! It's in *assembler*! * * The "iret" instruction is used to return from an interrupt or trap. The * stack looks like this: @@ -148,15 +171,18 @@ ENTRY(lg_restore_fl) * return to userspace or wherever. Our solution to this is to surround the * code with lguest_noirq_start: and lguest_noirq_end: labels. We tell the * Host that it is *never* to interrupt us there, even if interrupts seem to be - * enabled. */ + * enabled. + */ ENTRY(lguest_iret) pushl %eax movl 12(%esp), %eax lguest_noirq_start: - /* Note the %ss: segment prefix here. Normal data accesses use the + /* + * Note the %ss: segment prefix here. Normal data accesses use the * "ds" segment, but that will have already been restored for whatever * we're returning to (such as userspace): we can't trust it. The %ss: - * prefix makes sure we use the stack segment, which is still valid. */ + * prefix makes sure we use the stack segment, which is still valid. + */ movl %eax,%ss:lguest_data+LGUEST_DATA_irq_enabled popl %eax iret diff --git a/drivers/lguest/core.c b/drivers/lguest/core.c index a6974e9b8ebf..cd058bc903ff 100644 --- a/drivers/lguest/core.c +++ b/drivers/lguest/core.c @@ -1,6 +1,8 @@ -/*P:400 This contains run_guest() which actually calls into the Host<->Guest +/*P:400 + * This contains run_guest() which actually calls into the Host<->Guest * Switcher and analyzes the return, such as determining if the Guest wants the - * Host to do something. This file also contains useful helper routines. :*/ + * Host to do something. This file also contains useful helper routines. +:*/ #include #include #include @@ -24,7 +26,8 @@ static struct page **switcher_page; /* This One Big lock protects all inter-guest data structures. */ DEFINE_MUTEX(lguest_lock); -/*H:010 We need to set up the Switcher at a high virtual address. Remember the +/*H:010 + * We need to set up the Switcher at a high virtual address. Remember the * Switcher is a few hundred bytes of assembler code which actually changes the * CPU to run the Guest, and then changes back to the Host when a trap or * interrupt happens. @@ -33,7 +36,8 @@ DEFINE_MUTEX(lguest_lock); * Host since it will be running as the switchover occurs. * * Trying to map memory at a particular address is an unusual thing to do, so - * it's not a simple one-liner. */ + * it's not a simple one-liner. + */ static __init int map_switcher(void) { int i, err; @@ -47,8 +51,10 @@ static __init int map_switcher(void) * easy. */ - /* We allocate an array of struct page pointers. map_vm_area() wants - * this, rather than just an array of pages. */ + /* + * We allocate an array of struct page pointers. map_vm_area() wants + * this, rather than just an array of pages. + */ switcher_page = kmalloc(sizeof(switcher_page[0])*TOTAL_SWITCHER_PAGES, GFP_KERNEL); if (!switcher_page) { @@ -56,8 +62,10 @@ static __init int map_switcher(void) goto out; } - /* Now we actually allocate the pages. The Guest will see these pages, - * so we make sure they're zeroed. */ + /* + * Now we actually allocate the pages. The Guest will see these pages, + * so we make sure they're zeroed. + */ for (i = 0; i < TOTAL_SWITCHER_PAGES; i++) { unsigned long addr = get_zeroed_page(GFP_KERNEL); if (!addr) { @@ -67,19 +75,23 @@ static __init int map_switcher(void) switcher_page[i] = virt_to_page(addr); } - /* First we check that the Switcher won't overlap the fixmap area at + /* + * First we check that the Switcher won't overlap the fixmap area at * the top of memory. It's currently nowhere near, but it could have - * very strange effects if it ever happened. */ + * very strange effects if it ever happened. + */ if (SWITCHER_ADDR + (TOTAL_SWITCHER_PAGES+1)*PAGE_SIZE > FIXADDR_START){ err = -ENOMEM; printk("lguest: mapping switcher would thwack fixmap\n"); goto free_pages; } - /* Now we reserve the "virtual memory area" we want: 0xFFC00000 + /* + * Now we reserve the "virtual memory area" we want: 0xFFC00000 * (SWITCHER_ADDR). We might not get it in theory, but in practice * it's worked so far. The end address needs +1 because __get_vm_area - * allocates an extra guard page, so we need space for that. */ + * allocates an extra guard page, so we need space for that. + */ switcher_vma = __get_vm_area(TOTAL_SWITCHER_PAGES * PAGE_SIZE, VM_ALLOC, SWITCHER_ADDR, SWITCHER_ADDR + (TOTAL_SWITCHER_PAGES+1) * PAGE_SIZE); @@ -89,11 +101,13 @@ static __init int map_switcher(void) goto free_pages; } - /* This code actually sets up the pages we've allocated to appear at + /* + * This code actually sets up the pages we've allocated to appear at * SWITCHER_ADDR. map_vm_area() takes the vma we allocated above, the * kind of pages we're mapping (kernel pages), and a pointer to our * array of struct pages. It increments that pointer, but we don't - * care. */ + * care. + */ pagep = switcher_page; err = map_vm_area(switcher_vma, PAGE_KERNEL_EXEC, &pagep); if (err) { @@ -101,8 +115,10 @@ static __init int map_switcher(void) goto free_vma; } - /* Now the Switcher is mapped at the right address, we can't fail! - * Copy in the compiled-in Switcher code (from _switcher.S). */ + /* + * Now the Switcher is mapped at the right address, we can't fail! + * Copy in the compiled-in Switcher code (from _switcher.S). + */ memcpy(switcher_vma->addr, start_switcher_text, end_switcher_text - start_switcher_text); @@ -124,8 +140,7 @@ out: } /*:*/ -/* Cleaning up the mapping when the module is unloaded is almost... - * too easy. */ +/* Cleaning up the mapping when the module is unloaded is almost... too easy. */ static void unmap_switcher(void) { unsigned int i; @@ -151,16 +166,19 @@ static void unmap_switcher(void) * But we can't trust the Guest: it might be trying to access the Launcher * code. We have to check that the range is below the pfn_limit the Launcher * gave us. We have to make sure that addr + len doesn't give us a false - * positive by overflowing, too. */ + * positive by overflowing, too. + */ bool lguest_address_ok(const struct lguest *lg, unsigned long addr, unsigned long len) { return (addr+len) / PAGE_SIZE < lg->pfn_limit && (addr+len >= addr); } -/* This routine copies memory from the Guest. Here we can see how useful the +/* + * This routine copies memory from the Guest. Here we can see how useful the * kill_lguest() routine we met in the Launcher can be: we return a random - * value (all zeroes) instead of needing to return an error. */ + * value (all zeroes) instead of needing to return an error. + */ void __lgread(struct lg_cpu *cpu, void *b, unsigned long addr, unsigned bytes) { if (!lguest_address_ok(cpu->lg, addr, bytes) @@ -181,9 +199,11 @@ void __lgwrite(struct lg_cpu *cpu, unsigned long addr, const void *b, } /*:*/ -/*H:030 Let's jump straight to the the main loop which runs the Guest. +/*H:030 + * Let's jump straight to the the main loop which runs the Guest. * Remember, this is called by the Launcher reading /dev/lguest, and we keep - * going around and around until something interesting happens. */ + * going around and around until something interesting happens. + */ int run_guest(struct lg_cpu *cpu, unsigned long __user *user) { /* We stop running once the Guest is dead. */ @@ -195,8 +215,10 @@ int run_guest(struct lg_cpu *cpu, unsigned long __user *user) if (cpu->hcall) do_hypercalls(cpu); - /* It's possible the Guest did a NOTIFY hypercall to the - * Launcher, in which case we return from the read() now. */ + /* + * It's possible the Guest did a NOTIFY hypercall to the + * Launcher, in which case we return from the read() now. + */ if (cpu->pending_notify) { if (!send_notify_to_eventfd(cpu)) { if (put_user(cpu->pending_notify, user)) @@ -209,29 +231,39 @@ int run_guest(struct lg_cpu *cpu, unsigned long __user *user) if (signal_pending(current)) return -ERESTARTSYS; - /* Check if there are any interrupts which can be delivered now: + /* + * Check if there are any interrupts which can be delivered now: * if so, this sets up the hander to be executed when we next - * run the Guest. */ + * run the Guest. + */ irq = interrupt_pending(cpu, &more); if (irq < LGUEST_IRQS) try_deliver_interrupt(cpu, irq, more); - /* All long-lived kernel loops need to check with this horrible + /* + * All long-lived kernel loops need to check with this horrible * thing called the freezer. If the Host is trying to suspend, - * it stops us. */ + * it stops us. + */ try_to_freeze(); - /* Just make absolutely sure the Guest is still alive. One of - * those hypercalls could have been fatal, for example. */ + /* + * Just make absolutely sure the Guest is still alive. One of + * those hypercalls could have been fatal, for example. + */ if (cpu->lg->dead) break; - /* If the Guest asked to be stopped, we sleep. The Guest's - * clock timer will wake us. */ + /* + * If the Guest asked to be stopped, we sleep. The Guest's + * clock timer will wake us. + */ if (cpu->halted) { set_current_state(TASK_INTERRUPTIBLE); - /* Just before we sleep, make sure no interrupt snuck in - * which we should be doing. */ + /* + * Just before we sleep, make sure no interrupt snuck in + * which we should be doing. + */ if (interrupt_pending(cpu, &more) < LGUEST_IRQS) set_current_state(TASK_RUNNING); else @@ -239,8 +271,10 @@ int run_guest(struct lg_cpu *cpu, unsigned long __user *user) continue; } - /* OK, now we're ready to jump into the Guest. First we put up - * the "Do Not Disturb" sign: */ + /* + * OK, now we're ready to jump into the Guest. First we put up + * the "Do Not Disturb" sign: + */ local_irq_disable(); /* Actually run the Guest until something happens. */ @@ -327,8 +361,10 @@ static void __exit fini(void) } /*:*/ -/* The Host side of lguest can be a module. This is a nice way for people to - * play with it. */ +/* + * The Host side of lguest can be a module. This is a nice way for people to + * play with it. + */ module_init(init); module_exit(fini); MODULE_LICENSE("GPL"); diff --git a/drivers/lguest/hypercalls.c b/drivers/lguest/hypercalls.c index c29ffa19cb74..787ab4bc09f0 100644 --- a/drivers/lguest/hypercalls.c +++ b/drivers/lguest/hypercalls.c @@ -1,8 +1,10 @@ -/*P:500 Just as userspace programs request kernel operations through a system +/*P:500 + * Just as userspace programs request kernel operations through a system * call, the Guest requests Host operations through a "hypercall". You might * notice this nomenclature doesn't really follow any logic, but the name has * been around for long enough that we're stuck with it. As you'd expect, this - * code is basically a one big switch statement. :*/ + * code is basically a one big switch statement. +:*/ /* Copyright (C) 2006 Rusty Russell IBM Corporation @@ -28,30 +30,41 @@ #include #include "lg.h" -/*H:120 This is the core hypercall routine: where the Guest gets what it wants. - * Or gets killed. Or, in the case of LHCALL_SHUTDOWN, both. */ +/*H:120 + * This is the core hypercall routine: where the Guest gets what it wants. + * Or gets killed. Or, in the case of LHCALL_SHUTDOWN, both. + */ static void do_hcall(struct lg_cpu *cpu, struct hcall_args *args) { switch (args->arg0) { case LHCALL_FLUSH_ASYNC: - /* This call does nothing, except by breaking out of the Guest - * it makes us process all the asynchronous hypercalls. */ + /* + * This call does nothing, except by breaking out of the Guest + * it makes us process all the asynchronous hypercalls. + */ break; case LHCALL_SEND_INTERRUPTS: - /* This call does nothing too, but by breaking out of the Guest - * it makes us process any pending interrupts. */ + /* + * This call does nothing too, but by breaking out of the Guest + * it makes us process any pending interrupts. + */ break; case LHCALL_LGUEST_INIT: - /* You can't get here unless you're already initialized. Don't - * do that. */ + /* + * You can't get here unless you're already initialized. Don't + * do that. + */ kill_guest(cpu, "already have lguest_data"); break; case LHCALL_SHUTDOWN: { - /* Shutdown is such a trivial hypercall that we do it in four - * lines right here. */ char msg[128]; - /* If the lgread fails, it will call kill_guest() itself; the - * kill_guest() with the message will be ignored. */ + /* + * Shutdown is such a trivial hypercall that we do it in four + * lines right here. + * + * If the lgread fails, it will call kill_guest() itself; the + * kill_guest() with the message will be ignored. + */ __lgread(cpu, msg, args->arg1, sizeof(msg)); msg[sizeof(msg)-1] = '\0'; kill_guest(cpu, "CRASH: %s", msg); @@ -60,16 +73,17 @@ static void do_hcall(struct lg_cpu *cpu, struct hcall_args *args) break; } case LHCALL_FLUSH_TLB: - /* FLUSH_TLB comes in two flavors, depending on the - * argument: */ + /* FLUSH_TLB comes in two flavors, depending on the argument: */ if (args->arg1) guest_pagetable_clear_all(cpu); else guest_pagetable_flush_user(cpu); break; - /* All these calls simply pass the arguments through to the right - * routines. */ + /* + * All these calls simply pass the arguments through to the right + * routines. + */ case LHCALL_NEW_PGTABLE: guest_new_pagetable(cpu, args->arg1); break; @@ -112,15 +126,16 @@ static void do_hcall(struct lg_cpu *cpu, struct hcall_args *args) kill_guest(cpu, "Bad hypercall %li\n", args->arg0); } } -/*:*/ -/*H:124 Asynchronous hypercalls are easy: we just look in the array in the +/*H:124 + * Asynchronous hypercalls are easy: we just look in the array in the * Guest's "struct lguest_data" to see if any new ones are marked "ready". * * We are careful to do these in order: obviously we respect the order the * Guest put them in the ring, but we also promise the Guest that they will * happen before any normal hypercall (which is why we check this before - * checking for a normal hcall). */ + * checking for a normal hcall). + */ static void do_async_hcalls(struct lg_cpu *cpu) { unsigned int i; @@ -133,22 +148,28 @@ static void do_async_hcalls(struct lg_cpu *cpu) /* We process "struct lguest_data"s hcalls[] ring once. */ for (i = 0; i < ARRAY_SIZE(st); i++) { struct hcall_args args; - /* We remember where we were up to from last time. This makes + /* + * We remember where we were up to from last time. This makes * sure that the hypercalls are done in the order the Guest - * places them in the ring. */ + * places them in the ring. + */ unsigned int n = cpu->next_hcall; /* 0xFF means there's no call here (yet). */ if (st[n] == 0xFF) break; - /* OK, we have hypercall. Increment the "next_hcall" cursor, - * and wrap back to 0 if we reach the end. */ + /* + * OK, we have hypercall. Increment the "next_hcall" cursor, + * and wrap back to 0 if we reach the end. + */ if (++cpu->next_hcall == LHCALL_RING_SIZE) cpu->next_hcall = 0; - /* Copy the hypercall arguments into a local copy of - * the hcall_args struct. */ + /* + * Copy the hypercall arguments into a local copy of the + * hcall_args struct. + */ if (copy_from_user(&args, &cpu->lg->lguest_data->hcalls[n], sizeof(struct hcall_args))) { kill_guest(cpu, "Fetching async hypercalls"); @@ -164,19 +185,25 @@ static void do_async_hcalls(struct lg_cpu *cpu) break; } - /* Stop doing hypercalls if they want to notify the Launcher: - * it needs to service this first. */ + /* + * Stop doing hypercalls if they want to notify the Launcher: + * it needs to service this first. + */ if (cpu->pending_notify) break; } } -/* Last of all, we look at what happens first of all. The very first time the - * Guest makes a hypercall, we end up here to set things up: */ +/* + * Last of all, we look at what happens first of all. The very first time the + * Guest makes a hypercall, we end up here to set things up: + */ static void initialize(struct lg_cpu *cpu) { - /* You can't do anything until you're initialized. The Guest knows the - * rules, so we're unforgiving here. */ + /* + * You can't do anything until you're initialized. The Guest knows the + * rules, so we're unforgiving here. + */ if (cpu->hcall->arg0 != LHCALL_LGUEST_INIT) { kill_guest(cpu, "hypercall %li before INIT", cpu->hcall->arg0); return; @@ -185,32 +212,40 @@ static void initialize(struct lg_cpu *cpu) if (lguest_arch_init_hypercalls(cpu)) kill_guest(cpu, "bad guest page %p", cpu->lg->lguest_data); - /* The Guest tells us where we're not to deliver interrupts by putting - * the range of addresses into "struct lguest_data". */ + /* + * The Guest tells us where we're not to deliver interrupts by putting + * the range of addresses into "struct lguest_data". + */ if (get_user(cpu->lg->noirq_start, &cpu->lg->lguest_data->noirq_start) || get_user(cpu->lg->noirq_end, &cpu->lg->lguest_data->noirq_end)) kill_guest(cpu, "bad guest page %p", cpu->lg->lguest_data); - /* We write the current time into the Guest's data page once so it can - * set its clock. */ + /* + * We write the current time into the Guest's data page once so it can + * set its clock. + */ write_timestamp(cpu); /* page_tables.c will also do some setup. */ page_table_guest_data_init(cpu); - /* This is the one case where the above accesses might have been the + /* + * This is the one case where the above accesses might have been the * first write to a Guest page. This may have caused a copy-on-write * fault, but the old page might be (read-only) in the Guest - * pagetable. */ + * pagetable. + */ guest_pagetable_clear_all(cpu); } /*:*/ -/*M:013 If a Guest reads from a page (so creates a mapping) that it has never +/*M:013 + * If a Guest reads from a page (so creates a mapping) that it has never * written to, and then the Launcher writes to it (ie. the output of a virtual * device), the Guest will still see the old page. In practice, this never * happens: why would the Guest read a page which it has never written to? But - * a similar scenario might one day bite us, so it's worth mentioning. :*/ + * a similar scenario might one day bite us, so it's worth mentioning. +:*/ /*H:100 * Hypercalls @@ -229,17 +264,22 @@ void do_hypercalls(struct lg_cpu *cpu) return; } - /* The Guest has initialized. + /* + * The Guest has initialized. * - * Look in the hypercall ring for the async hypercalls: */ + * Look in the hypercall ring for the async hypercalls: + */ do_async_hcalls(cpu); - /* If we stopped reading the hypercall ring because the Guest did a + /* + * If we stopped reading the hypercall ring because the Guest did a * NOTIFY to the Launcher, we want to return now. Otherwise we do - * the hypercall. */ + * the hypercall. + */ if (!cpu->pending_notify) { do_hcall(cpu, cpu->hcall); - /* Tricky point: we reset the hcall pointer to mark the + /* + * Tricky point: we reset the hcall pointer to mark the * hypercall as "done". We use the hcall pointer rather than * the trap number to indicate a hypercall is pending. * Normally it doesn't matter: the Guest will run again and @@ -248,13 +288,16 @@ void do_hypercalls(struct lg_cpu *cpu) * However, if we are signalled or the Guest sends I/O to the * Launcher, the run_guest() loop will exit without running the * Guest. When it comes back it would try to re-run the - * hypercall. Finding that bug sucked. */ + * hypercall. Finding that bug sucked. + */ cpu->hcall = NULL; } } -/* This routine supplies the Guest with time: it's used for wallclock time at - * initial boot and as a rough time source if the TSC isn't available. */ +/* + * This routine supplies the Guest with time: it's used for wallclock time at + * initial boot and as a rough time source if the TSC isn't available. + */ void write_timestamp(struct lg_cpu *cpu) { struct timespec now; diff --git a/drivers/lguest/interrupts_and_traps.c b/drivers/lguest/interrupts_and_traps.c index 0e9067b0d507..18648180db02 100644 --- a/drivers/lguest/interrupts_and_traps.c +++ b/drivers/lguest/interrupts_and_traps.c @@ -1,4 +1,5 @@ -/*P:800 Interrupts (traps) are complicated enough to earn their own file. +/*P:800 + * Interrupts (traps) are complicated enough to earn their own file. * There are three classes of interrupts: * * 1) Real hardware interrupts which occur while we're running the Guest, @@ -10,7 +11,8 @@ * just like real hardware would deliver them. Traps from the Guest can be set * up to go directly back into the Guest, but sometimes the Host wants to see * them first, so we also have a way of "reflecting" them into the Guest as if - * they had been delivered to it directly. :*/ + * they had been delivered to it directly. +:*/ #include #include #include @@ -26,8 +28,10 @@ static unsigned long idt_address(u32 lo, u32 hi) return (lo & 0x0000FFFF) | (hi & 0xFFFF0000); } -/* The "type" of the interrupt handler is a 4 bit field: we only support a - * couple of types. */ +/* + * The "type" of the interrupt handler is a 4 bit field: we only support a + * couple of types. + */ static int idt_type(u32 lo, u32 hi) { return (hi >> 8) & 0xF; @@ -39,8 +43,10 @@ static bool idt_present(u32 lo, u32 hi) return (hi & 0x8000); } -/* We need a helper to "push" a value onto the Guest's stack, since that's a - * big part of what delivering an interrupt does. */ +/* + * We need a helper to "push" a value onto the Guest's stack, since that's a + * big part of what delivering an interrupt does. + */ static void push_guest_stack(struct lg_cpu *cpu, unsigned long *gstack, u32 val) { /* Stack grows upwards: move stack then write value. */ @@ -48,7 +54,8 @@ static void push_guest_stack(struct lg_cpu *cpu, unsigned long *gstack, u32 val) lgwrite(cpu, *gstack, u32, val); } -/*H:210 The set_guest_interrupt() routine actually delivers the interrupt or +/*H:210 + * The set_guest_interrupt() routine actually delivers the interrupt or * trap. The mechanics of delivering traps and interrupts to the Guest are the * same, except some traps have an "error code" which gets pushed onto the * stack as well: the caller tells us if this is one. @@ -59,7 +66,8 @@ static void push_guest_stack(struct lg_cpu *cpu, unsigned long *gstack, u32 val) * * We set up the stack just like the CPU does for a real interrupt, so it's * identical for the Guest (and the standard "iret" instruction will undo - * it). */ + * it). + */ static void set_guest_interrupt(struct lg_cpu *cpu, u32 lo, u32 hi, bool has_err) { @@ -67,20 +75,26 @@ static void set_guest_interrupt(struct lg_cpu *cpu, u32 lo, u32 hi, u32 eflags, ss, irq_enable; unsigned long virtstack; - /* There are two cases for interrupts: one where the Guest is already + /* + * There are two cases for interrupts: one where the Guest is already * in the kernel, and a more complex one where the Guest is in - * userspace. We check the privilege level to find out. */ + * userspace. We check the privilege level to find out. + */ if ((cpu->regs->ss&0x3) != GUEST_PL) { - /* The Guest told us their kernel stack with the SET_STACK - * hypercall: both the virtual address and the segment */ + /* + * The Guest told us their kernel stack with the SET_STACK + * hypercall: both the virtual address and the segment. + */ virtstack = cpu->esp1; ss = cpu->ss1; origstack = gstack = guest_pa(cpu, virtstack); - /* We push the old stack segment and pointer onto the new + /* + * We push the old stack segment and pointer onto the new * stack: when the Guest does an "iret" back from the interrupt * handler the CPU will notice they're dropping privilege - * levels and expect these here. */ + * levels and expect these here. + */ push_guest_stack(cpu, &gstack, cpu->regs->ss); push_guest_stack(cpu, &gstack, cpu->regs->esp); } else { @@ -91,18 +105,22 @@ static void set_guest_interrupt(struct lg_cpu *cpu, u32 lo, u32 hi, origstack = gstack = guest_pa(cpu, virtstack); } - /* Remember that we never let the Guest actually disable interrupts, so + /* + * Remember that we never let the Guest actually disable interrupts, so * the "Interrupt Flag" bit is always set. We copy that bit from the * Guest's "irq_enabled" field into the eflags word: we saw the Guest - * copy it back in "lguest_iret". */ + * copy it back in "lguest_iret". + */ eflags = cpu->regs->eflags; if (get_user(irq_enable, &cpu->lg->lguest_data->irq_enabled) == 0 && !(irq_enable & X86_EFLAGS_IF)) eflags &= ~X86_EFLAGS_IF; - /* An interrupt is expected to push three things on the stack: the old + /* + * An interrupt is expected to push three things on the stack: the old * "eflags" word, the old code segment, and the old instruction - * pointer. */ + * pointer. + */ push_guest_stack(cpu, &gstack, eflags); push_guest_stack(cpu, &gstack, cpu->regs->cs); push_guest_stack(cpu, &gstack, cpu->regs->eip); @@ -111,15 +129,19 @@ static void set_guest_interrupt(struct lg_cpu *cpu, u32 lo, u32 hi, if (has_err) push_guest_stack(cpu, &gstack, cpu->regs->errcode); - /* Now we've pushed all the old state, we change the stack, the code - * segment and the address to execute. */ + /* + * Now we've pushed all the old state, we change the stack, the code + * segment and the address to execute. + */ cpu->regs->ss = ss; cpu->regs->esp = virtstack + (gstack - origstack); cpu->regs->cs = (__KERNEL_CS|GUEST_PL); cpu->regs->eip = idt_address(lo, hi); - /* There are two kinds of interrupt handlers: 0xE is an "interrupt - * gate" which expects interrupts to be disabled on entry. */ + /* + * There are two kinds of interrupt handlers: 0xE is an "interrupt + * gate" which expects interrupts to be disabled on entry. + */ if (idt_type(lo, hi) == 0xE) if (put_user(0, &cpu->lg->lguest_data->irq_enabled)) kill_guest(cpu, "Disabling interrupts"); @@ -130,7 +152,8 @@ static void set_guest_interrupt(struct lg_cpu *cpu, u32 lo, u32 hi, * * interrupt_pending() returns the first pending interrupt which isn't blocked * by the Guest. It is called before every entry to the Guest, and just before - * we go to sleep when the Guest has halted itself. */ + * we go to sleep when the Guest has halted itself. + */ unsigned int interrupt_pending(struct lg_cpu *cpu, bool *more) { unsigned int irq; @@ -140,8 +163,10 @@ unsigned int interrupt_pending(struct lg_cpu *cpu, bool *more) if (!cpu->lg->lguest_data) return LGUEST_IRQS; - /* Take our "irqs_pending" array and remove any interrupts the Guest - * wants blocked: the result ends up in "blk". */ + /* + * Take our "irqs_pending" array and remove any interrupts the Guest + * wants blocked: the result ends up in "blk". + */ if (copy_from_user(&blk, cpu->lg->lguest_data->blocked_interrupts, sizeof(blk))) return LGUEST_IRQS; @@ -154,16 +179,20 @@ unsigned int interrupt_pending(struct lg_cpu *cpu, bool *more) return irq; } -/* This actually diverts the Guest to running an interrupt handler, once an - * interrupt has been identified by interrupt_pending(). */ +/* + * This actually diverts the Guest to running an interrupt handler, once an + * interrupt has been identified by interrupt_pending(). + */ void try_deliver_interrupt(struct lg_cpu *cpu, unsigned int irq, bool more) { struct desc_struct *idt; BUG_ON(irq >= LGUEST_IRQS); - /* They may be in the middle of an iret, where they asked us never to - * deliver interrupts. */ + /* + * They may be in the middle of an iret, where they asked us never to + * deliver interrupts. + */ if (cpu->regs->eip >= cpu->lg->noirq_start && (cpu->regs->eip < cpu->lg->noirq_end)) return; @@ -187,29 +216,37 @@ void try_deliver_interrupt(struct lg_cpu *cpu, unsigned int irq, bool more) } } - /* Look at the IDT entry the Guest gave us for this interrupt. The + /* + * Look at the IDT entry the Guest gave us for this interrupt. The * first 32 (FIRST_EXTERNAL_VECTOR) entries are for traps, so we skip - * over them. */ + * over them. + */ idt = &cpu->arch.idt[FIRST_EXTERNAL_VECTOR+irq]; /* If they don't have a handler (yet?), we just ignore it */ if (idt_present(idt->a, idt->b)) { /* OK, mark it no longer pending and deliver it. */ clear_bit(irq, cpu->irqs_pending); - /* set_guest_interrupt() takes the interrupt descriptor and a + /* + * set_guest_interrupt() takes the interrupt descriptor and a * flag to say whether this interrupt pushes an error code onto - * the stack as well: virtual interrupts never do. */ + * the stack as well: virtual interrupts never do. + */ set_guest_interrupt(cpu, idt->a, idt->b, false); } - /* Every time we deliver an interrupt, we update the timestamp in the + /* + * Every time we deliver an interrupt, we update the timestamp in the * Guest's lguest_data struct. It would be better for the Guest if we * did this more often, but it can actually be quite slow: doing it * here is a compromise which means at least it gets updated every - * timer interrupt. */ + * timer interrupt. + */ write_timestamp(cpu); - /* If there are no other interrupts we want to deliver, clear - * the pending flag. */ + /* + * If there are no other interrupts we want to deliver, clear + * the pending flag. + */ if (!more) put_user(0, &cpu->lg->lguest_data->irq_pending); } @@ -217,24 +254,29 @@ void try_deliver_interrupt(struct lg_cpu *cpu, unsigned int irq, bool more) /* And this is the routine when we want to set an interrupt for the Guest. */ void set_interrupt(struct lg_cpu *cpu, unsigned int irq) { - /* Next time the Guest runs, the core code will see if it can deliver - * this interrupt. */ + /* + * Next time the Guest runs, the core code will see if it can deliver + * this interrupt. + */ set_bit(irq, cpu->irqs_pending); - /* Make sure it sees it; it might be asleep (eg. halted), or - * running the Guest right now, in which case kick_process() - * will knock it out. */ + /* + * Make sure it sees it; it might be asleep (eg. halted), or running + * the Guest right now, in which case kick_process() will knock it out. + */ if (!wake_up_process(cpu->tsk)) kick_process(cpu->tsk); } /*:*/ -/* Linux uses trap 128 for system calls. Plan9 uses 64, and Ron Minnich sent +/* + * Linux uses trap 128 for system calls. Plan9 uses 64, and Ron Minnich sent * me a patch, so we support that too. It'd be a big step for lguest if half * the Plan 9 user base were to start using it. * * Actually now I think of it, it's possible that Ron *is* half the Plan 9 - * userbase. Oh well. */ + * userbase. Oh well. + */ static bool could_be_syscall(unsigned int num) { /* Normal Linux SYSCALL_VECTOR or reserved vector? */ @@ -274,9 +316,11 @@ void free_interrupts(void) clear_bit(syscall_vector, used_vectors); } -/*H:220 Now we've got the routines to deliver interrupts, delivering traps like +/*H:220 + * Now we've got the routines to deliver interrupts, delivering traps like * page fault is easy. The only trick is that Intel decided that some traps - * should have error codes: */ + * should have error codes: + */ static bool has_err(unsigned int trap) { return (trap == 8 || (trap >= 10 && trap <= 14) || trap == 17); @@ -285,13 +329,17 @@ static bool has_err(unsigned int trap) /* deliver_trap() returns true if it could deliver the trap. */ bool deliver_trap(struct lg_cpu *cpu, unsigned int num) { - /* Trap numbers are always 8 bit, but we set an impossible trap number - * for traps inside the Switcher, so check that here. */ + /* + * Trap numbers are always 8 bit, but we set an impossible trap number + * for traps inside the Switcher, so check that here. + */ if (num >= ARRAY_SIZE(cpu->arch.idt)) return false; - /* Early on the Guest hasn't set the IDT entries (or maybe it put a - * bogus one in): if we fail here, the Guest will be killed. */ + /* + * Early on the Guest hasn't set the IDT entries (or maybe it put a + * bogus one in): if we fail here, the Guest will be killed. + */ if (!idt_present(cpu->arch.idt[num].a, cpu->arch.idt[num].b)) return false; set_guest_interrupt(cpu, cpu->arch.idt[num].a, @@ -299,7 +347,8 @@ bool deliver_trap(struct lg_cpu *cpu, unsigned int num) return true; } -/*H:250 Here's the hard part: returning to the Host every time a trap happens +/*H:250 + * Here's the hard part: returning to the Host every time a trap happens * and then calling deliver_trap() and re-entering the Guest is slow. * Particularly because Guest userspace system calls are traps (usually trap * 128). @@ -311,69 +360,87 @@ bool deliver_trap(struct lg_cpu *cpu, unsigned int num) * the other hypervisors would beat it up at lunchtime. * * This routine indicates if a particular trap number could be delivered - * directly. */ + * directly. + */ static bool direct_trap(unsigned int num) { - /* Hardware interrupts don't go to the Guest at all (except system - * call). */ + /* + * Hardware interrupts don't go to the Guest at all (except system + * call). + */ if (num >= FIRST_EXTERNAL_VECTOR && !could_be_syscall(num)) return false; - /* The Host needs to see page faults (for shadow paging and to save the + /* + * The Host needs to see page faults (for shadow paging and to save the * fault address), general protection faults (in/out emulation) and * device not available (TS handling), invalid opcode fault (kvm hcall), - * and of course, the hypercall trap. */ + * and of course, the hypercall trap. + */ return num != 14 && num != 13 && num != 7 && num != 6 && num != LGUEST_TRAP_ENTRY; } /*:*/ -/*M:005 The Guest has the ability to turn its interrupt gates into trap gates, +/*M:005 + * The Guest has the ability to turn its interrupt gates into trap gates, * if it is careful. The Host will let trap gates can go directly to the * Guest, but the Guest needs the interrupts atomically disabled for an * interrupt gate. It can do this by pointing the trap gate at instructions - * within noirq_start and noirq_end, where it can safely disable interrupts. */ + * within noirq_start and noirq_end, where it can safely disable interrupts. + */ -/*M:006 The Guests do not use the sysenter (fast system call) instruction, +/*M:006 + * The Guests do not use the sysenter (fast system call) instruction, * because it's hardcoded to enter privilege level 0 and so can't go direct. * It's about twice as fast as the older "int 0x80" system call, so it might * still be worthwhile to handle it in the Switcher and lcall down to the * Guest. The sysenter semantics are hairy tho: search for that keyword in - * entry.S :*/ + * entry.S +:*/ -/*H:260 When we make traps go directly into the Guest, we need to make sure +/*H:260 + * When we make traps go directly into the Guest, we need to make sure * the kernel stack is valid (ie. mapped in the page tables). Otherwise, the * CPU trying to deliver the trap will fault while trying to push the interrupt * words on the stack: this is called a double fault, and it forces us to kill * the Guest. * - * Which is deeply unfair, because (literally!) it wasn't the Guests' fault. */ + * Which is deeply unfair, because (literally!) it wasn't the Guests' fault. + */ void pin_stack_pages(struct lg_cpu *cpu) { unsigned int i; - /* Depending on the CONFIG_4KSTACKS option, the Guest can have one or - * two pages of stack space. */ + /* + * Depending on the CONFIG_4KSTACKS option, the Guest can have one or + * two pages of stack space. + */ for (i = 0; i < cpu->lg->stack_pages; i++) - /* The stack grows *upwards*, so the address we're given is the + /* + * The stack grows *upwards*, so the address we're given is the * start of the page after the kernel stack. Subtract one to * get back onto the first stack page, and keep subtracting to - * get to the rest of the stack pages. */ + * get to the rest of the stack pages. + */ pin_page(cpu, cpu->esp1 - 1 - i * PAGE_SIZE); } -/* Direct traps also mean that we need to know whenever the Guest wants to use +/* + * Direct traps also mean that we need to know whenever the Guest wants to use * a different kernel stack, so we can change the IDT entries to use that * stack. The IDT entries expect a virtual address, so unlike most addresses * the Guest gives us, the "esp" (stack pointer) value here is virtual, not * physical. * * In Linux each process has its own kernel stack, so this happens a lot: we - * change stacks on each context switch. */ + * change stacks on each context switch. + */ void guest_set_stack(struct lg_cpu *cpu, u32 seg, u32 esp, unsigned int pages) { - /* You are not allowed have a stack segment with privilege level 0: bad - * Guest! */ + /* + * You're not allowed a stack segment with privilege level 0: bad Guest! + */ if ((seg & 0x3) != GUEST_PL) kill_guest(cpu, "bad stack segment %i", seg); /* We only expect one or two stack pages. */ @@ -387,11 +454,15 @@ void guest_set_stack(struct lg_cpu *cpu, u32 seg, u32 esp, unsigned int pages) pin_stack_pages(cpu); } -/* All this reference to mapping stacks leads us neatly into the other complex - * part of the Host: page table handling. */ +/* + * All this reference to mapping stacks leads us neatly into the other complex + * part of the Host: page table handling. + */ -/*H:235 This is the routine which actually checks the Guest's IDT entry and - * transfers it into the entry in "struct lguest": */ +/*H:235 + * This is the routine which actually checks the Guest's IDT entry and + * transfers it into the entry in "struct lguest": + */ static void set_trap(struct lg_cpu *cpu, struct desc_struct *trap, unsigned int num, u32 lo, u32 hi) { @@ -407,30 +478,38 @@ static void set_trap(struct lg_cpu *cpu, struct desc_struct *trap, if (type != 0xE && type != 0xF) kill_guest(cpu, "bad IDT type %i", type); - /* We only copy the handler address, present bit, privilege level and + /* + * We only copy the handler address, present bit, privilege level and * type. The privilege level controls where the trap can be triggered * manually with an "int" instruction. This is usually GUEST_PL, - * except for system calls which userspace can use. */ + * except for system calls which userspace can use. + */ trap->a = ((__KERNEL_CS|GUEST_PL)<<16) | (lo&0x0000FFFF); trap->b = (hi&0xFFFFEF00); } -/*H:230 While we're here, dealing with delivering traps and interrupts to the +/*H:230 + * While we're here, dealing with delivering traps and interrupts to the * Guest, we might as well complete the picture: how the Guest tells us where * it wants them to go. This would be simple, except making traps fast * requires some tricks. * * We saw the Guest setting Interrupt Descriptor Table (IDT) entries with the - * LHCALL_LOAD_IDT_ENTRY hypercall before: that comes here. */ + * LHCALL_LOAD_IDT_ENTRY hypercall before: that comes here. + */ void load_guest_idt_entry(struct lg_cpu *cpu, unsigned int num, u32 lo, u32 hi) { - /* Guest never handles: NMI, doublefault, spurious interrupt or - * hypercall. We ignore when it tries to set them. */ + /* + * Guest never handles: NMI, doublefault, spurious interrupt or + * hypercall. We ignore when it tries to set them. + */ if (num == 2 || num == 8 || num == 15 || num == LGUEST_TRAP_ENTRY) return; - /* Mark the IDT as changed: next time the Guest runs we'll know we have - * to copy this again. */ + /* + * Mark the IDT as changed: next time the Guest runs we'll know we have + * to copy this again. + */ cpu->changed |= CHANGED_IDT; /* Check that the Guest doesn't try to step outside the bounds. */ @@ -440,9 +519,11 @@ void load_guest_idt_entry(struct lg_cpu *cpu, unsigned int num, u32 lo, u32 hi) set_trap(cpu, &cpu->arch.idt[num], num, lo, hi); } -/* The default entry for each interrupt points into the Switcher routines which +/* + * The default entry for each interrupt points into the Switcher routines which * simply return to the Host. The run_guest() loop will then call - * deliver_trap() to bounce it back into the Guest. */ + * deliver_trap() to bounce it back into the Guest. + */ static void default_idt_entry(struct desc_struct *idt, int trap, const unsigned long handler, @@ -451,13 +532,17 @@ static void default_idt_entry(struct desc_struct *idt, /* A present interrupt gate. */ u32 flags = 0x8e00; - /* Set the privilege level on the entry for the hypercall: this allows - * the Guest to use the "int" instruction to trigger it. */ + /* + * Set the privilege level on the entry for the hypercall: this allows + * the Guest to use the "int" instruction to trigger it. + */ if (trap == LGUEST_TRAP_ENTRY) flags |= (GUEST_PL << 13); else if (base) - /* Copy priv. level from what Guest asked for. This allows - * debug (int 3) traps from Guest userspace, for example. */ + /* + * Copy privilege level from what Guest asked for. This allows + * debug (int 3) traps from Guest userspace, for example. + */ flags |= (base->b & 0x6000); /* Now pack it into the IDT entry in its weird format. */ @@ -475,16 +560,20 @@ void setup_default_idt_entries(struct lguest_ro_state *state, default_idt_entry(&state->guest_idt[i], i, def[i], NULL); } -/*H:240 We don't use the IDT entries in the "struct lguest" directly, instead +/*H:240 + * We don't use the IDT entries in the "struct lguest" directly, instead * we copy them into the IDT which we've set up for Guests on this CPU, just - * before we run the Guest. This routine does that copy. */ + * before we run the Guest. This routine does that copy. + */ void copy_traps(const struct lg_cpu *cpu, struct desc_struct *idt, const unsigned long *def) { unsigned int i; - /* We can simply copy the direct traps, otherwise we use the default - * ones in the Switcher: they will return to the Host. */ + /* + * We can simply copy the direct traps, otherwise we use the default + * ones in the Switcher: they will return to the Host. + */ for (i = 0; i < ARRAY_SIZE(cpu->arch.idt); i++) { const struct desc_struct *gidt = &cpu->arch.idt[i]; @@ -492,14 +581,16 @@ void copy_traps(const struct lg_cpu *cpu, struct desc_struct *idt, if (!direct_trap(i)) continue; - /* Only trap gates (type 15) can go direct to the Guest. + /* + * Only trap gates (type 15) can go direct to the Guest. * Interrupt gates (type 14) disable interrupts as they are * entered, which we never let the Guest do. Not present * entries (type 0x0) also can't go direct, of course. * * If it can't go direct, we still need to copy the priv. level: * they might want to give userspace access to a software - * interrupt. */ + * interrupt. + */ if (idt_type(gidt->a, gidt->b) == 0xF) idt[i] = *gidt; else @@ -518,7 +609,8 @@ void copy_traps(const struct lg_cpu *cpu, struct desc_struct *idt, * the next timer interrupt (in nanoseconds). We use the high-resolution timer * infrastructure to set a callback at that time. * - * 0 means "turn off the clock". */ + * 0 means "turn off the clock". + */ void guest_set_clockevent(struct lg_cpu *cpu, unsigned long delta) { ktime_t expires; @@ -529,9 +621,11 @@ void guest_set_clockevent(struct lg_cpu *cpu, unsigned long delta) return; } - /* We use wallclock time here, so the Guest might not be running for + /* + * We use wallclock time here, so the Guest might not be running for * all the time between now and the timer interrupt it asked for. This - * is almost always the right thing to do. */ + * is almost always the right thing to do. + */ expires = ktime_add_ns(ktime_get_real(), delta); hrtimer_start(&cpu->hrt, expires, HRTIMER_MODE_ABS); } diff --git a/drivers/lguest/lg.h b/drivers/lguest/lg.h index 01c591923793..74c0db691b53 100644 --- a/drivers/lguest/lg.h +++ b/drivers/lguest/lg.h @@ -54,13 +54,13 @@ struct lg_cpu { unsigned long pending_notify; /* pfn from LHCALL_NOTIFY */ - /* At end of a page shared mapped over lguest_pages in guest. */ + /* At end of a page shared mapped over lguest_pages in guest. */ unsigned long regs_page; struct lguest_regs *regs; struct lguest_pages *last_pages; - int cpu_pgd; /* which pgd this cpu is currently using */ + int cpu_pgd; /* Which pgd this cpu is currently using */ /* If a hypercall was asked for, this points to the arguments. */ struct hcall_args *hcall; @@ -96,8 +96,11 @@ struct lguest unsigned int nr_cpus; u32 pfn_limit; - /* This provides the offset to the base of guest-physical - * memory in the Launcher. */ + + /* + * This provides the offset to the base of guest-physical memory in the + * Launcher. + */ void __user *mem_base; unsigned long kernel_address; @@ -122,11 +125,13 @@ bool lguest_address_ok(const struct lguest *lg, void __lgread(struct lg_cpu *, void *, unsigned long, unsigned); void __lgwrite(struct lg_cpu *, unsigned long, const void *, unsigned); -/*H:035 Using memory-copy operations like that is usually inconvient, so we +/*H:035 + * Using memory-copy operations like that is usually inconvient, so we * have the following helper macros which read and write a specific type (often * an unsigned long). * - * This reads into a variable of the given type then returns that. */ + * This reads into a variable of the given type then returns that. + */ #define lgread(cpu, addr, type) \ ({ type _v; __lgread((cpu), &_v, (addr), sizeof(_v)); _v; }) @@ -140,9 +145,11 @@ void __lgwrite(struct lg_cpu *, unsigned long, const void *, unsigned); int run_guest(struct lg_cpu *cpu, unsigned long __user *user); -/* Helper macros to obtain the first 12 or the last 20 bits, this is only the +/* + * Helper macros to obtain the first 12 or the last 20 bits, this is only the * first step in the migration to the kernel types. pte_pfn is already defined - * in the kernel. */ + * in the kernel. + */ #define pgd_flags(x) (pgd_val(x) & ~PAGE_MASK) #define pgd_pfn(x) (pgd_val(x) >> PAGE_SHIFT) #define pmd_flags(x) (pmd_val(x) & ~PAGE_MASK) diff --git a/drivers/lguest/lguest_device.c b/drivers/lguest/lguest_device.c index e082cdac88b4..cc000e79c3d1 100644 --- a/drivers/lguest/lguest_device.c +++ b/drivers/lguest/lguest_device.c @@ -1,10 +1,12 @@ -/*P:050 Lguest guests use a very simple method to describe devices. It's a +/*P:050 + * Lguest guests use a very simple method to describe devices. It's a * series of device descriptors contained just above the top of normal Guest * memory. * * We use the standard "virtio" device infrastructure, which provides us with a * console, a network and a block driver. Each one expects some configuration - * information and a "virtqueue" or two to send and receive data. :*/ + * information and a "virtqueue" or two to send and receive data. +:*/ #include #include #include @@ -20,8 +22,10 @@ /* The pointer to our (page) of device descriptions. */ static void *lguest_devices; -/* For Guests, device memory can be used as normal memory, so we cast away the - * __iomem to quieten sparse. */ +/* + * For Guests, device memory can be used as normal memory, so we cast away the + * __iomem to quieten sparse. + */ static inline void *lguest_map(unsigned long phys_addr, unsigned long pages) { return (__force void *)ioremap_cache(phys_addr, PAGE_SIZE*pages); @@ -32,8 +36,10 @@ static inline void lguest_unmap(void *addr) iounmap((__force void __iomem *)addr); } -/*D:100 Each lguest device is just a virtio device plus a pointer to its entry - * in the lguest_devices page. */ +/*D:100 + * Each lguest device is just a virtio device plus a pointer to its entry + * in the lguest_devices page. + */ struct lguest_device { struct virtio_device vdev; @@ -41,9 +47,11 @@ struct lguest_device { struct lguest_device_desc *desc; }; -/* Since the virtio infrastructure hands us a pointer to the virtio_device all +/* + * Since the virtio infrastructure hands us a pointer to the virtio_device all * the time, it helps to have a curt macro to get a pointer to the struct - * lguest_device it's enclosed in. */ + * lguest_device it's enclosed in. + */ #define to_lgdev(vd) container_of(vd, struct lguest_device, vdev) /*D:130 @@ -55,7 +63,8 @@ struct lguest_device { * the driver will look at them during setup. * * A convenient routine to return the device's virtqueue config array: - * immediately after the descriptor. */ + * immediately after the descriptor. + */ static struct lguest_vqconfig *lg_vq(const struct lguest_device_desc *desc) { return (void *)(desc + 1); @@ -98,10 +107,12 @@ static u32 lg_get_features(struct virtio_device *vdev) return features; } -/* The virtio core takes the features the Host offers, and copies the - * ones supported by the driver into the vdev->features array. Once - * that's all sorted out, this routine is called so we can tell the - * Host which features we understand and accept. */ +/* + * The virtio core takes the features the Host offers, and copies the ones + * supported by the driver into the vdev->features array. Once that's all + * sorted out, this routine is called so we can tell the Host which features we + * understand and accept. + */ static void lg_finalize_features(struct virtio_device *vdev) { unsigned int i, bits; @@ -112,10 +123,11 @@ static void lg_finalize_features(struct virtio_device *vdev) /* Give virtio_ring a chance to accept features. */ vring_transport_features(vdev); - /* The vdev->feature array is a Linux bitmask: this isn't the - * same as a the simple array of bits used by lguest devices - * for features. So we do this slow, manual conversion which is - * completely general. */ + /* + * The vdev->feature array is a Linux bitmask: this isn't the same as a + * the simple array of bits used by lguest devices for features. So we + * do this slow, manual conversion which is completely general. + */ memset(out_features, 0, desc->feature_len); bits = min_t(unsigned, desc->feature_len, sizeof(vdev->features)) * 8; for (i = 0; i < bits; i++) { @@ -146,15 +158,19 @@ static void lg_set(struct virtio_device *vdev, unsigned int offset, memcpy(lg_config(desc) + offset, buf, len); } -/* The operations to get and set the status word just access the status field - * of the device descriptor. */ +/* + * The operations to get and set the status word just access the status field + * of the device descriptor. + */ static u8 lg_get_status(struct virtio_device *vdev) { return to_lgdev(vdev)->desc->status; } -/* To notify on status updates, we (ab)use the NOTIFY hypercall, with the - * descriptor address of the device. A zero status means "reset". */ +/* + * To notify on status updates, we (ab)use the NOTIFY hypercall, with the + * descriptor address of the device. A zero status means "reset". + */ static void set_status(struct virtio_device *vdev, u8 status) { unsigned long offset = (void *)to_lgdev(vdev)->desc - lguest_devices; @@ -200,13 +216,17 @@ struct lguest_vq_info void *pages; }; -/* When the virtio_ring code wants to prod the Host, it calls us here and we +/* + * When the virtio_ring code wants to prod the Host, it calls us here and we * make a hypercall. We hand the physical address of the virtqueue so the Host - * knows which virtqueue we're talking about. */ + * knows which virtqueue we're talking about. + */ static void lg_notify(struct virtqueue *vq) { - /* We store our virtqueue information in the "priv" pointer of the - * virtqueue structure. */ + /* + * We store our virtqueue information in the "priv" pointer of the + * virtqueue structure. + */ struct lguest_vq_info *lvq = vq->priv; kvm_hypercall1(LHCALL_NOTIFY, lvq->config.pfn << PAGE_SHIFT); @@ -215,7 +235,8 @@ static void lg_notify(struct virtqueue *vq) /* An extern declaration inside a C file is bad form. Don't do it. */ extern void lguest_setup_irq(unsigned int irq); -/* This routine finds the first virtqueue described in the configuration of +/* + * This routine finds the first virtqueue described in the configuration of * this device and sets it up. * * This is kind of an ugly duckling. It'd be nicer to have a standard @@ -225,7 +246,8 @@ extern void lguest_setup_irq(unsigned int irq); * simpler for the Host to simply tell us where the pages are. * * So we provide drivers with a "find the Nth virtqueue and set it up" - * function. */ + * function. + */ static struct virtqueue *lg_find_vq(struct virtio_device *vdev, unsigned index, void (*callback)(struct virtqueue *vq), @@ -244,9 +266,11 @@ static struct virtqueue *lg_find_vq(struct virtio_device *vdev, if (!lvq) return ERR_PTR(-ENOMEM); - /* Make a copy of the "struct lguest_vqconfig" entry, which sits after + /* + * Make a copy of the "struct lguest_vqconfig" entry, which sits after * the descriptor. We need a copy because the config space might not - * be aligned correctly. */ + * be aligned correctly. + */ memcpy(&lvq->config, lg_vq(ldev->desc)+index, sizeof(lvq->config)); printk("Mapping virtqueue %i addr %lx\n", index, @@ -261,8 +285,10 @@ static struct virtqueue *lg_find_vq(struct virtio_device *vdev, goto free_lvq; } - /* OK, tell virtio_ring.c to set up a virtqueue now we know its size - * and we've got a pointer to its pages. */ + /* + * OK, tell virtio_ring.c to set up a virtqueue now we know its size + * and we've got a pointer to its pages. + */ vq = vring_new_virtqueue(lvq->config.num, LGUEST_VRING_ALIGN, vdev, lvq->pages, lg_notify, callback, name); if (!vq) { @@ -273,18 +299,23 @@ static struct virtqueue *lg_find_vq(struct virtio_device *vdev, /* Make sure the interrupt is allocated. */ lguest_setup_irq(lvq->config.irq); - /* Tell the interrupt for this virtqueue to go to the virtio_ring - * interrupt handler. */ - /* FIXME: We used to have a flag for the Host to tell us we could use + /* + * Tell the interrupt for this virtqueue to go to the virtio_ring + * interrupt handler. + * + * FIXME: We used to have a flag for the Host to tell us we could use * the interrupt as a source of randomness: it'd be nice to have that - * back.. */ + * back. + */ err = request_irq(lvq->config.irq, vring_interrupt, IRQF_SHARED, dev_name(&vdev->dev), vq); if (err) goto destroy_vring; - /* Last of all we hook up our 'struct lguest_vq_info" to the - * virtqueue's priv pointer. */ + /* + * Last of all we hook up our 'struct lguest_vq_info" to the + * virtqueue's priv pointer. + */ vq->priv = lvq; return vq; @@ -358,11 +389,14 @@ static struct virtio_config_ops lguest_config_ops = { .del_vqs = lg_del_vqs, }; -/* The root device for the lguest virtio devices. This makes them appear as - * /sys/devices/lguest/0,1,2 not /sys/devices/0,1,2. */ +/* + * The root device for the lguest virtio devices. This makes them appear as + * /sys/devices/lguest/0,1,2 not /sys/devices/0,1,2. + */ static struct device *lguest_root; -/*D:120 This is the core of the lguest bus: actually adding a new device. +/*D:120 + * This is the core of the lguest bus: actually adding a new device. * It's a separate function because it's neater that way, and because an * earlier version of the code supported hotplug and unplug. They were removed * early on because they were never used. @@ -371,14 +405,14 @@ static struct device *lguest_root; * * It's worth reading this carefully: we start with a pointer to the new device * descriptor in the "lguest_devices" page, and the offset into the device - * descriptor page so we can uniquely identify it if things go badly wrong. */ + * descriptor page so we can uniquely identify it if things go badly wrong. + */ static void add_lguest_device(struct lguest_device_desc *d, unsigned int offset) { struct lguest_device *ldev; - /* Start with zeroed memory; Linux's device layer seems to count on - * it. */ + /* Start with zeroed memory; Linux's device layer counts on it. */ ldev = kzalloc(sizeof(*ldev), GFP_KERNEL); if (!ldev) { printk(KERN_EMERG "Cannot allocate lguest dev %u type %u\n", @@ -390,15 +424,19 @@ static void add_lguest_device(struct lguest_device_desc *d, ldev->vdev.dev.parent = lguest_root; /* We have a unique device index thanks to the dev_index counter. */ ldev->vdev.id.device = d->type; - /* We have a simple set of routines for querying the device's - * configuration information and setting its status. */ + /* + * We have a simple set of routines for querying the device's + * configuration information and setting its status. + */ ldev->vdev.config = &lguest_config_ops; /* And we remember the device's descriptor for lguest_config_ops. */ ldev->desc = d; - /* register_virtio_device() sets up the generic fields for the struct + /* + * register_virtio_device() sets up the generic fields for the struct * virtio_device and calls device_register(). This makes the bus - * infrastructure look for a matching driver. */ + * infrastructure look for a matching driver. + */ if (register_virtio_device(&ldev->vdev) != 0) { printk(KERN_ERR "Failed to register lguest dev %u type %u\n", offset, d->type); @@ -406,8 +444,10 @@ static void add_lguest_device(struct lguest_device_desc *d, } } -/*D:110 scan_devices() simply iterates through the device page. The type 0 is - * reserved to mean "end of devices". */ +/*D:110 + * scan_devices() simply iterates through the device page. The type 0 is + * reserved to mean "end of devices". + */ static void scan_devices(void) { unsigned int i; @@ -426,7 +466,8 @@ static void scan_devices(void) } } -/*D:105 Fairly early in boot, lguest_devices_init() is called to set up the +/*D:105 + * Fairly early in boot, lguest_devices_init() is called to set up the * lguest device infrastructure. We check that we are a Guest by checking * pv_info.name: there are other ways of checking, but this seems most * obvious to me. @@ -437,7 +478,8 @@ static void scan_devices(void) * correct sysfs incantation). * * Finally we call scan_devices() which adds all the devices found in the - * lguest_devices page. */ + * lguest_devices page. + */ static int __init lguest_devices_init(void) { if (strcmp(pv_info.name, "lguest") != 0) @@ -456,11 +498,13 @@ static int __init lguest_devices_init(void) /* We do this after core stuff, but before the drivers. */ postcore_initcall(lguest_devices_init); -/*D:150 At this point in the journey we used to now wade through the lguest +/*D:150 + * At this point in the journey we used to now wade through the lguest * devices themselves: net, block and console. Since they're all now virtio * devices rather than lguest-specific, I've decided to ignore them. Mostly, * they're kind of boring. But this does mean you'll never experience the * thrill of reading the forbidden love scene buried deep in the block driver. * * "make Launcher" beckons, where we answer questions like "Where do Guests - * come from?", and "What do you do when someone asks for optimization?". */ + * come from?", and "What do you do when someone asks for optimization?". + */ diff --git a/drivers/lguest/lguest_user.c b/drivers/lguest/lguest_user.c index 407722a8e0c4..7e92017103dc 100644 --- a/drivers/lguest/lguest_user.c +++ b/drivers/lguest/lguest_user.c @@ -1,8 +1,10 @@ -/*P:200 This contains all the /dev/lguest code, whereby the userspace launcher +/*P:200 + * This contains all the /dev/lguest code, whereby the userspace launcher * controls and communicates with the Guest. For example, the first write will * tell us the Guest's memory layout, pagetable, entry point and kernel address * offset. A read will run the Guest until something happens, such as a signal - * or the Guest doing a NOTIFY out to the Launcher. :*/ + * or the Guest doing a NOTIFY out to the Launcher. +:*/ #include #include #include @@ -37,8 +39,10 @@ static int add_eventfd(struct lguest *lg, unsigned long addr, int fd) if (!addr) return -EINVAL; - /* Replace the old array with the new one, carefully: others can - * be accessing it at the same time */ + /* + * Replace the old array with the new one, carefully: others can + * be accessing it at the same time. + */ new = kmalloc(sizeof(*new) + sizeof(new->map[0]) * (old->num + 1), GFP_KERNEL); if (!new) @@ -61,8 +65,10 @@ static int add_eventfd(struct lguest *lg, unsigned long addr, int fd) /* Now put new one in place. */ rcu_assign_pointer(lg->eventfds, new); - /* We're not in a big hurry. Wait until noone's looking at old - * version, then delete it. */ + /* + * We're not in a big hurry. Wait until noone's looking at old + * version, then delete it. + */ synchronize_rcu(); kfree(old); @@ -87,8 +93,10 @@ static int attach_eventfd(struct lguest *lg, const unsigned long __user *input) return err; } -/*L:050 Sending an interrupt is done by writing LHREQ_IRQ and an interrupt - * number to /dev/lguest. */ +/*L:050 + * Sending an interrupt is done by writing LHREQ_IRQ and an interrupt + * number to /dev/lguest. + */ static int user_send_irq(struct lg_cpu *cpu, const unsigned long __user *input) { unsigned long irq; @@ -102,8 +110,10 @@ static int user_send_irq(struct lg_cpu *cpu, const unsigned long __user *input) return 0; } -/*L:040 Once our Guest is initialized, the Launcher makes it run by reading - * from /dev/lguest. */ +/*L:040 + * Once our Guest is initialized, the Launcher makes it run by reading + * from /dev/lguest. + */ static ssize_t read(struct file *file, char __user *user, size_t size,loff_t*o) { struct lguest *lg = file->private_data; @@ -139,8 +149,10 @@ static ssize_t read(struct file *file, char __user *user, size_t size,loff_t*o) return len; } - /* If we returned from read() last time because the Guest sent I/O, - * clear the flag. */ + /* + * If we returned from read() last time because the Guest sent I/O, + * clear the flag. + */ if (cpu->pending_notify) cpu->pending_notify = 0; @@ -148,8 +160,10 @@ static ssize_t read(struct file *file, char __user *user, size_t size,loff_t*o) return run_guest(cpu, (unsigned long __user *)user); } -/*L:025 This actually initializes a CPU. For the moment, a Guest is only - * uniprocessor, so "id" is always 0. */ +/*L:025 + * This actually initializes a CPU. For the moment, a Guest is only + * uniprocessor, so "id" is always 0. + */ static int lg_cpu_start(struct lg_cpu *cpu, unsigned id, unsigned long start_ip) { /* We have a limited number the number of CPUs in the lguest struct. */ @@ -164,8 +178,10 @@ static int lg_cpu_start(struct lg_cpu *cpu, unsigned id, unsigned long start_ip) /* Each CPU has a timer it can set. */ init_clockdev(cpu); - /* We need a complete page for the Guest registers: they are accessible - * to the Guest and we can only grant it access to whole pages. */ + /* + * We need a complete page for the Guest registers: they are accessible + * to the Guest and we can only grant it access to whole pages. + */ cpu->regs_page = get_zeroed_page(GFP_KERNEL); if (!cpu->regs_page) return -ENOMEM; @@ -173,29 +189,38 @@ static int lg_cpu_start(struct lg_cpu *cpu, unsigned id, unsigned long start_ip) /* We actually put the registers at the bottom of the page. */ cpu->regs = (void *)cpu->regs_page + PAGE_SIZE - sizeof(*cpu->regs); - /* Now we initialize the Guest's registers, handing it the start - * address. */ + /* + * Now we initialize the Guest's registers, handing it the start + * address. + */ lguest_arch_setup_regs(cpu, start_ip); - /* We keep a pointer to the Launcher task (ie. current task) for when - * other Guests want to wake this one (eg. console input). */ + /* + * We keep a pointer to the Launcher task (ie. current task) for when + * other Guests want to wake this one (eg. console input). + */ cpu->tsk = current; - /* We need to keep a pointer to the Launcher's memory map, because if + /* + * We need to keep a pointer to the Launcher's memory map, because if * the Launcher dies we need to clean it up. If we don't keep a - * reference, it is destroyed before close() is called. */ + * reference, it is destroyed before close() is called. + */ cpu->mm = get_task_mm(cpu->tsk); - /* We remember which CPU's pages this Guest used last, for optimization - * when the same Guest runs on the same CPU twice. */ + /* + * We remember which CPU's pages this Guest used last, for optimization + * when the same Guest runs on the same CPU twice. + */ cpu->last_pages = NULL; /* No error == success. */ return 0; } -/*L:020 The initialization write supplies 3 pointer sized (32 or 64 bit) - * values (in addition to the LHREQ_INITIALIZE value). These are: +/*L:020 + * The initialization write supplies 3 pointer sized (32 or 64 bit) values (in + * addition to the LHREQ_INITIALIZE value). These are: * * base: The start of the Guest-physical memory inside the Launcher memory. * @@ -207,14 +232,15 @@ static int lg_cpu_start(struct lg_cpu *cpu, unsigned id, unsigned long start_ip) */ static int initialize(struct file *file, const unsigned long __user *input) { - /* "struct lguest" contains everything we (the Host) know about a - * Guest. */ + /* "struct lguest" contains all we (the Host) know about a Guest. */ struct lguest *lg; int err; unsigned long args[3]; - /* We grab the Big Lguest lock, which protects against multiple - * simultaneous initializations. */ + /* + * We grab the Big Lguest lock, which protects against multiple + * simultaneous initializations. + */ mutex_lock(&lguest_lock); /* You can't initialize twice! Close the device and start again... */ if (file->private_data) { @@ -249,8 +275,10 @@ static int initialize(struct file *file, const unsigned long __user *input) if (err) goto free_eventfds; - /* Initialize the Guest's shadow page tables, using the toplevel - * address the Launcher gave us. This allocates memory, so can fail. */ + /* + * Initialize the Guest's shadow page tables, using the toplevel + * address the Launcher gave us. This allocates memory, so can fail. + */ err = init_guest_pagetable(lg); if (err) goto free_regs; @@ -275,7 +303,8 @@ unlock: return err; } -/*L:010 The first operation the Launcher does must be a write. All writes +/*L:010 + * The first operation the Launcher does must be a write. All writes * start with an unsigned long number: for the first write this must be * LHREQ_INITIALIZE to set up the Guest. After that the Launcher can use * writes of other values to send interrupts. @@ -283,12 +312,15 @@ unlock: * Note that we overload the "offset" in the /dev/lguest file to indicate what * CPU number we're dealing with. Currently this is always 0, since we only * support uniprocessor Guests, but you can see the beginnings of SMP support - * here. */ + * here. + */ static ssize_t write(struct file *file, const char __user *in, size_t size, loff_t *off) { - /* Once the Guest is initialized, we hold the "struct lguest" in the - * file private data. */ + /* + * Once the Guest is initialized, we hold the "struct lguest" in the + * file private data. + */ struct lguest *lg = file->private_data; const unsigned long __user *input = (const unsigned long __user *)in; unsigned long req; @@ -323,13 +355,15 @@ static ssize_t write(struct file *file, const char __user *in, } } -/*L:060 The final piece of interface code is the close() routine. It reverses +/*L:060 + * The final piece of interface code is the close() routine. It reverses * everything done in initialize(). This is usually called because the * Launcher exited. * * Note that the close routine returns 0 or a negative error number: it can't * really fail, but it can whine. I blame Sun for this wart, and K&R C for - * letting them do it. :*/ + * letting them do it. +:*/ static int close(struct inode *inode, struct file *file) { struct lguest *lg = file->private_data; @@ -339,8 +373,10 @@ static int close(struct inode *inode, struct file *file) if (!lg) return 0; - /* We need the big lock, to protect from inter-guest I/O and other - * Launchers initializing guests. */ + /* + * We need the big lock, to protect from inter-guest I/O and other + * Launchers initializing guests. + */ mutex_lock(&lguest_lock); /* Free up the shadow page tables for the Guest. */ @@ -351,8 +387,10 @@ static int close(struct inode *inode, struct file *file) hrtimer_cancel(&lg->cpus[i].hrt); /* We can free up the register page we allocated. */ free_page(lg->cpus[i].regs_page); - /* Now all the memory cleanups are done, it's safe to release - * the Launcher's memory management structure. */ + /* + * Now all the memory cleanups are done, it's safe to release + * the Launcher's memory management structure. + */ mmput(lg->cpus[i].mm); } @@ -361,8 +399,10 @@ static int close(struct inode *inode, struct file *file) eventfd_ctx_put(lg->eventfds->map[i].event); kfree(lg->eventfds); - /* If lg->dead doesn't contain an error code it will be NULL or a - * kmalloc()ed string, either of which is ok to hand to kfree(). */ + /* + * If lg->dead doesn't contain an error code it will be NULL or a + * kmalloc()ed string, either of which is ok to hand to kfree(). + */ if (!IS_ERR(lg->dead)) kfree(lg->dead); /* Free the memory allocated to the lguest_struct */ @@ -386,7 +426,8 @@ static int close(struct inode *inode, struct file *file) * * We begin our understanding with the Host kernel interface which the Launcher * uses: reading and writing a character device called /dev/lguest. All the - * work happens in the read(), write() and close() routines: */ + * work happens in the read(), write() and close() routines: + */ static struct file_operations lguest_fops = { .owner = THIS_MODULE, .release = close, @@ -394,8 +435,10 @@ static struct file_operations lguest_fops = { .read = read, }; -/* This is a textbook example of a "misc" character device. Populate a "struct - * miscdevice" and register it with misc_register(). */ +/* + * This is a textbook example of a "misc" character device. Populate a "struct + * miscdevice" and register it with misc_register(). + */ static struct miscdevice lguest_dev = { .minor = MISC_DYNAMIC_MINOR, .name = "lguest", diff --git a/drivers/lguest/page_tables.c b/drivers/lguest/page_tables.c index a6fe1abda240..3da902e4b4cb 100644 --- a/drivers/lguest/page_tables.c +++ b/drivers/lguest/page_tables.c @@ -1,9 +1,11 @@ -/*P:700 The pagetable code, on the other hand, still shows the scars of +/*P:700 + * The pagetable code, on the other hand, still shows the scars of * previous encounters. It's functional, and as neat as it can be in the * circumstances, but be wary, for these things are subtle and break easily. * The Guest provides a virtual to physical mapping, but we can neither trust * it nor use it: we verify and convert it here then point the CPU to the - * converted Guest pages when running the Guest. :*/ + * converted Guest pages when running the Guest. +:*/ /* Copyright (C) Rusty Russell IBM Corporation 2006. * GPL v2 and any later version */ @@ -17,10 +19,12 @@ #include #include "lg.h" -/*M:008 We hold reference to pages, which prevents them from being swapped. +/*M:008 + * We hold reference to pages, which prevents them from being swapped. * It'd be nice to have a callback in the "struct mm_struct" when Linux wants * to swap out. If we had this, and a shrinker callback to trim PTE pages, we - * could probably consider launching Guests as non-root. :*/ + * could probably consider launching Guests as non-root. +:*/ /*H:300 * The Page Table Code @@ -45,16 +49,19 @@ * (v) Flushing (throwing away) page tables, * (vi) Mapping the Switcher when the Guest is about to run, * (vii) Setting up the page tables initially. - :*/ +:*/ - -/* 1024 entries in a page table page maps 1024 pages: 4MB. The Switcher is +/* + * 1024 entries in a page table page maps 1024 pages: 4MB. The Switcher is * conveniently placed at the top 4MB, so it uses a separate, complete PTE - * page. */ + * page. + */ #define SWITCHER_PGD_INDEX (PTRS_PER_PGD - 1) -/* For PAE we need the PMD index as well. We use the last 2MB, so we - * will need the last pmd entry of the last pmd page. */ +/* + * For PAE we need the PMD index as well. We use the last 2MB, so we + * will need the last pmd entry of the last pmd page. + */ #ifdef CONFIG_X86_PAE #define SWITCHER_PMD_INDEX (PTRS_PER_PMD - 1) #define RESERVE_MEM 2U @@ -64,13 +71,16 @@ #define CHECK_GPGD_MASK _PAGE_TABLE #endif -/* We actually need a separate PTE page for each CPU. Remember that after the +/* + * We actually need a separate PTE page for each CPU. Remember that after the * Switcher code itself comes two pages for each CPU, and we don't want this - * CPU's guest to see the pages of any other CPU. */ + * CPU's guest to see the pages of any other CPU. + */ static DEFINE_PER_CPU(pte_t *, switcher_pte_pages); #define switcher_pte_page(cpu) per_cpu(switcher_pte_pages, cpu) -/*H:320 The page table code is curly enough to need helper functions to keep it +/*H:320 + * The page table code is curly enough to need helper functions to keep it * clear and clean. * * There are two functions which return pointers to the shadow (aka "real") @@ -79,7 +89,8 @@ static DEFINE_PER_CPU(pte_t *, switcher_pte_pages); * spgd_addr() takes the virtual address and returns a pointer to the top-level * page directory entry (PGD) for that address. Since we keep track of several * page tables, the "i" argument tells us which one we're interested in (it's - * usually the current one). */ + * usually the current one). + */ static pgd_t *spgd_addr(struct lg_cpu *cpu, u32 i, unsigned long vaddr) { unsigned int index = pgd_index(vaddr); @@ -96,9 +107,11 @@ static pgd_t *spgd_addr(struct lg_cpu *cpu, u32 i, unsigned long vaddr) } #ifdef CONFIG_X86_PAE -/* This routine then takes the PGD entry given above, which contains the +/* + * This routine then takes the PGD entry given above, which contains the * address of the PMD page. It then returns a pointer to the PMD entry for the - * given address. */ + * given address. + */ static pmd_t *spmd_addr(struct lg_cpu *cpu, pgd_t spgd, unsigned long vaddr) { unsigned int index = pmd_index(vaddr); @@ -119,9 +132,11 @@ static pmd_t *spmd_addr(struct lg_cpu *cpu, pgd_t spgd, unsigned long vaddr) } #endif -/* This routine then takes the page directory entry returned above, which +/* + * This routine then takes the page directory entry returned above, which * contains the address of the page table entry (PTE) page. It then returns a - * pointer to the PTE entry for the given address. */ + * pointer to the PTE entry for the given address. + */ static pte_t *spte_addr(struct lg_cpu *cpu, pgd_t spgd, unsigned long vaddr) { #ifdef CONFIG_X86_PAE @@ -139,8 +154,10 @@ static pte_t *spte_addr(struct lg_cpu *cpu, pgd_t spgd, unsigned long vaddr) return &page[pte_index(vaddr)]; } -/* These two functions just like the above two, except they access the Guest - * page tables. Hence they return a Guest address. */ +/* + * These two functions just like the above two, except they access the Guest + * page tables. Hence they return a Guest address. + */ static unsigned long gpgd_addr(struct lg_cpu *cpu, unsigned long vaddr) { unsigned int index = vaddr >> (PGDIR_SHIFT); @@ -175,17 +192,21 @@ static unsigned long gpte_addr(struct lg_cpu *cpu, #endif /*:*/ -/*M:014 get_pfn is slow: we could probably try to grab batches of pages here as - * an optimization (ie. pre-faulting). :*/ +/*M:014 + * get_pfn is slow: we could probably try to grab batches of pages here as + * an optimization (ie. pre-faulting). +:*/ -/*H:350 This routine takes a page number given by the Guest and converts it to +/*H:350 + * This routine takes a page number given by the Guest and converts it to * an actual, physical page number. It can fail for several reasons: the * virtual address might not be mapped by the Launcher, the write flag is set * and the page is read-only, or the write flag was set and the page was * shared so had to be copied, but we ran out of memory. * * This holds a reference to the page, so release_pte() is careful to put that - * back. */ + * back. + */ static unsigned long get_pfn(unsigned long virtpfn, int write) { struct page *page; @@ -198,33 +219,41 @@ static unsigned long get_pfn(unsigned long virtpfn, int write) return -1UL; } -/*H:340 Converting a Guest page table entry to a shadow (ie. real) page table +/*H:340 + * Converting a Guest page table entry to a shadow (ie. real) page table * entry can be a little tricky. The flags are (almost) the same, but the * Guest PTE contains a virtual page number: the CPU needs the real page - * number. */ + * number. + */ static pte_t gpte_to_spte(struct lg_cpu *cpu, pte_t gpte, int write) { unsigned long pfn, base, flags; - /* The Guest sets the global flag, because it thinks that it is using + /* + * The Guest sets the global flag, because it thinks that it is using * PGE. We only told it to use PGE so it would tell us whether it was * flushing a kernel mapping or a userspace mapping. We don't actually - * use the global bit, so throw it away. */ + * use the global bit, so throw it away. + */ flags = (pte_flags(gpte) & ~_PAGE_GLOBAL); /* The Guest's pages are offset inside the Launcher. */ base = (unsigned long)cpu->lg->mem_base / PAGE_SIZE; - /* We need a temporary "unsigned long" variable to hold the answer from + /* + * We need a temporary "unsigned long" variable to hold the answer from * get_pfn(), because it returns 0xFFFFFFFF on failure, which wouldn't * fit in spte.pfn. get_pfn() finds the real physical number of the - * page, given the virtual number. */ + * page, given the virtual number. + */ pfn = get_pfn(base + pte_pfn(gpte), write); if (pfn == -1UL) { kill_guest(cpu, "failed to get page %lu", pte_pfn(gpte)); - /* When we destroy the Guest, we'll go through the shadow page + /* + * When we destroy the Guest, we'll go through the shadow page * tables and release_pte() them. Make sure we don't think - * this one is valid! */ + * this one is valid! + */ flags = 0; } /* Now we assemble our shadow PTE from the page number and flags. */ @@ -234,8 +263,10 @@ static pte_t gpte_to_spte(struct lg_cpu *cpu, pte_t gpte, int write) /*H:460 And to complete the chain, release_pte() looks like this: */ static void release_pte(pte_t pte) { - /* Remember that get_user_pages_fast() took a reference to the page, in - * get_pfn()? We have to put it back now. */ + /* + * Remember that get_user_pages_fast() took a reference to the page, in + * get_pfn()? We have to put it back now. + */ if (pte_flags(pte) & _PAGE_PRESENT) put_page(pte_page(pte)); } @@ -273,7 +304,8 @@ static void check_gpmd(struct lg_cpu *cpu, pmd_t gpmd) * and return to the Guest without it knowing. * * If we fixed up the fault (ie. we mapped the address), this routine returns - * true. Otherwise, it was a real fault and we need to tell the Guest. */ + * true. Otherwise, it was a real fault and we need to tell the Guest. + */ bool demand_page(struct lg_cpu *cpu, unsigned long vaddr, int errcode) { pgd_t gpgd; @@ -298,22 +330,26 @@ bool demand_page(struct lg_cpu *cpu, unsigned long vaddr, int errcode) if (!(pgd_flags(*spgd) & _PAGE_PRESENT)) { /* No shadow entry: allocate a new shadow PTE page. */ unsigned long ptepage = get_zeroed_page(GFP_KERNEL); - /* This is not really the Guest's fault, but killing it is - * simple for this corner case. */ + /* + * This is not really the Guest's fault, but killing it is + * simple for this corner case. + */ if (!ptepage) { kill_guest(cpu, "out of memory allocating pte page"); return false; } /* We check that the Guest pgd is OK. */ check_gpgd(cpu, gpgd); - /* And we copy the flags to the shadow PGD entry. The page - * number in the shadow PGD is the page we just allocated. */ + /* + * And we copy the flags to the shadow PGD entry. The page + * number in the shadow PGD is the page we just allocated. + */ set_pgd(spgd, __pgd(__pa(ptepage) | pgd_flags(gpgd))); } #ifdef CONFIG_X86_PAE gpmd = lgread(cpu, gpmd_addr(gpgd, vaddr), pmd_t); - /* middle level not present? We can't map it in. */ + /* Middle level not present? We can't map it in. */ if (!(pmd_flags(gpmd) & _PAGE_PRESENT)) return false; @@ -324,8 +360,10 @@ bool demand_page(struct lg_cpu *cpu, unsigned long vaddr, int errcode) /* No shadow entry: allocate a new shadow PTE page. */ unsigned long ptepage = get_zeroed_page(GFP_KERNEL); - /* This is not really the Guest's fault, but killing it is - * simple for this corner case. */ + /* + * This is not really the Guest's fault, but killing it is + * simple for this corner case. + */ if (!ptepage) { kill_guest(cpu, "out of memory allocating pte page"); return false; @@ -334,17 +372,23 @@ bool demand_page(struct lg_cpu *cpu, unsigned long vaddr, int errcode) /* We check that the Guest pmd is OK. */ check_gpmd(cpu, gpmd); - /* And we copy the flags to the shadow PMD entry. The page - * number in the shadow PMD is the page we just allocated. */ + /* + * And we copy the flags to the shadow PMD entry. The page + * number in the shadow PMD is the page we just allocated. + */ native_set_pmd(spmd, __pmd(__pa(ptepage) | pmd_flags(gpmd))); } - /* OK, now we look at the lower level in the Guest page table: keep its - * address, because we might update it later. */ + /* + * OK, now we look at the lower level in the Guest page table: keep its + * address, because we might update it later. + */ gpte_ptr = gpte_addr(cpu, gpmd, vaddr); #else - /* OK, now we look at the lower level in the Guest page table: keep its - * address, because we might update it later. */ + /* + * OK, now we look at the lower level in the Guest page table: keep its + * address, because we might update it later. + */ gpte_ptr = gpte_addr(cpu, gpgd, vaddr); #endif gpte = lgread(cpu, gpte_ptr, pte_t); @@ -353,8 +397,10 @@ bool demand_page(struct lg_cpu *cpu, unsigned long vaddr, int errcode) if (!(pte_flags(gpte) & _PAGE_PRESENT)) return false; - /* Check they're not trying to write to a page the Guest wants - * read-only (bit 2 of errcode == write). */ + /* + * Check they're not trying to write to a page the Guest wants + * read-only (bit 2 of errcode == write). + */ if ((errcode & 2) && !(pte_flags(gpte) & _PAGE_RW)) return false; @@ -362,8 +408,10 @@ bool demand_page(struct lg_cpu *cpu, unsigned long vaddr, int errcode) if ((errcode & 4) && !(pte_flags(gpte) & _PAGE_USER)) return false; - /* Check that the Guest PTE flags are OK, and the page number is below - * the pfn_limit (ie. not mapping the Launcher binary). */ + /* + * Check that the Guest PTE flags are OK, and the page number is below + * the pfn_limit (ie. not mapping the Launcher binary). + */ check_gpte(cpu, gpte); /* Add the _PAGE_ACCESSED and (for a write) _PAGE_DIRTY flag */ @@ -373,29 +421,40 @@ bool demand_page(struct lg_cpu *cpu, unsigned long vaddr, int errcode) /* Get the pointer to the shadow PTE entry we're going to set. */ spte = spte_addr(cpu, *spgd, vaddr); - /* If there was a valid shadow PTE entry here before, we release it. - * This can happen with a write to a previously read-only entry. */ + + /* + * If there was a valid shadow PTE entry here before, we release it. + * This can happen with a write to a previously read-only entry. + */ release_pte(*spte); - /* If this is a write, we insist that the Guest page is writable (the - * final arg to gpte_to_spte()). */ + /* + * If this is a write, we insist that the Guest page is writable (the + * final arg to gpte_to_spte()). + */ if (pte_dirty(gpte)) *spte = gpte_to_spte(cpu, gpte, 1); else - /* If this is a read, don't set the "writable" bit in the page + /* + * If this is a read, don't set the "writable" bit in the page * table entry, even if the Guest says it's writable. That way * we will come back here when a write does actually occur, so - * we can update the Guest's _PAGE_DIRTY flag. */ + * we can update the Guest's _PAGE_DIRTY flag. + */ native_set_pte(spte, gpte_to_spte(cpu, pte_wrprotect(gpte), 0)); - /* Finally, we write the Guest PTE entry back: we've set the - * _PAGE_ACCESSED and maybe the _PAGE_DIRTY flags. */ + /* + * Finally, we write the Guest PTE entry back: we've set the + * _PAGE_ACCESSED and maybe the _PAGE_DIRTY flags. + */ lgwrite(cpu, gpte_ptr, pte_t, gpte); - /* The fault is fixed, the page table is populated, the mapping + /* + * The fault is fixed, the page table is populated, the mapping * manipulated, the result returned and the code complete. A small * delay and a trace of alliteration are the only indications the Guest - * has that a page fault occurred at all. */ + * has that a page fault occurred at all. + */ return true; } @@ -408,7 +467,8 @@ bool demand_page(struct lg_cpu *cpu, unsigned long vaddr, int errcode) * mapped, so it's overkill. * * This is a quick version which answers the question: is this virtual address - * mapped by the shadow page tables, and is it writable? */ + * mapped by the shadow page tables, and is it writable? + */ static bool page_writable(struct lg_cpu *cpu, unsigned long vaddr) { pgd_t *spgd; @@ -428,16 +488,20 @@ static bool page_writable(struct lg_cpu *cpu, unsigned long vaddr) return false; #endif - /* Check the flags on the pte entry itself: it must be present and - * writable. */ + /* + * Check the flags on the pte entry itself: it must be present and + * writable. + */ flags = pte_flags(*(spte_addr(cpu, *spgd, vaddr))); return (flags & (_PAGE_PRESENT|_PAGE_RW)) == (_PAGE_PRESENT|_PAGE_RW); } -/* So, when pin_stack_pages() asks us to pin a page, we check if it's already +/* + * So, when pin_stack_pages() asks us to pin a page, we check if it's already * in the page tables, and if not, we call demand_page() with error code 2 - * (meaning "write"). */ + * (meaning "write"). + */ void pin_page(struct lg_cpu *cpu, unsigned long vaddr) { if (!page_writable(cpu, vaddr) && !demand_page(cpu, vaddr, 2)) @@ -485,9 +549,11 @@ static void release_pgd(pgd_t *spgd) /* If the entry's not present, there's nothing to release. */ if (pgd_flags(*spgd) & _PAGE_PRESENT) { unsigned int i; - /* Converting the pfn to find the actual PTE page is easy: turn + /* + * Converting the pfn to find the actual PTE page is easy: turn * the page number into a physical address, then convert to a - * virtual address (easy for kernel pages like this one). */ + * virtual address (easy for kernel pages like this one). + */ pte_t *ptepage = __va(pgd_pfn(*spgd) << PAGE_SHIFT); /* For each entry in the page, we might need to release it. */ for (i = 0; i < PTRS_PER_PTE; i++) @@ -499,9 +565,12 @@ static void release_pgd(pgd_t *spgd) } } #endif -/*H:445 We saw flush_user_mappings() twice: once from the flush_user_mappings() + +/*H:445 + * We saw flush_user_mappings() twice: once from the flush_user_mappings() * hypercall and once in new_pgdir() when we re-used a top-level pgdir page. - * It simply releases every PTE page from 0 up to the Guest's kernel address. */ + * It simply releases every PTE page from 0 up to the Guest's kernel address. + */ static void flush_user_mappings(struct lguest *lg, int idx) { unsigned int i; @@ -510,10 +579,12 @@ static void flush_user_mappings(struct lguest *lg, int idx) release_pgd(lg->pgdirs[idx].pgdir + i); } -/*H:440 (v) Flushing (throwing away) page tables, +/*H:440 + * (v) Flushing (throwing away) page tables, * * The Guest has a hypercall to throw away the page tables: it's used when a - * large number of mappings have been changed. */ + * large number of mappings have been changed. + */ void guest_pagetable_flush_user(struct lg_cpu *cpu) { /* Drop the userspace part of the current page table. */ @@ -551,9 +622,11 @@ unsigned long guest_pa(struct lg_cpu *cpu, unsigned long vaddr) return pte_pfn(gpte) * PAGE_SIZE | (vaddr & ~PAGE_MASK); } -/* We keep several page tables. This is a simple routine to find the page +/* + * We keep several page tables. This is a simple routine to find the page * table (if any) corresponding to this top-level address the Guest has given - * us. */ + * us. + */ static unsigned int find_pgdir(struct lguest *lg, unsigned long pgtable) { unsigned int i; @@ -563,9 +636,11 @@ static unsigned int find_pgdir(struct lguest *lg, unsigned long pgtable) return i; } -/*H:435 And this is us, creating the new page directory. If we really do +/*H:435 + * And this is us, creating the new page directory. If we really do * allocate a new one (and so the kernel parts are not there), we set - * blank_pgdir. */ + * blank_pgdir. + */ static unsigned int new_pgdir(struct lg_cpu *cpu, unsigned long gpgdir, int *blank_pgdir) @@ -575,8 +650,10 @@ static unsigned int new_pgdir(struct lg_cpu *cpu, pmd_t *pmd_table; #endif - /* We pick one entry at random to throw out. Choosing the Least - * Recently Used might be better, but this is easy. */ + /* + * We pick one entry at random to throw out. Choosing the Least + * Recently Used might be better, but this is easy. + */ next = random32() % ARRAY_SIZE(cpu->lg->pgdirs); /* If it's never been allocated at all before, try now. */ if (!cpu->lg->pgdirs[next].pgdir) { @@ -587,8 +664,10 @@ static unsigned int new_pgdir(struct lg_cpu *cpu, next = cpu->cpu_pgd; else { #ifdef CONFIG_X86_PAE - /* In PAE mode, allocate a pmd page and populate the - * last pgd entry. */ + /* + * In PAE mode, allocate a pmd page and populate the + * last pgd entry. + */ pmd_table = (pmd_t *)get_zeroed_page(GFP_KERNEL); if (!pmd_table) { free_page((long)cpu->lg->pgdirs[next].pgdir); @@ -598,8 +677,10 @@ static unsigned int new_pgdir(struct lg_cpu *cpu, set_pgd(cpu->lg->pgdirs[next].pgdir + SWITCHER_PGD_INDEX, __pgd(__pa(pmd_table) | _PAGE_PRESENT)); - /* This is a blank page, so there are no kernel - * mappings: caller must map the stack! */ + /* + * This is a blank page, so there are no kernel + * mappings: caller must map the stack! + */ *blank_pgdir = 1; } #else @@ -615,19 +696,23 @@ static unsigned int new_pgdir(struct lg_cpu *cpu, return next; } -/*H:430 (iv) Switching page tables +/*H:430 + * (iv) Switching page tables * * Now we've seen all the page table setting and manipulation, let's see * what happens when the Guest changes page tables (ie. changes the top-level - * pgdir). This occurs on almost every context switch. */ + * pgdir). This occurs on almost every context switch. + */ void guest_new_pagetable(struct lg_cpu *cpu, unsigned long pgtable) { int newpgdir, repin = 0; /* Look to see if we have this one already. */ newpgdir = find_pgdir(cpu->lg, pgtable); - /* If not, we allocate or mug an existing one: if it's a fresh one, - * repin gets set to 1. */ + /* + * If not, we allocate or mug an existing one: if it's a fresh one, + * repin gets set to 1. + */ if (newpgdir == ARRAY_SIZE(cpu->lg->pgdirs)) newpgdir = new_pgdir(cpu, pgtable, &repin); /* Change the current pgd index to the new one. */ @@ -637,9 +722,11 @@ void guest_new_pagetable(struct lg_cpu *cpu, unsigned long pgtable) pin_stack_pages(cpu); } -/*H:470 Finally, a routine which throws away everything: all PGD entries in all +/*H:470 + * Finally, a routine which throws away everything: all PGD entries in all * the shadow page tables, including the Guest's kernel mappings. This is used - * when we destroy the Guest. */ + * when we destroy the Guest. + */ static void release_all_pagetables(struct lguest *lg) { unsigned int i, j; @@ -656,8 +743,10 @@ static void release_all_pagetables(struct lguest *lg) spgd = lg->pgdirs[i].pgdir + SWITCHER_PGD_INDEX; pmdpage = __va(pgd_pfn(*spgd) << PAGE_SHIFT); - /* And release the pmd entries of that pmd page, - * except for the switcher pmd. */ + /* + * And release the pmd entries of that pmd page, + * except for the switcher pmd. + */ for (k = 0; k < SWITCHER_PMD_INDEX; k++) release_pmd(&pmdpage[k]); #endif @@ -667,10 +756,12 @@ static void release_all_pagetables(struct lguest *lg) } } -/* We also throw away everything when a Guest tells us it's changed a kernel +/* + * We also throw away everything when a Guest tells us it's changed a kernel * mapping. Since kernel mappings are in every page table, it's easiest to * throw them all away. This traps the Guest in amber for a while as - * everything faults back in, but it's rare. */ + * everything faults back in, but it's rare. + */ void guest_pagetable_clear_all(struct lg_cpu *cpu) { release_all_pagetables(cpu->lg); @@ -678,15 +769,19 @@ void guest_pagetable_clear_all(struct lg_cpu *cpu) pin_stack_pages(cpu); } /*:*/ -/*M:009 Since we throw away all mappings when a kernel mapping changes, our + +/*M:009 + * Since we throw away all mappings when a kernel mapping changes, our * performance sucks for guests using highmem. In fact, a guest with * PAGE_OFFSET 0xc0000000 (the default) and more than about 700MB of RAM is * usually slower than a Guest with less memory. * * This, of course, cannot be fixed. It would take some kind of... well, I - * don't know, but the term "puissant code-fu" comes to mind. :*/ + * don't know, but the term "puissant code-fu" comes to mind. +:*/ -/*H:420 This is the routine which actually sets the page table entry for then +/*H:420 + * This is the routine which actually sets the page table entry for then * "idx"'th shadow page table. * * Normally, we can just throw out the old entry and replace it with 0: if they @@ -715,31 +810,36 @@ static void do_set_pte(struct lg_cpu *cpu, int idx, spmd = spmd_addr(cpu, *spgd, vaddr); if (pmd_flags(*spmd) & _PAGE_PRESENT) { #endif - /* Otherwise, we start by releasing - * the existing entry. */ + /* Otherwise, start by releasing the existing entry. */ pte_t *spte = spte_addr(cpu, *spgd, vaddr); release_pte(*spte); - /* If they're setting this entry as dirty or accessed, - * we might as well put that entry they've given us - * in now. This shaves 10% off a - * copy-on-write micro-benchmark. */ + /* + * If they're setting this entry as dirty or accessed, + * we might as well put that entry they've given us in + * now. This shaves 10% off a copy-on-write + * micro-benchmark. + */ if (pte_flags(gpte) & (_PAGE_DIRTY | _PAGE_ACCESSED)) { check_gpte(cpu, gpte); native_set_pte(spte, gpte_to_spte(cpu, gpte, pte_flags(gpte) & _PAGE_DIRTY)); - } else - /* Otherwise kill it and we can demand_page() - * it in later. */ + } else { + /* + * Otherwise kill it and we can demand_page() + * it in later. + */ native_set_pte(spte, __pte(0)); + } #ifdef CONFIG_X86_PAE } #endif } } -/*H:410 Updating a PTE entry is a little trickier. +/*H:410 + * Updating a PTE entry is a little trickier. * * We keep track of several different page tables (the Guest uses one for each * process, so it makes sense to cache at least a few). Each of these have @@ -748,12 +848,15 @@ static void do_set_pte(struct lg_cpu *cpu, int idx, * all the page tables, not just the current one. This is rare. * * The benefit is that when we have to track a new page table, we can keep all - * the kernel mappings. This speeds up context switch immensely. */ + * the kernel mappings. This speeds up context switch immensely. + */ void guest_set_pte(struct lg_cpu *cpu, unsigned long gpgdir, unsigned long vaddr, pte_t gpte) { - /* Kernel mappings must be changed on all top levels. Slow, but doesn't - * happen often. */ + /* + * Kernel mappings must be changed on all top levels. Slow, but doesn't + * happen often. + */ if (vaddr >= cpu->lg->kernel_address) { unsigned int i; for (i = 0; i < ARRAY_SIZE(cpu->lg->pgdirs); i++) @@ -802,12 +905,14 @@ void guest_set_pmd(struct lguest *lg, unsigned long pmdp, u32 idx) } #endif -/* Once we know how much memory we have we can construct simple identity - * (which set virtual == physical) and linear mappings - * which will get the Guest far enough into the boot to create its own. +/* + * Once we know how much memory we have we can construct simple identity (which + * set virtual == physical) and linear mappings which will get the Guest far + * enough into the boot to create its own. * * We lay them out of the way, just below the initrd (which is why we need to - * know its size here). */ + * know its size here). + */ static unsigned long setup_pagetables(struct lguest *lg, unsigned long mem, unsigned long initrd_size) @@ -825,8 +930,10 @@ static unsigned long setup_pagetables(struct lguest *lg, unsigned int phys_linear; #endif - /* We have mapped_pages frames to map, so we need - * linear_pages page tables to map them. */ + /* + * We have mapped_pages frames to map, so we need linear_pages page + * tables to map them. + */ mapped_pages = mem / PAGE_SIZE; linear_pages = (mapped_pages + PTRS_PER_PTE - 1) / PTRS_PER_PTE; @@ -839,8 +946,10 @@ static unsigned long setup_pagetables(struct lguest *lg, #ifdef CONFIG_X86_PAE pmds = (void *)linear - PAGE_SIZE; #endif - /* Linear mapping is easy: put every page's address into the - * mapping in order. */ + /* + * Linear mapping is easy: put every page's address into the + * mapping in order. + */ for (i = 0; i < mapped_pages; i++) { pte_t pte; pte = pfn_pte(i, __pgprot(_PAGE_PRESENT|_PAGE_RW|_PAGE_USER)); @@ -848,8 +957,10 @@ static unsigned long setup_pagetables(struct lguest *lg, return -EFAULT; } - /* The top level points to the linear page table pages above. - * We setup the identity and linear mappings here. */ + /* + * The top level points to the linear page table pages above. + * We setup the identity and linear mappings here. + */ #ifdef CONFIG_X86_PAE for (i = j = 0; i < mapped_pages && j < PTRS_PER_PMD; i += PTRS_PER_PTE, j++) { @@ -880,15 +991,19 @@ static unsigned long setup_pagetables(struct lguest *lg, } #endif - /* We return the top level (guest-physical) address: remember where - * this is. */ + /* + * We return the top level (guest-physical) address: remember where + * this is. + */ return (unsigned long)pgdir - mem_base; } -/*H:500 (vii) Setting up the page tables initially. +/*H:500 + * (vii) Setting up the page tables initially. * * When a Guest is first created, the Launcher tells us where the toplevel of - * its first page table is. We set some things up here: */ + * its first page table is. We set some things up here: + */ int init_guest_pagetable(struct lguest *lg) { u64 mem; @@ -898,14 +1013,18 @@ int init_guest_pagetable(struct lguest *lg) pgd_t *pgd; pmd_t *pmd_table; #endif - /* Get the Guest memory size and the ramdisk size from the boot header - * located at lg->mem_base (Guest address 0). */ + /* + * Get the Guest memory size and the ramdisk size from the boot header + * located at lg->mem_base (Guest address 0). + */ if (copy_from_user(&mem, &boot->e820_map[0].size, sizeof(mem)) || get_user(initrd_size, &boot->hdr.ramdisk_size)) return -EFAULT; - /* We start on the first shadow page table, and give it a blank PGD - * page. */ + /* + * We start on the first shadow page table, and give it a blank PGD + * page. + */ lg->pgdirs[0].gpgdir = setup_pagetables(lg, mem, initrd_size); if (IS_ERR_VALUE(lg->pgdirs[0].gpgdir)) return lg->pgdirs[0].gpgdir; @@ -931,17 +1050,21 @@ void page_table_guest_data_init(struct lg_cpu *cpu) /* We get the kernel address: above this is all kernel memory. */ if (get_user(cpu->lg->kernel_address, &cpu->lg->lguest_data->kernel_address) - /* We tell the Guest that it can't use the top 2 or 4 MB - * of virtual addresses used by the Switcher. */ + /* + * We tell the Guest that it can't use the top 2 or 4 MB + * of virtual addresses used by the Switcher. + */ || put_user(RESERVE_MEM * 1024 * 1024, &cpu->lg->lguest_data->reserve_mem) || put_user(cpu->lg->pgdirs[0].gpgdir, &cpu->lg->lguest_data->pgdir)) kill_guest(cpu, "bad guest page %p", cpu->lg->lguest_data); - /* In flush_user_mappings() we loop from 0 to + /* + * In flush_user_mappings() we loop from 0 to * "pgd_index(lg->kernel_address)". This assumes it won't hit the - * Switcher mappings, so check that now. */ + * Switcher mappings, so check that now. + */ #ifdef CONFIG_X86_PAE if (pgd_index(cpu->lg->kernel_address) == SWITCHER_PGD_INDEX && pmd_index(cpu->lg->kernel_address) == SWITCHER_PMD_INDEX) @@ -964,12 +1087,14 @@ void free_guest_pagetable(struct lguest *lg) free_page((long)lg->pgdirs[i].pgdir); } -/*H:480 (vi) Mapping the Switcher when the Guest is about to run. +/*H:480 + * (vi) Mapping the Switcher when the Guest is about to run. * * The Switcher and the two pages for this CPU need to be visible in the * Guest (and not the pages for other CPUs). We have the appropriate PTE pages * for each CPU already set up, we just need to hook them in now we know which - * Guest is about to run on this CPU. */ + * Guest is about to run on this CPU. + */ void map_switcher_in_guest(struct lg_cpu *cpu, struct lguest_pages *pages) { pte_t *switcher_pte_page = __get_cpu_var(switcher_pte_pages); @@ -990,20 +1115,24 @@ void map_switcher_in_guest(struct lg_cpu *cpu, struct lguest_pages *pages) #else pgd_t switcher_pgd; - /* Make the last PGD entry for this Guest point to the Switcher's PTE - * page for this CPU (with appropriate flags). */ + /* + * Make the last PGD entry for this Guest point to the Switcher's PTE + * page for this CPU (with appropriate flags). + */ switcher_pgd = __pgd(__pa(switcher_pte_page) | __PAGE_KERNEL_EXEC); cpu->lg->pgdirs[cpu->cpu_pgd].pgdir[SWITCHER_PGD_INDEX] = switcher_pgd; #endif - /* We also change the Switcher PTE page. When we're running the Guest, + /* + * We also change the Switcher PTE page. When we're running the Guest, * we want the Guest's "regs" page to appear where the first Switcher * page for this CPU is. This is an optimization: when the Switcher * saves the Guest registers, it saves them into the first page of this * CPU's "struct lguest_pages": if we make sure the Guest's register * page is already mapped there, we don't have to copy them out - * again. */ + * again. + */ pfn = __pa(cpu->regs_page) >> PAGE_SHIFT; native_set_pte(®s_pte, pfn_pte(pfn, PAGE_KERNEL)); native_set_pte(&switcher_pte_page[pte_index((unsigned long)pages)], @@ -1019,10 +1148,12 @@ static void free_switcher_pte_pages(void) free_page((long)switcher_pte_page(i)); } -/*H:520 Setting up the Switcher PTE page for given CPU is fairly easy, given +/*H:520 + * Setting up the Switcher PTE page for given CPU is fairly easy, given * the CPU number and the "struct page"s for the Switcher code itself. * - * Currently the Switcher is less than a page long, so "pages" is always 1. */ + * Currently the Switcher is less than a page long, so "pages" is always 1. + */ static __init void populate_switcher_pte_page(unsigned int cpu, struct page *switcher_page[], unsigned int pages) @@ -1043,13 +1174,16 @@ static __init void populate_switcher_pte_page(unsigned int cpu, native_set_pte(&pte[i], pfn_pte(page_to_pfn(switcher_page[i]), __pgprot(_PAGE_PRESENT|_PAGE_ACCESSED|_PAGE_RW))); - /* The second page contains the "struct lguest_ro_state", and is - * read-only. */ + /* + * The second page contains the "struct lguest_ro_state", and is + * read-only. + */ native_set_pte(&pte[i+1], pfn_pte(page_to_pfn(switcher_page[i+1]), __pgprot(_PAGE_PRESENT|_PAGE_ACCESSED))); } -/* We've made it through the page table code. Perhaps our tired brains are +/* + * We've made it through the page table code. Perhaps our tired brains are * still processing the details, or perhaps we're simply glad it's over. * * If nothing else, note that all this complexity in juggling shadow page tables @@ -1058,10 +1192,13 @@ static __init void populate_switcher_pte_page(unsigned int cpu, * uses exotic direct Guest pagetable manipulation, and why both Intel and AMD * have implemented shadow page table support directly into hardware. * - * There is just one file remaining in the Host. */ + * There is just one file remaining in the Host. + */ -/*H:510 At boot or module load time, init_pagetables() allocates and populates - * the Switcher PTE page for each CPU. */ +/*H:510 + * At boot or module load time, init_pagetables() allocates and populates + * the Switcher PTE page for each CPU. + */ __init int init_pagetables(struct page **switcher_page, unsigned int pages) { unsigned int i; diff --git a/drivers/lguest/segments.c b/drivers/lguest/segments.c index 482ed5a18750..951c57b0a7e0 100644 --- a/drivers/lguest/segments.c +++ b/drivers/lguest/segments.c @@ -1,4 +1,5 @@ -/*P:600 The x86 architecture has segments, which involve a table of descriptors +/*P:600 + * The x86 architecture has segments, which involve a table of descriptors * which can be used to do funky things with virtual address interpretation. * We originally used to use segments so the Guest couldn't alter the * Guest<->Host Switcher, and then we had to trim Guest segments, and restore @@ -8,7 +9,8 @@ * * In these modern times, the segment handling code consists of simple sanity * checks, and the worst you'll experience reading this code is butterfly-rash - * from frolicking through its parklike serenity. :*/ + * from frolicking through its parklike serenity. +:*/ #include "lg.h" /*H:600 @@ -41,10 +43,12 @@ * begin. */ -/* There are several entries we don't let the Guest set. The TSS entry is the +/* + * There are several entries we don't let the Guest set. The TSS entry is the * "Task State Segment" which controls all kinds of delicate things. The * LGUEST_CS and LGUEST_DS entries are reserved for the Switcher, and the - * the Guest can't be trusted to deal with double faults. */ + * the Guest can't be trusted to deal with double faults. + */ static bool ignored_gdt(unsigned int num) { return (num == GDT_ENTRY_TSS @@ -53,42 +57,52 @@ static bool ignored_gdt(unsigned int num) || num == GDT_ENTRY_DOUBLEFAULT_TSS); } -/*H:630 Once the Guest gave us new GDT entries, we fix them up a little. We +/*H:630 + * Once the Guest gave us new GDT entries, we fix them up a little. We * don't care if they're invalid: the worst that can happen is a General * Protection Fault in the Switcher when it restores a Guest segment register * which tries to use that entry. Then we kill the Guest for causing such a - * mess: the message will be "unhandled trap 256". */ + * mess: the message will be "unhandled trap 256". + */ static void fixup_gdt_table(struct lg_cpu *cpu, unsigned start, unsigned end) { unsigned int i; for (i = start; i < end; i++) { - /* We never copy these ones to real GDT, so we don't care what - * they say */ + /* + * We never copy these ones to real GDT, so we don't care what + * they say + */ if (ignored_gdt(i)) continue; - /* Segment descriptors contain a privilege level: the Guest is + /* + * Segment descriptors contain a privilege level: the Guest is * sometimes careless and leaves this as 0, even though it's - * running at privilege level 1. If so, we fix it here. */ + * running at privilege level 1. If so, we fix it here. + */ if ((cpu->arch.gdt[i].b & 0x00006000) == 0) cpu->arch.gdt[i].b |= (GUEST_PL << 13); - /* Each descriptor has an "accessed" bit. If we don't set it + /* + * Each descriptor has an "accessed" bit. If we don't set it * now, the CPU will try to set it when the Guest first loads * that entry into a segment register. But the GDT isn't - * writable by the Guest, so bad things can happen. */ + * writable by the Guest, so bad things can happen. + */ cpu->arch.gdt[i].b |= 0x00000100; } } -/*H:610 Like the IDT, we never simply use the GDT the Guest gives us. We keep +/*H:610 + * Like the IDT, we never simply use the GDT the Guest gives us. We keep * a GDT for each CPU, and copy across the Guest's entries each time we want to * run the Guest on that CPU. * * This routine is called at boot or modprobe time for each CPU to set up the * constant GDT entries: the ones which are the same no matter what Guest we're - * running. */ + * running. + */ void setup_default_gdt_entries(struct lguest_ro_state *state) { struct desc_struct *gdt = state->guest_gdt; @@ -98,30 +112,37 @@ void setup_default_gdt_entries(struct lguest_ro_state *state) gdt[GDT_ENTRY_LGUEST_CS] = FULL_EXEC_SEGMENT; gdt[GDT_ENTRY_LGUEST_DS] = FULL_SEGMENT; - /* The TSS segment refers to the TSS entry for this particular CPU. + /* + * The TSS segment refers to the TSS entry for this particular CPU. * Forgive the magic flags: the 0x8900 means the entry is Present, it's * privilege level 0 Available 386 TSS system segment, and the 0x67 - * means Saturn is eclipsed by Mercury in the twelfth house. */ + * means Saturn is eclipsed by Mercury in the twelfth house. + */ gdt[GDT_ENTRY_TSS].a = 0x00000067 | (tss << 16); gdt[GDT_ENTRY_TSS].b = 0x00008900 | (tss & 0xFF000000) | ((tss >> 16) & 0x000000FF); } -/* This routine sets up the initial Guest GDT for booting. All entries start - * as 0 (unusable). */ +/* + * This routine sets up the initial Guest GDT for booting. All entries start + * as 0 (unusable). + */ void setup_guest_gdt(struct lg_cpu *cpu) { - /* Start with full 0-4G segments... */ + /* + * Start with full 0-4G segments...except the Guest is allowed to use + * them, so set the privilege level appropriately in the flags. + */ cpu->arch.gdt[GDT_ENTRY_KERNEL_CS] = FULL_EXEC_SEGMENT; cpu->arch.gdt[GDT_ENTRY_KERNEL_DS] = FULL_SEGMENT; - /* ...except the Guest is allowed to use them, so set the privilege - * level appropriately in the flags. */ cpu->arch.gdt[GDT_ENTRY_KERNEL_CS].b |= (GUEST_PL << 13); cpu->arch.gdt[GDT_ENTRY_KERNEL_DS].b |= (GUEST_PL << 13); } -/*H:650 An optimization of copy_gdt(), for just the three "thead-local storage" - * entries. */ +/*H:650 + * An optimization of copy_gdt(), for just the three "thead-local storage" + * entries. + */ void copy_gdt_tls(const struct lg_cpu *cpu, struct desc_struct *gdt) { unsigned int i; @@ -130,26 +151,34 @@ void copy_gdt_tls(const struct lg_cpu *cpu, struct desc_struct *gdt) gdt[i] = cpu->arch.gdt[i]; } -/*H:640 When the Guest is run on a different CPU, or the GDT entries have - * changed, copy_gdt() is called to copy the Guest's GDT entries across to this - * CPU's GDT. */ +/*H:640 + * When the Guest is run on a different CPU, or the GDT entries have changed, + * copy_gdt() is called to copy the Guest's GDT entries across to this CPU's + * GDT. + */ void copy_gdt(const struct lg_cpu *cpu, struct desc_struct *gdt) { unsigned int i; - /* The default entries from setup_default_gdt_entries() are not - * replaced. See ignored_gdt() above. */ + /* + * The default entries from setup_default_gdt_entries() are not + * replaced. See ignored_gdt() above. + */ for (i = 0; i < GDT_ENTRIES; i++) if (!ignored_gdt(i)) gdt[i] = cpu->arch.gdt[i]; } -/*H:620 This is where the Guest asks us to load a new GDT entry - * (LHCALL_LOAD_GDT_ENTRY). We tweak the entry and copy it in. */ +/*H:620 + * This is where the Guest asks us to load a new GDT entry + * (LHCALL_LOAD_GDT_ENTRY). We tweak the entry and copy it in. + */ void load_guest_gdt_entry(struct lg_cpu *cpu, u32 num, u32 lo, u32 hi) { - /* We assume the Guest has the same number of GDT entries as the - * Host, otherwise we'd have to dynamically allocate the Guest GDT. */ + /* + * We assume the Guest has the same number of GDT entries as the + * Host, otherwise we'd have to dynamically allocate the Guest GDT. + */ if (num >= ARRAY_SIZE(cpu->arch.gdt)) kill_guest(cpu, "too many gdt entries %i", num); @@ -157,15 +186,19 @@ void load_guest_gdt_entry(struct lg_cpu *cpu, u32 num, u32 lo, u32 hi) cpu->arch.gdt[num].a = lo; cpu->arch.gdt[num].b = hi; fixup_gdt_table(cpu, num, num+1); - /* Mark that the GDT changed so the core knows it has to copy it again, - * even if the Guest is run on the same CPU. */ + /* + * Mark that the GDT changed so the core knows it has to copy it again, + * even if the Guest is run on the same CPU. + */ cpu->changed |= CHANGED_GDT; } -/* This is the fast-track version for just changing the three TLS entries. +/* + * This is the fast-track version for just changing the three TLS entries. * Remember that this happens on every context switch, so it's worth * optimizing. But wouldn't it be neater to have a single hypercall to cover - * both cases? */ + * both cases? + */ void guest_load_tls(struct lg_cpu *cpu, unsigned long gtls) { struct desc_struct *tls = &cpu->arch.gdt[GDT_ENTRY_TLS_MIN]; @@ -175,7 +208,6 @@ void guest_load_tls(struct lg_cpu *cpu, unsigned long gtls) /* Note that just the TLS entries have changed. */ cpu->changed |= CHANGED_GDT_TLS; } -/*:*/ /*H:660 * With this, we have finished the Host. diff --git a/drivers/lguest/x86/core.c b/drivers/lguest/x86/core.c index eaf722fe309a..96f7d88ec7f8 100644 --- a/drivers/lguest/x86/core.c +++ b/drivers/lguest/x86/core.c @@ -17,13 +17,15 @@ * along with this program; if not, write to the Free Software * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. */ -/*P:450 This file contains the x86-specific lguest code. It used to be all +/*P:450 + * This file contains the x86-specific lguest code. It used to be all * mixed in with drivers/lguest/core.c but several foolhardy code slashers * wrestled most of the dependencies out to here in preparation for porting * lguest to other architectures (see what I mean by foolhardy?). * * This also contains a couple of non-obvious setup and teardown pieces which - * were implemented after days of debugging pain. :*/ + * were implemented after days of debugging pain. +:*/ #include #include #include @@ -82,25 +84,33 @@ static DEFINE_PER_CPU(struct lg_cpu *, last_cpu); */ static void copy_in_guest_info(struct lg_cpu *cpu, struct lguest_pages *pages) { - /* Copying all this data can be quite expensive. We usually run the + /* + * Copying all this data can be quite expensive. We usually run the * same Guest we ran last time (and that Guest hasn't run anywhere else * meanwhile). If that's not the case, we pretend everything in the - * Guest has changed. */ + * Guest has changed. + */ if (__get_cpu_var(last_cpu) != cpu || cpu->last_pages != pages) { __get_cpu_var(last_cpu) = cpu; cpu->last_pages = pages; cpu->changed = CHANGED_ALL; } - /* These copies are pretty cheap, so we do them unconditionally: */ - /* Save the current Host top-level page directory. */ + /* + * These copies are pretty cheap, so we do them unconditionally: */ + /* Save the current Host top-level page directory. + */ pages->state.host_cr3 = __pa(current->mm->pgd); - /* Set up the Guest's page tables to see this CPU's pages (and no - * other CPU's pages). */ + /* + * Set up the Guest's page tables to see this CPU's pages (and no + * other CPU's pages). + */ map_switcher_in_guest(cpu, pages); - /* Set up the two "TSS" members which tell the CPU what stack to use + /* + * Set up the two "TSS" members which tell the CPU what stack to use * for traps which do directly into the Guest (ie. traps at privilege - * level 1). */ + * level 1). + */ pages->state.guest_tss.sp1 = cpu->esp1; pages->state.guest_tss.ss1 = cpu->ss1; @@ -125,40 +135,53 @@ static void run_guest_once(struct lg_cpu *cpu, struct lguest_pages *pages) /* This is a dummy value we need for GCC's sake. */ unsigned int clobber; - /* Copy the guest-specific information into this CPU's "struct - * lguest_pages". */ + /* + * Copy the guest-specific information into this CPU's "struct + * lguest_pages". + */ copy_in_guest_info(cpu, pages); - /* Set the trap number to 256 (impossible value). If we fault while + /* + * Set the trap number to 256 (impossible value). If we fault while * switching to the Guest (bad segment registers or bug), this will - * cause us to abort the Guest. */ + * cause us to abort the Guest. + */ cpu->regs->trapnum = 256; - /* Now: we push the "eflags" register on the stack, then do an "lcall". + /* + * Now: we push the "eflags" register on the stack, then do an "lcall". * This is how we change from using the kernel code segment to using * the dedicated lguest code segment, as well as jumping into the * Switcher. * * The lcall also pushes the old code segment (KERNEL_CS) onto the * stack, then the address of this call. This stack layout happens to - * exactly match the stack layout created by an interrupt... */ + * exactly match the stack layout created by an interrupt... + */ asm volatile("pushf; lcall *lguest_entry" - /* This is how we tell GCC that %eax ("a") and %ebx ("b") - * are changed by this routine. The "=" means output. */ + /* + * This is how we tell GCC that %eax ("a") and %ebx ("b") + * are changed by this routine. The "=" means output. + */ : "=a"(clobber), "=b"(clobber) - /* %eax contains the pages pointer. ("0" refers to the + /* + * %eax contains the pages pointer. ("0" refers to the * 0-th argument above, ie "a"). %ebx contains the * physical address of the Guest's top-level page - * directory. */ + * directory. + */ : "0"(pages), "1"(__pa(cpu->lg->pgdirs[cpu->cpu_pgd].pgdir)) - /* We tell gcc that all these registers could change, + /* + * We tell gcc that all these registers could change, * which means we don't have to save and restore them in - * the Switcher. */ + * the Switcher. + */ : "memory", "%edx", "%ecx", "%edi", "%esi"); } /*:*/ -/*M:002 There are hooks in the scheduler which we can register to tell when we +/*M:002 + * There are hooks in the scheduler which we can register to tell when we * get kicked off the CPU (preempt_notifier_register()). This would allow us * to lazily disable SYSENTER which would regain some performance, and should * also simplify copy_in_guest_info(). Note that we'd still need to restore @@ -166,56 +189,72 @@ static void run_guest_once(struct lg_cpu *cpu, struct lguest_pages *pages) * * We could also try using this hooks for PGE, but that might be too expensive. * - * The hooks were designed for KVM, but we can also put them to good use. :*/ + * The hooks were designed for KVM, but we can also put them to good use. +:*/ -/*H:040 This is the i386-specific code to setup and run the Guest. Interrupts - * are disabled: we own the CPU. */ +/*H:040 + * This is the i386-specific code to setup and run the Guest. Interrupts + * are disabled: we own the CPU. + */ void lguest_arch_run_guest(struct lg_cpu *cpu) { - /* Remember the awfully-named TS bit? If the Guest has asked to set it + /* + * Remember the awfully-named TS bit? If the Guest has asked to set it * we set it now, so we can trap and pass that trap to the Guest if it - * uses the FPU. */ + * uses the FPU. + */ if (cpu->ts) unlazy_fpu(current); - /* SYSENTER is an optimized way of doing system calls. We can't allow + /* + * SYSENTER is an optimized way of doing system calls. We can't allow * it because it always jumps to privilege level 0. A normal Guest * won't try it because we don't advertise it in CPUID, but a malicious * Guest (or malicious Guest userspace program) could, so we tell the - * CPU to disable it before running the Guest. */ + * CPU to disable it before running the Guest. + */ if (boot_cpu_has(X86_FEATURE_SEP)) wrmsr(MSR_IA32_SYSENTER_CS, 0, 0); - /* Now we actually run the Guest. It will return when something + /* + * Now we actually run the Guest. It will return when something * interesting happens, and we can examine its registers to see what it - * was doing. */ + * was doing. + */ run_guest_once(cpu, lguest_pages(raw_smp_processor_id())); - /* Note that the "regs" structure contains two extra entries which are + /* + * Note that the "regs" structure contains two extra entries which are * not really registers: a trap number which says what interrupt or * trap made the switcher code come back, and an error code which some - * traps set. */ + * traps set. + */ /* Restore SYSENTER if it's supposed to be on. */ if (boot_cpu_has(X86_FEATURE_SEP)) wrmsr(MSR_IA32_SYSENTER_CS, __KERNEL_CS, 0); - /* If the Guest page faulted, then the cr2 register will tell us the + /* + * If the Guest page faulted, then the cr2 register will tell us the * bad virtual address. We have to grab this now, because once we * re-enable interrupts an interrupt could fault and thus overwrite - * cr2, or we could even move off to a different CPU. */ + * cr2, or we could even move off to a different CPU. + */ if (cpu->regs->trapnum == 14) cpu->arch.last_pagefault = read_cr2(); - /* Similarly, if we took a trap because the Guest used the FPU, + /* + * Similarly, if we took a trap because the Guest used the FPU, * we have to restore the FPU it expects to see. * math_state_restore() may sleep and we may even move off to * a different CPU. So all the critical stuff should be done - * before this. */ + * before this. + */ else if (cpu->regs->trapnum == 7) math_state_restore(); } -/*H:130 Now we've examined the hypercall code; our Guest can make requests. +/*H:130 + * Now we've examined the hypercall code; our Guest can make requests. * Our Guest is usually so well behaved; it never tries to do things it isn't * allowed to, and uses hypercalls instead. Unfortunately, Linux's paravirtual * infrastructure isn't quite complete, because it doesn't contain replacements @@ -225,26 +264,33 @@ void lguest_arch_run_guest(struct lg_cpu *cpu) * * When the Guest uses one of these instructions, we get a trap (General * Protection Fault) and come here. We see if it's one of those troublesome - * instructions and skip over it. We return true if we did. */ + * instructions and skip over it. We return true if we did. + */ static int emulate_insn(struct lg_cpu *cpu) { u8 insn; unsigned int insnlen = 0, in = 0, shift = 0; - /* The eip contains the *virtual* address of the Guest's instruction: - * guest_pa just subtracts the Guest's page_offset. */ + /* + * The eip contains the *virtual* address of the Guest's instruction: + * guest_pa just subtracts the Guest's page_offset. + */ unsigned long physaddr = guest_pa(cpu, cpu->regs->eip); - /* This must be the Guest kernel trying to do something, not userspace! + /* + * This must be the Guest kernel trying to do something, not userspace! * The bottom two bits of the CS segment register are the privilege - * level. */ + * level. + */ if ((cpu->regs->cs & 3) != GUEST_PL) return 0; /* Decoding x86 instructions is icky. */ insn = lgread(cpu, physaddr, u8); - /* 0x66 is an "operand prefix". It means it's using the upper 16 bits - of the eax register. */ + /* + * 0x66 is an "operand prefix". It means it's using the upper 16 bits + * of the eax register. + */ if (insn == 0x66) { shift = 16; /* The instruction is 1 byte so far, read the next byte. */ @@ -252,8 +298,10 @@ static int emulate_insn(struct lg_cpu *cpu) insn = lgread(cpu, physaddr + insnlen, u8); } - /* We can ignore the lower bit for the moment and decode the 4 opcodes - * we need to emulate. */ + /* + * We can ignore the lower bit for the moment and decode the 4 opcodes + * we need to emulate. + */ switch (insn & 0xFE) { case 0xE4: /* in ,%al */ insnlen += 2; @@ -274,9 +322,11 @@ static int emulate_insn(struct lg_cpu *cpu) return 0; } - /* If it was an "IN" instruction, they expect the result to be read + /* + * If it was an "IN" instruction, they expect the result to be read * into %eax, so we change %eax. We always return all-ones, which - * traditionally means "there's nothing there". */ + * traditionally means "there's nothing there". + */ if (in) { /* Lower bit tells is whether it's a 16 or 32 bit access */ if (insn & 0x1) @@ -290,7 +340,8 @@ static int emulate_insn(struct lg_cpu *cpu) return 1; } -/* Our hypercalls mechanism used to be based on direct software interrupts. +/* + * Our hypercalls mechanism used to be based on direct software interrupts. * After Anthony's "Refactor hypercall infrastructure" kvm patch, we decided to * change over to using kvm hypercalls. * @@ -318,16 +369,20 @@ static int emulate_insn(struct lg_cpu *cpu) */ static void rewrite_hypercall(struct lg_cpu *cpu) { - /* This are the opcodes we use to patch the Guest. The opcode for "int + /* + * This are the opcodes we use to patch the Guest. The opcode for "int * $0x1f" is "0xcd 0x1f" but vmcall instruction is 3 bytes long, so we - * complete the sequence with a NOP (0x90). */ + * complete the sequence with a NOP (0x90). + */ u8 insn[3] = {0xcd, 0x1f, 0x90}; __lgwrite(cpu, guest_pa(cpu, cpu->regs->eip), insn, sizeof(insn)); - /* The above write might have caused a copy of that page to be made + /* + * The above write might have caused a copy of that page to be made * (if it was read-only). We need to make sure the Guest has * up-to-date pagetables. As this doesn't happen often, we can just - * drop them all. */ + * drop them all. + */ guest_pagetable_clear_all(cpu); } @@ -335,9 +390,11 @@ static bool is_hypercall(struct lg_cpu *cpu) { u8 insn[3]; - /* This must be the Guest kernel trying to do something. + /* + * This must be the Guest kernel trying to do something. * The bottom two bits of the CS segment register are the privilege - * level. */ + * level. + */ if ((cpu->regs->cs & 3) != GUEST_PL) return false; @@ -351,86 +408,105 @@ void lguest_arch_handle_trap(struct lg_cpu *cpu) { switch (cpu->regs->trapnum) { case 13: /* We've intercepted a General Protection Fault. */ - /* Check if this was one of those annoying IN or OUT + /* + * Check if this was one of those annoying IN or OUT * instructions which we need to emulate. If so, we just go - * back into the Guest after we've done it. */ + * back into the Guest after we've done it. + */ if (cpu->regs->errcode == 0) { if (emulate_insn(cpu)) return; } - /* If KVM is active, the vmcall instruction triggers a - * General Protection Fault. Normally it triggers an - * invalid opcode fault (6): */ + /* + * If KVM is active, the vmcall instruction triggers a General + * Protection Fault. Normally it triggers an invalid opcode + * fault (6): + */ case 6: - /* We need to check if ring == GUEST_PL and - * faulting instruction == vmcall. */ + /* + * We need to check if ring == GUEST_PL and faulting + * instruction == vmcall. + */ if (is_hypercall(cpu)) { rewrite_hypercall(cpu); return; } break; case 14: /* We've intercepted a Page Fault. */ - /* The Guest accessed a virtual address that wasn't mapped. + /* + * The Guest accessed a virtual address that wasn't mapped. * This happens a lot: we don't actually set up most of the page * tables for the Guest at all when we start: as it runs it asks * for more and more, and we set them up as required. In this * case, we don't even tell the Guest that the fault happened. * * The errcode tells whether this was a read or a write, and - * whether kernel or userspace code. */ + * whether kernel or userspace code. + */ if (demand_page(cpu, cpu->arch.last_pagefault, cpu->regs->errcode)) return; - /* OK, it's really not there (or not OK): the Guest needs to + /* + * OK, it's really not there (or not OK): the Guest needs to * know. We write out the cr2 value so it knows where the * fault occurred. * * Note that if the Guest were really messed up, this could * happen before it's done the LHCALL_LGUEST_INIT hypercall, so - * lg->lguest_data could be NULL */ + * lg->lguest_data could be NULL + */ if (cpu->lg->lguest_data && put_user(cpu->arch.last_pagefault, &cpu->lg->lguest_data->cr2)) kill_guest(cpu, "Writing cr2"); break; case 7: /* We've intercepted a Device Not Available fault. */ - /* If the Guest doesn't want to know, we already restored the - * Floating Point Unit, so we just continue without telling - * it. */ + /* + * If the Guest doesn't want to know, we already restored the + * Floating Point Unit, so we just continue without telling it. + */ if (!cpu->ts) return; break; case 32 ... 255: - /* These values mean a real interrupt occurred, in which case + /* + * These values mean a real interrupt occurred, in which case * the Host handler has already been run. We just do a * friendly check if another process should now be run, then - * return to run the Guest again */ + * return to run the Guest again + */ cond_resched(); return; case LGUEST_TRAP_ENTRY: - /* Our 'struct hcall_args' maps directly over our regs: we set - * up the pointer now to indicate a hypercall is pending. */ + /* + * Our 'struct hcall_args' maps directly over our regs: we set + * up the pointer now to indicate a hypercall is pending. + */ cpu->hcall = (struct hcall_args *)cpu->regs; return; } /* We didn't handle the trap, so it needs to go to the Guest. */ if (!deliver_trap(cpu, cpu->regs->trapnum)) - /* If the Guest doesn't have a handler (either it hasn't + /* + * If the Guest doesn't have a handler (either it hasn't * registered any yet, or it's one of the faults we don't let - * it handle), it dies with this cryptic error message. */ + * it handle), it dies with this cryptic error message. + */ kill_guest(cpu, "unhandled trap %li at %#lx (%#lx)", cpu->regs->trapnum, cpu->regs->eip, cpu->regs->trapnum == 14 ? cpu->arch.last_pagefault : cpu->regs->errcode); } -/* Now we can look at each of the routines this calls, in increasing order of +/* + * Now we can look at each of the routines this calls, in increasing order of * complexity: do_hypercalls(), emulate_insn(), maybe_do_interrupt(), * deliver_trap() and demand_page(). After all those, we'll be ready to * examine the Switcher, and our philosophical understanding of the Host/Guest - * duality will be complete. :*/ + * duality will be complete. +:*/ static void adjust_pge(void *on) { if (on) @@ -439,13 +515,16 @@ static void adjust_pge(void *on) write_cr4(read_cr4() & ~X86_CR4_PGE); } -/*H:020 Now the Switcher is mapped and every thing else is ready, we need to do - * some more i386-specific initialization. */ +/*H:020 + * Now the Switcher is mapped and every thing else is ready, we need to do + * some more i386-specific initialization. + */ void __init lguest_arch_host_init(void) { int i; - /* Most of the i386/switcher.S doesn't care that it's been moved; on + /* + * Most of the i386/switcher.S doesn't care that it's been moved; on * Intel, jumps are relative, and it doesn't access any references to * external code or data. * @@ -453,7 +532,8 @@ void __init lguest_arch_host_init(void) * addresses are placed in a table (default_idt_entries), so we need to * update the table with the new addresses. switcher_offset() is a * convenience function which returns the distance between the - * compiled-in switcher code and the high-mapped copy we just made. */ + * compiled-in switcher code and the high-mapped copy we just made. + */ for (i = 0; i < IDT_ENTRIES; i++) default_idt_entries[i] += switcher_offset(); @@ -468,63 +548,81 @@ void __init lguest_arch_host_init(void) for_each_possible_cpu(i) { /* lguest_pages() returns this CPU's two pages. */ struct lguest_pages *pages = lguest_pages(i); - /* This is a convenience pointer to make the code fit one - * statement to a line. */ + /* This is a convenience pointer to make the code neater. */ struct lguest_ro_state *state = &pages->state; - /* The Global Descriptor Table: the Host has a different one + /* + * The Global Descriptor Table: the Host has a different one * for each CPU. We keep a descriptor for the GDT which says * where it is and how big it is (the size is actually the last - * byte, not the size, hence the "-1"). */ + * byte, not the size, hence the "-1"). + */ state->host_gdt_desc.size = GDT_SIZE-1; state->host_gdt_desc.address = (long)get_cpu_gdt_table(i); - /* All CPUs on the Host use the same Interrupt Descriptor + /* + * All CPUs on the Host use the same Interrupt Descriptor * Table, so we just use store_idt(), which gets this CPU's IDT - * descriptor. */ + * descriptor. + */ store_idt(&state->host_idt_desc); - /* The descriptors for the Guest's GDT and IDT can be filled + /* + * The descriptors for the Guest's GDT and IDT can be filled * out now, too. We copy the GDT & IDT into ->guest_gdt and - * ->guest_idt before actually running the Guest. */ + * ->guest_idt before actually running the Guest. + */ state->guest_idt_desc.size = sizeof(state->guest_idt)-1; state->guest_idt_desc.address = (long)&state->guest_idt; state->guest_gdt_desc.size = sizeof(state->guest_gdt)-1; state->guest_gdt_desc.address = (long)&state->guest_gdt; - /* We know where we want the stack to be when the Guest enters + /* + * We know where we want the stack to be when the Guest enters * the Switcher: in pages->regs. The stack grows upwards, so - * we start it at the end of that structure. */ + * we start it at the end of that structure. + */ state->guest_tss.sp0 = (long)(&pages->regs + 1); - /* And this is the GDT entry to use for the stack: we keep a - * couple of special LGUEST entries. */ + /* + * And this is the GDT entry to use for the stack: we keep a + * couple of special LGUEST entries. + */ state->guest_tss.ss0 = LGUEST_DS; - /* x86 can have a finegrained bitmap which indicates what I/O + /* + * x86 can have a finegrained bitmap which indicates what I/O * ports the process can use. We set it to the end of our - * structure, meaning "none". */ + * structure, meaning "none". + */ state->guest_tss.io_bitmap_base = sizeof(state->guest_tss); - /* Some GDT entries are the same across all Guests, so we can - * set them up now. */ + /* + * Some GDT entries are the same across all Guests, so we can + * set them up now. + */ setup_default_gdt_entries(state); /* Most IDT entries are the same for all Guests, too.*/ setup_default_idt_entries(state, default_idt_entries); - /* The Host needs to be able to use the LGUEST segments on this - * CPU, too, so put them in the Host GDT. */ + /* + * The Host needs to be able to use the LGUEST segments on this + * CPU, too, so put them in the Host GDT. + */ get_cpu_gdt_table(i)[GDT_ENTRY_LGUEST_CS] = FULL_EXEC_SEGMENT; get_cpu_gdt_table(i)[GDT_ENTRY_LGUEST_DS] = FULL_SEGMENT; } - /* In the Switcher, we want the %cs segment register to use the + /* + * In the Switcher, we want the %cs segment register to use the * LGUEST_CS GDT entry: we've put that in the Host and Guest GDTs, so * it will be undisturbed when we switch. To change %cs and jump we - * need this structure to feed to Intel's "lcall" instruction. */ + * need this structure to feed to Intel's "lcall" instruction. + */ lguest_entry.offset = (long)switch_to_guest + switcher_offset(); lguest_entry.segment = LGUEST_CS; - /* Finally, we need to turn off "Page Global Enable". PGE is an + /* + * Finally, we need to turn off "Page Global Enable". PGE is an * optimization where page table entries are specially marked to show * they never change. The Host kernel marks all the kernel pages this * way because it's always present, even when userspace is running. @@ -534,16 +632,21 @@ void __init lguest_arch_host_init(void) * you'll get really weird bugs that you'll chase for two days. * * I used to turn PGE off every time we switched to the Guest and back - * on when we return, but that slowed the Switcher down noticibly. */ + * on when we return, but that slowed the Switcher down noticibly. + */ - /* We don't need the complexity of CPUs coming and going while we're - * doing this. */ + /* + * We don't need the complexity of CPUs coming and going while we're + * doing this. + */ get_online_cpus(); if (cpu_has_pge) { /* We have a broader idea of "global". */ /* Remember that this was originally set (for cleanup). */ cpu_had_pge = 1; - /* adjust_pge is a helper function which sets or unsets the PGE - * bit on its CPU, depending on the argument (0 == unset). */ + /* + * adjust_pge is a helper function which sets or unsets the PGE + * bit on its CPU, depending on the argument (0 == unset). + */ on_each_cpu(adjust_pge, (void *)0, 1); /* Turn off the feature in the global feature set. */ clear_cpu_cap(&boot_cpu_data, X86_FEATURE_PGE); @@ -590,26 +693,32 @@ int lguest_arch_init_hypercalls(struct lg_cpu *cpu) { u32 tsc_speed; - /* The pointer to the Guest's "struct lguest_data" is the only argument. - * We check that address now. */ + /* + * The pointer to the Guest's "struct lguest_data" is the only argument. + * We check that address now. + */ if (!lguest_address_ok(cpu->lg, cpu->hcall->arg1, sizeof(*cpu->lg->lguest_data))) return -EFAULT; - /* Having checked it, we simply set lg->lguest_data to point straight + /* + * Having checked it, we simply set lg->lguest_data to point straight * into the Launcher's memory at the right place and then use * copy_to_user/from_user from now on, instead of lgread/write. I put * this in to show that I'm not immune to writing stupid - * optimizations. */ + * optimizations. + */ cpu->lg->lguest_data = cpu->lg->mem_base + cpu->hcall->arg1; - /* We insist that the Time Stamp Counter exist and doesn't change with + /* + * We insist that the Time Stamp Counter exist and doesn't change with * cpu frequency. Some devious chip manufacturers decided that TSC * changes could be handled in software. I decided that time going * backwards might be good for benchmarks, but it's bad for users. * * We also insist that the TSC be stable: the kernel detects unreliable - * TSCs for its own purposes, and we use that here. */ + * TSCs for its own purposes, and we use that here. + */ if (boot_cpu_has(X86_FEATURE_CONSTANT_TSC) && !check_tsc_unstable()) tsc_speed = tsc_khz; else @@ -625,38 +734,47 @@ int lguest_arch_init_hypercalls(struct lg_cpu *cpu) } /*:*/ -/*L:030 lguest_arch_setup_regs() +/*L:030 + * lguest_arch_setup_regs() * * Most of the Guest's registers are left alone: we used get_zeroed_page() to - * allocate the structure, so they will be 0. */ + * allocate the structure, so they will be 0. + */ void lguest_arch_setup_regs(struct lg_cpu *cpu, unsigned long start) { struct lguest_regs *regs = cpu->regs; - /* There are four "segment" registers which the Guest needs to boot: + /* + * There are four "segment" registers which the Guest needs to boot: * The "code segment" register (cs) refers to the kernel code segment * __KERNEL_CS, and the "data", "extra" and "stack" segment registers * refer to the kernel data segment __KERNEL_DS. * * The privilege level is packed into the lower bits. The Guest runs - * at privilege level 1 (GUEST_PL).*/ + * at privilege level 1 (GUEST_PL). + */ regs->ds = regs->es = regs->ss = __KERNEL_DS|GUEST_PL; regs->cs = __KERNEL_CS|GUEST_PL; - /* The "eflags" register contains miscellaneous flags. Bit 1 (0x002) + /* + * The "eflags" register contains miscellaneous flags. Bit 1 (0x002) * is supposed to always be "1". Bit 9 (0x200) controls whether * interrupts are enabled. We always leave interrupts enabled while - * running the Guest. */ + * running the Guest. + */ regs->eflags = X86_EFLAGS_IF | 0x2; - /* The "Extended Instruction Pointer" register says where the Guest is - * running. */ + /* + * The "Extended Instruction Pointer" register says where the Guest is + * running. + */ regs->eip = start; - /* %esi points to our boot information, at physical address 0, so don't - * touch it. */ + /* + * %esi points to our boot information, at physical address 0, so don't + * touch it. + */ - /* There are a couple of GDT entries the Guest expects when first - * booting. */ + /* There are a couple of GDT entries the Guest expects at boot. */ setup_guest_gdt(cpu); } diff --git a/drivers/lguest/x86/switcher_32.S b/drivers/lguest/x86/switcher_32.S index 3fc15318a80f..6dec09793836 100644 --- a/drivers/lguest/x86/switcher_32.S +++ b/drivers/lguest/x86/switcher_32.S @@ -1,12 +1,15 @@ -/*P:900 This is the Switcher: code which sits at 0xFFC00000 astride both the +/*P:900 + * This is the Switcher: code which sits at 0xFFC00000 astride both the * Host and Guest to do the low-level Guest<->Host switch. It is as simple as * it can be made, but it's naturally very specific to x86. * * You have now completed Preparation. If this has whet your appetite; if you * are feeling invigorated and refreshed then the next, more challenging stage - * can be found in "make Guest". :*/ + * can be found in "make Guest". + :*/ -/*M:012 Lguest is meant to be simple: my rule of thumb is that 1% more LOC must +/*M:012 + * Lguest is meant to be simple: my rule of thumb is that 1% more LOC must * gain at least 1% more performance. Since neither LOC nor performance can be * measured beforehand, it generally means implementing a feature then deciding * if it's worth it. And once it's implemented, who can say no? @@ -31,11 +34,14 @@ * Host (which is actually really easy). * * Two questions remain. Would the performance gain outweigh the complexity? - * And who would write the verse documenting it? :*/ + * And who would write the verse documenting it? +:*/ -/*M:011 Lguest64 handles NMI. This gave me NMI envy (until I looked at their +/*M:011 + * Lguest64 handles NMI. This gave me NMI envy (until I looked at their * code). It's worth doing though, since it would let us use oprofile in the - * Host when a Guest is running. :*/ + * Host when a Guest is running. +:*/ /*S:100 * Welcome to the Switcher itself! diff --git a/include/linux/lguest.h b/include/linux/lguest.h index dbf2479e808e..0a3a11afd64b 100644 --- a/include/linux/lguest.h +++ b/include/linux/lguest.h @@ -1,5 +1,7 @@ -/* Things the lguest guest needs to know. Note: like all lguest interfaces, - * this is subject to wild and random change between versions. */ +/* + * Things the lguest guest needs to know. Note: like all lguest interfaces, + * this is subject to wild and random change between versions. + */ #ifndef _LINUX_LGUEST_H #define _LINUX_LGUEST_H @@ -11,32 +13,42 @@ #define LG_CLOCK_MIN_DELTA 100UL #define LG_CLOCK_MAX_DELTA ULONG_MAX -/*G:031 The second method of communicating with the Host is to via "struct +/*G:031 + * The second method of communicating with the Host is to via "struct * lguest_data". Once the Guest's initialization hypercall tells the Host where - * this is, the Guest and Host both publish information in it. :*/ + * this is, the Guest and Host both publish information in it. +:*/ struct lguest_data { - /* 512 == enabled (same as eflags in normal hardware). The Guest - * changes interrupts so often that a hypercall is too slow. */ + /* + * 512 == enabled (same as eflags in normal hardware). The Guest + * changes interrupts so often that a hypercall is too slow. + */ unsigned int irq_enabled; /* Fine-grained interrupt disabling by the Guest */ DECLARE_BITMAP(blocked_interrupts, LGUEST_IRQS); - /* The Host writes the virtual address of the last page fault here, + /* + * The Host writes the virtual address of the last page fault here, * which saves the Guest a hypercall. CR2 is the native register where - * this address would normally be found. */ + * this address would normally be found. + */ unsigned long cr2; /* Wallclock time set by the Host. */ struct timespec time; - /* Interrupt pending set by the Host. The Guest should do a hypercall - * if it re-enables interrupts and sees this set (to X86_EFLAGS_IF). */ + /* + * Interrupt pending set by the Host. The Guest should do a hypercall + * if it re-enables interrupts and sees this set (to X86_EFLAGS_IF). + */ int irq_pending; - /* Async hypercall ring. Instead of directly making hypercalls, we can + /* + * Async hypercall ring. Instead of directly making hypercalls, we can * place them in here for processing the next time the Host wants. - * This batching can be quite efficient. */ + * This batching can be quite efficient. + */ /* 0xFF == done (set by Host), 0 == pending (set by Guest). */ u8 hcall_status[LHCALL_RING_SIZE]; diff --git a/include/linux/lguest_launcher.h b/include/linux/lguest_launcher.h index bfefbdf7498a..495203ff221c 100644 --- a/include/linux/lguest_launcher.h +++ b/include/linux/lguest_launcher.h @@ -29,8 +29,10 @@ struct lguest_device_desc { __u8 type; /* The number of virtqueues (first in config array) */ __u8 num_vq; - /* The number of bytes of feature bits. Multiply by 2: one for host - * features and one for Guest acknowledgements. */ + /* + * The number of bytes of feature bits. Multiply by 2: one for host + * features and one for Guest acknowledgements. + */ __u8 feature_len; /* The number of bytes of the config array after virtqueues. */ __u8 config_len; @@ -39,8 +41,10 @@ struct lguest_device_desc { __u8 config[0]; }; -/*D:135 This is how we expect the device configuration field for a virtqueue - * to be laid out in config space. */ +/*D:135 + * This is how we expect the device configuration field for a virtqueue + * to be laid out in config space. + */ struct lguest_vqconfig { /* The number of entries in the virtio_ring */ __u16 num; @@ -61,7 +65,9 @@ enum lguest_req LHREQ_EVENTFD, /* + address, fd. */ }; -/* The alignment to use between consumer and producer parts of vring. - * x86 pagesize for historical reasons. */ +/* + * The alignment to use between consumer and producer parts of vring. + * x86 pagesize for historical reasons. + */ #define LGUEST_VRING_ALIGN 4096 #endif /* _LINUX_LGUEST_LAUNCHER */ -- cgit v1.2.3 From 1842f23c05b6a866be831aa60bc8a8731c58ddd0 Mon Sep 17 00:00:00 2001 From: Rusty Russell Date: Thu, 30 Jul 2009 16:03:46 -0600 Subject: lguest and virtio: cleanup struct definitions to Linux style. I've been doing this for years, and akpm picked me up on it about 12 months ago. lguest partly serves as example code, so let's do it Right. Also, remove two unused fields in struct vblk_info in the example launcher. Signed-off-by: Rusty Russell Cc: Ingo Molnar --- Documentation/lguest/lguest.c | 21 +++++---------------- drivers/lguest/lg.h | 9 +++------ drivers/lguest/lguest_device.c | 3 +-- include/linux/lguest.h | 3 +-- include/linux/virtio_blk.h | 6 ++---- include/linux/virtio_config.h | 3 +-- include/linux/virtio_net.h | 6 ++---- include/linux/virtio_ring.h | 12 ++++-------- 8 files changed, 19 insertions(+), 44 deletions(-) (limited to 'include') diff --git a/Documentation/lguest/lguest.c b/Documentation/lguest/lguest.c index 45163651b519..950cde6d6e58 100644 --- a/Documentation/lguest/lguest.c +++ b/Documentation/lguest/lguest.c @@ -93,8 +93,7 @@ static int lguest_fd; static unsigned int __thread cpu_id; /* This is our list of devices. */ -struct device_list -{ +struct device_list { /* Counter to assign interrupt numbers. */ unsigned int next_irq; @@ -114,8 +113,7 @@ struct device_list static struct device_list devices; /* The device structure describes a single device. */ -struct device -{ +struct device { /* The linked-list pointer. */ struct device *next; @@ -140,8 +138,7 @@ struct device }; /* The virtqueue structure describes a queue attached to a device. */ -struct virtqueue -{ +struct virtqueue { struct virtqueue *next; /* Which device owns me. */ @@ -779,8 +776,7 @@ static void add_used_and_trigger(struct virtqueue *vq, unsigned head, int len) * * We associate some data with the console for our exit hack. */ -struct console_abort -{ +struct console_abort { /* How many times have they hit ^C? */ int count; /* When did they start? */ @@ -1570,20 +1566,13 @@ static void setup_tun_net(char *arg) /*:*/ /* This hangs off device->priv. */ -struct vblk_info -{ +struct vblk_info { /* The size of the file. */ off64_t len; /* The file descriptor for the file. */ int fd; - /* IO thread listens on this file descriptor [0]. */ - int workpipe[2]; - - /* IO thread writes to this file descriptor to mark it done, then - * Launcher triggers interrupt to Guest. */ - int done_fd; }; /*L:210 diff --git a/drivers/lguest/lg.h b/drivers/lguest/lg.h index 74c0db691b53..bc28745d05af 100644 --- a/drivers/lguest/lg.h +++ b/drivers/lguest/lg.h @@ -16,15 +16,13 @@ void free_pagetables(void); int init_pagetables(struct page **switcher_page, unsigned int pages); -struct pgdir -{ +struct pgdir { unsigned long gpgdir; pgd_t *pgdir; }; /* We have two pages shared with guests, per cpu. */ -struct lguest_pages -{ +struct lguest_pages { /* This is the stack page mapped rw in guest */ char spare[PAGE_SIZE - sizeof(struct lguest_regs)]; struct lguest_regs regs; @@ -89,8 +87,7 @@ struct lg_eventfd_map { }; /* The private info the thread maintains about the guest. */ -struct lguest -{ +struct lguest { struct lguest_data __user *lguest_data; struct lg_cpu cpus[NR_CPUS]; unsigned int nr_cpus; diff --git a/drivers/lguest/lguest_device.c b/drivers/lguest/lguest_device.c index 1401c1ace1ec..b6200bc39b58 100644 --- a/drivers/lguest/lguest_device.c +++ b/drivers/lguest/lguest_device.c @@ -207,8 +207,7 @@ static void lg_reset(struct virtio_device *vdev) */ /*D:140 This is the information we remember about each virtqueue. */ -struct lguest_vq_info -{ +struct lguest_vq_info { /* A copy of the information contained in the device config. */ struct lguest_vqconfig config; diff --git a/include/linux/lguest.h b/include/linux/lguest.h index 0a3a11afd64b..2fb1dcbcb5aa 100644 --- a/include/linux/lguest.h +++ b/include/linux/lguest.h @@ -18,8 +18,7 @@ * lguest_data". Once the Guest's initialization hypercall tells the Host where * this is, the Guest and Host both publish information in it. :*/ -struct lguest_data -{ +struct lguest_data { /* * 512 == enabled (same as eflags in normal hardware). The Guest * changes interrupts so often that a hypercall is too slow. diff --git a/include/linux/virtio_blk.h b/include/linux/virtio_blk.h index be7d255fc7cf..8dab9f2b8832 100644 --- a/include/linux/virtio_blk.h +++ b/include/linux/virtio_blk.h @@ -20,8 +20,7 @@ #define VIRTIO_BLK_ID_BYTES (sizeof(__u16[256])) /* IDENTIFY DATA */ -struct virtio_blk_config -{ +struct virtio_blk_config { /* The capacity (in 512-byte sectors). */ __u64 capacity; /* The maximum segment size (if VIRTIO_BLK_F_SIZE_MAX) */ @@ -50,8 +49,7 @@ struct virtio_blk_config #define VIRTIO_BLK_T_BARRIER 0x80000000 /* This is the first element of the read scatter-gather list. */ -struct virtio_blk_outhdr -{ +struct virtio_blk_outhdr { /* VIRTIO_BLK_T* */ __u32 type; /* io priority. */ diff --git a/include/linux/virtio_config.h b/include/linux/virtio_config.h index 99f514575f6a..e547e3c8ee9a 100644 --- a/include/linux/virtio_config.h +++ b/include/linux/virtio_config.h @@ -79,8 +79,7 @@ * the dev->feature bits if it wants. */ typedef void vq_callback_t(struct virtqueue *); -struct virtio_config_ops -{ +struct virtio_config_ops { void (*get)(struct virtio_device *vdev, unsigned offset, void *buf, unsigned len); void (*set)(struct virtio_device *vdev, unsigned offset, diff --git a/include/linux/virtio_net.h b/include/linux/virtio_net.h index 9c543d6ac535..d8dd539c9f48 100644 --- a/include/linux/virtio_net.h +++ b/include/linux/virtio_net.h @@ -31,8 +31,7 @@ #define VIRTIO_NET_S_LINK_UP 1 /* Link is up */ -struct virtio_net_config -{ +struct virtio_net_config { /* The config defining mac address (if VIRTIO_NET_F_MAC) */ __u8 mac[6]; /* See VIRTIO_NET_F_STATUS and VIRTIO_NET_S_* above */ @@ -41,8 +40,7 @@ struct virtio_net_config /* This is the first element of the scatter-gather list. If you don't * specify GSO or CSUM features, you can simply ignore the header. */ -struct virtio_net_hdr -{ +struct virtio_net_hdr { #define VIRTIO_NET_HDR_F_NEEDS_CSUM 1 // Use csum_start, csum_offset __u8 flags; #define VIRTIO_NET_HDR_GSO_NONE 0 // Not a GSO frame diff --git a/include/linux/virtio_ring.h b/include/linux/virtio_ring.h index 693e0ec5afa6..e4d144b132b5 100644 --- a/include/linux/virtio_ring.h +++ b/include/linux/virtio_ring.h @@ -30,8 +30,7 @@ #define VIRTIO_RING_F_INDIRECT_DESC 28 /* Virtio ring descriptors: 16 bytes. These can chain together via "next". */ -struct vring_desc -{ +struct vring_desc { /* Address (guest-physical). */ __u64 addr; /* Length. */ @@ -42,24 +41,21 @@ struct vring_desc __u16 next; }; -struct vring_avail -{ +struct vring_avail { __u16 flags; __u16 idx; __u16 ring[]; }; /* u32 is used here for ids for padding reasons. */ -struct vring_used_elem -{ +struct vring_used_elem { /* Index of start of used descriptor chain. */ __u32 id; /* Total length of the descriptor chain which was used (written to) */ __u32 len; }; -struct vring_used -{ +struct vring_used { __u16 flags; __u16 idx; struct vring_used_elem ring[]; -- cgit v1.2.3 From e5f5ccb646bc6009572b5c23201b5e81638ff150 Mon Sep 17 00:00:00 2001 From: Daniel Mack Date: Thu, 23 Jul 2009 20:35:53 +0200 Subject: power_supply: get_by_name and set_charged functionality This adds a function that indicates that a battery is fully charged. It also includes functions to get a power_supply device from the class of registered devices by name reference. These can be used to find a specific battery to call power_supply_set_battery_charged() on. Some battery drivers might need this information to calibrate themselves. Signed-off-by: Daniel Mack Cc: Ian Molton Cc: Anton Vorontsov Cc: Matt Reimer Signed-off-by: Anton Vorontsov --- drivers/power/power_supply_core.c | 28 ++++++++++++++++++++++++++++ include/linux/power_supply.h | 3 +++ 2 files changed, 31 insertions(+) (limited to 'include') diff --git a/drivers/power/power_supply_core.c b/drivers/power/power_supply_core.c index 12cd6e36ff1d..cce75b40b435 100644 --- a/drivers/power/power_supply_core.c +++ b/drivers/power/power_supply_core.c @@ -116,6 +116,34 @@ int power_supply_is_system_supplied(void) } EXPORT_SYMBOL_GPL(power_supply_is_system_supplied); +int power_supply_set_battery_charged(struct power_supply *psy) +{ + if (psy->type == POWER_SUPPLY_TYPE_BATTERY && psy->set_charged) { + psy->set_charged(psy); + return 0; + } + + return -EINVAL; +} +EXPORT_SYMBOL_GPL(power_supply_set_battery_charged); + +static int power_supply_match_device_by_name(struct device *dev, void *data) +{ + const char *name = data; + struct power_supply *psy = dev_get_drvdata(dev); + + return strcmp(psy->name, name) == 0; +} + +struct power_supply *power_supply_get_by_name(char *name) +{ + struct device *dev = class_find_device(power_supply_class, NULL, name, + power_supply_match_device_by_name); + + return dev ? dev_get_drvdata(dev) : NULL; +} +EXPORT_SYMBOL_GPL(power_supply_get_by_name); + int power_supply_register(struct device *parent, struct power_supply *psy) { int rc = 0; diff --git a/include/linux/power_supply.h b/include/linux/power_supply.h index 4c7c6fc35487..b5d096d3a9be 100644 --- a/include/linux/power_supply.h +++ b/include/linux/power_supply.h @@ -144,6 +144,7 @@ struct power_supply { enum power_supply_property psp, union power_supply_propval *val); void (*external_power_changed)(struct power_supply *psy); + void (*set_charged)(struct power_supply *psy); /* For APM emulation, think legacy userspace. */ int use_for_apm; @@ -183,8 +184,10 @@ struct power_supply_info { int use_for_apm; }; +extern struct power_supply *power_supply_get_by_name(char *name); extern void power_supply_changed(struct power_supply *psy); extern int power_supply_am_i_supplied(struct power_supply *psy); +extern int power_supply_set_battery_charged(struct power_supply *psy); #if defined(CONFIG_POWER_SUPPLY) || defined(CONFIG_POWER_SUPPLY_MODULE) extern int power_supply_is_system_supplied(void); -- cgit v1.2.3 From 9aada7ac047f789ffb27540cc1695989897b2dfe Mon Sep 17 00:00:00 2001 From: Hannes Eder Date: Thu, 30 Jul 2009 14:29:44 -0700 Subject: IPVS: use pr_fmt While being at it cleanup whitespace. Signed-off-by: Hannes Eder Signed-off-by: David S. Miller --- include/net/ip_vs.h | 119 ++++++++++++++++---------------- net/netfilter/ipvs/ip_vs_app.c | 3 + net/netfilter/ipvs/ip_vs_conn.c | 3 + net/netfilter/ipvs/ip_vs_core.c | 3 + net/netfilter/ipvs/ip_vs_ctl.c | 3 + net/netfilter/ipvs/ip_vs_dh.c | 3 + net/netfilter/ipvs/ip_vs_est.c | 4 ++ net/netfilter/ipvs/ip_vs_ftp.c | 3 + net/netfilter/ipvs/ip_vs_lblc.c | 3 + net/netfilter/ipvs/ip_vs_lblcr.c | 3 + net/netfilter/ipvs/ip_vs_lc.c | 3 + net/netfilter/ipvs/ip_vs_nq.c | 3 + net/netfilter/ipvs/ip_vs_proto.c | 7 +- net/netfilter/ipvs/ip_vs_proto_ah_esp.c | 7 +- net/netfilter/ipvs/ip_vs_proto_tcp.c | 3 + net/netfilter/ipvs/ip_vs_proto_udp.c | 3 + net/netfilter/ipvs/ip_vs_rr.c | 3 + net/netfilter/ipvs/ip_vs_sched.c | 3 + net/netfilter/ipvs/ip_vs_sed.c | 3 + net/netfilter/ipvs/ip_vs_sh.c | 3 + net/netfilter/ipvs/ip_vs_sync.c | 3 + net/netfilter/ipvs/ip_vs_wlc.c | 3 + net/netfilter/ipvs/ip_vs_wrr.c | 3 + net/netfilter/ipvs/ip_vs_xmit.c | 3 + 24 files changed, 134 insertions(+), 63 deletions(-) (limited to 'include') diff --git a/include/net/ip_vs.h b/include/net/ip_vs.h index bbae1e87efcd..910820327bc4 100644 --- a/include/net/ip_vs.h +++ b/include/net/ip_vs.h @@ -99,47 +99,47 @@ static inline const char *ip_vs_dbg_addr(int af, char *buf, size_t buf_len, return &buf[*idx - len]; } -#define IP_VS_DBG_BUF(level, msg...) \ - do { \ - char ip_vs_dbg_buf[160]; \ - int ip_vs_dbg_idx = 0; \ - if (level <= ip_vs_get_debug_level()) \ - printk(KERN_DEBUG "IPVS: " msg); \ - } while (0) -#define IP_VS_ERR_BUF(msg...) \ - do { \ - char ip_vs_dbg_buf[160]; \ - int ip_vs_dbg_idx = 0; \ - printk(KERN_ERR "IPVS: " msg); \ - } while (0) +#define IP_VS_DBG_BUF(level, msg, ...) \ + do { \ + char ip_vs_dbg_buf[160]; \ + int ip_vs_dbg_idx = 0; \ + if (level <= ip_vs_get_debug_level()) \ + printk(KERN_DEBUG pr_fmt(msg), ##__VA_ARGS__); \ + } while (0) +#define IP_VS_ERR_BUF(msg...) \ + do { \ + char ip_vs_dbg_buf[160]; \ + int ip_vs_dbg_idx = 0; \ + pr_err(msg); \ + } while (0) /* Only use from within IP_VS_DBG_BUF() or IP_VS_ERR_BUF macros */ -#define IP_VS_DBG_ADDR(af, addr) \ - ip_vs_dbg_addr(af, ip_vs_dbg_buf, \ - sizeof(ip_vs_dbg_buf), addr, \ - &ip_vs_dbg_idx) - -#define IP_VS_DBG(level, msg...) \ - do { \ - if (level <= ip_vs_get_debug_level()) \ - printk(KERN_DEBUG "IPVS: " msg); \ - } while (0) -#define IP_VS_DBG_RL(msg...) \ - do { \ - if (net_ratelimit()) \ - printk(KERN_DEBUG "IPVS: " msg); \ - } while (0) -#define IP_VS_DBG_PKT(level, pp, skb, ofs, msg) \ - do { \ - if (level <= ip_vs_get_debug_level()) \ - pp->debug_packet(pp, skb, ofs, msg); \ - } while (0) -#define IP_VS_DBG_RL_PKT(level, pp, skb, ofs, msg) \ - do { \ - if (level <= ip_vs_get_debug_level() && \ - net_ratelimit()) \ - pp->debug_packet(pp, skb, ofs, msg); \ - } while (0) +#define IP_VS_DBG_ADDR(af, addr) \ + ip_vs_dbg_addr(af, ip_vs_dbg_buf, \ + sizeof(ip_vs_dbg_buf), addr, \ + &ip_vs_dbg_idx) + +#define IP_VS_DBG(level, msg, ...) \ + do { \ + if (level <= ip_vs_get_debug_level()) \ + printk(KERN_DEBUG pr_fmt(msg), ##__VA_ARGS__); \ + } while (0) +#define IP_VS_DBG_RL(msg, ...) \ + do { \ + if (net_ratelimit()) \ + printk(KERN_DEBUG pr_fmt(msg), ##__VA_ARGS__); \ + } while (0) +#define IP_VS_DBG_PKT(level, pp, skb, ofs, msg) \ + do { \ + if (level <= ip_vs_get_debug_level()) \ + pp->debug_packet(pp, skb, ofs, msg); \ + } while (0) +#define IP_VS_DBG_RL_PKT(level, pp, skb, ofs, msg) \ + do { \ + if (level <= ip_vs_get_debug_level() && \ + net_ratelimit()) \ + pp->debug_packet(pp, skb, ofs, msg); \ + } while (0) #else /* NO DEBUGGING at ALL */ #define IP_VS_DBG_BUF(level, msg...) do {} while (0) #define IP_VS_ERR_BUF(msg...) do {} while (0) @@ -150,29 +150,30 @@ static inline const char *ip_vs_dbg_addr(int af, char *buf, size_t buf_len, #endif #define IP_VS_BUG() BUG() -#define IP_VS_ERR(msg...) printk(KERN_ERR "IPVS: " msg) -#define IP_VS_INFO(msg...) printk(KERN_INFO "IPVS: " msg) -#define IP_VS_WARNING(msg...) \ - printk(KERN_WARNING "IPVS: " msg) -#define IP_VS_ERR_RL(msg...) \ - do { \ - if (net_ratelimit()) \ - printk(KERN_ERR "IPVS: " msg); \ - } while (0) +#define IP_VS_ERR(msg...) pr_err(msg) +#define IP_VS_INFO(msg...) pr_info(msg) +#define IP_VS_WARNING(msg...) pr_warning(msg) +#define IP_VS_ERR_RL(msg...) \ + do { \ + if (net_ratelimit()) \ + pr_err(msg); \ + } while (0) #ifdef CONFIG_IP_VS_DEBUG #define EnterFunction(level) \ - do { \ - if (level <= ip_vs_get_debug_level()) \ - printk(KERN_DEBUG "Enter: %s, %s line %i\n", \ - __func__, __FILE__, __LINE__); \ - } while (0) -#define LeaveFunction(level) \ - do { \ - if (level <= ip_vs_get_debug_level()) \ - printk(KERN_DEBUG "Leave: %s, %s line %i\n", \ - __func__, __FILE__, __LINE__); \ - } while (0) + do { \ + if (level <= ip_vs_get_debug_level()) \ + printk(KERN_DEBUG \ + pr_fmt("Enter: %s, %s line %i\n"), \ + __func__, __FILE__, __LINE__); \ + } while (0) +#define LeaveFunction(level) \ + do { \ + if (level <= ip_vs_get_debug_level()) \ + printk(KERN_DEBUG \ + pr_fmt("Leave: %s, %s line %i\n"), \ + __func__, __FILE__, __LINE__); \ + } while (0) #else #define EnterFunction(level) do {} while (0) #define LeaveFunction(level) do {} while (0) diff --git a/net/netfilter/ipvs/ip_vs_app.c b/net/netfilter/ipvs/ip_vs_app.c index 201b8ea3020d..c1781f80daf2 100644 --- a/net/netfilter/ipvs/ip_vs_app.c +++ b/net/netfilter/ipvs/ip_vs_app.c @@ -18,6 +18,9 @@ * */ +#define KMSG_COMPONENT "IPVS" +#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt + #include #include #include diff --git a/net/netfilter/ipvs/ip_vs_conn.c b/net/netfilter/ipvs/ip_vs_conn.c index 77bfdfeb966e..4173d7b1d4cc 100644 --- a/net/netfilter/ipvs/ip_vs_conn.c +++ b/net/netfilter/ipvs/ip_vs_conn.c @@ -22,6 +22,9 @@ * */ +#define KMSG_COMPONENT "IPVS" +#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt + #include #include #include diff --git a/net/netfilter/ipvs/ip_vs_core.c b/net/netfilter/ipvs/ip_vs_core.c index 8dddb17a947a..6811dcaca0f6 100644 --- a/net/netfilter/ipvs/ip_vs_core.c +++ b/net/netfilter/ipvs/ip_vs_core.c @@ -24,6 +24,9 @@ * */ +#define KMSG_COMPONENT "IPVS" +#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt + #include #include #include diff --git a/net/netfilter/ipvs/ip_vs_ctl.c b/net/netfilter/ipvs/ip_vs_ctl.c index 2d24d81474ce..e6133ea1ea4c 100644 --- a/net/netfilter/ipvs/ip_vs_ctl.c +++ b/net/netfilter/ipvs/ip_vs_ctl.c @@ -18,6 +18,9 @@ * */ +#define KMSG_COMPONENT "IPVS" +#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt + #include #include #include diff --git a/net/netfilter/ipvs/ip_vs_dh.c b/net/netfilter/ipvs/ip_vs_dh.c index a9dac74bb13f..d0c0594d1e2e 100644 --- a/net/netfilter/ipvs/ip_vs_dh.c +++ b/net/netfilter/ipvs/ip_vs_dh.c @@ -35,6 +35,9 @@ * */ +#define KMSG_COMPONENT "IPVS" +#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt + #include #include #include diff --git a/net/netfilter/ipvs/ip_vs_est.c b/net/netfilter/ipvs/ip_vs_est.c index 2eb2860dabb5..702b53ca937c 100644 --- a/net/netfilter/ipvs/ip_vs_est.c +++ b/net/netfilter/ipvs/ip_vs_est.c @@ -11,6 +11,10 @@ * Changes: * */ + +#define KMSG_COMPONENT "IPVS" +#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt + #include #include #include diff --git a/net/netfilter/ipvs/ip_vs_ftp.c b/net/netfilter/ipvs/ip_vs_ftp.c index 428edbf481cc..9c16a3f64c1b 100644 --- a/net/netfilter/ipvs/ip_vs_ftp.c +++ b/net/netfilter/ipvs/ip_vs_ftp.c @@ -22,6 +22,9 @@ * */ +#define KMSG_COMPONENT "IPVS" +#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt + #include #include #include diff --git a/net/netfilter/ipvs/ip_vs_lblc.c b/net/netfilter/ipvs/ip_vs_lblc.c index 3eb5e2660c49..98fb185d890b 100644 --- a/net/netfilter/ipvs/ip_vs_lblc.c +++ b/net/netfilter/ipvs/ip_vs_lblc.c @@ -39,6 +39,9 @@ * me to write this module. */ +#define KMSG_COMPONENT "IPVS" +#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt + #include #include #include diff --git a/net/netfilter/ipvs/ip_vs_lblcr.c b/net/netfilter/ipvs/ip_vs_lblcr.c index c04ce56c7f0f..5f5e5f4bad5e 100644 --- a/net/netfilter/ipvs/ip_vs_lblcr.c +++ b/net/netfilter/ipvs/ip_vs_lblcr.c @@ -37,6 +37,9 @@ * */ +#define KMSG_COMPONENT "IPVS" +#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt + #include #include #include diff --git a/net/netfilter/ipvs/ip_vs_lc.c b/net/netfilter/ipvs/ip_vs_lc.c index d0dadc8a65fd..4ecd5e19c39a 100644 --- a/net/netfilter/ipvs/ip_vs_lc.c +++ b/net/netfilter/ipvs/ip_vs_lc.c @@ -14,6 +14,9 @@ * */ +#define KMSG_COMPONENT "IPVS" +#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt + #include #include diff --git a/net/netfilter/ipvs/ip_vs_nq.c b/net/netfilter/ipvs/ip_vs_nq.c index 694952db5026..2224478bdea8 100644 --- a/net/netfilter/ipvs/ip_vs_nq.c +++ b/net/netfilter/ipvs/ip_vs_nq.c @@ -31,6 +31,9 @@ * */ +#define KMSG_COMPONENT "IPVS" +#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt + #include #include diff --git a/net/netfilter/ipvs/ip_vs_proto.c b/net/netfilter/ipvs/ip_vs_proto.c index a01520e3d6b8..a95bc4021c90 100644 --- a/net/netfilter/ipvs/ip_vs_proto.c +++ b/net/netfilter/ipvs/ip_vs_proto.c @@ -13,6 +13,9 @@ * */ +#define KMSG_COMPONENT "IPVS" +#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt + #include #include #include @@ -181,7 +184,7 @@ ip_vs_tcpudp_debug_packet_v4(struct ip_vs_protocol *pp, &ih->daddr, ntohs(pptr[1])); } - printk(KERN_DEBUG "IPVS: %s: %s\n", msg, buf); + pr_debug("%s: %s\n", msg, buf); } #ifdef CONFIG_IP_VS_IPV6 @@ -215,7 +218,7 @@ ip_vs_tcpudp_debug_packet_v6(struct ip_vs_protocol *pp, &ih->daddr, ntohs(pptr[1])); } - printk(KERN_DEBUG "IPVS: %s: %s\n", msg, buf); + pr_debug("%s: %s\n", msg, buf); } #endif diff --git a/net/netfilter/ipvs/ip_vs_proto_ah_esp.c b/net/netfilter/ipvs/ip_vs_proto_ah_esp.c index 79f56c1e7c19..c30b43c36cd7 100644 --- a/net/netfilter/ipvs/ip_vs_proto_ah_esp.c +++ b/net/netfilter/ipvs/ip_vs_proto_ah_esp.c @@ -10,6 +10,9 @@ * */ +#define KMSG_COMPONENT "IPVS" +#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt + #include #include #include @@ -138,7 +141,7 @@ ah_esp_debug_packet_v4(struct ip_vs_protocol *pp, const struct sk_buff *skb, sprintf(buf, "%s %pI4->%pI4", pp->name, &ih->saddr, &ih->daddr); - printk(KERN_DEBUG "IPVS: %s: %s\n", msg, buf); + pr_debug("%s: %s\n", msg, buf); } #ifdef CONFIG_IP_VS_IPV6 @@ -156,7 +159,7 @@ ah_esp_debug_packet_v6(struct ip_vs_protocol *pp, const struct sk_buff *skb, sprintf(buf, "%s %pI6->%pI6", pp->name, &ih->saddr, &ih->daddr); - printk(KERN_DEBUG "IPVS: %s: %s\n", msg, buf); + pr_debug("%s: %s\n", msg, buf); } #endif diff --git a/net/netfilter/ipvs/ip_vs_proto_tcp.c b/net/netfilter/ipvs/ip_vs_proto_tcp.c index 8cba41802850..c36c80d3a2b4 100644 --- a/net/netfilter/ipvs/ip_vs_proto_tcp.c +++ b/net/netfilter/ipvs/ip_vs_proto_tcp.c @@ -13,6 +13,9 @@ * */ +#define KMSG_COMPONENT "IPVS" +#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt + #include #include #include /* for tcphdr */ diff --git a/net/netfilter/ipvs/ip_vs_proto_udp.c b/net/netfilter/ipvs/ip_vs_proto_udp.c index d2930a71084b..96ebe40bc537 100644 --- a/net/netfilter/ipvs/ip_vs_proto_udp.c +++ b/net/netfilter/ipvs/ip_vs_proto_udp.c @@ -13,6 +13,9 @@ * */ +#define KMSG_COMPONENT "IPVS" +#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt + #include #include #include diff --git a/net/netfilter/ipvs/ip_vs_rr.c b/net/netfilter/ipvs/ip_vs_rr.c index 2d16ab7f8c1e..b01007e1c11e 100644 --- a/net/netfilter/ipvs/ip_vs_rr.c +++ b/net/netfilter/ipvs/ip_vs_rr.c @@ -19,6 +19,9 @@ * */ +#define KMSG_COMPONENT "IPVS" +#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt + #include #include diff --git a/net/netfilter/ipvs/ip_vs_sched.c b/net/netfilter/ipvs/ip_vs_sched.c index a46ad9e35016..87bc5ea0ef29 100644 --- a/net/netfilter/ipvs/ip_vs_sched.c +++ b/net/netfilter/ipvs/ip_vs_sched.c @@ -17,6 +17,9 @@ * */ +#define KMSG_COMPONENT "IPVS" +#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt + #include #include #include diff --git a/net/netfilter/ipvs/ip_vs_sed.c b/net/netfilter/ipvs/ip_vs_sed.c index 20e4657d2f3b..4f745dd86dd8 100644 --- a/net/netfilter/ipvs/ip_vs_sed.c +++ b/net/netfilter/ipvs/ip_vs_sed.c @@ -35,6 +35,9 @@ * */ +#define KMSG_COMPONENT "IPVS" +#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt + #include #include diff --git a/net/netfilter/ipvs/ip_vs_sh.c b/net/netfilter/ipvs/ip_vs_sh.c index 75709ebeb630..fb4d2d23f2fe 100644 --- a/net/netfilter/ipvs/ip_vs_sh.c +++ b/net/netfilter/ipvs/ip_vs_sh.c @@ -32,6 +32,9 @@ * */ +#define KMSG_COMPONENT "IPVS" +#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt + #include #include #include diff --git a/net/netfilter/ipvs/ip_vs_sync.c b/net/netfilter/ipvs/ip_vs_sync.c index 5c48378a852f..cc04c99815fd 100644 --- a/net/netfilter/ipvs/ip_vs_sync.c +++ b/net/netfilter/ipvs/ip_vs_sync.c @@ -17,6 +17,9 @@ * Justin Ossevoort : Fix endian problem on sync message size. */ +#define KMSG_COMPONENT "IPVS" +#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt + #include #include #include diff --git a/net/netfilter/ipvs/ip_vs_wlc.c b/net/netfilter/ipvs/ip_vs_wlc.c index 8e942565b47d..bbddfdb10db2 100644 --- a/net/netfilter/ipvs/ip_vs_wlc.c +++ b/net/netfilter/ipvs/ip_vs_wlc.c @@ -19,6 +19,9 @@ * */ +#define KMSG_COMPONENT "IPVS" +#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt + #include #include diff --git a/net/netfilter/ipvs/ip_vs_wrr.c b/net/netfilter/ipvs/ip_vs_wrr.c index f7d74ef1ecf9..c39ebb6c5a54 100644 --- a/net/netfilter/ipvs/ip_vs_wrr.c +++ b/net/netfilter/ipvs/ip_vs_wrr.c @@ -18,6 +18,9 @@ * */ +#define KMSG_COMPONENT "IPVS" +#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt + #include #include #include diff --git a/net/netfilter/ipvs/ip_vs_xmit.c b/net/netfilter/ipvs/ip_vs_xmit.c index 5874657af7f2..061e76dfdad9 100644 --- a/net/netfilter/ipvs/ip_vs_xmit.c +++ b/net/netfilter/ipvs/ip_vs_xmit.c @@ -13,6 +13,9 @@ * */ +#define KMSG_COMPONENT "IPVS" +#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt + #include #include /* for tcphdr */ #include -- cgit v1.2.3 From 78ddb2785980fc01b129c3547463266cae9c6ca9 Mon Sep 17 00:00:00 2001 From: Ben Dooks Date: Thu, 30 Jul 2009 23:23:34 +0100 Subject: ARM: PL093: Header file for PL093 SSMC PrimeCell Header to define the standard registers for an PL093 SSMC memory controller. Signed-off-by: Ben Dooks Signed-off-by: Ben Dooks --- include/linux/amba/pl093.h | 80 ++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 80 insertions(+) create mode 100644 include/linux/amba/pl093.h (limited to 'include') diff --git a/include/linux/amba/pl093.h b/include/linux/amba/pl093.h new file mode 100644 index 000000000000..2983e3671adb --- /dev/null +++ b/include/linux/amba/pl093.h @@ -0,0 +1,80 @@ +/* linux/amba/pl093.h + * + * Copyright (c) 2008 Simtec Electronics + * http://armlinux.simtec.co.uk/ + * Ben Dooks + * + * AMBA PL093 SSMC (synchronous static memory controller) + * See DDI0236.pdf (r0p4) for more details + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. +*/ + +#define SMB_BANK(x) ((x) * 0x20) /* each bank control set is 0x20 apart */ + +/* Offsets for SMBxxxxRy registers */ + +#define SMBIDCYR (0x00) +#define SMBWSTRDR (0x04) +#define SMBWSTWRR (0x08) +#define SMBWSTOENR (0x0C) +#define SMBWSTWENR (0x10) +#define SMBCR (0x14) +#define SMBSR (0x18) +#define SMBWSTBRDR (0x1C) + +/* Masks for SMB registers */ +#define IDCY_MASK (0xf) +#define WSTRD_MASK (0xf) +#define WSTWR_MASK (0xf) +#define WSTOEN_MASK (0xf) +#define WSTWEN_MASK (0xf) + +/* Notes from datasheet: + * WSTOEN <= WSTRD + * WSTWEN <= WSTWR + * + * WSTOEN is not used with nWAIT + */ + +/* SMBCR bit definitions */ +#define SMBCR_BIWRITEEN (1 << 21) +#define SMBCR_ADDRVALIDWRITEEN (1 << 20) +#define SMBCR_SYNCWRITE (1 << 17) +#define SMBCR_BMWRITE (1 << 16) +#define SMBCR_WRAPREAD (1 << 14) +#define SMBCR_BIREADEN (1 << 13) +#define SMBCR_ADDRVALIDREADEN (1 << 12) +#define SMBCR_SYNCREAD (1 << 9) +#define SMBCR_BMREAD (1 << 8) +#define SMBCR_SMBLSPOL (1 << 6) +#define SMBCR_WP (1 << 3) +#define SMBCR_WAITEN (1 << 2) +#define SMBCR_WAITPOL (1 << 1) +#define SMBCR_RBLE (1 << 0) + +#define SMBCR_BURSTLENWRITE_MASK (3 << 18) +#define SMBCR_BURSTLENWRITE_4 (0 << 18) +#define SMBCR_BURSTLENWRITE_8 (1 << 18) +#define SMBCR_BURSTLENWRITE_RESERVED (2 << 18) +#define SMBCR_BURSTLENWRITE_CONTINUOUS (3 << 18) + +#define SMBCR_BURSTLENREAD_MASK (3 << 10) +#define SMBCR_BURSTLENREAD_4 (0 << 10) +#define SMBCR_BURSTLENREAD_8 (1 << 10) +#define SMBCR_BURSTLENREAD_16 (2 << 10) +#define SMBCR_BURSTLENREAD_CONTINUOUS (3 << 10) + +#define SMBCR_MW_MASK (3 << 4) +#define SMBCR_MW_8BIT (0 << 4) +#define SMBCR_MW_16BIT (1 << 4) +#define SMBCR_MW_M32BIT (2 << 4) + +/* SSMC status registers */ +#define SSMCCSR (0x200) +#define SSMCCR (0x204) +#define SSMCITCR (0x208) +#define SSMCITIP (0x20C) +#define SSMCITIOP (0x210) -- cgit v1.2.3 From a33bc5c15154c835aae26f16e6a3a7d9ad4acb45 Mon Sep 17 00:00:00 2001 From: Neil Horman Date: Thu, 30 Jul 2009 18:52:15 -0700 Subject: xfrm: select sane defaults for xfrm[4|6] gc_thresh Choose saner defaults for xfrm[4|6] gc_thresh values on init Currently, the xfrm[4|6] code has hard-coded initial gc_thresh values (set to 1024). Given that the ipv4 and ipv6 routing caches are sized dynamically at boot time, the static selections can be non-sensical. This patch dynamically selects an appropriate gc threshold based on the corresponding main routing table size, using the assumption that we should in the worst case be able to handle as many connections as the routing table can. For ipv4, the maximum route cache size is 16 * the number of hash buckets in the route cache. Given that xfrm4 starts garbage collection at the gc_thresh and prevents new allocations at 2 * gc_thresh, we set gc_thresh to half the maximum route cache size. For ipv6, its a bit trickier. there is no maximum route cache size, but the ipv6 dst_ops gc_thresh is statically set to 1024. It seems sane to select a simmilar gc_thresh for the xfrm6 code that is half the number of hash buckets in the v6 route cache times 16 (like the v4 code does). Signed-off-by: Neil Horman Signed-off-by: David S. Miller --- include/net/ip6_fib.h | 6 ++++++ include/net/xfrm.h | 2 +- net/ipv4/route.c | 2 +- net/ipv4/xfrm4_policy.c | 13 ++++++++++++- net/ipv6/ip6_fib.c | 16 +++++----------- net/ipv6/xfrm6_policy.c | 15 +++++++++++++++ 6 files changed, 40 insertions(+), 14 deletions(-) (limited to 'include') diff --git a/include/net/ip6_fib.h b/include/net/ip6_fib.h index 7c5c0f79168a..15b492a9aa79 100644 --- a/include/net/ip6_fib.h +++ b/include/net/ip6_fib.h @@ -22,6 +22,12 @@ #include #include +#ifdef CONFIG_IPV6_MULTIPLE_TABLES +#define FIB6_TABLE_HASHSZ 256 +#else +#define FIB6_TABLE_HASHSZ 1 +#endif + struct rt6_info; struct fib6_config diff --git a/include/net/xfrm.h b/include/net/xfrm.h index 9e3a3f4c1f60..223e90a44824 100644 --- a/include/net/xfrm.h +++ b/include/net/xfrm.h @@ -1280,7 +1280,7 @@ struct xfrm6_tunnel { }; extern void xfrm_init(void); -extern void xfrm4_init(void); +extern void xfrm4_init(int rt_hash_size); extern int xfrm_state_init(struct net *net); extern void xfrm_state_fini(struct net *net); extern void xfrm4_state_init(void); diff --git a/net/ipv4/route.c b/net/ipv4/route.c index 278f46f5011b..fafbe163e2b5 100644 --- a/net/ipv4/route.c +++ b/net/ipv4/route.c @@ -3442,7 +3442,7 @@ int __init ip_rt_init(void) printk(KERN_ERR "Unable to create route proc files\n"); #ifdef CONFIG_XFRM xfrm_init(); - xfrm4_init(); + xfrm4_init(ip_rt_max_size); #endif rtnl_register(PF_INET, RTM_GETROUTE, inet_rtm_getroute, NULL); diff --git a/net/ipv4/xfrm4_policy.c b/net/ipv4/xfrm4_policy.c index 26496babdf3a..1ba44742ebbf 100644 --- a/net/ipv4/xfrm4_policy.c +++ b/net/ipv4/xfrm4_policy.c @@ -290,10 +290,21 @@ static void __exit xfrm4_policy_fini(void) xfrm_policy_unregister_afinfo(&xfrm4_policy_afinfo); } -void __init xfrm4_init(void) +void __init xfrm4_init(int rt_max_size) { xfrm4_state_init(); xfrm4_policy_init(); + /* + * Select a default value for the gc_thresh based on the main route + * table hash size. It seems to me the worst case scenario is when + * we have ipsec operating in transport mode, in which we create a + * dst_entry per socket. The xfrm gc algorithm starts trying to remove + * entries at gc_thresh, and prevents new allocations as 2*gc_thresh + * so lets set an initial xfrm gc_thresh value at the rt_max_size/2. + * That will let us store an ipsec connection per route table entry, + * and start cleaning when were 1/2 full + */ + xfrm4_dst_ops.gc_thresh = rt_max_size/2; sysctl_hdr = register_net_sysctl_table(&init_net, net_ipv4_ctl_path, xfrm4_policy_table); } diff --git a/net/ipv6/ip6_fib.c b/net/ipv6/ip6_fib.c index 52ee1dced2ff..0e93ca56eb69 100644 --- a/net/ipv6/ip6_fib.c +++ b/net/ipv6/ip6_fib.c @@ -164,12 +164,6 @@ static __inline__ void rt6_release(struct rt6_info *rt) dst_free(&rt->u.dst); } -#ifdef CONFIG_IPV6_MULTIPLE_TABLES -#define FIB_TABLE_HASHSZ 256 -#else -#define FIB_TABLE_HASHSZ 1 -#endif - static void fib6_link_table(struct net *net, struct fib6_table *tb) { unsigned int h; @@ -180,7 +174,7 @@ static void fib6_link_table(struct net *net, struct fib6_table *tb) */ rwlock_init(&tb->tb6_lock); - h = tb->tb6_id & (FIB_TABLE_HASHSZ - 1); + h = tb->tb6_id & (FIB6_TABLE_HASHSZ - 1); /* * No protection necessary, this is the only list mutatation @@ -231,7 +225,7 @@ struct fib6_table *fib6_get_table(struct net *net, u32 id) if (id == 0) id = RT6_TABLE_MAIN; - h = id & (FIB_TABLE_HASHSZ - 1); + h = id & (FIB6_TABLE_HASHSZ - 1); rcu_read_lock(); head = &net->ipv6.fib_table_hash[h]; hlist_for_each_entry_rcu(tb, node, head, tb6_hlist) { @@ -382,7 +376,7 @@ static int inet6_dump_fib(struct sk_buff *skb, struct netlink_callback *cb) arg.net = net; w->args = &arg; - for (h = s_h; h < FIB_TABLE_HASHSZ; h++, s_e = 0) { + for (h = s_h; h < FIB6_TABLE_HASHSZ; h++, s_e = 0) { e = 0; head = &net->ipv6.fib_table_hash[h]; hlist_for_each_entry(tb, node, head, tb6_hlist) { @@ -1368,7 +1362,7 @@ void fib6_clean_all(struct net *net, int (*func)(struct rt6_info *, void *arg), unsigned int h; rcu_read_lock(); - for (h = 0; h < FIB_TABLE_HASHSZ; h++) { + for (h = 0; h < FIB6_TABLE_HASHSZ; h++) { head = &net->ipv6.fib_table_hash[h]; hlist_for_each_entry_rcu(table, node, head, tb6_hlist) { write_lock_bh(&table->tb6_lock); @@ -1483,7 +1477,7 @@ static int fib6_net_init(struct net *net) if (!net->ipv6.rt6_stats) goto out_timer; - net->ipv6.fib_table_hash = kcalloc(FIB_TABLE_HASHSZ, + net->ipv6.fib_table_hash = kcalloc(FIB6_TABLE_HASHSZ, sizeof(*net->ipv6.fib_table_hash), GFP_KERNEL); if (!net->ipv6.fib_table_hash) diff --git a/net/ipv6/xfrm6_policy.c b/net/ipv6/xfrm6_policy.c index 4acc308eac7f..611cffcf554f 100644 --- a/net/ipv6/xfrm6_policy.c +++ b/net/ipv6/xfrm6_policy.c @@ -323,6 +323,7 @@ static struct ctl_table_header *sysctl_hdr; int __init xfrm6_init(void) { int ret; + unsigned int gc_thresh; ret = xfrm6_policy_init(); if (ret) @@ -331,6 +332,20 @@ int __init xfrm6_init(void) ret = xfrm6_state_init(); if (ret) goto out_policy; + /* + * We need a good default value for the xfrm6 gc threshold. + * In ipv4 we set it to the route hash table size * 8, which + * is half the size of the maximaum route cache for ipv4. It + * would be good to do the same thing for v6, except the table is + * constructed differently here. Here each table for a net namespace + * can have FIB_TABLE_HASHSZ entries, so lets go with the same + * computation that we used for ipv4 here. Also, lets keep the initial + * gc_thresh to a minimum of 1024, since, the ipv6 route cache defaults + * to that as a minimum as well + */ + gc_thresh = FIB6_TABLE_HASHSZ * 8; + xfrm6_dst_ops.gc_thresh = (gc_thresh < 1024) ? 1024 : gc_thresh; + sysctl_hdr = register_net_sysctl_table(&init_net, net_ipv6_ctl_path, xfrm6_policy_table); out: -- cgit v1.2.3 From cbb4f2646d77b536ed2b1500ef6641083228ed8f Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Fri, 31 Jul 2009 08:55:48 +0200 Subject: io context: fix ref counting Commit d9c7d394a8ebacb60097b192939ae9f15235225e ("block: prevent possible io_context->refcount overflow") mistakenly changed atomic_inc(&ioc->nr_tasks) to atomic_long_inc(&ioc->refcount). Signed-off-by: Li Zefan Acked-by: Nikanth Karthikesan Signed-off-by: Jens Axboe --- include/linux/iocontext.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/iocontext.h b/include/linux/iocontext.h index dd05434fa45f..4da4a75c3f1e 100644 --- a/include/linux/iocontext.h +++ b/include/linux/iocontext.h @@ -92,7 +92,7 @@ static inline struct io_context *ioc_task_link(struct io_context *ioc) * a race). */ if (ioc && atomic_long_inc_not_zero(&ioc->refcount)) { - atomic_long_inc(&ioc->refcount); + atomic_inc(&ioc->nr_tasks); return ioc; } -- cgit v1.2.3 From 6de7e356faf54aa75de5b624bbce28a5b776dfa8 Mon Sep 17 00:00:00 2001 From: Sebastian Andrzej Siewior Date: Thu, 18 Jun 2009 10:19:12 +0200 Subject: lib/scatterlist: add a flags to signalize mapping direction sg_miter_start() is currently unaware of the direction of the copy process (to or from the scatter list). It is important to know the direction because the page has to be flushed in case the data written is seen on a different mapping in user land on cache incoherent architectures. Signed-off-by: Sebastian Andrzej Siewior Acked-by: FUJITA Tomonori Acked-by: Tejun Heo Signed-off-by: Pierre Ossman --- include/linux/scatterlist.h | 2 ++ lib/scatterlist.c | 16 ++++++++++++---- 2 files changed, 14 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/include/linux/scatterlist.h b/include/linux/scatterlist.h index e5996984ddd0..9aaf5bfdad1a 100644 --- a/include/linux/scatterlist.h +++ b/include/linux/scatterlist.h @@ -242,6 +242,8 @@ size_t sg_copy_to_buffer(struct scatterlist *sgl, unsigned int nents, */ #define SG_MITER_ATOMIC (1 << 0) /* use kmap_atomic */ +#define SG_MITER_TO_SG (1 << 1) /* flush back to phys on unmap */ +#define SG_MITER_FROM_SG (1 << 2) /* nop */ struct sg_mapping_iter { /* the following three fields can be accessed directly */ diff --git a/lib/scatterlist.c b/lib/scatterlist.c index a295e404e908..0d475d8167bf 100644 --- a/lib/scatterlist.c +++ b/lib/scatterlist.c @@ -314,6 +314,7 @@ void sg_miter_start(struct sg_mapping_iter *miter, struct scatterlist *sgl, miter->__sg = sgl; miter->__nents = nents; miter->__offset = 0; + WARN_ON(!(flags & (SG_MITER_TO_SG | SG_MITER_FROM_SG))); miter->__flags = flags; } EXPORT_SYMBOL(sg_miter_start); @@ -394,6 +395,9 @@ void sg_miter_stop(struct sg_mapping_iter *miter) if (miter->addr) { miter->__offset += miter->consumed; + if (miter->__flags & SG_MITER_TO_SG) + flush_kernel_dcache_page(miter->page); + if (miter->__flags & SG_MITER_ATOMIC) { WARN_ON(!irqs_disabled()); kunmap_atomic(miter->addr, KM_BIO_SRC_IRQ); @@ -426,8 +430,14 @@ static size_t sg_copy_buffer(struct scatterlist *sgl, unsigned int nents, unsigned int offset = 0; struct sg_mapping_iter miter; unsigned long flags; + unsigned int sg_flags = SG_MITER_ATOMIC; + + if (to_buffer) + sg_flags |= SG_MITER_FROM_SG; + else + sg_flags |= SG_MITER_TO_SG; - sg_miter_start(&miter, sgl, nents, SG_MITER_ATOMIC); + sg_miter_start(&miter, sgl, nents, sg_flags); local_irq_save(flags); @@ -438,10 +448,8 @@ static size_t sg_copy_buffer(struct scatterlist *sgl, unsigned int nents, if (to_buffer) memcpy(buf + offset, miter.addr, len); - else { + else memcpy(miter.addr, buf + offset, len); - flush_kernel_dcache_page(miter.page); - } offset += len; } -- cgit v1.2.3 From 4b2a108cd0d34880fe9d932258ca5b2ccebcd05e Mon Sep 17 00:00:00 2001 From: Sebastian Andrzej Siewior Date: Mon, 22 Jun 2009 09:18:05 +0200 Subject: cb710: use SG_MITER_TO_SG/SG_MITER_FROM_SG MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit the code allready uses flush_kernel_dcache_page(). This patch updates the driver to the recent sg API changes which require that either SG_MITER_TO_SG or SG_MITER_FROM_SG is set. SG_MITER_TO_SG calls flush_kernel_dcache_page() in sg_mitter_stop() Signed-off-by: Sebastian Andrzej Siewior Acked-by: MichaÅ‚ MirosÅ‚aw Signed-off-by: Pierre Ossman --- drivers/misc/cb710/sgbuf2.c | 4 ---- drivers/mmc/host/cb710-mmc.c | 6 +++--- include/linux/cb710.h | 29 +++-------------------------- 3 files changed, 6 insertions(+), 33 deletions(-) (limited to 'include') diff --git a/drivers/misc/cb710/sgbuf2.c b/drivers/misc/cb710/sgbuf2.c index d38a7acdb6ec..d019746551f3 100644 --- a/drivers/misc/cb710/sgbuf2.c +++ b/drivers/misc/cb710/sgbuf2.c @@ -114,7 +114,6 @@ static void sg_dwiter_write_slow(struct sg_mapping_iter *miter, uint32_t data) if (!left) return; addr += len; - flush_kernel_dcache_page(miter->page); } while (sg_dwiter_next(miter)); } @@ -142,9 +141,6 @@ void cb710_sg_dwiter_write_next_block(struct sg_mapping_iter *miter, uint32_t da return; } else sg_dwiter_write_slow(miter, data); - - if (miter->length == miter->consumed) - flush_kernel_dcache_page(miter->page); } EXPORT_SYMBOL_GPL(cb710_sg_dwiter_write_next_block); diff --git a/drivers/mmc/host/cb710-mmc.c b/drivers/mmc/host/cb710-mmc.c index 11efefb1af51..4e72964a7b43 100644 --- a/drivers/mmc/host/cb710-mmc.c +++ b/drivers/mmc/host/cb710-mmc.c @@ -278,7 +278,7 @@ static int cb710_mmc_receive(struct cb710_slot *slot, struct mmc_data *data) if (unlikely(data->blksz & 15 && (data->blocks != 1 || data->blksz != 8))) return -EINVAL; - sg_miter_start(&miter, data->sg, data->sg_len, 0); + sg_miter_start(&miter, data->sg, data->sg_len, SG_MITER_TO_SG); cb710_modify_port_8(slot, CB710_MMC_CONFIG2_PORT, 15, CB710_MMC_C2_READ_PIO_SIZE_MASK); @@ -307,7 +307,7 @@ static int cb710_mmc_receive(struct cb710_slot *slot, struct mmc_data *data) goto out; } out: - cb710_sg_miter_stop_writing(&miter); + sg_miter_stop(&miter); return err; } @@ -322,7 +322,7 @@ static int cb710_mmc_send(struct cb710_slot *slot, struct mmc_data *data) if (unlikely(data->blocks > 1 && data->blksz & 15)) return -EINVAL; - sg_miter_start(&miter, data->sg, data->sg_len, 0); + sg_miter_start(&miter, data->sg, data->sg_len, SG_MITER_FROM_SG); cb710_modify_port_8(slot, CB710_MMC_CONFIG2_PORT, 0, CB710_MMC_C2_READ_PIO_SIZE_MASK); diff --git a/include/linux/cb710.h b/include/linux/cb710.h index 63bc9a4d2926..8cc10411bab2 100644 --- a/include/linux/cb710.h +++ b/include/linux/cb710.h @@ -140,29 +140,6 @@ void cb710_dump_regs(struct cb710_chip *chip, unsigned dump); #include #include -/** - * cb710_sg_miter_stop_writing - stop mapping iteration after writing - * @miter: sg mapping iter to be stopped - * - * Description: - * Stops mapping iterator @miter. @miter should have been started - * started using sg_miter_start(). A stopped iteration can be - * resumed by calling sg_miter_next() on it. This is useful when - * resources (kmap) need to be released during iteration. - * - * This is a convenience wrapper that will be optimized out for arches - * that don't need flush_kernel_dcache_page(). - * - * Context: - * IRQ disabled if the SG_MITER_ATOMIC is set. Don't care otherwise. - */ -static inline void cb710_sg_miter_stop_writing(struct sg_mapping_iter *miter) -{ - if (miter->page) - flush_kernel_dcache_page(miter->page); - sg_miter_stop(miter); -} - /* * 32-bit PIO mapping sg iterator * @@ -171,12 +148,12 @@ static inline void cb710_sg_miter_stop_writing(struct sg_mapping_iter *miter) * without DMA support). * * Best-case reading (transfer from device): - * sg_miter_start(); + * sg_miter_start(, SG_MITER_TO_SG); * cb710_sg_dwiter_write_from_io(); - * cb710_sg_miter_stop_writing(); + * sg_miter_stop(); * * Best-case writing (transfer to device): - * sg_miter_start(); + * sg_miter_start(, SG_MITER_FROM_SG); * cb710_sg_dwiter_read_to_io(); * sg_miter_stop(); */ -- cgit v1.2.3 From c7121843685de2bf7f3afd3ae1d6a146010bf1fc Mon Sep 17 00:00:00 2001 From: Magnus Damm Date: Tue, 28 Jul 2009 14:09:55 -0700 Subject: clocksource: Save mult_orig in clocksource_disable() To fix the common case where ->enable() does not set up mult, make sure mult_orig is saved in mult on disable. Also add comments to explain why we do this. Signed-off-by: Magnus Damm Cc: johnstul@us.ibm.com Cc: lethal@linux-sh.org Cc: akpm@linux-foundation.org LKML-Reference: <20090618152432.10136.9932.sendpatchset@rx1.opensource.se> Signed-off-by: Thomas Gleixner --- include/linux/clocksource.h | 14 +++++++++++++- 1 file changed, 13 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/clocksource.h b/include/linux/clocksource.h index c56457c8334e..1219be4fb42e 100644 --- a/include/linux/clocksource.h +++ b/include/linux/clocksource.h @@ -293,7 +293,12 @@ static inline int clocksource_enable(struct clocksource *cs) if (cs->enable) ret = cs->enable(cs); - /* save mult_orig on enable */ + /* + * The frequency may have changed while the clocksource + * was disabled. If so the code in ->enable() must update + * the mult value to reflect the new frequency. Make sure + * mult_orig follows this change. + */ cs->mult_orig = cs->mult; return ret; @@ -309,6 +314,13 @@ static inline int clocksource_enable(struct clocksource *cs) */ static inline void clocksource_disable(struct clocksource *cs) { + /* + * Save mult_orig in mult so clocksource_enable() can + * restore the value regardless if ->enable() updates + * the value of mult or not. + */ + cs->mult = cs->mult_orig; + if (cs->disable) cs->disable(cs); } -- cgit v1.2.3 From 7c958e32649e0c35801762878fb0b6da8c55a515 Mon Sep 17 00:00:00 2001 From: "Martin K. Petersen" Date: Fri, 31 Jul 2009 11:49:11 -0400 Subject: block: Add a wrapper for setting minimum request size without a queue Introduce blk_limits_io_min() and make blk_queue_io_min() call it. Signed-off-by: Mike Snitzer Signed-off-by: Martin K. Petersen Signed-off-by: Jens Axboe --- block/blk-settings.c | 31 ++++++++++++++++++++++++------- include/linux/blkdev.h | 1 + 2 files changed, 25 insertions(+), 7 deletions(-) (limited to 'include') diff --git a/block/blk-settings.c b/block/blk-settings.c index 8e86e2d2b147..1f7197434166 100644 --- a/block/blk-settings.c +++ b/block/blk-settings.c @@ -383,6 +383,29 @@ void blk_queue_alignment_offset(struct request_queue *q, unsigned int offset) } EXPORT_SYMBOL(blk_queue_alignment_offset); +/** + * blk_limits_io_min - set minimum request size for a device + * @limits: the queue limits + * @min: smallest I/O size in bytes + * + * Description: + * Some devices have an internal block size bigger than the reported + * hardware sector size. This function can be used to signal the + * smallest I/O the device can perform without incurring a performance + * penalty. + */ +void blk_limits_io_min(struct queue_limits *limits, unsigned int min) +{ + limits->io_min = min; + + if (limits->io_min < limits->logical_block_size) + limits->io_min = limits->logical_block_size; + + if (limits->io_min < limits->physical_block_size) + limits->io_min = limits->physical_block_size; +} +EXPORT_SYMBOL(blk_limits_io_min); + /** * blk_queue_io_min - set minimum request size for the queue * @q: the request queue for the device @@ -396,13 +419,7 @@ EXPORT_SYMBOL(blk_queue_alignment_offset); */ void blk_queue_io_min(struct request_queue *q, unsigned int min) { - q->limits.io_min = min; - - if (q->limits.io_min < q->limits.logical_block_size) - q->limits.io_min = q->limits.logical_block_size; - - if (q->limits.io_min < q->limits.physical_block_size) - q->limits.io_min = q->limits.physical_block_size; + blk_limits_io_min(&q->limits, min); } EXPORT_SYMBOL(blk_queue_io_min); diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index e7cb5dbf6c26..69103e053c92 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -913,6 +913,7 @@ extern void blk_queue_logical_block_size(struct request_queue *, unsigned short) extern void blk_queue_physical_block_size(struct request_queue *, unsigned short); extern void blk_queue_alignment_offset(struct request_queue *q, unsigned int alignment); +extern void blk_limits_io_min(struct queue_limits *limits, unsigned int min); extern void blk_queue_io_min(struct request_queue *q, unsigned int min); extern void blk_queue_io_opt(struct request_queue *q, unsigned int opt); extern void blk_set_default_limits(struct queue_limits *lim); -- cgit v1.2.3 From 9f498cc5be7e013d8d6e4c616980ed0ffc8680d2 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Thu, 23 Jul 2009 14:46:33 +0200 Subject: perf_counter: Full task tracing In order to be able to distinguish between no samples due to inactivity and no samples due to task ended, Arjan asked for PERF_EVENT_EXIT events. This is useful to the boot delay instrumentation (bootchart) app. This patch changes the PERF_EVENT_FORK to be emitted on every clone, and adds PERF_EVENT_EXIT to be emitted on task exit, after the task's counters have been closed. This task tracing is controlled through: attr.comm || attr.mmap and through the new attr.task field. Suggested-by: Arjan van de Ven Cc: Paul Mackerras Cc: Anton Blanchard Signed-off-by: Peter Zijlstra [ cleaned up perf_counter.h a bit ] Signed-off-by: Ingo Molnar --- include/linux/perf_counter.h | 13 ++++++- kernel/fork.c | 4 +- kernel/perf_counter.c | 87 +++++++++++++++++++++++++++++--------------- 3 files changed, 71 insertions(+), 33 deletions(-) (limited to 'include') diff --git a/include/linux/perf_counter.h b/include/linux/perf_counter.h index bd15d7a5f5ce..e604e6ef72dd 100644 --- a/include/linux/perf_counter.h +++ b/include/linux/perf_counter.h @@ -181,8 +181,9 @@ struct perf_counter_attr { freq : 1, /* use freq, not period */ inherit_stat : 1, /* per task counts */ enable_on_exec : 1, /* next exec enables */ + task : 1, /* trace fork/exit */ - __reserved_1 : 51; + __reserved_1 : 50; __u32 wakeup_events; /* wakeup every n events */ __u32 __reserved_2; @@ -308,6 +309,15 @@ enum perf_event_type { */ PERF_EVENT_COMM = 3, + /* + * struct { + * struct perf_event_header header; + * u32 pid, ppid; + * u32 tid, ptid; + * }; + */ + PERF_EVENT_EXIT = 4, + /* * struct { * struct perf_event_header header; @@ -323,6 +333,7 @@ enum perf_event_type { * struct { * struct perf_event_header header; * u32 pid, ppid; + * u32 tid, ptid; * }; */ PERF_EVENT_FORK = 7, diff --git a/kernel/fork.c b/kernel/fork.c index 29b532e718f7..466531eb92cc 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -1269,6 +1269,7 @@ static struct task_struct *copy_process(unsigned long clone_flags, write_unlock_irq(&tasklist_lock); proc_fork_connector(p); cgroup_post_fork(p); + perf_counter_fork(p); return p; bad_fork_free_pid: @@ -1410,9 +1411,6 @@ long do_fork(unsigned long clone_flags, init_completion(&vfork); } - if (!(clone_flags & CLONE_THREAD)) - perf_counter_fork(p); - audit_finish_fork(p); tracehook_report_clone(regs, clone_flags, nr, p); diff --git a/kernel/perf_counter.c b/kernel/perf_counter.c index 48471d75ae01..199ed4771315 100644 --- a/kernel/perf_counter.c +++ b/kernel/perf_counter.c @@ -42,6 +42,7 @@ static int perf_overcommit __read_mostly = 1; static atomic_t nr_counters __read_mostly; static atomic_t nr_mmap_counters __read_mostly; static atomic_t nr_comm_counters __read_mostly; +static atomic_t nr_task_counters __read_mostly; /* * perf counter paranoia level: @@ -1654,6 +1655,8 @@ static void free_counter(struct perf_counter *counter) atomic_dec(&nr_mmap_counters); if (counter->attr.comm) atomic_dec(&nr_comm_counters); + if (counter->attr.task) + atomic_dec(&nr_task_counters); } if (counter->destroy) @@ -2831,10 +2834,12 @@ perf_counter_read_event(struct perf_counter *counter, } /* - * fork tracking + * task tracking -- fork/exit + * + * enabled by: attr.comm | attr.mmap | attr.task */ -struct perf_fork_event { +struct perf_task_event { struct task_struct *task; struct { @@ -2842,37 +2847,42 @@ struct perf_fork_event { u32 pid; u32 ppid; + u32 tid; + u32 ptid; } event; }; -static void perf_counter_fork_output(struct perf_counter *counter, - struct perf_fork_event *fork_event) +static void perf_counter_task_output(struct perf_counter *counter, + struct perf_task_event *task_event) { struct perf_output_handle handle; - int size = fork_event->event.header.size; - struct task_struct *task = fork_event->task; + int size = task_event->event.header.size; + struct task_struct *task = task_event->task; int ret = perf_output_begin(&handle, counter, size, 0, 0); if (ret) return; - fork_event->event.pid = perf_counter_pid(counter, task); - fork_event->event.ppid = perf_counter_pid(counter, task->real_parent); + task_event->event.pid = perf_counter_pid(counter, task); + task_event->event.ppid = perf_counter_pid(counter, task->real_parent); - perf_output_put(&handle, fork_event->event); + task_event->event.tid = perf_counter_tid(counter, task); + task_event->event.ptid = perf_counter_tid(counter, task->real_parent); + + perf_output_put(&handle, task_event->event); perf_output_end(&handle); } -static int perf_counter_fork_match(struct perf_counter *counter) +static int perf_counter_task_match(struct perf_counter *counter) { - if (counter->attr.comm || counter->attr.mmap) + if (counter->attr.comm || counter->attr.mmap || counter->attr.task) return 1; return 0; } -static void perf_counter_fork_ctx(struct perf_counter_context *ctx, - struct perf_fork_event *fork_event) +static void perf_counter_task_ctx(struct perf_counter_context *ctx, + struct perf_task_event *task_event) { struct perf_counter *counter; @@ -2881,19 +2891,19 @@ static void perf_counter_fork_ctx(struct perf_counter_context *ctx, rcu_read_lock(); list_for_each_entry_rcu(counter, &ctx->event_list, event_entry) { - if (perf_counter_fork_match(counter)) - perf_counter_fork_output(counter, fork_event); + if (perf_counter_task_match(counter)) + perf_counter_task_output(counter, task_event); } rcu_read_unlock(); } -static void perf_counter_fork_event(struct perf_fork_event *fork_event) +static void perf_counter_task_event(struct perf_task_event *task_event) { struct perf_cpu_context *cpuctx; struct perf_counter_context *ctx; cpuctx = &get_cpu_var(perf_cpu_context); - perf_counter_fork_ctx(&cpuctx->ctx, fork_event); + perf_counter_task_ctx(&cpuctx->ctx, task_event); put_cpu_var(perf_cpu_context); rcu_read_lock(); @@ -2903,32 +2913,40 @@ static void perf_counter_fork_event(struct perf_fork_event *fork_event) */ ctx = rcu_dereference(current->perf_counter_ctxp); if (ctx) - perf_counter_fork_ctx(ctx, fork_event); + perf_counter_task_ctx(ctx, task_event); rcu_read_unlock(); } -void perf_counter_fork(struct task_struct *task) +static void perf_counter_task(struct task_struct *task, int new) { - struct perf_fork_event fork_event; + struct perf_task_event task_event; if (!atomic_read(&nr_comm_counters) && - !atomic_read(&nr_mmap_counters)) + !atomic_read(&nr_mmap_counters) && + !atomic_read(&nr_task_counters)) return; - fork_event = (struct perf_fork_event){ + task_event = (struct perf_task_event){ .task = task, .event = { .header = { - .type = PERF_EVENT_FORK, + .type = new ? PERF_EVENT_FORK : PERF_EVENT_EXIT, .misc = 0, - .size = sizeof(fork_event.event), + .size = sizeof(task_event.event), }, /* .pid */ /* .ppid */ + /* .tid */ + /* .ptid */ }, }; - perf_counter_fork_event(&fork_event); + perf_counter_task_event(&task_event); +} + +void perf_counter_fork(struct task_struct *task) +{ + perf_counter_task(task, 1); } /* @@ -3887,6 +3905,8 @@ done: atomic_inc(&nr_mmap_counters); if (counter->attr.comm) atomic_inc(&nr_comm_counters); + if (counter->attr.task) + atomic_inc(&nr_task_counters); } return counter; @@ -4248,8 +4268,10 @@ void perf_counter_exit_task(struct task_struct *child) struct perf_counter_context *child_ctx; unsigned long flags; - if (likely(!child->perf_counter_ctxp)) + if (likely(!child->perf_counter_ctxp)) { + perf_counter_task(child, 0); return; + } local_irq_save(flags); /* @@ -4267,15 +4289,22 @@ void perf_counter_exit_task(struct task_struct *child) * incremented the context's refcount before we do put_ctx below. */ spin_lock(&child_ctx->lock); - child->perf_counter_ctxp = NULL; /* * If this context is a clone; unclone it so it can't get * swapped to another process while we're removing all * the counters from it. */ unclone_ctx(child_ctx); - spin_unlock(&child_ctx->lock); - local_irq_restore(flags); + spin_unlock_irqrestore(&child_ctx->lock, flags); + + /* + * Report the task dead after unscheduling the counters so that we + * won't get any samples after PERF_EVENT_EXIT. We can however still + * get a few PERF_EVENT_READ events. + */ + perf_counter_task(child, 0); + + child->perf_counter_ctxp = NULL; /* * We can recurse on the same lock type through: -- cgit v1.2.3 From 716a42348cdaf04534b15fbdc9c83e25baebfed5 Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Fri, 24 Jul 2009 20:05:23 +0200 Subject: sched: Fix cond_resched_lock() in !CONFIG_PREEMPT The might_sleep() test inside cond_resched_lock() assumes the spinlock is held and then preemption is disabled. This is true with CONFIG_PREEMPT but the preempt_count() doesn't change otherwise. Check by starting from the appropriate preempt offset depending on the config. Reported-by: Li Zefan Signed-off-by: Frederic Weisbecker Signed-off-by: Peter Zijlstra LKML-Reference: <1248458723-12146-1-git-send-email-fweisbec@gmail.com> Signed-off-by: Ingo Molnar --- include/linux/sched.h | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/sched.h b/include/linux/sched.h index cbbfca69aa4a..c472414953bf 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -2293,8 +2293,14 @@ extern int _cond_resched(void); extern int __cond_resched_lock(spinlock_t *lock); +#ifdef CONFIG_PREEMPT +#define PREEMPT_LOCK_OFFSET PREEMPT_OFFSET +#else +#define PREEMPT_LOCK_OFFSET 0 +#endif + #define cond_resched_lock(lock) ({ \ - __might_sleep(__FILE__, __LINE__, PREEMPT_OFFSET); \ + __might_sleep(__FILE__, __LINE__, PREEMPT_LOCK_OFFSET); \ __cond_resched_lock(lock); \ }) -- cgit v1.2.3 From 3f029d3c6d62068d59301d90c18dbde8ee402107 Mon Sep 17 00:00:00 2001 From: Gregory Haskins Date: Wed, 29 Jul 2009 11:08:47 -0400 Subject: sched: Enhance the pre/post scheduling logic We currently have an explicit "needs_post" vtable method which returns a stack variable for whether we should later run post-schedule. This leads to an awkward exchange of the variable as it bubbles back up out of the context switch. Peter Zijlstra observed that this information could be stored in the run-queue itself instead of handled on the stack. Therefore, we revert to the method of having context_switch return void, and update an internal rq->post_schedule variable when we require further processing. In addition, we fix a race condition where we try to access current->sched_class without holding the rq->lock. This is technically racy, as the sched-class could change out from under us. Instead, we reference the per-rq post_schedule variable with the runqueue unlocked, but with preemption disabled to see if we need to reacquire the rq->lock. Finally, we clean the code up slightly by removing the #ifdef CONFIG_SMP conditionals from the schedule() call, and implement some inline helper functions instead. This patch passes checkpatch, and rt-migrate. Signed-off-by: Gregory Haskins Signed-off-by: Peter Zijlstra LKML-Reference: <20090729150422.17691.55590.stgit@dev.haskins.net> Signed-off-by: Ingo Molnar --- include/linux/sched.h | 1 - kernel/sched.c | 82 +++++++++++++++++++++++++++++++-------------------- kernel/sched_rt.c | 31 +++++++------------ 3 files changed, 61 insertions(+), 53 deletions(-) (limited to 'include') diff --git a/include/linux/sched.h b/include/linux/sched.h index 2c35bc29d2a9..195d72d5c102 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -1047,7 +1047,6 @@ struct sched_class { struct rq *busiest, struct sched_domain *sd, enum cpu_idle_type idle); void (*pre_schedule) (struct rq *this_rq, struct task_struct *task); - int (*needs_post_schedule) (struct rq *this_rq); void (*post_schedule) (struct rq *this_rq); void (*task_wake_up) (struct rq *this_rq, struct task_struct *task); diff --git a/kernel/sched.c b/kernel/sched.c index a030d4514cdc..613fee54fc89 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -616,6 +616,7 @@ struct rq { unsigned char idle_at_tick; /* For active balancing */ + int post_schedule; int active_balance; int push_cpu; /* cpu of this runqueue: */ @@ -2839,17 +2840,11 @@ prepare_task_switch(struct rq *rq, struct task_struct *prev, * with the lock held can cause deadlocks; see schedule() for * details.) */ -static int finish_task_switch(struct rq *rq, struct task_struct *prev) +static void finish_task_switch(struct rq *rq, struct task_struct *prev) __releases(rq->lock) { struct mm_struct *mm = rq->prev_mm; long prev_state; - int post_schedule = 0; - -#ifdef CONFIG_SMP - if (current->sched_class->needs_post_schedule) - post_schedule = current->sched_class->needs_post_schedule(rq); -#endif rq->prev_mm = NULL; @@ -2880,10 +2875,44 @@ static int finish_task_switch(struct rq *rq, struct task_struct *prev) kprobe_flush_task(prev); put_task_struct(prev); } +} + +#ifdef CONFIG_SMP + +/* assumes rq->lock is held */ +static inline void pre_schedule(struct rq *rq, struct task_struct *prev) +{ + if (prev->sched_class->pre_schedule) + prev->sched_class->pre_schedule(rq, prev); +} + +/* rq->lock is NOT held, but preemption is disabled */ +static inline void post_schedule(struct rq *rq) +{ + if (rq->post_schedule) { + unsigned long flags; + + spin_lock_irqsave(&rq->lock, flags); + if (rq->curr->sched_class->post_schedule) + rq->curr->sched_class->post_schedule(rq); + spin_unlock_irqrestore(&rq->lock, flags); + + rq->post_schedule = 0; + } +} + +#else - return post_schedule; +static inline void pre_schedule(struct rq *rq, struct task_struct *p) +{ +} + +static inline void post_schedule(struct rq *rq) +{ } +#endif + /** * schedule_tail - first thing a freshly forked thread must call. * @prev: the thread we just switched away from. @@ -2892,14 +2921,14 @@ asmlinkage void schedule_tail(struct task_struct *prev) __releases(rq->lock) { struct rq *rq = this_rq(); - int post_schedule; - post_schedule = finish_task_switch(rq, prev); + finish_task_switch(rq, prev); -#ifdef CONFIG_SMP - if (post_schedule) - current->sched_class->post_schedule(rq); -#endif + /* + * FIXME: do we need to worry about rq being invalidated by the + * task_switch? + */ + post_schedule(rq); #ifdef __ARCH_WANT_UNLOCKED_CTXSW /* In this case, finish_task_switch does not reenable preemption */ @@ -2913,7 +2942,7 @@ asmlinkage void schedule_tail(struct task_struct *prev) * context_switch - switch to the new MM and the new * thread's register state. */ -static inline int +static inline void context_switch(struct rq *rq, struct task_struct *prev, struct task_struct *next) { @@ -2960,7 +2989,7 @@ context_switch(struct rq *rq, struct task_struct *prev, * CPUs since it called schedule(), thus the 'rq' on its stack * frame will be invalid. */ - return finish_task_switch(this_rq(), prev); + finish_task_switch(this_rq(), prev); } /* @@ -5371,7 +5400,6 @@ asmlinkage void __sched schedule(void) { struct task_struct *prev, *next; unsigned long *switch_count; - int post_schedule = 0; struct rq *rq; int cpu; @@ -5403,10 +5431,7 @@ need_resched_nonpreemptible: switch_count = &prev->nvcsw; } -#ifdef CONFIG_SMP - if (prev->sched_class->pre_schedule) - prev->sched_class->pre_schedule(rq, prev); -#endif + pre_schedule(rq, prev); if (unlikely(!rq->nr_running)) idle_balance(cpu, rq); @@ -5422,25 +5447,17 @@ need_resched_nonpreemptible: rq->curr = next; ++*switch_count; - post_schedule = context_switch(rq, prev, next); /* unlocks the rq */ + context_switch(rq, prev, next); /* unlocks the rq */ /* * the context switch might have flipped the stack from under * us, hence refresh the local variables. */ cpu = smp_processor_id(); rq = cpu_rq(cpu); - } else { -#ifdef CONFIG_SMP - if (current->sched_class->needs_post_schedule) - post_schedule = current->sched_class->needs_post_schedule(rq); -#endif + } else spin_unlock_irq(&rq->lock); - } -#ifdef CONFIG_SMP - if (post_schedule) - current->sched_class->post_schedule(rq); -#endif + post_schedule(rq); if (unlikely(reacquire_kernel_lock(current) < 0)) goto need_resched_nonpreemptible; @@ -9403,6 +9420,7 @@ void __init sched_init(void) #ifdef CONFIG_SMP rq->sd = NULL; rq->rd = NULL; + rq->post_schedule = 0; rq->active_balance = 0; rq->next_balance = jiffies; rq->push_cpu = 0; diff --git a/kernel/sched_rt.c b/kernel/sched_rt.c index 3918e01994e0..a8f89bc3e5eb 100644 --- a/kernel/sched_rt.c +++ b/kernel/sched_rt.c @@ -1056,6 +1056,11 @@ static struct task_struct *_pick_next_task_rt(struct rq *rq) return p; } +static inline int has_pushable_tasks(struct rq *rq) +{ + return !plist_head_empty(&rq->rt.pushable_tasks); +} + static struct task_struct *pick_next_task_rt(struct rq *rq) { struct task_struct *p = _pick_next_task_rt(rq); @@ -1064,6 +1069,12 @@ static struct task_struct *pick_next_task_rt(struct rq *rq) if (p) dequeue_pushable_task(rq, p); + /* + * We detect this state here so that we can avoid taking the RQ + * lock again later if there is no need to push + */ + rq->post_schedule = has_pushable_tasks(rq); + return p; } @@ -1262,11 +1273,6 @@ static struct rq *find_lock_lowest_rq(struct task_struct *task, struct rq *rq) return lowest_rq; } -static inline int has_pushable_tasks(struct rq *rq) -{ - return !plist_head_empty(&rq->rt.pushable_tasks); -} - static struct task_struct *pick_next_pushable_task(struct rq *rq) { struct task_struct *p; @@ -1466,23 +1472,9 @@ static void pre_schedule_rt(struct rq *rq, struct task_struct *prev) pull_rt_task(rq); } -/* - * assumes rq->lock is held - */ -static int needs_post_schedule_rt(struct rq *rq) -{ - return has_pushable_tasks(rq); -} - static void post_schedule_rt(struct rq *rq) { - /* - * This is only called if needs_post_schedule_rt() indicates that - * we need to push tasks away - */ - spin_lock_irq(&rq->lock); push_rt_tasks(rq); - spin_unlock_irq(&rq->lock); } /* @@ -1758,7 +1750,6 @@ static const struct sched_class rt_sched_class = { .rq_online = rq_online_rt, .rq_offline = rq_offline_rt, .pre_schedule = pre_schedule_rt, - .needs_post_schedule = needs_post_schedule_rt, .post_schedule = post_schedule_rt, .task_wake_up = task_wake_up_rt, .switched_from = switched_from_rt, -- cgit v1.2.3 From f607c6685774811b8112e124f10a053d77015485 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Mon, 20 Jul 2009 19:16:29 +0200 Subject: lockdep: Introduce lockdep_assert_held() Add a lockdep helper to validate that we indeed are the owner of a lock. Signed-off-by: Peter Zijlstra Signed-off-by: Ingo Molnar --- include/linux/lockdep.h | 8 ++++++++ kernel/lockdep.c | 33 +++++++++++++++++++++++++++++++++ kernel/sched.c | 2 ++ 3 files changed, 43 insertions(+) (limited to 'include') diff --git a/include/linux/lockdep.h b/include/linux/lockdep.h index 12aabfcb45f6..a6d5e5e4d084 100644 --- a/include/linux/lockdep.h +++ b/include/linux/lockdep.h @@ -296,6 +296,10 @@ extern void lock_acquire(struct lockdep_map *lock, unsigned int subclass, extern void lock_release(struct lockdep_map *lock, int nested, unsigned long ip); +#define lockdep_is_held(lock) lock_is_held(&(lock)->dep_map) + +extern int lock_is_held(struct lockdep_map *lock); + extern void lock_set_class(struct lockdep_map *lock, const char *name, struct lock_class_key *key, unsigned int subclass, unsigned long ip); @@ -314,6 +318,8 @@ extern void lockdep_trace_alloc(gfp_t mask); #define lockdep_depth(tsk) (debug_locks ? (tsk)->lockdep_depth : 0) +#define lockdep_assert_held(l) WARN_ON(debug_locks && !lockdep_is_held(l)) + #else /* !LOCKDEP */ static inline void lockdep_off(void) @@ -358,6 +364,8 @@ struct lock_class_key { }; #define lockdep_depth(tsk) (0) +#define lockdep_assert_held(l) do { } while (0) + #endif /* !LOCKDEP */ #ifdef CONFIG_LOCK_STAT diff --git a/kernel/lockdep.c b/kernel/lockdep.c index 4b6cebe8ab31..28914a509914 100644 --- a/kernel/lockdep.c +++ b/kernel/lockdep.c @@ -3059,6 +3059,19 @@ __lock_release(struct lockdep_map *lock, int nested, unsigned long ip) check_chain_key(curr); } +static int __lock_is_held(struct lockdep_map *lock) +{ + struct task_struct *curr = current; + int i; + + for (i = 0; i < curr->lockdep_depth; i++) { + if (curr->held_locks[i].instance == lock) + return 1; + } + + return 0; +} + /* * Check whether we follow the irq-flags state precisely: */ @@ -3160,6 +3173,26 @@ void lock_release(struct lockdep_map *lock, int nested, } EXPORT_SYMBOL_GPL(lock_release); +int lock_is_held(struct lockdep_map *lock) +{ + unsigned long flags; + int ret = 0; + + if (unlikely(current->lockdep_recursion)) + return ret; + + raw_local_irq_save(flags); + check_flags(flags); + + current->lockdep_recursion = 1; + ret = __lock_is_held(lock); + current->lockdep_recursion = 0; + raw_local_irq_restore(flags); + + return ret; +} +EXPORT_SYMBOL_GPL(lock_is_held); + void lockdep_set_current_reclaim_state(gfp_t gfp_mask) { current->lockdep_reclaim_gfp = gfp_mask; diff --git a/kernel/sched.c b/kernel/sched.c index 1b59e265273b..2c75f7daa439 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -6609,6 +6609,8 @@ int cond_resched_lock(spinlock_t *lock) int resched = should_resched(); int ret = 0; + lockdep_assert_held(lock); + if (spin_needbreak(lock) || resched) { spin_unlock(lock); if (resched) -- cgit v1.2.3 From bb97a91e2549a7f2df9c21d32542582f549ab3ec Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Mon, 20 Jul 2009 19:15:35 +0200 Subject: lockdep: Deal with many similar locks spin_lock_nest_lock() allows to take many instances of the same class, this can easily lead to overflow of MAX_LOCK_DEPTH. To avoid this overflow, we'll stop accounting instances but start reference counting the class in the held_lock structure. [ We could maintain a list of instances, if we'd move the hlock stuff into __lock_acquired(), but that would require significant modifications to the current code. ] We restrict this mode to spin_lock_nest_lock() only, because it degrades the lockdep quality due to lost of instance. For lockstat this means we don't track lock statistics for any but the first lock in the series. Currently nesting is limited to 11 bits because that was the spare space available in held_lock. This yields a 2048 instances maximium. Signed-off-by: Peter Zijlstra Cc: Marcelo Tosatti Cc: Linus Torvalds Signed-off-by: Ingo Molnar --- include/linux/lockdep.h | 4 ++- kernel/lockdep.c | 89 ++++++++++++++++++++++++++++++++++++++++++------- 2 files changed, 80 insertions(+), 13 deletions(-) (limited to 'include') diff --git a/include/linux/lockdep.h b/include/linux/lockdep.h index a6d5e5e4d084..47d42eff6124 100644 --- a/include/linux/lockdep.h +++ b/include/linux/lockdep.h @@ -213,10 +213,12 @@ struct held_lock { * interrupt context: */ unsigned int irq_context:2; /* bit 0 - soft, bit 1 - hard */ - unsigned int trylock:1; + unsigned int trylock:1; /* 16 bits */ + unsigned int read:2; /* see lock_acquire() comment */ unsigned int check:2; /* see lock_acquire() comment */ unsigned int hardirqs_off:1; + unsigned int references:11; /* 32 bits */ }; /* diff --git a/kernel/lockdep.c b/kernel/lockdep.c index 28914a509914..0bb246e21cd7 100644 --- a/kernel/lockdep.c +++ b/kernel/lockdep.c @@ -2708,13 +2708,15 @@ EXPORT_SYMBOL_GPL(lockdep_init_map); */ static int __lock_acquire(struct lockdep_map *lock, unsigned int subclass, int trylock, int read, int check, int hardirqs_off, - struct lockdep_map *nest_lock, unsigned long ip) + struct lockdep_map *nest_lock, unsigned long ip, + int references) { struct task_struct *curr = current; struct lock_class *class = NULL; struct held_lock *hlock; unsigned int depth, id; int chain_head = 0; + int class_idx; u64 chain_key; if (!prove_locking) @@ -2762,10 +2764,24 @@ static int __lock_acquire(struct lockdep_map *lock, unsigned int subclass, if (DEBUG_LOCKS_WARN_ON(depth >= MAX_LOCK_DEPTH)) return 0; + class_idx = class - lock_classes + 1; + + if (depth) { + hlock = curr->held_locks + depth - 1; + if (hlock->class_idx == class_idx && nest_lock) { + if (hlock->references) + hlock->references++; + else + hlock->references = 2; + + return 1; + } + } + hlock = curr->held_locks + depth; if (DEBUG_LOCKS_WARN_ON(!class)) return 0; - hlock->class_idx = class - lock_classes + 1; + hlock->class_idx = class_idx; hlock->acquire_ip = ip; hlock->instance = lock; hlock->nest_lock = nest_lock; @@ -2773,6 +2789,7 @@ static int __lock_acquire(struct lockdep_map *lock, unsigned int subclass, hlock->read = read; hlock->check = check; hlock->hardirqs_off = !!hardirqs_off; + hlock->references = references; #ifdef CONFIG_LOCK_STAT hlock->waittime_stamp = 0; hlock->holdtime_stamp = sched_clock(); @@ -2881,6 +2898,30 @@ static int check_unlock(struct task_struct *curr, struct lockdep_map *lock, return 1; } +static int match_held_lock(struct held_lock *hlock, struct lockdep_map *lock) +{ + if (hlock->instance == lock) + return 1; + + if (hlock->references) { + struct lock_class *class = lock->class_cache; + + if (!class) + class = look_up_lock_class(lock, 0); + + if (DEBUG_LOCKS_WARN_ON(!class)) + return 0; + + if (DEBUG_LOCKS_WARN_ON(!hlock->nest_lock)) + return 0; + + if (hlock->class_idx == class - lock_classes + 1) + return 1; + } + + return 0; +} + static int __lock_set_class(struct lockdep_map *lock, const char *name, struct lock_class_key *key, unsigned int subclass, @@ -2904,7 +2945,7 @@ __lock_set_class(struct lockdep_map *lock, const char *name, */ if (prev_hlock && prev_hlock->irq_context != hlock->irq_context) break; - if (hlock->instance == lock) + if (match_held_lock(hlock, lock)) goto found_it; prev_hlock = hlock; } @@ -2923,7 +2964,8 @@ found_it: if (!__lock_acquire(hlock->instance, hlock_class(hlock)->subclass, hlock->trylock, hlock->read, hlock->check, hlock->hardirqs_off, - hlock->nest_lock, hlock->acquire_ip)) + hlock->nest_lock, hlock->acquire_ip, + hlock->references)) return 0; } @@ -2962,20 +3004,34 @@ lock_release_non_nested(struct task_struct *curr, */ if (prev_hlock && prev_hlock->irq_context != hlock->irq_context) break; - if (hlock->instance == lock) + if (match_held_lock(hlock, lock)) goto found_it; prev_hlock = hlock; } return print_unlock_inbalance_bug(curr, lock, ip); found_it: - lock_release_holdtime(hlock); + if (hlock->instance == lock) + lock_release_holdtime(hlock); + + if (hlock->references) { + hlock->references--; + if (hlock->references) { + /* + * We had, and after removing one, still have + * references, the current lock stack is still + * valid. We're done! + */ + return 1; + } + } /* * We have the right lock to unlock, 'hlock' points to it. * Now we remove it from the stack, and add back the other * entries (if any), recalculating the hash along the way: */ + curr->lockdep_depth = i; curr->curr_chain_key = hlock->prev_chain_key; @@ -2984,7 +3040,8 @@ found_it: if (!__lock_acquire(hlock->instance, hlock_class(hlock)->subclass, hlock->trylock, hlock->read, hlock->check, hlock->hardirqs_off, - hlock->nest_lock, hlock->acquire_ip)) + hlock->nest_lock, hlock->acquire_ip, + hlock->references)) return 0; } @@ -3014,7 +3071,7 @@ static int lock_release_nested(struct task_struct *curr, /* * Is the unlock non-nested: */ - if (hlock->instance != lock) + if (hlock->instance != lock || hlock->references) return lock_release_non_nested(curr, lock, ip); curr->lockdep_depth--; @@ -3065,7 +3122,9 @@ static int __lock_is_held(struct lockdep_map *lock) int i; for (i = 0; i < curr->lockdep_depth; i++) { - if (curr->held_locks[i].instance == lock) + struct held_lock *hlock = curr->held_locks + i; + + if (match_held_lock(hlock, lock)) return 1; } @@ -3148,7 +3207,7 @@ void lock_acquire(struct lockdep_map *lock, unsigned int subclass, current->lockdep_recursion = 1; __lock_acquire(lock, subclass, trylock, read, check, - irqs_disabled_flags(flags), nest_lock, ip); + irqs_disabled_flags(flags), nest_lock, ip, 0); current->lockdep_recursion = 0; raw_local_irq_restore(flags); } @@ -3252,7 +3311,7 @@ __lock_contended(struct lockdep_map *lock, unsigned long ip) */ if (prev_hlock && prev_hlock->irq_context != hlock->irq_context) break; - if (hlock->instance == lock) + if (match_held_lock(hlock, lock)) goto found_it; prev_hlock = hlock; } @@ -3260,6 +3319,9 @@ __lock_contended(struct lockdep_map *lock, unsigned long ip) return; found_it: + if (hlock->instance != lock) + return; + hlock->waittime_stamp = sched_clock(); contention_point = lock_point(hlock_class(hlock)->contention_point, ip); @@ -3299,7 +3361,7 @@ __lock_acquired(struct lockdep_map *lock, unsigned long ip) */ if (prev_hlock && prev_hlock->irq_context != hlock->irq_context) break; - if (hlock->instance == lock) + if (match_held_lock(hlock, lock)) goto found_it; prev_hlock = hlock; } @@ -3307,6 +3369,9 @@ __lock_acquired(struct lockdep_map *lock, unsigned long ip) return; found_it: + if (hlock->instance != lock) + return; + cpu = smp_processor_id(); if (hlock->waittime_stamp) { now = sched_clock(); -- cgit v1.2.3 From e351b660fddd4df76cc4635f896d311ed0ff3752 Mon Sep 17 00:00:00 2001 From: Ming Lei Date: Wed, 22 Jul 2009 22:48:09 +0800 Subject: lockdep: Reintroduce generation count to make BFS faster We still can apply DaveM's generation count optimization to BFS, based on the following idea: - before doing each BFS, increase the global generation id by 1 - if one node in the graph has been visited, mark it as visited by storing the current global generation id into the node's dep_gen_id field - so we can decide if one node has been visited already, by comparing the node's dep_gen_id with the global generation id. By applying DaveM's generation count optimization to current implementation of BFS, we gain the following advantages: - we save MAX_LOCKDEP_ENTRIES/8 bytes memory; - we remove the bitmap_zero(bfs_accessed, MAX_LOCKDEP_ENTRIES); in each BFS, which is very time-consuming since MAX_LOCKDEP_ENTRIES may be very large.(16384UL) Signed-off-by: Ming Lei Signed-off-by: Peter Zijlstra Cc: "David S. Miller" LKML-Reference: <1248274089-6358-1-git-send-email-tom.leiming@gmail.com> Signed-off-by: Ingo Molnar --- include/linux/lockdep.h | 1 + kernel/lockdep.c | 11 ++++++----- 2 files changed, 7 insertions(+), 5 deletions(-) (limited to 'include') diff --git a/include/linux/lockdep.h b/include/linux/lockdep.h index 47d42eff6124..9ccf0e286b2a 100644 --- a/include/linux/lockdep.h +++ b/include/linux/lockdep.h @@ -58,6 +58,7 @@ struct lock_class { struct lockdep_subclass_key *key; unsigned int subclass; + unsigned int dep_gen_id; /* * IRQ/softirq usage tracking bits: diff --git a/kernel/lockdep.c b/kernel/lockdep.c index 0bb246e21cd7..2b443b5090a1 100644 --- a/kernel/lockdep.c +++ b/kernel/lockdep.c @@ -861,14 +861,15 @@ struct circular_queue { }; static struct circular_queue lock_cq; -static unsigned long bfs_accessed[BITS_TO_LONGS(MAX_LOCKDEP_ENTRIES)]; unsigned int max_bfs_queue_depth; +static unsigned int lockdep_dependency_gen_id; + static inline void __cq_init(struct circular_queue *cq) { cq->front = cq->rear = 0; - bitmap_zero(bfs_accessed, MAX_LOCKDEP_ENTRIES); + lockdep_dependency_gen_id++; } static inline int __cq_empty(struct circular_queue *cq) @@ -914,7 +915,7 @@ static inline void mark_lock_accessed(struct lock_list *lock, nr = lock - list_entries; WARN_ON(nr >= nr_list_entries); lock->parent = parent; - set_bit(nr, bfs_accessed); + lock->class->dep_gen_id = lockdep_dependency_gen_id; } static inline unsigned long lock_accessed(struct lock_list *lock) @@ -923,7 +924,7 @@ static inline unsigned long lock_accessed(struct lock_list *lock) nr = lock - list_entries; WARN_ON(nr >= nr_list_entries); - return test_bit(nr, bfs_accessed); + return lock->class->dep_gen_id == lockdep_dependency_gen_id; } static inline struct lock_list *get_lock_parent(struct lock_list *child) @@ -3604,7 +3605,7 @@ void __init lockdep_info(void) sizeof(struct lock_chain) * MAX_LOCKDEP_CHAINS + sizeof(struct list_head) * CHAINHASH_SIZE) / 1024 #ifdef CONFIG_PROVE_LOCKING - + sizeof(struct circular_queue) + sizeof(bfs_accessed) + + sizeof(struct circular_queue) #endif ); -- cgit v1.2.3 From 1e3e238e9c4bf9987b19185235cd0cdc21ea038c Mon Sep 17 00:00:00 2001 From: Hannes Eder Date: Sun, 2 Aug 2009 11:05:41 +0000 Subject: IPVS: use pr_err and friends instead of IP_VS_ERR and friends Since pr_err and friends are used instead of printk there is no point in keeping IP_VS_ERR and friends. Furthermore make use of '__func__' instead of hard coded function names. Signed-off-by: Hannes Eder Acked-by: Simon Horman Signed-off-by: David S. Miller --- include/net/ip_vs.h | 7 ++-- net/netfilter/ipvs/ip_vs_app.c | 16 +++++----- net/netfilter/ipvs/ip_vs_conn.c | 14 ++++---- net/netfilter/ipvs/ip_vs_core.c | 24 +++++++------- net/netfilter/ipvs/ip_vs_ctl.c | 62 +++++++++++++++++------------------- net/netfilter/ipvs/ip_vs_dh.c | 4 +-- net/netfilter/ipvs/ip_vs_ftp.c | 4 +-- net/netfilter/ipvs/ip_vs_lblc.c | 6 ++-- net/netfilter/ipvs/ip_vs_lblcr.c | 14 ++++---- net/netfilter/ipvs/ip_vs_lc.c | 2 +- net/netfilter/ipvs/ip_vs_nq.c | 2 +- net/netfilter/ipvs/ip_vs_proto.c | 2 +- net/netfilter/ipvs/ip_vs_proto_tcp.c | 2 +- net/netfilter/ipvs/ip_vs_proto_udp.c | 2 +- net/netfilter/ipvs/ip_vs_rr.c | 2 +- net/netfilter/ipvs/ip_vs_sched.c | 38 +++++++++++----------- net/netfilter/ipvs/ip_vs_sed.c | 2 +- net/netfilter/ipvs/ip_vs_sh.c | 2 +- net/netfilter/ipvs/ip_vs_sync.c | 50 ++++++++++++++--------------- net/netfilter/ipvs/ip_vs_wrr.c | 4 +-- net/netfilter/ipvs/ip_vs_xmit.c | 40 +++++++++++------------ 21 files changed, 147 insertions(+), 152 deletions(-) (limited to 'include') diff --git a/include/net/ip_vs.h b/include/net/ip_vs.h index 910820327bc4..1c8ee1b13651 100644 --- a/include/net/ip_vs.h +++ b/include/net/ip_vs.h @@ -150,13 +150,10 @@ static inline const char *ip_vs_dbg_addr(int af, char *buf, size_t buf_len, #endif #define IP_VS_BUG() BUG() -#define IP_VS_ERR(msg...) pr_err(msg) -#define IP_VS_INFO(msg...) pr_info(msg) -#define IP_VS_WARNING(msg...) pr_warning(msg) -#define IP_VS_ERR_RL(msg...) \ +#define IP_VS_ERR_RL(msg, ...) \ do { \ if (net_ratelimit()) \ - pr_err(msg); \ + pr_err(msg, ##__VA_ARGS__); \ } while (0) #ifdef CONFIG_IP_VS_DEBUG diff --git a/net/netfilter/ipvs/ip_vs_app.c b/net/netfilter/ipvs/ip_vs_app.c index c1781f80daf2..3c7e42735b60 100644 --- a/net/netfilter/ipvs/ip_vs_app.c +++ b/net/netfilter/ipvs/ip_vs_app.c @@ -265,12 +265,12 @@ static inline void vs_fix_seq(const struct ip_vs_seq *vseq, struct tcphdr *th) if (vseq->delta || vseq->previous_delta) { if(after(seq, vseq->init_seq)) { th->seq = htonl(seq + vseq->delta); - IP_VS_DBG(9, "vs_fix_seq(): added delta (%d) to seq\n", - vseq->delta); + IP_VS_DBG(9, "%s(): added delta (%d) to seq\n", + __func__, vseq->delta); } else { th->seq = htonl(seq + vseq->previous_delta); - IP_VS_DBG(9, "vs_fix_seq(): added previous_delta " - "(%d) to seq\n", vseq->previous_delta); + IP_VS_DBG(9, "%s(): added previous_delta (%d) to seq\n", + __func__, vseq->previous_delta); } } } @@ -294,14 +294,14 @@ vs_fix_ack_seq(const struct ip_vs_seq *vseq, struct tcphdr *th) to receive next, so compare it with init_seq+delta */ if(after(ack_seq, vseq->init_seq+vseq->delta)) { th->ack_seq = htonl(ack_seq - vseq->delta); - IP_VS_DBG(9, "vs_fix_ack_seq(): subtracted delta " - "(%d) from ack_seq\n", vseq->delta); + IP_VS_DBG(9, "%s(): subtracted delta " + "(%d) from ack_seq\n", __func__, vseq->delta); } else { th->ack_seq = htonl(ack_seq - vseq->previous_delta); - IP_VS_DBG(9, "vs_fix_ack_seq(): subtracted " + IP_VS_DBG(9, "%s(): subtracted " "previous_delta (%d) from ack_seq\n", - vseq->previous_delta); + __func__, vseq->previous_delta); } } } diff --git a/net/netfilter/ipvs/ip_vs_conn.c b/net/netfilter/ipvs/ip_vs_conn.c index 4173d7b1d4cc..27c30cf933da 100644 --- a/net/netfilter/ipvs/ip_vs_conn.c +++ b/net/netfilter/ipvs/ip_vs_conn.c @@ -153,8 +153,8 @@ static inline int ip_vs_conn_hash(struct ip_vs_conn *cp) atomic_inc(&cp->refcnt); ret = 1; } else { - IP_VS_ERR("ip_vs_conn_hash(): request for already hashed, " - "called from %p\n", __builtin_return_address(0)); + pr_err("%s(): request for already hashed, called from %pF\n", + __func__, __builtin_return_address(0)); ret = 0; } @@ -692,7 +692,7 @@ ip_vs_conn_new(int af, int proto, const union nf_inet_addr *caddr, __be16 cport, cp = kmem_cache_zalloc(ip_vs_conn_cachep, GFP_ATOMIC); if (cp == NULL) { - IP_VS_ERR_RL("ip_vs_conn_new: no memory available.\n"); + IP_VS_ERR_RL("%s(): no memory\n", __func__); return NULL; } @@ -1076,10 +1076,10 @@ int __init ip_vs_conn_init(void) return -ENOMEM; } - IP_VS_INFO("Connection hash table configured " - "(size=%d, memory=%ldKbytes)\n", - IP_VS_CONN_TAB_SIZE, - (long)(IP_VS_CONN_TAB_SIZE*sizeof(struct list_head))/1024); + pr_info("Connection hash table configured " + "(size=%d, memory=%ldKbytes)\n", + IP_VS_CONN_TAB_SIZE, + (long)(IP_VS_CONN_TAB_SIZE*sizeof(struct list_head))/1024); IP_VS_DBG(0, "Each connection entry needs %Zd bytes at least\n", sizeof(struct ip_vs_conn)); diff --git a/net/netfilter/ipvs/ip_vs_core.c b/net/netfilter/ipvs/ip_vs_core.c index 6811dcaca0f6..b227750af752 100644 --- a/net/netfilter/ipvs/ip_vs_core.c +++ b/net/netfilter/ipvs/ip_vs_core.c @@ -391,9 +391,9 @@ ip_vs_schedule(struct ip_vs_service *svc, const struct sk_buff *skb) */ if (!svc->fwmark && pptr[1] != svc->port) { if (!svc->port) - IP_VS_ERR("Schedule: port zero only supported " - "in persistent services, " - "check your ipvs configuration\n"); + pr_err("Schedule: port zero only supported " + "in persistent services, " + "check your ipvs configuration\n"); return NULL; } @@ -465,7 +465,7 @@ int ip_vs_leave(struct ip_vs_service *svc, struct sk_buff *skb, ip_vs_service_put(svc); /* create a new connection entry */ - IP_VS_DBG(6, "ip_vs_leave: create a cache_bypass entry\n"); + IP_VS_DBG(6, "%s(): create a cache_bypass entry\n", __func__); cp = ip_vs_conn_new(svc->af, iph.protocol, &iph.saddr, pptr[0], &iph.daddr, pptr[1], @@ -667,8 +667,8 @@ static int handle_response_icmp(int af, struct sk_buff *skb, unsigned int verdict = NF_DROP; if (IP_VS_FWD_METHOD(cp) != 0) { - IP_VS_ERR("shouldn't reach here, because the box is on the " - "half connection in the tun/dr module.\n"); + pr_err("shouldn't reach here, because the box is on the " + "half connection in the tun/dr module.\n"); } /* Ensure the checksum is correct */ @@ -1490,7 +1490,7 @@ static int __init ip_vs_init(void) ret = ip_vs_control_init(); if (ret < 0) { - IP_VS_ERR("can't setup control.\n"); + pr_err("can't setup control.\n"); goto cleanup_estimator; } @@ -1498,23 +1498,23 @@ static int __init ip_vs_init(void) ret = ip_vs_app_init(); if (ret < 0) { - IP_VS_ERR("can't setup application helper.\n"); + pr_err("can't setup application helper.\n"); goto cleanup_protocol; } ret = ip_vs_conn_init(); if (ret < 0) { - IP_VS_ERR("can't setup connection table.\n"); + pr_err("can't setup connection table.\n"); goto cleanup_app; } ret = nf_register_hooks(ip_vs_ops, ARRAY_SIZE(ip_vs_ops)); if (ret < 0) { - IP_VS_ERR("can't register hooks.\n"); + pr_err("can't register hooks.\n"); goto cleanup_conn; } - IP_VS_INFO("ipvs loaded.\n"); + pr_info("ipvs loaded.\n"); return ret; cleanup_conn: @@ -1537,7 +1537,7 @@ static void __exit ip_vs_cleanup(void) ip_vs_protocol_cleanup(); ip_vs_control_cleanup(); ip_vs_estimator_cleanup(); - IP_VS_INFO("ipvs unloaded.\n"); + pr_info("ipvs unloaded.\n"); } module_init(ip_vs_init); diff --git a/net/netfilter/ipvs/ip_vs_ctl.c b/net/netfilter/ipvs/ip_vs_ctl.c index e6133ea1ea4c..fba2892b99e1 100644 --- a/net/netfilter/ipvs/ip_vs_ctl.c +++ b/net/netfilter/ipvs/ip_vs_ctl.c @@ -343,8 +343,8 @@ static int ip_vs_svc_hash(struct ip_vs_service *svc) unsigned hash; if (svc->flags & IP_VS_SVC_F_HASHED) { - IP_VS_ERR("ip_vs_svc_hash(): request for already hashed, " - "called from %p\n", __builtin_return_address(0)); + pr_err("%s(): request for already hashed, called from %pF\n", + __func__, __builtin_return_address(0)); return 0; } @@ -377,8 +377,8 @@ static int ip_vs_svc_hash(struct ip_vs_service *svc) static int ip_vs_svc_unhash(struct ip_vs_service *svc) { if (!(svc->flags & IP_VS_SVC_F_HASHED)) { - IP_VS_ERR("ip_vs_svc_unhash(): request for unhash flagged, " - "called from %p\n", __builtin_return_address(0)); + pr_err("%s(): request for unhash flagged, called from %pF\n", + __func__, __builtin_return_address(0)); return 0; } @@ -844,7 +844,7 @@ ip_vs_new_dest(struct ip_vs_service *svc, struct ip_vs_dest_user_kern *udest, dest = kzalloc(sizeof(struct ip_vs_dest), GFP_ATOMIC); if (dest == NULL) { - IP_VS_ERR("ip_vs_new_dest: kmalloc failed.\n"); + pr_err("%s(): no memory.\n", __func__); return -ENOMEM; } @@ -888,13 +888,13 @@ ip_vs_add_dest(struct ip_vs_service *svc, struct ip_vs_dest_user_kern *udest) EnterFunction(2); if (udest->weight < 0) { - IP_VS_ERR("ip_vs_add_dest(): server weight less than zero\n"); + pr_err("%s(): server weight less than zero\n", __func__); return -ERANGE; } if (udest->l_threshold > udest->u_threshold) { - IP_VS_ERR("ip_vs_add_dest(): lower threshold is higher than " - "upper threshold\n"); + pr_err("%s(): lower threshold is higher than upper threshold\n", + __func__); return -ERANGE; } @@ -906,7 +906,7 @@ ip_vs_add_dest(struct ip_vs_service *svc, struct ip_vs_dest_user_kern *udest) dest = ip_vs_lookup_dest(svc, &daddr, dport); if (dest != NULL) { - IP_VS_DBG(1, "ip_vs_add_dest(): dest already exists\n"); + IP_VS_DBG(1, "%s(): dest already exists\n", __func__); return -EEXIST; } @@ -1000,13 +1000,13 @@ ip_vs_edit_dest(struct ip_vs_service *svc, struct ip_vs_dest_user_kern *udest) EnterFunction(2); if (udest->weight < 0) { - IP_VS_ERR("ip_vs_edit_dest(): server weight less than zero\n"); + pr_err("%s(): server weight less than zero\n", __func__); return -ERANGE; } if (udest->l_threshold > udest->u_threshold) { - IP_VS_ERR("ip_vs_edit_dest(): lower threshold is higher than " - "upper threshold\n"); + pr_err("%s(): lower threshold is higher than upper threshold\n", + __func__); return -ERANGE; } @@ -1018,7 +1018,7 @@ ip_vs_edit_dest(struct ip_vs_service *svc, struct ip_vs_dest_user_kern *udest) dest = ip_vs_lookup_dest(svc, &daddr, dport); if (dest == NULL) { - IP_VS_DBG(1, "ip_vs_edit_dest(): dest doesn't exist\n"); + IP_VS_DBG(1, "%s(): dest doesn't exist\n", __func__); return -ENOENT; } @@ -1118,7 +1118,7 @@ ip_vs_del_dest(struct ip_vs_service *svc, struct ip_vs_dest_user_kern *udest) dest = ip_vs_lookup_dest(svc, &udest->addr, dport); if (dest == NULL) { - IP_VS_DBG(1, "ip_vs_del_dest(): destination not found!\n"); + IP_VS_DBG(1, "%s(): destination not found!\n", __func__); return -ENOENT; } @@ -1164,8 +1164,7 @@ ip_vs_add_service(struct ip_vs_service_user_kern *u, /* Lookup the scheduler by 'u->sched_name' */ sched = ip_vs_scheduler_get(u->sched_name); if (sched == NULL) { - IP_VS_INFO("Scheduler module ip_vs_%s not found\n", - u->sched_name); + pr_info("Scheduler module ip_vs_%s not found\n", u->sched_name); ret = -ENOENT; goto out_mod_dec; } @@ -1179,7 +1178,7 @@ ip_vs_add_service(struct ip_vs_service_user_kern *u, svc = kzalloc(sizeof(struct ip_vs_service), GFP_ATOMIC); if (svc == NULL) { - IP_VS_DBG(1, "ip_vs_add_service: kmalloc failed.\n"); + IP_VS_DBG(1, "%s(): no memory\n", __func__); ret = -ENOMEM; goto out_err; } @@ -1262,8 +1261,7 @@ ip_vs_edit_service(struct ip_vs_service *svc, struct ip_vs_service_user_kern *u) */ sched = ip_vs_scheduler_get(u->sched_name); if (sched == NULL) { - IP_VS_INFO("Scheduler module ip_vs_%s not found\n", - u->sched_name); + pr_info("Scheduler module ip_vs_%s not found\n", u->sched_name); return -ENOENT; } old_sched = sched; @@ -2080,8 +2078,8 @@ do_ip_vs_set_ctl(struct sock *sk, int cmd, void __user *user, unsigned int len) return -EPERM; if (len != set_arglen[SET_CMDID(cmd)]) { - IP_VS_ERR("set_ctl: len %u != %u\n", - len, set_arglen[SET_CMDID(cmd)]); + pr_err("set_ctl: len %u != %u\n", + len, set_arglen[SET_CMDID(cmd)]); return -EINVAL; } @@ -2132,9 +2130,9 @@ do_ip_vs_set_ctl(struct sock *sk, int cmd, void __user *user, unsigned int len) /* Check for valid protocol: TCP or UDP, even for fwmark!=0 */ if (usvc.protocol != IPPROTO_TCP && usvc.protocol != IPPROTO_UDP) { - IP_VS_ERR("set_ctl: invalid protocol: %d %pI4:%d %s\n", - usvc.protocol, &usvc.addr.ip, - ntohs(usvc.port), usvc.sched_name); + pr_err("set_ctl: invalid protocol: %d %pI4:%d %s\n", + usvc.protocol, &usvc.addr.ip, + ntohs(usvc.port), usvc.sched_name); ret = -EFAULT; goto out_unlock; } @@ -2359,8 +2357,8 @@ do_ip_vs_get_ctl(struct sock *sk, int cmd, void __user *user, int *len) return -EPERM; if (*len < get_arglen[GET_CMDID(cmd)]) { - IP_VS_ERR("get_ctl: len %u < %u\n", - *len, get_arglen[GET_CMDID(cmd)]); + pr_err("get_ctl: len %u < %u\n", + *len, get_arglen[GET_CMDID(cmd)]); return -EINVAL; } @@ -2405,7 +2403,7 @@ do_ip_vs_get_ctl(struct sock *sk, int cmd, void __user *user, int *len) size = sizeof(*get) + sizeof(struct ip_vs_service_entry) * get->num_services; if (*len != size) { - IP_VS_ERR("length: %u != %u\n", *len, size); + pr_err("length: %u != %u\n", *len, size); ret = -EINVAL; goto out; } @@ -2445,7 +2443,7 @@ do_ip_vs_get_ctl(struct sock *sk, int cmd, void __user *user, int *len) size = sizeof(*get) + sizeof(struct ip_vs_dest_entry) * get->num_dests; if (*len != size) { - IP_VS_ERR("length: %u != %u\n", *len, size); + pr_err("length: %u != %u\n", *len, size); ret = -EINVAL; goto out; } @@ -3173,7 +3171,7 @@ static int ip_vs_genl_get_cmd(struct sk_buff *skb, struct genl_info *info) else if (cmd == IPVS_CMD_GET_CONFIG) reply_cmd = IPVS_CMD_SET_CONFIG; else { - IP_VS_ERR("unknown Generic Netlink command\n"); + pr_err("unknown Generic Netlink command\n"); return -EINVAL; } @@ -3238,7 +3236,7 @@ static int ip_vs_genl_get_cmd(struct sk_buff *skb, struct genl_info *info) goto out; nla_put_failure: - IP_VS_ERR("not enough space in Netlink message\n"); + pr_err("not enough space in Netlink message\n"); ret = -EMSGSIZE; out_err: @@ -3369,13 +3367,13 @@ int __init ip_vs_control_init(void) ret = nf_register_sockopt(&ip_vs_sockopts); if (ret) { - IP_VS_ERR("cannot register sockopt.\n"); + pr_err("cannot register sockopt.\n"); return ret; } ret = ip_vs_genl_register(); if (ret) { - IP_VS_ERR("cannot register Generic Netlink interface.\n"); + pr_err("cannot register Generic Netlink interface.\n"); nf_unregister_sockopt(&ip_vs_sockopts); return ret; } diff --git a/net/netfilter/ipvs/ip_vs_dh.c b/net/netfilter/ipvs/ip_vs_dh.c index d0c0594d1e2e..fe3e18834b91 100644 --- a/net/netfilter/ipvs/ip_vs_dh.c +++ b/net/netfilter/ipvs/ip_vs_dh.c @@ -150,7 +150,7 @@ static int ip_vs_dh_init_svc(struct ip_vs_service *svc) tbl = kmalloc(sizeof(struct ip_vs_dh_bucket)*IP_VS_DH_TAB_SIZE, GFP_ATOMIC); if (tbl == NULL) { - IP_VS_ERR("ip_vs_dh_init_svc(): no memory\n"); + pr_err("%s(): no memory\n", __func__); return -ENOMEM; } svc->sched_data = tbl; @@ -217,7 +217,7 @@ ip_vs_dh_schedule(struct ip_vs_service *svc, const struct sk_buff *skb) ip_vs_fill_iphdr(svc->af, skb_network_header(skb), &iph); - IP_VS_DBG(6, "ip_vs_dh_schedule(): Scheduling...\n"); + IP_VS_DBG(6, "%s(): Scheduling...\n", __func__); tbl = (struct ip_vs_dh_bucket *)svc->sched_data; dest = ip_vs_dh_get(svc->af, tbl, &iph.daddr); diff --git a/net/netfilter/ipvs/ip_vs_ftp.c b/net/netfilter/ipvs/ip_vs_ftp.c index 9c16a3f64c1b..33e2c799cba7 100644 --- a/net/netfilter/ipvs/ip_vs_ftp.c +++ b/net/netfilter/ipvs/ip_vs_ftp.c @@ -385,8 +385,8 @@ static int __init ip_vs_ftp_init(void) ret = register_ip_vs_app_inc(app, app->protocol, ports[i]); if (ret) break; - IP_VS_INFO("%s: loaded support on port[%d] = %d\n", - app->name, i, ports[i]); + pr_info("%s: loaded support on port[%d] = %d\n", + app->name, i, ports[i]); } if (ret) diff --git a/net/netfilter/ipvs/ip_vs_lblc.c b/net/netfilter/ipvs/ip_vs_lblc.c index 98fb185d890b..c1757f3620cd 100644 --- a/net/netfilter/ipvs/ip_vs_lblc.c +++ b/net/netfilter/ipvs/ip_vs_lblc.c @@ -202,7 +202,7 @@ ip_vs_lblc_new(struct ip_vs_lblc_table *tbl, const union nf_inet_addr *daddr, if (!en) { en = kmalloc(sizeof(*en), GFP_ATOMIC); if (!en) { - IP_VS_ERR("ip_vs_lblc_new(): no memory\n"); + pr_err("%s(): no memory\n", __func__); return NULL; } @@ -335,7 +335,7 @@ static int ip_vs_lblc_init_svc(struct ip_vs_service *svc) */ tbl = kmalloc(sizeof(*tbl), GFP_ATOMIC); if (tbl == NULL) { - IP_VS_ERR("ip_vs_lblc_init_svc(): no memory\n"); + pr_err("%s(): no memory\n", __func__); return -ENOMEM; } svc->sched_data = tbl; @@ -480,7 +480,7 @@ ip_vs_lblc_schedule(struct ip_vs_service *svc, const struct sk_buff *skb) ip_vs_fill_iphdr(svc->af, skb_network_header(skb), &iph); - IP_VS_DBG(6, "ip_vs_lblc_schedule(): Scheduling...\n"); + IP_VS_DBG(6, "%s(): Scheduling...\n", __func__); /* First look in our cache */ read_lock(&svc->sched_lock); diff --git a/net/netfilter/ipvs/ip_vs_lblcr.c b/net/netfilter/ipvs/ip_vs_lblcr.c index 5f5e5f4bad5e..715b57f9540d 100644 --- a/net/netfilter/ipvs/ip_vs_lblcr.c +++ b/net/netfilter/ipvs/ip_vs_lblcr.c @@ -111,7 +111,7 @@ ip_vs_dest_set_insert(struct ip_vs_dest_set *set, struct ip_vs_dest *dest) e = kmalloc(sizeof(*e), GFP_ATOMIC); if (e == NULL) { - IP_VS_ERR("ip_vs_dest_set_insert(): no memory\n"); + pr_err("%s(): no memory\n", __func__); return NULL; } @@ -205,8 +205,9 @@ static inline struct ip_vs_dest *ip_vs_dest_set_min(struct ip_vs_dest_set *set) } } - IP_VS_DBG_BUF(6, "ip_vs_dest_set_min: server %s:%d " + IP_VS_DBG_BUF(6, "%s(): server %s:%d " "activeconns %d refcnt %d weight %d overhead %d\n", + __func__, IP_VS_DBG_ADDR(least->af, &least->addr), ntohs(least->port), atomic_read(&least->activeconns), @@ -252,8 +253,9 @@ static inline struct ip_vs_dest *ip_vs_dest_set_max(struct ip_vs_dest_set *set) } } - IP_VS_DBG_BUF(6, "ip_vs_dest_set_max: server %s:%d " + IP_VS_DBG_BUF(6, "%s(): server %s:%d " "activeconns %d refcnt %d weight %d overhead %d\n", + __func__, IP_VS_DBG_ADDR(most->af, &most->addr), ntohs(most->port), atomic_read(&most->activeconns), atomic_read(&most->refcnt), @@ -377,7 +379,7 @@ ip_vs_lblcr_new(struct ip_vs_lblcr_table *tbl, const union nf_inet_addr *daddr, if (!en) { en = kmalloc(sizeof(*en), GFP_ATOMIC); if (!en) { - IP_VS_ERR("ip_vs_lblcr_new(): no memory\n"); + pr_err("%s(): no memory\n", __func__); return NULL; } @@ -511,7 +513,7 @@ static int ip_vs_lblcr_init_svc(struct ip_vs_service *svc) */ tbl = kmalloc(sizeof(*tbl), GFP_ATOMIC); if (tbl == NULL) { - IP_VS_ERR("ip_vs_lblcr_init_svc(): no memory\n"); + pr_err("%s(): no memory\n", __func__); return -ENOMEM; } svc->sched_data = tbl; @@ -657,7 +659,7 @@ ip_vs_lblcr_schedule(struct ip_vs_service *svc, const struct sk_buff *skb) ip_vs_fill_iphdr(svc->af, skb_network_header(skb), &iph); - IP_VS_DBG(6, "ip_vs_lblcr_schedule(): Scheduling...\n"); + IP_VS_DBG(6, "%s(): Scheduling...\n", __func__); /* First look in our cache */ read_lock(&svc->sched_lock); diff --git a/net/netfilter/ipvs/ip_vs_lc.c b/net/netfilter/ipvs/ip_vs_lc.c index 4ecd5e19c39a..4f69db1fac56 100644 --- a/net/netfilter/ipvs/ip_vs_lc.c +++ b/net/netfilter/ipvs/ip_vs_lc.c @@ -47,7 +47,7 @@ ip_vs_lc_schedule(struct ip_vs_service *svc, const struct sk_buff *skb) struct ip_vs_dest *dest, *least = NULL; unsigned int loh = 0, doh; - IP_VS_DBG(6, "ip_vs_lc_schedule(): Scheduling...\n"); + IP_VS_DBG(6, "%s(): Scheduling...\n", __func__); /* * Simply select the server with the least number of diff --git a/net/netfilter/ipvs/ip_vs_nq.c b/net/netfilter/ipvs/ip_vs_nq.c index 2224478bdea8..c413e1830823 100644 --- a/net/netfilter/ipvs/ip_vs_nq.c +++ b/net/netfilter/ipvs/ip_vs_nq.c @@ -60,7 +60,7 @@ ip_vs_nq_schedule(struct ip_vs_service *svc, const struct sk_buff *skb) struct ip_vs_dest *dest, *least = NULL; unsigned int loh = 0, doh; - IP_VS_DBG(6, "ip_vs_nq_schedule(): Scheduling...\n"); + IP_VS_DBG(6, "%s(): Scheduling...\n", __func__); /* * We calculate the load of each dest server as follows: diff --git a/net/netfilter/ipvs/ip_vs_proto.c b/net/netfilter/ipvs/ip_vs_proto.c index a95bc4021c90..85c8892e1e8b 100644 --- a/net/netfilter/ipvs/ip_vs_proto.c +++ b/net/netfilter/ipvs/ip_vs_proto.c @@ -262,7 +262,7 @@ int __init ip_vs_protocol_init(void) #ifdef CONFIG_IP_VS_PROTO_ESP REGISTER_PROTOCOL(&ip_vs_protocol_esp); #endif - IP_VS_INFO("Registered protocols (%s)\n", &protocols[2]); + pr_info("Registered protocols (%s)\n", &protocols[2]); return 0; } diff --git a/net/netfilter/ipvs/ip_vs_proto_tcp.c b/net/netfilter/ipvs/ip_vs_proto_tcp.c index c36c80d3a2b4..2278e141489e 100644 --- a/net/netfilter/ipvs/ip_vs_proto_tcp.c +++ b/net/netfilter/ipvs/ip_vs_proto_tcp.c @@ -664,7 +664,7 @@ tcp_app_conn_bind(struct ip_vs_conn *cp) break; spin_unlock(&tcp_app_lock); - IP_VS_DBG_BUF(9, "%s: Binding conn %s:%u->" + IP_VS_DBG_BUF(9, "%s(): Binding conn %s:%u->" "%s:%u to app %s on port %u\n", __func__, IP_VS_DBG_ADDR(cp->af, &cp->caddr), diff --git a/net/netfilter/ipvs/ip_vs_proto_udp.c b/net/netfilter/ipvs/ip_vs_proto_udp.c index 96ebe40bc537..33a05d3684d9 100644 --- a/net/netfilter/ipvs/ip_vs_proto_udp.c +++ b/net/netfilter/ipvs/ip_vs_proto_udp.c @@ -445,7 +445,7 @@ static int udp_app_conn_bind(struct ip_vs_conn *cp) break; spin_unlock(&udp_app_lock); - IP_VS_DBG_BUF(9, "%s: Binding conn %s:%u->" + IP_VS_DBG_BUF(9, "%s(): Binding conn %s:%u->" "%s:%u to app %s on port %u\n", __func__, IP_VS_DBG_ADDR(cp->af, &cp->caddr), diff --git a/net/netfilter/ipvs/ip_vs_rr.c b/net/netfilter/ipvs/ip_vs_rr.c index b01007e1c11e..e210f37d8ea2 100644 --- a/net/netfilter/ipvs/ip_vs_rr.c +++ b/net/netfilter/ipvs/ip_vs_rr.c @@ -51,7 +51,7 @@ ip_vs_rr_schedule(struct ip_vs_service *svc, const struct sk_buff *skb) struct list_head *p, *q; struct ip_vs_dest *dest; - IP_VS_DBG(6, "ip_vs_rr_schedule(): Scheduling...\n"); + IP_VS_DBG(6, "%s(): Scheduling...\n", __func__); write_lock(&svc->sched_lock); p = (struct list_head *)svc->sched_data; diff --git a/net/netfilter/ipvs/ip_vs_sched.c b/net/netfilter/ipvs/ip_vs_sched.c index 87bc5ea0ef29..bbc1ac795952 100644 --- a/net/netfilter/ipvs/ip_vs_sched.c +++ b/net/netfilter/ipvs/ip_vs_sched.c @@ -47,11 +47,11 @@ int ip_vs_bind_scheduler(struct ip_vs_service *svc, int ret; if (svc == NULL) { - IP_VS_ERR("ip_vs_bind_scheduler(): svc arg NULL\n"); + pr_err("%s(): svc arg NULL\n", __func__); return -EINVAL; } if (scheduler == NULL) { - IP_VS_ERR("ip_vs_bind_scheduler(): scheduler arg NULL\n"); + pr_err("%s(): scheduler arg NULL\n", __func__); return -EINVAL; } @@ -60,7 +60,7 @@ int ip_vs_bind_scheduler(struct ip_vs_service *svc, if (scheduler->init_service) { ret = scheduler->init_service(svc); if (ret) { - IP_VS_ERR("ip_vs_bind_scheduler(): init error\n"); + pr_err("%s(): init error\n", __func__); return ret; } } @@ -77,19 +77,19 @@ int ip_vs_unbind_scheduler(struct ip_vs_service *svc) struct ip_vs_scheduler *sched; if (svc == NULL) { - IP_VS_ERR("ip_vs_unbind_scheduler(): svc arg NULL\n"); + pr_err("%s(): svc arg NULL\n", __func__); return -EINVAL; } sched = svc->scheduler; if (sched == NULL) { - IP_VS_ERR("ip_vs_unbind_scheduler(): svc isn't bound\n"); + pr_err("%s(): svc isn't bound\n", __func__); return -EINVAL; } if (sched->done_service) { if (sched->done_service(svc) != 0) { - IP_VS_ERR("ip_vs_unbind_scheduler(): done error\n"); + pr_err("%s(): done error\n", __func__); return -EINVAL; } } @@ -106,8 +106,7 @@ static struct ip_vs_scheduler *ip_vs_sched_getbyname(const char *sched_name) { struct ip_vs_scheduler *sched; - IP_VS_DBG(2, "ip_vs_sched_getbyname(): sched_name \"%s\"\n", - sched_name); + IP_VS_DBG(2, "%s(): sched_name \"%s\"\n", __func__, sched_name); read_lock_bh(&__ip_vs_sched_lock); @@ -173,12 +172,12 @@ int register_ip_vs_scheduler(struct ip_vs_scheduler *scheduler) struct ip_vs_scheduler *sched; if (!scheduler) { - IP_VS_ERR("register_ip_vs_scheduler(): NULL arg\n"); + pr_err("%s(): NULL arg\n", __func__); return -EINVAL; } if (!scheduler->name) { - IP_VS_ERR("register_ip_vs_scheduler(): NULL scheduler_name\n"); + pr_err("%s(): NULL scheduler_name\n", __func__); return -EINVAL; } @@ -190,8 +189,8 @@ int register_ip_vs_scheduler(struct ip_vs_scheduler *scheduler) if (!list_empty(&scheduler->n_list)) { write_unlock_bh(&__ip_vs_sched_lock); ip_vs_use_count_dec(); - IP_VS_ERR("register_ip_vs_scheduler(): [%s] scheduler " - "already linked\n", scheduler->name); + pr_err("%s(): [%s] scheduler already linked\n", + __func__, scheduler->name); return -EINVAL; } @@ -203,9 +202,8 @@ int register_ip_vs_scheduler(struct ip_vs_scheduler *scheduler) if (strcmp(scheduler->name, sched->name) == 0) { write_unlock_bh(&__ip_vs_sched_lock); ip_vs_use_count_dec(); - IP_VS_ERR("register_ip_vs_scheduler(): [%s] scheduler " - "already existed in the system\n", - scheduler->name); + pr_err("%s(): [%s] scheduler already existed " + "in the system\n", __func__, scheduler->name); return -EINVAL; } } @@ -215,7 +213,7 @@ int register_ip_vs_scheduler(struct ip_vs_scheduler *scheduler) list_add(&scheduler->n_list, &ip_vs_schedulers); write_unlock_bh(&__ip_vs_sched_lock); - IP_VS_INFO("[%s] scheduler registered.\n", scheduler->name); + pr_info("[%s] scheduler registered.\n", scheduler->name); return 0; } @@ -227,15 +225,15 @@ int register_ip_vs_scheduler(struct ip_vs_scheduler *scheduler) int unregister_ip_vs_scheduler(struct ip_vs_scheduler *scheduler) { if (!scheduler) { - IP_VS_ERR( "unregister_ip_vs_scheduler(): NULL arg\n"); + pr_err("%s(): NULL arg\n", __func__); return -EINVAL; } write_lock_bh(&__ip_vs_sched_lock); if (list_empty(&scheduler->n_list)) { write_unlock_bh(&__ip_vs_sched_lock); - IP_VS_ERR("unregister_ip_vs_scheduler(): [%s] scheduler " - "is not in the list. failed\n", scheduler->name); + pr_err("%s(): [%s] scheduler is not in the list. failed\n", + __func__, scheduler->name); return -EINVAL; } @@ -248,7 +246,7 @@ int unregister_ip_vs_scheduler(struct ip_vs_scheduler *scheduler) /* decrease the module use count */ ip_vs_use_count_dec(); - IP_VS_INFO("[%s] scheduler unregistered.\n", scheduler->name); + pr_info("[%s] scheduler unregistered.\n", scheduler->name); return 0; } diff --git a/net/netfilter/ipvs/ip_vs_sed.c b/net/netfilter/ipvs/ip_vs_sed.c index 4f745dd86dd8..1ab75a9dc400 100644 --- a/net/netfilter/ipvs/ip_vs_sed.c +++ b/net/netfilter/ipvs/ip_vs_sed.c @@ -64,7 +64,7 @@ ip_vs_sed_schedule(struct ip_vs_service *svc, const struct sk_buff *skb) struct ip_vs_dest *dest, *least; unsigned int loh, doh; - IP_VS_DBG(6, "ip_vs_sed_schedule(): Scheduling...\n"); + IP_VS_DBG(6, "%s(): Scheduling...\n", __func__); /* * We calculate the load of each dest server as follows: diff --git a/net/netfilter/ipvs/ip_vs_sh.c b/net/netfilter/ipvs/ip_vs_sh.c index fb4d2d23f2fe..8e6cfd36e6f0 100644 --- a/net/netfilter/ipvs/ip_vs_sh.c +++ b/net/netfilter/ipvs/ip_vs_sh.c @@ -147,7 +147,7 @@ static int ip_vs_sh_init_svc(struct ip_vs_service *svc) tbl = kmalloc(sizeof(struct ip_vs_sh_bucket)*IP_VS_SH_TAB_SIZE, GFP_ATOMIC); if (tbl == NULL) { - IP_VS_ERR("ip_vs_sh_init_svc(): no memory\n"); + pr_err("%s(): no memory\n", __func__); return -ENOMEM; } svc->sched_data = tbl; diff --git a/net/netfilter/ipvs/ip_vs_sync.c b/net/netfilter/ipvs/ip_vs_sync.c index cc04c99815fd..e177f0dc2084 100644 --- a/net/netfilter/ipvs/ip_vs_sync.c +++ b/net/netfilter/ipvs/ip_vs_sync.c @@ -246,7 +246,7 @@ void ip_vs_sync_conn(struct ip_vs_conn *cp) if (!curr_sb) { if (!(curr_sb=ip_vs_sync_buff_create())) { spin_unlock(&curr_sb_lock); - IP_VS_ERR("ip_vs_sync_buff_create failed.\n"); + pr_err("ip_vs_sync_buff_create failed.\n"); return; } } @@ -412,7 +412,7 @@ static void ip_vs_process_message(const char *buffer, const size_t buflen) if (dest) atomic_dec(&dest->refcnt); if (!cp) { - IP_VS_ERR("ip_vs_conn_new failed\n"); + pr_err("ip_vs_conn_new failed\n"); return; } } else if (!cp->dest) { @@ -580,8 +580,8 @@ static int bind_mcastif_addr(struct socket *sock, char *ifname) addr = inet_select_addr(dev, 0, RT_SCOPE_UNIVERSE); if (!addr) - IP_VS_ERR("You probably need to specify IP address on " - "multicast interface.\n"); + pr_err("You probably need to specify IP address on " + "multicast interface.\n"); IP_VS_DBG(7, "binding socket with (%s) %pI4\n", ifname, &addr); @@ -605,13 +605,13 @@ static struct socket * make_send_sock(void) /* First create a socket */ result = sock_create_kern(PF_INET, SOCK_DGRAM, IPPROTO_UDP, &sock); if (result < 0) { - IP_VS_ERR("Error during creation of socket; terminating\n"); + pr_err("Error during creation of socket; terminating\n"); return ERR_PTR(result); } result = set_mcast_if(sock->sk, ip_vs_master_mcast_ifn); if (result < 0) { - IP_VS_ERR("Error setting outbound mcast interface\n"); + pr_err("Error setting outbound mcast interface\n"); goto error; } @@ -620,14 +620,14 @@ static struct socket * make_send_sock(void) result = bind_mcastif_addr(sock, ip_vs_master_mcast_ifn); if (result < 0) { - IP_VS_ERR("Error binding address of the mcast interface\n"); + pr_err("Error binding address of the mcast interface\n"); goto error; } result = sock->ops->connect(sock, (struct sockaddr *) &mcast_addr, sizeof(struct sockaddr), 0); if (result < 0) { - IP_VS_ERR("Error connecting to the multicast addr\n"); + pr_err("Error connecting to the multicast addr\n"); goto error; } @@ -650,7 +650,7 @@ static struct socket * make_receive_sock(void) /* First create a socket */ result = sock_create_kern(PF_INET, SOCK_DGRAM, IPPROTO_UDP, &sock); if (result < 0) { - IP_VS_ERR("Error during creation of socket; terminating\n"); + pr_err("Error during creation of socket; terminating\n"); return ERR_PTR(result); } @@ -660,7 +660,7 @@ static struct socket * make_receive_sock(void) result = sock->ops->bind(sock, (struct sockaddr *) &mcast_addr, sizeof(struct sockaddr)); if (result < 0) { - IP_VS_ERR("Error binding to the multicast addr\n"); + pr_err("Error binding to the multicast addr\n"); goto error; } @@ -669,7 +669,7 @@ static struct socket * make_receive_sock(void) (struct in_addr *) &mcast_addr.sin_addr, ip_vs_backup_mcast_ifn); if (result < 0) { - IP_VS_ERR("Error joining to the multicast group\n"); + pr_err("Error joining to the multicast group\n"); goto error; } @@ -709,7 +709,7 @@ ip_vs_send_sync_msg(struct socket *sock, struct ip_vs_sync_mesg *msg) msg->size = htons(msg->size); if (ip_vs_send_async(sock, (char *)msg, msize) != msize) - IP_VS_ERR("ip_vs_send_async error\n"); + pr_err("ip_vs_send_async error\n"); } static int @@ -740,9 +740,9 @@ static int sync_thread_master(void *data) struct ip_vs_sync_thread_data *tinfo = data; struct ip_vs_sync_buff *sb; - IP_VS_INFO("sync thread started: state = MASTER, mcast_ifn = %s, " - "syncid = %d\n", - ip_vs_master_mcast_ifn, ip_vs_master_syncid); + pr_info("sync thread started: state = MASTER, mcast_ifn = %s, " + "syncid = %d\n", + ip_vs_master_mcast_ifn, ip_vs_master_syncid); while (!kthread_should_stop()) { while ((sb = sb_dequeue())) { @@ -783,9 +783,9 @@ static int sync_thread_backup(void *data) struct ip_vs_sync_thread_data *tinfo = data; int len; - IP_VS_INFO("sync thread started: state = BACKUP, mcast_ifn = %s, " - "syncid = %d\n", - ip_vs_backup_mcast_ifn, ip_vs_backup_syncid); + pr_info("sync thread started: state = BACKUP, mcast_ifn = %s, " + "syncid = %d\n", + ip_vs_backup_mcast_ifn, ip_vs_backup_syncid); while (!kthread_should_stop()) { wait_event_interruptible(*tinfo->sock->sk->sk_sleep, @@ -797,7 +797,7 @@ static int sync_thread_backup(void *data) len = ip_vs_receive(tinfo->sock, tinfo->buf, sync_recv_mesg_maxlen); if (len <= 0) { - IP_VS_ERR("receiving message error\n"); + pr_err("receiving message error\n"); break; } @@ -827,7 +827,7 @@ int start_sync_thread(int state, char *mcast_ifn, __u8 syncid) int (*threadfn)(void *data); int result = -ENOMEM; - IP_VS_DBG(7, "%s: pid %d\n", __func__, task_pid_nr(current)); + IP_VS_DBG(7, "%s(): pid %d\n", __func__, task_pid_nr(current)); IP_VS_DBG(7, "Each ip_vs_sync_conn entry needs %Zd bytes\n", sizeof(struct ip_vs_sync_conn)); @@ -904,14 +904,14 @@ out: int stop_sync_thread(int state) { - IP_VS_DBG(7, "%s: pid %d\n", __func__, task_pid_nr(current)); + IP_VS_DBG(7, "%s(): pid %d\n", __func__, task_pid_nr(current)); if (state == IP_VS_STATE_MASTER) { if (!sync_master_thread) return -ESRCH; - IP_VS_INFO("stopping master sync thread %d ...\n", - task_pid_nr(sync_master_thread)); + pr_info("stopping master sync thread %d ...\n", + task_pid_nr(sync_master_thread)); /* * The lock synchronizes with sb_queue_tail(), so that we don't @@ -928,8 +928,8 @@ int stop_sync_thread(int state) if (!sync_backup_thread) return -ESRCH; - IP_VS_INFO("stopping backup sync thread %d ...\n", - task_pid_nr(sync_backup_thread)); + pr_info("stopping backup sync thread %d ...\n", + task_pid_nr(sync_backup_thread)); ip_vs_sync_state &= ~IP_VS_STATE_BACKUP; kthread_stop(sync_backup_thread); diff --git a/net/netfilter/ipvs/ip_vs_wrr.c b/net/netfilter/ipvs/ip_vs_wrr.c index c39ebb6c5a54..70ff82cda57d 100644 --- a/net/netfilter/ipvs/ip_vs_wrr.c +++ b/net/netfilter/ipvs/ip_vs_wrr.c @@ -97,7 +97,7 @@ static int ip_vs_wrr_init_svc(struct ip_vs_service *svc) */ mark = kmalloc(sizeof(struct ip_vs_wrr_mark), GFP_ATOMIC); if (mark == NULL) { - IP_VS_ERR("ip_vs_wrr_init_svc(): no memory\n"); + pr_err("%s(): no memory\n", __func__); return -ENOMEM; } mark->cl = &svc->destinations; @@ -144,7 +144,7 @@ ip_vs_wrr_schedule(struct ip_vs_service *svc, const struct sk_buff *skb) struct ip_vs_wrr_mark *mark = svc->sched_data; struct list_head *p; - IP_VS_DBG(6, "ip_vs_wrr_schedule(): Scheduling...\n"); + IP_VS_DBG(6, "%s(): Scheduling...\n", __func__); /* * This loop will always terminate, because mark->cw in (0, max_weight] diff --git a/net/netfilter/ipvs/ip_vs_xmit.c b/net/netfilter/ipvs/ip_vs_xmit.c index 061e76dfdad9..30b3189bd29c 100644 --- a/net/netfilter/ipvs/ip_vs_xmit.c +++ b/net/netfilter/ipvs/ip_vs_xmit.c @@ -238,8 +238,8 @@ ip_vs_bypass_xmit(struct sk_buff *skb, struct ip_vs_conn *cp, EnterFunction(10); if (ip_route_output_key(&init_net, &rt, &fl)) { - IP_VS_DBG_RL("ip_vs_bypass_xmit(): ip_route_output error, dest: %pI4\n", - &iph->daddr); + IP_VS_DBG_RL("%s(): ip_route_output error, dest: %pI4\n", + __func__, &iph->daddr); goto tx_error_icmp; } @@ -248,7 +248,7 @@ ip_vs_bypass_xmit(struct sk_buff *skb, struct ip_vs_conn *cp, if ((skb->len > mtu) && (iph->frag_off & htons(IP_DF))) { ip_rt_put(rt); icmp_send(skb, ICMP_DEST_UNREACH,ICMP_FRAG_NEEDED, htonl(mtu)); - IP_VS_DBG_RL("ip_vs_bypass_xmit(): frag needed\n"); + IP_VS_DBG_RL("%s(): frag needed\n", __func__); goto tx_error; } @@ -302,8 +302,8 @@ ip_vs_bypass_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp, rt = (struct rt6_info *)ip6_route_output(&init_net, NULL, &fl); if (!rt) { - IP_VS_DBG_RL("ip_vs_bypass_xmit_v6(): ip6_route_output error, dest: %pI6\n", - &iph->daddr); + IP_VS_DBG_RL("%s(): ip6_route_output error, dest: %pI6\n", + __func__, &iph->daddr); goto tx_error_icmp; } @@ -312,7 +312,7 @@ ip_vs_bypass_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp, if (skb->len > mtu) { dst_release(&rt->u.dst); icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu, skb->dev); - IP_VS_DBG_RL("ip_vs_bypass_xmit_v6(): frag needed\n"); + IP_VS_DBG_RL("%s(): frag needed\n", __func__); goto tx_error; } @@ -539,9 +539,9 @@ ip_vs_tunnel_xmit(struct sk_buff *skb, struct ip_vs_conn *cp, EnterFunction(10); if (skb->protocol != htons(ETH_P_IP)) { - IP_VS_DBG_RL("ip_vs_tunnel_xmit(): protocol error, " + IP_VS_DBG_RL("%s(): protocol error, " "ETH_P_IP: %d, skb protocol: %d\n", - htons(ETH_P_IP), skb->protocol); + __func__, htons(ETH_P_IP), skb->protocol); goto tx_error; } @@ -553,7 +553,7 @@ ip_vs_tunnel_xmit(struct sk_buff *skb, struct ip_vs_conn *cp, mtu = dst_mtu(&rt->u.dst) - sizeof(struct iphdr); if (mtu < 68) { ip_rt_put(rt); - IP_VS_DBG_RL("ip_vs_tunnel_xmit(): mtu less than 68\n"); + IP_VS_DBG_RL("%s(): mtu less than 68\n", __func__); goto tx_error; } if (skb_dst(skb)) @@ -565,7 +565,7 @@ ip_vs_tunnel_xmit(struct sk_buff *skb, struct ip_vs_conn *cp, && mtu < ntohs(old_iph->tot_len)) { icmp_send(skb, ICMP_DEST_UNREACH,ICMP_FRAG_NEEDED, htonl(mtu)); ip_rt_put(rt); - IP_VS_DBG_RL("ip_vs_tunnel_xmit(): frag needed\n"); + IP_VS_DBG_RL("%s(): frag needed\n", __func__); goto tx_error; } @@ -581,7 +581,7 @@ ip_vs_tunnel_xmit(struct sk_buff *skb, struct ip_vs_conn *cp, if (!new_skb) { ip_rt_put(rt); kfree_skb(skb); - IP_VS_ERR_RL("ip_vs_tunnel_xmit(): no memory\n"); + IP_VS_ERR_RL("%s(): no memory\n", __func__); return NF_STOLEN; } kfree_skb(skb); @@ -649,9 +649,9 @@ ip_vs_tunnel_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp, EnterFunction(10); if (skb->protocol != htons(ETH_P_IPV6)) { - IP_VS_DBG_RL("ip_vs_tunnel_xmit_v6(): protocol error, " + IP_VS_DBG_RL("%s(): protocol error, " "ETH_P_IPV6: %d, skb protocol: %d\n", - htons(ETH_P_IPV6), skb->protocol); + __func__, htons(ETH_P_IPV6), skb->protocol); goto tx_error; } @@ -665,7 +665,7 @@ ip_vs_tunnel_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp, /* TODO IPv6: do we need this check in IPv6? */ if (mtu < 1280) { dst_release(&rt->u.dst); - IP_VS_DBG_RL("ip_vs_tunnel_xmit_v6(): mtu less than 1280\n"); + IP_VS_DBG_RL("%s(): mtu less than 1280\n", __func__); goto tx_error; } if (skb_dst(skb)) @@ -674,7 +674,7 @@ ip_vs_tunnel_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp, if (mtu < ntohs(old_iph->payload_len) + sizeof(struct ipv6hdr)) { icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu, skb->dev); dst_release(&rt->u.dst); - IP_VS_DBG_RL("ip_vs_tunnel_xmit_v6(): frag needed\n"); + IP_VS_DBG_RL("%s(): frag needed\n", __func__); goto tx_error; } @@ -690,7 +690,7 @@ ip_vs_tunnel_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp, if (!new_skb) { dst_release(&rt->u.dst); kfree_skb(skb); - IP_VS_ERR_RL("ip_vs_tunnel_xmit_v6(): no memory\n"); + IP_VS_ERR_RL("%s(): no memory\n", __func__); return NF_STOLEN; } kfree_skb(skb); @@ -763,7 +763,7 @@ ip_vs_dr_xmit(struct sk_buff *skb, struct ip_vs_conn *cp, if ((iph->frag_off & htons(IP_DF)) && skb->len > mtu) { icmp_send(skb, ICMP_DEST_UNREACH,ICMP_FRAG_NEEDED, htonl(mtu)); ip_rt_put(rt); - IP_VS_DBG_RL("ip_vs_dr_xmit(): frag needed\n"); + IP_VS_DBG_RL("%s(): frag needed\n", __func__); goto tx_error; } @@ -816,7 +816,7 @@ ip_vs_dr_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp, if (skb->len > mtu) { icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu, skb->dev); dst_release(&rt->u.dst); - IP_VS_DBG_RL("ip_vs_dr_xmit_v6(): frag needed\n"); + IP_VS_DBG_RL("%s(): frag needed\n", __func__); goto tx_error; } @@ -891,7 +891,7 @@ ip_vs_icmp_xmit(struct sk_buff *skb, struct ip_vs_conn *cp, if ((skb->len > mtu) && (ip_hdr(skb)->frag_off & htons(IP_DF))) { ip_rt_put(rt); icmp_send(skb, ICMP_DEST_UNREACH, ICMP_FRAG_NEEDED, htonl(mtu)); - IP_VS_DBG_RL("ip_vs_in_icmp(): frag needed\n"); + IP_VS_DBG_RL("%s(): frag needed\n", __func__); goto tx_error; } @@ -966,7 +966,7 @@ ip_vs_icmp_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp, if (skb->len > mtu) { dst_release(&rt->u.dst); icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu, skb->dev); - IP_VS_DBG_RL("ip_vs_in_icmp(): frag needed\n"); + IP_VS_DBG_RL("%s(): frag needed\n", __func__); goto tx_error; } -- cgit v1.2.3 From e4c4e448cf557921ffbbbd6d6ddac81fdceacb4f Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 30 Jul 2009 03:15:07 +0000 Subject: neigh: Convert garbage collection from softirq to workqueue Current neigh_periodic_timer() function is fired by timer IRQ, and scans one hash bucket each round (very litle work in fact) As we are supposed to scan whole hash table in 15 seconds, this means neigh_periodic_timer() can be fired very often. (depending on the number of concurrent hash entries we stored in this table) Converting this to a workqueue permits scanning whole table, minimizing icache pollution, and firing this work every 15 seconds, independantly of hash table size. This 15 seconds delay is not a hard number, as work is a deferrable one. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- include/net/neighbour.h | 4 +-- net/core/neighbour.c | 89 ++++++++++++++++++++++++------------------------- 2 files changed, 45 insertions(+), 48 deletions(-) (limited to 'include') diff --git a/include/net/neighbour.h b/include/net/neighbour.h index d8d790e56d3d..18b69b6cecaf 100644 --- a/include/net/neighbour.h +++ b/include/net/neighbour.h @@ -24,6 +24,7 @@ #include #include +#include #include /* @@ -167,7 +168,7 @@ struct neigh_table int gc_thresh2; int gc_thresh3; unsigned long last_flush; - struct timer_list gc_timer; + struct delayed_work gc_work; struct timer_list proxy_timer; struct sk_buff_head proxy_queue; atomic_t entries; @@ -178,7 +179,6 @@ struct neigh_table struct neighbour **hash_buckets; unsigned int hash_mask; __u32 hash_rnd; - unsigned int hash_chain_gc; struct pneigh_entry **phash_buckets; }; diff --git a/net/core/neighbour.c b/net/core/neighbour.c index c6f9ad8e4c7a..e587e6819698 100644 --- a/net/core/neighbour.c +++ b/net/core/neighbour.c @@ -692,75 +692,74 @@ static void neigh_connect(struct neighbour *neigh) hh->hh_output = neigh->ops->hh_output; } -static void neigh_periodic_timer(unsigned long arg) +static void neigh_periodic_work(struct work_struct *work) { - struct neigh_table *tbl = (struct neigh_table *)arg; + struct neigh_table *tbl = container_of(work, struct neigh_table, gc_work.work); struct neighbour *n, **np; - unsigned long expire, now = jiffies; + unsigned int i; NEIGH_CACHE_STAT_INC(tbl, periodic_gc_runs); - write_lock(&tbl->lock); + write_lock_bh(&tbl->lock); /* * periodically recompute ReachableTime from random function */ - if (time_after(now, tbl->last_rand + 300 * HZ)) { + if (time_after(jiffies, tbl->last_rand + 300 * HZ)) { struct neigh_parms *p; - tbl->last_rand = now; + tbl->last_rand = jiffies; for (p = &tbl->parms; p; p = p->next) p->reachable_time = neigh_rand_reach_time(p->base_reachable_time); } - np = &tbl->hash_buckets[tbl->hash_chain_gc]; - tbl->hash_chain_gc = ((tbl->hash_chain_gc + 1) & tbl->hash_mask); + for (i = 0 ; i <= tbl->hash_mask; i++) { + np = &tbl->hash_buckets[i]; - while ((n = *np) != NULL) { - unsigned int state; + while ((n = *np) != NULL) { + unsigned int state; - write_lock(&n->lock); + write_lock(&n->lock); - state = n->nud_state; - if (state & (NUD_PERMANENT | NUD_IN_TIMER)) { - write_unlock(&n->lock); - goto next_elt; - } + state = n->nud_state; + if (state & (NUD_PERMANENT | NUD_IN_TIMER)) { + write_unlock(&n->lock); + goto next_elt; + } - if (time_before(n->used, n->confirmed)) - n->used = n->confirmed; + if (time_before(n->used, n->confirmed)) + n->used = n->confirmed; - if (atomic_read(&n->refcnt) == 1 && - (state == NUD_FAILED || - time_after(now, n->used + n->parms->gc_staletime))) { - *np = n->next; - n->dead = 1; + if (atomic_read(&n->refcnt) == 1 && + (state == NUD_FAILED || + time_after(jiffies, n->used + n->parms->gc_staletime))) { + *np = n->next; + n->dead = 1; + write_unlock(&n->lock); + neigh_cleanup_and_release(n); + continue; + } write_unlock(&n->lock); - neigh_cleanup_and_release(n); - continue; - } - write_unlock(&n->lock); next_elt: - np = &n->next; + np = &n->next; + } + /* + * It's fine to release lock here, even if hash table + * grows while we are preempted. + */ + write_unlock_bh(&tbl->lock); + cond_resched(); + write_lock_bh(&tbl->lock); } - /* Cycle through all hash buckets every base_reachable_time/2 ticks. * ARP entry timeouts range from 1/2 base_reachable_time to 3/2 * base_reachable_time. */ - expire = tbl->parms.base_reachable_time >> 1; - expire /= (tbl->hash_mask + 1); - if (!expire) - expire = 1; - - if (expire>HZ) - mod_timer(&tbl->gc_timer, round_jiffies(now + expire)); - else - mod_timer(&tbl->gc_timer, now + expire); - - write_unlock(&tbl->lock); + schedule_delayed_work(&tbl->gc_work, + tbl->parms.base_reachable_time >> 1); + write_unlock_bh(&tbl->lock); } static __inline__ int neigh_max_probes(struct neighbour *n) @@ -1442,10 +1441,8 @@ void neigh_table_init_no_netlink(struct neigh_table *tbl) get_random_bytes(&tbl->hash_rnd, sizeof(tbl->hash_rnd)); rwlock_init(&tbl->lock); - setup_timer(&tbl->gc_timer, neigh_periodic_timer, (unsigned long)tbl); - tbl->gc_timer.expires = now + 1; - add_timer(&tbl->gc_timer); - + INIT_DELAYED_WORK_DEFERRABLE(&tbl->gc_work, neigh_periodic_work); + schedule_delayed_work(&tbl->gc_work, tbl->parms.reachable_time); setup_timer(&tbl->proxy_timer, neigh_proxy_process, (unsigned long)tbl); skb_queue_head_init_class(&tbl->proxy_queue, &neigh_table_proxy_queue_class); @@ -1482,7 +1479,8 @@ int neigh_table_clear(struct neigh_table *tbl) struct neigh_table **tp; /* It is not clean... Fix it to unload IPv6 module safely */ - del_timer_sync(&tbl->gc_timer); + cancel_delayed_work(&tbl->gc_work); + flush_scheduled_work(); del_timer_sync(&tbl->proxy_timer); pneigh_queue_purge(&tbl->proxy_queue); neigh_ifdown(tbl, NULL); @@ -1752,7 +1750,6 @@ static int neightbl_fill_info(struct sk_buff *skb, struct neigh_table *tbl, .ndtc_last_rand = jiffies_to_msecs(rand_delta), .ndtc_hash_rnd = tbl->hash_rnd, .ndtc_hash_mask = tbl->hash_mask, - .ndtc_hash_chain_gc = tbl->hash_chain_gc, .ndtc_proxy_qlen = tbl->proxy_queue.qlen, }; -- cgit v1.2.3 From 47cab6a722d44c71c4f8224017ef548522243cf4 Mon Sep 17 00:00:00 2001 From: Ingo Molnar Date: Mon, 3 Aug 2009 09:31:54 +0200 Subject: debug lockups: Improve lockup detection, fix generic arch fallback As Andrew noted, my previous patch ("debug lockups: Improve lockup detection") broke/removed SysRq-L support from architecture that do not provide a __trigger_all_cpu_backtrace implementation. Restore a fallback path and clean up the SysRq-L machinery a bit: - Rename the arch method to arch_trigger_all_cpu_backtrace() - Simplify the define - Document the method a bit - in the hope of more architectures adding support for it. [ The patch touches Sparc code for the rename. ] Cc: Paul E. McKenney Cc: Peter Zijlstra Cc: Andrew Morton Cc: Linus Torvalds Cc: "David S. Miller" LKML-Reference: <20090802140809.7ec4bb6b.akpm@linux-foundation.org> Signed-off-by: Ingo Molnar --- arch/sparc/include/asm/irq_64.h | 4 ++-- arch/sparc/kernel/process_64.c | 4 ++-- arch/x86/include/asm/nmi.h | 4 ++-- arch/x86/kernel/apic/nmi.c | 2 +- drivers/char/sysrq.c | 15 ++++++++++++++- include/linux/nmi.h | 19 +++++++++++++++++-- 6 files changed, 38 insertions(+), 10 deletions(-) (limited to 'include') diff --git a/arch/sparc/include/asm/irq_64.h b/arch/sparc/include/asm/irq_64.h index 1934f2cbf513..a0b443cb3c1f 100644 --- a/arch/sparc/include/asm/irq_64.h +++ b/arch/sparc/include/asm/irq_64.h @@ -89,8 +89,8 @@ static inline unsigned long get_softint(void) return retval; } -void __trigger_all_cpu_backtrace(void); -#define trigger_all_cpu_backtrace() __trigger_all_cpu_backtrace() +void arch_trigger_all_cpu_backtrace(void); +#define arch_trigger_all_cpu_backtrace arch_trigger_all_cpu_backtrace extern void *hardirq_stack[NR_CPUS]; extern void *softirq_stack[NR_CPUS]; diff --git a/arch/sparc/kernel/process_64.c b/arch/sparc/kernel/process_64.c index 4041f94e7724..18d67854a1b8 100644 --- a/arch/sparc/kernel/process_64.c +++ b/arch/sparc/kernel/process_64.c @@ -251,7 +251,7 @@ static void __global_reg_poll(struct global_reg_snapshot *gp) } } -void __trigger_all_cpu_backtrace(void) +void arch_trigger_all_cpu_backtrace(void) { struct thread_info *tp = current_thread_info(); struct pt_regs *regs = get_irq_regs(); @@ -304,7 +304,7 @@ void __trigger_all_cpu_backtrace(void) static void sysrq_handle_globreg(int key, struct tty_struct *tty) { - __trigger_all_cpu_backtrace(); + arch_trigger_all_cpu_backtrace(); } static struct sysrq_key_op sparc_globalreg_op = { diff --git a/arch/x86/include/asm/nmi.h b/arch/x86/include/asm/nmi.h index c86e5ed4af51..e63cf7d441e1 100644 --- a/arch/x86/include/asm/nmi.h +++ b/arch/x86/include/asm/nmi.h @@ -45,8 +45,8 @@ extern int proc_nmi_enabled(struct ctl_table *, int , struct file *, void __user *, size_t *, loff_t *); extern int unknown_nmi_panic; -void __trigger_all_cpu_backtrace(void); -#define trigger_all_cpu_backtrace() __trigger_all_cpu_backtrace() +void arch_trigger_all_cpu_backtrace(void); +#define arch_trigger_all_cpu_backtrace arch_trigger_all_cpu_backtrace static inline void localise_nmi_watchdog(void) { diff --git a/arch/x86/kernel/apic/nmi.c b/arch/x86/kernel/apic/nmi.c index 1bb1ac20e9ec..db7220220d09 100644 --- a/arch/x86/kernel/apic/nmi.c +++ b/arch/x86/kernel/apic/nmi.c @@ -554,7 +554,7 @@ int do_nmi_callback(struct pt_regs *regs, int cpu) return 0; } -void __trigger_all_cpu_backtrace(void) +void arch_trigger_all_cpu_backtrace(void) { int i; diff --git a/drivers/char/sysrq.c b/drivers/char/sysrq.c index 165f307f30e8..50eecfe1d724 100644 --- a/drivers/char/sysrq.c +++ b/drivers/char/sysrq.c @@ -223,7 +223,20 @@ static DECLARE_WORK(sysrq_showallcpus, sysrq_showregs_othercpus); static void sysrq_handle_showallcpus(int key, struct tty_struct *tty) { - trigger_all_cpu_backtrace(); + /* + * Fall back to the workqueue based printing if the + * backtrace printing did not succeed or the + * architecture has no support for it: + */ + if (!trigger_all_cpu_backtrace()) { + struct pt_regs *regs = get_irq_regs(); + + if (regs) { + printk(KERN_INFO "CPU%d:\n", smp_processor_id()); + show_regs(regs); + } + schedule_work(&sysrq_showallcpus); + } } static struct sysrq_key_op sysrq_showallcpus_op = { diff --git a/include/linux/nmi.h b/include/linux/nmi.h index 29af2d5df097..b752e807adde 100644 --- a/include/linux/nmi.h +++ b/include/linux/nmi.h @@ -28,8 +28,23 @@ static inline void acpi_nmi_disable(void) { } static inline void acpi_nmi_enable(void) { } #endif -#ifndef trigger_all_cpu_backtrace -#define trigger_all_cpu_backtrace() do { } while (0) +/* + * Create trigger_all_cpu_backtrace() out of the arch-provided + * base function. Return whether such support was available, + * to allow calling code to fall back to some other mechanism: + */ +#ifdef arch_trigger_all_cpu_backtrace +static inline bool trigger_all_cpu_backtrace(void) +{ + arch_trigger_all_cpu_backtrace(); + + return true; +} +#else +static inline bool trigger_all_cpu_backtrace(void) +{ + return false; +} #endif #endif -- cgit v1.2.3 From ff663cf8705bea101d5f73cf471855c85242575e Mon Sep 17 00:00:00 2001 From: Zhenyu Wang Date: Thu, 23 Jul 2009 17:25:49 +0100 Subject: agp: Add generic support for graphics dma remapping New driver hooks for support graphics memory dma remapping are introduced in this patch. It makes generic code can tell if current device needs dma remapping, then call driver provided interfaces for mapping and unmapping. Change has also been made to handle scratch_page in remapping case. Signed-off-by: Zhenyu Wang Signed-off-by: David Woodhouse --- drivers/char/agp/agp.h | 6 ++++++ drivers/char/agp/backend.c | 20 ++++++++++++++++++++ drivers/char/agp/generic.c | 9 +++++++++ include/linux/agp_backend.h | 6 +++++- 4 files changed, 40 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/drivers/char/agp/agp.h b/drivers/char/agp/agp.h index ce110a3bf298..17e6d0d3ba36 100644 --- a/drivers/char/agp/agp.h +++ b/drivers/char/agp/agp.h @@ -121,6 +121,11 @@ struct agp_bridge_driver { void (*agp_destroy_pages)(struct agp_memory *); int (*agp_type_to_mask_type) (struct agp_bridge_data *, int); void (*chipset_flush)(struct agp_bridge_data *); + + int (*agp_map_page)(void *addr, dma_addr_t *ret); + void (*agp_unmap_page)(void *addr, dma_addr_t dma); + int (*agp_map_memory)(struct agp_memory *mem); + void (*agp_unmap_memory)(struct agp_memory *mem); }; struct agp_bridge_data { @@ -135,6 +140,7 @@ struct agp_bridge_data { u32 *gatt_table_real; unsigned long scratch_page; unsigned long scratch_page_real; + dma_addr_t scratch_page_dma; unsigned long gart_bus_addr; unsigned long gatt_bus_addr; u32 mode; diff --git a/drivers/char/agp/backend.c b/drivers/char/agp/backend.c index 3bd7e503de41..19ac3663acdc 100644 --- a/drivers/char/agp/backend.c +++ b/drivers/char/agp/backend.c @@ -152,6 +152,15 @@ static int agp_backend_initialize(struct agp_bridge_data *bridge) bridge->scratch_page_real = phys_to_gart(page_to_phys(page)); bridge->scratch_page = bridge->driver->mask_memory(bridge, phys_to_gart(page_to_phys(page)), 0); + + if (bridge->driver->agp_map_page && + bridge->driver->agp_map_page(phys_to_virt(page_to_phys(page)), + &bridge->scratch_page_dma)) { + dev_err(&bridge->dev->dev, + "unable to dma-map scratch page\n"); + rc = -ENOMEM; + goto err_out_nounmap; + } } size_value = bridge->driver->fetch_size(); @@ -191,6 +200,13 @@ static int agp_backend_initialize(struct agp_bridge_data *bridge) return 0; err_out: + if (bridge->driver->needs_scratch_page && + bridge->driver->agp_unmap_page) { + void *va = gart_to_virt(bridge->scratch_page_real); + + bridge->driver->agp_unmap_page(va, bridge->scratch_page_dma); + } +err_out_nounmap: if (bridge->driver->needs_scratch_page) { void *va = gart_to_virt(bridge->scratch_page_real); @@ -221,6 +237,10 @@ static void agp_backend_cleanup(struct agp_bridge_data *bridge) bridge->driver->needs_scratch_page) { void *va = gart_to_virt(bridge->scratch_page_real); + if (bridge->driver->agp_unmap_page) + bridge->driver->agp_unmap_page(va, + bridge->scratch_page_dma); + bridge->driver->agp_destroy_page(va, AGP_PAGE_DESTROY_UNMAP); bridge->driver->agp_destroy_page(va, AGP_PAGE_DESTROY_FREE); } diff --git a/drivers/char/agp/generic.c b/drivers/char/agp/generic.c index a3bcc7ef42f9..28f0208c66a6 100644 --- a/drivers/char/agp/generic.c +++ b/drivers/char/agp/generic.c @@ -437,6 +437,12 @@ int agp_bind_memory(struct agp_memory *curr, off_t pg_start) curr->bridge->driver->cache_flush(); curr->is_flushed = true; } + + if (curr->bridge->driver->agp_map_memory) { + ret_val = curr->bridge->driver->agp_map_memory(curr); + if (ret_val) + return ret_val; + } ret_val = curr->bridge->driver->insert_memory(curr, pg_start, curr->type); if (ret_val != 0) @@ -478,6 +484,9 @@ int agp_unbind_memory(struct agp_memory *curr) if (ret_val != 0) return ret_val; + if (curr->bridge->driver->agp_unmap_memory) + curr->bridge->driver->agp_unmap_memory(curr); + curr->is_bound = false; curr->pg_start = 0; spin_lock(&curr->bridge->mapped_lock); diff --git a/include/linux/agp_backend.h b/include/linux/agp_backend.h index 76fa794fdac0..8a294d65b9b1 100644 --- a/include/linux/agp_backend.h +++ b/include/linux/agp_backend.h @@ -79,9 +79,13 @@ struct agp_memory { u32 physical; bool is_bound; bool is_flushed; - bool vmalloc_flag; + bool vmalloc_flag; + bool sg_vmalloc_flag; /* list of agp_memory mapped to the aperture */ struct list_head mapped_list; + /* DMA-mapped addresses */ + struct scatterlist *sg_list; + int num_sg; }; #define AGP_NORMAL_MEMORY 0 -- cgit v1.2.3 From f692775d7e0a22477143cd884e45c955448ac7d2 Mon Sep 17 00:00:00 2001 From: David Woodhouse Date: Wed, 29 Jul 2009 09:28:45 +0100 Subject: intel-agp: fix sglist allocation to avoid vmalloc() Signed-off-by: David Woodhouse --- drivers/char/agp/intel-agp.c | 29 ++++++++++------------------- include/linux/agp_backend.h | 1 - 2 files changed, 10 insertions(+), 20 deletions(-) (limited to 'include') diff --git a/drivers/char/agp/intel-agp.c b/drivers/char/agp/intel-agp.c index b9d9886ff3c3..d8c80d8be5e2 100644 --- a/drivers/char/agp/intel-agp.c +++ b/drivers/char/agp/intel-agp.c @@ -198,39 +198,30 @@ static void intel_agp_unmap_page(struct page *page, dma_addr_t dma) static void intel_agp_free_sglist(struct agp_memory *mem) { + struct sg_table st; + + st.sgl = mem->sg_list; + st.orig_nents = st.nents = mem->page_count; + + sg_free_table(&st); - if (mem->sg_vmalloc_flag) - vfree(mem->sg_list); - else - kfree(mem->sg_list); - mem->sg_vmalloc_flag = 0; mem->sg_list = NULL; mem->num_sg = 0; } static int intel_agp_map_memory(struct agp_memory *mem) { + struct sg_table st; struct scatterlist *sg; int i; DBG("try mapping %lu pages\n", (unsigned long)mem->page_count); - if ((mem->page_count * sizeof(*mem->sg_list)) < 2*PAGE_SIZE) - mem->sg_list = kcalloc(mem->page_count, sizeof(*mem->sg_list), - GFP_KERNEL); - - if (mem->sg_list == NULL) { - mem->sg_list = vmalloc(mem->page_count * sizeof(*mem->sg_list)); - mem->sg_vmalloc_flag = 1; - } - - if (!mem->sg_list) { - mem->sg_vmalloc_flag = 0; + if (sg_alloc_table(&st, mem->page_count, GFP_KERNEL)) return -ENOMEM; - } - sg_init_table(mem->sg_list, mem->page_count); - sg = mem->sg_list; + mem->sg_list = sg = st.sgl; + for (i = 0 ; i < mem->page_count; i++, sg = sg_next(sg)) sg_set_page(sg, mem->pages[i], PAGE_SIZE, 0); diff --git a/include/linux/agp_backend.h b/include/linux/agp_backend.h index 8a294d65b9b1..880130f7311f 100644 --- a/include/linux/agp_backend.h +++ b/include/linux/agp_backend.h @@ -80,7 +80,6 @@ struct agp_memory { bool is_bound; bool is_flushed; bool vmalloc_flag; - bool sg_vmalloc_flag; /* list of agp_memory mapped to the aperture */ struct list_head mapped_list; /* DMA-mapped addresses */ -- cgit v1.2.3 From 42c4ab41a176ee784c0f28c0b29025a8fc34f05a Mon Sep 17 00:00:00 2001 From: Stanislaw Gruszka Date: Wed, 29 Jul 2009 12:15:26 +0200 Subject: itimers: Merge ITIMER_VIRT and ITIMER_PROF Both cpu itimers have same data flow in the few places, this patch make unification of code related with VIRT and PROF itimers. Signed-off-by: Stanislaw Gruszka Acked-by: Peter Zijlstra Acked-by: Thomas Gleixner Cc: Oleg Nesterov Cc: Andrew Morton Cc: Paul Mackerras Cc: Benjamin Herrenschmidt LKML-Reference: <1248862529-6063-2-git-send-email-sgruszka@redhat.com> Signed-off-by: Ingo Molnar --- include/linux/sched.h | 14 ++++- kernel/fork.c | 9 +-- kernel/itimer.c | 146 +++++++++++++++++++++------------------------- kernel/posix-cpu-timers.c | 98 +++++++++++++++---------------- 4 files changed, 130 insertions(+), 137 deletions(-) (limited to 'include') diff --git a/include/linux/sched.h b/include/linux/sched.h index 3ab08e4bb6b8..3b3efaddd953 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -470,6 +470,11 @@ struct pacct_struct { unsigned long ac_minflt, ac_majflt; }; +struct cpu_itimer { + cputime_t expires; + cputime_t incr; +}; + /** * struct task_cputime - collected CPU time counts * @utime: time spent in user mode, in &cputime_t units @@ -564,9 +569,12 @@ struct signal_struct { struct pid *leader_pid; ktime_t it_real_incr; - /* ITIMER_PROF and ITIMER_VIRTUAL timers for the process */ - cputime_t it_prof_expires, it_virt_expires; - cputime_t it_prof_incr, it_virt_incr; + /* + * ITIMER_PROF and ITIMER_VIRTUAL timers for the process, we use + * CPUCLOCK_PROF and CPUCLOCK_VIRT for indexing array as these + * values are defined to 0 and 1 respectively + */ + struct cpu_itimer it[2]; /* * Thread group totals for process CPU timers. diff --git a/kernel/fork.c b/kernel/fork.c index 29b532e718f7..893ab0bf5e39 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -62,6 +62,7 @@ #include #include #include +#include #include #include @@ -790,10 +791,10 @@ static void posix_cpu_timers_init_group(struct signal_struct *sig) thread_group_cputime_init(sig); /* Expiration times and increments. */ - sig->it_virt_expires = cputime_zero; - sig->it_virt_incr = cputime_zero; - sig->it_prof_expires = cputime_zero; - sig->it_prof_incr = cputime_zero; + sig->it[CPUCLOCK_PROF].expires = cputime_zero; + sig->it[CPUCLOCK_PROF].incr = cputime_zero; + sig->it[CPUCLOCK_VIRT].expires = cputime_zero; + sig->it[CPUCLOCK_VIRT].incr = cputime_zero; /* Cached expiration times. */ sig->cputime_expires.prof_exp = cputime_zero; diff --git a/kernel/itimer.c b/kernel/itimer.c index 58762f7077ec..852c88ddd1f0 100644 --- a/kernel/itimer.c +++ b/kernel/itimer.c @@ -41,10 +41,43 @@ static struct timeval itimer_get_remtime(struct hrtimer *timer) return ktime_to_timeval(rem); } +static void get_cpu_itimer(struct task_struct *tsk, unsigned int clock_id, + struct itimerval *value) +{ + cputime_t cval, cinterval; + struct cpu_itimer *it = &tsk->signal->it[clock_id]; + + spin_lock_irq(&tsk->sighand->siglock); + + cval = it->expires; + cinterval = it->incr; + if (!cputime_eq(cval, cputime_zero)) { + struct task_cputime cputime; + cputime_t t; + + thread_group_cputimer(tsk, &cputime); + if (clock_id == CPUCLOCK_PROF) + t = cputime_add(cputime.utime, cputime.stime); + else + /* CPUCLOCK_VIRT */ + t = cputime.utime; + + if (cputime_le(cval, t)) + /* about to fire */ + cval = jiffies_to_cputime(1); + else + cval = cputime_sub(cval, t); + } + + spin_unlock_irq(&tsk->sighand->siglock); + + cputime_to_timeval(cval, &value->it_value); + cputime_to_timeval(cinterval, &value->it_interval); +} + int do_getitimer(int which, struct itimerval *value) { struct task_struct *tsk = current; - cputime_t cinterval, cval; switch (which) { case ITIMER_REAL: @@ -55,44 +88,10 @@ int do_getitimer(int which, struct itimerval *value) spin_unlock_irq(&tsk->sighand->siglock); break; case ITIMER_VIRTUAL: - spin_lock_irq(&tsk->sighand->siglock); - cval = tsk->signal->it_virt_expires; - cinterval = tsk->signal->it_virt_incr; - if (!cputime_eq(cval, cputime_zero)) { - struct task_cputime cputime; - cputime_t utime; - - thread_group_cputimer(tsk, &cputime); - utime = cputime.utime; - if (cputime_le(cval, utime)) { /* about to fire */ - cval = jiffies_to_cputime(1); - } else { - cval = cputime_sub(cval, utime); - } - } - spin_unlock_irq(&tsk->sighand->siglock); - cputime_to_timeval(cval, &value->it_value); - cputime_to_timeval(cinterval, &value->it_interval); + get_cpu_itimer(tsk, CPUCLOCK_VIRT, value); break; case ITIMER_PROF: - spin_lock_irq(&tsk->sighand->siglock); - cval = tsk->signal->it_prof_expires; - cinterval = tsk->signal->it_prof_incr; - if (!cputime_eq(cval, cputime_zero)) { - struct task_cputime times; - cputime_t ptime; - - thread_group_cputimer(tsk, ×); - ptime = cputime_add(times.utime, times.stime); - if (cputime_le(cval, ptime)) { /* about to fire */ - cval = jiffies_to_cputime(1); - } else { - cval = cputime_sub(cval, ptime); - } - } - spin_unlock_irq(&tsk->sighand->siglock); - cputime_to_timeval(cval, &value->it_value); - cputime_to_timeval(cinterval, &value->it_interval); + get_cpu_itimer(tsk, CPUCLOCK_PROF, value); break; default: return(-EINVAL); @@ -128,6 +127,36 @@ enum hrtimer_restart it_real_fn(struct hrtimer *timer) return HRTIMER_NORESTART; } +static void set_cpu_itimer(struct task_struct *tsk, unsigned int clock_id, + struct itimerval *value, struct itimerval *ovalue) +{ + cputime_t cval, cinterval, nval, ninterval; + struct cpu_itimer *it = &tsk->signal->it[clock_id]; + + nval = timeval_to_cputime(&value->it_value); + ninterval = timeval_to_cputime(&value->it_interval); + + spin_lock_irq(&tsk->sighand->siglock); + + cval = it->expires; + cinterval = it->incr; + if (!cputime_eq(cval, cputime_zero) || + !cputime_eq(nval, cputime_zero)) { + if (cputime_gt(nval, cputime_zero)) + nval = cputime_add(nval, jiffies_to_cputime(1)); + set_process_cpu_timer(tsk, clock_id, &nval, &cval); + } + it->expires = nval; + it->incr = ninterval; + + spin_unlock_irq(&tsk->sighand->siglock); + + if (ovalue) { + cputime_to_timeval(cval, &ovalue->it_value); + cputime_to_timeval(cinterval, &ovalue->it_interval); + } +} + /* * Returns true if the timeval is in canonical form */ @@ -139,7 +168,6 @@ int do_setitimer(int which, struct itimerval *value, struct itimerval *ovalue) struct task_struct *tsk = current; struct hrtimer *timer; ktime_t expires; - cputime_t cval, cinterval, nval, ninterval; /* * Validate the timevals in value. @@ -174,48 +202,10 @@ again: spin_unlock_irq(&tsk->sighand->siglock); break; case ITIMER_VIRTUAL: - nval = timeval_to_cputime(&value->it_value); - ninterval = timeval_to_cputime(&value->it_interval); - spin_lock_irq(&tsk->sighand->siglock); - cval = tsk->signal->it_virt_expires; - cinterval = tsk->signal->it_virt_incr; - if (!cputime_eq(cval, cputime_zero) || - !cputime_eq(nval, cputime_zero)) { - if (cputime_gt(nval, cputime_zero)) - nval = cputime_add(nval, - jiffies_to_cputime(1)); - set_process_cpu_timer(tsk, CPUCLOCK_VIRT, - &nval, &cval); - } - tsk->signal->it_virt_expires = nval; - tsk->signal->it_virt_incr = ninterval; - spin_unlock_irq(&tsk->sighand->siglock); - if (ovalue) { - cputime_to_timeval(cval, &ovalue->it_value); - cputime_to_timeval(cinterval, &ovalue->it_interval); - } + set_cpu_itimer(tsk, CPUCLOCK_VIRT, value, ovalue); break; case ITIMER_PROF: - nval = timeval_to_cputime(&value->it_value); - ninterval = timeval_to_cputime(&value->it_interval); - spin_lock_irq(&tsk->sighand->siglock); - cval = tsk->signal->it_prof_expires; - cinterval = tsk->signal->it_prof_incr; - if (!cputime_eq(cval, cputime_zero) || - !cputime_eq(nval, cputime_zero)) { - if (cputime_gt(nval, cputime_zero)) - nval = cputime_add(nval, - jiffies_to_cputime(1)); - set_process_cpu_timer(tsk, CPUCLOCK_PROF, - &nval, &cval); - } - tsk->signal->it_prof_expires = nval; - tsk->signal->it_prof_incr = ninterval; - spin_unlock_irq(&tsk->sighand->siglock); - if (ovalue) { - cputime_to_timeval(cval, &ovalue->it_value); - cputime_to_timeval(cinterval, &ovalue->it_interval); - } + set_cpu_itimer(tsk, CPUCLOCK_PROF, value, ovalue); break; default: return -EINVAL; diff --git a/kernel/posix-cpu-timers.c b/kernel/posix-cpu-timers.c index bece7c0b67b2..9b2d5e4dc8c4 100644 --- a/kernel/posix-cpu-timers.c +++ b/kernel/posix-cpu-timers.c @@ -14,11 +14,11 @@ */ void update_rlimit_cpu(unsigned long rlim_new) { - cputime_t cputime; + cputime_t cputime = secs_to_cputime(rlim_new); + struct signal_struct *const sig = current->signal; - cputime = secs_to_cputime(rlim_new); - if (cputime_eq(current->signal->it_prof_expires, cputime_zero) || - cputime_gt(current->signal->it_prof_expires, cputime)) { + if (cputime_eq(sig->it[CPUCLOCK_PROF].expires, cputime_zero) || + cputime_gt(sig->it[CPUCLOCK_PROF].expires, cputime)) { spin_lock_irq(¤t->sighand->siglock); set_process_cpu_timer(current, CPUCLOCK_PROF, &cputime, NULL); spin_unlock_irq(¤t->sighand->siglock); @@ -613,6 +613,9 @@ static void arm_timer(struct k_itimer *timer, union cpu_time_count now) break; } } else { + struct signal_struct *const sig = p->signal; + union cpu_time_count *exp = &timer->it.cpu.expires; + /* * For a process timer, set the cached expiration time. */ @@ -620,30 +623,27 @@ static void arm_timer(struct k_itimer *timer, union cpu_time_count now) default: BUG(); case CPUCLOCK_VIRT: - if (!cputime_eq(p->signal->it_virt_expires, + if (!cputime_eq(sig->it[CPUCLOCK_VIRT].expires, cputime_zero) && - cputime_lt(p->signal->it_virt_expires, - timer->it.cpu.expires.cpu)) + cputime_lt(sig->it[CPUCLOCK_VIRT].expires, + exp->cpu)) break; - p->signal->cputime_expires.virt_exp = - timer->it.cpu.expires.cpu; + sig->cputime_expires.virt_exp = exp->cpu; break; case CPUCLOCK_PROF: - if (!cputime_eq(p->signal->it_prof_expires, + if (!cputime_eq(sig->it[CPUCLOCK_PROF].expires, cputime_zero) && - cputime_lt(p->signal->it_prof_expires, - timer->it.cpu.expires.cpu)) + cputime_lt(sig->it[CPUCLOCK_PROF].expires, + exp->cpu)) break; - i = p->signal->rlim[RLIMIT_CPU].rlim_cur; + i = sig->rlim[RLIMIT_CPU].rlim_cur; if (i != RLIM_INFINITY && - i <= cputime_to_secs(timer->it.cpu.expires.cpu)) + i <= cputime_to_secs(exp->cpu)) break; - p->signal->cputime_expires.prof_exp = - timer->it.cpu.expires.cpu; + sig->cputime_expires.prof_exp = exp->cpu; break; case CPUCLOCK_SCHED: - p->signal->cputime_expires.sched_exp = - timer->it.cpu.expires.sched; + sig->cputime_expires.sched_exp = exp->sched; break; } } @@ -1070,6 +1070,27 @@ static void stop_process_timers(struct task_struct *tsk) spin_unlock_irqrestore(&cputimer->lock, flags); } +static void check_cpu_itimer(struct task_struct *tsk, struct cpu_itimer *it, + cputime_t *expires, cputime_t cur_time, int signo) +{ + if (cputime_eq(it->expires, cputime_zero)) + return; + + if (cputime_ge(cur_time, it->expires)) { + it->expires = it->incr; + if (!cputime_eq(it->expires, cputime_zero)) + it->expires = cputime_add(it->expires, cur_time); + + __group_send_sig_info(signo, SEND_SIG_PRIV, tsk); + } + + if (!cputime_eq(it->expires, cputime_zero) && + (cputime_eq(*expires, cputime_zero) || + cputime_lt(it->expires, *expires))) { + *expires = it->expires; + } +} + /* * Check for any per-thread CPU timers that have fired and move them * off the tsk->*_timers list onto the firing list. Per-thread timers @@ -1089,10 +1110,10 @@ static void check_process_timers(struct task_struct *tsk, * Don't sample the current process CPU clocks if there are no timers. */ if (list_empty(&timers[CPUCLOCK_PROF]) && - cputime_eq(sig->it_prof_expires, cputime_zero) && + cputime_eq(sig->it[CPUCLOCK_PROF].expires, cputime_zero) && sig->rlim[RLIMIT_CPU].rlim_cur == RLIM_INFINITY && list_empty(&timers[CPUCLOCK_VIRT]) && - cputime_eq(sig->it_virt_expires, cputime_zero) && + cputime_eq(sig->it[CPUCLOCK_VIRT].expires, cputime_zero) && list_empty(&timers[CPUCLOCK_SCHED])) { stop_process_timers(tsk); return; @@ -1152,38 +1173,11 @@ static void check_process_timers(struct task_struct *tsk, /* * Check for the special case process timers. */ - if (!cputime_eq(sig->it_prof_expires, cputime_zero)) { - if (cputime_ge(ptime, sig->it_prof_expires)) { - /* ITIMER_PROF fires and reloads. */ - sig->it_prof_expires = sig->it_prof_incr; - if (!cputime_eq(sig->it_prof_expires, cputime_zero)) { - sig->it_prof_expires = cputime_add( - sig->it_prof_expires, ptime); - } - __group_send_sig_info(SIGPROF, SEND_SIG_PRIV, tsk); - } - if (!cputime_eq(sig->it_prof_expires, cputime_zero) && - (cputime_eq(prof_expires, cputime_zero) || - cputime_lt(sig->it_prof_expires, prof_expires))) { - prof_expires = sig->it_prof_expires; - } - } - if (!cputime_eq(sig->it_virt_expires, cputime_zero)) { - if (cputime_ge(utime, sig->it_virt_expires)) { - /* ITIMER_VIRTUAL fires and reloads. */ - sig->it_virt_expires = sig->it_virt_incr; - if (!cputime_eq(sig->it_virt_expires, cputime_zero)) { - sig->it_virt_expires = cputime_add( - sig->it_virt_expires, utime); - } - __group_send_sig_info(SIGVTALRM, SEND_SIG_PRIV, tsk); - } - if (!cputime_eq(sig->it_virt_expires, cputime_zero) && - (cputime_eq(virt_expires, cputime_zero) || - cputime_lt(sig->it_virt_expires, virt_expires))) { - virt_expires = sig->it_virt_expires; - } - } + check_cpu_itimer(tsk, &sig->it[CPUCLOCK_PROF], &prof_expires, ptime, + SIGPROF); + check_cpu_itimer(tsk, &sig->it[CPUCLOCK_VIRT], &virt_expires, utime, + SIGVTALRM); + if (sig->rlim[RLIMIT_CPU].rlim_cur != RLIM_INFINITY) { unsigned long psecs = cputime_to_secs(ptime); cputime_t x; -- cgit v1.2.3 From 8356b5f9c424e5831715abbce747197c30d1fd71 Mon Sep 17 00:00:00 2001 From: Stanislaw Gruszka Date: Wed, 29 Jul 2009 12:15:27 +0200 Subject: itimers: Fix periodic tics precision Measure ITIMER_PROF and ITIMER_VIRT timers interval error between real ticks and requested by user. Take it into account when scheduling next tick. This patch introduce possibility where time between two consecutive tics is smaller then requested interval, it preserve however dependency that n tick is generated not earlier than n*interval time - counting from the beginning of periodic signal generation. Signed-off-by: Stanislaw Gruszka Acked-by: Peter Zijlstra Acked-by: Thomas Gleixner Cc: Oleg Nesterov Cc: Andrew Morton Cc: Paul Mackerras Cc: Benjamin Herrenschmidt LKML-Reference: <1248862529-6063-3-git-send-email-sgruszka@redhat.com> Signed-off-by: Ingo Molnar --- include/linux/sched.h | 2 ++ kernel/itimer.c | 24 +++++++++++++++++++++--- kernel/posix-cpu-timers.c | 20 +++++++++++++++++--- 3 files changed, 40 insertions(+), 6 deletions(-) (limited to 'include') diff --git a/include/linux/sched.h b/include/linux/sched.h index 3b3efaddd953..a069e65e8bb7 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -473,6 +473,8 @@ struct pacct_struct { struct cpu_itimer { cputime_t expires; cputime_t incr; + u32 error; + u32 incr_error; }; /** diff --git a/kernel/itimer.c b/kernel/itimer.c index 852c88ddd1f0..21adff7b2a17 100644 --- a/kernel/itimer.c +++ b/kernel/itimer.c @@ -42,7 +42,7 @@ static struct timeval itimer_get_remtime(struct hrtimer *timer) } static void get_cpu_itimer(struct task_struct *tsk, unsigned int clock_id, - struct itimerval *value) + struct itimerval *const value) { cputime_t cval, cinterval; struct cpu_itimer *it = &tsk->signal->it[clock_id]; @@ -127,14 +127,32 @@ enum hrtimer_restart it_real_fn(struct hrtimer *timer) return HRTIMER_NORESTART; } +static inline u32 cputime_sub_ns(cputime_t ct, s64 real_ns) +{ + struct timespec ts; + s64 cpu_ns; + + cputime_to_timespec(ct, &ts); + cpu_ns = timespec_to_ns(&ts); + + return (cpu_ns <= real_ns) ? 0 : cpu_ns - real_ns; +} + static void set_cpu_itimer(struct task_struct *tsk, unsigned int clock_id, - struct itimerval *value, struct itimerval *ovalue) + const struct itimerval *const value, + struct itimerval *const ovalue) { - cputime_t cval, cinterval, nval, ninterval; + cputime_t cval, nval, cinterval, ninterval; + s64 ns_ninterval, ns_nval; struct cpu_itimer *it = &tsk->signal->it[clock_id]; nval = timeval_to_cputime(&value->it_value); + ns_nval = timeval_to_ns(&value->it_value); ninterval = timeval_to_cputime(&value->it_interval); + ns_ninterval = timeval_to_ns(&value->it_interval); + + it->incr_error = cputime_sub_ns(ninterval, ns_ninterval); + it->error = cputime_sub_ns(nval, ns_nval); spin_lock_irq(&tsk->sighand->siglock); diff --git a/kernel/posix-cpu-timers.c b/kernel/posix-cpu-timers.c index 9b2d5e4dc8c4..b60d644ea4b3 100644 --- a/kernel/posix-cpu-timers.c +++ b/kernel/posix-cpu-timers.c @@ -1070,6 +1070,8 @@ static void stop_process_timers(struct task_struct *tsk) spin_unlock_irqrestore(&cputimer->lock, flags); } +static u32 onecputick; + static void check_cpu_itimer(struct task_struct *tsk, struct cpu_itimer *it, cputime_t *expires, cputime_t cur_time, int signo) { @@ -1077,9 +1079,16 @@ static void check_cpu_itimer(struct task_struct *tsk, struct cpu_itimer *it, return; if (cputime_ge(cur_time, it->expires)) { - it->expires = it->incr; - if (!cputime_eq(it->expires, cputime_zero)) - it->expires = cputime_add(it->expires, cur_time); + if (!cputime_eq(it->incr, cputime_zero)) { + it->expires = cputime_add(it->expires, it->incr); + it->error += it->incr_error; + if (it->error >= onecputick) { + it->expires = cputime_sub(it->expires, + jiffies_to_cputime(1)); + it->error -= onecputick; + } + } else + it->expires = cputime_zero; __group_send_sig_info(signo, SEND_SIG_PRIV, tsk); } @@ -1696,10 +1705,15 @@ static __init int init_posix_cpu_timers(void) .nsleep = thread_cpu_nsleep, .nsleep_restart = thread_cpu_nsleep_restart, }; + struct timespec ts; register_posix_clock(CLOCK_PROCESS_CPUTIME_ID, &process); register_posix_clock(CLOCK_THREAD_CPUTIME_ID, &thread); + cputime_to_timespec(jiffies_to_cputime(1), &ts); + onecputick = ts.tv_nsec; + WARN_ON(ts.tv_sec != 0); + return 0; } __initcall(init_posix_cpu_timers); -- cgit v1.2.3 From a42548a18866e87092db93b771e6c5b060d78401 Mon Sep 17 00:00:00 2001 From: Stanislaw Gruszka Date: Wed, 29 Jul 2009 12:15:29 +0200 Subject: cputime: Optimize jiffies_to_cputime(1) For powerpc with CONFIG_VIRT_CPU_ACCOUNTING jiffies_to_cputime(1) is not compile time constant and run time calculations are quite expensive. To optimize we use precomputed value. For all other architectures is is preprocessor definition. Signed-off-by: Stanislaw Gruszka Acked-by: Peter Zijlstra Acked-by: Thomas Gleixner Cc: Oleg Nesterov Cc: Andrew Morton Cc: Paul Mackerras Cc: Benjamin Herrenschmidt LKML-Reference: <1248862529-6063-5-git-send-email-sgruszka@redhat.com> Signed-off-by: Ingo Molnar --- arch/ia64/include/asm/cputime.h | 1 + arch/powerpc/include/asm/cputime.h | 13 +++++++++++++ arch/powerpc/kernel/time.c | 4 ++++ arch/s390/include/asm/cputime.h | 1 + include/asm-generic/cputime.h | 1 + kernel/itimer.c | 4 ++-- kernel/posix-cpu-timers.c | 6 +++--- kernel/sched.c | 9 ++++----- 8 files changed, 29 insertions(+), 10 deletions(-) (limited to 'include') diff --git a/arch/ia64/include/asm/cputime.h b/arch/ia64/include/asm/cputime.h index d20b998cb91d..7fa8a8594660 100644 --- a/arch/ia64/include/asm/cputime.h +++ b/arch/ia64/include/asm/cputime.h @@ -30,6 +30,7 @@ typedef u64 cputime_t; typedef u64 cputime64_t; #define cputime_zero ((cputime_t)0) +#define cputime_one_jiffy jiffies_to_cputime(1) #define cputime_max ((~((cputime_t)0) >> 1) - 1) #define cputime_add(__a, __b) ((__a) + (__b)) #define cputime_sub(__a, __b) ((__a) - (__b)) diff --git a/arch/powerpc/include/asm/cputime.h b/arch/powerpc/include/asm/cputime.h index f42e623030ee..fa19f3fe05ff 100644 --- a/arch/powerpc/include/asm/cputime.h +++ b/arch/powerpc/include/asm/cputime.h @@ -18,6 +18,9 @@ #ifndef CONFIG_VIRT_CPU_ACCOUNTING #include +#ifdef __KERNEL__ +static inline void setup_cputime_one_jiffy(void) { } +#endif #else #include @@ -48,6 +51,11 @@ typedef u64 cputime64_t; #ifdef __KERNEL__ +/* + * One jiffy in timebase units computed during initialization + */ +extern cputime_t cputime_one_jiffy; + /* * Convert cputime <-> jiffies */ @@ -89,6 +97,11 @@ static inline cputime_t jiffies_to_cputime(const unsigned long jif) return ct; } +static inline void setup_cputime_one_jiffy(void) +{ + cputime_one_jiffy = jiffies_to_cputime(1); +} + static inline cputime64_t jiffies64_to_cputime64(const u64 jif) { cputime_t ct; diff --git a/arch/powerpc/kernel/time.c b/arch/powerpc/kernel/time.c index eae4511ceeac..211d7b0cd370 100644 --- a/arch/powerpc/kernel/time.c +++ b/arch/powerpc/kernel/time.c @@ -193,6 +193,8 @@ EXPORT_SYMBOL(__cputime_clockt_factor); DEFINE_PER_CPU(unsigned long, cputime_last_delta); DEFINE_PER_CPU(unsigned long, cputime_scaled_last_delta); +cputime_t cputime_one_jiffy; + static void calc_cputime_factors(void) { struct div_result res; @@ -500,6 +502,7 @@ static int __init iSeries_tb_recal(void) tb_to_xs = divres.result_low; vdso_data->tb_ticks_per_sec = tb_ticks_per_sec; vdso_data->tb_to_xs = tb_to_xs; + setup_cputime_one_jiffy(); } else { printk( "Titan recalibrate: FAILED (difference > 4 percent)\n" @@ -945,6 +948,7 @@ void __init time_init(void) tb_ticks_per_usec = ppc_tb_freq / 1000000; tb_to_us = mulhwu_scale_factor(ppc_tb_freq, 1000000); calc_cputime_factors(); + setup_cputime_one_jiffy(); /* * Calculate the length of each tick in ns. It will not be diff --git a/arch/s390/include/asm/cputime.h b/arch/s390/include/asm/cputime.h index 7a3817a656df..24b1244aadb9 100644 --- a/arch/s390/include/asm/cputime.h +++ b/arch/s390/include/asm/cputime.h @@ -42,6 +42,7 @@ __div(unsigned long long n, unsigned int base) #endif /* __s390x__ */ #define cputime_zero (0ULL) +#define cputime_one_jiffy jiffies_to_cputime(1) #define cputime_max ((~0UL >> 1) - 1) #define cputime_add(__a, __b) ((__a) + (__b)) #define cputime_sub(__a, __b) ((__a) - (__b)) diff --git a/include/asm-generic/cputime.h b/include/asm-generic/cputime.h index 1c1fa422d18a..ca0f239f0e13 100644 --- a/include/asm-generic/cputime.h +++ b/include/asm-generic/cputime.h @@ -7,6 +7,7 @@ typedef unsigned long cputime_t; #define cputime_zero (0UL) +#define cputime_one_jiffy jiffies_to_cputime(1) #define cputime_max ((~0UL >> 1) - 1) #define cputime_add(__a, __b) ((__a) + (__b)) #define cputime_sub(__a, __b) ((__a) - (__b)) diff --git a/kernel/itimer.c b/kernel/itimer.c index 21adff7b2a17..8078a32d3b10 100644 --- a/kernel/itimer.c +++ b/kernel/itimer.c @@ -64,7 +64,7 @@ static void get_cpu_itimer(struct task_struct *tsk, unsigned int clock_id, if (cputime_le(cval, t)) /* about to fire */ - cval = jiffies_to_cputime(1); + cval = cputime_one_jiffy; else cval = cputime_sub(cval, t); } @@ -161,7 +161,7 @@ static void set_cpu_itimer(struct task_struct *tsk, unsigned int clock_id, if (!cputime_eq(cval, cputime_zero) || !cputime_eq(nval, cputime_zero)) { if (cputime_gt(nval, cputime_zero)) - nval = cputime_add(nval, jiffies_to_cputime(1)); + nval = cputime_add(nval, cputime_one_jiffy); set_process_cpu_timer(tsk, clock_id, &nval, &cval); } it->expires = nval; diff --git a/kernel/posix-cpu-timers.c b/kernel/posix-cpu-timers.c index 69c92374355f..18bdde6f676f 100644 --- a/kernel/posix-cpu-timers.c +++ b/kernel/posix-cpu-timers.c @@ -1086,7 +1086,7 @@ static void check_cpu_itimer(struct task_struct *tsk, struct cpu_itimer *it, it->error += it->incr_error; if (it->error >= onecputick) { it->expires = cputime_sub(it->expires, - jiffies_to_cputime(1)); + cputime_one_jiffy); it->error -= onecputick; } } else @@ -1461,7 +1461,7 @@ void set_process_cpu_timer(struct task_struct *tsk, unsigned int clock_idx, if (!cputime_eq(*oldval, cputime_zero)) { if (cputime_le(*oldval, now.cpu)) { /* Just about to fire. */ - *oldval = jiffies_to_cputime(1); + *oldval = cputime_one_jiffy; } else { *oldval = cputime_sub(*oldval, now.cpu); } @@ -1712,7 +1712,7 @@ static __init int init_posix_cpu_timers(void) register_posix_clock(CLOCK_PROCESS_CPUTIME_ID, &process); register_posix_clock(CLOCK_THREAD_CPUTIME_ID, &thread); - cputime_to_timespec(jiffies_to_cputime(1), &ts); + cputime_to_timespec(cputime_one_jiffy, &ts); onecputick = ts.tv_nsec; WARN_ON(ts.tv_sec != 0); diff --git a/kernel/sched.c b/kernel/sched.c index 1b59e265273b..8f977d5cc515 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -5031,17 +5031,16 @@ void account_idle_time(cputime_t cputime) */ void account_process_tick(struct task_struct *p, int user_tick) { - cputime_t one_jiffy = jiffies_to_cputime(1); - cputime_t one_jiffy_scaled = cputime_to_scaled(one_jiffy); + cputime_t one_jiffy_scaled = cputime_to_scaled(cputime_one_jiffy); struct rq *rq = this_rq(); if (user_tick) - account_user_time(p, one_jiffy, one_jiffy_scaled); + account_user_time(p, cputime_one_jiffy, one_jiffy_scaled); else if ((p != rq->idle) || (irq_count() != HARDIRQ_OFFSET)) - account_system_time(p, HARDIRQ_OFFSET, one_jiffy, + account_system_time(p, HARDIRQ_OFFSET, cputime_one_jiffy, one_jiffy_scaled); else - account_idle_time(one_jiffy); + account_idle_time(cputime_one_jiffy); } /* -- cgit v1.2.3 From 7699ad35ed06044c4fc1be162553880f98658616 Mon Sep 17 00:00:00 2001 From: Nicolas Pitre Date: Mon, 15 Jun 2009 01:10:18 -0400 Subject: mtd: let include/linux/mtd/partitions.h stand on its own When declaring static MTD partitions in board specific code, only including should suffice without gcc nagging us with: In file included from arch/arm/mach-kirkwood/sheevaplug-setup.c:14: include/linux/mtd/partitions.h:50: warning: 'struct mtd_info' declared inside parameter list include/linux/mtd/partitions.h:50: warning: its scope is only this definition or declaration, which is probably not what you want include/linux/mtd/partitions.h:51: warning: 'struct mtd_info' declared inside parameter list include/linux/mtd/partitions.h:61: warning: 'struct mtd_info' declared inside parameter list include/linux/mtd/partitions.h:67: warning: 'struct mtd_info' declared inside parameter list Signed-off-by: Nicolas Pitre Signed-off-by: Artem Bityutskiy Signed-off-by: David Woodhouse --- include/linux/mtd/partitions.h | 2 ++ 1 file changed, 2 insertions(+) (limited to 'include') diff --git a/include/linux/mtd/partitions.h b/include/linux/mtd/partitions.h index af6dcb992bc3..b70313d33ff8 100644 --- a/include/linux/mtd/partitions.h +++ b/include/linux/mtd/partitions.h @@ -47,6 +47,8 @@ struct mtd_partition { #define MTDPART_SIZ_FULL (0) +struct mtd_info; + int add_mtd_partitions(struct mtd_info *, const struct mtd_partition *, int); int del_mtd_partitions(struct mtd_info *); -- cgit v1.2.3 From 6afc4fdb3e94ba60cd566cb878b60c6c01979277 Mon Sep 17 00:00:00 2001 From: Saeed Bishara Date: Tue, 28 Jul 2009 04:56:43 -0700 Subject: mtd: fix the conversion from dev to mtd_info The patch fixes a bug when converting dev to mtd_info by using the drvdata of the dev, the previous code used container_of(dev, struct mtd_info, dev), but won't work for the mtdXro devices as they created without being contained inside mtd_info structure. Signed-off-by: Saeed Bishara Signed-off-by: David Woodhouse --- drivers/mtd/mtdcore.c | 7 ++++--- include/linux/mtd/mtd.h | 2 +- 2 files changed, 5 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/drivers/mtd/mtdcore.c b/drivers/mtd/mtdcore.c index fac54a3fa3f1..00ebf7af7467 100644 --- a/drivers/mtd/mtdcore.c +++ b/drivers/mtd/mtdcore.c @@ -65,8 +65,8 @@ static void mtd_release(struct device *dev) static int mtd_cls_suspend(struct device *dev, pm_message_t state) { struct mtd_info *mtd = dev_to_mtd(dev); - - if (mtd->suspend) + + if (mtd && mtd->suspend) return mtd->suspend(mtd); else return 0; @@ -76,7 +76,7 @@ static int mtd_cls_resume(struct device *dev) { struct mtd_info *mtd = dev_to_mtd(dev); - if (mtd->resume) + if (mtd && mtd->resume) mtd->resume(mtd); return 0; } @@ -298,6 +298,7 @@ int add_mtd_device(struct mtd_info *mtd) mtd->dev.class = &mtd_class; mtd->dev.devt = MTD_DEVT(i); dev_set_name(&mtd->dev, "mtd%d", i); + dev_set_drvdata(&mtd->dev, mtd); if (device_register(&mtd->dev) != 0) { mtd_table[i] = NULL; break; diff --git a/include/linux/mtd/mtd.h b/include/linux/mtd/mtd.h index 5675b63a0631..0f32a9b6ff55 100644 --- a/include/linux/mtd/mtd.h +++ b/include/linux/mtd/mtd.h @@ -251,7 +251,7 @@ struct mtd_info { static inline struct mtd_info *dev_to_mtd(struct device *dev) { - return dev ? container_of(dev, struct mtd_info, dev) : NULL; + return dev ? dev_get_drvdata(dev) : NULL; } static inline uint32_t mtd_div_by_eb(uint64_t sz, struct mtd_info *mtd) -- cgit v1.2.3 From 3e352aa8ee2bd48f1a19c7742810b3a4a7ba605e Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Mon, 3 Aug 2009 14:10:11 +0900 Subject: x86, percpu: Fix DECLARE/DEFINE_PER_CPU_PAGE_ALIGNED() DECLARE/DEFINE_PER_CPU_PAGE_ALIGNED() put percpu variables in .page_aligned section without adding any alignment restrictions. Currently, this doesn't cause any problem because all users of the macros have explicit page alignment and page-sized but it's much safer to enforce page alignment from the macros. After all, it's what they claim to do. Add __aligned(PAGE_SIZE) to DECLARE/DEFINE_PER_CPU_PAGE_ALIGNED() and drop explicit alignment from it users. Signed-off-by: Tejun Heo Cc: Ingo Molnar Signed-off-by: H. Peter Anvin --- arch/x86/kernel/cpu/common.c | 3 +-- include/linux/percpu-defs.h | 8 +++++--- 2 files changed, 6 insertions(+), 5 deletions(-) (limited to 'include') diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c index f1961c07af9a..12493c5485e5 100644 --- a/arch/x86/kernel/cpu/common.c +++ b/arch/x86/kernel/cpu/common.c @@ -1008,8 +1008,7 @@ static const unsigned int exception_stack_sizes[N_EXCEPTION_STACKS] = { }; static DEFINE_PER_CPU_PAGE_ALIGNED(char, exception_stacks - [(N_EXCEPTION_STACKS - 1) * EXCEPTION_STKSZ + DEBUG_STKSZ]) - __aligned(PAGE_SIZE); + [(N_EXCEPTION_STACKS - 1) * EXCEPTION_STKSZ + DEBUG_STKSZ]); /* May not be marked __init: used by software suspend */ void syscall_init(void) diff --git a/include/linux/percpu-defs.h b/include/linux/percpu-defs.h index 68438e18fff4..afd5f8b7061f 100644 --- a/include/linux/percpu-defs.h +++ b/include/linux/percpu-defs.h @@ -69,11 +69,13 @@ /* * Declaration/definition used for per-CPU variables that must be page aligned. */ -#define DECLARE_PER_CPU_PAGE_ALIGNED(type, name) \ - DECLARE_PER_CPU_SECTION(type, name, ".page_aligned") +#define DECLARE_PER_CPU_PAGE_ALIGNED(type, name) \ + DECLARE_PER_CPU_SECTION(type, name, ".page_aligned") \ + __aligned(PAGE_SIZE) #define DEFINE_PER_CPU_PAGE_ALIGNED(type, name) \ - DEFINE_PER_CPU_SECTION(type, name, ".page_aligned") + DEFINE_PER_CPU_SECTION(type, name, ".page_aligned") \ + __aligned(PAGE_SIZE) /* * Intermodule exports for per-CPU variables. -- cgit v1.2.3 From af0d3b103bcfa877343ee338de12002cd50c9ee5 Mon Sep 17 00:00:00 2001 From: Dave Young Date: Mon, 3 Aug 2009 04:26:16 +0000 Subject: bluetooth: rfcomm_init bug fix rfcomm tty may be used before rfcomm_tty_driver initilized, The problem is that now socket layer init before tty layer, if userspace program do socket callback right here then oops will happen. reporting in: http://marc.info/?l=linux-bluetooth&m=124404919324542&w=2 make 3 changes: 1. remove #ifdef in rfcomm/core.c, make it blank function when rfcomm tty not selected in rfcomm.h 2. tune the rfcomm_init error patch to ensure tty driver initilized before rfcomm socket usage. 3. remove __exit for rfcomm_cleanup_sockets because above change need call it in a __init function. Reported-by: Oliver Hartkopp Tested-by: Oliver Hartkopp Signed-off-by: Dave Young Signed-off-by: David S. Miller --- include/net/bluetooth/rfcomm.h | 12 +++++++++++- net/bluetooth/rfcomm/core.c | 27 +++++++++++++++++++-------- net/bluetooth/rfcomm/sock.c | 2 +- 3 files changed, 31 insertions(+), 10 deletions(-) (limited to 'include') diff --git a/include/net/bluetooth/rfcomm.h b/include/net/bluetooth/rfcomm.h index 80072611d26a..c274993234e3 100644 --- a/include/net/bluetooth/rfcomm.h +++ b/include/net/bluetooth/rfcomm.h @@ -355,7 +355,17 @@ struct rfcomm_dev_list_req { }; int rfcomm_dev_ioctl(struct sock *sk, unsigned int cmd, void __user *arg); + +#ifdef CONFIG_BT_RFCOMM_TTY int rfcomm_init_ttys(void); void rfcomm_cleanup_ttys(void); - +#else +static inline int rfcomm_init_ttys(void) +{ + return 0; +} +static inline void rfcomm_cleanup_ttys(void) +{ +} +#endif #endif /* __RFCOMM_H */ diff --git a/net/bluetooth/rfcomm/core.c b/net/bluetooth/rfcomm/core.c index e50566ebf9f9..94b3388c188b 100644 --- a/net/bluetooth/rfcomm/core.c +++ b/net/bluetooth/rfcomm/core.c @@ -2080,28 +2080,41 @@ static CLASS_ATTR(rfcomm_dlc, S_IRUGO, rfcomm_dlc_sysfs_show, NULL); /* ---- Initialization ---- */ static int __init rfcomm_init(void) { + int ret; + l2cap_load(); hci_register_cb(&rfcomm_cb); rfcomm_thread = kthread_run(rfcomm_run, NULL, "krfcommd"); if (IS_ERR(rfcomm_thread)) { - hci_unregister_cb(&rfcomm_cb); - return PTR_ERR(rfcomm_thread); + ret = PTR_ERR(rfcomm_thread); + goto out_thread; } if (class_create_file(bt_class, &class_attr_rfcomm_dlc) < 0) BT_ERR("Failed to create RFCOMM info file"); - rfcomm_init_sockets(); + ret = rfcomm_init_ttys(); + if (ret) + goto out_tty; -#ifdef CONFIG_BT_RFCOMM_TTY - rfcomm_init_ttys(); -#endif + ret = rfcomm_init_sockets(); + if (ret) + goto out_sock; BT_INFO("RFCOMM ver %s", VERSION); return 0; + +out_sock: + rfcomm_cleanup_ttys(); +out_tty: + kthread_stop(rfcomm_thread); +out_thread: + hci_unregister_cb(&rfcomm_cb); + + return ret; } static void __exit rfcomm_exit(void) @@ -2112,9 +2125,7 @@ static void __exit rfcomm_exit(void) kthread_stop(rfcomm_thread); -#ifdef CONFIG_BT_RFCOMM_TTY rfcomm_cleanup_ttys(); -#endif rfcomm_cleanup_sockets(); } diff --git a/net/bluetooth/rfcomm/sock.c b/net/bluetooth/rfcomm/sock.c index 7f482784e9f7..0b85e8116859 100644 --- a/net/bluetooth/rfcomm/sock.c +++ b/net/bluetooth/rfcomm/sock.c @@ -1132,7 +1132,7 @@ error: return err; } -void __exit rfcomm_cleanup_sockets(void) +void rfcomm_cleanup_sockets(void) { class_remove_file(bt_class, &class_attr_rfcomm); -- cgit v1.2.3 From 371842448c05b42d11a4be1c8e4e81d62ecc7534 Mon Sep 17 00:00:00 2001 From: "Luis R. Rodriguez" Date: Thu, 30 Jul 2009 17:43:48 -0700 Subject: cfg80211: fix regression on beacon world roaming feature A regression was added through patch a4ed90d6: "cfg80211: respect API on orig_flags on channel for beacon hint" We did indeed respect _orig flags but the intention was not clearly stated in the commit log. This patch fixes firmware issues picked up by iwlwifi when we lift passive scan of beaconing restrictions on channels its EEPROM has been configured to always enable. By doing so though we also disallowed beacon hints on devices registering their wiphy with custom world regulatory domains enabled, this happens to be currently ath5k, ath9k and ar9170. The passive scan and beacon restrictions on those devices would never be lifted even if we did find a beacon and the hardware did support such enhancements when world roaming. Since Johannes indicates iwlwifi firmware cannot be changed to allow beacon hinting we set up a flag now to specifically allow drivers to disable beacon hints for devices which cannot use them. We enable the flag on iwlwifi to disable beacon hints and by default enable it for all other drivers. It should be noted beacon hints lift passive scan flags and beacon restrictions when we receive a beacon from an AP on any 5 GHz non-DFS channels, and channels 12-14 on the 2.4 GHz band. We don't bother with channels 1-11 as those channels are allowed world wide. This should fix world roaming for ath5k, ath9k and ar9170, thereby improving scan time when we receive the first beacon from any AP, and also enabling beaconing operation (AP/IBSS/Mesh) on cards which would otherwise not be allowed to do so. Drivers not using custom regulatory stuff (wiphy_apply_custom_regulatory()) were not affected by this as the orig_flags for the channels would have been cleared upon wiphy registration. I tested this with a world roaming ath5k card. Cc: Jouni Malinen Signed-off-by: Luis R. Rodriguez Reviewed-by: Johannes Berg Signed-off-by: John W. Linville --- drivers/net/wireless/iwlwifi/iwl-core.c | 3 +++ drivers/net/wireless/iwlwifi/iwl3945-base.c | 3 +++ include/net/cfg80211.h | 5 +++++ net/wireless/reg.c | 9 +++++---- net/wireless/reg.h | 3 ++- 5 files changed, 18 insertions(+), 5 deletions(-) (limited to 'include') diff --git a/drivers/net/wireless/iwlwifi/iwl-core.c b/drivers/net/wireless/iwlwifi/iwl-core.c index 6ab07165ea28..18b135f510e5 100644 --- a/drivers/net/wireless/iwlwifi/iwl-core.c +++ b/drivers/net/wireless/iwlwifi/iwl-core.c @@ -1332,6 +1332,9 @@ int iwl_setup_mac(struct iwl_priv *priv) hw->wiphy->custom_regulatory = true; + /* Firmware does not support this */ + hw->wiphy->disable_beacon_hints = true; + hw->wiphy->max_scan_ssids = PROBE_OPTION_MAX; /* we create the 802.11 header and a zero-length SSID element */ hw->wiphy->max_scan_ie_len = IWL_MAX_PROBE_REQUEST - 24 - 2; diff --git a/drivers/net/wireless/iwlwifi/iwl3945-base.c b/drivers/net/wireless/iwlwifi/iwl3945-base.c index 2f50ab60bfdf..523843369ca2 100644 --- a/drivers/net/wireless/iwlwifi/iwl3945-base.c +++ b/drivers/net/wireless/iwlwifi/iwl3945-base.c @@ -3968,6 +3968,9 @@ static int iwl3945_setup_mac(struct iwl_priv *priv) hw->wiphy->custom_regulatory = true; + /* Firmware does not support this */ + hw->wiphy->disable_beacon_hints = true; + hw->wiphy->max_scan_ssids = PROBE_OPTION_MAX_3945; /* we create the 802.11 header and a zero-length SSID element */ hw->wiphy->max_scan_ie_len = IWL_MAX_PROBE_REQUEST - 24 - 2; diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index 1a21895b732b..d1892d66701a 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -979,6 +979,10 @@ struct cfg80211_ops { * channels at a later time. This can be used for devices which do not * have calibration information gauranteed for frequencies or settings * outside of its regulatory domain. + * @disable_beacon_hints: enable this if your driver needs to ensure that + * passive scan flags and beaconing flags may not be lifted by cfg80211 + * due to regulatory beacon hints. For more information on beacon + * hints read the documenation for regulatory_hint_found_beacon() * @reg_notifier: the driver's regulatory notification callback * @regd: the driver's regulatory domain, if one was requested via * the regulatory_hint() API. This can be used by the driver @@ -1004,6 +1008,7 @@ struct wiphy { bool custom_regulatory; bool strict_regulatory; + bool disable_beacon_hints; enum cfg80211_signal_type signal_type; diff --git a/net/wireless/reg.c b/net/wireless/reg.c index 5e14371cda70..75a406d33619 100644 --- a/net/wireless/reg.c +++ b/net/wireless/reg.c @@ -1089,17 +1089,18 @@ static void handle_reg_beacon(struct wiphy *wiphy, chan->beacon_found = true; + if (wiphy->disable_beacon_hints) + return; + chan_before.center_freq = chan->center_freq; chan_before.flags = chan->flags; - if ((chan->flags & IEEE80211_CHAN_PASSIVE_SCAN) && - !(chan->orig_flags & IEEE80211_CHAN_PASSIVE_SCAN)) { + if (chan->flags & IEEE80211_CHAN_PASSIVE_SCAN) { chan->flags &= ~IEEE80211_CHAN_PASSIVE_SCAN; channel_changed = true; } - if ((chan->flags & IEEE80211_CHAN_NO_IBSS) && - !(chan->orig_flags & IEEE80211_CHAN_NO_IBSS)) { + if (chan->flags & IEEE80211_CHAN_NO_IBSS) { chan->flags &= ~IEEE80211_CHAN_NO_IBSS; channel_changed = true; } diff --git a/net/wireless/reg.h b/net/wireless/reg.h index e37829a49dc4..4e167a8e11be 100644 --- a/net/wireless/reg.h +++ b/net/wireless/reg.h @@ -30,7 +30,8 @@ int set_regdom(const struct ieee80211_regdomain *rd); * non-radar 5 GHz channels. * * Drivers do not need to call this, cfg80211 will do it for after a scan - * on a newly found BSS. + * on a newly found BSS. If you cannot make use of this feature you can + * set the wiphy->disable_beacon_hints to true. */ int regulatory_hint_found_beacon(struct wiphy *wiphy, struct ieee80211_channel *beacon_chan, -- cgit v1.2.3 From 7320700df1864b601cef5adbddce8654a0e3f78b Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Mon, 3 Aug 2009 17:01:53 -0400 Subject: drm/radeon: add some new r7xx pci ids Signed-off-by: Alex Deucher Signed-off-by: Dave Airlie --- include/drm/drm_pciids.h | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'include') diff --git a/include/drm/drm_pciids.h b/include/drm/drm_pciids.h index 7174818c2c13..9d4c00491547 100644 --- a/include/drm/drm_pciids.h +++ b/include/drm/drm_pciids.h @@ -257,9 +257,12 @@ {0x1002, 0x940F, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_R600|RADEON_NEW_MEMMAP}, \ {0x1002, 0x94A0, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV740|RADEON_IS_MOBILITY|RADEON_NEW_MEMMAP}, \ {0x1002, 0x94A1, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV740|RADEON_IS_MOBILITY|RADEON_NEW_MEMMAP}, \ + {0x1002, 0x94A3, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV740|RADEON_IS_MOBILITY|RADEON_NEW_MEMMAP}, \ {0x1002, 0x94B1, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV740|RADEON_NEW_MEMMAP}, \ {0x1002, 0x94B3, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV740|RADEON_NEW_MEMMAP}, \ + {0x1002, 0x94B4, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV740|RADEON_NEW_MEMMAP}, \ {0x1002, 0x94B5, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV740|RADEON_NEW_MEMMAP}, \ + {0x1002, 0x94B9, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV740|RADEON_IS_MOBILITY|RADEON_NEW_MEMMAP}, \ {0x1002, 0x9440, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV770|RADEON_NEW_MEMMAP}, \ {0x1002, 0x9441, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV770|RADEON_NEW_MEMMAP}, \ {0x1002, 0x9442, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV770|RADEON_NEW_MEMMAP}, \ @@ -288,6 +291,7 @@ {0x1002, 0x948F, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV730|RADEON_NEW_MEMMAP}, \ {0x1002, 0x9490, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV730|RADEON_NEW_MEMMAP}, \ {0x1002, 0x9491, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV730|RADEON_IS_MOBILITY|RADEON_NEW_MEMMAP}, \ + {0x1002, 0x9495, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV730|RADEON_NEW_MEMMAP}, \ {0x1002, 0x9498, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV730|RADEON_NEW_MEMMAP}, \ {0x1002, 0x949C, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV730|RADEON_NEW_MEMMAP}, \ {0x1002, 0x949E, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV730|RADEON_NEW_MEMMAP}, \ @@ -325,6 +329,7 @@ {0x1002, 0x9552, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV710|RADEON_IS_MOBILITY|RADEON_NEW_MEMMAP}, \ {0x1002, 0x9553, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV710|RADEON_IS_MOBILITY|RADEON_NEW_MEMMAP}, \ {0x1002, 0x9555, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV710|RADEON_IS_MOBILITY|RADEON_NEW_MEMMAP}, \ + {0x1002, 0x9557, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV710|RADEON_IS_MOBILITY|RADEON_NEW_MEMMAP}, \ {0x1002, 0x9580, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV630|RADEON_NEW_MEMMAP}, \ {0x1002, 0x9581, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV630|RADEON_IS_MOBILITY|RADEON_NEW_MEMMAP}, \ {0x1002, 0x9583, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RV630|RADEON_IS_MOBILITY|RADEON_NEW_MEMMAP}, \ -- cgit v1.2.3 From 8a4c47f346cc7a12d0897c05eb3cc1add26b487f Mon Sep 17 00:00:00 2001 From: Zhao Yakui Date: Mon, 20 Jul 2009 13:48:04 +0800 Subject: drm: Remove the unused prefix in DRM_DEBUG_KMS/DRIVER/MODE We will have to add a prefix when using the macro defintion of DRM_DEBUG_KMS /DRM_DEBUG_DRIVER/MODE. It is not convenient. We should use the DRM_NAME as default prefix. So remove the prefix in the macro definition of DRM_DEBUG_KMS/DRIVER/MODE. Signed-off-by: Zhao Yakui Acked-by: Ian Romanick Signed-off-by: Dave Airlie --- drivers/gpu/drm/drm_modes.c | 8 +++----- drivers/gpu/drm/i915/i915_dma.c | 35 +++++++++++++++-------------------- drivers/gpu/drm/i915/intel_lvds.c | 10 +++------- drivers/gpu/drm/i915/intel_sdvo.c | 35 ++++++++++++++++------------------- include/drm/drmP.h | 18 +++++++++--------- 5 files changed, 46 insertions(+), 60 deletions(-) (limited to 'include') diff --git a/drivers/gpu/drm/drm_modes.c b/drivers/gpu/drm/drm_modes.c index fd489d76fbbc..5eca2d5c5f23 100644 --- a/drivers/gpu/drm/drm_modes.c +++ b/drivers/gpu/drm/drm_modes.c @@ -40,7 +40,6 @@ #include "drm.h" #include "drm_crtc.h" -#define DRM_MODESET_DEBUG "drm_mode" /** * drm_mode_debug_printmodeline - debug print a mode * @dev: DRM device @@ -53,8 +52,8 @@ */ void drm_mode_debug_printmodeline(struct drm_display_mode *mode) { - DRM_DEBUG_MODE(DRM_MODESET_DEBUG, - "Modeline %d:\"%s\" %d %d %d %d %d %d %d %d %d %d 0x%x 0x%x\n", + DRM_DEBUG_MODE("Modeline %d:\"%s\" %d %d %d %d %d %d %d %d %d %d " + "0x%x 0x%x\n", mode->base.id, mode->name, mode->vrefresh, mode->clock, mode->hdisplay, mode->hsync_start, mode->hsync_end, mode->htotal, @@ -819,8 +818,7 @@ void drm_mode_prune_invalid(struct drm_device *dev, list_del(&mode->head); if (verbose) { drm_mode_debug_printmodeline(mode); - DRM_DEBUG_MODE(DRM_MODESET_DEBUG, - "Not using %s mode %d\n", + DRM_DEBUG_MODE("Not using %s mode %d\n", mode->name, mode->status); } drm_mode_destroy(dev, mode); diff --git a/drivers/gpu/drm/i915/i915_dma.c b/drivers/gpu/drm/i915/i915_dma.c index 8c4783180bf6..14625e146f18 100644 --- a/drivers/gpu/drm/i915/i915_dma.c +++ b/drivers/gpu/drm/i915/i915_dma.c @@ -33,8 +33,6 @@ #include "i915_drm.h" #include "i915_drv.h" -#define I915_DRV "i915_drv" - /* Really want an OS-independent resettable timer. Would like to have * this loop run for (eg) 3 sec, but have the timer reset every time * the head pointer changes, so that EBUSY only happens if the ring @@ -101,7 +99,7 @@ static int i915_init_phys_hws(struct drm_device *dev) memset(dev_priv->hw_status_page, 0, PAGE_SIZE); I915_WRITE(HWS_PGA, dev_priv->dma_status_page); - DRM_DEBUG_DRIVER(I915_DRV, "Enabled hardware status page\n"); + DRM_DEBUG_DRIVER("Enabled hardware status page\n"); return 0; } @@ -187,8 +185,7 @@ static int i915_initialize(struct drm_device * dev, drm_i915_init_t * init) master_priv->sarea_priv = (drm_i915_sarea_t *) ((u8 *)master_priv->sarea->handle + init->sarea_priv_offset); } else { - DRM_DEBUG_DRIVER(I915_DRV, - "sarea not found assuming DRI2 userspace\n"); + DRM_DEBUG_DRIVER("sarea not found assuming DRI2 userspace\n"); } if (init->ring_size != 0) { @@ -238,7 +235,7 @@ static int i915_dma_resume(struct drm_device * dev) { drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private; - DRM_DEBUG_DRIVER(I915_DRV, "%s\n", __func__); + DRM_DEBUG_DRIVER("%s\n", __func__); if (dev_priv->ring.map.handle == NULL) { DRM_ERROR("can not ioremap virtual address for" @@ -251,14 +248,14 @@ static int i915_dma_resume(struct drm_device * dev) DRM_ERROR("Can not find hardware status page\n"); return -EINVAL; } - DRM_DEBUG_DRIVER(I915_DRV, "hw status page @ %p\n", + DRM_DEBUG_DRIVER("hw status page @ %p\n", dev_priv->hw_status_page); if (dev_priv->status_gfx_addr != 0) I915_WRITE(HWS_PGA, dev_priv->status_gfx_addr); else I915_WRITE(HWS_PGA, dev_priv->dma_status_page); - DRM_DEBUG_DRIVER(I915_DRV, "Enabled hardware status page\n"); + DRM_DEBUG_DRIVER("Enabled hardware status page\n"); return 0; } @@ -552,7 +549,7 @@ static int i915_dispatch_flip(struct drm_device * dev) if (!master_priv->sarea_priv) return -EINVAL; - DRM_DEBUG_DRIVER(I915_DRV, "%s: page=%d pfCurrentPage=%d\n", + DRM_DEBUG_DRIVER("%s: page=%d pfCurrentPage=%d\n", __func__, dev_priv->current_page, master_priv->sarea_priv->pf_current_page); @@ -633,8 +630,7 @@ static int i915_batchbuffer(struct drm_device *dev, void *data, return -EINVAL; } - DRM_DEBUG_DRIVER(I915_DRV, - "i915 batchbuffer, start %x used %d cliprects %d\n", + DRM_DEBUG_DRIVER("i915 batchbuffer, start %x used %d cliprects %d\n", batch->start, batch->used, batch->num_cliprects); RING_LOCK_TEST_WITH_RETURN(dev, file_priv); @@ -681,8 +677,7 @@ static int i915_cmdbuffer(struct drm_device *dev, void *data, void *batch_data; int ret; - DRM_DEBUG_DRIVER(I915_DRV, - "i915 cmdbuffer, buf %p sz %d cliprects %d\n", + DRM_DEBUG_DRIVER("i915 cmdbuffer, buf %p sz %d cliprects %d\n", cmdbuf->buf, cmdbuf->sz, cmdbuf->num_cliprects); RING_LOCK_TEST_WITH_RETURN(dev, file_priv); @@ -735,7 +730,7 @@ static int i915_flip_bufs(struct drm_device *dev, void *data, { int ret; - DRM_DEBUG_DRIVER(I915_DRV, "%s\n", __func__); + DRM_DEBUG_DRIVER("%s\n", __func__); RING_LOCK_TEST_WITH_RETURN(dev, file_priv); @@ -778,7 +773,7 @@ static int i915_getparam(struct drm_device *dev, void *data, value = dev_priv->num_fence_regs - dev_priv->fence_reg_start; break; default: - DRM_DEBUG_DRIVER(I915_DRV, "Unknown parameter %d\n", + DRM_DEBUG_DRIVER("Unknown parameter %d\n", param->param); return -EINVAL; } @@ -819,7 +814,7 @@ static int i915_setparam(struct drm_device *dev, void *data, dev_priv->fence_reg_start = param->value; break; default: - DRM_DEBUG_DRIVER(I915_DRV, "unknown parameter %d\n", + DRM_DEBUG_DRIVER("unknown parameter %d\n", param->param); return -EINVAL; } @@ -846,7 +841,7 @@ static int i915_set_status_page(struct drm_device *dev, void *data, return 0; } - DRM_DEBUG("set status page addr 0x%08x\n", (u32)hws->addr); + DRM_DEBUG_DRIVER("set status page addr 0x%08x\n", (u32)hws->addr); dev_priv->status_gfx_addr = hws->addr & (0x1ffff<<12); @@ -868,9 +863,9 @@ static int i915_set_status_page(struct drm_device *dev, void *data, memset(dev_priv->hw_status_page, 0, PAGE_SIZE); I915_WRITE(HWS_PGA, dev_priv->status_gfx_addr); - DRM_DEBUG_DRIVER(I915_DRV, "load hws HWS_PGA with gfx mem 0x%x\n", + DRM_DEBUG_DRIVER("load hws HWS_PGA with gfx mem 0x%x\n", dev_priv->status_gfx_addr); - DRM_DEBUG_DRIVER(I915_DRV, "load hws at %p\n", + DRM_DEBUG_DRIVER("load hws at %p\n", dev_priv->hw_status_page); return 0; } @@ -1310,7 +1305,7 @@ int i915_driver_open(struct drm_device *dev, struct drm_file *file_priv) { struct drm_i915_file_private *i915_file_priv; - DRM_DEBUG_DRIVER(I915_DRV, "\n"); + DRM_DEBUG_DRIVER("\n"); i915_file_priv = (struct drm_i915_file_private *) kmalloc(sizeof(*i915_file_priv), GFP_KERNEL); diff --git a/drivers/gpu/drm/i915/intel_lvds.c b/drivers/gpu/drm/i915/intel_lvds.c index 9ab38efffecf..b59c65d19d81 100644 --- a/drivers/gpu/drm/i915/intel_lvds.c +++ b/drivers/gpu/drm/i915/intel_lvds.c @@ -38,8 +38,6 @@ #include "i915_drv.h" #include -#define I915_LVDS "i915_lvds" - /* * the following four scaling options are defined. * #define DRM_MODE_SCALE_NON_GPU 0 @@ -673,8 +671,7 @@ static int intel_lvds_set_property(struct drm_connector *connector, struct drm_crtc *crtc = connector->encoder->crtc; struct intel_lvds_priv *lvds_priv = intel_output->dev_priv; if (value == DRM_MODE_SCALE_NON_GPU) { - DRM_DEBUG_KMS(I915_LVDS, - "non_GPU property is unsupported\n"); + DRM_DEBUG_KMS("non_GPU property is unsupported\n"); return 0; } if (lvds_priv->fitting_mode == value) { @@ -731,8 +728,7 @@ static const struct drm_encoder_funcs intel_lvds_enc_funcs = { static int __init intel_no_lvds_dmi_callback(const struct dmi_system_id *id) { - DRM_DEBUG_KMS(I915_LVDS, - "Skipping LVDS initialization for %s\n", id->ident); + DRM_DEBUG_KMS("Skipping LVDS initialization for %s\n", id->ident); return 1; } @@ -1013,7 +1009,7 @@ out: return; failed: - DRM_DEBUG_KMS(I915_LVDS, "No LVDS modes found, disabling.\n"); + DRM_DEBUG_KMS("No LVDS modes found, disabling.\n"); if (intel_output->ddc_bus) intel_i2c_destroy(intel_output->ddc_bus); drm_connector_cleanup(connector); diff --git a/drivers/gpu/drm/i915/intel_sdvo.c b/drivers/gpu/drm/i915/intel_sdvo.c index 4f0c30948bc4..abef69c8a49a 100644 --- a/drivers/gpu/drm/i915/intel_sdvo.c +++ b/drivers/gpu/drm/i915/intel_sdvo.c @@ -36,7 +36,6 @@ #include "intel_sdvo_regs.h" #undef SDVO_DEBUG -#define I915_SDVO "i915_sdvo" struct intel_sdvo_priv { u8 slave_addr; @@ -178,7 +177,7 @@ static bool intel_sdvo_read_byte(struct intel_output *intel_output, u8 addr, return true; } - DRM_DEBUG("i2c transfer returned %d\n", ret); + DRM_DEBUG_KMS("i2c transfer returned %d\n", ret); return false; } @@ -288,7 +287,7 @@ static void intel_sdvo_debug_write(struct intel_output *intel_output, u8 cmd, struct intel_sdvo_priv *sdvo_priv = intel_output->dev_priv; int i; - DRM_DEBUG_KMS(I915_SDVO, "%s: W: %02X ", + DRM_DEBUG_KMS("%s: W: %02X ", SDVO_NAME(sdvo_priv), cmd); for (i = 0; i < args_len; i++) DRM_LOG_KMS("%02X ", ((u8 *)args)[i]); @@ -341,7 +340,7 @@ static void intel_sdvo_debug_response(struct intel_output *intel_output, struct intel_sdvo_priv *sdvo_priv = intel_output->dev_priv; int i; - DRM_DEBUG_KMS(I915_SDVO, "%s: R: ", SDVO_NAME(sdvo_priv)); + DRM_DEBUG_KMS("%s: R: ", SDVO_NAME(sdvo_priv)); for (i = 0; i < response_len; i++) DRM_LOG_KMS("%02X ", ((u8 *)response)[i]); for (; i < 8; i++) @@ -658,10 +657,10 @@ static int intel_sdvo_get_clock_rate_mult(struct intel_output *intel_output) status = intel_sdvo_read_response(intel_output, &response, 1); if (status != SDVO_CMD_STATUS_SUCCESS) { - DRM_DEBUG("Couldn't get SDVO clock rate multiplier\n"); + DRM_DEBUG_KMS("Couldn't get SDVO clock rate multiplier\n"); return SDVO_CLOCK_RATE_MULT_1X; } else { - DRM_DEBUG("Current clock rate multiplier: %d\n", response); + DRM_DEBUG_KMS("Current clock rate multiplier: %d\n", response); } return response; @@ -942,14 +941,14 @@ static void intel_sdvo_set_tv_format(struct intel_output *output) format = &sdvo_priv->tv_format; memset(&unset, 0, sizeof(unset)); if (memcmp(format, &unset, sizeof(*format))) { - DRM_DEBUG("%s: Choosing default TV format of NTSC-M\n", + DRM_DEBUG_KMS("%s: Choosing default TV format of NTSC-M\n", SDVO_NAME(sdvo_priv)); format->ntsc_m = 1; intel_sdvo_write_cmd(output, SDVO_CMD_SET_TV_FORMAT, format, sizeof(*format)); status = intel_sdvo_read_response(output, NULL, 0); if (status != SDVO_CMD_STATUS_SUCCESS) - DRM_DEBUG("%s: Failed to set TV format\n", + DRM_DEBUG_KMS("%s: Failed to set TV format\n", SDVO_NAME(sdvo_priv)); } } @@ -1220,8 +1219,8 @@ static void intel_sdvo_dpms(struct drm_encoder *encoder, int mode) * a given it the status is a success, we succeeded. */ if (status == SDVO_CMD_STATUS_SUCCESS && !input1) { - DRM_DEBUG("First %s output reported failure to sync\n", - SDVO_NAME(sdvo_priv)); + DRM_DEBUG_KMS("First %s output reported failure to " + "sync\n", SDVO_NAME(sdvo_priv)); } if (0) @@ -1316,8 +1315,8 @@ static void intel_sdvo_restore(struct drm_connector *connector) intel_wait_for_vblank(dev); status = intel_sdvo_get_trained_inputs(intel_output, &input1, &input2); if (status == SDVO_CMD_STATUS_SUCCESS && !input1) - DRM_DEBUG("First %s output reported failure to sync\n", - SDVO_NAME(sdvo_priv)); + DRM_DEBUG_KMS("First %s output reported failure to " + "sync\n", SDVO_NAME(sdvo_priv)); } intel_sdvo_set_active_outputs(intel_output, sdvo_priv->save_active_outputs); @@ -1395,7 +1394,7 @@ int intel_sdvo_supports_hotplug(struct drm_connector *connector) u8 response[2]; u8 status; struct intel_output *intel_output; - DRM_DEBUG("\n"); + DRM_DEBUG_KMS("\n"); if (!connector) return 0; @@ -1460,7 +1459,7 @@ static enum drm_connector_status intel_sdvo_detect(struct drm_connector *connect intel_sdvo_write_cmd(intel_output, SDVO_CMD_GET_ATTACHED_DISPLAYS, NULL, 0); status = intel_sdvo_read_response(intel_output, &response, 2); - DRM_DEBUG("SDVO response %d %d\n", response[0], response[1]); + DRM_DEBUG_KMS("SDVO response %d %d\n", response[0], response[1]); if (status != SDVO_CMD_STATUS_SUCCESS) return connector_status_unknown; @@ -1905,8 +1904,7 @@ bool intel_sdvo_init(struct drm_device *dev, int output_device) /* Read the regs to test if we can talk to the device */ for (i = 0; i < 0x40; i++) { if (!intel_sdvo_read_byte(intel_output, i, &ch[i])) { - DRM_DEBUG_KMS(I915_SDVO, - "No SDVO device found on SDVO%c\n", + DRM_DEBUG_KMS("No SDVO device found on SDVO%c\n", output_device == SDVOB ? 'B' : 'C'); goto err_i2c; } @@ -1989,8 +1987,7 @@ bool intel_sdvo_init(struct drm_device *dev, int output_device) sdvo_priv->controlled_output = 0; memcpy (bytes, &sdvo_priv->caps.output_flags, 2); - DRM_DEBUG_KMS(I915_SDVO, - "%s: Unknown SDVO output type (0x%02x%02x)\n", + DRM_DEBUG_KMS("%s: Unknown SDVO output type (0x%02x%02x)\n", SDVO_NAME(sdvo_priv), bytes[0], bytes[1]); encoder_type = DRM_MODE_ENCODER_NONE; @@ -2022,7 +2019,7 @@ bool intel_sdvo_init(struct drm_device *dev, int output_device) &sdvo_priv->pixel_clock_max); - DRM_DEBUG_KMS(I915_SDVO, "%s device VID/DID: %02X:%02X.%02X, " + DRM_DEBUG_KMS("%s device VID/DID: %02X:%02X.%02X, " "clock range %dMHz - %dMHz, " "input 1: %c, input 2: %c, " "output 1: %c, output 2: %c\n", diff --git a/include/drm/drmP.h b/include/drm/drmP.h index 45b67d9c39c1..edbdb02a7a3f 100644 --- a/include/drm/drmP.h +++ b/include/drm/drmP.h @@ -174,19 +174,19 @@ extern void drm_ut_debug_printk(unsigned int request_level, __func__, fmt, ##args); \ } while (0) -#define DRM_DEBUG_DRIVER(prefix, fmt, args...) \ +#define DRM_DEBUG_DRIVER(fmt, args...) \ do { \ - drm_ut_debug_printk(DRM_UT_DRIVER, prefix, \ + drm_ut_debug_printk(DRM_UT_DRIVER, DRM_NAME, \ __func__, fmt, ##args); \ } while (0) -#define DRM_DEBUG_KMS(prefix, fmt, args...) \ +#define DRM_DEBUG_KMS(fmt, args...) \ do { \ - drm_ut_debug_printk(DRM_UT_KMS, prefix, \ + drm_ut_debug_printk(DRM_UT_KMS, DRM_NAME, \ __func__, fmt, ##args); \ } while (0) -#define DRM_DEBUG_MODE(prefix, fmt, args...) \ +#define DRM_DEBUG_MODE(fmt, args...) \ do { \ - drm_ut_debug_printk(DRM_UT_MODE, prefix, \ + drm_ut_debug_printk(DRM_UT_MODE, DRM_NAME, \ __func__, fmt, ##args); \ } while (0) #define DRM_LOG(fmt, args...) \ @@ -210,9 +210,9 @@ extern void drm_ut_debug_printk(unsigned int request_level, NULL, fmt, ##args); \ } while (0) #else -#define DRM_DEBUG_DRIVER(prefix, fmt, args...) do { } while (0) -#define DRM_DEBUG_KMS(prefix, fmt, args...) do { } while (0) -#define DRM_DEBUG_MODE(prefix, fmt, args...) do { } while (0) +#define DRM_DEBUG_DRIVER(fmt, args...) do { } while (0) +#define DRM_DEBUG_KMS(fmt, args...) do { } while (0) +#define DRM_DEBUG_MODE(fmt, args...) do { } while (0) #define DRM_DEBUG(fmt, arg...) do { } while (0) #define DRM_LOG(fmt, arg...) do { } while (0) #define DRM_LOG_KMS(fmt, args...) do { } while (0) -- cgit v1.2.3 From f940f37f022f7392ab81a35516222cbd46110b42 Mon Sep 17 00:00:00 2001 From: Zhao Yakui Date: Mon, 20 Jul 2009 13:48:05 +0800 Subject: drm: Remove the macro defintion of DRM_DEBUG_MODE Two macro definitions of DRM_DEBUG_KMS/MODE can be used to add the debug info related with KMS. It is confusing. So remove the macro definition of DRM_DEBUG_MODE. Instead it can be replaced by the DRM_DEBUG_KMS. Signed-off-by: Zhao Yakui Acked-by: Ian Romanick Signed-off-by: Dave Airlie --- drivers/gpu/drm/drm_modes.c | 4 ++-- include/drm/drmP.h | 7 ------- 2 files changed, 2 insertions(+), 9 deletions(-) (limited to 'include') diff --git a/drivers/gpu/drm/drm_modes.c b/drivers/gpu/drm/drm_modes.c index 5eca2d5c5f23..6b4d2dc3cdd9 100644 --- a/drivers/gpu/drm/drm_modes.c +++ b/drivers/gpu/drm/drm_modes.c @@ -52,7 +52,7 @@ */ void drm_mode_debug_printmodeline(struct drm_display_mode *mode) { - DRM_DEBUG_MODE("Modeline %d:\"%s\" %d %d %d %d %d %d %d %d %d %d " + DRM_DEBUG_KMS("Modeline %d:\"%s\" %d %d %d %d %d %d %d %d %d %d " "0x%x 0x%x\n", mode->base.id, mode->name, mode->vrefresh, mode->clock, mode->hdisplay, mode->hsync_start, @@ -818,7 +818,7 @@ void drm_mode_prune_invalid(struct drm_device *dev, list_del(&mode->head); if (verbose) { drm_mode_debug_printmodeline(mode); - DRM_DEBUG_MODE("Not using %s mode %d\n", + DRM_DEBUG_KMS("Not using %s mode %d\n", mode->name, mode->status); } drm_mode_destroy(dev, mode); diff --git a/include/drm/drmP.h b/include/drm/drmP.h index edbdb02a7a3f..6513d16cd029 100644 --- a/include/drm/drmP.h +++ b/include/drm/drmP.h @@ -88,7 +88,6 @@ struct drm_device; #define DRM_UT_CORE 0x01 #define DRM_UT_DRIVER 0x02 #define DRM_UT_KMS 0x04 -#define DRM_UT_MODE 0x08 extern void drm_ut_debug_printk(unsigned int request_level, const char *prefix, @@ -184,11 +183,6 @@ extern void drm_ut_debug_printk(unsigned int request_level, drm_ut_debug_printk(DRM_UT_KMS, DRM_NAME, \ __func__, fmt, ##args); \ } while (0) -#define DRM_DEBUG_MODE(fmt, args...) \ - do { \ - drm_ut_debug_printk(DRM_UT_MODE, DRM_NAME, \ - __func__, fmt, ##args); \ - } while (0) #define DRM_LOG(fmt, args...) \ do { \ drm_ut_debug_printk(DRM_UT_CORE, NULL, \ @@ -212,7 +206,6 @@ extern void drm_ut_debug_printk(unsigned int request_level, #else #define DRM_DEBUG_DRIVER(fmt, args...) do { } while (0) #define DRM_DEBUG_KMS(fmt, args...) do { } while (0) -#define DRM_DEBUG_MODE(fmt, args...) do { } while (0) #define DRM_DEBUG(fmt, arg...) do { } while (0) #define DRM_LOG(fmt, arg...) do { } while (0) #define DRM_LOG_KMS(fmt, args...) do { } while (0) -- cgit v1.2.3 From 87fdff81cd2d770f0adc742e21eb5e062ad20def Mon Sep 17 00:00:00 2001 From: Zhao Yakui Date: Mon, 20 Jul 2009 13:48:06 +0800 Subject: DRM: Add the explanation about DRM debug level Add the explanation about DRM debug level in the drmP header file. This is to explain how/where to use the different DRM debug level. Signed-off-by: Zhao Yakui Signed-off-by: Dave Airlie --- include/drm/drmP.h | 31 +++++++++++++++++++++++++++++++ 1 file changed, 31 insertions(+) (limited to 'include') diff --git a/include/drm/drmP.h b/include/drm/drmP.h index 6513d16cd029..e0f1c1fee58b 100644 --- a/include/drm/drmP.h +++ b/include/drm/drmP.h @@ -88,6 +88,37 @@ struct drm_device; #define DRM_UT_CORE 0x01 #define DRM_UT_DRIVER 0x02 #define DRM_UT_KMS 0x04 +/* + * Three debug levels are defined. + * drm_core, drm_driver, drm_kms + * drm_core level can be used in the generic drm code. For example: + * drm_ioctl, drm_mm, drm_memory + * The macro definiton of DRM_DEBUG is used. + * DRM_DEBUG(fmt, args...) + * The debug info by using the DRM_DEBUG can be obtained by adding + * the boot option of "drm.debug=1". + * + * drm_driver level can be used in the specific drm driver. It is used + * to add the debug info related with the drm driver. For example: + * i915_drv, i915_dma, i915_gem, radeon_drv, + * The macro definition of DRM_DEBUG_DRIVER can be used. + * DRM_DEBUG_DRIVER(fmt, args...) + * The debug info by using the DRM_DEBUG_DRIVER can be obtained by + * adding the boot option of "drm.debug=0x02" + * + * drm_kms level can be used in the KMS code related with specific drm driver. + * It is used to add the debug info related with KMS mode. For example: + * the connector/crtc , + * The macro definition of DRM_DEBUG_KMS can be used. + * DRM_DEBUG_KMS(fmt, args...) + * The debug info by using the DRM_DEBUG_KMS can be obtained by + * adding the boot option of "drm.debug=0x04" + * + * If we add the boot option of "drm.debug=0x06", we can get the debug info by + * using the DRM_DEBUG_KMS and DRM_DEBUG_DRIVER. + * If we add the boot option of "drm.debug=0x05", we can get the debug info by + * using the DRM_DEBUG_KMS and DRM_DEBUG. + */ extern void drm_ut_debug_printk(unsigned int request_level, const char *prefix, -- cgit v1.2.3 From 2066facca4c7dfe9f5068ece0200a4dbf10f49e1 Mon Sep 17 00:00:00 2001 From: Francisco Jerez Date: Sun, 2 Aug 2009 04:19:17 +0200 Subject: drm/kms: slave encoder interface. Define some helper functions to make easier to detach a KMS encoder implementation from the drm module of the GPU it's used in. This is mainly useful for some external I2C encoders known to be present on cards with GPUs from several different manufacturers. Signed-off-by: Francisco Jerez Signed-off-by: Dave Airlie --- drivers/gpu/drm/Makefile | 2 +- drivers/gpu/drm/drm_encoder_slave.c | 116 ++++++++++++++++++++++++++ include/drm/drm_encoder_slave.h | 162 ++++++++++++++++++++++++++++++++++++ 3 files changed, 279 insertions(+), 1 deletion(-) create mode 100644 drivers/gpu/drm/drm_encoder_slave.c create mode 100644 include/drm/drm_encoder_slave.h (limited to 'include') diff --git a/drivers/gpu/drm/Makefile b/drivers/gpu/drm/Makefile index fe23f29f7cba..5f0aec4f082a 100644 --- a/drivers/gpu/drm/Makefile +++ b/drivers/gpu/drm/Makefile @@ -11,7 +11,7 @@ drm-y := drm_auth.o drm_bufs.o drm_cache.o \ drm_agpsupport.o drm_scatter.o ati_pcigart.o drm_pci.o \ drm_sysfs.o drm_hashtab.o drm_sman.o drm_mm.o \ drm_crtc.o drm_crtc_helper.o drm_modes.o drm_edid.o \ - drm_info.o drm_debugfs.o + drm_info.o drm_debugfs.o drm_encoder_slave.o drm-$(CONFIG_COMPAT) += drm_ioc32.o diff --git a/drivers/gpu/drm/drm_encoder_slave.c b/drivers/gpu/drm/drm_encoder_slave.c new file mode 100644 index 000000000000..6ffd600ccfae --- /dev/null +++ b/drivers/gpu/drm/drm_encoder_slave.c @@ -0,0 +1,116 @@ +/* + * Copyright (C) 2009 Francisco Jerez. + * All Rights Reserved. + * + * Permission is hereby granted, free of charge, to any person obtaining + * a copy of this software and associated documentation files (the + * "Software"), to deal in the Software without restriction, including + * without limitation the rights to use, copy, modify, merge, publish, + * distribute, sublicense, and/or sell copies of the Software, and to + * permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice (including the + * next paragraph) shall be included in all copies or substantial + * portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. + * IN NO EVENT SHALL THE COPYRIGHT OWNER(S) AND/OR ITS SUPPLIERS BE + * LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION + * OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION + * WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. + * + */ + +#include + +/** + * drm_i2c_encoder_init - Initialize an I2C slave encoder + * @dev: DRM device. + * @encoder: Encoder to be attached to the I2C device. You aren't + * required to have called drm_encoder_init() before. + * @adap: I2C adapter that will be used to communicate with + * the device. + * @info: Information that will be used to create the I2C device. + * Required fields are @addr and @type. + * + * Create an I2C device on the specified bus (the module containing its + * driver is transparently loaded) and attach it to the specified + * &drm_encoder_slave. The @slave_funcs field will be initialized with + * the hooks provided by the slave driver. + * + * Returns 0 on success or a negative errno on failure, in particular, + * -ENODEV is returned when no matching driver is found. + */ +int drm_i2c_encoder_init(struct drm_device *dev, + struct drm_encoder_slave *encoder, + struct i2c_adapter *adap, + const struct i2c_board_info *info) +{ + char modalias[sizeof(I2C_MODULE_PREFIX) + + I2C_NAME_SIZE]; + struct module *module = NULL; + struct i2c_client *client; + struct drm_i2c_encoder_driver *encoder_drv; + int err = 0; + + snprintf(modalias, sizeof(modalias), + "%s%s", I2C_MODULE_PREFIX, info->type); + request_module(modalias); + + client = i2c_new_device(adap, info); + if (!client) { + err = -ENOMEM; + goto fail; + } + + if (!client->driver) { + err = -ENODEV; + goto fail_unregister; + } + + module = client->driver->driver.owner; + if (!try_module_get(module)) { + err = -ENODEV; + goto fail_unregister; + } + + encoder->bus_priv = client; + + encoder_drv = to_drm_i2c_encoder_driver(client->driver); + + err = encoder_drv->encoder_init(client, dev, encoder); + if (err) + goto fail_unregister; + + return 0; + +fail_unregister: + i2c_unregister_device(client); + module_put(module); +fail: + return err; +} +EXPORT_SYMBOL(drm_i2c_encoder_init); + +/** + * drm_i2c_encoder_destroy - Unregister the I2C device backing an encoder + * @drm_encoder: Encoder to be unregistered. + * + * This should be called from the @destroy method of an I2C slave + * encoder driver once I2C access is no longer needed. + */ +void drm_i2c_encoder_destroy(struct drm_encoder *drm_encoder) +{ + struct drm_encoder_slave *encoder = to_encoder_slave(drm_encoder); + struct i2c_client *client = drm_i2c_encoder_get_client(drm_encoder); + struct module *module = client->driver->driver.owner; + + i2c_unregister_device(client); + encoder->bus_priv = NULL; + + module_put(module); +} +EXPORT_SYMBOL(drm_i2c_encoder_destroy); diff --git a/include/drm/drm_encoder_slave.h b/include/drm/drm_encoder_slave.h new file mode 100644 index 000000000000..821ec40c17d8 --- /dev/null +++ b/include/drm/drm_encoder_slave.h @@ -0,0 +1,162 @@ +/* + * Copyright (C) 2009 Francisco Jerez. + * All Rights Reserved. + * + * Permission is hereby granted, free of charge, to any person obtaining + * a copy of this software and associated documentation files (the + * "Software"), to deal in the Software without restriction, including + * without limitation the rights to use, copy, modify, merge, publish, + * distribute, sublicense, and/or sell copies of the Software, and to + * permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice (including the + * next paragraph) shall be included in all copies or substantial + * portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. + * IN NO EVENT SHALL THE COPYRIGHT OWNER(S) AND/OR ITS SUPPLIERS BE + * LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION + * OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION + * WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. + * + */ + +#ifndef __DRM_ENCODER_SLAVE_H__ +#define __DRM_ENCODER_SLAVE_H__ + +#include +#include + +/** + * struct drm_encoder_slave_funcs - Entry points exposed by a slave encoder driver + * @set_config: Initialize any encoder-specific modesetting parameters. + * The meaning of the @params parameter is implementation + * dependent. It will usually be a structure with DVO port + * data format settings or timings. It's not required for + * the new parameters to take effect until the next mode + * is set. + * + * Most of its members are analogous to the function pointers in + * &drm_encoder_helper_funcs and they can optionally be used to + * initialize the latter. Connector-like methods (e.g. @get_modes and + * @set_property) will typically be wrapped around and only be called + * if the encoder is the currently selected one for the connector. + */ +struct drm_encoder_slave_funcs { + void (*set_config)(struct drm_encoder *encoder, + void *params); + + void (*destroy)(struct drm_encoder *encoder); + void (*dpms)(struct drm_encoder *encoder, int mode); + void (*save)(struct drm_encoder *encoder); + void (*restore)(struct drm_encoder *encoder); + bool (*mode_fixup)(struct drm_encoder *encoder, + struct drm_display_mode *mode, + struct drm_display_mode *adjusted_mode); + int (*mode_valid)(struct drm_encoder *encoder, + struct drm_display_mode *mode); + void (*mode_set)(struct drm_encoder *encoder, + struct drm_display_mode *mode, + struct drm_display_mode *adjusted_mode); + + enum drm_connector_status (*detect)(struct drm_encoder *encoder, + struct drm_connector *connector); + int (*get_modes)(struct drm_encoder *encoder, + struct drm_connector *connector); + int (*create_resources)(struct drm_encoder *encoder, + struct drm_connector *connector); + int (*set_property)(struct drm_encoder *encoder, + struct drm_connector *connector, + struct drm_property *property, + uint64_t val); + +}; + +/** + * struct drm_encoder_slave - Slave encoder struct + * @base: DRM encoder object. + * @slave_funcs: Slave encoder callbacks. + * @slave_priv: Slave encoder private data. + * @bus_priv: Bus specific data. + * + * A &drm_encoder_slave has two sets of callbacks, @slave_funcs and the + * ones in @base. The former are never actually called by the common + * CRTC code, it's just a convenience for splitting the encoder + * functions in an upper, GPU-specific layer and a (hopefully) + * GPU-agnostic lower layer: It's the GPU driver responsibility to + * call the slave methods when appropriate. + * + * drm_i2c_encoder_init() provides a way to get an implementation of + * this. + */ +struct drm_encoder_slave { + struct drm_encoder base; + + struct drm_encoder_slave_funcs *slave_funcs; + void *slave_priv; + void *bus_priv; +}; +#define to_encoder_slave(x) container_of((x), struct drm_encoder_slave, base) + +int drm_i2c_encoder_init(struct drm_device *dev, + struct drm_encoder_slave *encoder, + struct i2c_adapter *adap, + const struct i2c_board_info *info); + + +/** + * struct drm_i2c_encoder_driver + * + * Describes a device driver for an encoder connected to the GPU + * through an I2C bus. In addition to the entry points in @i2c_driver + * an @encoder_init function should be provided. It will be called to + * give the driver an opportunity to allocate any per-encoder data + * structures and to initialize the @slave_funcs and (optionally) + * @slave_priv members of @encoder. + */ +struct drm_i2c_encoder_driver { + struct i2c_driver i2c_driver; + + int (*encoder_init)(struct i2c_client *client, + struct drm_device *dev, + struct drm_encoder_slave *encoder); + +}; +#define to_drm_i2c_encoder_driver(x) container_of((x), \ + struct drm_i2c_encoder_driver, \ + i2c_driver) + +/** + * drm_i2c_encoder_get_client - Get the I2C client corresponding to an encoder + */ +static inline struct i2c_client *drm_i2c_encoder_get_client(struct drm_encoder *encoder) +{ + return (struct i2c_client *)to_encoder_slave(encoder)->bus_priv; +} + +/** + * drm_i2c_encoder_register - Register an I2C encoder driver + * @owner: Module containing the driver. + * @driver: Driver to be registered. + */ +static inline int drm_i2c_encoder_register(struct module *owner, + struct drm_i2c_encoder_driver *driver) +{ + return i2c_register_driver(owner, &driver->i2c_driver); +} + +/** + * drm_i2c_encoder_unregister - Unregister an I2C encoder driver + * @driver: Driver to be unregistered. + */ +static inline void drm_i2c_encoder_unregister(struct drm_i2c_encoder_driver *driver) +{ + return i2c_del_driver(&driver->i2c_driver); +} + +void drm_i2c_encoder_destroy(struct drm_encoder *encoder); + +#endif -- cgit v1.2.3 From 74bd3c26b90f39b9dcc05c471333da8998572b5d Mon Sep 17 00:00:00 2001 From: Francisco Jerez Date: Sun, 2 Aug 2009 04:19:18 +0200 Subject: drm: Define DRM_MODE_CONNECTOR_TV The existing TV connector types are often unsuitable either because there is no way to probe them until they're actually plugged in or because they can change during run time (e.g. 7-pin DIN connectors that behave as S-Video, Component, Composite or SCART depending on the adaptor plugged in). Signed-off-by: Francisco Jerez Signed-off-by: Dave Airlie --- drivers/gpu/drm/drm_crtc.c | 1 + drivers/gpu/drm/drm_sysfs.c | 3 +++ include/drm/drm_mode.h | 1 + 3 files changed, 5 insertions(+) (limited to 'include') diff --git a/drivers/gpu/drm/drm_crtc.c b/drivers/gpu/drm/drm_crtc.c index 9c758305472c..c7ab80b45e3f 100644 --- a/drivers/gpu/drm/drm_crtc.c +++ b/drivers/gpu/drm/drm_crtc.c @@ -146,6 +146,7 @@ static struct drm_conn_prop_enum_list drm_connector_enum_list[] = { DRM_MODE_CONNECTOR_DisplayPort, "DisplayPort", 0 }, { DRM_MODE_CONNECTOR_HDMIA, "HDMI Type A", 0 }, { DRM_MODE_CONNECTOR_HDMIB, "HDMI Type B", 0 }, + { DRM_MODE_CONNECTOR_TV, "TV", 0 }, }; static struct drm_prop_enum_list drm_encoder_enum_list[] = diff --git a/drivers/gpu/drm/drm_sysfs.c b/drivers/gpu/drm/drm_sysfs.c index 85ec31b3ff00..adc179459c25 100644 --- a/drivers/gpu/drm/drm_sysfs.c +++ b/drivers/gpu/drm/drm_sysfs.c @@ -247,6 +247,7 @@ static ssize_t subconnector_show(struct device *device, case DRM_MODE_CONNECTOR_Composite: case DRM_MODE_CONNECTOR_SVIDEO: case DRM_MODE_CONNECTOR_Component: + case DRM_MODE_CONNECTOR_TV: prop = dev->mode_config.tv_subconnector_property; is_tv = 1; break; @@ -287,6 +288,7 @@ static ssize_t select_subconnector_show(struct device *device, case DRM_MODE_CONNECTOR_Composite: case DRM_MODE_CONNECTOR_SVIDEO: case DRM_MODE_CONNECTOR_Component: + case DRM_MODE_CONNECTOR_TV: prop = dev->mode_config.tv_select_subconnector_property; is_tv = 1; break; @@ -385,6 +387,7 @@ int drm_sysfs_connector_add(struct drm_connector *connector) case DRM_MODE_CONNECTOR_Composite: case DRM_MODE_CONNECTOR_SVIDEO: case DRM_MODE_CONNECTOR_Component: + case DRM_MODE_CONNECTOR_TV: for (i = 0; i < ARRAY_SIZE(connector_attrs_opt1); i++) { ret = device_create_file(&connector->kdev, &connector_attrs_opt1[i]); if (ret) diff --git a/include/drm/drm_mode.h b/include/drm/drm_mode.h index ae304cc73c90..c51e9f528c8f 100644 --- a/include/drm/drm_mode.h +++ b/include/drm/drm_mode.h @@ -155,6 +155,7 @@ struct drm_mode_get_encoder { #define DRM_MODE_CONNECTOR_DisplayPort 10 #define DRM_MODE_CONNECTOR_HDMIA 11 #define DRM_MODE_CONNECTOR_HDMIB 12 +#define DRM_MODE_CONNECTOR_TV 13 struct drm_mode_get_connector { -- cgit v1.2.3 From aeaa1ad3ff32be833680e484d99ec29d892da1ff Mon Sep 17 00:00:00 2001 From: Francisco Jerez Date: Sun, 2 Aug 2009 04:19:19 +0200 Subject: drm: Define DRM_MODE_SUBCONNECTOR_SCART Signed-off-by: Francisco Jerez Signed-off-by: Dave Airlie --- drivers/gpu/drm/drm_crtc.c | 2 ++ include/drm/drm_mode.h | 1 + 2 files changed, 3 insertions(+) (limited to 'include') diff --git a/drivers/gpu/drm/drm_crtc.c b/drivers/gpu/drm/drm_crtc.c index c7ab80b45e3f..ed53c5c37ac4 100644 --- a/drivers/gpu/drm/drm_crtc.c +++ b/drivers/gpu/drm/drm_crtc.c @@ -108,6 +108,7 @@ static struct drm_prop_enum_list drm_tv_select_enum_list[] = { DRM_MODE_SUBCONNECTOR_Composite, "Composite" }, /* TV-out */ { DRM_MODE_SUBCONNECTOR_SVIDEO, "SVIDEO" }, /* TV-out */ { DRM_MODE_SUBCONNECTOR_Component, "Component" }, /* TV-out */ + { DRM_MODE_SUBCONNECTOR_SCART, "SCART" }, /* TV-out */ }; DRM_ENUM_NAME_FN(drm_get_tv_select_name, drm_tv_select_enum_list) @@ -118,6 +119,7 @@ static struct drm_prop_enum_list drm_tv_subconnector_enum_list[] = { DRM_MODE_SUBCONNECTOR_Composite, "Composite" }, /* TV-out */ { DRM_MODE_SUBCONNECTOR_SVIDEO, "SVIDEO" }, /* TV-out */ { DRM_MODE_SUBCONNECTOR_Component, "Component" }, /* TV-out */ + { DRM_MODE_SUBCONNECTOR_SCART, "SCART" }, /* TV-out */ }; DRM_ENUM_NAME_FN(drm_get_tv_subconnector_name, diff --git a/include/drm/drm_mode.h b/include/drm/drm_mode.h index c51e9f528c8f..616aeb42b773 100644 --- a/include/drm/drm_mode.h +++ b/include/drm/drm_mode.h @@ -141,6 +141,7 @@ struct drm_mode_get_encoder { #define DRM_MODE_SUBCONNECTOR_Composite 5 #define DRM_MODE_SUBCONNECTOR_SVIDEO 6 #define DRM_MODE_SUBCONNECTOR_Component 8 +#define DRM_MODE_SUBCONNECTOR_SCART 9 #define DRM_MODE_CONNECTOR_Unknown 0 #define DRM_MODE_CONNECTOR_VGA 1 -- cgit v1.2.3 From b6b7902e54c7e8abbc213d8bdc290350c00ccfe5 Mon Sep 17 00:00:00 2001 From: Francisco Jerez Date: Sun, 2 Aug 2009 04:19:20 +0200 Subject: drm: Define some new standard TV properties. Namely "brightness", "contrast" and "flicker reduction". Signed-off-by: Francisco Jerez Signed-off-by: Dave Airlie --- drivers/gpu/drm/drm_crtc.c | 18 ++++++++++++++++++ include/drm/drm_crtc.h | 3 +++ 2 files changed, 21 insertions(+) (limited to 'include') diff --git a/drivers/gpu/drm/drm_crtc.c b/drivers/gpu/drm/drm_crtc.c index ed53c5c37ac4..a8c831134fc3 100644 --- a/drivers/gpu/drm/drm_crtc.c +++ b/drivers/gpu/drm/drm_crtc.c @@ -718,6 +718,24 @@ int drm_mode_create_tv_properties(struct drm_device *dev, int num_modes, drm_property_add_enum(dev->mode_config.tv_mode_property, i, i, modes[i]); + dev->mode_config.tv_brightness_property = + drm_property_create(dev, DRM_MODE_PROP_RANGE, + "brightness", 2); + dev->mode_config.tv_brightness_property->values[0] = 0; + dev->mode_config.tv_brightness_property->values[1] = 100; + + dev->mode_config.tv_contrast_property = + drm_property_create(dev, DRM_MODE_PROP_RANGE, + "contrast", 2); + dev->mode_config.tv_contrast_property->values[0] = 0; + dev->mode_config.tv_contrast_property->values[1] = 100; + + dev->mode_config.tv_flicker_reduction_property = + drm_property_create(dev, DRM_MODE_PROP_RANGE, + "flicker reduction", 2); + dev->mode_config.tv_flicker_reduction_property->values[0] = 0; + dev->mode_config.tv_flicker_reduction_property->values[1] = 100; + return 0; } EXPORT_SYMBOL(drm_mode_create_tv_properties); diff --git a/include/drm/drm_crtc.h b/include/drm/drm_crtc.h index 125994d8ac0b..5f2cc0ca4c7d 100644 --- a/include/drm/drm_crtc.h +++ b/include/drm/drm_crtc.h @@ -572,6 +572,9 @@ struct drm_mode_config { struct drm_property *tv_right_margin_property; struct drm_property *tv_top_margin_property; struct drm_property *tv_bottom_margin_property; + struct drm_property *tv_brightness_property; + struct drm_property *tv_contrast_property; + struct drm_property *tv_flicker_reduction_property; /* Optional properties */ struct drm_property *scaling_mode_property; -- cgit v1.2.3 From 42935ecaf4e784d0815afa9a7e5fe7e141157ca3 Mon Sep 17 00:00:00 2001 From: "Luis R. Rodriguez" Date: Wed, 29 Jul 2009 20:08:07 -0400 Subject: mac80211: redefine usage of the mac80211 workqueue The mac80211 workqueue exists to enable mac80211 and drivers to queue their own work on a single threaded workqueue. mac80211 takes care to flush the workqueue during suspend but we never really had requirements on drivers for how they should use the workqueue in consideration for suspend. We extend mac80211 to document how the mac80211 workqueue should be used, how it should not be used and finally move raw access to the workqueue to mac80211 only. Drivers and mac80211 use helpers to queue work onto the mac80211 workqueue: * ieee80211_queue_work() * ieee80211_queue_delayed_work() These helpers will now warn if mac80211 already completed its suspend cycle and someone is trying to queue work. mac80211 flushes the mac80211 workqueue prior to suspend a few times, but we haven't taken the care to ensure drivers won't add more work after suspend. To help with this we add a warning when someone tries to add work and mac80211 already completed the suspend cycle. Drivers should ensure they cancel any work or delayed work in the mac80211 stop() callback. Signed-off-by: Luis R. Rodriguez Signed-off-by: John W. Linville --- drivers/net/wireless/at76c50x-usb.c | 10 +++--- drivers/net/wireless/ath/ar9170/led.c | 11 ++++--- drivers/net/wireless/ath/ar9170/main.c | 26 ++++++++------- drivers/net/wireless/ath/ath9k/main.c | 15 +++++---- drivers/net/wireless/ath/ath9k/virtual.c | 17 +++++----- drivers/net/wireless/ath/ath9k/xmit.c | 2 +- drivers/net/wireless/b43/main.c | 8 ++--- drivers/net/wireless/b43/phy_common.c | 2 +- drivers/net/wireless/b43/pio.c | 2 +- drivers/net/wireless/b43legacy/main.c | 8 ++--- drivers/net/wireless/p54/led.c | 5 ++- drivers/net/wireless/p54/main.c | 2 +- drivers/net/wireless/p54/p54spi.c | 4 +-- drivers/net/wireless/p54/txrx.c | 2 +- drivers/net/wireless/rt2x00/rt2x00dev.c | 2 +- drivers/net/wireless/rt2x00/rt2x00link.c | 8 ++--- drivers/net/wireless/rt2x00/rt2x00mac.c | 2 +- drivers/net/wireless/rtl818x/rtl8187_dev.c | 2 +- drivers/net/wireless/rtl818x/rtl8187_leds.c | 10 +++--- drivers/net/wireless/zd1211rw/zd_mac.c | 2 +- include/net/mac80211.h | 50 ++++++++++++++++++++++++----- net/mac80211/ibss.c | 6 ++-- net/mac80211/ieee80211_i.h | 6 ++++ net/mac80211/iface.c | 4 +-- net/mac80211/main.c | 8 ++--- net/mac80211/mesh.c | 10 +++--- net/mac80211/mesh_hwmp.c | 4 +-- net/mac80211/mlme.c | 48 +++++++++++++-------------- net/mac80211/pm.c | 4 +-- net/mac80211/scan.c | 8 ++--- net/mac80211/tx.c | 2 +- net/mac80211/util.c | 41 +++++++++++++++++++++++ 32 files changed, 208 insertions(+), 123 deletions(-) (limited to 'include') diff --git a/drivers/net/wireless/at76c50x-usb.c b/drivers/net/wireless/at76c50x-usb.c index aff09a1cf64f..7218dbabad3e 100644 --- a/drivers/net/wireless/at76c50x-usb.c +++ b/drivers/net/wireless/at76c50x-usb.c @@ -1875,8 +1875,8 @@ static void at76_dwork_hw_scan(struct work_struct *work) /* FIXME: add maximum time for scan to complete */ if (ret != CMD_STATUS_COMPLETE) { - queue_delayed_work(priv->hw->workqueue, &priv->dwork_hw_scan, - SCAN_POLL_INTERVAL); + ieee80211_queue_delayed_work(priv->hw, &priv->dwork_hw_scan, + SCAN_POLL_INTERVAL); mutex_unlock(&priv->mtx); return; } @@ -1937,8 +1937,8 @@ static int at76_hw_scan(struct ieee80211_hw *hw, goto exit; } - queue_delayed_work(priv->hw->workqueue, &priv->dwork_hw_scan, - SCAN_POLL_INTERVAL); + ieee80211_queue_delayed_work(priv->hw, &priv->dwork_hw_scan, + SCAN_POLL_INTERVAL); exit: mutex_unlock(&priv->mtx); @@ -2027,7 +2027,7 @@ static void at76_configure_filter(struct ieee80211_hw *hw, } else return; - queue_work(hw->workqueue, &priv->work_set_promisc); + ieee80211_queue_work(hw, &priv->work_set_promisc); } static int at76_set_key(struct ieee80211_hw *hw, enum set_key_cmd cmd, diff --git a/drivers/net/wireless/ath/ar9170/led.c b/drivers/net/wireless/ath/ar9170/led.c index 63fda6cd2101..86c4e79f6bc8 100644 --- a/drivers/net/wireless/ath/ar9170/led.c +++ b/drivers/net/wireless/ath/ar9170/led.c @@ -90,9 +90,12 @@ static void ar9170_update_leds(struct work_struct *work) ar9170_set_leds_state(ar, led_val); mutex_unlock(&ar->mutex); - if (rerun) - queue_delayed_work(ar->hw->workqueue, &ar->led_work, - msecs_to_jiffies(blink_delay)); + if (!rerun) + return; + + ieee80211_queue_delayed_work(ar->hw, + &ar->led_work, + msecs_to_jiffies(blink_delay)); } static void ar9170_led_brightness_set(struct led_classdev *led, @@ -110,7 +113,7 @@ static void ar9170_led_brightness_set(struct led_classdev *led, } if (likely(IS_ACCEPTING_CMD(ar) && arl->toggled)) - queue_delayed_work(ar->hw->workqueue, &ar->led_work, HZ/10); + ieee80211_queue_delayed_work(ar->hw, &ar->led_work, HZ/10); } static int ar9170_register_led(struct ar9170 *ar, int i, char *name, diff --git a/drivers/net/wireless/ath/ar9170/main.c b/drivers/net/wireless/ath/ar9170/main.c index 099ed3c3ba28..4fc389ae74b4 100644 --- a/drivers/net/wireless/ath/ar9170/main.c +++ b/drivers/net/wireless/ath/ar9170/main.c @@ -595,10 +595,12 @@ static void ar9170_tx_janitor(struct work_struct *work) ar9170_tx_fake_ampdu_status(ar); - if (resched) - queue_delayed_work(ar->hw->workqueue, - &ar->tx_janitor, - msecs_to_jiffies(AR9170_JANITOR_DELAY)); + if (!resched) + return; + + ieee80211_queue_delayed_work(ar->hw, + &ar->tx_janitor, + msecs_to_jiffies(AR9170_JANITOR_DELAY)); } void ar9170_handle_command_response(struct ar9170 *ar, void *buf, u32 len) @@ -648,7 +650,7 @@ void ar9170_handle_command_response(struct ar9170 *ar, void *buf, u32 len) * pre-TBTT event */ if (ar->vif && ar->vif->type == NL80211_IFTYPE_AP) - queue_work(ar->hw->workqueue, &ar->beacon_work); + ieee80211_queue_work(ar->hw, &ar->beacon_work); break; case 0xc2: @@ -1825,10 +1827,12 @@ static void ar9170_tx(struct ar9170 *ar) } } - if (schedule_garbagecollector) - queue_delayed_work(ar->hw->workqueue, - &ar->tx_janitor, - msecs_to_jiffies(AR9170_JANITOR_DELAY)); + if (!schedule_garbagecollector) + return; + + ieee80211_queue_delayed_work(ar->hw, + &ar->tx_janitor, + msecs_to_jiffies(AR9170_JANITOR_DELAY)); } static bool ar9170_tx_ampdu_queue(struct ar9170 *ar, struct sk_buff *skb) @@ -2157,7 +2161,7 @@ static void ar9170_op_configure_filter(struct ieee80211_hw *hw, } if (likely(IS_STARTED(ar))) - queue_work(ar->hw->workqueue, &ar->filter_config_work); + ieee80211_queue_work(ar->hw, &ar->filter_config_work); } static void ar9170_op_bss_info_changed(struct ieee80211_hw *hw, @@ -2415,7 +2419,7 @@ static void ar9170_sta_notify(struct ieee80211_hw *hw, } if (IS_STARTED(ar) && ar->filter_changed) - queue_work(ar->hw->workqueue, &ar->filter_config_work); + ieee80211_queue_work(ar->hw, &ar->filter_config_work); } static int ar9170_get_stats(struct ieee80211_hw *hw, diff --git a/drivers/net/wireless/ath/ath9k/main.c b/drivers/net/wireless/ath/ath9k/main.c index cf44623b5cd2..292ac2b41891 100644 --- a/drivers/net/wireless/ath/ath9k/main.c +++ b/drivers/net/wireless/ath/ath9k/main.c @@ -973,10 +973,11 @@ static void ath_led_blink_work(struct work_struct *work) ath9k_hw_set_gpio(sc->sc_ah, ATH_LED_PIN, (sc->sc_flags & SC_OP_LED_ON) ? 1 : 0); - queue_delayed_work(sc->hw->workqueue, &sc->ath_led_blink_work, - (sc->sc_flags & SC_OP_LED_ON) ? - msecs_to_jiffies(sc->led_off_duration) : - msecs_to_jiffies(sc->led_on_duration)); + ieee80211_queue_delayed_work(sc->hw, + &sc->ath_led_blink_work, + (sc->sc_flags & SC_OP_LED_ON) ? + msecs_to_jiffies(sc->led_off_duration) : + msecs_to_jiffies(sc->led_on_duration)); sc->led_on_duration = sc->led_on_cnt ? max((ATH_LED_ON_DURATION_IDLE - sc->led_on_cnt), 25) : @@ -1013,8 +1014,8 @@ static void ath_led_brightness(struct led_classdev *led_cdev, case LED_FULL: if (led->led_type == ATH_LED_ASSOC) { sc->sc_flags |= SC_OP_LED_ASSOCIATED; - queue_delayed_work(sc->hw->workqueue, - &sc->ath_led_blink_work, 0); + ieee80211_queue_delayed_work(sc->hw, + &sc->ath_led_blink_work, 0); } else if (led->led_type == ATH_LED_RADIO) { ath9k_hw_set_gpio(sc->sc_ah, ATH_LED_PIN, 0); sc->sc_flags |= SC_OP_LED_ON; @@ -1972,7 +1973,7 @@ static int ath9k_start(struct ieee80211_hw *hw) ieee80211_wake_queues(hw); - queue_delayed_work(sc->hw->workqueue, &sc->tx_complete_work, 0); + ieee80211_queue_delayed_work(sc->hw, &sc->tx_complete_work, 0); mutex_unlock: mutex_unlock(&sc->mutex); diff --git a/drivers/net/wireless/ath/ath9k/virtual.c b/drivers/net/wireless/ath/ath9k/virtual.c index e1d419e02b4a..19b88f8177fd 100644 --- a/drivers/net/wireless/ath/ath9k/virtual.c +++ b/drivers/net/wireless/ath/ath9k/virtual.c @@ -351,7 +351,7 @@ void ath9k_tx_status(struct ieee80211_hw *hw, struct sk_buff *skb) * Drop from tasklet to work to allow mutex for channel * change. */ - queue_work(aphy->sc->hw->workqueue, + ieee80211_queue_work(aphy->sc->hw, &aphy->sc->chan_work); } } @@ -367,7 +367,7 @@ static void ath9k_mark_paused(struct ath_wiphy *aphy) struct ath_softc *sc = aphy->sc; aphy->state = ATH_WIPHY_PAUSED; if (!__ath9k_wiphy_pausing(sc)) - queue_work(sc->hw->workqueue, &sc->chan_work); + ieee80211_queue_work(sc->hw, &sc->chan_work); } static void ath9k_pause_iter(void *data, u8 *mac, struct ieee80211_vif *vif) @@ -521,7 +521,7 @@ int ath9k_wiphy_select(struct ath_wiphy *aphy) spin_unlock_bh(&sc->wiphy_lock); ath_radio_disable(sc); ath_radio_enable(sc); - queue_work(aphy->sc->hw->workqueue, + ieee80211_queue_work(aphy->sc->hw, &aphy->sc->chan_work); return -EBUSY; /* previous select still in progress */ } @@ -541,7 +541,7 @@ int ath9k_wiphy_select(struct ath_wiphy *aphy) if (now) { /* Ready to request channel change immediately */ - queue_work(aphy->sc->hw->workqueue, &aphy->sc->chan_work); + ieee80211_queue_work(aphy->sc->hw, &aphy->sc->chan_work); } /* @@ -648,8 +648,9 @@ try_again: "change\n"); } - queue_delayed_work(sc->hw->workqueue, &sc->wiphy_work, - sc->wiphy_scheduler_int); + ieee80211_queue_delayed_work(sc->hw, + &sc->wiphy_work, + sc->wiphy_scheduler_int); } void ath9k_wiphy_set_scheduler(struct ath_softc *sc, unsigned int msec_int) @@ -657,8 +658,8 @@ void ath9k_wiphy_set_scheduler(struct ath_softc *sc, unsigned int msec_int) cancel_delayed_work_sync(&sc->wiphy_work); sc->wiphy_scheduler_int = msecs_to_jiffies(msec_int); if (sc->wiphy_scheduler_int) - queue_delayed_work(sc->hw->workqueue, &sc->wiphy_work, - sc->wiphy_scheduler_int); + ieee80211_queue_delayed_work(sc->hw, &sc->wiphy_work, + sc->wiphy_scheduler_int); } /* caller must hold wiphy_lock */ diff --git a/drivers/net/wireless/ath/ath9k/xmit.c b/drivers/net/wireless/ath/ath9k/xmit.c index b7806e2ca0e1..87762da0383b 100644 --- a/drivers/net/wireless/ath/ath9k/xmit.c +++ b/drivers/net/wireless/ath/ath9k/xmit.c @@ -2063,7 +2063,7 @@ static void ath_tx_complete_poll_work(struct work_struct *work) ath_reset(sc, false); } - queue_delayed_work(sc->hw->workqueue, &sc->tx_complete_work, + ieee80211_queue_delayed_work(sc->hw, &sc->tx_complete_work, msecs_to_jiffies(ATH_TX_COMPLETE_POLL_INT)); } diff --git a/drivers/net/wireless/b43/main.c b/drivers/net/wireless/b43/main.c index 3f4360ad0e4e..f985938962e3 100644 --- a/drivers/net/wireless/b43/main.c +++ b/drivers/net/wireless/b43/main.c @@ -1654,7 +1654,7 @@ static void b43_update_templates(struct b43_wl *wl) wl->current_beacon = beacon; wl->beacon0_uploaded = 0; wl->beacon1_uploaded = 0; - queue_work(wl->hw->workqueue, &wl->beacon_update_trigger); + ieee80211_queue_work(wl->hw, &wl->beacon_update_trigger); } static void b43_set_beacon_int(struct b43_wldev *dev, u16 beacon_int) @@ -2914,7 +2914,7 @@ out_requeue: delay = msecs_to_jiffies(50); else delay = round_jiffies_relative(HZ * 15); - queue_delayed_work(wl->hw->workqueue, &dev->periodic_work, delay); + ieee80211_queue_delayed_work(wl->hw, &dev->periodic_work, delay); out: mutex_unlock(&wl->mutex); } @@ -2925,7 +2925,7 @@ static void b43_periodic_tasks_setup(struct b43_wldev *dev) dev->periodic_state = 0; INIT_DELAYED_WORK(work, b43_periodic_work_handler); - queue_delayed_work(dev->wl->hw->workqueue, work, 0); + ieee80211_queue_delayed_work(dev->wl->hw, work, 0); } /* Check if communication with the device works correctly. */ @@ -4871,7 +4871,7 @@ void b43_controller_restart(struct b43_wldev *dev, const char *reason) if (b43_status(dev) < B43_STAT_INITIALIZED) return; b43info(dev->wl, "Controller RESET (%s) ...\n", reason); - queue_work(dev->wl->hw->workqueue, &dev->restart_work); + ieee80211_queue_work(dev->wl->hw, &dev->restart_work); } #ifdef CONFIG_PM diff --git a/drivers/net/wireless/b43/phy_common.c b/drivers/net/wireless/b43/phy_common.c index 6d241622210e..f537bfef690a 100644 --- a/drivers/net/wireless/b43/phy_common.c +++ b/drivers/net/wireless/b43/phy_common.c @@ -352,7 +352,7 @@ void b43_phy_txpower_check(struct b43_wldev *dev, unsigned int flags) /* We must adjust the transmission power in hardware. * Schedule b43_phy_txpower_adjust_work(). */ - queue_work(dev->wl->hw->workqueue, &dev->wl->txpower_adjust_work); + ieee80211_queue_work(dev->wl->hw, &dev->wl->txpower_adjust_work); } int b43_phy_shm_tssi_read(struct b43_wldev *dev, u16 shm_offset) diff --git a/drivers/net/wireless/b43/pio.c b/drivers/net/wireless/b43/pio.c index 69138e8c1db6..73c047d8de40 100644 --- a/drivers/net/wireless/b43/pio.c +++ b/drivers/net/wireless/b43/pio.c @@ -783,7 +783,7 @@ void b43_pio_rx(struct b43_pio_rxqueue *q) { /* Due to latency issues we must run the RX path in * a workqueue to be able to schedule between packets. */ - queue_work(q->dev->wl->hw->workqueue, &q->rx_work); + ieee80211_queue_work(q->dev->wl->hw, &q->rx_work); } static void b43_pio_tx_suspend_queue(struct b43_pio_txqueue *q) diff --git a/drivers/net/wireless/b43legacy/main.c b/drivers/net/wireless/b43legacy/main.c index c4973c1942bf..b1435594921a 100644 --- a/drivers/net/wireless/b43legacy/main.c +++ b/drivers/net/wireless/b43legacy/main.c @@ -1252,7 +1252,7 @@ static void b43legacy_update_templates(struct b43legacy_wl *wl) wl->current_beacon = beacon; wl->beacon0_uploaded = 0; wl->beacon1_uploaded = 0; - queue_work(wl->hw->workqueue, &wl->beacon_update_trigger); + ieee80211_queue_work(wl->hw, &wl->beacon_update_trigger); } static void b43legacy_set_beacon_int(struct b43legacy_wldev *dev, @@ -2300,7 +2300,7 @@ out_requeue: delay = msecs_to_jiffies(50); else delay = round_jiffies_relative(HZ * 15); - queue_delayed_work(wl->hw->workqueue, &dev->periodic_work, delay); + ieee80211_queue_delayed_work(wl->hw, &dev->periodic_work, delay); out: mutex_unlock(&wl->mutex); } @@ -2311,7 +2311,7 @@ static void b43legacy_periodic_tasks_setup(struct b43legacy_wldev *dev) dev->periodic_state = 0; INIT_DELAYED_WORK(work, b43legacy_periodic_work_handler); - queue_delayed_work(dev->wl->hw->workqueue, work, 0); + ieee80211_queue_delayed_work(dev->wl->hw, work, 0); } /* Validate access to the chip (SHM) */ @@ -3885,7 +3885,7 @@ void b43legacy_controller_restart(struct b43legacy_wldev *dev, if (b43legacy_status(dev) < B43legacy_STAT_INITIALIZED) return; b43legacyinfo(dev->wl, "Controller RESET (%s) ...\n", reason); - queue_work(dev->wl->hw->workqueue, &dev->restart_work); + ieee80211_queue_work(dev->wl->hw, &dev->restart_work); } #ifdef CONFIG_PM diff --git a/drivers/net/wireless/p54/led.c b/drivers/net/wireless/p54/led.c index c00115b206d4..9575ac033630 100644 --- a/drivers/net/wireless/p54/led.c +++ b/drivers/net/wireless/p54/led.c @@ -61,7 +61,7 @@ static void p54_update_leds(struct work_struct *work) wiphy_name(priv->hw->wiphy), err); if (rerun) - queue_delayed_work(priv->hw->workqueue, &priv->led_work, + ieee80211_queue_delayed_work(priv->hw, &priv->led_work, msecs_to_jiffies(blink_delay)); } @@ -78,8 +78,7 @@ static void p54_led_brightness_set(struct led_classdev *led_dev, if ((brightness) && (led->registered)) { led->toggled++; - queue_delayed_work(priv->hw->workqueue, &priv->led_work, - HZ/10); + ieee80211_queue_delayed_work(priv->hw, &priv->led_work, HZ/10); } } diff --git a/drivers/net/wireless/p54/main.c b/drivers/net/wireless/p54/main.c index 955f6d7ec16a..a0d0e726bc4e 100644 --- a/drivers/net/wireless/p54/main.c +++ b/drivers/net/wireless/p54/main.c @@ -180,7 +180,7 @@ static int p54_start(struct ieee80211_hw *dev) goto out; } - queue_delayed_work(dev->workqueue, &priv->work, 0); + ieee80211_queue_delayed_work(dev, &priv->work, 0); priv->softled_state = 0; err = p54_set_leds(priv); diff --git a/drivers/net/wireless/p54/p54spi.c b/drivers/net/wireless/p54/p54spi.c index eef532987d05..05458d9249ce 100644 --- a/drivers/net/wireless/p54/p54spi.c +++ b/drivers/net/wireless/p54/p54spi.c @@ -391,7 +391,7 @@ static irqreturn_t p54spi_interrupt(int irq, void *config) struct spi_device *spi = config; struct p54s_priv *priv = dev_get_drvdata(&spi->dev); - queue_work(priv->hw->workqueue, &priv->work); + ieee80211_queue_work(priv->hw, &priv->work); return IRQ_HANDLED; } @@ -479,7 +479,7 @@ static void p54spi_op_tx(struct ieee80211_hw *dev, struct sk_buff *skb) list_add_tail(&di->tx_list, &priv->tx_pending); spin_unlock_irqrestore(&priv->tx_lock, flags); - queue_work(priv->hw->workqueue, &priv->work); + ieee80211_queue_work(priv->hw, &priv->work); } static void p54spi_work(struct work_struct *work) diff --git a/drivers/net/wireless/p54/txrx.c b/drivers/net/wireless/p54/txrx.c index c32a0d2fa1f7..704685fab177 100644 --- a/drivers/net/wireless/p54/txrx.c +++ b/drivers/net/wireless/p54/txrx.c @@ -380,7 +380,7 @@ static int p54_rx_data(struct p54_common *priv, struct sk_buff *skb) ieee80211_rx_irqsafe(priv->hw, skb); - queue_delayed_work(priv->hw->workqueue, &priv->work, + ieee80211_queue_delayed_work(priv->hw, &priv->work, msecs_to_jiffies(P54_STATISTICS_UPDATE)); return -1; diff --git a/drivers/net/wireless/rt2x00/rt2x00dev.c b/drivers/net/wireless/rt2x00/rt2x00dev.c index 658a63bfb761..b717afbf3f38 100644 --- a/drivers/net/wireless/rt2x00/rt2x00dev.c +++ b/drivers/net/wireless/rt2x00/rt2x00dev.c @@ -215,7 +215,7 @@ void rt2x00lib_beacondone(struct rt2x00_dev *rt2x00dev) rt2x00lib_beacondone_iter, rt2x00dev); - queue_work(rt2x00dev->hw->workqueue, &rt2x00dev->intf_work); + ieee80211_queue_work(rt2x00dev->hw, &rt2x00dev->intf_work); } EXPORT_SYMBOL_GPL(rt2x00lib_beacondone); diff --git a/drivers/net/wireless/rt2x00/rt2x00link.c b/drivers/net/wireless/rt2x00/rt2x00link.c index 79915687e744..917831689ccd 100644 --- a/drivers/net/wireless/rt2x00/rt2x00link.c +++ b/drivers/net/wireless/rt2x00/rt2x00link.c @@ -351,8 +351,8 @@ void rt2x00link_start_tuner(struct rt2x00_dev *rt2x00dev) rt2x00link_reset_tuner(rt2x00dev, false); - queue_delayed_work(rt2x00dev->hw->workqueue, - &link->work, LINK_TUNE_INTERVAL); + ieee80211_queue_delayed_work(rt2x00dev->hw, + &link->work, LINK_TUNE_INTERVAL); } void rt2x00link_stop_tuner(struct rt2x00_dev *rt2x00dev) @@ -461,8 +461,8 @@ static void rt2x00link_tuner(struct work_struct *work) * Increase tuner counter, and reschedule the next link tuner run. */ link->count++; - queue_delayed_work(rt2x00dev->hw->workqueue, - &link->work, LINK_TUNE_INTERVAL); + ieee80211_queue_delayed_work(rt2x00dev->hw, + &link->work, LINK_TUNE_INTERVAL); } void rt2x00link_register(struct rt2x00_dev *rt2x00dev) diff --git a/drivers/net/wireless/rt2x00/rt2x00mac.c b/drivers/net/wireless/rt2x00/rt2x00mac.c index e92c8f99d695..81febdfd6639 100644 --- a/drivers/net/wireless/rt2x00/rt2x00mac.c +++ b/drivers/net/wireless/rt2x00/rt2x00mac.c @@ -431,7 +431,7 @@ void rt2x00mac_configure_filter(struct ieee80211_hw *hw, if (!test_bit(DRIVER_REQUIRE_SCHEDULED, &rt2x00dev->flags)) rt2x00dev->ops->lib->config_filter(rt2x00dev, *total_flags); else - queue_work(rt2x00dev->hw->workqueue, &rt2x00dev->filter_work); + ieee80211_queue_work(rt2x00dev->hw, &rt2x00dev->filter_work); } EXPORT_SYMBOL_GPL(rt2x00mac_configure_filter); diff --git a/drivers/net/wireless/rtl818x/rtl8187_dev.c b/drivers/net/wireless/rtl818x/rtl8187_dev.c index c9b9dbe584c6..53f57dc52226 100644 --- a/drivers/net/wireless/rtl818x/rtl8187_dev.c +++ b/drivers/net/wireless/rtl818x/rtl8187_dev.c @@ -220,7 +220,7 @@ static void rtl8187_tx_cb(struct urb *urb) * reading a register in the device. We are in interrupt mode * here, thus queue the skb and finish on a work queue. */ skb_queue_tail(&priv->b_tx_status.queue, skb); - queue_delayed_work(hw->workqueue, &priv->work, 0); + ieee80211_queue_delayed_work(hw, &priv->work, 0); } } diff --git a/drivers/net/wireless/rtl818x/rtl8187_leds.c b/drivers/net/wireless/rtl818x/rtl8187_leds.c index cf9f899fe0e6..a6cfb7e77994 100644 --- a/drivers/net/wireless/rtl818x/rtl8187_leds.c +++ b/drivers/net/wireless/rtl818x/rtl8187_leds.c @@ -108,11 +108,11 @@ static void rtl8187_led_brightness_set(struct led_classdev *led_dev, struct rtl8187_priv *priv = hw->priv; if (brightness == LED_OFF) { - queue_delayed_work(hw->workqueue, &priv->led_off, 0); + ieee80211_queue_delayed_work(hw, &priv->led_off, 0); /* The LED is off for 1/20 sec so that it just blinks. */ - queue_delayed_work(hw->workqueue, &priv->led_on, HZ / 20); + ieee80211_queue_delayed_work(hw, &priv->led_on, HZ / 20); } else - queue_delayed_work(hw->workqueue, &priv->led_on, 0); + ieee80211_queue_delayed_work(hw, &priv->led_on, 0); } static int rtl8187_register_led(struct ieee80211_hw *dev, @@ -193,7 +193,7 @@ void rtl8187_leds_init(struct ieee80211_hw *dev, u16 custid) err = rtl8187_register_led(dev, &priv->led_rx, name, ieee80211_get_rx_led_name(dev), ledpin); if (!err) { - queue_delayed_work(dev->workqueue, &priv->led_on, 0); + ieee80211_queue_delayed_work(dev, &priv->led_on, 0); return; } /* registration of RX LED failed - unregister TX */ @@ -209,7 +209,7 @@ void rtl8187_leds_exit(struct ieee80211_hw *dev) struct rtl8187_priv *priv = dev->priv; /* turn the LED off before exiting */ - queue_delayed_work(dev->workqueue, &priv->led_off, 0); + ieee80211_queue_delayed_work(dev, &priv->led_off, 0); cancel_delayed_work_sync(&priv->led_off); cancel_delayed_work_sync(&priv->led_on); rtl8187_unregister_led(&priv->led_rx); diff --git a/drivers/net/wireless/zd1211rw/zd_mac.c b/drivers/net/wireless/zd1211rw/zd_mac.c index 9600b72495da..54abdd0c0045 100644 --- a/drivers/net/wireless/zd1211rw/zd_mac.c +++ b/drivers/net/wireless/zd1211rw/zd_mac.c @@ -698,7 +698,7 @@ int zd_mac_rx(struct ieee80211_hw *hw, const u8 *buffer, unsigned int length) && !mac->pass_ctrl) return 0; - fc = *(__le16 *)buffer; + fc = get_unaligned((__le16*)buffer); need_padding = ieee80211_is_data_qos(fc) ^ ieee80211_has_a4(fc); skb = dev_alloc_skb(length + (need_padding ? 2 : 0)); diff --git a/include/net/mac80211.h b/include/net/mac80211.h index d4e09a06b4a2..5ed93f4406a8 100644 --- a/include/net/mac80211.h +++ b/include/net/mac80211.h @@ -72,6 +72,21 @@ * not do so then mac80211 may add this under certain circumstances. */ +/** + * DOC: mac80211 workqueue + * + * mac80211 provides its own workqueue for drivers and internal mac80211 use. + * The workqueue is a single threaded workqueue and can only be accessed by + * helpers for sanity checking. Drivers must ensure all work added onto the + * mac80211 workqueue should be cancelled on the driver stop() callback. + * + * mac80211 will flushed the workqueue upon interface removal and during + * suspend. + * + * All work performed on the mac80211 workqueue must not acquire the RTNL lock. + * + */ + /** * enum ieee80211_max_queues - maximum number of queues * @@ -913,12 +928,6 @@ enum ieee80211_hw_flags { * * @conf: &struct ieee80211_conf, device configuration, don't use. * - * @workqueue: single threaded workqueue available for driver use, - * allocated by mac80211 on registration and flushed when an - * interface is removed. - * NOTICE: All work performed on this workqueue must not - * acquire the RTNL lock. - * * @priv: pointer to private area that was allocated for driver use * along with this structure. * @@ -954,7 +963,6 @@ enum ieee80211_hw_flags { struct ieee80211_hw { struct ieee80211_conf conf; struct wiphy *wiphy; - struct workqueue_struct *workqueue; const char *rate_control_algorithm; void *priv; u32 flags; @@ -1301,7 +1309,8 @@ enum ieee80211_ampdu_mlme_action { * is disabled. This should turn off the hardware (at least * it must turn off frame reception.) * May be called right after add_interface if that rejects - * an interface. + * an interface. If you added any work onto the mac80211 workqueue + * you should ensure to cancel it on this callback. * Must be implemented. * * @add_interface: Called when a netdevice attached to the hardware is @@ -1927,6 +1936,31 @@ void ieee80211_iterate_active_interfaces_atomic(struct ieee80211_hw *hw, struct ieee80211_vif *vif), void *data); +/** + * ieee80211_queue_work - add work onto the mac80211 workqueue + * + * Drivers and mac80211 use this to add work onto the mac80211 workqueue. + * This helper ensures drivers are not queueing work when they should not be. + * + * @hw: the hardware struct for the interface we are adding work for + * @work: the work we want to add onto the mac80211 workqueue + */ +void ieee80211_queue_work(struct ieee80211_hw *hw, struct work_struct *work); + +/** + * ieee80211_queue_delayed_work - add work onto the mac80211 workqueue + * + * Drivers and mac80211 use this to queue delayed work onto the mac80211 + * workqueue. + * + * @hw: the hardware struct for the interface we are adding work for + * @dwork: delayable work to queue onto the mac80211 workqueue + * @delay: number of jiffies to wait before queueing + */ +void ieee80211_queue_delayed_work(struct ieee80211_hw *hw, + struct delayed_work *dwork, + unsigned long delay); + /** * ieee80211_start_tx_ba_session - Start a tx Block Ack session. * @hw: pointer as obtained from ieee80211_alloc_hw(). diff --git a/net/mac80211/ibss.c b/net/mac80211/ibss.c index 6e3cca65c460..920ec8792f4b 100644 --- a/net/mac80211/ibss.c +++ b/net/mac80211/ibss.c @@ -781,7 +781,7 @@ static void ieee80211_ibss_timer(unsigned long data) } set_bit(IEEE80211_IBSS_REQ_RUN, &ifibss->request); - queue_work(local->hw.workqueue, &ifibss->work); + ieee80211_queue_work(&local->hw, &ifibss->work); } #ifdef CONFIG_PM @@ -853,7 +853,7 @@ ieee80211_ibss_rx_mgmt(struct ieee80211_sub_if_data *sdata, struct sk_buff *skb) case IEEE80211_STYPE_PROBE_REQ: case IEEE80211_STYPE_AUTH: skb_queue_tail(&sdata->u.ibss.skb_queue, skb); - queue_work(local->hw.workqueue, &sdata->u.ibss.work); + ieee80211_queue_work(&local->hw, &sdata->u.ibss.work); return RX_QUEUED; } @@ -912,7 +912,7 @@ int ieee80211_ibss_join(struct ieee80211_sub_if_data *sdata, ieee80211_recalc_idle(sdata->local); set_bit(IEEE80211_IBSS_REQ_RUN, &sdata->u.ibss.request); - queue_work(sdata->local->hw.workqueue, &sdata->u.ibss.work); + ieee80211_queue_work(&sdata->local->hw, &sdata->u.ibss.work); return 0; } diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h index aec6853cb435..316825be2019 100644 --- a/net/mac80211/ieee80211_i.h +++ b/net/mac80211/ieee80211_i.h @@ -614,6 +614,12 @@ struct ieee80211_local { const struct ieee80211_ops *ops; + /* + * private workqueue to mac80211. mac80211 makes this accessible + * via ieee80211_queue_work() + */ + struct workqueue_struct *workqueue; + unsigned long queue_stop_reasons[IEEE80211_MAX_QUEUES]; /* also used to protect ampdu_ac_queue and amdpu_ac_stop_refcnt */ spinlock_t queue_stop_reason_lock; diff --git a/net/mac80211/iface.c b/net/mac80211/iface.c index a83087f4237d..8c1284d45e69 100644 --- a/net/mac80211/iface.c +++ b/net/mac80211/iface.c @@ -312,7 +312,7 @@ static int ieee80211_open(struct net_device *dev) * to fix this. */ if (sdata->vif.type == NL80211_IFTYPE_STATION) - queue_work(local->hw.workqueue, &sdata->u.mgd.work); + ieee80211_queue_work(&local->hw, &sdata->u.mgd.work); netif_tx_start_all_queues(dev); @@ -551,7 +551,7 @@ static int ieee80211_stop(struct net_device *dev) ieee80211_led_radio(local, false); - flush_workqueue(local->hw.workqueue); + flush_workqueue(local->workqueue); tasklet_disable(&local->tx_pending_tasklet); tasklet_disable(&local->tasklet); diff --git a/net/mac80211/main.c b/net/mac80211/main.c index 5e76dd1daf71..22e07385ff60 100644 --- a/net/mac80211/main.c +++ b/net/mac80211/main.c @@ -821,9 +821,9 @@ int ieee80211_register_hw(struct ieee80211_hw *hw) if (hw->queues > IEEE80211_MAX_QUEUES) hw->queues = IEEE80211_MAX_QUEUES; - local->hw.workqueue = + local->workqueue = create_singlethread_workqueue(wiphy_name(local->hw.wiphy)); - if (!local->hw.workqueue) { + if (!local->workqueue) { result = -ENOMEM; goto fail_workqueue; } @@ -913,7 +913,7 @@ int ieee80211_register_hw(struct ieee80211_hw *hw) sta_info_stop(local); fail_sta_info: debugfs_hw_del(local); - destroy_workqueue(local->hw.workqueue); + destroy_workqueue(local->workqueue); fail_workqueue: wiphy_unregister(local->hw.wiphy); fail_wiphy_register: @@ -955,7 +955,7 @@ void ieee80211_unregister_hw(struct ieee80211_hw *hw) skb_queue_purge(&local->skb_queue); skb_queue_purge(&local->skb_queue_unreliable); - destroy_workqueue(local->hw.workqueue); + destroy_workqueue(local->workqueue); wiphy_unregister(local->hw.wiphy); ieee80211_wep_free(local); ieee80211_led_exit(local); diff --git a/net/mac80211/mesh.c b/net/mac80211/mesh.c index 9a3826978b1c..2f4f518ab45c 100644 --- a/net/mac80211/mesh.c +++ b/net/mac80211/mesh.c @@ -54,7 +54,7 @@ static void ieee80211_mesh_housekeeping_timer(unsigned long data) return; } - queue_work(local->hw.workqueue, &ifmsh->work); + ieee80211_queue_work(local->hw.workqueue, &ifmsh->work); } /** @@ -357,7 +357,7 @@ static void ieee80211_mesh_path_timer(unsigned long data) return; } - queue_work(local->hw.workqueue, &ifmsh->work); + ieee80211_queue_work(local->hw.workqueue, &ifmsh->work); } struct mesh_table *mesh_table_grow(struct mesh_table *tbl) @@ -471,7 +471,7 @@ void ieee80211_start_mesh(struct ieee80211_sub_if_data *sdata) struct ieee80211_local *local = sdata->local; ifmsh->housekeeping = true; - queue_work(local->hw.workqueue, &ifmsh->work); + ieee80211_queue_work(local->hw.workqueue, &ifmsh->work); ieee80211_bss_info_change_notify(sdata, BSS_CHANGED_BEACON | BSS_CHANGED_BEACON_ENABLED); } @@ -619,7 +619,7 @@ void ieee80211_mesh_notify_scan_completed(struct ieee80211_local *local) rcu_read_lock(); list_for_each_entry_rcu(sdata, &local->interfaces, list) if (ieee80211_vif_is_mesh(&sdata->vif)) - queue_work(local->hw.workqueue, &sdata->u.mesh.work); + ieee80211_queue_work(local->hw.workqueue, &sdata->u.mesh.work); rcu_read_unlock(); } @@ -692,7 +692,7 @@ ieee80211_mesh_rx_mgmt(struct ieee80211_sub_if_data *sdata, struct sk_buff *skb) case IEEE80211_STYPE_PROBE_RESP: case IEEE80211_STYPE_BEACON: skb_queue_tail(&ifmsh->skb_queue, skb); - queue_work(local->hw.workqueue, &ifmsh->work); + ieee80211_queue_work(local->hw.workqueue, &ifmsh->work); return RX_QUEUED; } diff --git a/net/mac80211/mesh_hwmp.c b/net/mac80211/mesh_hwmp.c index e93c37ef6a48..11ab71a68ff9 100644 --- a/net/mac80211/mesh_hwmp.c +++ b/net/mac80211/mesh_hwmp.c @@ -660,14 +660,14 @@ static void mesh_queue_preq(struct mesh_path *mpath, u8 flags) spin_unlock(&ifmsh->mesh_preq_queue_lock); if (time_after(jiffies, ifmsh->last_preq + min_preq_int_jiff(sdata))) - queue_work(sdata->local->hw.workqueue, &ifmsh->work); + ieee80211_queue_work(sdata->local->hw.workqueue, &ifmsh->work); else if (time_before(jiffies, ifmsh->last_preq)) { /* avoid long wait if did not send preqs for a long time * and jiffies wrapped around */ ifmsh->last_preq = jiffies - min_preq_int_jiff(sdata) - 1; - queue_work(sdata->local->hw.workqueue, &ifmsh->work); + ieee80211_queue_work(sdata->local->hw.workqueue, &ifmsh->work); } else mod_timer(&ifmsh->mesh_path_timer, ifmsh->last_preq + min_preq_int_jiff(sdata)); diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index ee83125ed179..0779ba150b26 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -565,7 +565,7 @@ static void ieee80211_chswitch_timer(unsigned long data) return; } - queue_work(sdata->local->hw.workqueue, &ifmgd->chswitch_work); + ieee80211_queue_work(&sdata->local->hw, &ifmgd->chswitch_work); } void ieee80211_sta_process_chanswitch(struct ieee80211_sub_if_data *sdata, @@ -597,7 +597,7 @@ void ieee80211_sta_process_chanswitch(struct ieee80211_sub_if_data *sdata, sdata->local->csa_channel = new_ch; if (sw_elem->count <= 1) { - queue_work(sdata->local->hw.workqueue, &ifmgd->chswitch_work); + ieee80211_queue_work(&sdata->local->hw, &ifmgd->chswitch_work); } else { ieee80211_stop_queues_by_reason(&sdata->local->hw, IEEE80211_QUEUE_STOP_REASON_CSA); @@ -763,7 +763,7 @@ void ieee80211_dynamic_ps_timer(unsigned long data) if (local->quiescing || local->suspended) return; - queue_work(local->hw.workqueue, &local->dynamic_ps_enable_work); + ieee80211_queue_work(&local->hw, &local->dynamic_ps_enable_work); } /* MLME */ @@ -950,7 +950,7 @@ ieee80211_direct_probe(struct ieee80211_sub_if_data *sdata, * due to work needing to be done. Hence, queue the STAs work * again for that. */ - queue_work(local->hw.workqueue, &ifmgd->work); + ieee80211_queue_work(&local->hw, &ifmgd->work); return RX_MGMT_CFG80211_AUTH_TO; } @@ -995,7 +995,7 @@ ieee80211_authenticate(struct ieee80211_sub_if_data *sdata, * due to work needing to be done. Hence, queue the STAs work * again for that. */ - queue_work(local->hw.workqueue, &ifmgd->work); + ieee80211_queue_work(&local->hw, &ifmgd->work); return RX_MGMT_CFG80211_AUTH_TO; } @@ -1124,7 +1124,7 @@ ieee80211_associate(struct ieee80211_sub_if_data *sdata, * due to work needing to be done. Hence, queue the STAs work * again for that. */ - queue_work(local->hw.workqueue, &ifmgd->work); + ieee80211_queue_work(&local->hw, &ifmgd->work); return RX_MGMT_CFG80211_ASSOC_TO; } @@ -1232,8 +1232,7 @@ void ieee80211_beacon_loss(struct ieee80211_vif *vif) { struct ieee80211_sub_if_data *sdata = vif_to_sdata(vif); - queue_work(sdata->local->hw.workqueue, - &sdata->u.mgd.beacon_loss_work); + ieee80211_queue_work(&sdata->local->hw, &sdata->u.mgd.beacon_loss_work); } EXPORT_SYMBOL(ieee80211_beacon_loss); @@ -1888,7 +1887,7 @@ ieee80211_rx_result ieee80211_sta_rx_mgmt(struct ieee80211_sub_if_data *sdata, case IEEE80211_STYPE_DISASSOC: case IEEE80211_STYPE_ACTION: skb_queue_tail(&sdata->u.mgd.skb_queue, skb); - queue_work(local->hw.workqueue, &sdata->u.mgd.work); + ieee80211_queue_work(&local->hw, &sdata->u.mgd.work); return RX_QUEUED; } @@ -2026,7 +2025,7 @@ static void ieee80211_sta_timer(unsigned long data) return; } - queue_work(local->hw.workqueue, &ifmgd->work); + ieee80211_queue_work(&local->hw, &ifmgd->work); } static void ieee80211_sta_work(struct work_struct *work) @@ -2051,13 +2050,11 @@ static void ieee80211_sta_work(struct work_struct *work) return; /* - * Nothing should have been stuffed into the workqueue during - * the suspend->resume cycle. If this WARN is seen then there - * is a bug with either the driver suspend or something in - * mac80211 stuffing into the workqueue which we haven't yet - * cleared during mac80211's suspend cycle. + * ieee80211_queue_work() should have picked up most cases, + * here we'll pick the the rest. */ - if (WARN_ON(local->suspended)) + if (WARN(local->suspended, "STA MLME work scheduled while " + "going to suspend\n")) return; ifmgd = &sdata->u.mgd; @@ -2113,9 +2110,9 @@ static void ieee80211_sta_work(struct work_struct *work) mutex_unlock(&ifmgd->mtx); if (test_and_clear_bit(IEEE80211_STA_REQ_SCAN, &ifmgd->request)) - queue_delayed_work(local->hw.workqueue, - &local->scan_work, - round_jiffies_relative(0)); + ieee80211_queue_delayed_work(&local->hw, + &local->scan_work, + round_jiffies_relative(0)); return; } @@ -2196,8 +2193,7 @@ static void ieee80211_sta_bcn_mon_timer(unsigned long data) if (local->quiescing) return; - queue_work(sdata->local->hw.workqueue, - &sdata->u.mgd.beacon_loss_work); + ieee80211_queue_work(&sdata->local->hw, &sdata->u.mgd.beacon_loss_work); } static void ieee80211_sta_conn_mon_timer(unsigned long data) @@ -2210,7 +2206,7 @@ static void ieee80211_sta_conn_mon_timer(unsigned long data) if (local->quiescing) return; - queue_work(local->hw.workqueue, &ifmgd->monitor_work); + ieee80211_queue_work(&local->hw, &ifmgd->monitor_work); } static void ieee80211_sta_monitor_work(struct work_struct *work) @@ -2229,10 +2225,10 @@ static void ieee80211_restart_sta_timer(struct ieee80211_sub_if_data *sdata) IEEE80211_STA_CONNECTION_POLL); /* let's probe the connection once */ - queue_work(sdata->local->hw.workqueue, + ieee80211_queue_work(&sdata->local->hw, &sdata->u.mgd.monitor_work); /* and do all the other regular work too */ - queue_work(sdata->local->hw.workqueue, + ieee80211_queue_work(&sdata->local->hw, &sdata->u.mgd.work); } } @@ -2393,7 +2389,7 @@ int ieee80211_mgd_auth(struct ieee80211_sub_if_data *sdata, list_add(&wk->list, &sdata->u.mgd.work_list); mutex_unlock(&ifmgd->mtx); - queue_work(sdata->local->hw.workqueue, &sdata->u.mgd.work); + ieee80211_queue_work(&sdata->local->hw, &sdata->u.mgd.work); return 0; } @@ -2467,7 +2463,7 @@ int ieee80211_mgd_assoc(struct ieee80211_sub_if_data *sdata, else ifmgd->flags &= ~IEEE80211_STA_CONTROL_PORT; - queue_work(sdata->local->hw.workqueue, &sdata->u.mgd.work); + ieee80211_queue_work(&sdata->local->hw, &sdata->u.mgd.work); err = 0; diff --git a/net/mac80211/pm.c b/net/mac80211/pm.c index 3320f7daaf25..a5d2f1fb4417 100644 --- a/net/mac80211/pm.c +++ b/net/mac80211/pm.c @@ -26,7 +26,7 @@ int __ieee80211_suspend(struct ieee80211_hw *hw) /* make quiescing visible to timers everywhere */ mb(); - flush_workqueue(local->hw.workqueue); + flush_workqueue(local->workqueue); /* Don't try to run timers while suspended. */ del_timer_sync(&local->sta_cleanup); @@ -117,7 +117,7 @@ int __ieee80211_suspend(struct ieee80211_hw *hw) * shouldn't be doing (or cancel everything in the * stop callback) that but better safe than sorry. */ - flush_workqueue(local->hw.workqueue); + flush_workqueue(local->workqueue); local->suspended = true; /* need suspended to be visible before quiescing is false */ diff --git a/net/mac80211/scan.c b/net/mac80211/scan.c index 45731000eb8d..244f53f3c8b4 100644 --- a/net/mac80211/scan.c +++ b/net/mac80211/scan.c @@ -385,8 +385,9 @@ static int ieee80211_start_sw_scan(struct ieee80211_local *local) spin_unlock_bh(&local->filter_lock); /* TODO: start scan as soon as all nullfunc frames are ACKed */ - queue_delayed_work(local->hw.workqueue, &local->scan_work, - IEEE80211_CHANNEL_TIME); + ieee80211_queue_delayed_work(&local->hw, + &local->scan_work, + IEEE80211_CHANNEL_TIME); return 0; } @@ -715,8 +716,7 @@ void ieee80211_scan_work(struct work_struct *work) } } while (next_delay == 0); - queue_delayed_work(local->hw.workqueue, &local->scan_work, - next_delay); + ieee80211_queue_delayed_work(&local->hw, &local->scan_work, next_delay); } int ieee80211_request_scan(struct ieee80211_sub_if_data *sdata, diff --git a/net/mac80211/tx.c b/net/mac80211/tx.c index 4e1b2ba122cd..7cffaa046b33 100644 --- a/net/mac80211/tx.c +++ b/net/mac80211/tx.c @@ -1400,7 +1400,7 @@ static void ieee80211_xmit(struct ieee80211_sub_if_data *sdata, if (local->hw.conf.flags & IEEE80211_CONF_PS) { ieee80211_stop_queues_by_reason(&local->hw, IEEE80211_QUEUE_STOP_REASON_PS); - queue_work(local->hw.workqueue, + ieee80211_queue_work(&local->hw, &local->dynamic_ps_disable_work); } diff --git a/net/mac80211/util.c b/net/mac80211/util.c index 8502936e5314..e55d57f559ec 100644 --- a/net/mac80211/util.c +++ b/net/mac80211/util.c @@ -511,6 +511,46 @@ void ieee80211_iterate_active_interfaces_atomic( } EXPORT_SYMBOL_GPL(ieee80211_iterate_active_interfaces_atomic); +/* + * Nothing should have been stuffed into the workqueue during + * the suspend->resume cycle. If this WARN is seen then there + * is a bug with either the driver suspend or something in + * mac80211 stuffing into the workqueue which we haven't yet + * cleared during mac80211's suspend cycle. + */ +static bool ieee80211_can_queue_work(struct ieee80211_local *local) +{ + if (WARN(local->suspended, "queueing ieee80211 work while " + "going to suspend\n")) + return false; + + return true; +} + +void ieee80211_queue_work(struct ieee80211_hw *hw, struct work_struct *work) +{ + struct ieee80211_local *local = hw_to_local(hw); + + if (!ieee80211_can_queue_work(local)) + return; + + queue_work(local->workqueue, work); +} +EXPORT_SYMBOL(ieee80211_queue_work); + +void ieee80211_queue_delayed_work(struct ieee80211_hw *hw, + struct delayed_work *dwork, + unsigned long delay) +{ + struct ieee80211_local *local = hw_to_local(hw); + + if (!ieee80211_can_queue_work(local)) + return; + + queue_delayed_work(local->workqueue, dwork, delay); +} +EXPORT_SYMBOL(ieee80211_queue_delayed_work); + void ieee802_11_parse_elems(u8 *start, size_t len, struct ieee802_11_elems *elems) { @@ -1114,3 +1154,4 @@ int ieee80211_reconfig(struct ieee80211_local *local) #endif return 0; } + -- cgit v1.2.3 From 1487cd5e76337555737cbc55d7d83f41460d198f Mon Sep 17 00:00:00 2001 From: Jussi Kivilinna Date: Thu, 30 Jul 2009 19:41:20 +0300 Subject: usbnet: allow "minidriver" to prevent urb unlinking on usbnet_stop rndis_wlan devices freeze after running usbnet_stop several times. It appears that firmware freezes in state where it does not respond to any RNDIS commands and device have to be physically unplugged/replugged. This patch lets minidrivers to disable unlink_urbs on usbnet_stop through new info flag. Signed-off-by: Jussi Kivilinna Cc: David Brownell Signed-off-by: John W. Linville --- drivers/net/usb/usbnet.c | 32 ++++++++++++++++++-------------- drivers/net/wireless/rndis_wlan.c | 9 ++++++--- include/linux/usb/usbnet.h | 1 + 3 files changed, 25 insertions(+), 17 deletions(-) (limited to 'include') diff --git a/drivers/net/usb/usbnet.c b/drivers/net/usb/usbnet.c index 25e435c49040..af1fe4696509 100644 --- a/drivers/net/usb/usbnet.c +++ b/drivers/net/usb/usbnet.c @@ -601,21 +601,25 @@ int usbnet_stop (struct net_device *net) info->description); } - // ensure there are no more active urbs - add_wait_queue (&unlink_wakeup, &wait); - dev->wait = &unlink_wakeup; - temp = unlink_urbs (dev, &dev->txq) + unlink_urbs (dev, &dev->rxq); - - // maybe wait for deletions to finish. - while (!skb_queue_empty(&dev->rxq) - && !skb_queue_empty(&dev->txq) - && !skb_queue_empty(&dev->done)) { - msleep(UNLINK_TIMEOUT_MS); - if (netif_msg_ifdown (dev)) - devdbg (dev, "waited for %d urb completions", temp); + if (!(info->flags & FLAG_AVOID_UNLINK_URBS)) { + /* ensure there are no more active urbs */ + add_wait_queue(&unlink_wakeup, &wait); + dev->wait = &unlink_wakeup; + temp = unlink_urbs(dev, &dev->txq) + + unlink_urbs(dev, &dev->rxq); + + /* maybe wait for deletions to finish. */ + while (!skb_queue_empty(&dev->rxq) + && !skb_queue_empty(&dev->txq) + && !skb_queue_empty(&dev->done)) { + msleep(UNLINK_TIMEOUT_MS); + if (netif_msg_ifdown(dev)) + devdbg(dev, "waited for %d urb completions", + temp); + } + dev->wait = NULL; + remove_wait_queue(&unlink_wakeup, &wait); } - dev->wait = NULL; - remove_wait_queue (&unlink_wakeup, &wait); usb_kill_urb(dev->interrupt); diff --git a/drivers/net/wireless/rndis_wlan.c b/drivers/net/wireless/rndis_wlan.c index 09c0702ae645..76c5ec6bbbc5 100644 --- a/drivers/net/wireless/rndis_wlan.c +++ b/drivers/net/wireless/rndis_wlan.c @@ -2513,7 +2513,8 @@ static int rndis_wlan_stop(struct usbnet *usbdev) static const struct driver_info bcm4320b_info = { .description = "Wireless RNDIS device, BCM4320b based", - .flags = FLAG_WLAN | FLAG_FRAMING_RN | FLAG_NO_SETINT, + .flags = FLAG_WLAN | FLAG_FRAMING_RN | FLAG_NO_SETINT | + FLAG_AVOID_UNLINK_URBS, .bind = rndis_wlan_bind, .unbind = rndis_wlan_unbind, .status = rndis_status, @@ -2527,7 +2528,8 @@ static const struct driver_info bcm4320b_info = { static const struct driver_info bcm4320a_info = { .description = "Wireless RNDIS device, BCM4320a based", - .flags = FLAG_WLAN | FLAG_FRAMING_RN | FLAG_NO_SETINT, + .flags = FLAG_WLAN | FLAG_FRAMING_RN | FLAG_NO_SETINT | + FLAG_AVOID_UNLINK_URBS, .bind = rndis_wlan_bind, .unbind = rndis_wlan_unbind, .status = rndis_status, @@ -2541,7 +2543,8 @@ static const struct driver_info bcm4320a_info = { static const struct driver_info rndis_wlan_info = { .description = "Wireless RNDIS device", - .flags = FLAG_WLAN | FLAG_FRAMING_RN | FLAG_NO_SETINT, + .flags = FLAG_WLAN | FLAG_FRAMING_RN | FLAG_NO_SETINT | + FLAG_AVOID_UNLINK_URBS, .bind = rndis_wlan_bind, .unbind = rndis_wlan_unbind, .status = rndis_status, diff --git a/include/linux/usb/usbnet.h b/include/linux/usb/usbnet.h index 7c17b2efba86..c642f78dd9cf 100644 --- a/include/linux/usb/usbnet.h +++ b/include/linux/usb/usbnet.h @@ -86,6 +86,7 @@ struct driver_info { #define FLAG_FRAMING_AX 0x0040 /* AX88772/178 packets */ #define FLAG_WLAN 0x0080 /* use "wlan%d" names */ +#define FLAG_AVOID_UNLINK_URBS 0x0100 /* don't unlink urbs at usbnet_stop() */ /* init device ... can sleep, or cause probe() failure */ -- cgit v1.2.3 From 2a4901bcbe9c122bd56e1f6c337fcb4ad75fafb7 Mon Sep 17 00:00:00 2001 From: Jussi Kivilinna Date: Thu, 30 Jul 2009 19:41:52 +0300 Subject: rndis_host: allow rndis_wlan to see all indications Allow rndis_wlan to see all indications. Currently rndis_host lets rndis_wlan to know about link state changes only, but there is whole set of other 802.11-specific indications that rndis_wlan should handle properly. So rename link_change() to indication() and convert rndis_wlan to use it. Signed-off-by: Jussi Kivilinna Cc: David Brownell Signed-off-by: John W. Linville --- drivers/net/usb/rndis_host.c | 50 +++++++++++++++++++++++---------------- drivers/net/wireless/rndis_wlan.c | 31 +++++++++++++++++++----- include/linux/usb/usbnet.h | 5 ++-- 3 files changed, 56 insertions(+), 30 deletions(-) (limited to 'include') diff --git a/drivers/net/usb/rndis_host.c b/drivers/net/usb/rndis_host.c index 2232232b7989..d032bba9bc4c 100644 --- a/drivers/net/usb/rndis_host.c +++ b/drivers/net/usb/rndis_host.c @@ -64,6 +64,32 @@ void rndis_status(struct usbnet *dev, struct urb *urb) } EXPORT_SYMBOL_GPL(rndis_status); +/* + * RNDIS indicate messages. + */ +static void rndis_msg_indicate(struct usbnet *dev, struct rndis_indicate *msg, + int buflen) +{ + struct cdc_state *info = (void *)&dev->data; + struct device *udev = &info->control->dev; + + if (dev->driver_info->indication) { + dev->driver_info->indication(dev, msg, buflen); + } else { + switch (msg->status) { + case RNDIS_STATUS_MEDIA_CONNECT: + dev_info(udev, "rndis media connect\n"); + break; + case RNDIS_STATUS_MEDIA_DISCONNECT: + dev_info(udev, "rndis media disconnect\n"); + break; + default: + dev_info(udev, "rndis indication: 0x%08x\n", + le32_to_cpu(msg->status)); + } + } +} + /* * RPC done RNDIS-style. Caller guarantees: * - message is properly byteswapped @@ -143,27 +169,9 @@ int rndis_command(struct usbnet *dev, struct rndis_msg_hdr *buf, int buflen) request_id, xid); /* then likely retry */ } else switch (buf->msg_type) { - case RNDIS_MSG_INDICATE: { /* fault/event */ - struct rndis_indicate *msg = (void *)buf; - int state = 0; - - switch (msg->status) { - case RNDIS_STATUS_MEDIA_CONNECT: - state = 1; - case RNDIS_STATUS_MEDIA_DISCONNECT: - dev_info(&info->control->dev, - "rndis media %sconnect\n", - !state?"dis":""); - if (dev->driver_info->link_change) - dev->driver_info->link_change( - dev, state); - break; - default: - dev_info(&info->control->dev, - "rndis indication: 0x%08x\n", - le32_to_cpu(msg->status)); - } - } + case RNDIS_MSG_INDICATE: /* fault/event */ + rndis_msg_indicate(dev, (void *)buf, buflen); + break; case RNDIS_MSG_KEEPALIVE: { /* ping */ struct rndis_keepalive_c *msg = (void *)buf; diff --git a/drivers/net/wireless/rndis_wlan.c b/drivers/net/wireless/rndis_wlan.c index f6dcbb168b78..6b6452b0e8c4 100644 --- a/drivers/net/wireless/rndis_wlan.c +++ b/drivers/net/wireless/rndis_wlan.c @@ -2211,13 +2211,32 @@ static void rndis_wlan_set_multicast_list(struct net_device *dev) queue_work(priv->workqueue, &priv->work); } -static void rndis_wlan_link_change(struct usbnet *usbdev, int state) +static void rndis_wlan_indication(struct usbnet *usbdev, void *ind, int buflen) { struct rndis_wlan_private *priv = get_rndis_wlan_priv(usbdev); + struct rndis_indicate *msg = ind; /* queue work to avoid recursive calls into rndis_command */ - set_bit(state ? WORK_LINK_UP : WORK_LINK_DOWN, &priv->work_pending); - queue_work(priv->workqueue, &priv->work); + switch (msg->status) { + case RNDIS_STATUS_MEDIA_CONNECT: + devinfo(usbdev, "media connect"); + + set_bit(WORK_LINK_UP, &priv->work_pending); + queue_work(priv->workqueue, &priv->work); + break; + + case RNDIS_STATUS_MEDIA_DISCONNECT: + devinfo(usbdev, "media disconnect"); + + set_bit(WORK_LINK_DOWN, &priv->work_pending); + queue_work(priv->workqueue, &priv->work); + break; + + default: + devinfo(usbdev, "indication: 0x%08x", + le32_to_cpu(msg->status)); + break; + } } @@ -2666,7 +2685,7 @@ static const struct driver_info bcm4320b_info = { .reset = rndis_wlan_reset, .stop = rndis_wlan_stop, .early_init = bcm4320b_early_init, - .link_change = rndis_wlan_link_change, + .indication = rndis_wlan_indication, }; static const struct driver_info bcm4320a_info = { @@ -2681,7 +2700,7 @@ static const struct driver_info bcm4320a_info = { .reset = rndis_wlan_reset, .stop = rndis_wlan_stop, .early_init = bcm4320a_early_init, - .link_change = rndis_wlan_link_change, + .indication = rndis_wlan_indication, }; static const struct driver_info rndis_wlan_info = { @@ -2696,7 +2715,7 @@ static const struct driver_info rndis_wlan_info = { .reset = rndis_wlan_reset, .stop = rndis_wlan_stop, .early_init = bcm4320a_early_init, - .link_change = rndis_wlan_link_change, + .indication = rndis_wlan_indication, }; /*-------------------------------------------------------------------------*/ diff --git a/include/linux/usb/usbnet.h b/include/linux/usb/usbnet.h index c642f78dd9cf..de8b4b18961b 100644 --- a/include/linux/usb/usbnet.h +++ b/include/linux/usb/usbnet.h @@ -122,9 +122,8 @@ struct driver_info { * right after minidriver have initialized hardware. */ int (*early_init)(struct usbnet *dev); - /* called by minidriver when link state changes, state: 0=disconnect, - * 1=connect */ - void (*link_change)(struct usbnet *dev, int state); + /* called by minidriver when receiving indication */ + void (*indication)(struct usbnet *dev, void *ind, int indlen); /* for new devices, use the descriptor-reading code instead */ int in; /* rx endpoint */ -- cgit v1.2.3 From 030645aceb3d9f10b1c3d2231c50f5a8bb3a9667 Mon Sep 17 00:00:00 2001 From: Jussi Kivilinna Date: Thu, 30 Jul 2009 19:41:58 +0300 Subject: rndis_wlan: handle 802.11 indications from device Add handling for 802.11 specific rndis indications. Signed-off-by: Jussi Kivilinna Signed-off-by: John W. Linville --- drivers/net/wireless/rndis_wlan.c | 233 +++++++++++++++++++++++++++++++++++++- include/linux/usb/rndis_host.h | 13 ++- 2 files changed, 239 insertions(+), 7 deletions(-) (limited to 'include') diff --git a/drivers/net/wireless/rndis_wlan.c b/drivers/net/wireless/rndis_wlan.c index 6b6452b0e8c4..7a50cfa18843 100644 --- a/drivers/net/wireless/rndis_wlan.c +++ b/drivers/net/wireless/rndis_wlan.c @@ -201,6 +201,24 @@ enum ndis_80211_priv_filter { NDIS_80211_PRIV_8021X_WEP }; +enum ndis_80211_status_type { + NDIS_80211_STATUSTYPE_AUTHENTICATION, + NDIS_80211_STATUSTYPE_MEDIASTREAMMODE, + NDIS_80211_STATUSTYPE_PMKID_CANDIDATELIST, + NDIS_80211_STATUSTYPE_RADIOSTATE, +}; + +enum ndis_80211_media_stream_mode { + NDIS_80211_MEDIA_STREAM_OFF, + NDIS_80211_MEDIA_STREAM_ON +}; + +enum ndis_80211_radio_status { + NDIS_80211_RADIO_STATUS_ON, + NDIS_80211_RADIO_STATUS_HARDWARE_OFF, + NDIS_80211_RADIO_STATUS_SOFTWARE_OFF, +}; + enum ndis_80211_addkey_bits { NDIS_80211_ADDKEY_8021X_AUTH = cpu_to_le32(1 << 28), NDIS_80211_ADDKEY_SET_INIT_RECV_SEQ = cpu_to_le32(1 << 29), @@ -213,6 +231,35 @@ enum ndis_80211_addwep_bits { NDIS_80211_ADDWEP_TRANSMIT_KEY = cpu_to_le32(1 << 31) }; +struct ndis_80211_auth_request { + __le32 length; + u8 bssid[6]; + u8 padding[2]; + __le32 flags; +} __attribute__((packed)); + +struct ndis_80211_pmkid_candidate { + u8 bssid[6]; + u8 padding[2]; + __le32 flags; +} __attribute__((packed)); + +struct ndis_80211_pmkid_cand_list { + __le32 version; + __le32 num_candidates; + struct ndis_80211_pmkid_candidate candidate_list[0]; +} __attribute__((packed)); + +struct ndis_80211_status_indication { + __le32 status_type; + union { + enum ndis_80211_media_stream_mode media_stream_mode; + enum ndis_80211_radio_status radio_status; + struct ndis_80211_auth_request auth_request[0]; + struct ndis_80211_pmkid_cand_list cand_list; + } u; +} __attribute__((packed)); + struct ndis_80211_ssid { __le32 length; u8 essid[NDIS_802_11_LENGTH_SSID]; @@ -2211,16 +2258,195 @@ static void rndis_wlan_set_multicast_list(struct net_device *dev) queue_work(priv->workqueue, &priv->work); } + +static void rndis_wlan_auth_indication(struct usbnet *usbdev, + struct ndis_80211_status_indication *indication, + int len) +{ + u8 *buf; + const char *type; + int flags, buflen; + bool pairwise_error, group_error; + struct ndis_80211_auth_request *auth_req; + + /* must have at least one array entry */ + if (len < offsetof(struct ndis_80211_status_indication, u) + + sizeof(struct ndis_80211_auth_request)) { + devinfo(usbdev, "authentication indication: " + "too short message (%i)", len); + return; + } + + buf = (void *)&indication->u.auth_request[0]; + buflen = len - offsetof(struct ndis_80211_status_indication, u); + + while (buflen >= sizeof(*auth_req)) { + auth_req = (void *)buf; + type = "unknown"; + flags = le32_to_cpu(auth_req->flags); + pairwise_error = false; + group_error = false; + + if (flags & 0x1) + type = "reauth request"; + if (flags & 0x2) + type = "key update request"; + if (flags & 0x6) { + pairwise_error = true; + type = "pairwise_error"; + } + if (flags & 0xe) { + group_error = true; + type = "group_error"; + } + + devinfo(usbdev, "authentication indication: %s (0x%08x)", type, + le32_to_cpu(auth_req->flags)); + + if (pairwise_error || group_error) { + union iwreq_data wrqu; + struct iw_michaelmicfailure micfailure; + + memset(&micfailure, 0, sizeof(micfailure)); + if (pairwise_error) + micfailure.flags |= IW_MICFAILURE_PAIRWISE; + if (group_error) + micfailure.flags |= IW_MICFAILURE_GROUP; + + memcpy(micfailure.src_addr.sa_data, auth_req->bssid, + ETH_ALEN); + + memset(&wrqu, 0, sizeof(wrqu)); + wrqu.data.length = sizeof(micfailure); + wireless_send_event(usbdev->net, IWEVMICHAELMICFAILURE, + &wrqu, (u8 *)&micfailure); + } + + buflen -= le32_to_cpu(auth_req->length); + buf += le32_to_cpu(auth_req->length); + } +} + +static void rndis_wlan_pmkid_cand_list_indication(struct usbnet *usbdev, + struct ndis_80211_status_indication *indication, + int len) +{ + struct ndis_80211_pmkid_cand_list *cand_list; + int list_len, expected_len, i; + + if (len < offsetof(struct ndis_80211_status_indication, u) + + sizeof(struct ndis_80211_pmkid_cand_list)) { + devinfo(usbdev, "pmkid candidate list indication: " + "too short message (%i)", len); + return; + } + + list_len = le32_to_cpu(indication->u.cand_list.num_candidates) * + sizeof(struct ndis_80211_pmkid_candidate); + expected_len = sizeof(struct ndis_80211_pmkid_cand_list) + list_len + + offsetof(struct ndis_80211_status_indication, u); + + if (len < expected_len) { + devinfo(usbdev, "pmkid candidate list indication: " + "list larger than buffer (%i < %i)", + len, expected_len); + return; + } + + cand_list = &indication->u.cand_list; + + devinfo(usbdev, "pmkid candidate list indication: " + "version %i, candidates %i", + le32_to_cpu(cand_list->version), + le32_to_cpu(cand_list->num_candidates)); + + if (le32_to_cpu(cand_list->version) != 1) + return; + + for (i = 0; i < le32_to_cpu(cand_list->num_candidates); i++) { + struct iw_pmkid_cand pcand; + union iwreq_data wrqu; + struct ndis_80211_pmkid_candidate *cand = + &cand_list->candidate_list[i]; + + devdbg(usbdev, "cand[%i]: flags: 0x%08x, bssid: %pM", + i, le32_to_cpu(cand->flags), cand->bssid); + + memset(&pcand, 0, sizeof(pcand)); + if (le32_to_cpu(cand->flags) & 0x01) + pcand.flags |= IW_PMKID_CAND_PREAUTH; + pcand.index = i; + memcpy(pcand.bssid.sa_data, cand->bssid, ETH_ALEN); + + memset(&wrqu, 0, sizeof(wrqu)); + wrqu.data.length = sizeof(pcand); + wireless_send_event(usbdev->net, IWEVPMKIDCAND, &wrqu, + (u8 *)&pcand); + } +} + +static void rndis_wlan_media_specific_indication(struct usbnet *usbdev, + struct rndis_indicate *msg, int buflen) +{ + struct ndis_80211_status_indication *indication; + int len, offset; + + offset = offsetof(struct rndis_indicate, status) + + le32_to_cpu(msg->offset); + len = le32_to_cpu(msg->length); + + if (len < 8) { + devinfo(usbdev, "media specific indication, " + "ignore too short message (%i < 8)", len); + return; + } + + if (offset + len > buflen) { + devinfo(usbdev, "media specific indication, " + "too large to fit to buffer (%i > %i)", + offset + len, buflen); + return; + } + + indication = (void *)((u8 *)msg + offset); + + switch (le32_to_cpu(indication->status_type)) { + case NDIS_80211_STATUSTYPE_RADIOSTATE: + devinfo(usbdev, "radio state indication: %i", + le32_to_cpu(indication->u.radio_status)); + return; + + case NDIS_80211_STATUSTYPE_MEDIASTREAMMODE: + devinfo(usbdev, "media stream mode indication: %i", + le32_to_cpu(indication->u.media_stream_mode)); + return; + + case NDIS_80211_STATUSTYPE_AUTHENTICATION: + rndis_wlan_auth_indication(usbdev, indication, len); + return; + + case NDIS_80211_STATUSTYPE_PMKID_CANDIDATELIST: + rndis_wlan_pmkid_cand_list_indication(usbdev, indication, len); + return; + + default: + devinfo(usbdev, "media specific indication: " + "unknown status type 0x%08x", + le32_to_cpu(indication->status_type)); + } +} + + static void rndis_wlan_indication(struct usbnet *usbdev, void *ind, int buflen) { struct rndis_wlan_private *priv = get_rndis_wlan_priv(usbdev); struct rndis_indicate *msg = ind; - /* queue work to avoid recursive calls into rndis_command */ switch (msg->status) { case RNDIS_STATUS_MEDIA_CONNECT: devinfo(usbdev, "media connect"); + /* queue work to avoid recursive calls into rndis_command */ set_bit(WORK_LINK_UP, &priv->work_pending); queue_work(priv->workqueue, &priv->work); break; @@ -2228,10 +2454,15 @@ static void rndis_wlan_indication(struct usbnet *usbdev, void *ind, int buflen) case RNDIS_STATUS_MEDIA_DISCONNECT: devinfo(usbdev, "media disconnect"); + /* queue work to avoid recursive calls into rndis_command */ set_bit(WORK_LINK_DOWN, &priv->work_pending); queue_work(priv->workqueue, &priv->work); break; + case RNDIS_STATUS_MEDIA_SPECIFIC_INDICATION: + rndis_wlan_media_specific_indication(usbdev, msg, buflen); + break; + default: devinfo(usbdev, "indication: 0x%08x", le32_to_cpu(msg->status)); diff --git a/include/linux/usb/rndis_host.h b/include/linux/usb/rndis_host.h index 37836b937d97..1ef1ebc2b04f 100644 --- a/include/linux/usb/rndis_host.h +++ b/include/linux/usb/rndis_host.h @@ -70,12 +70,13 @@ struct rndis_msg_hdr { #define RNDIS_MSG_KEEPALIVE_C (RNDIS_MSG_KEEPALIVE|RNDIS_MSG_COMPLETION) /* codes for "status" field of completion messages */ -#define RNDIS_STATUS_SUCCESS cpu_to_le32(0x00000000) -#define RNDIS_STATUS_FAILURE cpu_to_le32(0xc0000001) -#define RNDIS_STATUS_INVALID_DATA cpu_to_le32(0xc0010015) -#define RNDIS_STATUS_NOT_SUPPORTED cpu_to_le32(0xc00000bb) -#define RNDIS_STATUS_MEDIA_CONNECT cpu_to_le32(0x4001000b) -#define RNDIS_STATUS_MEDIA_DISCONNECT cpu_to_le32(0x4001000c) +#define RNDIS_STATUS_SUCCESS cpu_to_le32(0x00000000) +#define RNDIS_STATUS_FAILURE cpu_to_le32(0xc0000001) +#define RNDIS_STATUS_INVALID_DATA cpu_to_le32(0xc0010015) +#define RNDIS_STATUS_NOT_SUPPORTED cpu_to_le32(0xc00000bb) +#define RNDIS_STATUS_MEDIA_CONNECT cpu_to_le32(0x4001000b) +#define RNDIS_STATUS_MEDIA_DISCONNECT cpu_to_le32(0x4001000c) +#define RNDIS_STATUS_MEDIA_SPECIFIC_INDICATION cpu_to_le32(0x40010012) /* codes for OID_GEN_PHYSICAL_MEDIUM */ #define RNDIS_PHYSICAL_MEDIUM_UNSPECIFIED cpu_to_le32(0x00000000) -- cgit v1.2.3 From 8b19e6ca3bac7e04e93fb73f561d670e77c5fae6 Mon Sep 17 00:00:00 2001 From: "Luis R. Rodriguez" Date: Thu, 30 Jul 2009 17:38:09 -0700 Subject: cfg80211: enable country IE support to all cfg80211 drivers Since the bss is always set now once we are connected, if the bss has its own information element we refer to it and pass that instead of relying on mac80211's parsing. Now all cfg80211 drivers get country IE support, automatically and we reduce the call overhead that we had on mac80211 which called this upon every beacon and instead now call this only upon a successfull connection by a STA on cfg80211. Acked-by: Johannes Berg Signed-off-by: Luis R. Rodriguez Signed-off-by: John W. Linville --- include/net/cfg80211.h | 14 -------------- net/mac80211/mlme.c | 6 +----- net/wireless/reg.c | 6 +----- net/wireless/reg.h | 15 +++++++++++++++ net/wireless/sme.c | 16 ++++++++++++++++ 5 files changed, 33 insertions(+), 24 deletions(-) (limited to 'include') diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index e1b92358242b..fa729979de88 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -1513,20 +1513,6 @@ unsigned int cfg80211_classify8021d(struct sk_buff *skb); */ extern int regulatory_hint(struct wiphy *wiphy, const char *alpha2); -/** - * regulatory_hint_11d - hints a country IE as a regulatory domain - * @wiphy: the wireless device giving the hint (used only for reporting - * conflicts) - * @country_ie: pointer to the country IE - * @country_ie_len: length of the country IE - * - * We will intersect the rd with the what CRDA tells us should apply - * for the alpha2 this country IE belongs to, this prevents APs from - * sending us incorrect or outdated information against a country. - */ -extern void regulatory_hint_11d(struct wiphy *wiphy, - u8 *country_ie, - u8 country_ie_len); /** * wiphy_apply_custom_regulatory - apply a custom driver regulatory domain * @wiphy: the wireless device we want to process the regulatory domain on diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index 2d5edfda867a..c9e4091cd2bb 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -1845,12 +1845,8 @@ static void ieee80211_rx_mgmt_beacon(struct ieee80211_sub_if_data *sdata, bssid, ap_ht_cap_flags); } + /* Note: country IE parsing is done for us by cfg80211 */ if (elems.country_elem) { - /* Note we are only reviewing this on beacons - * for the BSSID we are associated to */ - regulatory_hint_11d(local->hw.wiphy, - elems.country_elem, elems.country_elem_len); - /* TODO: IBSS also needs this */ if (elems.pwr_constr_elem) ieee80211_handle_pwr_constr(sdata, diff --git a/net/wireless/reg.c b/net/wireless/reg.c index 6ab56f098de1..b3ac0aace0e5 100644 --- a/net/wireless/reg.c +++ b/net/wireless/reg.c @@ -1822,10 +1822,7 @@ void regulatory_hint_11d(struct wiphy *wiphy, env = ENVIRON_OUTDOOR; /* - * We will run this for *every* beacon processed for the BSSID, so - * we optimize an early check to exit out early if we don't have to - * do anything - * + * We will run this only upon a successful connection on cfg80211. * We leave conflict resolution to the workqueue, where can hold * cfg80211_mutex. */ @@ -1878,7 +1875,6 @@ free_rd_out: out: mutex_unlock(®_mutex); } -EXPORT_SYMBOL(regulatory_hint_11d); static bool freq_is_chan_12_13_14(u16 freq) { diff --git a/net/wireless/reg.h b/net/wireless/reg.h index e37829a49dc4..662a9dad76d5 100644 --- a/net/wireless/reg.h +++ b/net/wireless/reg.h @@ -36,4 +36,19 @@ int regulatory_hint_found_beacon(struct wiphy *wiphy, struct ieee80211_channel *beacon_chan, gfp_t gfp); +/** + * regulatory_hint_11d - hints a country IE as a regulatory domain + * @wiphy: the wireless device giving the hint (used only for reporting + * conflicts) + * @country_ie: pointer to the country IE + * @country_ie_len: length of the country IE + * + * We will intersect the rd with the what CRDA tells us should apply + * for the alpha2 this country IE belongs to, this prevents APs from + * sending us incorrect or outdated information against a country. + */ +void regulatory_hint_11d(struct wiphy *wiphy, + u8 *country_ie, + u8 country_ie_len); + #endif /* __NET_WIRELESS_REG_H */ diff --git a/net/wireless/sme.c b/net/wireless/sme.c index 3728d2b88b25..af91192eedf5 100644 --- a/net/wireless/sme.c +++ b/net/wireless/sme.c @@ -13,6 +13,7 @@ #include #include #include "nl80211.h" +#include "reg.h" struct cfg80211_conn { struct cfg80211_connect_params params; @@ -320,6 +321,7 @@ void __cfg80211_connect_result(struct net_device *dev, const u8 *bssid, struct cfg80211_bss *bss) { struct wireless_dev *wdev = dev->ieee80211_ptr; + u8 *country_ie; #ifdef CONFIG_WIRELESS_EXT union iwreq_data wrqu; #endif @@ -401,6 +403,20 @@ void __cfg80211_connect_result(struct net_device *dev, const u8 *bssid, wdev->sme_state = CFG80211_SME_CONNECTED; cfg80211_upload_connect_keys(wdev); + + country_ie = (u8 *) ieee80211_bss_get_ie(bss, WLAN_EID_COUNTRY); + + if (!country_ie) + return; + + /* + * ieee80211_bss_get_ie() ensures we can access: + * - country_ie + 2, the start of the country ie data, and + * - and country_ie[1] which is the IE length + */ + regulatory_hint_11d(wdev->wiphy, + country_ie + 2, + country_ie[1]); } void cfg80211_connect_result(struct net_device *dev, const u8 *bssid, -- cgit v1.2.3 From 3ad201496badddd8e1cda87ee6d29e8b3b8e1279 Mon Sep 17 00:00:00 2001 From: Tomas Winkler Date: Sun, 2 Aug 2009 02:36:49 +0300 Subject: rfkill: add the GPS radio type Althoug GPS is a technology w/o transmitting radio and thus not a primary candidate for rfkill switch, rfkill gives unified interface point for devices with wireless technology. The input key is not supplied as it is too be deprecated. Cc: johannes@sipsolutions.net Signed-off-by: Tomas Winkler Acked-by: Marcel Holtmann Signed-off-by: John W. Linville --- include/linux/rfkill.h | 1 + net/rfkill/core.c | 4 +++- 2 files changed, 4 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/rfkill.h b/include/linux/rfkill.h index 10202903141a..21ca51bf4dd2 100644 --- a/include/linux/rfkill.h +++ b/include/linux/rfkill.h @@ -47,6 +47,7 @@ enum rfkill_type { RFKILL_TYPE_UWB, RFKILL_TYPE_WIMAX, RFKILL_TYPE_WWAN, + RFKILL_TYPE_GPS, NUM_RFKILL_TYPES, }; diff --git a/net/rfkill/core.c b/net/rfkill/core.c index 044de1c6af3d..dbeaf2983822 100644 --- a/net/rfkill/core.c +++ b/net/rfkill/core.c @@ -589,11 +589,13 @@ static const char *rfkill_get_type_str(enum rfkill_type type) return "wimax"; case RFKILL_TYPE_WWAN: return "wwan"; + case RFKILL_TYPE_GPS: + return "gps"; default: BUG(); } - BUILD_BUG_ON(NUM_RFKILL_TYPES != RFKILL_TYPE_WWAN + 1); + BUILD_BUG_ON(NUM_RFKILL_TYPES != RFKILL_TYPE_GPS + 1); } static ssize_t rfkill_type_show(struct device *dev, -- cgit v1.2.3 From e3b90ca28412fb9dcc8c5ca38e179e78fec07eee Mon Sep 17 00:00:00 2001 From: Igor Perminov Date: Tue, 4 Aug 2009 16:48:51 +0400 Subject: mac80211: FIF_PSPOLL filter flag When an interface is configured in the AP mode, the mac80211 implementation doesn't inform the driver to receive PS Poll frames. It leads to inability to communicate with power-saving stations reliably. The FIF_CONTROL flag isn't passed by mac80211 to ieee80211_ops.configure_filter when an interface is in the AP mode. And it's ok, because we don't want to receive ACK frames and other control ones, but only PS Poll ones. This patch introduces the FIF_PSPOLL filter flag in addition to FIF_CONTROL, which means for the driver "pass PS Poll frames". This flag is passed to the driver: A) When an interface is configured in the AP mode. B) In all cases, when the FIF_CONTROL flag was passed earlier (in addition to it). Signed-off-by: Igor Perminov Signed-off-by: John W. Linville --- include/net/mac80211.h | 8 ++++++-- net/mac80211/ieee80211_i.h | 2 +- net/mac80211/iface.c | 18 ++++++++++++++++-- net/mac80211/main.c | 3 +++ 4 files changed, 26 insertions(+), 5 deletions(-) (limited to 'include') diff --git a/include/net/mac80211.h b/include/net/mac80211.h index 5ed93f4406a8..e2fb5767e1fa 100644 --- a/include/net/mac80211.h +++ b/include/net/mac80211.h @@ -1244,10 +1244,13 @@ ieee80211_get_alt_retry_rate(const struct ieee80211_hw *hw, * mac80211 needs to do and the amount of CPU wakeups, so you should * honour this flag if possible. * - * @FIF_CONTROL: pass control frames, if PROMISC_IN_BSS is not set then - * only those addressed to this station + * @FIF_CONTROL: pass control frames (except for PS Poll), if PROMISC_IN_BSS + * is not set then only those addressed to this station. * * @FIF_OTHER_BSS: pass frames destined to other BSSes + * + * @FIF_PSPOLL: pass PS Poll frames, if PROMISC_IN_BSS is not set then only + * those addressed to this station. */ enum ieee80211_filter_flags { FIF_PROMISC_IN_BSS = 1<<0, @@ -1257,6 +1260,7 @@ enum ieee80211_filter_flags { FIF_BCN_PRBRESP_PROMISC = 1<<4, FIF_CONTROL = 1<<5, FIF_OTHER_BSS = 1<<6, + FIF_PSPOLL = 1<<7, }; /** diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h index 8d790e40f3e9..630a438180fd 100644 --- a/net/mac80211/ieee80211_i.h +++ b/net/mac80211/ieee80211_i.h @@ -628,7 +628,7 @@ struct ieee80211_local { int open_count; int monitors, cooked_mntrs; /* number of interfaces with corresponding FIF_ flags */ - int fif_fcsfail, fif_plcpfail, fif_control, fif_other_bss; + int fif_fcsfail, fif_plcpfail, fif_control, fif_other_bss, fif_pspoll; unsigned int filter_flags; /* FIF_* */ struct iw_statistics wstats; diff --git a/net/mac80211/iface.c b/net/mac80211/iface.c index 8c1284d45e69..e8fb03b91a44 100644 --- a/net/mac80211/iface.c +++ b/net/mac80211/iface.c @@ -220,8 +220,10 @@ static int ieee80211_open(struct net_device *dev) local->fif_fcsfail++; if (sdata->u.mntr_flags & MONITOR_FLAG_PLCPFAIL) local->fif_plcpfail++; - if (sdata->u.mntr_flags & MONITOR_FLAG_CONTROL) + if (sdata->u.mntr_flags & MONITOR_FLAG_CONTROL) { local->fif_control++; + local->fif_pspoll++; + } if (sdata->u.mntr_flags & MONITOR_FLAG_OTHER_BSS) local->fif_other_bss++; @@ -244,7 +246,14 @@ static int ieee80211_open(struct net_device *dev) spin_unlock_bh(&local->filter_lock); ieee80211_start_mesh(sdata); + } else if (sdata->vif.type == NL80211_IFTYPE_AP) { + local->fif_pspoll++; + + spin_lock_bh(&local->filter_lock); + ieee80211_configure_filter(local); + spin_unlock_bh(&local->filter_lock); } + changed |= ieee80211_reset_erp_info(sdata); ieee80211_bss_info_change_notify(sdata, changed); ieee80211_enable_keys(sdata); @@ -388,6 +397,9 @@ static int ieee80211_stop(struct net_device *dev) if (sdata->flags & IEEE80211_SDATA_PROMISC) atomic_dec(&local->iff_promiscs); + if (sdata->vif.type == NL80211_IFTYPE_AP) + local->fif_pspoll--; + netif_addr_lock_bh(dev); spin_lock_bh(&local->filter_lock); __dev_addr_unsync(&local->mc_list, &local->mc_count, @@ -439,8 +451,10 @@ static int ieee80211_stop(struct net_device *dev) local->fif_fcsfail--; if (sdata->u.mntr_flags & MONITOR_FLAG_PLCPFAIL) local->fif_plcpfail--; - if (sdata->u.mntr_flags & MONITOR_FLAG_CONTROL) + if (sdata->u.mntr_flags & MONITOR_FLAG_CONTROL) { + local->fif_pspoll--; local->fif_control--; + } if (sdata->u.mntr_flags & MONITOR_FLAG_OTHER_BSS) local->fif_other_bss--; diff --git a/net/mac80211/main.c b/net/mac80211/main.c index 22e07385ff60..0c4f8e122ed6 100644 --- a/net/mac80211/main.c +++ b/net/mac80211/main.c @@ -77,6 +77,9 @@ void ieee80211_configure_filter(struct ieee80211_local *local) if (local->fif_other_bss) new_flags |= FIF_OTHER_BSS; + if (local->fif_pspoll) + new_flags |= FIF_PSPOLL; + changed_flags = local->filter_flags ^ new_flags; /* be a bit nasty */ -- cgit v1.2.3 From 18eac1cc100fa2afd5f39085aae6b694e417734b Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Mon, 3 Aug 2009 10:58:29 -0700 Subject: tty-ldisc: make refcount be atomic_t 'users' count This is pure preparation of changing the ldisc reference counting to be a true refcount that defines the lifetime of the ldisc. But this is a purely syntactic change for now to make the next steps easier. This patch should make no semantic changes at all. But I wanted to make the ldisc refcount be an atomic (I will be touching it without locks soon enough), and I wanted to rename it so that there isn't quite as much confusion between 'ldo->refcount' (ldisk operations refcount) and 'ld->refcount' (ldisc refcount itself) in the same file. So it's now an atomic 'ld->users' count. It still starts at zero, despite having a reference from 'tty->ldisc', but that will change once we turn it into a _real_ refcount. Signed-off-by: Linus Torvalds Tested-by: OGAWA Hirofumi Tested-by: Sergey Senozhatsky Acked-by: Alan Cox Signed-off-by: Greg Kroah-Hartman --- drivers/char/tty_ldisc.c | 18 ++++++++---------- include/linux/tty_ldisc.h | 2 +- 2 files changed, 9 insertions(+), 11 deletions(-) (limited to 'include') diff --git a/drivers/char/tty_ldisc.c b/drivers/char/tty_ldisc.c index acd76b767d4c..fd175e60aad5 100644 --- a/drivers/char/tty_ldisc.c +++ b/drivers/char/tty_ldisc.c @@ -142,7 +142,7 @@ static struct tty_ldisc *tty_ldisc_try_get(int disc) /* lock it */ ldops->refcount++; ld->ops = ldops; - ld->refcount = 0; + atomic_set(&ld->users, 0); err = 0; } } @@ -206,7 +206,7 @@ static void tty_ldisc_put(struct tty_ldisc *ld) ldo->refcount--; module_put(ldo->owner); spin_unlock_irqrestore(&tty_ldisc_lock, flags); - WARN_ON(ld->refcount); + WARN_ON(atomic_read(&ld->users)); kfree(ld); } @@ -297,7 +297,7 @@ static int tty_ldisc_try(struct tty_struct *tty) spin_lock_irqsave(&tty_ldisc_lock, flags); ld = tty->ldisc; if (test_bit(TTY_LDISC, &tty->flags)) { - ld->refcount++; + atomic_inc(&ld->users); ret = 1; } spin_unlock_irqrestore(&tty_ldisc_lock, flags); @@ -324,7 +324,7 @@ struct tty_ldisc *tty_ldisc_ref_wait(struct tty_struct *tty) { /* wait_event is a macro */ wait_event(tty_ldisc_wait, tty_ldisc_try(tty)); - WARN_ON(tty->ldisc->refcount == 0); + WARN_ON(atomic_read(&tty->ldisc->users) == 0); return tty->ldisc; } EXPORT_SYMBOL_GPL(tty_ldisc_ref_wait); @@ -365,11 +365,9 @@ void tty_ldisc_deref(struct tty_ldisc *ld) BUG_ON(ld == NULL); spin_lock_irqsave(&tty_ldisc_lock, flags); - if (ld->refcount == 0) + if (atomic_read(&ld->users) == 0) printk(KERN_ERR "tty_ldisc_deref: no references.\n"); - else - ld->refcount--; - if (ld->refcount == 0) + else if (atomic_dec_and_test(&ld->users)) wake_up(&tty_ldisc_wait); spin_unlock_irqrestore(&tty_ldisc_lock, flags); } @@ -536,10 +534,10 @@ static int tty_ldisc_wait_idle(struct tty_struct *tty) { unsigned long flags; spin_lock_irqsave(&tty_ldisc_lock, flags); - while (tty->ldisc->refcount) { + while (atomic_read(&tty->ldisc->users)) { spin_unlock_irqrestore(&tty_ldisc_lock, flags); if (wait_event_timeout(tty_ldisc_wait, - tty->ldisc->refcount == 0, 5 * HZ) == 0) + atomic_read(&tty->ldisc->users) == 0, 5 * HZ) == 0) return -EBUSY; spin_lock_irqsave(&tty_ldisc_lock, flags); } diff --git a/include/linux/tty_ldisc.h b/include/linux/tty_ldisc.h index 40f38d896777..0c4ee9b88f85 100644 --- a/include/linux/tty_ldisc.h +++ b/include/linux/tty_ldisc.h @@ -144,7 +144,7 @@ struct tty_ldisc_ops { struct tty_ldisc { struct tty_ldisc_ops *ops; - int refcount; + atomic_t users; }; #define TTY_LDISC_MAGIC 0x5403 -- cgit v1.2.3 From 6502fbfaf81b09b3f474bb7b3796257e9450273e Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Tue, 4 Aug 2009 11:24:24 -0400 Subject: drm/radeon: Add support for RS880 chips These are new AMD IGP chips Signed-off-by: Alex Deucher Signed-off-by: Dave Airlie --- drivers/gpu/drm/radeon/r600_cp.c | 22 +++++++++++++++------- drivers/gpu/drm/radeon/radeon_drv.h | 1 + include/drm/drm_pciids.h | 5 +++++ 3 files changed, 21 insertions(+), 7 deletions(-) (limited to 'include') diff --git a/drivers/gpu/drm/radeon/r600_cp.c b/drivers/gpu/drm/radeon/r600_cp.c index 146f3570af8e..20f17908b036 100644 --- a/drivers/gpu/drm/radeon/r600_cp.c +++ b/drivers/gpu/drm/radeon/r600_cp.c @@ -384,8 +384,9 @@ static void r600_cp_load_microcode(drm_radeon_private_t *dev_priv) DRM_INFO("Loading RV670 PFP Microcode\n"); for (i = 0; i < PFP_UCODE_SIZE; i++) RADEON_WRITE(R600_CP_PFP_UCODE_DATA, RV670_pfp_microcode[i]); - } else if (((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RS780)) { - DRM_INFO("Loading RS780 CP Microcode\n"); + } else if (((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RS780) || + ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RS880)) { + DRM_INFO("Loading RS780/RS880 CP Microcode\n"); for (i = 0; i < PM4_UCODE_SIZE; i++) { RADEON_WRITE(R600_CP_ME_RAM_DATA, RS780_cp_microcode[i][0]); @@ -396,7 +397,7 @@ static void r600_cp_load_microcode(drm_radeon_private_t *dev_priv) } RADEON_WRITE(R600_CP_PFP_UCODE_ADDR, 0); - DRM_INFO("Loading RS780 PFP Microcode\n"); + DRM_INFO("Loading RS780/RS880 PFP Microcode\n"); for (i = 0; i < PFP_UCODE_SIZE; i++) RADEON_WRITE(R600_CP_PFP_UCODE_DATA, RS780_pfp_microcode[i]); } @@ -783,6 +784,7 @@ static void r600_gfx_init(struct drm_device *dev, break; case CHIP_RV610: case CHIP_RS780: + case CHIP_RS880: case CHIP_RV620: dev_priv->r600_max_pipes = 1; dev_priv->r600_max_tile_pipes = 1; @@ -917,7 +919,8 @@ static void r600_gfx_init(struct drm_device *dev, ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RV630) || ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RV610) || ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RV620) || - ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RS780)) + ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RS780) || + ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RS880)) RADEON_WRITE(R600_DB_DEBUG, R600_PREZ_MUST_WAIT_FOR_POSTZ_DONE); else RADEON_WRITE(R600_DB_DEBUG, 0); @@ -935,7 +938,8 @@ static void r600_gfx_init(struct drm_device *dev, sq_ms_fifo_sizes = RADEON_READ(R600_SQ_MS_FIFO_SIZES); if (((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RV610) || ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RV620) || - ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RS780)) { + ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RS780) || + ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RS880)) { sq_ms_fifo_sizes = (R600_CACHE_FIFO_SIZE(0xa) | R600_FETCH_FIFO_HIWATER(0xa) | R600_DONE_FIFO_HIWATER(0xe0) | @@ -978,7 +982,8 @@ static void r600_gfx_init(struct drm_device *dev, R600_NUM_ES_STACK_ENTRIES(0)); } else if (((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RV610) || ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RV620) || - ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RS780)) { + ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RS780) || + ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RS880)) { /* no vertex cache */ sq_config &= ~R600_VC_ENABLE; @@ -1035,7 +1040,8 @@ static void r600_gfx_init(struct drm_device *dev, if (((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RV610) || ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RV620) || - ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RS780)) + ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RS780) || + ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RS880)) RADEON_WRITE(R600_VGT_CACHE_INVALIDATION, R600_CACHE_INVALIDATION(R600_TC_ONLY)); else RADEON_WRITE(R600_VGT_CACHE_INVALIDATION, R600_CACHE_INVALIDATION(R600_VC_AND_TC)); @@ -1078,6 +1084,7 @@ static void r600_gfx_init(struct drm_device *dev, break; case CHIP_RV610: case CHIP_RS780: + case CHIP_RS880: case CHIP_RV620: gs_prim_buffer_depth = 32; break; @@ -1123,6 +1130,7 @@ static void r600_gfx_init(struct drm_device *dev, switch (dev_priv->flags & RADEON_FAMILY_MASK) { case CHIP_RV610: case CHIP_RS780: + case CHIP_RS880: case CHIP_RV620: tc_cntl = R600_TC_L2_SIZE(8); break; diff --git a/drivers/gpu/drm/radeon/radeon_drv.h b/drivers/gpu/drm/radeon/radeon_drv.h index 127d0456f628..3933f8216a34 100644 --- a/drivers/gpu/drm/radeon/radeon_drv.h +++ b/drivers/gpu/drm/radeon/radeon_drv.h @@ -143,6 +143,7 @@ enum radeon_family { CHIP_RV635, CHIP_RV670, CHIP_RS780, + CHIP_RS880, CHIP_RV770, CHIP_RV730, CHIP_RV710, diff --git a/include/drm/drm_pciids.h b/include/drm/drm_pciids.h index 9d4c00491547..853508499d20 100644 --- a/include/drm/drm_pciids.h +++ b/include/drm/drm_pciids.h @@ -370,6 +370,11 @@ {0x1002, 0x9614, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RS780|RADEON_NEW_MEMMAP|RADEON_IS_IGP}, \ {0x1002, 0x9615, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RS780|RADEON_NEW_MEMMAP|RADEON_IS_IGP}, \ {0x1002, 0x9616, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RS780|RADEON_NEW_MEMMAP|RADEON_IS_IGP}, \ + {0x1002, 0x9710, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RS880|RADEON_NEW_MEMMAP|RADEON_IS_IGP}, \ + {0x1002, 0x9711, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RS880|RADEON_NEW_MEMMAP|RADEON_IS_IGP}, \ + {0x1002, 0x9712, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RS880|RADEON_IS_MOBILITY|RADEON_NEW_MEMMAP|RADEON_IS_IGP}, \ + {0x1002, 0x9713, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RS880|RADEON_IS_MOBILITY|RADEON_NEW_MEMMAP|RADEON_IS_IGP}, \ + {0x1002, 0x9714, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RS880|RADEON_NEW_MEMMAP|RADEON_IS_IGP}, \ {0, 0, 0} #define r128_PCI_IDS \ -- cgit v1.2.3 From 5116d8f6b977970ebefc1932c0f313163a6ec91f Mon Sep 17 00:00:00 2001 From: "Michael S. Tsirkin" Date: Sun, 26 Jul 2009 17:10:01 +0300 Subject: KVM: fix ack not being delivered when msi present kvm_notify_acked_irq does not check irq type, so that it sometimes interprets msi vector as irq. As a result, ack notifiers are not called, which typially hangs the guest. The fix is to track and check irq type. Signed-off-by: Michael S. Tsirkin Signed-off-by: Avi Kivity --- include/linux/kvm_host.h | 1 + virt/kvm/irq_comm.c | 4 +++- 2 files changed, 4 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h index 16713dc672e4..3060bdc35ffe 100644 --- a/include/linux/kvm_host.h +++ b/include/linux/kvm_host.h @@ -110,6 +110,7 @@ struct kvm_memory_slot { struct kvm_kernel_irq_routing_entry { u32 gsi; + u32 type; int (*set)(struct kvm_kernel_irq_routing_entry *e, struct kvm *kvm, int level); union { diff --git a/virt/kvm/irq_comm.c b/virt/kvm/irq_comm.c index a8bd466d00cc..ddc17f0e2f35 100644 --- a/virt/kvm/irq_comm.c +++ b/virt/kvm/irq_comm.c @@ -160,7 +160,8 @@ void kvm_notify_acked_irq(struct kvm *kvm, unsigned irqchip, unsigned pin) unsigned gsi = pin; list_for_each_entry(e, &kvm->irq_routing, link) - if (e->irqchip.irqchip == irqchip && + if (e->type == KVM_IRQ_ROUTING_IRQCHIP && + e->irqchip.irqchip == irqchip && e->irqchip.pin == pin) { gsi = e->gsi; break; @@ -259,6 +260,7 @@ static int setup_routing_entry(struct kvm_kernel_irq_routing_entry *e, int delta; e->gsi = ue->gsi; + e->type = ue->type; switch (ue->type) { case KVM_IRQ_ROUTING_IRQCHIP: delta = 0; -- cgit v1.2.3 From 36cbd3dcc10384f813ec0814255f576c84f2bcd4 Mon Sep 17 00:00:00 2001 From: Jan Engelhardt Date: Wed, 5 Aug 2009 10:42:58 -0700 Subject: net: mark read-only arrays as const String literals are constant, and usually, we can also tag the array of pointers const too, moving it to the .rodata section. Signed-off-by: Jan Engelhardt Signed-off-by: David S. Miller --- include/net/ip_vs.h | 3 ++- include/net/irda/ircomm_event.h | 2 +- include/net/irda/ircomm_tty_attach.h | 4 ++-- include/net/irda/irlap_event.h | 2 +- include/net/irda/irlmp_event.h | 4 ++-- include/net/sctp/constants.h | 4 +++- net/8021q/vlanproc.c | 2 +- net/atm/lec.c | 9 +++++---- net/atm/proc.c | 9 +++++---- net/bluetooth/af_bluetooth.c | 4 ++-- net/bridge/br_stp.c | 2 +- net/core/dev.c | 2 +- net/core/net-sysfs.c | 2 +- net/core/sock.c | 6 +++--- net/dccp/ccids/ccid3.c | 4 ++-- net/dccp/feat.c | 7 ++++--- net/dccp/proto.c | 4 ++-- net/ipv4/fib_trie.c | 2 +- net/ipv6/proc.c | 4 ++-- net/irda/ircomm/ircomm_event.c | 4 ++-- net/irda/ircomm/ircomm_tty_attach.c | 4 ++-- net/irda/iriap.c | 4 ++-- net/irda/irlan/irlan_common.c | 4 ++-- net/irda/irlap.c | 2 +- net/irda/irlap_event.c | 4 ++-- net/irda/irlmp_event.c | 6 +++--- net/llc/llc_proc.c | 2 +- net/netfilter/ipvs/ip_vs_proto.c | 3 ++- net/netfilter/ipvs/ip_vs_proto_tcp.c | 2 +- net/netfilter/ipvs/ip_vs_proto_udp.c | 2 +- net/rds/ib_stats.c | 2 +- net/rds/iw_stats.c | 2 +- net/rds/rds.h | 3 ++- net/rds/stats.c | 4 ++-- net/rxrpc/ar-ack.c | 2 +- net/sctp/debug.c | 14 +++++++------- 36 files changed, 74 insertions(+), 66 deletions(-) (limited to 'include') diff --git a/include/net/ip_vs.h b/include/net/ip_vs.h index 1c8ee1b13651..98978e73f666 100644 --- a/include/net/ip_vs.h +++ b/include/net/ip_vs.h @@ -738,7 +738,8 @@ extern void ip_vs_protocol_cleanup(void); extern void ip_vs_protocol_timeout_change(int flags); extern int *ip_vs_create_timeout_table(int *table, int size); extern int -ip_vs_set_state_timeout(int *table, int num, char **names, char *name, int to); +ip_vs_set_state_timeout(int *table, int num, const char *const *names, + const char *name, int to); extern void ip_vs_tcpudp_debug_packet(struct ip_vs_protocol *pp, const struct sk_buff *skb, int offset, const char *msg); diff --git a/include/net/irda/ircomm_event.h b/include/net/irda/ircomm_event.h index c290447872d1..bc0c6f31f1c6 100644 --- a/include/net/irda/ircomm_event.h +++ b/include/net/irda/ircomm_event.h @@ -74,7 +74,7 @@ struct ircomm_info { struct qos_info *qos; }; -extern char *ircomm_state[]; +extern const char *const ircomm_state[]; struct ircomm_cb; /* Forward decl. */ diff --git a/include/net/irda/ircomm_tty_attach.h b/include/net/irda/ircomm_tty_attach.h index f91a5695aa44..0a63bbb972d7 100644 --- a/include/net/irda/ircomm_tty_attach.h +++ b/include/net/irda/ircomm_tty_attach.h @@ -66,8 +66,8 @@ struct ircomm_tty_info { __u8 dlsap_sel; }; -extern char *ircomm_state[]; -extern char *ircomm_tty_state[]; +extern const char *const ircomm_state[]; +extern const char *const ircomm_tty_state[]; int ircomm_tty_do_event(struct ircomm_tty_cb *self, IRCOMM_TTY_EVENT event, struct sk_buff *skb, struct ircomm_tty_info *info); diff --git a/include/net/irda/irlap_event.h b/include/net/irda/irlap_event.h index 2ae2e119ef4b..4c90824c50fb 100644 --- a/include/net/irda/irlap_event.h +++ b/include/net/irda/irlap_event.h @@ -120,7 +120,7 @@ typedef enum { /* FIXME check the two first reason codes */ LAP_PRIMARY_CONFLICT, } LAP_REASON; -extern const char *irlap_state[]; +extern const char *const irlap_state[]; void irlap_do_event(struct irlap_cb *self, IRLAP_EVENT event, struct sk_buff *skb, struct irlap_info *info); diff --git a/include/net/irda/irlmp_event.h b/include/net/irda/irlmp_event.h index e03ae4ae3963..9e4ec17a7449 100644 --- a/include/net/irda/irlmp_event.h +++ b/include/net/irda/irlmp_event.h @@ -79,8 +79,8 @@ typedef enum { LM_LAP_IDLE_TIMEOUT, } IRLMP_EVENT; -extern const char *irlmp_state[]; -extern const char *irlsap_state[]; +extern const char *const irlmp_state[]; +extern const char *const irlsap_state[]; void irlmp_watchdog_timer_expired(void *data); void irlmp_discovery_timer_expired(void *data); diff --git a/include/net/sctp/constants.h b/include/net/sctp/constants.h index b05b0557211f..8bc25f7b04ce 100644 --- a/include/net/sctp/constants.h +++ b/include/net/sctp/constants.h @@ -241,7 +241,9 @@ const char *sctp_tname(const sctp_subtype_t); /* timeouts */ const char *sctp_pname(const sctp_subtype_t); /* primitives */ /* This is a table of printable names of sctp_state_t's. */ -extern const char *sctp_state_tbl[], *sctp_evttype_tbl[], *sctp_status_tbl[]; +extern const char *const sctp_state_tbl[]; +extern const char *const sctp_evttype_tbl[]; +extern const char *const sctp_status_tbl[]; /* Maximum chunk length considering padding requirements. */ enum { SCTP_MAX_CHUNK_LEN = ((1<<16) - sizeof(__u32)) }; diff --git a/net/8021q/vlanproc.c b/net/8021q/vlanproc.c index b55a091a33df..6262c335f3c2 100644 --- a/net/8021q/vlanproc.c +++ b/net/8021q/vlanproc.c @@ -107,7 +107,7 @@ static const struct file_operations vlandev_fops = { */ /* Strings */ -static const char *vlan_name_type_str[VLAN_NAME_TYPE_HIGHEST] = { +static const char *const vlan_name_type_str[VLAN_NAME_TYPE_HIGHEST] = { [VLAN_NAME_TYPE_RAW_PLUS_VID] = "VLAN_NAME_TYPE_RAW_PLUS_VID", [VLAN_NAME_TYPE_PLUS_VID_NO_PAD] = "VLAN_NAME_TYPE_PLUS_VID_NO_PAD", [VLAN_NAME_TYPE_RAW_PLUS_VID_NO_PAD] = "VLAN_NAME_TYPE_RAW_PLUS_VID_NO_PAD", diff --git a/net/atm/lec.c b/net/atm/lec.c index c463868c993b..8e723c2654cb 100644 --- a/net/atm/lec.c +++ b/net/atm/lec.c @@ -935,9 +935,9 @@ static int lecd_attach(struct atm_vcc *vcc, int arg) } #ifdef CONFIG_PROC_FS -static char *lec_arp_get_status_string(unsigned char status) +static const char *lec_arp_get_status_string(unsigned char status) { - static char *lec_arp_status_string[] = { + static const char *const lec_arp_status_string[] = { "ESI_UNKNOWN ", "ESI_ARP_PENDING ", "ESI_VC_PENDING ", @@ -1121,7 +1121,8 @@ static void *lec_seq_next(struct seq_file *seq, void *v, loff_t *pos) static int lec_seq_show(struct seq_file *seq, void *v) { - static char lec_banner[] = "Itf MAC ATM destination" + static const char lec_banner[] = + "Itf MAC ATM destination" " Status Flags " "VPI/VCI Recv VPI/VCI\n"; @@ -1505,7 +1506,7 @@ lec_arp_remove(struct lec_priv *priv, struct lec_arp_table *to_remove) } #if DEBUG_ARP_TABLE -static char *get_status_string(unsigned char st) +static const char *get_status_string(unsigned char st) { switch (st) { case ESI_UNKNOWN: diff --git a/net/atm/proc.c b/net/atm/proc.c index 38de5ff61ecd..ab8419a324b6 100644 --- a/net/atm/proc.c +++ b/net/atm/proc.c @@ -151,8 +151,9 @@ static void *vcc_seq_next(struct seq_file *seq, void *v, loff_t *pos) static void pvc_info(struct seq_file *seq, struct atm_vcc *vcc) { - static const char *class_name[] = { "off","UBR","CBR","VBR","ABR" }; - static const char *aal_name[] = { + static const char *const class_name[] = + {"off","UBR","CBR","VBR","ABR"}; + static const char *const aal_name[] = { "---", "1", "2", "3/4", /* 0- 3 */ "???", "5", "???", "???", /* 4- 7 */ "???", "???", "???", "???", /* 8-11 */ @@ -178,7 +179,7 @@ static void pvc_info(struct seq_file *seq, struct atm_vcc *vcc) static const char *vcc_state(struct atm_vcc *vcc) { - static const char *map[] = { ATM_VS2TXT_MAP }; + static const char *const map[] = { ATM_VS2TXT_MAP }; return map[ATM_VF2VS(vcc->flags)]; } @@ -335,7 +336,7 @@ static const struct file_operations vcc_seq_fops = { static int svc_seq_show(struct seq_file *seq, void *v) { - static char atm_svc_banner[] = + static const char atm_svc_banner[] = "Itf VPI VCI State Remote\n"; if (v == SEQ_START_TOKEN) diff --git a/net/bluetooth/af_bluetooth.c b/net/bluetooth/af_bluetooth.c index 0250e0600150..8cfb5a849841 100644 --- a/net/bluetooth/af_bluetooth.c +++ b/net/bluetooth/af_bluetooth.c @@ -49,7 +49,7 @@ static struct net_proto_family *bt_proto[BT_MAX_PROTO]; static DEFINE_RWLOCK(bt_proto_lock); static struct lock_class_key bt_lock_key[BT_MAX_PROTO]; -static const char *bt_key_strings[BT_MAX_PROTO] = { +static const char *const bt_key_strings[BT_MAX_PROTO] = { "sk_lock-AF_BLUETOOTH-BTPROTO_L2CAP", "sk_lock-AF_BLUETOOTH-BTPROTO_HCI", "sk_lock-AF_BLUETOOTH-BTPROTO_SCO", @@ -61,7 +61,7 @@ static const char *bt_key_strings[BT_MAX_PROTO] = { }; static struct lock_class_key bt_slock_key[BT_MAX_PROTO]; -static const char *bt_slock_key_strings[BT_MAX_PROTO] = { +static const char *const bt_slock_key_strings[BT_MAX_PROTO] = { "slock-AF_BLUETOOTH-BTPROTO_L2CAP", "slock-AF_BLUETOOTH-BTPROTO_HCI", "slock-AF_BLUETOOTH-BTPROTO_SCO", diff --git a/net/bridge/br_stp.c b/net/bridge/br_stp.c index 0660515f3992..fd3f8d6c0998 100644 --- a/net/bridge/br_stp.c +++ b/net/bridge/br_stp.c @@ -21,7 +21,7 @@ */ #define MESSAGE_AGE_INCR ((HZ < 256) ? 1 : (HZ/256)) -static const char *br_port_state_names[] = { +static const char *const br_port_state_names[] = { [BR_STATE_DISABLED] = "disabled", [BR_STATE_LISTENING] = "listening", [BR_STATE_LEARNING] = "learning", diff --git a/net/core/dev.c b/net/core/dev.c index 71347668c506..f01a9c41f112 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -272,7 +272,7 @@ static const unsigned short netdev_lock_type[] = ARPHRD_PHONET_PIPE, ARPHRD_IEEE802154, ARPHRD_IEEE802154_PHY, ARPHRD_VOID, ARPHRD_NONE}; -static const char *netdev_lock_name[] = +static const char *const netdev_lock_name[] = {"_xmit_NETROM", "_xmit_ETHER", "_xmit_EETHER", "_xmit_AX25", "_xmit_PRONET", "_xmit_CHAOS", "_xmit_IEEE802", "_xmit_ARCNET", "_xmit_APPLETLK", "_xmit_DLCI", "_xmit_ATM", "_xmit_METRICOM", diff --git a/net/core/net-sysfs.c b/net/core/net-sysfs.c index 3994680c08b9..ad91e9e5f475 100644 --- a/net/core/net-sysfs.c +++ b/net/core/net-sysfs.c @@ -141,7 +141,7 @@ static ssize_t show_dormant(struct device *dev, return -EINVAL; } -static const char *operstates[] = { +static const char *const operstates[] = { "unknown", "notpresent", /* currently unused */ "down", diff --git a/net/core/sock.c b/net/core/sock.c index bbb25be7ddfe..a324a80c163e 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -142,7 +142,7 @@ static struct lock_class_key af_family_slock_keys[AF_MAX]; * strings build-time, so that runtime initialization of socket * locks is fast): */ -static const char *af_family_key_strings[AF_MAX+1] = { +static const char *const af_family_key_strings[AF_MAX+1] = { "sk_lock-AF_UNSPEC", "sk_lock-AF_UNIX" , "sk_lock-AF_INET" , "sk_lock-AF_AX25" , "sk_lock-AF_IPX" , "sk_lock-AF_APPLETALK", "sk_lock-AF_NETROM", "sk_lock-AF_BRIDGE" , "sk_lock-AF_ATMPVC" , @@ -158,7 +158,7 @@ static const char *af_family_key_strings[AF_MAX+1] = { "sk_lock-AF_IEEE802154", "sk_lock-AF_MAX" }; -static const char *af_family_slock_key_strings[AF_MAX+1] = { +static const char *const af_family_slock_key_strings[AF_MAX+1] = { "slock-AF_UNSPEC", "slock-AF_UNIX" , "slock-AF_INET" , "slock-AF_AX25" , "slock-AF_IPX" , "slock-AF_APPLETALK", "slock-AF_NETROM", "slock-AF_BRIDGE" , "slock-AF_ATMPVC" , @@ -174,7 +174,7 @@ static const char *af_family_slock_key_strings[AF_MAX+1] = { "slock-AF_IEEE802154", "slock-AF_MAX" }; -static const char *af_family_clock_key_strings[AF_MAX+1] = { +static const char *const af_family_clock_key_strings[AF_MAX+1] = { "clock-AF_UNSPEC", "clock-AF_UNIX" , "clock-AF_INET" , "clock-AF_AX25" , "clock-AF_IPX" , "clock-AF_APPLETALK", "clock-AF_NETROM", "clock-AF_BRIDGE" , "clock-AF_ATMPVC" , diff --git a/net/dccp/ccids/ccid3.c b/net/dccp/ccids/ccid3.c index a27b7f4c19c5..f596ce149c3c 100644 --- a/net/dccp/ccids/ccid3.c +++ b/net/dccp/ccids/ccid3.c @@ -52,7 +52,7 @@ static int ccid3_debug; #ifdef CONFIG_IP_DCCP_CCID3_DEBUG static const char *ccid3_tx_state_name(enum ccid3_hc_tx_states state) { - static char *ccid3_state_names[] = { + static const char *const ccid3_state_names[] = { [TFRC_SSTATE_NO_SENT] = "NO_SENT", [TFRC_SSTATE_NO_FBACK] = "NO_FBACK", [TFRC_SSTATE_FBACK] = "FBACK", @@ -646,7 +646,7 @@ enum ccid3_fback_type { #ifdef CONFIG_IP_DCCP_CCID3_DEBUG static const char *ccid3_rx_state_name(enum ccid3_hc_rx_states state) { - static char *ccid3_rx_state_names[] = { + static const char *const ccid3_rx_state_names[] = { [TFRC_RSTATE_NO_DATA] = "NO_DATA", [TFRC_RSTATE_DATA] = "DATA", [TFRC_RSTATE_TERM] = "TERM", diff --git a/net/dccp/feat.c b/net/dccp/feat.c index b04160a2eea5..972b8dc918d6 100644 --- a/net/dccp/feat.c +++ b/net/dccp/feat.c @@ -213,7 +213,7 @@ static int dccp_feat_default_value(u8 feat_num) */ static const char *dccp_feat_fname(const u8 feat) { - static const char *feature_names[] = { + static const char *const feature_names[] = { [DCCPF_RESERVED] = "Reserved", [DCCPF_CCID] = "CCID", [DCCPF_SHORT_SEQNOS] = "Allow Short Seqnos", @@ -236,8 +236,9 @@ static const char *dccp_feat_fname(const u8 feat) return feature_names[feat]; } -static const char *dccp_feat_sname[] = { "DEFAULT", "INITIALISING", "CHANGING", - "UNSTABLE", "STABLE" }; +static const char *const dccp_feat_sname[] = { + "DEFAULT", "INITIALISING", "CHANGING", "UNSTABLE", "STABLE", +}; #ifdef CONFIG_IP_DCCP_DEBUG static const char *dccp_feat_oname(const u8 opt) diff --git a/net/dccp/proto.c b/net/dccp/proto.c index 94ca8eaace7d..37b3b4293ef4 100644 --- a/net/dccp/proto.c +++ b/net/dccp/proto.c @@ -124,7 +124,7 @@ EXPORT_SYMBOL_GPL(dccp_done); const char *dccp_packet_name(const int type) { - static const char *dccp_packet_names[] = { + static const char *const dccp_packet_names[] = { [DCCP_PKT_REQUEST] = "REQUEST", [DCCP_PKT_RESPONSE] = "RESPONSE", [DCCP_PKT_DATA] = "DATA", @@ -147,7 +147,7 @@ EXPORT_SYMBOL_GPL(dccp_packet_name); const char *dccp_state_name(const int state) { - static char *dccp_state_names[] = { + static const char *const dccp_state_names[] = { [DCCP_OPEN] = "OPEN", [DCCP_REQUESTING] = "REQUESTING", [DCCP_PARTOPEN] = "PARTOPEN", diff --git a/net/ipv4/fib_trie.c b/net/ipv4/fib_trie.c index d58b49115386..fe3c846b99a6 100644 --- a/net/ipv4/fib_trie.c +++ b/net/ipv4/fib_trie.c @@ -2421,7 +2421,7 @@ static inline const char *rtn_scope(char *buf, size_t len, enum rt_scope_t s) } } -static const char *rtn_type_names[__RTN_MAX] = { +static const char *const rtn_type_names[__RTN_MAX] = { [RTN_UNSPEC] = "UNSPEC", [RTN_UNICAST] = "UNICAST", [RTN_LOCAL] = "LOCAL", diff --git a/net/ipv6/proc.c b/net/ipv6/proc.c index 590ddefb7ffc..c9605c3ad91f 100644 --- a/net/ipv6/proc.c +++ b/net/ipv6/proc.c @@ -101,7 +101,7 @@ static struct snmp_mib snmp6_icmp6_list[] = { }; /* RFC 4293 v6 ICMPMsgStatsTable; named items for RFC 2466 compatibility */ -static char *icmp6type2name[256] = { +static const char *const icmp6type2name[256] = { [ICMPV6_DEST_UNREACH] = "DestUnreachs", [ICMPV6_PKT_TOOBIG] = "PktTooBigs", [ICMPV6_TIME_EXCEED] = "TimeExcds", @@ -144,7 +144,7 @@ static void snmp6_seq_show_icmpv6msg(struct seq_file *seq, void **mib) /* print by name -- deprecated items */ for (i = 0; i < ICMP6MSG_MIB_MAX; i++) { int icmptype; - char *p; + const char *p; icmptype = i & 0xff; p = icmp6type2name[icmptype]; diff --git a/net/irda/ircomm/ircomm_event.c b/net/irda/ircomm/ircomm_event.c index c35b3ef5c2f0..d78554fedbac 100644 --- a/net/irda/ircomm/ircomm_event.c +++ b/net/irda/ircomm/ircomm_event.c @@ -49,7 +49,7 @@ static int ircomm_state_waitr(struct ircomm_cb *self, IRCOMM_EVENT event, static int ircomm_state_conn(struct ircomm_cb *self, IRCOMM_EVENT event, struct sk_buff *skb, struct ircomm_info *info); -char *ircomm_state[] = { +const char *const ircomm_state[] = { "IRCOMM_IDLE", "IRCOMM_WAITI", "IRCOMM_WAITR", @@ -57,7 +57,7 @@ char *ircomm_state[] = { }; #ifdef CONFIG_IRDA_DEBUG -static char *ircomm_event[] = { +static const char *const ircomm_event[] = { "IRCOMM_CONNECT_REQUEST", "IRCOMM_CONNECT_RESPONSE", "IRCOMM_TTP_CONNECT_INDICATION", diff --git a/net/irda/ircomm/ircomm_tty_attach.c b/net/irda/ircomm/ircomm_tty_attach.c index 9032a1d1190d..eafc010907c2 100644 --- a/net/irda/ircomm/ircomm_tty_attach.c +++ b/net/irda/ircomm/ircomm_tty_attach.c @@ -80,7 +80,7 @@ static int ircomm_tty_state_ready(struct ircomm_tty_cb *self, struct sk_buff *skb, struct ircomm_tty_info *info); -char *ircomm_tty_state[] = { +const char *const ircomm_tty_state[] = { "IRCOMM_TTY_IDLE", "IRCOMM_TTY_SEARCH", "IRCOMM_TTY_QUERY_PARAMETERS", @@ -91,7 +91,7 @@ char *ircomm_tty_state[] = { }; #ifdef CONFIG_IRDA_DEBUG -static char *ircomm_tty_event[] = { +static const char *const ircomm_tty_event[] = { "IRCOMM_TTY_ATTACH_CABLE", "IRCOMM_TTY_DETACH_CABLE", "IRCOMM_TTY_DATA_REQUEST", diff --git a/net/irda/iriap.c b/net/irda/iriap.c index 4a105dc32dcd..294e34d3517c 100644 --- a/net/irda/iriap.c +++ b/net/irda/iriap.c @@ -44,7 +44,7 @@ #ifdef CONFIG_IRDA_DEBUG /* FIXME: This one should go in irlmp.c */ -static const char *ias_charset_types[] = { +static const char *const ias_charset_types[] = { "CS_ASCII", "CS_ISO_8859_1", "CS_ISO_8859_2", @@ -966,7 +966,7 @@ static void iriap_watchdog_timer_expired(void *data) #ifdef CONFIG_PROC_FS -static const char *ias_value_types[] = { +static const char *const ias_value_types[] = { "IAS_MISSING", "IAS_INTEGER", "IAS_OCT_SEQ", diff --git a/net/irda/irlan/irlan_common.c b/net/irda/irlan/irlan_common.c index 774d73a76852..62116829b817 100644 --- a/net/irda/irlan/irlan_common.c +++ b/net/irda/irlan/irlan_common.c @@ -69,14 +69,14 @@ static int eth; /* Use "eth" or "irlan" name for devices */ static int access = ACCESS_PEER; /* PEER, DIRECT or HOSTED */ #ifdef CONFIG_PROC_FS -static const char *irlan_access[] = { +static const char *const irlan_access[] = { "UNKNOWN", "DIRECT", "PEER", "HOSTED" }; -static const char *irlan_media[] = { +static const char *const irlan_media[] = { "UNKNOWN", "802.3", "802.5" diff --git a/net/irda/irlap.c b/net/irda/irlap.c index e4965b764b9b..356e65b1dc42 100644 --- a/net/irda/irlap.c +++ b/net/irda/irlap.c @@ -63,7 +63,7 @@ static void irlap_init_qos_capabilities(struct irlap_cb *self, struct qos_info *qos_user); #ifdef CONFIG_IRDA_DEBUG -static char *lap_reasons[] = { +static const char *const lap_reasons[] = { "ERROR, NOT USED", "LAP_DISC_INDICATION", "LAP_NO_RESPONSE", diff --git a/net/irda/irlap_event.c b/net/irda/irlap_event.c index 16c4ef0f5c1a..c5c51959e3ce 100644 --- a/net/irda/irlap_event.c +++ b/net/irda/irlap_event.c @@ -78,7 +78,7 @@ static int irlap_state_reset_check(struct irlap_cb *, IRLAP_EVENT event, struct sk_buff *, struct irlap_info *); #ifdef CONFIG_IRDA_DEBUG -static const char *irlap_event[] = { +static const char *const irlap_event[] = { "DISCOVERY_REQUEST", "CONNECT_REQUEST", "CONNECT_RESPONSE", @@ -120,7 +120,7 @@ static const char *irlap_event[] = { }; #endif /* CONFIG_IRDA_DEBUG */ -const char *irlap_state[] = { +const char *const irlap_state[] = { "LAP_NDM", "LAP_QUERY", "LAP_REPLY", diff --git a/net/irda/irlmp_event.c b/net/irda/irlmp_event.c index 78cce0cb073f..c1fb5db81042 100644 --- a/net/irda/irlmp_event.c +++ b/net/irda/irlmp_event.c @@ -33,13 +33,13 @@ #include #include -const char *irlmp_state[] = { +const char *const irlmp_state[] = { "LAP_STANDBY", "LAP_U_CONNECT", "LAP_ACTIVE", }; -const char *irlsap_state[] = { +const char *const irlsap_state[] = { "LSAP_DISCONNECTED", "LSAP_CONNECT", "LSAP_CONNECT_PEND", @@ -49,7 +49,7 @@ const char *irlsap_state[] = { }; #ifdef CONFIG_IRDA_DEBUG -static const char *irlmp_event[] = { +static const char *const irlmp_event[] = { "LM_CONNECT_REQUEST", "LM_CONNECT_CONFIRM", "LM_CONNECT_RESPONSE", diff --git a/net/llc/llc_proc.c b/net/llc/llc_proc.c index f97be471fe2e..be47ac427f6b 100644 --- a/net/llc/llc_proc.c +++ b/net/llc/llc_proc.c @@ -143,7 +143,7 @@ out: return 0; } -static char *llc_conn_state_names[] = { +static const char *const llc_conn_state_names[] = { [LLC_CONN_STATE_ADM] = "adm", [LLC_CONN_STATE_SETUP] = "setup", [LLC_CONN_STATE_NORMAL] = "normal", diff --git a/net/netfilter/ipvs/ip_vs_proto.c b/net/netfilter/ipvs/ip_vs_proto.c index 85c8892e1e8b..3e7671674549 100644 --- a/net/netfilter/ipvs/ip_vs_proto.c +++ b/net/netfilter/ipvs/ip_vs_proto.c @@ -127,7 +127,8 @@ ip_vs_create_timeout_table(int *table, int size) * Set timeout value for state specified by name */ int -ip_vs_set_state_timeout(int *table, int num, char **names, char *name, int to) +ip_vs_set_state_timeout(int *table, int num, const char *const *names, + const char *name, int to) { int i; diff --git a/net/netfilter/ipvs/ip_vs_proto_tcp.c b/net/netfilter/ipvs/ip_vs_proto_tcp.c index 2278e141489e..91d28e073742 100644 --- a/net/netfilter/ipvs/ip_vs_proto_tcp.c +++ b/net/netfilter/ipvs/ip_vs_proto_tcp.c @@ -377,7 +377,7 @@ static int tcp_timeouts[IP_VS_TCP_S_LAST+1] = { [IP_VS_TCP_S_LAST] = 2*HZ, }; -static char * tcp_state_name_table[IP_VS_TCP_S_LAST+1] = { +static const char *const tcp_state_name_table[IP_VS_TCP_S_LAST+1] = { [IP_VS_TCP_S_NONE] = "NONE", [IP_VS_TCP_S_ESTABLISHED] = "ESTABLISHED", [IP_VS_TCP_S_SYN_SENT] = "SYN_SENT", diff --git a/net/netfilter/ipvs/ip_vs_proto_udp.c b/net/netfilter/ipvs/ip_vs_proto_udp.c index 33a05d3684d9..e7a6885e0167 100644 --- a/net/netfilter/ipvs/ip_vs_proto_udp.c +++ b/net/netfilter/ipvs/ip_vs_proto_udp.c @@ -472,7 +472,7 @@ static int udp_timeouts[IP_VS_UDP_S_LAST+1] = { [IP_VS_UDP_S_LAST] = 2*HZ, }; -static char * udp_state_name_table[IP_VS_UDP_S_LAST+1] = { +static const char *const udp_state_name_table[IP_VS_UDP_S_LAST+1] = { [IP_VS_UDP_S_NORMAL] = "UDP", [IP_VS_UDP_S_LAST] = "BUG!", }; diff --git a/net/rds/ib_stats.c b/net/rds/ib_stats.c index 02e3e3d50d4a..8d8488306fe4 100644 --- a/net/rds/ib_stats.c +++ b/net/rds/ib_stats.c @@ -39,7 +39,7 @@ DEFINE_PER_CPU(struct rds_ib_statistics, rds_ib_stats) ____cacheline_aligned; -static char *rds_ib_stat_names[] = { +static const char *const rds_ib_stat_names[] = { "ib_connect_raced", "ib_listen_closed_stale", "ib_tx_cq_call", diff --git a/net/rds/iw_stats.c b/net/rds/iw_stats.c index ccc7e8f0bf0e..d33ea790484e 100644 --- a/net/rds/iw_stats.c +++ b/net/rds/iw_stats.c @@ -39,7 +39,7 @@ DEFINE_PER_CPU(struct rds_iw_statistics, rds_iw_stats) ____cacheline_aligned; -static char *rds_iw_stat_names[] = { +static const char *const rds_iw_stat_names[] = { "iw_connect_raced", "iw_listen_closed_stale", "iw_tx_cq_call", diff --git a/net/rds/rds.h b/net/rds/rds.h index dbe111236783..290566c69d28 100644 --- a/net/rds/rds.h +++ b/net/rds/rds.h @@ -652,7 +652,8 @@ DECLARE_PER_CPU_SHARED_ALIGNED(struct rds_statistics, rds_stats); int __init rds_stats_init(void); void rds_stats_exit(void); void rds_stats_info_copy(struct rds_info_iterator *iter, - uint64_t *values, char **names, size_t nr); + uint64_t *values, const char *const *names, + size_t nr); /* sysctl.c */ int __init rds_sysctl_init(void); diff --git a/net/rds/stats.c b/net/rds/stats.c index 637146893cf3..91d8c58b8335 100644 --- a/net/rds/stats.c +++ b/net/rds/stats.c @@ -40,7 +40,7 @@ DEFINE_PER_CPU_SHARED_ALIGNED(struct rds_statistics, rds_stats); /* :.,$s/unsigned long\>.*\ Date: Tue, 4 Aug 2009 07:28:28 +0000 Subject: net: implement a SO_PROTOCOL getsockoption Similar to SO_TYPE returning the socket type, SO_PROTOCOL allows to retrieve the protocol used with a given socket. I am not quite sure why we have that-many copies of socket.h, and why the values are not the same on all arches either, but for where hex numbers dominate, I use 0x1029 for SO_PROTOCOL as that seems to be the next free unused number across a bunch of operating systems, or so Google results make me want to believe. SO_PROTOCOL for others just uses the next free Linux number, 38. Signed-off-by: Jan Engelhardt Signed-off-by: David S. Miller --- arch/alpha/include/asm/socket.h | 1 + arch/arm/include/asm/socket.h | 2 ++ arch/avr32/include/asm/socket.h | 2 ++ arch/cris/include/asm/socket.h | 2 ++ arch/frv/include/asm/socket.h | 2 ++ arch/h8300/include/asm/socket.h | 2 ++ arch/ia64/include/asm/socket.h | 2 ++ arch/m32r/include/asm/socket.h | 2 ++ arch/m68k/include/asm/socket.h | 2 ++ arch/microblaze/include/asm/socket.h | 2 ++ arch/mips/include/asm/socket.h | 1 + arch/mn10300/include/asm/socket.h | 2 ++ arch/parisc/include/asm/socket.h | 1 + arch/powerpc/include/asm/socket.h | 2 ++ arch/s390/include/asm/socket.h | 2 ++ arch/sparc/include/asm/socket.h | 2 ++ arch/x86/include/asm/socket.h | 2 ++ arch/xtensa/include/asm/socket.h | 2 ++ include/asm-generic/socket.h | 2 ++ net/core/sock.c | 5 +++++ 20 files changed, 40 insertions(+) (limited to 'include') diff --git a/arch/alpha/include/asm/socket.h b/arch/alpha/include/asm/socket.h index 3641ec1452f4..2f8b4d377749 100644 --- a/arch/alpha/include/asm/socket.h +++ b/arch/alpha/include/asm/socket.h @@ -32,6 +32,7 @@ #define SO_RCVTIMEO 0x1012 #define SO_SNDTIMEO 0x1013 #define SO_ACCEPTCONN 0x1014 +#define SO_PROTOCOL 0x1028 /* linux-specific, might as well be the same as on i386 */ #define SO_NO_CHECK 11 diff --git a/arch/arm/include/asm/socket.h b/arch/arm/include/asm/socket.h index 537de4e0ef50..7f47454ffbf3 100644 --- a/arch/arm/include/asm/socket.h +++ b/arch/arm/include/asm/socket.h @@ -57,4 +57,6 @@ #define SO_TIMESTAMPING 37 #define SCM_TIMESTAMPING SO_TIMESTAMPING +#define SO_PROTOCOL 38 + #endif /* _ASM_SOCKET_H */ diff --git a/arch/avr32/include/asm/socket.h b/arch/avr32/include/asm/socket.h index 04c860619700..6af2866a4f00 100644 --- a/arch/avr32/include/asm/socket.h +++ b/arch/avr32/include/asm/socket.h @@ -57,4 +57,6 @@ #define SO_TIMESTAMPING 37 #define SCM_TIMESTAMPING SO_TIMESTAMPING +#define SO_PROTOCOL 38 + #endif /* __ASM_AVR32_SOCKET_H */ diff --git a/arch/cris/include/asm/socket.h b/arch/cris/include/asm/socket.h index d5cf74005408..f3859fb0990c 100644 --- a/arch/cris/include/asm/socket.h +++ b/arch/cris/include/asm/socket.h @@ -59,6 +59,8 @@ #define SO_TIMESTAMPING 37 #define SCM_TIMESTAMPING SO_TIMESTAMPING +#define SO_PROTOCOL 38 + #endif /* _ASM_SOCKET_H */ diff --git a/arch/frv/include/asm/socket.h b/arch/frv/include/asm/socket.h index 57c3d4054e8b..8dab3486ffa4 100644 --- a/arch/frv/include/asm/socket.h +++ b/arch/frv/include/asm/socket.h @@ -57,5 +57,7 @@ #define SO_TIMESTAMPING 37 #define SCM_TIMESTAMPING SO_TIMESTAMPING +#define SO_PROTOCOL 38 + #endif /* _ASM_SOCKET_H */ diff --git a/arch/h8300/include/asm/socket.h b/arch/h8300/include/asm/socket.h index 602518a70a1a..ba770d09cd63 100644 --- a/arch/h8300/include/asm/socket.h +++ b/arch/h8300/include/asm/socket.h @@ -57,4 +57,6 @@ #define SO_TIMESTAMPING 37 #define SCM_TIMESTAMPING SO_TIMESTAMPING +#define SO_PROTOCOL 38 + #endif /* _ASM_SOCKET_H */ diff --git a/arch/ia64/include/asm/socket.h b/arch/ia64/include/asm/socket.h index 745421225ec6..091cd9d47d0f 100644 --- a/arch/ia64/include/asm/socket.h +++ b/arch/ia64/include/asm/socket.h @@ -66,4 +66,6 @@ #define SO_TIMESTAMPING 37 #define SCM_TIMESTAMPING SO_TIMESTAMPING +#define SO_PROTOCOL 38 + #endif /* _ASM_IA64_SOCKET_H */ diff --git a/arch/m32r/include/asm/socket.h b/arch/m32r/include/asm/socket.h index be7ed589af5c..d36f5928fb79 100644 --- a/arch/m32r/include/asm/socket.h +++ b/arch/m32r/include/asm/socket.h @@ -57,4 +57,6 @@ #define SO_TIMESTAMPING 37 #define SCM_TIMESTAMPING SO_TIMESTAMPING +#define SO_PROTOCOL 38 + #endif /* _ASM_M32R_SOCKET_H */ diff --git a/arch/m68k/include/asm/socket.h b/arch/m68k/include/asm/socket.h index ca87f938b03f..060cb7ed024f 100644 --- a/arch/m68k/include/asm/socket.h +++ b/arch/m68k/include/asm/socket.h @@ -57,4 +57,6 @@ #define SO_TIMESTAMPING 37 #define SCM_TIMESTAMPING SO_TIMESTAMPING +#define SO_PROTOCOL 38 + #endif /* _ASM_SOCKET_H */ diff --git a/arch/microblaze/include/asm/socket.h b/arch/microblaze/include/asm/socket.h index 825936860314..96bf8bfa935e 100644 --- a/arch/microblaze/include/asm/socket.h +++ b/arch/microblaze/include/asm/socket.h @@ -66,4 +66,6 @@ #define SO_TIMESTAMPING 37 #define SCM_TIMESTAMPING SO_TIMESTAMPING +#define SO_PROTOCOL 38 + #endif /* _ASM_MICROBLAZE_SOCKET_H */ diff --git a/arch/mips/include/asm/socket.h b/arch/mips/include/asm/socket.h index 2abca1780169..289ce5f5f2a3 100644 --- a/arch/mips/include/asm/socket.h +++ b/arch/mips/include/asm/socket.h @@ -42,6 +42,7 @@ To add: #define SO_REUSEPORT 0x0200 /* Allow local address and port reuse. */ #define SO_SNDTIMEO 0x1005 /* send timeout */ #define SO_RCVTIMEO 0x1006 /* receive timeout */ #define SO_ACCEPTCONN 0x1009 +#define SO_PROTOCOL 0x1028 /* protocol type */ /* linux-specific, might as well be the same as on i386 */ #define SO_NO_CHECK 11 diff --git a/arch/mn10300/include/asm/socket.h b/arch/mn10300/include/asm/socket.h index fb5daf438ec9..19d7cf709b77 100644 --- a/arch/mn10300/include/asm/socket.h +++ b/arch/mn10300/include/asm/socket.h @@ -57,4 +57,6 @@ #define SO_TIMESTAMPING 37 #define SCM_TIMESTAMPING SO_TIMESTAMPING +#define SO_PROTOCOL 38 + #endif /* _ASM_SOCKET_H */ diff --git a/arch/parisc/include/asm/socket.h b/arch/parisc/include/asm/socket.h index 885472bf7b78..a658b09df624 100644 --- a/arch/parisc/include/asm/socket.h +++ b/arch/parisc/include/asm/socket.h @@ -24,6 +24,7 @@ #define SO_RCVTIMEO 0x1006 #define SO_ERROR 0x1007 #define SO_TYPE 0x1008 +#define SO_PROTOCOL 0x1028 #define SO_PEERNAME 0x2000 #define SO_NO_CHECK 0x400b diff --git a/arch/powerpc/include/asm/socket.h b/arch/powerpc/include/asm/socket.h index 1e5cfad0e3f7..609049d7117e 100644 --- a/arch/powerpc/include/asm/socket.h +++ b/arch/powerpc/include/asm/socket.h @@ -64,4 +64,6 @@ #define SO_TIMESTAMPING 37 #define SCM_TIMESTAMPING SO_TIMESTAMPING +#define SO_PROTOCOL 38 + #endif /* _ASM_POWERPC_SOCKET_H */ diff --git a/arch/s390/include/asm/socket.h b/arch/s390/include/asm/socket.h index 02330c50241b..65baa9a83abc 100644 --- a/arch/s390/include/asm/socket.h +++ b/arch/s390/include/asm/socket.h @@ -65,4 +65,6 @@ #define SO_TIMESTAMPING 37 #define SCM_TIMESTAMPING SO_TIMESTAMPING +#define SO_PROTOCOL 38 + #endif /* _ASM_SOCKET_H */ diff --git a/arch/sparc/include/asm/socket.h b/arch/sparc/include/asm/socket.h index 982a12f959f4..9cbbfafd0538 100644 --- a/arch/sparc/include/asm/socket.h +++ b/arch/sparc/include/asm/socket.h @@ -29,6 +29,8 @@ #define SO_RCVBUFFORCE 0x100b #define SO_ERROR 0x1007 #define SO_TYPE 0x1008 +#define SO_PROTOCOL 0x1028 + /* Linux specific, keep the same. */ #define SO_NO_CHECK 0x000b diff --git a/arch/x86/include/asm/socket.h b/arch/x86/include/asm/socket.h index ca8bf2cd0ba9..1077d2535a32 100644 --- a/arch/x86/include/asm/socket.h +++ b/arch/x86/include/asm/socket.h @@ -57,4 +57,6 @@ #define SO_TIMESTAMPING 37 #define SCM_TIMESTAMPING SO_TIMESTAMPING +#define SO_PROTOCOL 38 + #endif /* _ASM_X86_SOCKET_H */ diff --git a/arch/xtensa/include/asm/socket.h b/arch/xtensa/include/asm/socket.h index dd1a7a4a1cea..e47f172142f1 100644 --- a/arch/xtensa/include/asm/socket.h +++ b/arch/xtensa/include/asm/socket.h @@ -68,4 +68,6 @@ #define SO_TIMESTAMPING 37 #define SCM_TIMESTAMPING SO_TIMESTAMPING +#define SO_PROTOCOL 38 + #endif /* _XTENSA_SOCKET_H */ diff --git a/include/asm-generic/socket.h b/include/asm-generic/socket.h index 5d79e409241c..7e24d186616e 100644 --- a/include/asm-generic/socket.h +++ b/include/asm-generic/socket.h @@ -60,4 +60,6 @@ #define SO_TIMESTAMPING 37 #define SCM_TIMESTAMPING SO_TIMESTAMPING +#define SO_PROTOCOL 38 + #endif /* __ASM_GENERIC_SOCKET_H */ diff --git a/net/core/sock.c b/net/core/sock.c index a324a80c163e..ebce661234ac 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -482,6 +482,7 @@ int sock_setsockopt(struct socket *sock, int level, int optname, sk->sk_reuse = valbool; break; case SO_TYPE: + case SO_PROTOCOL: case SO_ERROR: ret = -ENOPROTOOPT; break; @@ -764,6 +765,10 @@ int sock_getsockopt(struct socket *sock, int level, int optname, v.val = sk->sk_type; break; + case SO_PROTOCOL: + v.val = sk->sk_protocol; + break; + case SO_ERROR: v.val = -sock_error(sk); if (v.val == 0) -- cgit v1.2.3 From 0d6038ee76f2e06b79d0465807f67e86bf4025de Mon Sep 17 00:00:00 2001 From: Jan Engelhardt Date: Tue, 4 Aug 2009 07:28:29 +0000 Subject: net: implement a SO_DOMAIN getsockoption MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This sockopt goes in line with SO_TYPE and SO_PROTOCOL. It makes it possible for userspace programs to pass around file descriptors — I am referring to arguments-to-functions, but it may even work for the fd passing over UNIX sockets — without needing to also pass the auxiliary information (PF_INET6/IPPROTO_TCP). Signed-off-by: Jan Engelhardt Signed-off-by: David S. Miller --- arch/alpha/include/asm/socket.h | 1 + arch/arm/include/asm/socket.h | 1 + arch/avr32/include/asm/socket.h | 1 + arch/cris/include/asm/socket.h | 1 + arch/frv/include/asm/socket.h | 1 + arch/h8300/include/asm/socket.h | 1 + arch/ia64/include/asm/socket.h | 1 + arch/m32r/include/asm/socket.h | 1 + arch/m68k/include/asm/socket.h | 1 + arch/microblaze/include/asm/socket.h | 1 + arch/mips/include/asm/socket.h | 1 + arch/mn10300/include/asm/socket.h | 1 + arch/parisc/include/asm/socket.h | 1 + arch/powerpc/include/asm/socket.h | 1 + arch/s390/include/asm/socket.h | 1 + arch/sparc/include/asm/socket.h | 1 + arch/x86/include/asm/socket.h | 1 + arch/xtensa/include/asm/socket.h | 1 + include/asm-generic/socket.h | 1 + net/core/sock.c | 5 +++++ 20 files changed, 24 insertions(+) (limited to 'include') diff --git a/arch/alpha/include/asm/socket.h b/arch/alpha/include/asm/socket.h index 2f8b4d377749..26773e3246e2 100644 --- a/arch/alpha/include/asm/socket.h +++ b/arch/alpha/include/asm/socket.h @@ -33,6 +33,7 @@ #define SO_SNDTIMEO 0x1013 #define SO_ACCEPTCONN 0x1014 #define SO_PROTOCOL 0x1028 +#define SO_DOMAIN 0x1029 /* linux-specific, might as well be the same as on i386 */ #define SO_NO_CHECK 11 diff --git a/arch/arm/include/asm/socket.h b/arch/arm/include/asm/socket.h index 7f47454ffbf3..92ac61d294fd 100644 --- a/arch/arm/include/asm/socket.h +++ b/arch/arm/include/asm/socket.h @@ -58,5 +58,6 @@ #define SCM_TIMESTAMPING SO_TIMESTAMPING #define SO_PROTOCOL 38 +#define SO_DOMAIN 39 #endif /* _ASM_SOCKET_H */ diff --git a/arch/avr32/include/asm/socket.h b/arch/avr32/include/asm/socket.h index 6af2866a4f00..fe863f9794d5 100644 --- a/arch/avr32/include/asm/socket.h +++ b/arch/avr32/include/asm/socket.h @@ -58,5 +58,6 @@ #define SCM_TIMESTAMPING SO_TIMESTAMPING #define SO_PROTOCOL 38 +#define SO_DOMAIN 39 #endif /* __ASM_AVR32_SOCKET_H */ diff --git a/arch/cris/include/asm/socket.h b/arch/cris/include/asm/socket.h index f3859fb0990c..45ec49bdb7b1 100644 --- a/arch/cris/include/asm/socket.h +++ b/arch/cris/include/asm/socket.h @@ -60,6 +60,7 @@ #define SCM_TIMESTAMPING SO_TIMESTAMPING #define SO_PROTOCOL 38 +#define SO_DOMAIN 39 #endif /* _ASM_SOCKET_H */ diff --git a/arch/frv/include/asm/socket.h b/arch/frv/include/asm/socket.h index 8dab3486ffa4..2dea726095c2 100644 --- a/arch/frv/include/asm/socket.h +++ b/arch/frv/include/asm/socket.h @@ -58,6 +58,7 @@ #define SCM_TIMESTAMPING SO_TIMESTAMPING #define SO_PROTOCOL 38 +#define SO_DOMAIN 39 #endif /* _ASM_SOCKET_H */ diff --git a/arch/h8300/include/asm/socket.h b/arch/h8300/include/asm/socket.h index ba770d09cd63..1547f01c8e22 100644 --- a/arch/h8300/include/asm/socket.h +++ b/arch/h8300/include/asm/socket.h @@ -58,5 +58,6 @@ #define SCM_TIMESTAMPING SO_TIMESTAMPING #define SO_PROTOCOL 38 +#define SO_DOMAIN 39 #endif /* _ASM_SOCKET_H */ diff --git a/arch/ia64/include/asm/socket.h b/arch/ia64/include/asm/socket.h index 091cd9d47d0f..0b0d5ff062e5 100644 --- a/arch/ia64/include/asm/socket.h +++ b/arch/ia64/include/asm/socket.h @@ -67,5 +67,6 @@ #define SCM_TIMESTAMPING SO_TIMESTAMPING #define SO_PROTOCOL 38 +#define SO_DOMAIN 39 #endif /* _ASM_IA64_SOCKET_H */ diff --git a/arch/m32r/include/asm/socket.h b/arch/m32r/include/asm/socket.h index d36f5928fb79..3390a864f224 100644 --- a/arch/m32r/include/asm/socket.h +++ b/arch/m32r/include/asm/socket.h @@ -58,5 +58,6 @@ #define SCM_TIMESTAMPING SO_TIMESTAMPING #define SO_PROTOCOL 38 +#define SO_DOMAIN 39 #endif /* _ASM_M32R_SOCKET_H */ diff --git a/arch/m68k/include/asm/socket.h b/arch/m68k/include/asm/socket.h index 060cb7ed024f..eee01cce921b 100644 --- a/arch/m68k/include/asm/socket.h +++ b/arch/m68k/include/asm/socket.h @@ -58,5 +58,6 @@ #define SCM_TIMESTAMPING SO_TIMESTAMPING #define SO_PROTOCOL 38 +#define SO_DOMAIN 39 #endif /* _ASM_SOCKET_H */ diff --git a/arch/microblaze/include/asm/socket.h b/arch/microblaze/include/asm/socket.h index 96bf8bfa935e..7361ae7cfcde 100644 --- a/arch/microblaze/include/asm/socket.h +++ b/arch/microblaze/include/asm/socket.h @@ -67,5 +67,6 @@ #define SCM_TIMESTAMPING SO_TIMESTAMPING #define SO_PROTOCOL 38 +#define SO_DOMAIN 39 #endif /* _ASM_MICROBLAZE_SOCKET_H */ diff --git a/arch/mips/include/asm/socket.h b/arch/mips/include/asm/socket.h index 289ce5f5f2a3..ae05accd9fe4 100644 --- a/arch/mips/include/asm/socket.h +++ b/arch/mips/include/asm/socket.h @@ -43,6 +43,7 @@ To add: #define SO_REUSEPORT 0x0200 /* Allow local address and port reuse. */ #define SO_RCVTIMEO 0x1006 /* receive timeout */ #define SO_ACCEPTCONN 0x1009 #define SO_PROTOCOL 0x1028 /* protocol type */ +#define SO_DOMAIN 0x1029 /* domain/socket family */ /* linux-specific, might as well be the same as on i386 */ #define SO_NO_CHECK 11 diff --git a/arch/mn10300/include/asm/socket.h b/arch/mn10300/include/asm/socket.h index 19d7cf709b77..4df75af29d76 100644 --- a/arch/mn10300/include/asm/socket.h +++ b/arch/mn10300/include/asm/socket.h @@ -58,5 +58,6 @@ #define SCM_TIMESTAMPING SO_TIMESTAMPING #define SO_PROTOCOL 38 +#define SO_DOMAIN 39 #endif /* _ASM_SOCKET_H */ diff --git a/arch/parisc/include/asm/socket.h b/arch/parisc/include/asm/socket.h index a658b09df624..960b1e5d8e16 100644 --- a/arch/parisc/include/asm/socket.h +++ b/arch/parisc/include/asm/socket.h @@ -25,6 +25,7 @@ #define SO_ERROR 0x1007 #define SO_TYPE 0x1008 #define SO_PROTOCOL 0x1028 +#define SO_DOMAIN 0x1029 #define SO_PEERNAME 0x2000 #define SO_NO_CHECK 0x400b diff --git a/arch/powerpc/include/asm/socket.h b/arch/powerpc/include/asm/socket.h index 609049d7117e..3ab8b3e6feb0 100644 --- a/arch/powerpc/include/asm/socket.h +++ b/arch/powerpc/include/asm/socket.h @@ -65,5 +65,6 @@ #define SCM_TIMESTAMPING SO_TIMESTAMPING #define SO_PROTOCOL 38 +#define SO_DOMAIN 39 #endif /* _ASM_POWERPC_SOCKET_H */ diff --git a/arch/s390/include/asm/socket.h b/arch/s390/include/asm/socket.h index 65baa9a83abc..e42df89a0b85 100644 --- a/arch/s390/include/asm/socket.h +++ b/arch/s390/include/asm/socket.h @@ -66,5 +66,6 @@ #define SCM_TIMESTAMPING SO_TIMESTAMPING #define SO_PROTOCOL 38 +#define SO_DOMAIN 39 #endif /* _ASM_SOCKET_H */ diff --git a/arch/sparc/include/asm/socket.h b/arch/sparc/include/asm/socket.h index 9cbbfafd0538..3a5ae3d12088 100644 --- a/arch/sparc/include/asm/socket.h +++ b/arch/sparc/include/asm/socket.h @@ -30,6 +30,7 @@ #define SO_ERROR 0x1007 #define SO_TYPE 0x1008 #define SO_PROTOCOL 0x1028 +#define SO_DOMAIN 0x1029 /* Linux specific, keep the same. */ diff --git a/arch/x86/include/asm/socket.h b/arch/x86/include/asm/socket.h index 1077d2535a32..b2a8c74f2d06 100644 --- a/arch/x86/include/asm/socket.h +++ b/arch/x86/include/asm/socket.h @@ -58,5 +58,6 @@ #define SCM_TIMESTAMPING SO_TIMESTAMPING #define SO_PROTOCOL 38 +#define SO_DOMAIN 39 #endif /* _ASM_X86_SOCKET_H */ diff --git a/arch/xtensa/include/asm/socket.h b/arch/xtensa/include/asm/socket.h index e47f172142f1..beb3a6bdb61d 100644 --- a/arch/xtensa/include/asm/socket.h +++ b/arch/xtensa/include/asm/socket.h @@ -69,5 +69,6 @@ #define SCM_TIMESTAMPING SO_TIMESTAMPING #define SO_PROTOCOL 38 +#define SO_DOMAIN 39 #endif /* _XTENSA_SOCKET_H */ diff --git a/include/asm-generic/socket.h b/include/asm-generic/socket.h index 7e24d186616e..538991cef6f0 100644 --- a/include/asm-generic/socket.h +++ b/include/asm-generic/socket.h @@ -61,5 +61,6 @@ #define SCM_TIMESTAMPING SO_TIMESTAMPING #define SO_PROTOCOL 38 +#define SO_DOMAIN 39 #endif /* __ASM_GENERIC_SOCKET_H */ diff --git a/net/core/sock.c b/net/core/sock.c index ebce661234ac..3ac34ea6ec05 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -483,6 +483,7 @@ int sock_setsockopt(struct socket *sock, int level, int optname, break; case SO_TYPE: case SO_PROTOCOL: + case SO_DOMAIN: case SO_ERROR: ret = -ENOPROTOOPT; break; @@ -769,6 +770,10 @@ int sock_getsockopt(struct socket *sock, int level, int optname, v.val = sk->sk_protocol; break; + case SO_DOMAIN: + v.val = sk->sk_family; + break; + case SO_ERROR: v.val = -sock_error(sk); if (v.val == 0) -- cgit v1.2.3 From 7c73875e7dda627040b12c19b01db634fa7f0fd1 Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Fri, 31 Jul 2009 12:53:58 -0400 Subject: Capabilities: move cap_file_mmap to commoncap.c Currently we duplicate the mmap_min_addr test in cap_file_mmap and in security_file_mmap if !CONFIG_SECURITY. This patch moves cap_file_mmap into commoncap.c and then calls that function directly from security_file_mmap ifndef CONFIG_SECURITY like all of the other capability checks are done. Signed-off-by: Eric Paris Acked-by: Serge Hallyn Signed-off-by: James Morris --- include/linux/security.h | 7 ++++--- security/capability.c | 9 --------- security/commoncap.c | 30 ++++++++++++++++++++++++++++++ 3 files changed, 34 insertions(+), 12 deletions(-) (limited to 'include') diff --git a/include/linux/security.h b/include/linux/security.h index 145909165dbf..963a48fc3005 100644 --- a/include/linux/security.h +++ b/include/linux/security.h @@ -66,6 +66,9 @@ extern int cap_inode_setxattr(struct dentry *dentry, const char *name, extern int cap_inode_removexattr(struct dentry *dentry, const char *name); extern int cap_inode_need_killpriv(struct dentry *dentry); extern int cap_inode_killpriv(struct dentry *dentry); +extern int cap_file_mmap(struct file *file, unsigned long reqprot, + unsigned long prot, unsigned long flags, + unsigned long addr, unsigned long addr_only); extern int cap_task_fix_setuid(struct cred *new, const struct cred *old, int flags); extern int cap_task_prctl(int option, unsigned long arg2, unsigned long arg3, unsigned long arg4, unsigned long arg5); @@ -2197,9 +2200,7 @@ static inline int security_file_mmap(struct file *file, unsigned long reqprot, unsigned long addr, unsigned long addr_only) { - if ((addr < mmap_min_addr) && !capable(CAP_SYS_RAWIO)) - return -EACCES; - return 0; + return cap_file_mmap(file, reqprot, prot, flags, addr, addr_only); } static inline int security_file_mprotect(struct vm_area_struct *vma, diff --git a/security/capability.c b/security/capability.c index f218dd361647..ec0573054024 100644 --- a/security/capability.c +++ b/security/capability.c @@ -330,15 +330,6 @@ static int cap_file_ioctl(struct file *file, unsigned int command, return 0; } -static int cap_file_mmap(struct file *file, unsigned long reqprot, - unsigned long prot, unsigned long flags, - unsigned long addr, unsigned long addr_only) -{ - if ((addr < mmap_min_addr) && !capable(CAP_SYS_RAWIO)) - return -EACCES; - return 0; -} - static int cap_file_mprotect(struct vm_area_struct *vma, unsigned long reqprot, unsigned long prot) { diff --git a/security/commoncap.c b/security/commoncap.c index aa97704564d4..3852e9432801 100644 --- a/security/commoncap.c +++ b/security/commoncap.c @@ -984,3 +984,33 @@ int cap_vm_enough_memory(struct mm_struct *mm, long pages) cap_sys_admin = 1; return __vm_enough_memory(mm, pages, cap_sys_admin); } + +/* + * cap_file_mmap - check if able to map given addr + * @file: unused + * @reqprot: unused + * @prot: unused + * @flags: unused + * @addr: address attempting to be mapped + * @addr_only: unused + * + * If the process is attempting to map memory below mmap_min_addr they need + * CAP_SYS_RAWIO. The other parameters to this function are unused by the + * capability security module. Returns 0 if this mapping should be allowed + * -EPERM if not. + */ +int cap_file_mmap(struct file *file, unsigned long reqprot, + unsigned long prot, unsigned long flags, + unsigned long addr, unsigned long addr_only) +{ + int ret = 0; + + if (addr < mmap_min_addr) { + ret = cap_capable(current, current_cred(), CAP_SYS_RAWIO, + SECURITY_CAP_AUDIT); + /* set PF_SUPERPRIV if it turns out we allow the low mmap */ + if (ret == 0) + current->flags |= PF_SUPERPRIV; + } + return ret; +} -- cgit v1.2.3 From a2551df7ec568d87793d2eea4ca744e86318f205 Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Fri, 31 Jul 2009 12:54:11 -0400 Subject: Security/SELinux: seperate lsm specific mmap_min_addr Currently SELinux enforcement of controls on the ability to map low memory is determined by the mmap_min_addr tunable. This patch causes SELinux to ignore the tunable and instead use a seperate Kconfig option specific to how much space the LSM should protect. The tunable will now only control the need for CAP_SYS_RAWIO and SELinux permissions will always protect the amount of low memory designated by CONFIG_LSM_MMAP_MIN_ADDR. This allows users who need to disable the mmap_min_addr controls (usual reason being they run WINE as a non-root user) to do so and still have SELinux controls preventing confined domains (like a web server) from being able to map some area of low memory. Signed-off-by: Eric Paris Signed-off-by: James Morris --- include/linux/mm.h | 15 --------------- include/linux/security.h | 17 +++++++++++++++++ kernel/sysctl.c | 7 ++++--- mm/Kconfig | 6 +++--- mm/mmap.c | 3 --- mm/nommu.c | 3 --- security/Kconfig | 16 ++++++++++++++++ security/Makefile | 2 +- security/commoncap.c | 2 +- security/min_addr.c | 49 ++++++++++++++++++++++++++++++++++++++++++++++++ security/selinux/hooks.c | 2 +- 11 files changed, 92 insertions(+), 30 deletions(-) create mode 100644 security/min_addr.c (limited to 'include') diff --git a/include/linux/mm.h b/include/linux/mm.h index ba3a7cb1eaa0..9a72cc78e6b8 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -34,8 +34,6 @@ extern int sysctl_legacy_va_layout; #define sysctl_legacy_va_layout 0 #endif -extern unsigned long mmap_min_addr; - #include #include #include @@ -574,19 +572,6 @@ static inline void set_page_links(struct page *page, enum zone_type zone, set_page_section(page, pfn_to_section_nr(pfn)); } -/* - * If a hint addr is less than mmap_min_addr change hint to be as - * low as possible but still greater than mmap_min_addr - */ -static inline unsigned long round_hint_to_min(unsigned long hint) -{ - hint &= PAGE_MASK; - if (((void *)hint != NULL) && - (hint < mmap_min_addr)) - return PAGE_ALIGN(mmap_min_addr); - return hint; -} - /* * Some inline functions in vmstat.h depend on page_zone() */ diff --git a/include/linux/security.h b/include/linux/security.h index 963a48fc3005..7b431155e392 100644 --- a/include/linux/security.h +++ b/include/linux/security.h @@ -28,6 +28,7 @@ #include #include #include +#include /* PAGE_ALIGN */ #include #include #include @@ -95,6 +96,7 @@ extern int cap_netlink_send(struct sock *sk, struct sk_buff *skb); extern int cap_netlink_recv(struct sk_buff *skb, int cap); extern unsigned long mmap_min_addr; +extern unsigned long dac_mmap_min_addr; /* * Values used in the task_security_ops calls */ @@ -147,6 +149,21 @@ static inline void security_free_mnt_opts(struct security_mnt_opts *opts) opts->num_mnt_opts = 0; } +/* + * If a hint addr is less than mmap_min_addr change hint to be as + * low as possible but still greater than mmap_min_addr + */ +static inline unsigned long round_hint_to_min(unsigned long hint) +{ + hint &= PAGE_MASK; + if (((void *)hint != NULL) && + (hint < mmap_min_addr)) + return PAGE_ALIGN(mmap_min_addr); + return hint; +} + +extern int mmap_min_addr_handler(struct ctl_table *table, int write, struct file *filp, + void __user *buffer, size_t *lenp, loff_t *ppos); /** * struct security_operations - main security structure * diff --git a/kernel/sysctl.c b/kernel/sysctl.c index 98e02328c67d..58be76017fd0 100644 --- a/kernel/sysctl.c +++ b/kernel/sysctl.c @@ -49,6 +49,7 @@ #include #include #include +#include #include #include @@ -1306,10 +1307,10 @@ static struct ctl_table vm_table[] = { { .ctl_name = CTL_UNNUMBERED, .procname = "mmap_min_addr", - .data = &mmap_min_addr, - .maxlen = sizeof(unsigned long), + .data = &dac_mmap_min_addr, + .maxlen = sizeof(unsigned long), .mode = 0644, - .proc_handler = &proc_doulongvec_minmax, + .proc_handler = &mmap_min_addr_handler, }, #ifdef CONFIG_NUMA { diff --git a/mm/Kconfig b/mm/Kconfig index c948d4ca8bde..fe5f674d7a7d 100644 --- a/mm/Kconfig +++ b/mm/Kconfig @@ -225,9 +225,9 @@ config DEFAULT_MMAP_MIN_ADDR For most ia64, ppc64 and x86 users with lots of address space a value of 65536 is reasonable and should cause no problems. On arm and other archs it should not be higher than 32768. - Programs which use vm86 functionality would either need additional - permissions from either the LSM or the capabilities module or have - this protection disabled. + Programs which use vm86 functionality or have some need to map + this low address space will need CAP_SYS_RAWIO or disable this + protection by setting the value to 0. This value can be changed after boot using the /proc/sys/vm/mmap_min_addr tunable. diff --git a/mm/mmap.c b/mm/mmap.c index 34579b23ebd5..8101de490c73 100644 --- a/mm/mmap.c +++ b/mm/mmap.c @@ -88,9 +88,6 @@ int sysctl_overcommit_ratio = 50; /* default is 50% */ int sysctl_max_map_count __read_mostly = DEFAULT_MAX_MAP_COUNT; struct percpu_counter vm_committed_as; -/* amount of vm to protect from userspace access */ -unsigned long mmap_min_addr = CONFIG_DEFAULT_MMAP_MIN_ADDR; - /* * Check that a process has enough memory to allocate a new virtual * mapping. 0 means there is enough memory for the allocation to diff --git a/mm/nommu.c b/mm/nommu.c index 53cab10fece4..28754c40be98 100644 --- a/mm/nommu.c +++ b/mm/nommu.c @@ -69,9 +69,6 @@ int sysctl_max_map_count = DEFAULT_MAX_MAP_COUNT; int sysctl_nr_trim_pages = CONFIG_NOMMU_INITIAL_TRIM_EXCESS; int heap_stack_gap = 0; -/* amount of vm to protect from userspace access */ -unsigned long mmap_min_addr = CONFIG_DEFAULT_MMAP_MIN_ADDR; - atomic_long_t mmap_pages_allocated; EXPORT_SYMBOL(mem_map); diff --git a/security/Kconfig b/security/Kconfig index d23c839038f0..9c60c346a91d 100644 --- a/security/Kconfig +++ b/security/Kconfig @@ -113,6 +113,22 @@ config SECURITY_ROOTPLUG If you are unsure how to answer this question, answer N. +config LSM_MMAP_MIN_ADDR + int "Low address space for LSM to from user allocation" + depends on SECURITY && SECURITY_SELINUX + default 65535 + help + This is the portion of low virtual memory which should be protected + from userspace allocation. Keeping a user from writing to low pages + can help reduce the impact of kernel NULL pointer bugs. + + For most ia64, ppc64 and x86 users with lots of address space + a value of 65536 is reasonable and should cause no problems. + On arm and other archs it should not be higher than 32768. + Programs which use vm86 functionality or have some need to map + this low address space will need the permission specific to the + systems running LSM. + source security/selinux/Kconfig source security/smack/Kconfig source security/tomoyo/Kconfig diff --git a/security/Makefile b/security/Makefile index c67557cdaa85..b56e7f9ecbc2 100644 --- a/security/Makefile +++ b/security/Makefile @@ -8,7 +8,7 @@ subdir-$(CONFIG_SECURITY_SMACK) += smack subdir-$(CONFIG_SECURITY_TOMOYO) += tomoyo # always enable default capabilities -obj-y += commoncap.o +obj-y += commoncap.o min_addr.o # Object file lists obj-$(CONFIG_SECURITY) += security.o capability.o diff --git a/security/commoncap.c b/security/commoncap.c index 3852e9432801..fe30751a6cd9 100644 --- a/security/commoncap.c +++ b/security/commoncap.c @@ -1005,7 +1005,7 @@ int cap_file_mmap(struct file *file, unsigned long reqprot, { int ret = 0; - if (addr < mmap_min_addr) { + if (addr < dac_mmap_min_addr) { ret = cap_capable(current, current_cred(), CAP_SYS_RAWIO, SECURITY_CAP_AUDIT); /* set PF_SUPERPRIV if it turns out we allow the low mmap */ diff --git a/security/min_addr.c b/security/min_addr.c new file mode 100644 index 000000000000..14cc7b3b8d03 --- /dev/null +++ b/security/min_addr.c @@ -0,0 +1,49 @@ +#include +#include +#include +#include + +/* amount of vm to protect from userspace access by both DAC and the LSM*/ +unsigned long mmap_min_addr; +/* amount of vm to protect from userspace using CAP_SYS_RAWIO (DAC) */ +unsigned long dac_mmap_min_addr = CONFIG_DEFAULT_MMAP_MIN_ADDR; +/* amount of vm to protect from userspace using the LSM = CONFIG_LSM_MMAP_MIN_ADDR */ + +/* + * Update mmap_min_addr = max(dac_mmap_min_addr, CONFIG_LSM_MMAP_MIN_ADDR) + */ +static void update_mmap_min_addr(void) +{ +#ifdef CONFIG_LSM_MMAP_MIN_ADDR + if (dac_mmap_min_addr > CONFIG_LSM_MMAP_MIN_ADDR) + mmap_min_addr = dac_mmap_min_addr; + else + mmap_min_addr = CONFIG_LSM_MMAP_MIN_ADDR; +#else + mmap_min_addr = dac_mmap_min_addr; +#endif +} + +/* + * sysctl handler which just sets dac_mmap_min_addr = the new value and then + * calls update_mmap_min_addr() so non MAP_FIXED hints get rounded properly + */ +int mmap_min_addr_handler(struct ctl_table *table, int write, struct file *filp, + void __user *buffer, size_t *lenp, loff_t *ppos) +{ + int ret; + + ret = proc_doulongvec_minmax(table, write, filp, buffer, lenp, ppos); + + update_mmap_min_addr(); + + return ret; +} + +int __init init_mmap_min_addr(void) +{ + update_mmap_min_addr(); + + return 0; +} +pure_initcall(init_mmap_min_addr); diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c index 8a78f584f46e..5dee88362e71 100644 --- a/security/selinux/hooks.c +++ b/security/selinux/hooks.c @@ -3040,7 +3040,7 @@ static int selinux_file_mmap(struct file *file, unsigned long reqprot, * at bad behaviour/exploit that we always want to get the AVC, even * if DAC would have also denied the operation. */ - if (addr < mmap_min_addr) { + if (addr < CONFIG_LSM_MMAP_MIN_ADDR) { rc = avc_has_perm(sid, sid, SECCLASS_MEMPROTECT, MEMPROTECT__MMAP_ZERO, NULL); if (rc) -- cgit v1.2.3 From af6af30c0fcd77e621638e53ef8b176bca8bd3b4 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Wed, 5 Aug 2009 20:41:04 +0200 Subject: ftrace: Fix perf-tracepoint OOPS Not all tracepoints are created equal, in specific the ftrace tracepoints are created with TRACE_EVENT_FORMAT() which does not generate the needed bits to tie them into perf counters. For those events, don't create the 'id' file and fail ->profile_enable when their ID is specified through other means. Reported-by: Chris Mason Signed-off-by: Peter Zijlstra Cc: Steven Rostedt LKML-Reference: <1249497664.5890.4.camel@laptop> [ v2: fix build error in the !CONFIG_EVENT_PROFILE case ] Signed-off-by: Ingo Molnar --- include/linux/ftrace_event.h | 8 +++----- kernel/trace/trace_event_profile.c | 2 +- kernel/trace/trace_events.c | 2 +- 3 files changed, 5 insertions(+), 7 deletions(-) (limited to 'include') diff --git a/include/linux/ftrace_event.h b/include/linux/ftrace_event.h index 5c093ffc655b..d7cd193c2277 100644 --- a/include/linux/ftrace_event.h +++ b/include/linux/ftrace_event.h @@ -119,11 +119,9 @@ struct ftrace_event_call { void *filter; void *mod; -#ifdef CONFIG_EVENT_PROFILE - atomic_t profile_count; - int (*profile_enable)(struct ftrace_event_call *); - void (*profile_disable)(struct ftrace_event_call *); -#endif + atomic_t profile_count; + int (*profile_enable)(struct ftrace_event_call *); + void (*profile_disable)(struct ftrace_event_call *); }; #define MAX_FILTER_PRED 32 diff --git a/kernel/trace/trace_event_profile.c b/kernel/trace/trace_event_profile.c index 5b5895afecfe..11ba5bb4ed0a 100644 --- a/kernel/trace/trace_event_profile.c +++ b/kernel/trace/trace_event_profile.c @@ -14,7 +14,7 @@ int ftrace_profile_enable(int event_id) mutex_lock(&event_mutex); list_for_each_entry(event, &ftrace_events, list) { - if (event->id == event_id) { + if (event->id == event_id && event->profile_enable) { ret = event->profile_enable(event); break; } diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c index 23d2972b22d6..e75276a49cf5 100644 --- a/kernel/trace/trace_events.c +++ b/kernel/trace/trace_events.c @@ -940,7 +940,7 @@ event_create_dir(struct ftrace_event_call *call, struct dentry *d_events, entry = trace_create_file("enable", 0644, call->dir, call, enable); - if (call->id) + if (call->id && call->profile_enable) entry = trace_create_file("id", 0444, call->dir, call, id); -- cgit v1.2.3 From d82f1c35348cebe2fb2d4a4d31ce0ab0769e3d93 Mon Sep 17 00:00:00 2001 From: Eric Miao Date: Wed, 5 Aug 2009 01:24:41 -0700 Subject: Input: matrix_keypad - make matrix keymap size dynamic Remove assumption on the shift and size of rows/columns form matrix_keypad driver. Signed-off-by: Eric Miao Signed-off-by: Dmitry Torokhov --- drivers/input/keyboard/matrix_keypad.c | 18 +++++++++--------- include/linux/input/matrix_keypad.h | 13 +++++++------ 2 files changed, 16 insertions(+), 15 deletions(-) (limited to 'include') diff --git a/drivers/input/keyboard/matrix_keypad.c b/drivers/input/keyboard/matrix_keypad.c index e9b2e7cb05be..541b981ff075 100644 --- a/drivers/input/keyboard/matrix_keypad.c +++ b/drivers/input/keyboard/matrix_keypad.c @@ -27,6 +27,7 @@ struct matrix_keypad { const struct matrix_keypad_platform_data *pdata; struct input_dev *input_dev; unsigned short *keycodes; + unsigned int row_shift; uint32_t last_key_state[MATRIX_MAX_COLS]; struct delayed_work work; @@ -136,7 +137,7 @@ static void matrix_keypad_scan(struct work_struct *work) if ((bits_changed & (1 << row)) == 0) continue; - code = (row << 4) + col; + code = MATRIX_SCAN_CODE(row, col, keypad->row_shift); input_event(input_dev, EV_MSC, MSC_SCAN, code); input_report_key(input_dev, keypad->keycodes[code], @@ -317,6 +318,7 @@ static int __devinit matrix_keypad_probe(struct platform_device *pdev) struct matrix_keypad *keypad; struct input_dev *input_dev; unsigned short *keycodes; + unsigned int row_shift; int i; int err; @@ -332,14 +334,11 @@ static int __devinit matrix_keypad_probe(struct platform_device *pdev) return -EINVAL; } - if (!keymap_data->max_keymap_size) { - dev_err(&pdev->dev, "invalid keymap data supplied\n"); - return -EINVAL; - } + row_shift = get_count_order(pdata->num_col_gpios); keypad = kzalloc(sizeof(struct matrix_keypad), GFP_KERNEL); - keycodes = kzalloc(keymap_data->max_keymap_size * - sizeof(keypad->keycodes), + keycodes = kzalloc((pdata->num_row_gpios << row_shift) * + sizeof(*keycodes), GFP_KERNEL); input_dev = input_allocate_device(); if (!keypad || !keycodes || !input_dev) { @@ -350,6 +349,7 @@ static int __devinit matrix_keypad_probe(struct platform_device *pdev) keypad->input_dev = input_dev; keypad->pdata = pdata; keypad->keycodes = keycodes; + keypad->row_shift = row_shift; keypad->stopped = true; INIT_DELAYED_WORK(&keypad->work, matrix_keypad_scan); spin_lock_init(&keypad->lock); @@ -363,7 +363,7 @@ static int __devinit matrix_keypad_probe(struct platform_device *pdev) input_dev->keycode = keycodes; input_dev->keycodesize = sizeof(*keycodes); - input_dev->keycodemax = keymap_data->max_keymap_size; + input_dev->keycodemax = pdata->num_row_gpios << keypad->row_shift; for (i = 0; i < keymap_data->keymap_size; i++) { unsigned int key = keymap_data->keymap[i]; @@ -371,7 +371,7 @@ static int __devinit matrix_keypad_probe(struct platform_device *pdev) unsigned int col = KEY_COL(key); unsigned short code = KEY_VAL(key); - keycodes[(row << 4) + col] = code; + keycodes[MATRIX_SCAN_CODE(row, col, row_shift)] = code; __set_bit(code, input_dev->keybit); } __clear_bit(KEY_RESERVED, input_dev->keybit); diff --git a/include/linux/input/matrix_keypad.h b/include/linux/input/matrix_keypad.h index 7964516c6954..15d5903af2dd 100644 --- a/include/linux/input/matrix_keypad.h +++ b/include/linux/input/matrix_keypad.h @@ -15,12 +15,13 @@ #define KEY_COL(k) (((k) >> 16) & 0xff) #define KEY_VAL(k) ((k) & 0xffff) +#define MATRIX_SCAN_CODE(row, col, row_shift) (((row) << (row_shift)) + (col)) + /** * struct matrix_keymap_data - keymap for matrix keyboards * @keymap: pointer to array of uint32 values encoded with KEY() macro * representing keymap * @keymap_size: number of entries (initialized) in this keymap - * @max_keymap_size: maximum size of keymap supported by the device * * This structure is supposed to be used by platform code to supply * keymaps to drivers that implement matrix-like keypads/keyboards. @@ -28,14 +29,13 @@ struct matrix_keymap_data { const uint32_t *keymap; unsigned int keymap_size; - unsigned int max_keymap_size; }; /** * struct matrix_keypad_platform_data - platform-dependent keypad data * @keymap_data: pointer to &matrix_keymap_data - * @row_gpios: array of gpio numbers reporesenting rows - * @col_gpios: array of gpio numbers reporesenting colums + * @row_gpios: pointer to array of gpio numbers representing rows + * @col_gpios: pointer to array of gpio numbers reporesenting colums * @num_row_gpios: actual number of row gpios used by device * @num_col_gpios: actual number of col gpios used by device * @col_scan_delay_us: delay, measured in microseconds, that is @@ -48,8 +48,9 @@ struct matrix_keymap_data { struct matrix_keypad_platform_data { const struct matrix_keymap_data *keymap_data; - unsigned int row_gpios[MATRIX_MAX_ROWS]; - unsigned int col_gpios[MATRIX_MAX_COLS]; + const unsigned int *row_gpios; + const unsigned int *col_gpios; + unsigned int num_row_gpios; unsigned int num_col_gpios; -- cgit v1.2.3 From ace1366369841c9c3a9788f79baa4d73f1c53107 Mon Sep 17 00:00:00 2001 From: Huang Ying Date: Thu, 6 Aug 2009 15:35:20 +1000 Subject: crypto: cryptd - Add support to access underlaying shash cryptd_alloc_ahash() will allocate a cryptd-ed ahash for specified algorithm name. The new allocated one is guaranteed to be cryptd-ed ahash, so the shash underlying can be gotten via cryptd_ahash_child(). Signed-off-by: Huang Ying Signed-off-by: Herbert Xu --- crypto/cryptd.c | 35 +++++++++++++++++++++++++++++++++++ include/crypto/cryptd.h | 17 +++++++++++++++++ 2 files changed, 52 insertions(+) (limited to 'include') diff --git a/crypto/cryptd.c b/crypto/cryptd.c index 2eb705822f2b..35335825a4ef 100644 --- a/crypto/cryptd.c +++ b/crypto/cryptd.c @@ -682,6 +682,41 @@ void cryptd_free_ablkcipher(struct cryptd_ablkcipher *tfm) } EXPORT_SYMBOL_GPL(cryptd_free_ablkcipher); +struct cryptd_ahash *cryptd_alloc_ahash(const char *alg_name, + u32 type, u32 mask) +{ + char cryptd_alg_name[CRYPTO_MAX_ALG_NAME]; + struct crypto_ahash *tfm; + + if (snprintf(cryptd_alg_name, CRYPTO_MAX_ALG_NAME, + "cryptd(%s)", alg_name) >= CRYPTO_MAX_ALG_NAME) + return ERR_PTR(-EINVAL); + tfm = crypto_alloc_ahash(cryptd_alg_name, type, mask); + if (IS_ERR(tfm)) + return ERR_CAST(tfm); + if (tfm->base.__crt_alg->cra_module != THIS_MODULE) { + crypto_free_ahash(tfm); + return ERR_PTR(-EINVAL); + } + + return __cryptd_ahash_cast(tfm); +} +EXPORT_SYMBOL_GPL(cryptd_alloc_ahash); + +struct crypto_shash *cryptd_ahash_child(struct cryptd_ahash *tfm) +{ + struct cryptd_hash_ctx *ctx = crypto_ahash_ctx(&tfm->base); + + return ctx->child; +} +EXPORT_SYMBOL_GPL(cryptd_ahash_child); + +void cryptd_free_ahash(struct cryptd_ahash *tfm) +{ + crypto_free_ahash(&tfm->base); +} +EXPORT_SYMBOL_GPL(cryptd_free_ahash); + static int __init cryptd_init(void) { int err; diff --git a/include/crypto/cryptd.h b/include/crypto/cryptd.h index 55fa7bbdbc71..2f65a6e8ea4d 100644 --- a/include/crypto/cryptd.h +++ b/include/crypto/cryptd.h @@ -7,6 +7,7 @@ #include #include +#include struct cryptd_ablkcipher { struct crypto_ablkcipher base; @@ -24,4 +25,20 @@ struct cryptd_ablkcipher *cryptd_alloc_ablkcipher(const char *alg_name, struct crypto_blkcipher *cryptd_ablkcipher_child(struct cryptd_ablkcipher *tfm); void cryptd_free_ablkcipher(struct cryptd_ablkcipher *tfm); +struct cryptd_ahash { + struct crypto_ahash base; +}; + +static inline struct cryptd_ahash *__cryptd_ahash_cast( + struct crypto_ahash *tfm) +{ + return (struct cryptd_ahash *)tfm; +} + +/* alg_name should be algorithm to be cryptd-ed */ +struct cryptd_ahash *cryptd_alloc_ahash(const char *alg_name, + u32 type, u32 mask); +struct crypto_shash *cryptd_ahash_child(struct cryptd_ahash *tfm); +void cryptd_free_ahash(struct cryptd_ahash *tfm); + #endif -- cgit v1.2.3 From bbff2e433e80fae72c8d00d482927d52ec19ba33 Mon Sep 17 00:00:00 2001 From: Wu Fengguang Date: Thu, 6 Aug 2009 11:36:25 +0300 Subject: slab: remove duplicate kmem_cache_init_late() declarations kmem_cache_init_late() has been declared in slab.h CC: Nick Piggin CC: Matt Mackall CC: Christoph Lameter Signed-off-by: Wu Fengguang Signed-off-by: Pekka Enberg --- include/linux/slob_def.h | 5 ----- include/linux/slub_def.h | 2 -- mm/slob.c | 5 +++++ 3 files changed, 5 insertions(+), 7 deletions(-) (limited to 'include') diff --git a/include/linux/slob_def.h b/include/linux/slob_def.h index bb5368df4be8..0ec00b39d006 100644 --- a/include/linux/slob_def.h +++ b/include/linux/slob_def.h @@ -34,9 +34,4 @@ static __always_inline void *__kmalloc(size_t size, gfp_t flags) return kmalloc(size, flags); } -static inline void kmem_cache_init_late(void) -{ - /* Nothing to do */ -} - #endif /* __LINUX_SLOB_DEF_H */ diff --git a/include/linux/slub_def.h b/include/linux/slub_def.h index c1c862b1d01a..76c831693616 100644 --- a/include/linux/slub_def.h +++ b/include/linux/slub_def.h @@ -304,6 +304,4 @@ static __always_inline void *kmalloc_node(size_t size, gfp_t flags, int node) } #endif -void __init kmem_cache_init_late(void); - #endif /* _LINUX_SLUB_DEF_H */ diff --git a/mm/slob.c b/mm/slob.c index 9641da3d5e58..837ebd64cc34 100644 --- a/mm/slob.c +++ b/mm/slob.c @@ -692,3 +692,8 @@ void __init kmem_cache_init(void) { slob_ready = 1; } + +void __init kmem_cache_init_late(void) +{ + /* Nothing to do */ +} -- cgit v1.2.3 From bbd8a0d3a3b65d341437f8b99c828fa5cc29c739 Mon Sep 17 00:00:00 2001 From: Krishna Kumar Date: Thu, 6 Aug 2009 01:44:21 +0000 Subject: net: Avoid enqueuing skb for default qdiscs dev_queue_xmit enqueue's a skb and calls qdisc_run which dequeue's the skb and xmits it. In most cases, the skb that is enqueue'd is the same one that is dequeue'd (unless the queue gets stopped or multiple cpu's write to the same queue and ends in a race with qdisc_run). For default qdiscs, we can remove the redundant enqueue/dequeue and simply xmit the skb since the default qdisc is work-conserving. The patch uses a new flag - TCQ_F_CAN_BYPASS to identify the default fast queue. The controversial part of the patch is incrementing qlen when a skb is requeued - this is to avoid checks like the second line below: + } else if ((q->flags & TCQ_F_CAN_BYPASS) && !qdisc_qlen(q) && >> !q->gso_skb && + !test_and_set_bit(__QDISC_STATE_RUNNING, &q->state)) { Results of a 2 hour testing for multiple netperf sessions (1, 2, 4, 8, 12 sessions on a 4 cpu system-X). The BW numbers are aggregate Mb/s across iterations tested with this version on System-X boxes with Chelsio 10gbps cards: ---------------------------------- Size | ORG BW NEW BW | ---------------------------------- 128K | 156964 159381 | 256K | 158650 162042 | ---------------------------------- Changes from ver1: 1. Move sch_direct_xmit declaration from sch_generic.h to pkt_sched.h 2. Update qdisc basic statistics for direct xmit path. 3. Set qlen to zero in qdisc_reset. 4. Changed some function names to more meaningful ones. Signed-off-by: Krishna Kumar Signed-off-by: David S. Miller --- include/net/pkt_sched.h | 3 ++ include/net/sch_generic.h | 15 +++++++- net/core/dev.c | 48 +++++++++++++++++------- net/sched/sch_generic.c | 93 +++++++++++++++++++++++++++++------------------ 4 files changed, 108 insertions(+), 51 deletions(-) (limited to 'include') diff --git a/include/net/pkt_sched.h b/include/net/pkt_sched.h index 82a3191375f5..f911ec7598ef 100644 --- a/include/net/pkt_sched.h +++ b/include/net/pkt_sched.h @@ -87,6 +87,9 @@ extern struct qdisc_rate_table *qdisc_get_rtab(struct tc_ratespec *r, extern void qdisc_put_rtab(struct qdisc_rate_table *tab); extern void qdisc_put_stab(struct qdisc_size_table *tab); extern void qdisc_warn_nonwc(char *txt, struct Qdisc *qdisc); +extern int sch_direct_xmit(struct sk_buff *skb, struct Qdisc *q, + struct net_device *dev, struct netdev_queue *txq, + spinlock_t *root_lock); extern void __qdisc_run(struct Qdisc *q); diff --git a/include/net/sch_generic.h b/include/net/sch_generic.h index 964ffa0d8815..84b3fc2aef0f 100644 --- a/include/net/sch_generic.h +++ b/include/net/sch_generic.h @@ -45,6 +45,7 @@ struct Qdisc #define TCQ_F_BUILTIN 1 #define TCQ_F_THROTTLED 2 #define TCQ_F_INGRESS 4 +#define TCQ_F_CAN_BYPASS 8 #define TCQ_F_WARN_NONWC (1 << 16) int padded; struct Qdisc_ops *ops; @@ -182,6 +183,11 @@ struct qdisc_skb_cb { char data[]; }; +static inline int qdisc_qlen(struct Qdisc *q) +{ + return q->q.qlen; +} + static inline struct qdisc_skb_cb *qdisc_skb_cb(struct sk_buff *skb) { return (struct qdisc_skb_cb *)skb->cb; @@ -387,13 +393,18 @@ static inline int qdisc_enqueue_root(struct sk_buff *skb, struct Qdisc *sch) return qdisc_enqueue(skb, sch) & NET_XMIT_MASK; } +static inline void __qdisc_update_bstats(struct Qdisc *sch, unsigned int len) +{ + sch->bstats.bytes += len; + sch->bstats.packets++; +} + static inline int __qdisc_enqueue_tail(struct sk_buff *skb, struct Qdisc *sch, struct sk_buff_head *list) { __skb_queue_tail(list, skb); sch->qstats.backlog += qdisc_pkt_len(skb); - sch->bstats.bytes += qdisc_pkt_len(skb); - sch->bstats.packets++; + __qdisc_update_bstats(sch, qdisc_pkt_len(skb)); return NET_XMIT_SUCCESS; } diff --git a/net/core/dev.c b/net/core/dev.c index f01a9c41f112..a0bc087616a4 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -1786,6 +1786,40 @@ static struct netdev_queue *dev_pick_tx(struct net_device *dev, return netdev_get_tx_queue(dev, queue_index); } +static inline int __dev_xmit_skb(struct sk_buff *skb, struct Qdisc *q, + struct net_device *dev, + struct netdev_queue *txq) +{ + spinlock_t *root_lock = qdisc_lock(q); + int rc; + + spin_lock(root_lock); + if (unlikely(test_bit(__QDISC_STATE_DEACTIVATED, &q->state))) { + kfree_skb(skb); + rc = NET_XMIT_DROP; + } else if ((q->flags & TCQ_F_CAN_BYPASS) && !qdisc_qlen(q) && + !test_and_set_bit(__QDISC_STATE_RUNNING, &q->state)) { + /* + * This is a work-conserving queue; there are no old skbs + * waiting to be sent out; and the qdisc is not running - + * xmit the skb directly. + */ + __qdisc_update_bstats(q, skb->len); + if (sch_direct_xmit(skb, q, dev, txq, root_lock)) + __qdisc_run(q); + else + clear_bit(__QDISC_STATE_RUNNING, &q->state); + + rc = NET_XMIT_SUCCESS; + } else { + rc = qdisc_enqueue_root(skb, q); + qdisc_run(q); + } + spin_unlock(root_lock); + + return rc; +} + /** * dev_queue_xmit - transmit a buffer * @skb: buffer to transmit @@ -1859,19 +1893,7 @@ gso: skb->tc_verd = SET_TC_AT(skb->tc_verd,AT_EGRESS); #endif if (q->enqueue) { - spinlock_t *root_lock = qdisc_lock(q); - - spin_lock(root_lock); - - if (unlikely(test_bit(__QDISC_STATE_DEACTIVATED, &q->state))) { - kfree_skb(skb); - rc = NET_XMIT_DROP; - } else { - rc = qdisc_enqueue_root(skb, q); - qdisc_run(q); - } - spin_unlock(root_lock); - + rc = __dev_xmit_skb(skb, q, dev, txq); goto out; } diff --git a/net/sched/sch_generic.c b/net/sched/sch_generic.c index 27d03816ec3e..693df7ae33d8 100644 --- a/net/sched/sch_generic.c +++ b/net/sched/sch_generic.c @@ -37,15 +37,11 @@ * - updates to tree and tree walking are only done under the rtnl mutex. */ -static inline int qdisc_qlen(struct Qdisc *q) -{ - return q->q.qlen; -} - static inline int dev_requeue_skb(struct sk_buff *skb, struct Qdisc *q) { q->gso_skb = skb; q->qstats.requeues++; + q->q.qlen++; /* it's still part of the queue */ __netif_schedule(q); return 0; @@ -61,9 +57,11 @@ static inline struct sk_buff *dequeue_skb(struct Qdisc *q) /* check the reason of requeuing without tx lock first */ txq = netdev_get_tx_queue(dev, skb_get_queue_mapping(skb)); - if (!netif_tx_queue_stopped(txq) && !netif_tx_queue_frozen(txq)) + if (!netif_tx_queue_stopped(txq) && + !netif_tx_queue_frozen(txq)) { q->gso_skb = NULL; - else + q->q.qlen--; + } else skb = NULL; } else { skb = q->dequeue(q); @@ -103,44 +101,23 @@ static inline int handle_dev_cpu_collision(struct sk_buff *skb, } /* - * NOTE: Called under qdisc_lock(q) with locally disabled BH. - * - * __QDISC_STATE_RUNNING guarantees only one CPU can process - * this qdisc at a time. qdisc_lock(q) serializes queue accesses for - * this queue. - * - * netif_tx_lock serializes accesses to device driver. - * - * qdisc_lock(q) and netif_tx_lock are mutually exclusive, - * if one is grabbed, another must be free. - * - * Note, that this procedure can be called by a watchdog timer + * Transmit one skb, and handle the return status as required. Holding the + * __QDISC_STATE_RUNNING bit guarantees that only one CPU can execute this + * function. * * Returns to the caller: * 0 - queue is empty or throttled. * >0 - queue is not empty. - * */ -static inline int qdisc_restart(struct Qdisc *q) +int sch_direct_xmit(struct sk_buff *skb, struct Qdisc *q, + struct net_device *dev, struct netdev_queue *txq, + spinlock_t *root_lock) { - struct netdev_queue *txq; int ret = NETDEV_TX_BUSY; - struct net_device *dev; - spinlock_t *root_lock; - struct sk_buff *skb; - - /* Dequeue packet */ - if (unlikely((skb = dequeue_skb(q)) == NULL)) - return 0; - - root_lock = qdisc_lock(q); /* And release qdisc */ spin_unlock(root_lock); - dev = qdisc_dev(q); - txq = netdev_get_tx_queue(dev, skb_get_queue_mapping(skb)); - HARD_TX_LOCK(dev, txq, smp_processor_id()); if (!netif_tx_queue_stopped(txq) && !netif_tx_queue_frozen(txq)) @@ -177,6 +154,44 @@ static inline int qdisc_restart(struct Qdisc *q) return ret; } +/* + * NOTE: Called under qdisc_lock(q) with locally disabled BH. + * + * __QDISC_STATE_RUNNING guarantees only one CPU can process + * this qdisc at a time. qdisc_lock(q) serializes queue accesses for + * this queue. + * + * netif_tx_lock serializes accesses to device driver. + * + * qdisc_lock(q) and netif_tx_lock are mutually exclusive, + * if one is grabbed, another must be free. + * + * Note, that this procedure can be called by a watchdog timer + * + * Returns to the caller: + * 0 - queue is empty or throttled. + * >0 - queue is not empty. + * + */ +static inline int qdisc_restart(struct Qdisc *q) +{ + struct netdev_queue *txq; + struct net_device *dev; + spinlock_t *root_lock; + struct sk_buff *skb; + + /* Dequeue packet */ + skb = dequeue_skb(q); + if (unlikely(!skb)) + return 0; + + root_lock = qdisc_lock(q); + dev = qdisc_dev(q); + txq = netdev_get_tx_queue(dev, skb_get_queue_mapping(skb)); + + return sch_direct_xmit(skb, q, dev, txq, root_lock); +} + void __qdisc_run(struct Qdisc *q) { unsigned long start_time = jiffies; @@ -547,8 +562,11 @@ void qdisc_reset(struct Qdisc *qdisc) if (ops->reset) ops->reset(qdisc); - kfree_skb(qdisc->gso_skb); - qdisc->gso_skb = NULL; + if (qdisc->gso_skb) { + kfree_skb(qdisc->gso_skb); + qdisc->gso_skb = NULL; + qdisc->q.qlen = 0; + } } EXPORT_SYMBOL(qdisc_reset); @@ -605,6 +623,9 @@ static void attach_one_default_qdisc(struct net_device *dev, printk(KERN_INFO "%s: activation failed\n", dev->name); return; } + + /* Can by-pass the queue discipline for default qdisc */ + qdisc->flags |= TCQ_F_CAN_BYPASS; } else { qdisc = &noqueue_qdisc; } -- cgit v1.2.3 From 54e346215e4fe2ca8c94c54e546cc61902060510 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Fri, 7 Aug 2009 14:38:25 -0300 Subject: vfs: fix inode_init_always calling convention Currently inode_init_always calls into ->destroy_inode if the additional initialization fails. That's not only counter-intuitive because inode_init_always did not allocate the inode structure, but in case of XFS it's actively harmful as ->destroy_inode might delete the inode from a radix-tree that has never been added. This in turn might end up deleting the inode for the same inum that has been instanciated by another process and cause lots of cause subtile problems. Also in the case of re-initializing a reclaimable inode in XFS it would free an inode we still want to keep alive. Signed-off-by: Christoph Hellwig Reviewed-by: Eric Sandeen --- fs/inode.c | 30 +++++++++++++++++------------- fs/xfs/xfs_iget.c | 17 +++++------------ include/linux/fs.h | 2 +- 3 files changed, 23 insertions(+), 26 deletions(-) (limited to 'include') diff --git a/fs/inode.c b/fs/inode.c index 901bad1e5f12..af2c05235cc8 100644 --- a/fs/inode.c +++ b/fs/inode.c @@ -120,12 +120,11 @@ static void wake_up_inode(struct inode *inode) * These are initializations that need to be done on every inode * allocation as the fields are not initialised by slab allocation. */ -struct inode *inode_init_always(struct super_block *sb, struct inode *inode) +int inode_init_always(struct super_block *sb, struct inode *inode) { static const struct address_space_operations empty_aops; static struct inode_operations empty_iops; static const struct file_operations empty_fops; - struct address_space *const mapping = &inode->i_data; inode->i_sb = sb; @@ -152,7 +151,7 @@ struct inode *inode_init_always(struct super_block *sb, struct inode *inode) inode->dirtied_when = 0; if (security_inode_alloc(inode)) - goto out_free_inode; + goto out; /* allocate and initialize an i_integrity */ if (ima_inode_alloc(inode)) @@ -198,16 +197,12 @@ struct inode *inode_init_always(struct super_block *sb, struct inode *inode) inode->i_fsnotify_mask = 0; #endif - return inode; + return 0; out_free_security: security_inode_free(inode); -out_free_inode: - if (inode->i_sb->s_op->destroy_inode) - inode->i_sb->s_op->destroy_inode(inode); - else - kmem_cache_free(inode_cachep, (inode)); - return NULL; +out: + return -ENOMEM; } EXPORT_SYMBOL(inode_init_always); @@ -220,9 +215,18 @@ static struct inode *alloc_inode(struct super_block *sb) else inode = kmem_cache_alloc(inode_cachep, GFP_KERNEL); - if (inode) - return inode_init_always(sb, inode); - return NULL; + if (!inode) + return NULL; + + if (unlikely(inode_init_always(sb, inode))) { + if (inode->i_sb->s_op->destroy_inode) + inode->i_sb->s_op->destroy_inode(inode); + else + kmem_cache_free(inode_cachep, inode); + return NULL; + } + + return inode; } void destroy_inode(struct inode *inode) diff --git a/fs/xfs/xfs_iget.c b/fs/xfs/xfs_iget.c index 5fcec6f020a7..719c85b155f4 100644 --- a/fs/xfs/xfs_iget.c +++ b/fs/xfs/xfs_iget.c @@ -64,6 +64,10 @@ xfs_inode_alloc( ip = kmem_zone_alloc(xfs_inode_zone, KM_SLEEP); if (!ip) return NULL; + if (inode_init_always(mp->m_super, VFS_I(ip))) { + kmem_zone_free(xfs_inode_zone, ip); + return NULL; + } ASSERT(atomic_read(&ip->i_iocount) == 0); ASSERT(atomic_read(&ip->i_pincount) == 0); @@ -105,17 +109,6 @@ xfs_inode_alloc( #ifdef XFS_DIR2_TRACE ip->i_dir_trace = ktrace_alloc(XFS_DIR2_KTRACE_SIZE, KM_NOFS); #endif - /* - * Now initialise the VFS inode. We do this after the xfs_inode - * initialisation as internal failures will result in ->destroy_inode - * being called and that will pass down through the reclaim path and - * free the XFS inode. This path requires the XFS inode to already be - * initialised. Hence if this call fails, the xfs_inode has already - * been freed and we should not reference it at all in the error - * handling. - */ - if (!inode_init_always(mp->m_super, VFS_I(ip))) - return NULL; /* prevent anyone from using this yet */ VFS_I(ip)->i_state = I_NEW|I_LOCK; @@ -167,7 +160,7 @@ xfs_iget_cache_hit( * errors cleanly, then tag it so it can be set up correctly * later. */ - if (!inode_init_always(mp->m_super, VFS_I(ip))) { + if (inode_init_always(mp->m_super, VFS_I(ip))) { error = ENOMEM; goto out_error; } diff --git a/include/linux/fs.h b/include/linux/fs.h index a36ffa5a77a4..0c3b5e58a986 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -2137,7 +2137,7 @@ extern loff_t default_llseek(struct file *file, loff_t offset, int origin); extern loff_t vfs_llseek(struct file *file, loff_t offset, int origin); -extern struct inode * inode_init_always(struct super_block *, struct inode *); +extern int inode_init_always(struct super_block *, struct inode *); extern void inode_init_once(struct inode *); extern void inode_add_to_lists(struct super_block *, struct inode *); extern void iput(struct inode *); -- cgit v1.2.3 From 2e00c97e2c1d2ffc9e26252ca26b237678b0b772 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Fri, 7 Aug 2009 14:38:29 -0300 Subject: vfs: add __destroy_inode When we want to tear down an inode that lost the add to the cache race in XFS we must not call into ->destroy_inode because that would delete the inode that won the race from the inode cache radix tree. This patch provides the __destroy_inode helper needed to fix this, the actual fix will be in th next patch. As XFS was the only reason destroy_inode was exported we shift the export to the new __destroy_inode. Signed-off-by: Christoph Hellwig Reviewed-by: Eric Sandeen --- fs/inode.c | 10 +++++++--- include/linux/fs.h | 1 + 2 files changed, 8 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/fs/inode.c b/fs/inode.c index af2c05235cc8..ae7b67e48661 100644 --- a/fs/inode.c +++ b/fs/inode.c @@ -229,7 +229,7 @@ static struct inode *alloc_inode(struct super_block *sb) return inode; } -void destroy_inode(struct inode *inode) +void __destroy_inode(struct inode *inode) { BUG_ON(inode_has_buffers(inode)); ima_inode_free(inode); @@ -241,13 +241,17 @@ void destroy_inode(struct inode *inode) if (inode->i_default_acl && inode->i_default_acl != ACL_NOT_CACHED) posix_acl_release(inode->i_default_acl); #endif +} +EXPORT_SYMBOL(__destroy_inode); + +void destroy_inode(struct inode *inode) +{ + __destroy_inode(inode); if (inode->i_sb->s_op->destroy_inode) inode->i_sb->s_op->destroy_inode(inode); else kmem_cache_free(inode_cachep, (inode)); } -EXPORT_SYMBOL(destroy_inode); - /* * These are initializations that only need to be done diff --git a/include/linux/fs.h b/include/linux/fs.h index 0c3b5e58a986..67888a9e0655 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -2164,6 +2164,7 @@ extern void __iget(struct inode * inode); extern void iget_failed(struct inode *); extern void clear_inode(struct inode *); extern void destroy_inode(struct inode *); +extern void __destroy_inode(struct inode *); extern struct inode *new_inode(struct super_block *); extern int should_remove_suid(struct dentry *); extern int file_remove_suid(struct file *); -- cgit v1.2.3 From 4bfc44958e499af9a73f62201543b3a1f617cfeb Mon Sep 17 00:00:00 2001 From: KAMEZAWA Hiroyuki Date: Thu, 6 Aug 2009 15:07:33 -0700 Subject: mm: make set_mempolicy(MPOL_INTERLEAV) N_HIGH_MEMORY aware At first, init_task's mems_allowed is initialized as this. init_task->mems_allowed == node_state[N_POSSIBLE] And cpuset's top_cpuset mask is initialized as this top_cpuset->mems_allowed = node_state[N_HIGH_MEMORY] Before 2.6.29: policy's mems_allowed is initialized as this. 1. update tasks->mems_allowed by its cpuset->mems_allowed. 2. policy->mems_allowed = nodes_and(tasks->mems_allowed, user's mask) Updating task's mems_allowed in reference to top_cpuset's one. cpuset's mems_allowed is aware of N_HIGH_MEMORY, always. In 2.6.30: After commit 58568d2a8215cb6f55caf2332017d7bdff954e1c ("cpuset,mm: update tasks' mems_allowed in time"), policy's mems_allowed is initialized as this. 1. policy->mems_allowd = nodes_and(task->mems_allowed, user's mask) Here, if task is in top_cpuset, task->mems_allowed is not updated from init's one. Assume user excutes command as #numactrl --interleave=all ,.... policy->mems_allowd = nodes_and(N_POSSIBLE, ALL_SET_MASK) Then, policy's mems_allowd can includes a possible node, which has no pgdat. MPOL's INTERLEAVE just scans nodemask of task->mems_allowd and access this directly. NODE_DATA(nid)->zonelist even if NODE_DATA(nid)==NULL Then, what's we need is making policy->mems_allowed be aware of N_HIGH_MEMORY. This patch does that. But to do so, extra nodemask will be on statck. Because I know cpumask has a new interface of CPUMASK_ALLOC(), I added it to node. This patch stands on old behavior. But I feel this fix itself is just a Band-Aid. But to do fundametal fix, we have to take care of memory hotplug and it takes time. (task->mems_allowd should be N_HIGH_MEMORY, I think.) mpol_set_nodemask() should be aware of N_HIGH_MEMORY and policy's nodemask should be includes only online nodes. In old behavior, this is guaranteed by frequent reference to cpuset's code. Now, most of them are removed and mempolicy has to check it by itself. To do check, a few nodemask_t will be used for calculating nodemask. But, size of nodemask_t can be big and it's not good to allocate them on stack. Now, cpumask_t has CPUMASK_ALLOC/FREE an easy code for get scratch area. NODEMASK_ALLOC/FREE shoudl be there. [akpm@linux-foundation.org: cleanups & tweaks] Tested-by: KOSAKI Motohiro Signed-off-by: KAMEZAWA Hiroyuki Cc: Miao Xie Cc: Ingo Molnar Cc: Peter Zijlstra Cc: Christoph Lameter Cc: Paul Menage Cc: Nick Piggin Cc: Yasunori Goto Cc: Pekka Enberg Cc: David Rientjes Cc: Lee Schermerhorn Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/nodemask.h | 28 ++++++++++++++++ mm/mempolicy.c | 84 +++++++++++++++++++++++++++++++++--------------- 2 files changed, 86 insertions(+), 26 deletions(-) (limited to 'include') diff --git a/include/linux/nodemask.h b/include/linux/nodemask.h index 829b94b156f2..b359c4a9ec9e 100644 --- a/include/linux/nodemask.h +++ b/include/linux/nodemask.h @@ -82,6 +82,12 @@ * to generate slightly worse code. So use a simple one-line #define * for node_isset(), instead of wrapping an inline inside a macro, the * way we do the other calls. + * + * NODEMASK_SCRATCH + * When doing above logical AND, OR, XOR, Remap operations the callers tend to + * need temporary nodemask_t's on the stack. But if NODES_SHIFT is large, + * nodemask_t's consume too much stack space. NODEMASK_SCRATCH is a helper + * for such situations. See below and CPUMASK_ALLOC also. */ #include @@ -473,4 +479,26 @@ static inline int num_node_state(enum node_states state) #define for_each_node(node) for_each_node_state(node, N_POSSIBLE) #define for_each_online_node(node) for_each_node_state(node, N_ONLINE) +/* + * For nodemask scrach area.(See CPUMASK_ALLOC() in cpumask.h) + */ + +#if NODES_SHIFT > 8 /* nodemask_t > 64 bytes */ +#define NODEMASK_ALLOC(x, m) struct x *m = kmalloc(sizeof(*m), GFP_KERNEL) +#define NODEMASK_FREE(m) kfree(m) +#else +#define NODEMASK_ALLOC(x, m) struct x _m, *m = &_m +#define NODEMASK_FREE(m) +#endif + +/* A example struture for using NODEMASK_ALLOC, used in mempolicy. */ +struct nodemask_scratch { + nodemask_t mask1; + nodemask_t mask2; +}; + +#define NODEMASK_SCRATCH(x) NODEMASK_ALLOC(nodemask_scratch, x) +#define NODEMASK_SCRATCH_FREE(x) NODEMASK_FREE(x) + + #endif /* __LINUX_NODEMASK_H */ diff --git a/mm/mempolicy.c b/mm/mempolicy.c index e08e2c4da63a..7dd9d9f80694 100644 --- a/mm/mempolicy.c +++ b/mm/mempolicy.c @@ -191,25 +191,27 @@ static int mpol_new_bind(struct mempolicy *pol, const nodemask_t *nodes) * Must be called holding task's alloc_lock to protect task's mems_allowed * and mempolicy. May also be called holding the mmap_semaphore for write. */ -static int mpol_set_nodemask(struct mempolicy *pol, const nodemask_t *nodes) +static int mpol_set_nodemask(struct mempolicy *pol, + const nodemask_t *nodes, struct nodemask_scratch *nsc) { - nodemask_t cpuset_context_nmask; int ret; /* if mode is MPOL_DEFAULT, pol is NULL. This is right. */ if (pol == NULL) return 0; + /* Check N_HIGH_MEMORY */ + nodes_and(nsc->mask1, + cpuset_current_mems_allowed, node_states[N_HIGH_MEMORY]); VM_BUG_ON(!nodes); if (pol->mode == MPOL_PREFERRED && nodes_empty(*nodes)) nodes = NULL; /* explicit local allocation */ else { if (pol->flags & MPOL_F_RELATIVE_NODES) - mpol_relative_nodemask(&cpuset_context_nmask, nodes, - &cpuset_current_mems_allowed); + mpol_relative_nodemask(&nsc->mask2, nodes,&nsc->mask1); else - nodes_and(cpuset_context_nmask, *nodes, - cpuset_current_mems_allowed); + nodes_and(nsc->mask2, *nodes, nsc->mask1); + if (mpol_store_user_nodemask(pol)) pol->w.user_nodemask = *nodes; else @@ -217,8 +219,10 @@ static int mpol_set_nodemask(struct mempolicy *pol, const nodemask_t *nodes) cpuset_current_mems_allowed; } - ret = mpol_ops[pol->mode].create(pol, - nodes ? &cpuset_context_nmask : NULL); + if (nodes) + ret = mpol_ops[pol->mode].create(pol, &nsc->mask2); + else + ret = mpol_ops[pol->mode].create(pol, NULL); return ret; } @@ -620,12 +624,17 @@ static long do_set_mempolicy(unsigned short mode, unsigned short flags, { struct mempolicy *new, *old; struct mm_struct *mm = current->mm; + NODEMASK_SCRATCH(scratch); int ret; - new = mpol_new(mode, flags, nodes); - if (IS_ERR(new)) - return PTR_ERR(new); + if (!scratch) + return -ENOMEM; + new = mpol_new(mode, flags, nodes); + if (IS_ERR(new)) { + ret = PTR_ERR(new); + goto out; + } /* * prevent changing our mempolicy while show_numa_maps() * is using it. @@ -635,13 +644,13 @@ static long do_set_mempolicy(unsigned short mode, unsigned short flags, if (mm) down_write(&mm->mmap_sem); task_lock(current); - ret = mpol_set_nodemask(new, nodes); + ret = mpol_set_nodemask(new, nodes, scratch); if (ret) { task_unlock(current); if (mm) up_write(&mm->mmap_sem); mpol_put(new); - return ret; + goto out; } old = current->mempolicy; current->mempolicy = new; @@ -654,7 +663,10 @@ static long do_set_mempolicy(unsigned short mode, unsigned short flags, up_write(&mm->mmap_sem); mpol_put(old); - return 0; + ret = 0; +out: + NODEMASK_SCRATCH_FREE(scratch); + return ret; } /* @@ -1014,12 +1026,20 @@ static long do_mbind(unsigned long start, unsigned long len, if (err) return err; } - down_write(&mm->mmap_sem); - task_lock(current); - err = mpol_set_nodemask(new, nmask); - task_unlock(current); + { + NODEMASK_SCRATCH(scratch); + if (scratch) { + down_write(&mm->mmap_sem); + task_lock(current); + err = mpol_set_nodemask(new, nmask, scratch); + task_unlock(current); + if (err) + up_write(&mm->mmap_sem); + } else + err = -ENOMEM; + NODEMASK_SCRATCH_FREE(scratch); + } if (err) { - up_write(&mm->mmap_sem); mpol_put(new); return err; } @@ -1891,6 +1911,7 @@ restart: * Install non-NULL @mpol in inode's shared policy rb-tree. * On entry, the current task has a reference on a non-NULL @mpol. * This must be released on exit. + * This is called at get_inode() calls and we can use GFP_KERNEL. */ void mpol_shared_policy_init(struct shared_policy *sp, struct mempolicy *mpol) { @@ -1902,19 +1923,24 @@ void mpol_shared_policy_init(struct shared_policy *sp, struct mempolicy *mpol) if (mpol) { struct vm_area_struct pvma; struct mempolicy *new; + NODEMASK_SCRATCH(scratch); + if (!scratch) + return; /* contextualize the tmpfs mount point mempolicy */ new = mpol_new(mpol->mode, mpol->flags, &mpol->w.user_nodemask); if (IS_ERR(new)) { mpol_put(mpol); /* drop our ref on sb mpol */ + NODEMASK_SCRATCH_FREE(scratch); return; /* no valid nodemask intersection */ } task_lock(current); - ret = mpol_set_nodemask(new, &mpol->w.user_nodemask); + ret = mpol_set_nodemask(new, &mpol->w.user_nodemask, scratch); task_unlock(current); mpol_put(mpol); /* drop our ref on sb mpol */ if (ret) { + NODEMASK_SCRATCH_FREE(scratch); mpol_put(new); return; } @@ -1924,6 +1950,7 @@ void mpol_shared_policy_init(struct shared_policy *sp, struct mempolicy *mpol) pvma.vm_end = TASK_SIZE; /* policy covers entire file */ mpol_set_shared_policy(sp, &pvma, new); /* adds ref */ mpol_put(new); /* drop initial ref */ + NODEMASK_SCRATCH_FREE(scratch); } } @@ -2140,13 +2167,18 @@ int mpol_parse_str(char *str, struct mempolicy **mpol, int no_context) err = 1; else { int ret; - - task_lock(current); - ret = mpol_set_nodemask(new, &nodes); - task_unlock(current); - if (ret) + NODEMASK_SCRATCH(scratch); + if (scratch) { + task_lock(current); + ret = mpol_set_nodemask(new, &nodes, scratch); + task_unlock(current); + } else + ret = -ENOMEM; + NODEMASK_SCRATCH_FREE(scratch); + if (ret) { err = 1; - else if (no_context) { + mpol_put(new); + } else if (no_context) { /* save for contextualization */ new->w.user_nodemask = nodes; } -- cgit v1.2.3 From daeb6b6fbe27049f465c48a7d0ee5555c3b84064 Mon Sep 17 00:00:00 2001 From: Phillip Lougher Date: Thu, 6 Aug 2009 15:09:30 -0700 Subject: bzip2/lzma/gzip: fix comments describing decompressor API Fix and improve comments in decompress/generic.h that describe the decompressor API. Also remove an unused definition, and rename INBUF_LEN in lib/decompress_inflate.c to conform to bzip2/lzma naming. Signed-off-by: Phillip Lougher Cc: "H. Peter Anvin" Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/decompress/generic.h | 32 +++++++++++++++++++------------- lib/decompress_inflate.c | 8 ++++---- 2 files changed, 23 insertions(+), 17 deletions(-) (limited to 'include') diff --git a/include/linux/decompress/generic.h b/include/linux/decompress/generic.h index 6dfb856327bb..0c7111a55a1a 100644 --- a/include/linux/decompress/generic.h +++ b/include/linux/decompress/generic.h @@ -1,31 +1,37 @@ #ifndef DECOMPRESS_GENERIC_H #define DECOMPRESS_GENERIC_H -/* Minimal chunksize to be read. - *Bzip2 prefers at least 4096 - *Lzma prefers 0x10000 */ -#define COMPR_IOBUF_SIZE 4096 - typedef int (*decompress_fn) (unsigned char *inbuf, int len, int(*fill)(void*, unsigned int), - int(*writebb)(void*, unsigned int), - unsigned char *output, + int(*flush)(void*, unsigned int), + unsigned char *outbuf, int *posp, void(*error)(char *x)); /* inbuf - input buffer *len - len of pre-read data in inbuf - *fill - function to fill inbuf if empty - *writebb - function to write out outbug + *fill - function to fill inbuf when empty + *flush - function to write out outbuf + *outbuf - output buffer *posp - if non-null, input position (number of bytes read) will be * returned here * - *If len != 0, the inbuf is initialized (with as much data), and fill - *should not be called - *If len = 0, the inbuf is allocated, but empty. Its size is IOBUF_SIZE - *fill should be called (repeatedly...) to read data, at most IOBUF_SIZE + *If len != 0, inbuf should contain all the necessary input data, and fill + *should be NULL + *If len = 0, inbuf can be NULL, in which case the decompressor will allocate + *the input buffer. If inbuf != NULL it must be at least XXX_IOBUF_SIZE bytes. + *fill will be called (repeatedly...) to read data, at most XXX_IOBUF_SIZE + *bytes should be read per call. Replace XXX with the appropriate decompressor + *name, i.e. LZMA_IOBUF_SIZE. + * + *If flush = NULL, outbuf must be large enough to buffer all the expected + *output. If flush != NULL, the output buffer will be allocated by the + *decompressor (outbuf = NULL), and the flush function will be called to + *flush the output buffer at the appropriate time (decompressor and stream + *dependent). */ + /* Utility routine to detect the decompression method */ decompress_fn decompress_method(const unsigned char *inbuf, int len, const char **name); diff --git a/lib/decompress_inflate.c b/lib/decompress_inflate.c index e36b296fc9f8..bfe605ac534f 100644 --- a/lib/decompress_inflate.c +++ b/lib/decompress_inflate.c @@ -25,7 +25,7 @@ #include #include -#define INBUF_LEN (16*1024) +#define GZIP_IOBUF_SIZE (16*1024) /* Included from initramfs et al code */ STATIC int INIT gunzip(unsigned char *buf, int len, @@ -55,7 +55,7 @@ STATIC int INIT gunzip(unsigned char *buf, int len, if (buf) zbuf = buf; else { - zbuf = malloc(INBUF_LEN); + zbuf = malloc(GZIP_IOBUF_SIZE); len = 0; } if (!zbuf) { @@ -77,7 +77,7 @@ STATIC int INIT gunzip(unsigned char *buf, int len, } if (len == 0) - len = fill(zbuf, INBUF_LEN); + len = fill(zbuf, GZIP_IOBUF_SIZE); /* verify the gzip header */ if (len < 10 || @@ -113,7 +113,7 @@ STATIC int INIT gunzip(unsigned char *buf, int len, while (rc == Z_OK) { if (strm->avail_in == 0) { /* TODO: handle case where both pos and fill are set */ - len = fill(zbuf, INBUF_LEN); + len = fill(zbuf, GZIP_IOBUF_SIZE); if (len < 0) { rc = -1; error("read error"); -- cgit v1.2.3 From fa56d4cb4022c8b313c3b99236e1b87effc3655b Mon Sep 17 00:00:00 2001 From: Bartlomiej Zolnierkiewicz Date: Tue, 23 Jun 2009 11:29:11 +0000 Subject: ide: allow ide_dev_read_id() to be called from the IRQ context * Un-static __ide_wait_stat(). * Allow ide_dev_read_id() helper to be called from the IRQ context by adding irq_ctx flag and using mdelay()/__ide_wait_stat() when needed. * Switch ide_driveid_update() to set irq_ctx flag. This change is needed for the consecutive patch which fixes races in handling of user-space SET XFER commands but for improved bisectability and clarity it is better to do it in a separate patch. Signed-off-by: Bartlomiej Zolnierkiewicz Signed-off-by: David S. Miller --- drivers/ide/ide-iops.c | 6 +++--- drivers/ide/ide-probe.c | 31 +++++++++++++++++++++---------- include/linux/ide.h | 3 ++- 3 files changed, 26 insertions(+), 14 deletions(-) (limited to 'include') diff --git a/drivers/ide/ide-iops.c b/drivers/ide/ide-iops.c index 2892b242bbe1..b99873845d21 100644 --- a/drivers/ide/ide-iops.c +++ b/drivers/ide/ide-iops.c @@ -102,8 +102,8 @@ EXPORT_SYMBOL(ide_fixstring); * setting a timer to wake up at half second intervals thereafter, * until timeout is achieved, before timing out. */ -static int __ide_wait_stat(ide_drive_t *drive, u8 good, u8 bad, - unsigned long timeout, u8 *rstat) +int __ide_wait_stat(ide_drive_t *drive, u8 good, u8 bad, + unsigned long timeout, u8 *rstat) { ide_hwif_t *hwif = drive->hwif; const struct ide_tp_ops *tp_ops = hwif->tp_ops; @@ -316,7 +316,7 @@ int ide_driveid_update(ide_drive_t *drive) return 0; SELECT_MASK(drive, 1); - rc = ide_dev_read_id(drive, ATA_CMD_ID_ATA, id); + rc = ide_dev_read_id(drive, ATA_CMD_ID_ATA, id, 1); SELECT_MASK(drive, 0); if (rc) diff --git a/drivers/ide/ide-probe.c b/drivers/ide/ide-probe.c index 1bb106f6221a..8de442cbee94 100644 --- a/drivers/ide/ide-probe.c +++ b/drivers/ide/ide-probe.c @@ -238,6 +238,7 @@ static void do_identify(ide_drive_t *drive, u8 cmd, u16 *id) * @drive: drive to identify * @cmd: command to use * @id: buffer for IDENTIFY data + * @irq_ctx: flag set when called from the IRQ context * * Sends an ATA(PI) IDENTIFY request to a drive and waits for a response. * @@ -246,7 +247,7 @@ static void do_identify(ide_drive_t *drive, u8 cmd, u16 *id) * 2 device aborted the command (refused to identify itself) */ -int ide_dev_read_id(ide_drive_t *drive, u8 cmd, u16 *id) +int ide_dev_read_id(ide_drive_t *drive, u8 cmd, u16 *id, int irq_ctx) { ide_hwif_t *hwif = drive->hwif; struct ide_io_ports *io_ports = &hwif->io_ports; @@ -263,7 +264,10 @@ int ide_dev_read_id(ide_drive_t *drive, u8 cmd, u16 *id) tp_ops->write_devctl(hwif, ATA_NIEN | ATA_DEVCTL_OBS); /* take a deep breath */ - msleep(50); + if (irq_ctx) + mdelay(50); + else + msleep(50); if (io_ports->ctl_addr && (hwif->host_flags & IDE_HFLAG_BROKEN_ALTSTATUS) == 0) { @@ -295,12 +299,19 @@ int ide_dev_read_id(ide_drive_t *drive, u8 cmd, u16 *id) timeout = ((cmd == ATA_CMD_ID_ATA) ? WAIT_WORSTCASE : WAIT_PIDENTIFY) / 2; - if (ide_busy_sleep(drive, timeout, use_altstatus)) - return 1; - /* wait for IRQ and ATA_DRQ */ - msleep(50); - s = tp_ops->read_status(hwif); + if (irq_ctx) { + rc = __ide_wait_stat(drive, ATA_DRQ, BAD_R_STAT, timeout, &s); + if (rc) + return 1; + } else { + rc = ide_busy_sleep(drive, timeout, use_altstatus); + if (rc) + return 1; + + msleep(50); + s = tp_ops->read_status(hwif); + } if (OK_STAT(s, ATA_DRQ, BAD_R_STAT)) { /* drive returned ID */ @@ -406,10 +417,10 @@ static int do_probe (ide_drive_t *drive, u8 cmd) if (OK_STAT(stat, ATA_DRDY, ATA_BUSY) || present || cmd == ATA_CMD_ID_ATAPI) { - rc = ide_dev_read_id(drive, cmd, id); + rc = ide_dev_read_id(drive, cmd, id, 0); if (rc) /* failed: try again */ - rc = ide_dev_read_id(drive, cmd, id); + rc = ide_dev_read_id(drive, cmd, id, 0); stat = tp_ops->read_status(hwif); @@ -424,7 +435,7 @@ static int do_probe (ide_drive_t *drive, u8 cmd) msleep(50); tp_ops->exec_command(hwif, ATA_CMD_DEV_RESET); (void)ide_busy_sleep(drive, WAIT_WORSTCASE, 0); - rc = ide_dev_read_id(drive, cmd, id); + rc = ide_dev_read_id(drive, cmd, id, 0); } /* ensure drive IRQ is clear */ diff --git a/include/linux/ide.h b/include/linux/ide.h index edc93a6d931d..cb6cd0459a5e 100644 --- a/include/linux/ide.h +++ b/include/linux/ide.h @@ -1081,6 +1081,7 @@ extern void ide_fixstring(u8 *, const int, const int); int ide_busy_sleep(ide_drive_t *, unsigned long, int); +int __ide_wait_stat(ide_drive_t *, u8, u8, unsigned long, u8 *); int ide_wait_stat(ide_startstop_t *, ide_drive_t *, u8, u8, unsigned long); ide_startstop_t ide_do_park_unpark(ide_drive_t *, struct request *); @@ -1169,7 +1170,7 @@ int ide_no_data_taskfile(ide_drive_t *, struct ide_cmd *); int ide_taskfile_ioctl(ide_drive_t *, unsigned long); -int ide_dev_read_id(ide_drive_t *, u8, u16 *); +int ide_dev_read_id(ide_drive_t *, u8, u16 *, int); extern int ide_driveid_update(ide_drive_t *); extern int ide_config_drive_speed(ide_drive_t *, u8); -- cgit v1.2.3 From 665d66e8fad60a5a162c4615f27f916ad1a6d567 Mon Sep 17 00:00:00 2001 From: Bartlomiej Zolnierkiewicz Date: Tue, 23 Jun 2009 11:35:51 +0000 Subject: ide: fix races in handling of user-space SET XFER commands * Make cmd->tf_flags field 'u16' and add IDE_TFLAG_SET_XFER taskfile flag. * Update ide_finish_cmd() to set xfer / re-read id if the new flag is set. * Convert set_xfer_rate() (write handler for /proc/ide/hd?/current_speed) and ide_cmd_ioctl() (HDIO_DRIVE_CMD ioctl handler) to use the new flag. * Remove no longer needed disable_irq_nosync() + enable_irq() from ide_config_drive_speed(). Signed-off-by: Bartlomiej Zolnierkiewicz Signed-off-by: David S. Miller --- drivers/ide/ide-ioctls.c | 8 ++------ drivers/ide/ide-iops.c | 10 ---------- drivers/ide/ide-proc.c | 10 ++-------- drivers/ide/ide-taskfile.c | 9 ++++++++- include/linux/ide.h | 3 ++- 5 files changed, 14 insertions(+), 26 deletions(-) (limited to 'include') diff --git a/drivers/ide/ide-ioctls.c b/drivers/ide/ide-ioctls.c index e246d3d3fbcc..d3440b5010a5 100644 --- a/drivers/ide/ide-ioctls.c +++ b/drivers/ide/ide-ioctls.c @@ -167,6 +167,8 @@ static int ide_cmd_ioctl(ide_drive_t *drive, unsigned long arg) err = -EINVAL; goto abort; } + + cmd.tf_flags |= IDE_TFLAG_SET_XFER; } err = ide_raw_taskfile(drive, &cmd, buf, args[3]); @@ -174,12 +176,6 @@ static int ide_cmd_ioctl(ide_drive_t *drive, unsigned long arg) args[0] = tf->status; args[1] = tf->error; args[2] = tf->nsect; - - if (!err && xfer_rate) { - /* active-retuning-calls future */ - ide_set_xfer_rate(drive, xfer_rate); - ide_driveid_update(drive); - } abort: if (copy_to_user((void __user *)arg, &args, 4)) err = -EFAULT; diff --git a/drivers/ide/ide-iops.c b/drivers/ide/ide-iops.c index b99873845d21..b14fa9a87c49 100644 --- a/drivers/ide/ide-iops.c +++ b/drivers/ide/ide-iops.c @@ -363,14 +363,6 @@ int ide_config_drive_speed(ide_drive_t *drive, u8 speed) * this point (lost interrupt). */ - /* - * FIXME: we race against the running IRQ here if - * this is called from non IRQ context. If we use - * disable_irq() we hang on the error path. Work - * is needed. - */ - disable_irq_nosync(hwif->irq); - udelay(1); tp_ops->dev_select(drive); SELECT_MASK(drive, 1); @@ -394,8 +386,6 @@ int ide_config_drive_speed(ide_drive_t *drive, u8 speed) SELECT_MASK(drive, 0); - enable_irq(hwif->irq); - if (error) { (void) ide_dump_status(drive, "set_drive_speed_status", stat); return error; diff --git a/drivers/ide/ide-proc.c b/drivers/ide/ide-proc.c index 3242698832a4..021de41655e6 100644 --- a/drivers/ide/ide-proc.c +++ b/drivers/ide/ide-proc.c @@ -195,7 +195,6 @@ ide_devset_get(xfer_rate, current_speed); static int set_xfer_rate (ide_drive_t *drive, int arg) { struct ide_cmd cmd; - int err; if (arg < XFER_PIO_0 || arg > XFER_UDMA_6) return -EINVAL; @@ -206,14 +205,9 @@ static int set_xfer_rate (ide_drive_t *drive, int arg) cmd.tf.nsect = (u8)arg; cmd.valid.out.tf = IDE_VALID_FEATURE | IDE_VALID_NSECT; cmd.valid.in.tf = IDE_VALID_NSECT; + cmd.tf_flags = IDE_TFLAG_SET_XFER; - err = ide_no_data_taskfile(drive, &cmd); - - if (!err) { - ide_set_xfer_rate(drive, (u8) arg); - ide_driveid_update(drive); - } - return err; + return ide_no_data_taskfile(drive, &cmd); } ide_devset_rw(current_speed, xfer_rate); diff --git a/drivers/ide/ide-taskfile.c b/drivers/ide/ide-taskfile.c index 50336d51eebc..cc8633cbe133 100644 --- a/drivers/ide/ide-taskfile.c +++ b/drivers/ide/ide-taskfile.c @@ -324,10 +324,17 @@ static void ide_error_cmd(ide_drive_t *drive, struct ide_cmd *cmd) void ide_finish_cmd(ide_drive_t *drive, struct ide_cmd *cmd, u8 stat) { struct request *rq = drive->hwif->rq; - u8 err = ide_read_error(drive); + u8 err = ide_read_error(drive), nsect = cmd->tf.nsect; + u8 set_xfer = !!(cmd->tf_flags & IDE_TFLAG_SET_XFER); ide_complete_cmd(drive, cmd, stat, err); rq->errors = err; + + if (err == 0 && set_xfer) { + ide_set_xfer_rate(drive, nsect); + ide_driveid_update(drive); + } + ide_complete_rq(drive, err ? -EIO : 0, blk_rq_bytes(rq)); } diff --git a/include/linux/ide.h b/include/linux/ide.h index cb6cd0459a5e..803c1ae31237 100644 --- a/include/linux/ide.h +++ b/include/linux/ide.h @@ -258,6 +258,7 @@ enum { IDE_TFLAG_DYN = (1 << 5), IDE_TFLAG_FS = (1 << 6), IDE_TFLAG_MULTI_PIO = (1 << 7), + IDE_TFLAG_SET_XFER = (1 << 8), }; enum { @@ -294,7 +295,7 @@ struct ide_cmd { } out, in; } valid; - u8 tf_flags; + u16 tf_flags; u8 ftf_flags; /* for TASKFILE ioctl */ int protocol; -- cgit v1.2.3 From 3a6593050fbd8bbcaed3a44d01c31d907315c86c Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Tue, 21 Jul 2009 17:34:57 +0200 Subject: perf_counter, ftrace: Fix perf_counter integration Adds possible second part to the assign argument of TP_EVENT(). TP_perf_assign( __perf_count(foo); __perf_addr(bar); ) Which, when specified make the swcounter increment with @foo instead of the usual 1, and report @bar for PERF_SAMPLE_ADDR (data address associated with the event) when this triggers a counter overflow. Signed-off-by: Peter Zijlstra Acked-by: Steven Rostedt Cc: Frederic Weisbecker Cc: Jason Baron Cc: Paul Mackerras Signed-off-by: Ingo Molnar --- include/trace/ftrace.h | 110 ++++++++++++++++++++++++++++++++++++++----------- kernel/perf_counter.c | 6 +-- 2 files changed, 88 insertions(+), 28 deletions(-) (limited to 'include') diff --git a/include/trace/ftrace.h b/include/trace/ftrace.h index 1867553c61e5..fec71f8dbc48 100644 --- a/include/trace/ftrace.h +++ b/include/trace/ftrace.h @@ -144,6 +144,9 @@ #undef TP_fast_assign #define TP_fast_assign(args...) args +#undef TP_perf_assign +#define TP_perf_assign(args...) + #undef TRACE_EVENT #define TRACE_EVENT(call, proto, args, tstruct, func, print) \ static int \ @@ -345,6 +348,88 @@ static inline int ftrace_get_offsets_##call( \ #include TRACE_INCLUDE(TRACE_INCLUDE_FILE) +#ifdef CONFIG_EVENT_PROFILE + +/* + * Generate the functions needed for tracepoint perf_counter support. + * + * static void ftrace_profile_(proto) + * { + * extern void perf_tpcounter_event(int, u64, u64); + * u64 __addr = 0, __count = 1; + * + * <-- here we expand the TP_perf_assign() macro + * + * perf_tpcounter_event(event_.id, __addr, __count); + * } + * + * static int ftrace_profile_enable_(struct ftrace_event_call *event_call) + * { + * int ret = 0; + * + * if (!atomic_inc_return(&event_call->profile_count)) + * ret = register_trace_(ftrace_profile_); + * + * return ret; + * } + * + * static void ftrace_profile_disable_(struct ftrace_event_call *event_call) + * { + * if (atomic_add_negative(-1, &event->call->profile_count)) + * unregister_trace_(ftrace_profile_); + * } + * + */ + +#undef TP_fast_assign +#define TP_fast_assign(args...) + +#undef TP_perf_assign +#define TP_perf_assign(args...) args + +#undef __perf_addr +#define __perf_addr(a) __addr = (a) + +#undef __perf_count +#define __perf_count(c) __count = (c) + +#undef TRACE_EVENT +#define TRACE_EVENT(call, proto, args, tstruct, assign, print) \ + \ +static void ftrace_profile_##call(proto) \ +{ \ + extern void perf_tpcounter_event(int, u64, u64); \ + u64 __addr = 0, __count = 1; \ + { assign; } \ + perf_tpcounter_event(event_##call.id, __addr, __count); \ +} \ + \ +static int ftrace_profile_enable_##call(struct ftrace_event_call *event_call) \ +{ \ + int ret = 0; \ + \ + if (!atomic_inc_return(&event_call->profile_count)) \ + ret = register_trace_##call(ftrace_profile_##call); \ + \ + return ret; \ +} \ + \ +static void ftrace_profile_disable_##call(struct ftrace_event_call *event_call)\ +{ \ + if (atomic_add_negative(-1, &event_call->profile_count)) \ + unregister_trace_##call(ftrace_profile_##call); \ +} + +#include TRACE_INCLUDE(TRACE_INCLUDE_FILE) + +#undef TP_fast_assign +#define TP_fast_assign(args...) args + +#undef TP_perf_assign +#define TP_perf_assign(args...) + +#endif + /* * Stage 4 of the trace events. * @@ -447,28 +532,6 @@ static inline int ftrace_get_offsets_##call( \ #define TP_FMT(fmt, args...) fmt "\n", ##args #ifdef CONFIG_EVENT_PROFILE -#define _TRACE_PROFILE(call, proto, args) \ -static void ftrace_profile_##call(proto) \ -{ \ - extern void perf_tpcounter_event(int); \ - perf_tpcounter_event(event_##call.id); \ -} \ - \ -static int ftrace_profile_enable_##call(struct ftrace_event_call *event_call) \ -{ \ - int ret = 0; \ - \ - if (!atomic_inc_return(&event_call->profile_count)) \ - ret = register_trace_##call(ftrace_profile_##call); \ - \ - return ret; \ -} \ - \ -static void ftrace_profile_disable_##call(struct ftrace_event_call *event_call)\ -{ \ - if (atomic_add_negative(-1, &event_call->profile_count)) \ - unregister_trace_##call(ftrace_profile_##call); \ -} #define _TRACE_PROFILE_INIT(call) \ .profile_count = ATOMIC_INIT(-1), \ @@ -476,7 +539,6 @@ static void ftrace_profile_disable_##call(struct ftrace_event_call *event_call)\ .profile_disable = ftrace_profile_disable_##call, #else -#define _TRACE_PROFILE(call, proto, args) #define _TRACE_PROFILE_INIT(call) #endif @@ -502,7 +564,6 @@ static void ftrace_profile_disable_##call(struct ftrace_event_call *event_call)\ #undef TRACE_EVENT #define TRACE_EVENT(call, proto, args, tstruct, assign, print) \ -_TRACE_PROFILE(call, PARAMS(proto), PARAMS(args)) \ \ static struct ftrace_event_call event_##call; \ \ @@ -586,6 +647,5 @@ __attribute__((section("_ftrace_events"))) event_##call = { \ #include TRACE_INCLUDE(TRACE_INCLUDE_FILE) -#undef _TRACE_PROFILE #undef _TRACE_PROFILE_INIT diff --git a/kernel/perf_counter.c b/kernel/perf_counter.c index 673c1aaf7332..52eb4b68d34f 100644 --- a/kernel/perf_counter.c +++ b/kernel/perf_counter.c @@ -3703,17 +3703,17 @@ static const struct pmu perf_ops_task_clock = { }; #ifdef CONFIG_EVENT_PROFILE -void perf_tpcounter_event(int event_id) +void perf_tpcounter_event(int event_id, u64 addr, u64 count) { struct perf_sample_data data = { .regs = get_irq_regs(), - .addr = 0, + .addr = addr, }; if (!data.regs) data.regs = task_pt_regs(current); - do_perf_swcounter_event(PERF_TYPE_TRACEPOINT, event_id, 1, 1, &data); + do_perf_swcounter_event(PERF_TYPE_TRACEPOINT, event_id, count, 1, &data); } EXPORT_SYMBOL_GPL(perf_tpcounter_event); -- cgit v1.2.3 From f413cdb80ce00ec1a4d0ab949b5d96c81cae7f75 Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Fri, 7 Aug 2009 01:25:54 +0200 Subject: perf_counter: Fix/complete ftrace event records sampling This patch implements the kernel side support for ftrace event record sampling. A new counter sampling attribute is added: PERF_SAMPLE_TP_RECORD which requests ftrace events record sampling. In this case if a PERF_TYPE_TRACEPOINT counter is active and a tracepoint fires, we emit the tracepoint binary record to the perfcounter event buffer, as a sample. Result, after setting PERF_SAMPLE_TP_RECORD attribute from perf record: perf record -f -F 1 -a -e workqueue:workqueue_execution perf report -D 0x21e18 [0x48]: event: 9 . . ... raw event: size 72 bytes . 0000: 09 00 00 00 01 00 48 00 d0 c7 00 81 ff ff ff ff ......H........ . 0010: 0a 00 00 00 0a 00 00 00 21 00 00 00 00 00 00 00 ........!...... . 0020: 2b 00 01 02 0a 00 00 00 0a 00 00 00 65 76 65 6e +...........eve . 0030: 74 73 2f 31 00 00 00 00 00 00 00 00 0a 00 00 00 ts/1........... . 0040: e0 b1 31 81 ff ff ff ff ....... . 0x21e18 [0x48]: PERF_EVENT_SAMPLE (IP, 1): 10: 0xffffffff8100c7d0 period: 33 The raw ftrace binary record starts at offset 0020. Translation: struct trace_entry { type = 0x2b = 43; flags = 1; preempt_count = 2; pid = 0xa = 10; tgid = 0xa = 10; } thread_comm = "events/1" thread_pid = 0xa = 10; func = 0xffffffff8131b1e0 = flush_to_ldisc() What will come next? - Userspace support ('perf trace'), 'flight data recorder' mode for perf trace, etc. - The unconditional copy from the profiling callback brings some costs however if someone wants no such sampling to occur, and needs to be fixed in the future. For that we need to have an instant access to the perf counter attribute. This is a matter of a flag to add in the struct ftrace_event. - Take care of the events recursivity! Don't ever try to record a lock event for example, it seems some locking is used in the profiling fast path and lead to a tracing recursivity. That will be fixed using raw spinlock or recursivity protection. - [...] - Profit! :-) Signed-off-by: Frederic Weisbecker Cc: Li Zefan Cc: Tom Zanussi Cc: Arnaldo Carvalho de Melo Cc: Peter Zijlstra Cc: Mike Galbraith Cc: Steven Rostedt Cc: Paul Mackerras Cc: Pekka Enberg Cc: Gabriel Munteanu Cc: Lai Jiangshan Signed-off-by: Ingo Molnar --- include/linux/ftrace_event.h | 4 +- include/linux/perf_counter.h | 9 ++- include/trace/ftrace.h | 130 ++++++++++++++++++++++++++++++++----------- kernel/perf_counter.c | 18 +++++- kernel/trace/trace.c | 1 + kernel/trace/trace.h | 4 -- tools/perf/builtin-record.c | 1 + 7 files changed, 126 insertions(+), 41 deletions(-) (limited to 'include') diff --git a/include/linux/ftrace_event.h b/include/linux/ftrace_event.h index d7cd193c2277..a81170de7f6b 100644 --- a/include/linux/ftrace_event.h +++ b/include/linux/ftrace_event.h @@ -89,7 +89,9 @@ enum print_line_t { TRACE_TYPE_NO_CONSUME = 3 /* Handled but ask to not consume */ }; - +void tracing_generic_entry_update(struct trace_entry *entry, + unsigned long flags, + int pc); struct ring_buffer_event * trace_current_buffer_lock_reserve(int type, unsigned long len, unsigned long flags, int pc); diff --git a/include/linux/perf_counter.h b/include/linux/perf_counter.h index e604e6ef72dd..a67dd5c5b6d3 100644 --- a/include/linux/perf_counter.h +++ b/include/linux/perf_counter.h @@ -121,8 +121,9 @@ enum perf_counter_sample_format { PERF_SAMPLE_CPU = 1U << 7, PERF_SAMPLE_PERIOD = 1U << 8, PERF_SAMPLE_STREAM_ID = 1U << 9, + PERF_SAMPLE_TP_RECORD = 1U << 10, - PERF_SAMPLE_MAX = 1U << 10, /* non-ABI */ + PERF_SAMPLE_MAX = 1U << 11, /* non-ABI */ }; /* @@ -413,6 +414,11 @@ struct perf_callchain_entry { __u64 ip[PERF_MAX_STACK_DEPTH]; }; +struct perf_tracepoint_record { + int size; + char *record; +}; + struct task_struct; /** @@ -681,6 +687,7 @@ struct perf_sample_data { struct pt_regs *regs; u64 addr; u64 period; + void *private; }; extern int perf_counter_overflow(struct perf_counter *counter, int nmi, diff --git a/include/trace/ftrace.h b/include/trace/ftrace.h index fec71f8dbc48..7fb16d90e7b1 100644 --- a/include/trace/ftrace.h +++ b/include/trace/ftrace.h @@ -353,15 +353,7 @@ static inline int ftrace_get_offsets_##call( \ /* * Generate the functions needed for tracepoint perf_counter support. * - * static void ftrace_profile_(proto) - * { - * extern void perf_tpcounter_event(int, u64, u64); - * u64 __addr = 0, __count = 1; - * - * <-- here we expand the TP_perf_assign() macro - * - * perf_tpcounter_event(event_.id, __addr, __count); - * } + * NOTE: The insertion profile callback (ftrace_profile_) is defined later * * static int ftrace_profile_enable_(struct ftrace_event_call *event_call) * { @@ -381,28 +373,10 @@ static inline int ftrace_get_offsets_##call( \ * */ -#undef TP_fast_assign -#define TP_fast_assign(args...) - -#undef TP_perf_assign -#define TP_perf_assign(args...) args - -#undef __perf_addr -#define __perf_addr(a) __addr = (a) - -#undef __perf_count -#define __perf_count(c) __count = (c) - #undef TRACE_EVENT #define TRACE_EVENT(call, proto, args, tstruct, assign, print) \ \ -static void ftrace_profile_##call(proto) \ -{ \ - extern void perf_tpcounter_event(int, u64, u64); \ - u64 __addr = 0, __count = 1; \ - { assign; } \ - perf_tpcounter_event(event_##call.id, __addr, __count); \ -} \ +static void ftrace_profile_##call(proto); \ \ static int ftrace_profile_enable_##call(struct ftrace_event_call *event_call) \ { \ @@ -422,12 +396,6 @@ static void ftrace_profile_disable_##call(struct ftrace_event_call *event_call)\ #include TRACE_INCLUDE(TRACE_INCLUDE_FILE) -#undef TP_fast_assign -#define TP_fast_assign(args...) args - -#undef TP_perf_assign -#define TP_perf_assign(args...) - #endif /* @@ -647,5 +615,99 @@ __attribute__((section("_ftrace_events"))) event_##call = { \ #include TRACE_INCLUDE(TRACE_INCLUDE_FILE) +/* + * Define the insertion callback to profile events + * + * The job is very similar to ftrace_raw_event_ except that we don't + * insert in the ring buffer but in a perf counter. + * + * static void ftrace_profile_(proto) + * { + * struct ftrace_data_offsets_ __maybe_unused __data_offsets; + * struct ftrace_event_call *event_call = &event_; + * extern void perf_tpcounter_event(int, u64, u64, void *, int); + * struct ftrace_raw_##call *entry; + * u64 __addr = 0, __count = 1; + * unsigned long irq_flags; + * int __entry_size; + * int __data_size; + * int pc; + * + * local_save_flags(irq_flags); + * pc = preempt_count(); + * + * __data_size = ftrace_get_offsets_(&__data_offsets, args); + * __entry_size = __data_size + sizeof(*entry); + * + * do { + * char raw_data[__entry_size]; <- allocate our sample in the stack + * struct trace_entry *ent; + * + * entry = (struct ftrace_raw_ *)raw_data; + * ent = &entry->ent; + * tracing_generic_entry_update(ent, irq_flags, pc); + * ent->type = event_call->id; + * + * <- do some jobs with dynamic arrays + * + * <- affect our values + * + * perf_tpcounter_event(event_call->id, __addr, __count, entry, + * __entry_size); <- submit them to perf counter + * } while (0); + * + * } + */ + +#ifdef CONFIG_EVENT_PROFILE + +#undef __perf_addr +#define __perf_addr(a) __addr = (a) + +#undef __perf_count +#define __perf_count(c) __count = (c) + +#undef TRACE_EVENT +#define TRACE_EVENT(call, proto, args, tstruct, assign, print) \ +static void ftrace_profile_##call(proto) \ +{ \ + struct ftrace_data_offsets_##call __maybe_unused __data_offsets;\ + struct ftrace_event_call *event_call = &event_##call; \ + extern void perf_tpcounter_event(int, u64, u64, void *, int); \ + struct ftrace_raw_##call *entry; \ + u64 __addr = 0, __count = 1; \ + unsigned long irq_flags; \ + int __entry_size; \ + int __data_size; \ + int pc; \ + \ + local_save_flags(irq_flags); \ + pc = preempt_count(); \ + \ + __data_size = ftrace_get_offsets_##call(&__data_offsets, args); \ + __entry_size = ALIGN(__data_size + sizeof(*entry), sizeof(u64));\ + \ + do { \ + char raw_data[__entry_size]; \ + struct trace_entry *ent; \ + \ + entry = (struct ftrace_raw_##call *)raw_data; \ + ent = &entry->ent; \ + tracing_generic_entry_update(ent, irq_flags, pc); \ + ent->type = event_call->id; \ + \ + tstruct \ + \ + { assign; } \ + \ + perf_tpcounter_event(event_call->id, __addr, __count, entry,\ + __entry_size); \ + } while (0); \ + \ +} + +#include TRACE_INCLUDE(TRACE_INCLUDE_FILE) +#endif /* CONFIG_EVENT_PROFILE */ + #undef _TRACE_PROFILE_INIT diff --git a/kernel/perf_counter.c b/kernel/perf_counter.c index 52eb4b68d34f..868102172aa4 100644 --- a/kernel/perf_counter.c +++ b/kernel/perf_counter.c @@ -2646,6 +2646,7 @@ static void perf_counter_output(struct perf_counter *counter, int nmi, u64 counter; } group_entry; struct perf_callchain_entry *callchain = NULL; + struct perf_tracepoint_record *tp; int callchain_size = 0; u64 time; struct { @@ -2714,6 +2715,11 @@ static void perf_counter_output(struct perf_counter *counter, int nmi, header.size += sizeof(u64); } + if (sample_type & PERF_SAMPLE_TP_RECORD) { + tp = data->private; + header.size += tp->size; + } + ret = perf_output_begin(&handle, counter, header.size, nmi, 1); if (ret) return; @@ -2777,6 +2783,9 @@ static void perf_counter_output(struct perf_counter *counter, int nmi, } } + if (sample_type & PERF_SAMPLE_TP_RECORD) + perf_output_copy(&handle, tp->record, tp->size); + perf_output_end(&handle); } @@ -3703,11 +3712,18 @@ static const struct pmu perf_ops_task_clock = { }; #ifdef CONFIG_EVENT_PROFILE -void perf_tpcounter_event(int event_id, u64 addr, u64 count) +void perf_tpcounter_event(int event_id, u64 addr, u64 count, void *record, + int entry_size) { + struct perf_tracepoint_record tp = { + .size = entry_size, + .record = record, + }; + struct perf_sample_data data = { .regs = get_irq_regs(), .addr = addr, + .private = &tp, }; if (!data.regs) diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 8930e39b9d8c..c22b40f8f576 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -848,6 +848,7 @@ tracing_generic_entry_update(struct trace_entry *entry, unsigned long flags, ((pc & SOFTIRQ_MASK) ? TRACE_FLAG_SOFTIRQ : 0) | (need_resched() ? TRACE_FLAG_NEED_RESCHED : 0); } +EXPORT_SYMBOL_GPL(tracing_generic_entry_update); struct ring_buffer_event *trace_buffer_lock_reserve(struct trace_array *tr, int type, diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h index 3548ae5cc780..8b9f4f6e9559 100644 --- a/kernel/trace/trace.h +++ b/kernel/trace/trace.h @@ -438,10 +438,6 @@ struct trace_entry *tracing_get_trace_entry(struct trace_array *tr, struct trace_entry *trace_find_next_entry(struct trace_iterator *iter, int *ent_cpu, u64 *ent_ts); -void tracing_generic_entry_update(struct trace_entry *entry, - unsigned long flags, - int pc); - void default_wait_pipe(struct trace_iterator *iter); void poll_wait_pipe(struct trace_iterator *iter); diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c index 6da09928130f..90c98082af10 100644 --- a/tools/perf/builtin-record.c +++ b/tools/perf/builtin-record.c @@ -412,6 +412,7 @@ static void create_counter(int counter, int cpu, pid_t pid) if (call_graph) attr->sample_type |= PERF_SAMPLE_CALLCHAIN; + attr->mmap = track; attr->comm = track; attr->inherit = (cpu < 0) && inherit; -- cgit v1.2.3 From 3a43ce68ae1758fa6a839386025ef45acb6baa22 Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Sat, 8 Aug 2009 04:26:37 +0200 Subject: perf_counter: Fix tracepoint sampling to be part of generic sampling Based on Peter's comments, make tracepoint sampling generic just like all the other sampling bits are. This is a rename with no code changes: - PERF_SAMPLE_TP_RECORD to PERF_SAMPLE_RAW - struct perf_tracepoint_record to perf_raw_record We want the system in place that transport tracepoints raw samples events into the perf ring buffer to be generalized and usable by any type of counter. Reported-by; Peter Zijlstra Signed-off-by: Frederic Weisbecker Cc: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: Mike Galbraith Cc: Paul Mackerras LKML-Reference: <1249698400-5441-4-git-send-email-fweisbec@gmail.com> Signed-off-by: Ingo Molnar --- include/linux/perf_counter.h | 10 +++++----- kernel/perf_counter.c | 20 ++++++++++---------- 2 files changed, 15 insertions(+), 15 deletions(-) (limited to 'include') diff --git a/include/linux/perf_counter.h b/include/linux/perf_counter.h index a67dd5c5b6d3..2aabe43c1d04 100644 --- a/include/linux/perf_counter.h +++ b/include/linux/perf_counter.h @@ -121,7 +121,7 @@ enum perf_counter_sample_format { PERF_SAMPLE_CPU = 1U << 7, PERF_SAMPLE_PERIOD = 1U << 8, PERF_SAMPLE_STREAM_ID = 1U << 9, - PERF_SAMPLE_TP_RECORD = 1U << 10, + PERF_SAMPLE_RAW = 1U << 10, PERF_SAMPLE_MAX = 1U << 11, /* non-ABI */ }; @@ -414,9 +414,9 @@ struct perf_callchain_entry { __u64 ip[PERF_MAX_STACK_DEPTH]; }; -struct perf_tracepoint_record { - int size; - char *record; +struct perf_raw_record { + u32 size; + void *data; }; struct task_struct; @@ -687,7 +687,7 @@ struct perf_sample_data { struct pt_regs *regs; u64 addr; u64 period; - void *private; + struct perf_raw_record *raw; }; extern int perf_counter_overflow(struct perf_counter *counter, int nmi, diff --git a/kernel/perf_counter.c b/kernel/perf_counter.c index 117622cb73a3..002310540417 100644 --- a/kernel/perf_counter.c +++ b/kernel/perf_counter.c @@ -2646,7 +2646,7 @@ static void perf_counter_output(struct perf_counter *counter, int nmi, u64 counter; } group_entry; struct perf_callchain_entry *callchain = NULL; - struct perf_tracepoint_record *tp = NULL; + struct perf_raw_record *raw = NULL; int callchain_size = 0; u64 time; struct { @@ -2715,10 +2715,10 @@ static void perf_counter_output(struct perf_counter *counter, int nmi, header.size += sizeof(u64); } - if (sample_type & PERF_SAMPLE_TP_RECORD) { - tp = data->private; - if (tp) - header.size += tp->size; + if (sample_type & PERF_SAMPLE_RAW) { + raw = data->raw; + if (raw) + header.size += raw->size; } ret = perf_output_begin(&handle, counter, header.size, nmi, 1); @@ -2784,8 +2784,8 @@ static void perf_counter_output(struct perf_counter *counter, int nmi, } } - if ((sample_type & PERF_SAMPLE_TP_RECORD) && tp) - perf_output_copy(&handle, tp->record, tp->size); + if ((sample_type & PERF_SAMPLE_RAW) && raw) + perf_output_copy(&handle, raw->data, raw->size); perf_output_end(&handle); } @@ -3740,15 +3740,15 @@ static const struct pmu perf_ops_task_clock = { void perf_tpcounter_event(int event_id, u64 addr, u64 count, void *record, int entry_size) { - struct perf_tracepoint_record tp = { + struct perf_raw_record raw = { .size = entry_size, - .record = record, + .data = record, }; struct perf_sample_data data = { .regs = get_irq_regs(), .addr = addr, - .private = &tp, + .raw = &raw, }; if (!data.regs) -- cgit v1.2.3 From 8e5b59a2d728e6963b35dba8bb36e0b70267462e Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Thu, 6 Aug 2009 16:02:50 -0700 Subject: sched: Add default defines for PREEMPT_ACTIVE The PREEMPT_ACTIVE setting doesn't actually need to be arch-specific, so set up a sane default for all arches to (hopefully) migrate to. > if we look at linux/hardirq.h, it makes this claim: > * - bit 28 is the PREEMPT_ACTIVE flag > if that's true, then why are we letting any arch set this define ? a > quick survey shows that half the arches (11) are using 0x10000000 (bit > 28) while the other half (10) are using 0x4000000 (bit 26). and then > there is the ia64 oddity which uses bit 30. the exact value here > shouldnt really matter across arches though should it ? actually alpha, arm and avr32 also use bit 30 (0x40000000), there are only five (or eight, depending on how you count) architectures (blackfin, h8300, m68k, s390 and sparc) using bit 26. Signed-off-by: Arnd Bergmann Signed-off-by: Mike Frysinger Signed-off-by: Andrew Morton Signed-off-by: Ingo Molnar --- include/linux/hardirq.h | 6 ++++++ 1 file changed, 6 insertions(+) (limited to 'include') diff --git a/include/linux/hardirq.h b/include/linux/hardirq.h index 8246c697863d..0d885fd75111 100644 --- a/include/linux/hardirq.h +++ b/include/linux/hardirq.h @@ -64,6 +64,12 @@ #define HARDIRQ_OFFSET (1UL << HARDIRQ_SHIFT) #define NMI_OFFSET (1UL << NMI_SHIFT) +#ifndef PREEMPT_ACTIVE +#define PREEMPT_ACTIVE_BITS 1 +#define PREEMPT_ACTIVE_SHIFT (NMI_SHIFT + NMI_BITS) +#define PREEMPT_ACTIVE (__IRQ_MASK(PREEMPT_ACTIVE_BITS) << PREEMPT_ACTIVE_SHIFT) +#endif + #if PREEMPT_ACTIVE < (1 << (NMI_SHIFT + NMI_BITS)) #error PREEMPT_ACTIVE is too low! #endif -- cgit v1.2.3 From 62ab460cf5e450e1d207a98a9c6cf2e4a6a78fd1 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Sun, 9 Aug 2009 15:06:19 -0400 Subject: NFSv4: Add 'server capability' flags for NFSv4 recommended attributes If the NFSv4 server doesn't support a POSIX attribute, the generic NFS code needs to know that, so that it don't keep trying to poll for it. However, by the same count, if the NFSv4 server does support that attribute, then we should ensure that the inode metadata is appropriately labelled as being untrusted. For instance, if we don't know the correct value of the file's uid, we should certainly not be caching ACLs or ACCESS results. Signed-off-by: Trond Myklebust --- fs/nfs/client.c | 7 ++-- fs/nfs/inode.c | 92 +++++++++++++++++++++++++++++++++++++++-------- fs/nfs/nfs4proc.c | 22 ++++++++++++ include/linux/nfs_fs_sb.h | 9 +++++ 4 files changed, 114 insertions(+), 16 deletions(-) (limited to 'include') diff --git a/fs/nfs/client.c b/fs/nfs/client.c index 8d25ccb2d51d..8f34fd8028fc 100644 --- a/fs/nfs/client.c +++ b/fs/nfs/client.c @@ -809,6 +809,9 @@ static int nfs_init_server(struct nfs_server *server, /* Initialise the client representation from the mount data */ server->flags = data->flags; server->options = data->options; + server->caps |= NFS_CAP_HARDLINKS|NFS_CAP_SYMLINKS|NFS_CAP_FILEID| + NFS_CAP_MODE|NFS_CAP_NLINK|NFS_CAP_OWNER|NFS_CAP_OWNER_GROUP| + NFS_CAP_ATIME|NFS_CAP_CTIME|NFS_CAP_MTIME; if (data->rsize) server->rsize = nfs_block_size(data->rsize, NULL); @@ -1274,7 +1277,7 @@ static int nfs4_init_server(struct nfs_server *server, /* Initialise the client representation from the mount data */ server->flags = data->flags; - server->caps |= NFS_CAP_ATOMIC_OPEN; + server->caps |= NFS_CAP_ATOMIC_OPEN|NFS_CAP_CHANGE_ATTR; server->options = data->options; /* Get a client record */ @@ -1400,7 +1403,7 @@ struct nfs_server *nfs4_create_referral_server(struct nfs_clone_mount *data, /* Initialise the client representation from the parent server */ nfs_server_copy_userdata(server, parent_server); - server->caps |= NFS_CAP_ATOMIC_OPEN; + server->caps |= NFS_CAP_ATOMIC_OPEN|NFS_CAP_CHANGE_ATTR; /* Get a client representation. * Note: NFSv4 always uses TCP, */ diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c index bd7938eda6a8..fe5a8b45d867 100644 --- a/fs/nfs/inode.c +++ b/fs/nfs/inode.c @@ -286,6 +286,11 @@ nfs_fhget(struct super_block *sb, struct nfs_fh *fh, struct nfs_fattr *fattr) /* We can't support update_atime(), since the server will reset it */ inode->i_flags |= S_NOATIME|S_NOCMTIME; inode->i_mode = fattr->mode; + if ((fattr->valid & NFS_ATTR_FATTR_MODE) == 0 + && nfs_server_capable(inode, NFS_CAP_MODE)) + nfsi->cache_validity |= NFS_INO_INVALID_ATTR + | NFS_INO_INVALID_ACCESS + | NFS_INO_INVALID_ACL; /* Why so? Because we want revalidate for devices/FIFOs, and * that's precisely what we have in nfs_file_inode_operations. */ @@ -330,20 +335,46 @@ nfs_fhget(struct super_block *sb, struct nfs_fh *fh, struct nfs_fattr *fattr) nfsi->attr_gencount = fattr->gencount; if (fattr->valid & NFS_ATTR_FATTR_ATIME) inode->i_atime = fattr->atime; + else if (nfs_server_capable(inode, NFS_CAP_ATIME)) + nfsi->cache_validity |= NFS_INO_INVALID_ATTR; if (fattr->valid & NFS_ATTR_FATTR_MTIME) inode->i_mtime = fattr->mtime; + else if (nfs_server_capable(inode, NFS_CAP_MTIME)) + nfsi->cache_validity |= NFS_INO_INVALID_ATTR + | NFS_INO_INVALID_DATA; if (fattr->valid & NFS_ATTR_FATTR_CTIME) inode->i_ctime = fattr->ctime; + else if (nfs_server_capable(inode, NFS_CAP_CTIME)) + nfsi->cache_validity |= NFS_INO_INVALID_ATTR + | NFS_INO_INVALID_ACCESS + | NFS_INO_INVALID_ACL; if (fattr->valid & NFS_ATTR_FATTR_CHANGE) nfsi->change_attr = fattr->change_attr; + else if (nfs_server_capable(inode, NFS_CAP_CHANGE_ATTR)) + nfsi->cache_validity |= NFS_INO_INVALID_ATTR + | NFS_INO_INVALID_DATA; if (fattr->valid & NFS_ATTR_FATTR_SIZE) inode->i_size = nfs_size_to_loff_t(fattr->size); + else + nfsi->cache_validity |= NFS_INO_INVALID_ATTR + | NFS_INO_INVALID_DATA + | NFS_INO_REVAL_PAGECACHE; if (fattr->valid & NFS_ATTR_FATTR_NLINK) inode->i_nlink = fattr->nlink; + else if (nfs_server_capable(inode, NFS_CAP_NLINK)) + nfsi->cache_validity |= NFS_INO_INVALID_ATTR; if (fattr->valid & NFS_ATTR_FATTR_OWNER) inode->i_uid = fattr->uid; + else if (nfs_server_capable(inode, NFS_CAP_OWNER)) + nfsi->cache_validity |= NFS_INO_INVALID_ATTR + | NFS_INO_INVALID_ACCESS + | NFS_INO_INVALID_ACL; if (fattr->valid & NFS_ATTR_FATTR_GROUP) inode->i_gid = fattr->gid; + else if (nfs_server_capable(inode, NFS_CAP_OWNER_GROUP)) + nfsi->cache_validity |= NFS_INO_INVALID_ATTR + | NFS_INO_INVALID_ACCESS + | NFS_INO_INVALID_ACL; if (fattr->valid & NFS_ATTR_FATTR_BLOCKS_USED) inode->i_blocks = fattr->du.nfs2.blocks; if (fattr->valid & NFS_ATTR_FATTR_SPACE_USED) { @@ -1145,6 +1176,7 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr) loff_t cur_isize, new_isize; unsigned long invalid = 0; unsigned long now = jiffies; + unsigned long save_cache_validity; dfprintk(VFS, "NFS: %s(%s/%ld ct=%d info=0x%x)\n", __func__, inode->i_sb->s_id, inode->i_ino, @@ -1171,10 +1203,11 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr) */ nfsi->read_cache_jiffies = fattr->time_start; - if ((fattr->valid & NFS_ATTR_FATTR_CHANGE) || (fattr->valid & (NFS_ATTR_FATTR_MTIME|NFS_ATTR_FATTR_CTIME))) - nfsi->cache_validity &= ~(NFS_INO_INVALID_ATTR - | NFS_INO_INVALID_ATIME - | NFS_INO_REVAL_PAGECACHE); + save_cache_validity = nfsi->cache_validity; + nfsi->cache_validity &= ~(NFS_INO_INVALID_ATTR + | NFS_INO_INVALID_ATIME + | NFS_INO_REVAL_FORCED + | NFS_INO_REVAL_PAGECACHE); /* Do atomic weak cache consistency updates */ nfs_wcc_update_inode(inode, fattr); @@ -1189,7 +1222,8 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr) nfs_force_lookup_revalidate(inode); nfsi->change_attr = fattr->change_attr; } - } + } else if (server->caps & NFS_CAP_CHANGE_ATTR) + invalid |= save_cache_validity; if (fattr->valid & NFS_ATTR_FATTR_MTIME) { /* NFSv2/v3: Check if the mtime agrees */ @@ -1201,7 +1235,12 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr) nfs_force_lookup_revalidate(inode); memcpy(&inode->i_mtime, &fattr->mtime, sizeof(inode->i_mtime)); } - } + } else if (server->caps & NFS_CAP_MTIME) + invalid |= save_cache_validity & (NFS_INO_INVALID_ATTR + | NFS_INO_INVALID_DATA + | NFS_INO_REVAL_PAGECACHE + | NFS_INO_REVAL_FORCED); + if (fattr->valid & NFS_ATTR_FATTR_CTIME) { /* If ctime has changed we should definitely clear access+acl caches */ if (!timespec_equal(&inode->i_ctime, &fattr->ctime)) { @@ -1215,7 +1254,11 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr) } memcpy(&inode->i_ctime, &fattr->ctime, sizeof(inode->i_ctime)); } - } + } else if (server->caps & NFS_CAP_CTIME) + invalid |= save_cache_validity & (NFS_INO_INVALID_ATTR + | NFS_INO_INVALID_ACCESS + | NFS_INO_INVALID_ACL + | NFS_INO_REVAL_FORCED); /* Check if our cached file size is stale */ if (fattr->valid & NFS_ATTR_FATTR_SIZE) { @@ -1231,30 +1274,50 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr) dprintk("NFS: isize change on server for file %s/%ld\n", inode->i_sb->s_id, inode->i_ino); } - } + } else + invalid |= save_cache_validity & (NFS_INO_INVALID_ATTR + | NFS_INO_REVAL_PAGECACHE + | NFS_INO_REVAL_FORCED); if (fattr->valid & NFS_ATTR_FATTR_ATIME) memcpy(&inode->i_atime, &fattr->atime, sizeof(inode->i_atime)); + else if (server->caps & NFS_CAP_ATIME) + invalid |= save_cache_validity & (NFS_INO_INVALID_ATIME + | NFS_INO_REVAL_FORCED); if (fattr->valid & NFS_ATTR_FATTR_MODE) { if ((inode->i_mode & S_IALLUGO) != (fattr->mode & S_IALLUGO)) { invalid |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ACL; inode->i_mode = fattr->mode; } - } + } else if (server->caps & NFS_CAP_MODE) + invalid |= save_cache_validity & (NFS_INO_INVALID_ATTR + | NFS_INO_INVALID_ACCESS + | NFS_INO_INVALID_ACL + | NFS_INO_REVAL_FORCED); + if (fattr->valid & NFS_ATTR_FATTR_OWNER) { if (inode->i_uid != fattr->uid) { invalid |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ACL; inode->i_uid = fattr->uid; } - } + } else if (server->caps & NFS_CAP_OWNER) + invalid |= save_cache_validity & (NFS_INO_INVALID_ATTR + | NFS_INO_INVALID_ACCESS + | NFS_INO_INVALID_ACL + | NFS_INO_REVAL_FORCED); + if (fattr->valid & NFS_ATTR_FATTR_GROUP) { if (inode->i_gid != fattr->gid) { invalid |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ACL; inode->i_gid = fattr->gid; } - } + } else if (server->caps & NFS_CAP_OWNER_GROUP) + invalid |= save_cache_validity & (NFS_INO_INVALID_ATTR + | NFS_INO_INVALID_ACCESS + | NFS_INO_INVALID_ACL + | NFS_INO_REVAL_FORCED); if (fattr->valid & NFS_ATTR_FATTR_NLINK) { if (inode->i_nlink != fattr->nlink) { @@ -1263,7 +1326,9 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr) invalid |= NFS_INO_INVALID_DATA; inode->i_nlink = fattr->nlink; } - } + } else if (server->caps & NFS_CAP_NLINK) + invalid |= save_cache_validity & (NFS_INO_INVALID_ATTR + | NFS_INO_REVAL_FORCED); if (fattr->valid & NFS_ATTR_FATTR_SPACE_USED) { /* @@ -1293,9 +1358,8 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr) || S_ISLNK(inode->i_mode))) invalid &= ~NFS_INO_INVALID_DATA; if (!nfs_have_delegation(inode, FMODE_READ) || - (nfsi->cache_validity & NFS_INO_REVAL_FORCED)) + (save_cache_validity & NFS_INO_REVAL_FORCED)) nfsi->cache_validity |= invalid; - nfsi->cache_validity &= ~NFS_INO_REVAL_FORCED; return 0; out_changed: diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index d95f7f9e60c4..be6544aef41f 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -2003,12 +2003,34 @@ static int _nfs4_server_capabilities(struct nfs_server *server, struct nfs_fh *f status = nfs4_call_sync(server, &msg, &args, &res, 0); if (status == 0) { memcpy(server->attr_bitmask, res.attr_bitmask, sizeof(server->attr_bitmask)); + server->caps &= ~(NFS_CAP_ACLS|NFS_CAP_HARDLINKS| + NFS_CAP_SYMLINKS|NFS_CAP_FILEID| + NFS_CAP_MODE|NFS_CAP_NLINK|NFS_CAP_OWNER| + NFS_CAP_OWNER_GROUP|NFS_CAP_ATIME| + NFS_CAP_CTIME|NFS_CAP_MTIME); if (res.attr_bitmask[0] & FATTR4_WORD0_ACL) server->caps |= NFS_CAP_ACLS; if (res.has_links != 0) server->caps |= NFS_CAP_HARDLINKS; if (res.has_symlinks != 0) server->caps |= NFS_CAP_SYMLINKS; + if (res.attr_bitmask[0] & FATTR4_WORD0_FILEID) + server->caps |= NFS_CAP_FILEID; + if (res.attr_bitmask[1] & FATTR4_WORD1_MODE) + server->caps |= NFS_CAP_MODE; + if (res.attr_bitmask[1] & FATTR4_WORD1_NUMLINKS) + server->caps |= NFS_CAP_NLINK; + if (res.attr_bitmask[1] & FATTR4_WORD1_OWNER) + server->caps |= NFS_CAP_OWNER; + if (res.attr_bitmask[1] & FATTR4_WORD1_OWNER_GROUP) + server->caps |= NFS_CAP_OWNER_GROUP; + if (res.attr_bitmask[1] & FATTR4_WORD1_TIME_ACCESS) + server->caps |= NFS_CAP_ATIME; + if (res.attr_bitmask[1] & FATTR4_WORD1_TIME_METADATA) + server->caps |= NFS_CAP_CTIME; + if (res.attr_bitmask[1] & FATTR4_WORD1_TIME_MODIFY) + server->caps |= NFS_CAP_MTIME; + memcpy(server->cache_consistency_bitmask, res.attr_bitmask, sizeof(server->cache_consistency_bitmask)); server->cache_consistency_bitmask[0] &= FATTR4_WORD0_CHANGE|FATTR4_WORD0_SIZE; server->cache_consistency_bitmask[1] &= FATTR4_WORD1_TIME_METADATA|FATTR4_WORD1_TIME_MODIFY; diff --git a/include/linux/nfs_fs_sb.h b/include/linux/nfs_fs_sb.h index 19fe15d12042..320569eabe3b 100644 --- a/include/linux/nfs_fs_sb.h +++ b/include/linux/nfs_fs_sb.h @@ -167,6 +167,15 @@ struct nfs_server { #define NFS_CAP_SYMLINKS (1U << 2) #define NFS_CAP_ACLS (1U << 3) #define NFS_CAP_ATOMIC_OPEN (1U << 4) +#define NFS_CAP_CHANGE_ATTR (1U << 5) +#define NFS_CAP_FILEID (1U << 6) +#define NFS_CAP_MODE (1U << 7) +#define NFS_CAP_NLINK (1U << 8) +#define NFS_CAP_OWNER (1U << 9) +#define NFS_CAP_OWNER_GROUP (1U << 10) +#define NFS_CAP_ATIME (1U << 11) +#define NFS_CAP_CTIME (1U << 12) +#define NFS_CAP_MTIME (1U << 13) /* maximum number of slots to use */ -- cgit v1.2.3 From 169026a61e6f436dfc12c9d10d95455c4e9f945b Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Sun, 9 Aug 2009 15:09:32 -0400 Subject: SUNRPC: Clean up RPCBIND_MAXUADDRLEN definitions Clean up: Replace the single-integer definition of RPCBIND_MAXUADDRLEN with a definition that is based on previously defined address string sizes, and document the way this maximum is calculated. Also provide a separate macro for the size of the port number extension. Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- include/linux/sunrpc/msg_prot.h | 17 ++++++++++++++++- 1 file changed, 16 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/sunrpc/msg_prot.h b/include/linux/sunrpc/msg_prot.h index 70df4f1d8847..77e624883393 100644 --- a/include/linux/sunrpc/msg_prot.h +++ b/include/linux/sunrpc/msg_prot.h @@ -189,7 +189,22 @@ typedef __be32 rpc_fraghdr; * Additionally, the two alternative forms specified in Section 2.2 of * [RFC2373] are also acceptable. */ -#define RPCBIND_MAXUADDRLEN (56u) + +#include + +/* Maximum size of the port number part of a universal address */ +#define RPCBIND_MAXUADDRPLEN sizeof(".255.255") + +/* Maximum size of an IPv4 universal address */ +#define RPCBIND_MAXUADDR4LEN \ + (INET_ADDRSTRLEN + RPCBIND_MAXUADDRPLEN) + +/* Maximum size of an IPv6 universal address */ +#define RPCBIND_MAXUADDR6LEN \ + (INET6_ADDRSTRLEN + RPCBIND_MAXUADDRPLEN) + +/* Assume INET6_ADDRSTRLEN will always be larger than INET_ADDRSTRLEN... */ +#define RPCBIND_MAXUADDRLEN RPCBIND_MAXUADDR6LEN #endif /* __KERNEL__ */ #endif /* _LINUX_SUNRPC_MSGPROT_H_ */ -- cgit v1.2.3 From a02d692611348f11ee1bc37431a883c3ff2de23e Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Sun, 9 Aug 2009 15:09:34 -0400 Subject: SUNRPC: Provide functions for managing universal addresses Introduce a set of functions in the kernel's RPC implementation for converting between a socket address and either a standard presentation address string or an RPC universal address. The universal address functions will be used to encode and decode RPCB_FOO and NFSv4 SETCLIENTID arguments. The other functions are part of a previous promise to deliver shared functions that can be used by upper-layer protocols to display and manipulate IP addresses. The kernel's current address printf formatters were designed specifically for kernel to user-space APIs that require a particular string format for socket addresses, thus are somewhat limited for the purposes of sunrpc.ko. The formatter for IPv6 addresses, %pI6, does not support short-handing or scope IDs. Also, these printf formatters are unique per address family, so a separate formatter string is required for printing AF_INET and AF_INET6 addresses. Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- fs/nfs/internal.h | 2 - include/linux/sunrpc/clnt.h | 38 +++++ net/sunrpc/Makefile | 2 +- net/sunrpc/addr.c | 364 ++++++++++++++++++++++++++++++++++++++++++++ 4 files changed, 403 insertions(+), 3 deletions(-) create mode 100644 net/sunrpc/addr.c (limited to 'include') diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h index 8d2b71d57d29..ff68397f9b19 100644 --- a/fs/nfs/internal.h +++ b/fs/nfs/internal.h @@ -370,8 +370,6 @@ unsigned int nfs_page_array_len(unsigned int base, size_t len) PAGE_SIZE - 1) >> PAGE_SHIFT; } -#define IPV6_SCOPE_DELIMITER '%' - /* * Set the port number in an address. Be agnostic about the address * family. diff --git a/include/linux/sunrpc/clnt.h b/include/linux/sunrpc/clnt.h index 37881f1a0bd7..2636e8ad551c 100644 --- a/include/linux/sunrpc/clnt.h +++ b/include/linux/sunrpc/clnt.h @@ -9,6 +9,10 @@ #ifndef _LINUX_SUNRPC_CLNT_H #define _LINUX_SUNRPC_CLNT_H +#include +#include +#include + #include #include #include @@ -151,5 +155,39 @@ void rpc_force_rebind(struct rpc_clnt *); size_t rpc_peeraddr(struct rpc_clnt *, struct sockaddr *, size_t); const char *rpc_peeraddr2str(struct rpc_clnt *, enum rpc_display_format_t); +size_t rpc_ntop(const struct sockaddr *, char *, const size_t); +size_t rpc_pton(const char *, const size_t, + struct sockaddr *, const size_t); +char * rpc_sockaddr2uaddr(const struct sockaddr *); +size_t rpc_uaddr2sockaddr(const char *, const size_t, + struct sockaddr *, const size_t); + +static inline unsigned short rpc_get_port(const struct sockaddr *sap) +{ + switch (sap->sa_family) { + case AF_INET: + return ntohs(((struct sockaddr_in *)sap)->sin_port); + case AF_INET6: + return ntohs(((struct sockaddr_in6 *)sap)->sin6_port); + } + return 0; +} + +static inline void rpc_set_port(struct sockaddr *sap, + const unsigned short port) +{ + switch (sap->sa_family) { + case AF_INET: + ((struct sockaddr_in *)sap)->sin_port = htons(port); + break; + case AF_INET6: + ((struct sockaddr_in6 *)sap)->sin6_port = htons(port); + break; + } +} + +#define IPV6_SCOPE_DELIMITER '%' +#define IPV6_SCOPE_ID_LEN sizeof("%nnnnnnnnnn") + #endif /* __KERNEL__ */ #endif /* _LINUX_SUNRPC_CLNT_H */ diff --git a/net/sunrpc/Makefile b/net/sunrpc/Makefile index db73fd2a3f0e..9d2fca5ad14a 100644 --- a/net/sunrpc/Makefile +++ b/net/sunrpc/Makefile @@ -10,7 +10,7 @@ obj-$(CONFIG_SUNRPC_XPRT_RDMA) += xprtrdma/ sunrpc-y := clnt.o xprt.o socklib.o xprtsock.o sched.o \ auth.o auth_null.o auth_unix.o auth_generic.o \ svc.o svcsock.o svcauth.o svcauth_unix.o \ - rpcb_clnt.o timer.o xdr.o \ + addr.o rpcb_clnt.o timer.o xdr.o \ sunrpc_syms.o cache.o rpc_pipe.o \ svc_xprt.o sunrpc-$(CONFIG_NFS_V4_1) += backchannel_rqst.o bc_svc.o diff --git a/net/sunrpc/addr.c b/net/sunrpc/addr.c new file mode 100644 index 000000000000..22e8fd89477f --- /dev/null +++ b/net/sunrpc/addr.c @@ -0,0 +1,364 @@ +/* + * Copyright 2009, Oracle. All rights reserved. + * + * Convert socket addresses to presentation addresses and universal + * addresses, and vice versa. + * + * Universal addresses are introduced by RFC 1833 and further refined by + * recent RFCs describing NFSv4. The universal address format is part + * of the external (network) interface provided by rpcbind version 3 + * and 4, and by NFSv4. Such an address is a string containing a + * presentation format IP address followed by a port number in + * "hibyte.lobyte" format. + * + * IPv6 addresses can also include a scope ID, typically denoted by + * a '%' followed by a device name or a non-negative integer. Refer to + * RFC 4291, Section 2.2 for details on IPv6 presentation formats. + */ + +#include +#include + +#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE) + +static size_t rpc_ntop6_noscopeid(const struct sockaddr *sap, + char *buf, const int buflen) +{ + const struct sockaddr_in6 *sin6 = (struct sockaddr_in6 *)sap; + const struct in6_addr *addr = &sin6->sin6_addr; + + /* + * RFC 4291, Section 2.2.2 + * + * Shorthanded ANY address + */ + if (ipv6_addr_any(addr)) + return snprintf(buf, buflen, "::"); + + /* + * RFC 4291, Section 2.2.2 + * + * Shorthanded loopback address + */ + if (ipv6_addr_loopback(addr)) + return snprintf(buf, buflen, "::1"); + + /* + * RFC 4291, Section 2.2.3 + * + * Special presentation address format for mapped v4 + * addresses. + */ + if (ipv6_addr_v4mapped(addr)) + return snprintf(buf, buflen, "::ffff:%pI4", + &addr->s6_addr32[3]); + + /* + * RFC 4291, Section 2.2.1 + * + * To keep the result as short as possible, especially + * since we don't shorthand, we don't want leading zeros + * in each halfword, so avoid %pI6. + */ + return snprintf(buf, buflen, "%x:%x:%x:%x:%x:%x:%x:%x", + ntohs(addr->s6_addr16[0]), ntohs(addr->s6_addr16[1]), + ntohs(addr->s6_addr16[2]), ntohs(addr->s6_addr16[3]), + ntohs(addr->s6_addr16[4]), ntohs(addr->s6_addr16[5]), + ntohs(addr->s6_addr16[6]), ntohs(addr->s6_addr16[7])); +} + +static size_t rpc_ntop6(const struct sockaddr *sap, + char *buf, const size_t buflen) +{ + const struct sockaddr_in6 *sin6 = (struct sockaddr_in6 *)sap; + char scopebuf[IPV6_SCOPE_ID_LEN]; + size_t len; + int rc; + + len = rpc_ntop6_noscopeid(sap, buf, buflen); + if (unlikely(len == 0)) + return len; + + if (!(ipv6_addr_type(&sin6->sin6_addr) & IPV6_ADDR_LINKLOCAL) && + !(ipv6_addr_type(&sin6->sin6_addr) & IPV6_ADDR_SITELOCAL)) + return len; + + rc = snprintf(scopebuf, sizeof(scopebuf), "%c%u", + IPV6_SCOPE_DELIMITER, sin6->sin6_scope_id); + if (unlikely((size_t)rc > sizeof(scopebuf))) + return 0; + + len += rc; + if (unlikely(len > buflen)) + return 0; + + strcat(buf, scopebuf); + return len; +} + +#else /* !(defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE)) */ + +static size_t rpc_ntop6_noscopeid(const struct sockaddr *sap, + char *buf, const int buflen) +{ + return 0; +} + +static size_t rpc_ntop6(const struct sockaddr *sap, + char *buf, const size_t buflen) +{ + return 0; +} + +#endif /* !(defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE)) */ + +static int rpc_ntop4(const struct sockaddr *sap, + char *buf, const size_t buflen) +{ + const struct sockaddr_in *sin = (struct sockaddr_in *)sap; + + return snprintf(buf, buflen, "%pI4", &sin->sin_addr); +} + +/** + * rpc_ntop - construct a presentation address in @buf + * @sap: socket address + * @buf: construction area + * @buflen: size of @buf, in bytes + * + * Plants a %NUL-terminated string in @buf and returns the length + * of the string, excluding the %NUL. Otherwise zero is returned. + */ +size_t rpc_ntop(const struct sockaddr *sap, char *buf, const size_t buflen) +{ + switch (sap->sa_family) { + case AF_INET: + return rpc_ntop4(sap, buf, buflen); + case AF_INET6: + return rpc_ntop6(sap, buf, buflen); + } + + return 0; +} +EXPORT_SYMBOL_GPL(rpc_ntop); + +static size_t rpc_pton4(const char *buf, const size_t buflen, + struct sockaddr *sap, const size_t salen) +{ + struct sockaddr_in *sin = (struct sockaddr_in *)sap; + u8 *addr = (u8 *)&sin->sin_addr.s_addr; + + if (buflen > INET_ADDRSTRLEN || salen < sizeof(struct sockaddr_in)) + return 0; + + memset(sap, 0, sizeof(struct sockaddr_in)); + + if (in4_pton(buf, buflen, addr, '\0', NULL) == 0) + return 0; + + sin->sin_family = AF_INET; + return sizeof(struct sockaddr_in);; +} + +#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE) +static int rpc_parse_scope_id(const char *buf, const size_t buflen, + const char *delim, struct sockaddr_in6 *sin6) +{ + char *p; + size_t len; + + if ((buf + buflen) == delim) + return 1; + + if (*delim != IPV6_SCOPE_DELIMITER) + return 0; + + if (!(ipv6_addr_type(&sin6->sin6_addr) & IPV6_ADDR_LINKLOCAL) && + !(ipv6_addr_type(&sin6->sin6_addr) & IPV6_ADDR_SITELOCAL)) + return 0; + + len = (buf + buflen) - delim - 1; + p = kstrndup(delim + 1, len, GFP_KERNEL); + if (p) { + unsigned long scope_id = 0; + struct net_device *dev; + + dev = dev_get_by_name(&init_net, p); + if (dev != NULL) { + scope_id = dev->ifindex; + dev_put(dev); + } else { + if (strict_strtoul(p, 10, &scope_id) == 0) { + kfree(p); + return 0; + } + } + + kfree(p); + + sin6->sin6_scope_id = scope_id; + return 1; + } + + return 0; +} + +static size_t rpc_pton6(const char *buf, const size_t buflen, + struct sockaddr *sap, const size_t salen) +{ + struct sockaddr_in6 *sin6 = (struct sockaddr_in6 *)sap; + u8 *addr = (u8 *)&sin6->sin6_addr.in6_u; + const char *delim; + + if (buflen > (INET6_ADDRSTRLEN + IPV6_SCOPE_ID_LEN) || + salen < sizeof(struct sockaddr_in6)) + return 0; + + memset(sap, 0, sizeof(struct sockaddr_in6)); + + if (in6_pton(buf, buflen, addr, IPV6_SCOPE_DELIMITER, &delim) == 0) + return 0; + + if (!rpc_parse_scope_id(buf, buflen, delim, sin6)) + return 0; + + sin6->sin6_family = AF_INET6; + return sizeof(struct sockaddr_in6); +} +#else +static size_t rpc_pton6(const char *buf, const size_t buflen, + struct sockaddr *sap, const size_t salen) +{ + return 0; +} +#endif + +/** + * rpc_pton - Construct a sockaddr in @sap + * @buf: C string containing presentation format IP address + * @buflen: length of presentation address in bytes + * @sap: buffer into which to plant socket address + * @salen: size of buffer in bytes + * + * Returns the size of the socket address if successful; otherwise + * zero is returned. + * + * Plants a socket address in @sap and returns the size of the + * socket address, if successful. Returns zero if an error + * occurred. + */ +size_t rpc_pton(const char *buf, const size_t buflen, + struct sockaddr *sap, const size_t salen) +{ + unsigned int i; + + for (i = 0; i < buflen; i++) + if (buf[i] == ':') + return rpc_pton6(buf, buflen, sap, salen); + return rpc_pton4(buf, buflen, sap, salen); +} +EXPORT_SYMBOL_GPL(rpc_pton); + +/** + * rpc_sockaddr2uaddr - Construct a universal address string from @sap. + * @sap: socket address + * + * Returns a %NUL-terminated string in dynamically allocated memory; + * otherwise NULL is returned if an error occurred. Caller must + * free the returned string. + */ +char *rpc_sockaddr2uaddr(const struct sockaddr *sap) +{ + char portbuf[RPCBIND_MAXUADDRPLEN]; + char addrbuf[RPCBIND_MAXUADDRLEN]; + unsigned short port; + + switch (sap->sa_family) { + case AF_INET: + if (rpc_ntop4(sap, addrbuf, sizeof(addrbuf)) == 0) + return NULL; + port = ntohs(((struct sockaddr_in *)sap)->sin_port); + break; + case AF_INET6: + if (rpc_ntop6_noscopeid(sap, addrbuf, sizeof(addrbuf)) == 0) + return NULL; + port = ntohs(((struct sockaddr_in6 *)sap)->sin6_port); + break; + default: + return NULL; + } + + if (snprintf(portbuf, sizeof(portbuf), + ".%u.%u", port >> 8, port & 0xff) > (int)sizeof(portbuf)) + return NULL; + + if (strlcat(addrbuf, portbuf, sizeof(addrbuf)) > sizeof(addrbuf)) + return NULL; + + return kstrdup(addrbuf, GFP_KERNEL); +} +EXPORT_SYMBOL_GPL(rpc_sockaddr2uaddr); + +/** + * rpc_uaddr2sockaddr - convert a universal address to a socket address. + * @uaddr: C string containing universal address to convert + * @uaddr_len: length of universal address string + * @sap: buffer into which to plant socket address + * @salen: size of buffer + * + * Returns the size of the socket address if successful; otherwise + * zero is returned. + */ +size_t rpc_uaddr2sockaddr(const char *uaddr, const size_t uaddr_len, + struct sockaddr *sap, const size_t salen) +{ + char *c, buf[RPCBIND_MAXUADDRLEN]; + unsigned long portlo, porthi; + unsigned short port; + + if (uaddr_len > sizeof(buf)) + return 0; + + memcpy(buf, uaddr, uaddr_len); + + buf[uaddr_len] = '\n'; + buf[uaddr_len + 1] = '\0'; + + c = strrchr(buf, '.'); + if (unlikely(c == NULL)) + return 0; + if (unlikely(strict_strtoul(c + 1, 10, &portlo) != 0)) + return 0; + if (unlikely(portlo > 255)) + return 0; + + c[0] = '\n'; + c[1] = '\0'; + + c = strrchr(buf, '.'); + if (unlikely(c == NULL)) + return 0; + if (unlikely(strict_strtoul(c + 1, 10, &porthi) != 0)) + return 0; + if (unlikely(porthi > 255)) + return 0; + + port = (unsigned short)((porthi << 8) | portlo); + + c[0] = '\0'; + + if (rpc_pton(buf, strlen(buf), sap, salen) == 0) + return 0; + + switch (sap->sa_family) { + case AF_INET: + ((struct sockaddr_in *)sap)->sin_port = htons(port); + return sizeof(struct sockaddr_in); + case AF_INET6: + ((struct sockaddr_in6 *)sap)->sin6_port = htons(port); + return sizeof(struct sockaddr_in6); + } + + return 0; +} +EXPORT_SYMBOL_GPL(rpc_uaddr2sockaddr); -- cgit v1.2.3 From ba809130bc260fce04141aca01ef9e068d32af2a Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Sun, 9 Aug 2009 15:09:35 -0400 Subject: SUNRPC: Remove duplicate universal address generation RPC universal address generation is currently done in several places: rpcb_clnt.c, nfs4proc.c xprtsock.c, and xprtrdma.c. Remove the redundant cases that convert a socket address to a universal address. The nfs4proc.c case takes a pre-formatted presentation address string, not a socket address, so we'll leave that one. Because the new uaddr constructor uses the recently introduced rpc_ntop(), it now supports proper "::" shorthanding for IPv6 addresses. This allows the kernel to register properly formed universal addresses with the local rpcbind service, in _all_ cases. The kernel can now also send properly formed universal addresses in RPCB_GETADDR requests, and support link-local properly when encoding and decoding IPv6 addresses. Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- include/linux/sunrpc/xprt.h | 1 - net/sunrpc/rpcb_clnt.c | 48 +++++++++++++++++++++++++---------------- net/sunrpc/xprtrdma/transport.c | 8 ------- net/sunrpc/xprtsock.c | 18 ---------------- 4 files changed, 29 insertions(+), 46 deletions(-) (limited to 'include') diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h index 1175d58efc2e..65fad9534d93 100644 --- a/include/linux/sunrpc/xprt.h +++ b/include/linux/sunrpc/xprt.h @@ -41,7 +41,6 @@ enum rpc_display_format_t { RPC_DISPLAY_ALL, RPC_DISPLAY_HEX_ADDR, RPC_DISPLAY_HEX_PORT, - RPC_DISPLAY_UNIVERSAL_ADDR, RPC_DISPLAY_NETID, RPC_DISPLAY_MAX, }; diff --git a/net/sunrpc/rpcb_clnt.c b/net/sunrpc/rpcb_clnt.c index ad1d7315c498..1fb1c070c19d 100644 --- a/net/sunrpc/rpcb_clnt.c +++ b/net/sunrpc/rpcb_clnt.c @@ -153,6 +153,7 @@ static void rpcb_map_release(void *data) rpcb_wake_rpcbind_waiters(map->r_xprt, map->r_status); xprt_put(map->r_xprt); + kfree(map->r_addr); kfree(map); } @@ -299,12 +300,9 @@ static int rpcb_register_inet4(const struct sockaddr *sap, const struct sockaddr_in *sin = (const struct sockaddr_in *)sap; struct rpcbind_args *map = msg->rpc_argp; unsigned short port = ntohs(sin->sin_port); - char buf[32]; + int result; - /* Construct AF_INET universal address */ - snprintf(buf, sizeof(buf), "%pI4.%u.%u", - &sin->sin_addr.s_addr, port >> 8, port & 0xff); - map->r_addr = buf; + map->r_addr = rpc_sockaddr2uaddr(sap); dprintk("RPC: %sregistering [%u, %u, %s, '%s'] with " "local rpcbind\n", (port ? "" : "un"), @@ -315,7 +313,9 @@ static int rpcb_register_inet4(const struct sockaddr *sap, if (port) msg->rpc_proc = &rpcb_procedures4[RPCBPROC_SET]; - return rpcb_register_call(RPCBVERS_4, msg); + result = rpcb_register_call(RPCBVERS_4, msg); + kfree(map->r_addr); + return result; } /* @@ -327,16 +327,9 @@ static int rpcb_register_inet6(const struct sockaddr *sap, const struct sockaddr_in6 *sin6 = (const struct sockaddr_in6 *)sap; struct rpcbind_args *map = msg->rpc_argp; unsigned short port = ntohs(sin6->sin6_port); - char buf[64]; + int result; - /* Construct AF_INET6 universal address */ - if (ipv6_addr_any(&sin6->sin6_addr)) - snprintf(buf, sizeof(buf), "::.%u.%u", - port >> 8, port & 0xff); - else - snprintf(buf, sizeof(buf), "%pI6.%u.%u", - &sin6->sin6_addr, port >> 8, port & 0xff); - map->r_addr = buf; + map->r_addr = rpc_sockaddr2uaddr(sap); dprintk("RPC: %sregistering [%u, %u, %s, '%s'] with " "local rpcbind\n", (port ? "" : "un"), @@ -347,7 +340,9 @@ static int rpcb_register_inet6(const struct sockaddr *sap, if (port) msg->rpc_proc = &rpcb_procedures4[RPCBPROC_SET]; - return rpcb_register_call(RPCBVERS_4, msg); + result = rpcb_register_call(RPCBVERS_4, msg); + kfree(map->r_addr); + return result; } static int rpcb_unregister_all_protofamilies(struct rpc_message *msg) @@ -570,6 +565,7 @@ void rpcb_getport_async(struct rpc_task *task) goto bailout_nofree; } + /* Parent transport's destination address */ salen = rpc_peeraddr(clnt, sap, sizeof(addr)); /* Don't ever use rpcbind v2 for AF_INET6 requests */ @@ -620,11 +616,22 @@ void rpcb_getport_async(struct rpc_task *task) map->r_prot = xprt->prot; map->r_port = 0; map->r_xprt = xprt_get(xprt); - map->r_netid = rpc_peeraddr2str(clnt, RPC_DISPLAY_NETID); - map->r_addr = rpc_peeraddr2str(rpcb_clnt, RPC_DISPLAY_UNIVERSAL_ADDR); - map->r_owner = ""; map->r_status = -EIO; + switch (bind_version) { + case RPCBVERS_4: + case RPCBVERS_3: + map->r_netid = rpc_peeraddr2str(clnt, RPC_DISPLAY_NETID); + map->r_addr = rpc_sockaddr2uaddr(sap); + map->r_owner = ""; + break; + case RPCBVERS_2: + map->r_addr = NULL; + break; + default: + BUG(); + } + child = rpcb_call_async(rpcb_clnt, map, proc); rpc_release_client(rpcb_clnt); if (IS_ERR(child)) { @@ -722,6 +729,9 @@ static int rpcb_decode_set(struct rpc_rqst *req, __be32 *p, static int rpcb_encode_getaddr(struct rpc_rqst *req, __be32 *p, struct rpcbind_args *rpcb) { + if (rpcb->r_addr == NULL) + return -EIO; + dprintk("RPC: encoding rpcb request (%u, %u, %s)\n", rpcb->r_prog, rpcb->r_vers, rpcb->r_addr); *p++ = htonl(rpcb->r_prog); diff --git a/net/sunrpc/xprtrdma/transport.c b/net/sunrpc/xprtrdma/transport.c index 1dd6123070e9..537c210a8b92 100644 --- a/net/sunrpc/xprtrdma/transport.c +++ b/net/sunrpc/xprtrdma/transport.c @@ -202,14 +202,6 @@ xprt_rdma_format_addresses(struct rpc_xprt *xprt) snprintf(buf, 8, "%4hx", ntohs(addr->sin_port)); xprt->address_strings[RPC_DISPLAY_HEX_PORT] = buf; - buf = kzalloc(30, GFP_KERNEL); - if (buf) - snprintf(buf, 30, "%pI4.%u.%u", - &addr->sin_addr.s_addr, - ntohs(addr->sin_port) >> 8, - ntohs(addr->sin_port) & 0xff); - xprt->address_strings[RPC_DISPLAY_UNIVERSAL_ADDR] = buf; - /* netid */ xprt->address_strings[RPC_DISPLAY_NETID] = "rdma"; } diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index 83c73c4d017a..a42c2adda59f 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -341,15 +341,6 @@ static void xs_format_ipv4_peer_addresses(struct rpc_xprt *xprt, } xprt->address_strings[RPC_DISPLAY_HEX_PORT] = buf; - buf = kzalloc(30, GFP_KERNEL); - if (buf) { - snprintf(buf, 30, "%pI4.%u.%u", - &addr->sin_addr.s_addr, - ntohs(addr->sin_port) >> 8, - ntohs(addr->sin_port) & 0xff); - } - xprt->address_strings[RPC_DISPLAY_UNIVERSAL_ADDR] = buf; - xprt->address_strings[RPC_DISPLAY_NETID] = netid; } @@ -397,15 +388,6 @@ static void xs_format_ipv6_peer_addresses(struct rpc_xprt *xprt, } xprt->address_strings[RPC_DISPLAY_HEX_PORT] = buf; - buf = kzalloc(50, GFP_KERNEL); - if (buf) { - snprintf(buf, 50, "%pI6.%u.%u", - &addr->sin6_addr, - ntohs(addr->sin6_port) >> 8, - ntohs(addr->sin6_port) & 0xff); - } - xprt->address_strings[RPC_DISPLAY_UNIVERSAL_ADDR] = buf; - xprt->address_strings[RPC_DISPLAY_NETID] = netid; } -- cgit v1.2.3 From c740eff84bcfd63c0497ef880e80171931cb8222 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Sun, 9 Aug 2009 15:09:46 -0400 Subject: SUNRPC: Kill RPC_DISPLAY_ALL At some point, I recall that rpc_pipe_fs used RPC_DISPLAY_ALL. Currently there are no uses of RPC_DISPLAY_ALL outside the transport modules themselves, so we can safely get rid of it. Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- include/linux/sunrpc/xprt.h | 1 - net/sunrpc/xprtrdma/transport.c | 5 ----- net/sunrpc/xprtsock.c | 50 ++++++++++++++++++++++++++++------------- 3 files changed, 34 insertions(+), 22 deletions(-) (limited to 'include') diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h index 65fad9534d93..c090df442572 100644 --- a/include/linux/sunrpc/xprt.h +++ b/include/linux/sunrpc/xprt.h @@ -38,7 +38,6 @@ enum rpc_display_format_t { RPC_DISPLAY_ADDR = 0, RPC_DISPLAY_PORT, RPC_DISPLAY_PROTO, - RPC_DISPLAY_ALL, RPC_DISPLAY_HEX_ADDR, RPC_DISPLAY_HEX_PORT, RPC_DISPLAY_NETID, diff --git a/net/sunrpc/xprtrdma/transport.c b/net/sunrpc/xprtrdma/transport.c index 5f9b8676b6bd..9a63f669ece4 100644 --- a/net/sunrpc/xprtrdma/transport.c +++ b/net/sunrpc/xprtrdma/transport.c @@ -181,11 +181,6 @@ xprt_rdma_format_addresses(struct rpc_xprt *xprt) xprt->address_strings[RPC_DISPLAY_PROTO] = "rdma"; - (void)snprintf(buf, sizeof(buf), "addr=%s port=%s proto=rdma", - xprt->address_strings[RPC_DISPLAY_ADDR], - xprt->address_strings[RPC_DISPLAY_PORT]); - xprt->address_strings[RPC_DISPLAY_ALL] = kstrdup(buf, GFP_KERNEL); - (void)snprintf(buf, sizeof(buf), "%02x%02x%02x%02x", NIPQUAD(sin->sin_addr.s_addr)); xprt->address_strings[RPC_DISPLAY_HEX_ADDR] = kstrdup(buf, GFP_KERNEL); diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index 7bc3c178ccb3..eee5ac96e177 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -307,12 +307,6 @@ static void xs_format_common_peer_addresses(struct rpc_xprt *xprt) (void)snprintf(buf, sizeof(buf), "%u", rpc_get_port(sap)); xprt->address_strings[RPC_DISPLAY_PORT] = kstrdup(buf, GFP_KERNEL); - (void)snprintf(buf, sizeof(buf), "addr=%s port=%s proto=%s", - xprt->address_strings[RPC_DISPLAY_ADDR], - xprt->address_strings[RPC_DISPLAY_PORT], - xprt->address_strings[RPC_DISPLAY_PROTO]); - xprt->address_strings[RPC_DISPLAY_ALL] = kstrdup(buf, GFP_KERNEL); - (void)snprintf(buf, sizeof(buf), "%4hx", rpc_get_port(sap)); xprt->address_strings[RPC_DISPLAY_HEX_PORT] = kstrdup(buf, GFP_KERNEL); } @@ -1721,8 +1715,11 @@ static void xs_udp_connect_worker4(struct work_struct *work) goto out; } - dprintk("RPC: worker connecting xprt %p to address: %s\n", - xprt, xprt->address_strings[RPC_DISPLAY_ALL]); + dprintk("RPC: worker connecting xprt %p via %s to " + "%s (port %s)\n", xprt, + xprt->address_strings[RPC_DISPLAY_PROTO], + xprt->address_strings[RPC_DISPLAY_ADDR], + xprt->address_strings[RPC_DISPLAY_PORT]); xs_udp_finish_connecting(xprt, sock); status = 0; @@ -1763,8 +1760,11 @@ static void xs_udp_connect_worker6(struct work_struct *work) goto out; } - dprintk("RPC: worker connecting xprt %p to address: %s\n", - xprt, xprt->address_strings[RPC_DISPLAY_ALL]); + dprintk("RPC: worker connecting xprt %p via %s to " + "%s (port %s)\n", xprt, + xprt->address_strings[RPC_DISPLAY_PROTO], + xprt->address_strings[RPC_DISPLAY_ADDR], + xprt->address_strings[RPC_DISPLAY_PORT]); xs_udp_finish_connecting(xprt, sock); status = 0; @@ -1889,8 +1889,11 @@ static void xs_tcp_setup_socket(struct rpc_xprt *xprt, goto out_eagain; } - dprintk("RPC: worker connecting xprt %p to address: %s\n", - xprt, xprt->address_strings[RPC_DISPLAY_ALL]); + dprintk("RPC: worker connecting xprt %p via %s to " + "%s (port %s)\n", xprt, + xprt->address_strings[RPC_DISPLAY_PROTO], + xprt->address_strings[RPC_DISPLAY_ADDR], + xprt->address_strings[RPC_DISPLAY_PORT]); status = xs_tcp_finish_connecting(xprt, sock); dprintk("RPC: %p connect status %d connected %d sock state %d\n", @@ -2228,8 +2231,15 @@ static struct rpc_xprt *xs_setup_udp(struct xprt_create *args) return ERR_PTR(-EAFNOSUPPORT); } - dprintk("RPC: set up transport to address %s\n", - xprt->address_strings[RPC_DISPLAY_ALL]); + if (xprt_bound(xprt)) + dprintk("RPC: set up xprt to %s (port %s) via %s\n", + xprt->address_strings[RPC_DISPLAY_ADDR], + xprt->address_strings[RPC_DISPLAY_PORT], + xprt->address_strings[RPC_DISPLAY_PROTO]); + else + dprintk("RPC: set up xprt to %s (autobind) via %s\n", + xprt->address_strings[RPC_DISPLAY_ADDR], + xprt->address_strings[RPC_DISPLAY_PROTO]); if (try_module_get(THIS_MODULE)) return xprt; @@ -2293,8 +2303,16 @@ static struct rpc_xprt *xs_setup_tcp(struct xprt_create *args) return ERR_PTR(-EAFNOSUPPORT); } - dprintk("RPC: set up transport to address %s\n", - xprt->address_strings[RPC_DISPLAY_ALL]); + if (xprt_bound(xprt)) + dprintk("RPC: set up xprt to %s (port %s) via %s\n", + xprt->address_strings[RPC_DISPLAY_ADDR], + xprt->address_strings[RPC_DISPLAY_PORT], + xprt->address_strings[RPC_DISPLAY_PROTO]); + else + dprintk("RPC: set up xprt to %s (autobind) via %s\n", + xprt->address_strings[RPC_DISPLAY_ADDR], + xprt->address_strings[RPC_DISPLAY_PROTO]); + if (try_module_get(THIS_MODULE)) return xprt; -- cgit v1.2.3 From b693ba4a338da15db1db4b5ebaa36e4ab9781c82 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Sun, 9 Aug 2009 15:14:15 -0400 Subject: SUNRPC: Constify rpc_pipe_ops... Signed-off-by: Trond Myklebust --- fs/nfs/idmap.c | 2 +- include/linux/sunrpc/rpc_pipe_fs.h | 5 +++-- net/sunrpc/auth_gss/auth_gss.c | 8 ++++---- net/sunrpc/rpc_pipe.c | 7 ++++--- 4 files changed, 12 insertions(+), 10 deletions(-) (limited to 'include') diff --git a/fs/nfs/idmap.c b/fs/nfs/idmap.c index 86147b0ab2cf..fae0d3e52b44 100644 --- a/fs/nfs/idmap.c +++ b/fs/nfs/idmap.c @@ -101,7 +101,7 @@ static void idmap_pipe_destroy_msg(struct rpc_pipe_msg *); static unsigned int fnvhash32(const void *, size_t); -static struct rpc_pipe_ops idmap_upcall_ops = { +static const struct rpc_pipe_ops idmap_upcall_ops = { .upcall = idmap_pipe_upcall, .downcall = idmap_pipe_downcall, .destroy_msg = idmap_pipe_destroy_msg, diff --git a/include/linux/sunrpc/rpc_pipe_fs.h b/include/linux/sunrpc/rpc_pipe_fs.h index cea764c2359f..91f5b13389c5 100644 --- a/include/linux/sunrpc/rpc_pipe_fs.h +++ b/include/linux/sunrpc/rpc_pipe_fs.h @@ -32,8 +32,8 @@ struct rpc_inode { wait_queue_head_t waitq; #define RPC_PIPE_WAIT_FOR_OPEN 1 int flags; - struct rpc_pipe_ops *ops; struct delayed_work queue_timeout; + const struct rpc_pipe_ops *ops; }; static inline struct rpc_inode * @@ -46,7 +46,8 @@ extern int rpc_queue_upcall(struct inode *, struct rpc_pipe_msg *); extern struct dentry *rpc_mkdir(char *, struct rpc_clnt *); extern int rpc_rmdir(struct dentry *); -extern struct dentry *rpc_mkpipe(struct dentry *, const char *, void *, struct rpc_pipe_ops *, int flags); +extern struct dentry *rpc_mkpipe(struct dentry *, const char *, void *, + const struct rpc_pipe_ops *, int flags); extern int rpc_unlink(struct dentry *); extern struct vfsmount *rpc_get_mount(void); extern void rpc_put_mount(void); diff --git a/net/sunrpc/auth_gss/auth_gss.c b/net/sunrpc/auth_gss/auth_gss.c index 66d458fc6920..23eb3864ffc0 100644 --- a/net/sunrpc/auth_gss/auth_gss.c +++ b/net/sunrpc/auth_gss/auth_gss.c @@ -89,8 +89,8 @@ static struct rpc_wait_queue pipe_version_rpc_waitqueue; static DECLARE_WAIT_QUEUE_HEAD(pipe_version_waitqueue); static void gss_free_ctx(struct gss_cl_ctx *); -static struct rpc_pipe_ops gss_upcall_ops_v0; -static struct rpc_pipe_ops gss_upcall_ops_v1; +static const struct rpc_pipe_ops gss_upcall_ops_v0; +static const struct rpc_pipe_ops gss_upcall_ops_v1; static inline struct gss_cl_ctx * gss_get_ctx(struct gss_cl_ctx *ctx) @@ -1507,7 +1507,7 @@ static const struct rpc_credops gss_nullops = { .crunwrap_resp = gss_unwrap_resp, }; -static struct rpc_pipe_ops gss_upcall_ops_v0 = { +static const struct rpc_pipe_ops gss_upcall_ops_v0 = { .upcall = gss_pipe_upcall, .downcall = gss_pipe_downcall, .destroy_msg = gss_pipe_destroy_msg, @@ -1515,7 +1515,7 @@ static struct rpc_pipe_ops gss_upcall_ops_v0 = { .release_pipe = gss_pipe_release, }; -static struct rpc_pipe_ops gss_upcall_ops_v1 = { +static const struct rpc_pipe_ops gss_upcall_ops_v1 = { .upcall = gss_pipe_upcall, .downcall = gss_pipe_downcall, .destroy_msg = gss_pipe_destroy_msg, diff --git a/net/sunrpc/rpc_pipe.c b/net/sunrpc/rpc_pipe.c index 9ced0628d69c..f6f60f625e3f 100644 --- a/net/sunrpc/rpc_pipe.c +++ b/net/sunrpc/rpc_pipe.c @@ -125,7 +125,7 @@ static void rpc_close_pipes(struct inode *inode) { struct rpc_inode *rpci = RPC_I(inode); - struct rpc_pipe_ops *ops; + const struct rpc_pipe_ops *ops; int need_release; mutex_lock(&inode->i_mutex); @@ -776,8 +776,9 @@ rpc_rmdir(struct dentry *dentry) * The @private argument passed here will be available to all these methods * from the file pointer, via RPC_I(file->f_dentry->d_inode)->private. */ -struct dentry * -rpc_mkpipe(struct dentry *parent, const char *name, void *private, struct rpc_pipe_ops *ops, int flags) +struct dentry *rpc_mkpipe(struct dentry *parent, const char *name, + void *private, const struct rpc_pipe_ops *ops, + int flags) { struct dentry *dentry; struct inode *dir, *inode; -- cgit v1.2.3 From 458adb8ba9b26bfc66593866013adbb62a1a3d2e Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Sun, 9 Aug 2009 15:14:22 -0400 Subject: SUNRPC: Rename rpc_mkdir to rpc_create_client_dir() This reflects the fact that rpc_mkdir() as it stands today, can only create a RPC client type directory. Signed-off-by: Trond Myklebust --- include/linux/sunrpc/rpc_pipe_fs.h | 5 +++-- net/sunrpc/clnt.c | 6 +++--- net/sunrpc/rpc_pipe.c | 11 +++++------ 3 files changed, 11 insertions(+), 11 deletions(-) (limited to 'include') diff --git a/include/linux/sunrpc/rpc_pipe_fs.h b/include/linux/sunrpc/rpc_pipe_fs.h index 91f5b13389c5..8de0ac276499 100644 --- a/include/linux/sunrpc/rpc_pipe_fs.h +++ b/include/linux/sunrpc/rpc_pipe_fs.h @@ -44,8 +44,9 @@ RPC_I(struct inode *inode) extern int rpc_queue_upcall(struct inode *, struct rpc_pipe_msg *); -extern struct dentry *rpc_mkdir(char *, struct rpc_clnt *); -extern int rpc_rmdir(struct dentry *); +struct rpc_clnt; +extern struct dentry *rpc_create_client_dir(const char *, struct rpc_clnt *); +extern int rpc_remove_client_dir(struct dentry *); extern struct dentry *rpc_mkpipe(struct dentry *, const char *, void *, const struct rpc_pipe_ops *, int flags); extern int rpc_unlink(struct dentry *); diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index ebfcf9b89909..6ec37701a165 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -113,7 +113,7 @@ rpc_setup_pipedir(struct rpc_clnt *clnt, char *dir_name) "%s/clnt%x", dir_name, (unsigned int)clntid++); clnt->cl_pathname[sizeof(clnt->cl_pathname) - 1] = '\0'; - clnt->cl_dentry = rpc_mkdir(clnt->cl_pathname, clnt); + clnt->cl_dentry = rpc_create_client_dir(clnt->cl_pathname, clnt); if (!IS_ERR(clnt->cl_dentry)) return 0; error = PTR_ERR(clnt->cl_dentry); @@ -232,7 +232,7 @@ static struct rpc_clnt * rpc_new_client(const struct rpc_create_args *args, stru out_no_auth: if (!IS_ERR(clnt->cl_dentry)) { - rpc_rmdir(clnt->cl_dentry); + rpc_remove_client_dir(clnt->cl_dentry); rpc_put_mount(); } out_no_path: @@ -424,7 +424,7 @@ rpc_free_client(struct kref *kref) dprintk("RPC: destroying %s client for %s\n", clnt->cl_protname, clnt->cl_server); if (!IS_ERR(clnt->cl_dentry)) { - rpc_rmdir(clnt->cl_dentry); + rpc_remove_client_dir(clnt->cl_dentry); rpc_put_mount(); } if (clnt->cl_parent != clnt) { diff --git a/net/sunrpc/rpc_pipe.c b/net/sunrpc/rpc_pipe.c index 2a4e6eb0a3e9..08580bedc25f 100644 --- a/net/sunrpc/rpc_pipe.c +++ b/net/sunrpc/rpc_pipe.c @@ -755,7 +755,7 @@ out_bad: } /** - * rpc_mkdir - Create a new directory in rpc_pipefs + * rpc_create_client_dir - Create a new rpc_client directory in rpc_pipefs * @path: path from the rpc_pipefs root to the new directory * @rpc_client: rpc client to associate with this directory * @@ -764,8 +764,8 @@ out_bad: * information about the client, together with any "pipes" that may * later be created using rpc_mkpipe(). */ -struct dentry * -rpc_mkdir(char *path, struct rpc_clnt *rpc_client) +struct dentry *rpc_create_client_dir(const char *path, + struct rpc_clnt *rpc_client) { struct nameidata nd; struct dentry *dentry; @@ -797,11 +797,10 @@ out_err: } /** - * rpc_rmdir - Remove a directory created with rpc_mkdir() + * rpc_remove_client_dir - Remove a directory created with rpc_create_client_dir() * @dentry: directory to remove */ -int -rpc_rmdir(struct dentry *dentry) +int rpc_remove_client_dir(struct dentry *dentry) { struct dentry *parent; struct inode *dir; -- cgit v1.2.3 From 7d217caca5d704e48aa5e59aba0b3ad4c7af4fd2 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Sun, 9 Aug 2009 15:14:24 -0400 Subject: SUNRPC: Replace rpc_client->cl_dentry and cl_mnt, with a cl_path Signed-off-by: Trond Myklebust --- fs/nfs/idmap.c | 4 ++-- include/linux/sunrpc/clnt.h | 4 ++-- net/sunrpc/auth_gss/auth_gss.c | 4 ++-- net/sunrpc/clnt.c | 24 ++++++++++++------------ 4 files changed, 18 insertions(+), 18 deletions(-) (limited to 'include') diff --git a/fs/nfs/idmap.c b/fs/nfs/idmap.c index fae0d3e52b44..21a84d45916f 100644 --- a/fs/nfs/idmap.c +++ b/fs/nfs/idmap.c @@ -119,8 +119,8 @@ nfs_idmap_new(struct nfs_client *clp) if (idmap == NULL) return -ENOMEM; - idmap->idmap_dentry = rpc_mkpipe(clp->cl_rpcclient->cl_dentry, "idmap", - idmap, &idmap_upcall_ops, 0); + idmap->idmap_dentry = rpc_mkpipe(clp->cl_rpcclient->cl_path.dentry, + "idmap", idmap, &idmap_upcall_ops, 0); if (IS_ERR(idmap->idmap_dentry)) { error = PTR_ERR(idmap->idmap_dentry); kfree(idmap); diff --git a/include/linux/sunrpc/clnt.h b/include/linux/sunrpc/clnt.h index 37881f1a0bd7..38ad162330a2 100644 --- a/include/linux/sunrpc/clnt.h +++ b/include/linux/sunrpc/clnt.h @@ -17,6 +17,7 @@ #include #include #include +#include struct rpc_inode; @@ -51,8 +52,7 @@ struct rpc_clnt { int cl_nodelen; /* nodename length */ char cl_nodename[UNX_MAXNODENAME]; char cl_pathname[30];/* Path in rpc_pipe_fs */ - struct vfsmount * cl_vfsmnt; - struct dentry * cl_dentry; /* inode */ + struct path cl_path; struct rpc_clnt * cl_parent; /* Points to parent of clones */ struct rpc_rtt cl_rtt_default; struct rpc_timeout cl_timeout_default; diff --git a/net/sunrpc/auth_gss/auth_gss.c b/net/sunrpc/auth_gss/auth_gss.c index 23eb3864ffc0..fc6a43ccd950 100644 --- a/net/sunrpc/auth_gss/auth_gss.c +++ b/net/sunrpc/auth_gss/auth_gss.c @@ -777,7 +777,7 @@ gss_create(struct rpc_clnt *clnt, rpc_authflavor_t flavor) * that we supported only the old pipe. So we instead create * the new pipe first. */ - gss_auth->dentry[1] = rpc_mkpipe(clnt->cl_dentry, + gss_auth->dentry[1] = rpc_mkpipe(clnt->cl_path.dentry, "gssd", clnt, &gss_upcall_ops_v1, RPC_PIPE_WAIT_FOR_OPEN); @@ -786,7 +786,7 @@ gss_create(struct rpc_clnt *clnt, rpc_authflavor_t flavor) goto err_put_mech; } - gss_auth->dentry[0] = rpc_mkpipe(clnt->cl_dentry, + gss_auth->dentry[0] = rpc_mkpipe(clnt->cl_path.dentry, gss_auth->mech->gm_name, clnt, &gss_upcall_ops_v0, RPC_PIPE_WAIT_FOR_OPEN); diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index 6ec37701a165..b3f863346300 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -99,24 +99,24 @@ rpc_setup_pipedir(struct rpc_clnt *clnt, char *dir_name) static uint32_t clntid; int error; - clnt->cl_vfsmnt = ERR_PTR(-ENOENT); - clnt->cl_dentry = ERR_PTR(-ENOENT); + clnt->cl_path.mnt = ERR_PTR(-ENOENT); + clnt->cl_path.dentry = ERR_PTR(-ENOENT); if (dir_name == NULL) return 0; - clnt->cl_vfsmnt = rpc_get_mount(); - if (IS_ERR(clnt->cl_vfsmnt)) - return PTR_ERR(clnt->cl_vfsmnt); + clnt->cl_path.mnt = rpc_get_mount(); + if (IS_ERR(clnt->cl_path.mnt)) + return PTR_ERR(clnt->cl_path.mnt); for (;;) { snprintf(clnt->cl_pathname, sizeof(clnt->cl_pathname), "%s/clnt%x", dir_name, (unsigned int)clntid++); clnt->cl_pathname[sizeof(clnt->cl_pathname) - 1] = '\0'; - clnt->cl_dentry = rpc_create_client_dir(clnt->cl_pathname, clnt); - if (!IS_ERR(clnt->cl_dentry)) + clnt->cl_path.dentry = rpc_create_client_dir(clnt->cl_pathname, clnt); + if (!IS_ERR(clnt->cl_path.dentry)) return 0; - error = PTR_ERR(clnt->cl_dentry); + error = PTR_ERR(clnt->cl_path.dentry); if (error != -EEXIST) { printk(KERN_INFO "RPC: Couldn't create pipefs entry %s, error %d\n", clnt->cl_pathname, error); @@ -231,8 +231,8 @@ static struct rpc_clnt * rpc_new_client(const struct rpc_create_args *args, stru return clnt; out_no_auth: - if (!IS_ERR(clnt->cl_dentry)) { - rpc_remove_client_dir(clnt->cl_dentry); + if (!IS_ERR(clnt->cl_path.dentry)) { + rpc_remove_client_dir(clnt->cl_path.dentry); rpc_put_mount(); } out_no_path: @@ -423,8 +423,8 @@ rpc_free_client(struct kref *kref) dprintk("RPC: destroying %s client for %s\n", clnt->cl_protname, clnt->cl_server); - if (!IS_ERR(clnt->cl_dentry)) { - rpc_remove_client_dir(clnt->cl_dentry); + if (!IS_ERR(clnt->cl_path.dentry)) { + rpc_remove_client_dir(clnt->cl_path.dentry); rpc_put_mount(); } if (clnt->cl_parent != clnt) { -- cgit v1.2.3 From 23ac6581702ac6d029643328a7e6ea3baf834c5e Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Sun, 9 Aug 2009 15:14:25 -0400 Subject: SUNRPC: clean up rpc_setup_pipedir() There is still a little wart or two there: Since we've already got a vfsmount, we might as well pass that in to rpc_create_client_dir. Another point is that if we open code __rpc_lookup_path() here, then we can avoid looking up the entire parent directory path over and over again: it doesn't change. Also get rid of rpc_clnt->cl_pathname, since it has no users... Signed-off-by: Trond Myklebust --- include/linux/sunrpc/clnt.h | 1 - include/linux/sunrpc/rpc_pipe_fs.h | 2 +- net/sunrpc/clnt.c | 48 +++++++++++++++++++++---------- net/sunrpc/rpc_pipe.c | 58 ++------------------------------------ 4 files changed, 37 insertions(+), 72 deletions(-) (limited to 'include') diff --git a/include/linux/sunrpc/clnt.h b/include/linux/sunrpc/clnt.h index 38ad162330a2..1848d44922ef 100644 --- a/include/linux/sunrpc/clnt.h +++ b/include/linux/sunrpc/clnt.h @@ -51,7 +51,6 @@ struct rpc_clnt { int cl_nodelen; /* nodename length */ char cl_nodename[UNX_MAXNODENAME]; - char cl_pathname[30];/* Path in rpc_pipe_fs */ struct path cl_path; struct rpc_clnt * cl_parent; /* Points to parent of clones */ struct rpc_rtt cl_rtt_default; diff --git a/include/linux/sunrpc/rpc_pipe_fs.h b/include/linux/sunrpc/rpc_pipe_fs.h index 8de0ac276499..88332ef1e959 100644 --- a/include/linux/sunrpc/rpc_pipe_fs.h +++ b/include/linux/sunrpc/rpc_pipe_fs.h @@ -45,7 +45,7 @@ RPC_I(struct inode *inode) extern int rpc_queue_upcall(struct inode *, struct rpc_pipe_msg *); struct rpc_clnt; -extern struct dentry *rpc_create_client_dir(const char *, struct rpc_clnt *); +extern struct dentry *rpc_create_client_dir(struct dentry *, struct qstr *, struct rpc_clnt *); extern int rpc_remove_client_dir(struct dentry *); extern struct dentry *rpc_mkpipe(struct dentry *, const char *, void *, const struct rpc_pipe_ops *, int flags); diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index b3f863346300..c1e467e1b07d 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -27,6 +27,8 @@ #include #include #include +#include +#include #include #include #include @@ -97,6 +99,12 @@ static int rpc_setup_pipedir(struct rpc_clnt *clnt, char *dir_name) { static uint32_t clntid; + struct nameidata nd; + struct path path; + char name[15]; + struct qstr q = { + .name = name, + }; int error; clnt->cl_path.mnt = ERR_PTR(-ENOENT); @@ -104,26 +112,36 @@ rpc_setup_pipedir(struct rpc_clnt *clnt, char *dir_name) if (dir_name == NULL) return 0; - clnt->cl_path.mnt = rpc_get_mount(); - if (IS_ERR(clnt->cl_path.mnt)) - return PTR_ERR(clnt->cl_path.mnt); + path.mnt = rpc_get_mount(); + if (IS_ERR(path.mnt)) + return PTR_ERR(path.mnt); + error = vfs_path_lookup(path.mnt->mnt_root, path.mnt, dir_name, 0, &nd); + if (error) + goto err; for (;;) { - snprintf(clnt->cl_pathname, sizeof(clnt->cl_pathname), - "%s/clnt%x", dir_name, - (unsigned int)clntid++); - clnt->cl_pathname[sizeof(clnt->cl_pathname) - 1] = '\0'; - clnt->cl_path.dentry = rpc_create_client_dir(clnt->cl_pathname, clnt); - if (!IS_ERR(clnt->cl_path.dentry)) - return 0; - error = PTR_ERR(clnt->cl_path.dentry); + q.len = snprintf(name, sizeof(name), "clnt%x", (unsigned int)clntid++); + name[sizeof(name) - 1] = '\0'; + q.hash = full_name_hash(q.name, q.len); + path.dentry = rpc_create_client_dir(nd.path.dentry, &q, clnt); + if (!IS_ERR(path.dentry)) + break; + error = PTR_ERR(path.dentry); if (error != -EEXIST) { - printk(KERN_INFO "RPC: Couldn't create pipefs entry %s, error %d\n", - clnt->cl_pathname, error); - rpc_put_mount(); - return error; + printk(KERN_INFO "RPC: Couldn't create pipefs entry" + " %s/%s, error %d\n", + dir_name, name, error); + goto err_path_put; } } + path_put(&nd.path); + clnt->cl_path = path; + return 0; +err_path_put: + path_put(&nd.path); +err: + rpc_put_mount(); + return error; } static struct rpc_clnt * rpc_new_client(const struct rpc_create_args *args, struct rpc_xprt *xprt) diff --git a/net/sunrpc/rpc_pipe.c b/net/sunrpc/rpc_pipe.c index 6d152f69587e..1613d858ba38 100644 --- a/net/sunrpc/rpc_pipe.c +++ b/net/sunrpc/rpc_pipe.c @@ -443,42 +443,6 @@ static const struct dentry_operations rpc_dentry_operations = { .d_delete = rpc_delete_dentry, }; -static int __rpc_lookup_path(const char *pathname, unsigned flags, - struct nameidata *nd) -{ - struct vfsmount *mnt; - - if (pathname[0] == '\0') - return -ENOENT; - - mnt = rpc_get_mount(); - if (IS_ERR(mnt)) { - printk(KERN_WARNING "%s: %s failed to mount " - "pseudofilesystem \n", __FILE__, __func__); - return PTR_ERR(mnt); - } - - if (vfs_path_lookup(mnt->mnt_root, mnt, pathname, flags, nd)) { - printk(KERN_WARNING "%s: %s failed to find path %s\n", - __FILE__, __func__, pathname); - rpc_put_mount(); - return -ENOENT; - } - return 0; -} - -static int rpc_lookup_parent(const char *pathname, struct nameidata *nd) -{ - return __rpc_lookup_path(pathname, LOOKUP_PARENT, nd); -} - -static void -rpc_release_path(struct nameidata *nd) -{ - path_put(&nd->path); - rpc_put_mount(); -} - static struct inode * rpc_get_inode(struct super_block *sb, umode_t mode) { @@ -889,27 +853,11 @@ EXPORT_SYMBOL_GPL(rpc_unlink); * information about the client, together with any "pipes" that may * later be created using rpc_mkpipe(). */ -struct dentry *rpc_create_client_dir(const char *path, +struct dentry *rpc_create_client_dir(struct dentry *dentry, + struct qstr *name, struct rpc_clnt *rpc_client) { - struct nameidata nd; - struct dentry *ret; - struct inode *dir; - - ret = ERR_PTR(rpc_lookup_parent(path, &nd)); - if (IS_ERR(ret)) - goto out_err; - dir = nd.path.dentry->d_inode; - - ret = rpc_mkdir_populate(nd.path.dentry, &nd.last, - S_IRUGO | S_IXUGO, rpc_client); - rpc_release_path(&nd); - if (!IS_ERR(ret)) - return ret; -out_err: - printk(KERN_WARNING "%s: %s() failed to create directory %s (errno = %ld)\n", - __FILE__, __func__, path, PTR_ERR(ret)); - return ret; + return rpc_mkdir_populate(dentry, name, S_IRUGO | S_IXUGO, rpc_client); } /* -- cgit v1.2.3 From 2da8ca26c6bfad685bfddf39728eac1c83906aa9 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Sun, 9 Aug 2009 15:14:26 -0400 Subject: NFSD: Clean up the idmapper warning... What part of 'internal use' is so hard to understand? Signed-off-by: Trond Myklebust --- fs/nfsd/nfs4idmap.c | 4 ++-- include/linux/sunrpc/cache.h | 3 ++- net/sunrpc/cache.c | 2 +- 3 files changed, 5 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/fs/nfsd/nfs4idmap.c b/fs/nfsd/nfs4idmap.c index 5b398421b051..e9012ad36ac0 100644 --- a/fs/nfsd/nfs4idmap.c +++ b/fs/nfsd/nfs4idmap.c @@ -175,10 +175,10 @@ idtoname_show(struct seq_file *m, struct cache_detail *cd, struct cache_head *h) } static void -warn_no_idmapd(struct cache_detail *detail) +warn_no_idmapd(struct cache_detail *detail, int has_died) { printk("nfsd: nfsv4 idmapping failing: has idmapd %s?\n", - detail->last_close? "died" : "not been started"); + has_died ? "died" : "not been started"); } diff --git a/include/linux/sunrpc/cache.h b/include/linux/sunrpc/cache.h index 2d8b211b9324..3d1fad22185a 100644 --- a/include/linux/sunrpc/cache.h +++ b/include/linux/sunrpc/cache.h @@ -79,6 +79,8 @@ struct cache_detail { int (*cache_show)(struct seq_file *m, struct cache_detail *cd, struct cache_head *h); + void (*warn_no_listener)(struct cache_detail *cd, + int has_died); struct cache_head * (*alloc)(void); int (*match)(struct cache_head *orig, struct cache_head *new); @@ -102,7 +104,6 @@ struct cache_detail { atomic_t readers; /* how many time is /chennel open */ time_t last_close; /* if no readers, when did last close */ time_t last_warn; /* when we last warned about no readers */ - void (*warn_no_listener)(struct cache_detail *cd); }; diff --git a/net/sunrpc/cache.c b/net/sunrpc/cache.c index ff0c23053d2f..8ede4a6f384b 100644 --- a/net/sunrpc/cache.c +++ b/net/sunrpc/cache.c @@ -1020,7 +1020,7 @@ static void warn_no_listener(struct cache_detail *detail) if (detail->last_warn != detail->last_close) { detail->last_warn = detail->last_close; if (detail->warn_no_listener) - detail->warn_no_listener(detail); + detail->warn_no_listener(detail, detail->last_close != 0); } } -- cgit v1.2.3 From bc74b4f5e63a09fb78e245794a0de1e5a2716bbe Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Sun, 9 Aug 2009 15:14:29 -0400 Subject: SUNRPC: Allow the cache_detail to specify alternative upcall mechanisms For events that are rare, such as referral DNS lookups, it makes limited sense to have a daemon constantly listening for upcalls on a channel. An alternative in those cases might simply be to run the app that fills the cache using call_usermodehelper_exec() and friends. The following patch allows the cache_detail to specify alternative upcall mechanisms for these particular cases. Signed-off-by: Trond Myklebust --- fs/nfsd/export.c | 14 ++++++++++++-- fs/nfsd/nfs4idmap.c | 16 ++++++++++++++-- include/linux/sunrpc/cache.h | 13 ++++++++++--- net/sunrpc/auth_gss/svcauth_gss.c | 7 ++++++- net/sunrpc/cache.c | 26 ++++++++++++++++++-------- net/sunrpc/svcauth_unix.c | 14 ++++++++++++-- 6 files changed, 72 insertions(+), 18 deletions(-) (limited to 'include') diff --git a/fs/nfsd/export.c b/fs/nfsd/export.c index b92a27629fb7..d9462643155c 100644 --- a/fs/nfsd/export.c +++ b/fs/nfsd/export.c @@ -85,6 +85,11 @@ static void expkey_request(struct cache_detail *cd, (*bpp)[-1] = '\n'; } +static int expkey_upcall(struct cache_detail *cd, struct cache_head *h) +{ + return sunrpc_cache_pipe_upcall(cd, h, expkey_request); +} + static struct svc_expkey *svc_expkey_update(struct svc_expkey *new, struct svc_expkey *old); static struct svc_expkey *svc_expkey_lookup(struct svc_expkey *); static struct cache_detail svc_expkey_cache; @@ -259,7 +264,7 @@ static struct cache_detail svc_expkey_cache = { .hash_table = expkey_table, .name = "nfsd.fh", .cache_put = expkey_put, - .cache_request = expkey_request, + .cache_upcall = expkey_upcall, .cache_parse = expkey_parse, .cache_show = expkey_show, .match = expkey_match, @@ -355,6 +360,11 @@ static void svc_export_request(struct cache_detail *cd, (*bpp)[-1] = '\n'; } +static int svc_export_upcall(struct cache_detail *cd, struct cache_head *h) +{ + return sunrpc_cache_pipe_upcall(cd, h, svc_export_request); +} + static struct svc_export *svc_export_update(struct svc_export *new, struct svc_export *old); static struct svc_export *svc_export_lookup(struct svc_export *); @@ -724,7 +734,7 @@ struct cache_detail svc_export_cache = { .hash_table = export_table, .name = "nfsd.export", .cache_put = svc_export_put, - .cache_request = svc_export_request, + .cache_upcall = svc_export_upcall, .cache_parse = svc_export_parse, .cache_show = svc_export_show, .match = svc_export_match, diff --git a/fs/nfsd/nfs4idmap.c b/fs/nfsd/nfs4idmap.c index e9012ad36ac0..cdfa86fa1471 100644 --- a/fs/nfsd/nfs4idmap.c +++ b/fs/nfsd/nfs4idmap.c @@ -145,6 +145,12 @@ idtoname_request(struct cache_detail *cd, struct cache_head *ch, char **bpp, (*bpp)[-1] = '\n'; } +static int +idtoname_upcall(struct cache_detail *cd, struct cache_head *ch) +{ + return sunrpc_cache_pipe_upcall(cd, ch, idtoname_request); +} + static int idtoname_match(struct cache_head *ca, struct cache_head *cb) { @@ -192,7 +198,7 @@ static struct cache_detail idtoname_cache = { .hash_table = idtoname_table, .name = "nfs4.idtoname", .cache_put = ent_put, - .cache_request = idtoname_request, + .cache_upcall = idtoname_upcall, .cache_parse = idtoname_parse, .cache_show = idtoname_show, .warn_no_listener = warn_no_idmapd, @@ -324,6 +330,12 @@ nametoid_request(struct cache_detail *cd, struct cache_head *ch, char **bpp, (*bpp)[-1] = '\n'; } +static int +nametoid_upcall(struct cache_detail *cd, struct cache_head *ch) +{ + return sunrpc_cache_pipe_upcall(cd, ch, nametoid_request); +} + static int nametoid_match(struct cache_head *ca, struct cache_head *cb) { @@ -363,7 +375,7 @@ static struct cache_detail nametoid_cache = { .hash_table = nametoid_table, .name = "nfs4.nametoid", .cache_put = ent_put, - .cache_request = nametoid_request, + .cache_upcall = nametoid_upcall, .cache_parse = nametoid_parse, .cache_show = nametoid_show, .warn_no_listener = warn_no_idmapd, diff --git a/include/linux/sunrpc/cache.h b/include/linux/sunrpc/cache.h index 3d1fad22185a..23ee25981a0c 100644 --- a/include/linux/sunrpc/cache.h +++ b/include/linux/sunrpc/cache.h @@ -70,9 +70,9 @@ struct cache_detail { char *name; void (*cache_put)(struct kref *); - void (*cache_request)(struct cache_detail *cd, - struct cache_head *h, - char **bpp, int *blen); + int (*cache_upcall)(struct cache_detail *, + struct cache_head *); + int (*cache_parse)(struct cache_detail *, char *buf, int len); @@ -135,6 +135,13 @@ extern struct cache_head * sunrpc_cache_update(struct cache_detail *detail, struct cache_head *new, struct cache_head *old, int hash); +extern int +sunrpc_cache_pipe_upcall(struct cache_detail *detail, struct cache_head *h, + void (*cache_request)(struct cache_detail *, + struct cache_head *, + char **, + int *)); + extern void cache_clean_deferred(void *owner); diff --git a/net/sunrpc/auth_gss/svcauth_gss.c b/net/sunrpc/auth_gss/svcauth_gss.c index 2278a50c6444..2e6a148d277c 100644 --- a/net/sunrpc/auth_gss/svcauth_gss.c +++ b/net/sunrpc/auth_gss/svcauth_gss.c @@ -181,6 +181,11 @@ static void rsi_request(struct cache_detail *cd, (*bpp)[-1] = '\n'; } +static int rsi_upcall(struct cache_detail *cd, struct cache_head *h) +{ + return sunrpc_cache_pipe_upcall(cd, h, rsi_request); +} + static int rsi_parse(struct cache_detail *cd, char *mesg, int mlen) @@ -270,7 +275,7 @@ static struct cache_detail rsi_cache = { .hash_table = rsi_table, .name = "auth.rpcsec.init", .cache_put = rsi_put, - .cache_request = rsi_request, + .cache_upcall = rsi_upcall, .cache_parse = rsi_parse, .match = rsi_match, .init = rsi_init, diff --git a/net/sunrpc/cache.c b/net/sunrpc/cache.c index c8e7d2d07260..e438352bed7b 100644 --- a/net/sunrpc/cache.c +++ b/net/sunrpc/cache.c @@ -176,7 +176,13 @@ struct cache_head *sunrpc_cache_update(struct cache_detail *detail, } EXPORT_SYMBOL_GPL(sunrpc_cache_update); -static int cache_make_upcall(struct cache_detail *detail, struct cache_head *h); +static int cache_make_upcall(struct cache_detail *cd, struct cache_head *h) +{ + if (!cd->cache_upcall) + return -EINVAL; + return cd->cache_upcall(cd, h); +} + /* * This is the generic cache management routine for all * the authentication caches. @@ -322,7 +328,7 @@ static int create_cache_proc_entries(struct cache_detail *cd) if (p == NULL) goto out_nomem; - if (cd->cache_request || cd->cache_parse) { + if (cd->cache_upcall || cd->cache_parse) { p = proc_create_data("channel", S_IFREG|S_IRUSR|S_IWUSR, cd->proc_ent, &cache_file_operations, cd); cd->channel_ent = p; @@ -1080,10 +1086,16 @@ static void warn_no_listener(struct cache_detail *detail) } /* - * register an upcall request to user-space. + * register an upcall request to user-space and queue it up for read() by the + * upcall daemon. + * * Each request is at most one page long. */ -static int cache_make_upcall(struct cache_detail *detail, struct cache_head *h) +int sunrpc_cache_pipe_upcall(struct cache_detail *detail, struct cache_head *h, + void (*cache_request)(struct cache_detail *, + struct cache_head *, + char **, + int *)) { char *buf; @@ -1091,9 +1103,6 @@ static int cache_make_upcall(struct cache_detail *detail, struct cache_head *h) char *bp; int len; - if (detail->cache_request == NULL) - return -EINVAL; - if (atomic_read(&detail->readers) == 0 && detail->last_close < get_seconds() - 30) { warn_no_listener(detail); @@ -1112,7 +1121,7 @@ static int cache_make_upcall(struct cache_detail *detail, struct cache_head *h) bp = buf; len = PAGE_SIZE; - detail->cache_request(detail, h, &bp, &len); + cache_request(detail, h, &bp, &len); if (len < 0) { kfree(buf); @@ -1130,6 +1139,7 @@ static int cache_make_upcall(struct cache_detail *detail, struct cache_head *h) wake_up(&queue_wait); return 0; } +EXPORT_SYMBOL_GPL(sunrpc_cache_pipe_upcall); /* * parse a message from user-space and pass it diff --git a/net/sunrpc/svcauth_unix.c b/net/sunrpc/svcauth_unix.c index 5c865e2d299e..6caffa34ac01 100644 --- a/net/sunrpc/svcauth_unix.c +++ b/net/sunrpc/svcauth_unix.c @@ -171,6 +171,11 @@ static void ip_map_request(struct cache_detail *cd, (*bpp)[-1] = '\n'; } +static int ip_map_upcall(struct cache_detail *cd, struct cache_head *h) +{ + return sunrpc_cache_pipe_upcall(cd, h, ip_map_request); +} + static struct ip_map *ip_map_lookup(char *class, struct in6_addr *addr); static int ip_map_update(struct ip_map *ipm, struct unix_domain *udom, time_t expiry); @@ -289,7 +294,7 @@ struct cache_detail ip_map_cache = { .hash_table = ip_table, .name = "auth.unix.ip", .cache_put = ip_map_put, - .cache_request = ip_map_request, + .cache_upcall = ip_map_upcall, .cache_parse = ip_map_parse, .cache_show = ip_map_show, .match = ip_map_match, @@ -523,6 +528,11 @@ static void unix_gid_request(struct cache_detail *cd, (*bpp)[-1] = '\n'; } +static int unix_gid_upcall(struct cache_detail *cd, struct cache_head *h) +{ + return sunrpc_cache_pipe_upcall(cd, h, unix_gid_request); +} + static struct unix_gid *unix_gid_lookup(uid_t uid); extern struct cache_detail unix_gid_cache; @@ -622,7 +632,7 @@ struct cache_detail unix_gid_cache = { .hash_table = gid_table, .name = "auth.unix.gid", .cache_put = unix_gid_put, - .cache_request = unix_gid_request, + .cache_upcall = unix_gid_upcall, .cache_parse = unix_gid_parse, .cache_show = unix_gid_show, .match = unix_gid_match, -- cgit v1.2.3 From 173912a6add00f4715774dcecf9ee53274c5924c Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Sun, 9 Aug 2009 15:14:29 -0400 Subject: SUNRPC: Move procfs-specific stuff out of the generic sunrpc cache code Signed-off-by: Trond Myklebust --- include/linux/sunrpc/cache.h | 11 +- net/sunrpc/cache.c | 319 ++++++++++++++++++++++++++----------------- 2 files changed, 199 insertions(+), 131 deletions(-) (limited to 'include') diff --git a/include/linux/sunrpc/cache.h b/include/linux/sunrpc/cache.h index 23ee25981a0c..8e5bf3036652 100644 --- a/include/linux/sunrpc/cache.h +++ b/include/linux/sunrpc/cache.h @@ -59,6 +59,11 @@ struct cache_head { #define CACHE_NEW_EXPIRY 120 /* keep new things pending confirmation for 120 seconds */ +struct cache_detail_procfs { + struct proc_dir_entry *proc_ent; + struct proc_dir_entry *flush_ent, *channel_ent, *content_ent; +}; + struct cache_detail { struct module * owner; int hash_size; @@ -98,12 +103,14 @@ struct cache_detail { /* fields for communication over channel */ struct list_head queue; - struct proc_dir_entry *proc_ent; - struct proc_dir_entry *flush_ent, *channel_ent, *content_ent; atomic_t readers; /* how many time is /chennel open */ time_t last_close; /* if no readers, when did last close */ time_t last_warn; /* when we last warned about no readers */ + + union { + struct cache_detail_procfs procfs; + } u; }; diff --git a/net/sunrpc/cache.c b/net/sunrpc/cache.c index e438352bed7b..1cd82eda56d0 100644 --- a/net/sunrpc/cache.c +++ b/net/sunrpc/cache.c @@ -291,69 +291,9 @@ static DEFINE_SPINLOCK(cache_list_lock); static struct cache_detail *current_detail; static int current_index; -static const struct file_operations cache_file_operations; -static const struct file_operations content_file_operations; -static const struct file_operations cache_flush_operations; - static void do_cache_clean(struct work_struct *work); static DECLARE_DELAYED_WORK(cache_cleaner, do_cache_clean); -static void remove_cache_proc_entries(struct cache_detail *cd) -{ - if (cd->proc_ent == NULL) - return; - if (cd->flush_ent) - remove_proc_entry("flush", cd->proc_ent); - if (cd->channel_ent) - remove_proc_entry("channel", cd->proc_ent); - if (cd->content_ent) - remove_proc_entry("content", cd->proc_ent); - cd->proc_ent = NULL; - remove_proc_entry(cd->name, proc_net_rpc); -} - -#ifdef CONFIG_PROC_FS -static int create_cache_proc_entries(struct cache_detail *cd) -{ - struct proc_dir_entry *p; - - cd->proc_ent = proc_mkdir(cd->name, proc_net_rpc); - if (cd->proc_ent == NULL) - goto out_nomem; - cd->channel_ent = cd->content_ent = NULL; - - p = proc_create_data("flush", S_IFREG|S_IRUSR|S_IWUSR, - cd->proc_ent, &cache_flush_operations, cd); - cd->flush_ent = p; - if (p == NULL) - goto out_nomem; - - if (cd->cache_upcall || cd->cache_parse) { - p = proc_create_data("channel", S_IFREG|S_IRUSR|S_IWUSR, - cd->proc_ent, &cache_file_operations, cd); - cd->channel_ent = p; - if (p == NULL) - goto out_nomem; - } - if (cd->cache_show) { - p = proc_create_data("content", S_IFREG|S_IRUSR|S_IWUSR, - cd->proc_ent, &content_file_operations, cd); - cd->content_ent = p; - if (p == NULL) - goto out_nomem; - } - return 0; -out_nomem: - remove_cache_proc_entries(cd); - return -ENOMEM; -} -#else /* CONFIG_PROC_FS */ -static int create_cache_proc_entries(struct cache_detail *cd) -{ - return 0; -} -#endif - static void sunrpc_init_cache_detail(struct cache_detail *cd) { rwlock_init(&cd->hash_lock); @@ -395,25 +335,6 @@ out: printk(KERN_ERR "nfsd: failed to unregister %s cache\n", cd->name); } -int cache_register(struct cache_detail *cd) -{ - int ret; - - sunrpc_init_cache_detail(cd); - ret = create_cache_proc_entries(cd); - if (ret) - sunrpc_destroy_cache_detail(cd); - return ret; -} -EXPORT_SYMBOL_GPL(cache_register); - -void cache_unregister(struct cache_detail *cd) -{ - remove_cache_proc_entries(cd); - sunrpc_destroy_cache_detail(cd); -} -EXPORT_SYMBOL_GPL(cache_unregister); - /* clean cache tries to find something to clean * and cleans it. * It returns 1 if it cleaned something, @@ -704,13 +625,12 @@ struct cache_reader { int offset; /* if non-0, we have a refcnt on next request */ }; -static ssize_t -cache_read(struct file *filp, char __user *buf, size_t count, loff_t *ppos) +static ssize_t cache_read(struct file *filp, char __user *buf, size_t count, + loff_t *ppos, struct cache_detail *cd) { struct cache_reader *rp = filp->private_data; struct cache_request *rq; struct inode *inode = filp->f_path.dentry->d_inode; - struct cache_detail *cd = PDE(inode)->data; int err; if (count == 0) @@ -834,13 +754,12 @@ out_slow: return cache_slow_downcall(buf, count, cd); } -static ssize_t -cache_write(struct file *filp, const char __user *buf, size_t count, - loff_t *ppos) +static ssize_t cache_write(struct file *filp, const char __user *buf, + size_t count, loff_t *ppos, + struct cache_detail *cd) { struct address_space *mapping = filp->f_mapping; struct inode *inode = filp->f_path.dentry->d_inode; - struct cache_detail *cd = PDE(inode)->data; ssize_t ret = -EINVAL; if (!cd->cache_parse) @@ -855,13 +774,12 @@ out: static DECLARE_WAIT_QUEUE_HEAD(queue_wait); -static unsigned int -cache_poll(struct file *filp, poll_table *wait) +static unsigned int cache_poll(struct file *filp, poll_table *wait, + struct cache_detail *cd) { unsigned int mask; struct cache_reader *rp = filp->private_data; struct cache_queue *cq; - struct cache_detail *cd = PDE(filp->f_path.dentry->d_inode)->data; poll_wait(filp, &queue_wait, wait); @@ -883,14 +801,13 @@ cache_poll(struct file *filp, poll_table *wait) return mask; } -static int -cache_ioctl(struct inode *ino, struct file *filp, - unsigned int cmd, unsigned long arg) +static int cache_ioctl(struct inode *ino, struct file *filp, + unsigned int cmd, unsigned long arg, + struct cache_detail *cd) { int len = 0; struct cache_reader *rp = filp->private_data; struct cache_queue *cq; - struct cache_detail *cd = PDE(ino)->data; if (cmd != FIONREAD || !rp) return -EINVAL; @@ -913,15 +830,13 @@ cache_ioctl(struct inode *ino, struct file *filp, return put_user(len, (int __user *)arg); } -static int -cache_open(struct inode *inode, struct file *filp) +static int cache_open(struct inode *inode, struct file *filp, + struct cache_detail *cd) { struct cache_reader *rp = NULL; nonseekable_open(inode, filp); if (filp->f_mode & FMODE_READ) { - struct cache_detail *cd = PDE(inode)->data; - rp = kmalloc(sizeof(*rp), GFP_KERNEL); if (!rp) return -ENOMEM; @@ -936,11 +851,10 @@ cache_open(struct inode *inode, struct file *filp) return 0; } -static int -cache_release(struct inode *inode, struct file *filp) +static int cache_release(struct inode *inode, struct file *filp, + struct cache_detail *cd) { struct cache_reader *rp = filp->private_data; - struct cache_detail *cd = PDE(inode)->data; if (rp) { spin_lock(&queue_lock); @@ -969,18 +883,6 @@ cache_release(struct inode *inode, struct file *filp) -static const struct file_operations cache_file_operations = { - .owner = THIS_MODULE, - .llseek = no_llseek, - .read = cache_read, - .write = cache_write, - .poll = cache_poll, - .ioctl = cache_ioctl, /* for FIONREAD */ - .open = cache_open, - .release = cache_release, -}; - - static void queue_loose(struct cache_detail *detail, struct cache_head *ch) { struct cache_queue *cq; @@ -1307,10 +1209,10 @@ static const struct seq_operations cache_content_op = { .show = c_show, }; -static int content_open(struct inode *inode, struct file *file) +static int content_open(struct inode *inode, struct file *file, + struct cache_detail *cd) { struct handle *han; - struct cache_detail *cd = PDE(inode)->data; han = __seq_open_private(file, &cache_content_op, sizeof(*han)); if (han == NULL) @@ -1320,17 +1222,10 @@ static int content_open(struct inode *inode, struct file *file) return 0; } -static const struct file_operations content_file_operations = { - .open = content_open, - .read = seq_read, - .llseek = seq_lseek, - .release = seq_release_private, -}; - static ssize_t read_flush(struct file *file, char __user *buf, - size_t count, loff_t *ppos) + size_t count, loff_t *ppos, + struct cache_detail *cd) { - struct cache_detail *cd = PDE(file->f_path.dentry->d_inode)->data; char tbuf[20]; unsigned long p = *ppos; size_t len; @@ -1348,10 +1243,10 @@ static ssize_t read_flush(struct file *file, char __user *buf, return len; } -static ssize_t write_flush(struct file * file, const char __user * buf, - size_t count, loff_t *ppos) +static ssize_t write_flush(struct file *file, const char __user *buf, + size_t count, loff_t *ppos, + struct cache_detail *cd) { - struct cache_detail *cd = PDE(file->f_path.dentry->d_inode)->data; char tbuf[20]; char *ep; long flushtime; @@ -1372,8 +1267,174 @@ static ssize_t write_flush(struct file * file, const char __user * buf, return count; } -static const struct file_operations cache_flush_operations = { +static ssize_t cache_read_procfs(struct file *filp, char __user *buf, + size_t count, loff_t *ppos) +{ + struct cache_detail *cd = PDE(filp->f_path.dentry->d_inode)->data; + + return cache_read(filp, buf, count, ppos, cd); +} + +static ssize_t cache_write_procfs(struct file *filp, const char __user *buf, + size_t count, loff_t *ppos) +{ + struct cache_detail *cd = PDE(filp->f_path.dentry->d_inode)->data; + + return cache_write(filp, buf, count, ppos, cd); +} + +static unsigned int cache_poll_procfs(struct file *filp, poll_table *wait) +{ + struct cache_detail *cd = PDE(filp->f_path.dentry->d_inode)->data; + + return cache_poll(filp, wait, cd); +} + +static int cache_ioctl_procfs(struct inode *inode, struct file *filp, + unsigned int cmd, unsigned long arg) +{ + struct cache_detail *cd = PDE(inode)->data; + + return cache_ioctl(inode, filp, cmd, arg, cd); +} + +static int cache_open_procfs(struct inode *inode, struct file *filp) +{ + struct cache_detail *cd = PDE(inode)->data; + + return cache_open(inode, filp, cd); +} + +static int cache_release_procfs(struct inode *inode, struct file *filp) +{ + struct cache_detail *cd = PDE(inode)->data; + + return cache_release(inode, filp, cd); +} + +static const struct file_operations cache_file_operations_procfs = { + .owner = THIS_MODULE, + .llseek = no_llseek, + .read = cache_read_procfs, + .write = cache_write_procfs, + .poll = cache_poll_procfs, + .ioctl = cache_ioctl_procfs, /* for FIONREAD */ + .open = cache_open_procfs, + .release = cache_release_procfs, +}; + +static int content_open_procfs(struct inode *inode, struct file *filp) +{ + struct cache_detail *cd = PDE(inode)->data; + + return content_open(inode, filp, cd); +} + +static const struct file_operations content_file_operations_procfs = { + .open = content_open_procfs, + .read = seq_read, + .llseek = seq_lseek, + .release = seq_release_private, +}; + +static ssize_t read_flush_procfs(struct file *filp, char __user *buf, + size_t count, loff_t *ppos) +{ + struct cache_detail *cd = PDE(filp->f_path.dentry->d_inode)->data; + + return read_flush(filp, buf, count, ppos, cd); +} + +static ssize_t write_flush_procfs(struct file *filp, + const char __user *buf, + size_t count, loff_t *ppos) +{ + struct cache_detail *cd = PDE(filp->f_path.dentry->d_inode)->data; + + return write_flush(filp, buf, count, ppos, cd); +} + +static const struct file_operations cache_flush_operations_procfs = { .open = nonseekable_open, - .read = read_flush, - .write = write_flush, + .read = read_flush_procfs, + .write = write_flush_procfs, }; + +static void remove_cache_proc_entries(struct cache_detail *cd) +{ + if (cd->u.procfs.proc_ent == NULL) + return; + if (cd->u.procfs.flush_ent) + remove_proc_entry("flush", cd->u.procfs.proc_ent); + if (cd->u.procfs.channel_ent) + remove_proc_entry("channel", cd->u.procfs.proc_ent); + if (cd->u.procfs.content_ent) + remove_proc_entry("content", cd->u.procfs.proc_ent); + cd->u.procfs.proc_ent = NULL; + remove_proc_entry(cd->name, proc_net_rpc); +} + +#ifdef CONFIG_PROC_FS +static int create_cache_proc_entries(struct cache_detail *cd) +{ + struct proc_dir_entry *p; + + cd->u.procfs.proc_ent = proc_mkdir(cd->name, proc_net_rpc); + if (cd->u.procfs.proc_ent == NULL) + goto out_nomem; + cd->u.procfs.channel_ent = NULL; + cd->u.procfs.content_ent = NULL; + + p = proc_create_data("flush", S_IFREG|S_IRUSR|S_IWUSR, + cd->u.procfs.proc_ent, + &cache_flush_operations_procfs, cd); + cd->u.procfs.flush_ent = p; + if (p == NULL) + goto out_nomem; + + if (cd->cache_upcall || cd->cache_parse) { + p = proc_create_data("channel", S_IFREG|S_IRUSR|S_IWUSR, + cd->u.procfs.proc_ent, + &cache_file_operations_procfs, cd); + cd->u.procfs.channel_ent = p; + if (p == NULL) + goto out_nomem; + } + if (cd->cache_show) { + p = proc_create_data("content", S_IFREG|S_IRUSR|S_IWUSR, + cd->u.procfs.proc_ent, + &content_file_operations_procfs, cd); + cd->u.procfs.content_ent = p; + if (p == NULL) + goto out_nomem; + } + return 0; +out_nomem: + remove_cache_proc_entries(cd); + return -ENOMEM; +} +#else /* CONFIG_PROC_FS */ +static int create_cache_proc_entries(struct cache_detail *cd) +{ + return 0; +} +#endif + +int cache_register(struct cache_detail *cd) +{ + int ret; + + sunrpc_init_cache_detail(cd); + ret = create_cache_proc_entries(cd); + if (ret) + sunrpc_destroy_cache_detail(cd); + return ret; +} +EXPORT_SYMBOL_GPL(cache_register); + +void cache_unregister(struct cache_detail *cd) +{ + remove_cache_proc_entries(cd); + sunrpc_destroy_cache_detail(cd); +} +EXPORT_SYMBOL_GPL(cache_unregister); -- cgit v1.2.3 From 8854e82d9accc80f43c0bc3ff06b5979ac858185 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Sun, 9 Aug 2009 15:14:30 -0400 Subject: SUNRPC: Add an rpc_pipefs front end for the sunrpc cache code Signed-off-by: Trond Myklebust --- include/linux/sunrpc/cache.h | 13 ++++ include/linux/sunrpc/rpc_pipe_fs.h | 8 +++ net/sunrpc/cache.c | 126 +++++++++++++++++++++++++++++++++++++ net/sunrpc/rpc_pipe.c | 43 +++++++++++++ 4 files changed, 190 insertions(+) (limited to 'include') diff --git a/include/linux/sunrpc/cache.h b/include/linux/sunrpc/cache.h index 8e5bf3036652..6f52b4d7c447 100644 --- a/include/linux/sunrpc/cache.h +++ b/include/linux/sunrpc/cache.h @@ -64,6 +64,10 @@ struct cache_detail_procfs { struct proc_dir_entry *flush_ent, *channel_ent, *content_ent; }; +struct cache_detail_pipefs { + struct dentry *dir; +}; + struct cache_detail { struct module * owner; int hash_size; @@ -110,6 +114,7 @@ struct cache_detail { union { struct cache_detail_procfs procfs; + struct cache_detail_pipefs pipefs; } u; }; @@ -135,6 +140,10 @@ struct cache_deferred_req { }; +extern const struct file_operations cache_file_operations_pipefs; +extern const struct file_operations content_file_operations_pipefs; +extern const struct file_operations cache_flush_operations_pipefs; + extern struct cache_head * sunrpc_cache_lookup(struct cache_detail *detail, struct cache_head *key, int hash); @@ -186,6 +195,10 @@ extern void cache_purge(struct cache_detail *detail); extern int cache_register(struct cache_detail *cd); extern void cache_unregister(struct cache_detail *cd); +extern int sunrpc_cache_register_pipefs(struct dentry *parent, const char *, + mode_t, struct cache_detail *); +extern void sunrpc_cache_unregister_pipefs(struct cache_detail *); + extern void qword_add(char **bpp, int *lp, char *str); extern void qword_addhex(char **bpp, int *lp, char *buf, int blen); extern int qword_get(char **bpp, char *dest, int bufsize); diff --git a/include/linux/sunrpc/rpc_pipe_fs.h b/include/linux/sunrpc/rpc_pipe_fs.h index 88332ef1e959..a92571a34556 100644 --- a/include/linux/sunrpc/rpc_pipe_fs.h +++ b/include/linux/sunrpc/rpc_pipe_fs.h @@ -47,6 +47,14 @@ extern int rpc_queue_upcall(struct inode *, struct rpc_pipe_msg *); struct rpc_clnt; extern struct dentry *rpc_create_client_dir(struct dentry *, struct qstr *, struct rpc_clnt *); extern int rpc_remove_client_dir(struct dentry *); + +struct cache_detail; +extern struct dentry *rpc_create_cache_dir(struct dentry *, + struct qstr *, + mode_t umode, + struct cache_detail *); +extern void rpc_remove_cache_dir(struct dentry *); + extern struct dentry *rpc_mkpipe(struct dentry *, const char *, void *, const struct rpc_pipe_ops *, int flags); extern int rpc_unlink(struct dentry *); diff --git a/net/sunrpc/cache.c b/net/sunrpc/cache.c index 1cd82eda56d0..db7720e453c3 100644 --- a/net/sunrpc/cache.c +++ b/net/sunrpc/cache.c @@ -32,6 +32,7 @@ #include #include #include +#include #define RPCDBG_FACILITY RPCDBG_CACHE @@ -1438,3 +1439,128 @@ void cache_unregister(struct cache_detail *cd) sunrpc_destroy_cache_detail(cd); } EXPORT_SYMBOL_GPL(cache_unregister); + +static ssize_t cache_read_pipefs(struct file *filp, char __user *buf, + size_t count, loff_t *ppos) +{ + struct cache_detail *cd = RPC_I(filp->f_path.dentry->d_inode)->private; + + return cache_read(filp, buf, count, ppos, cd); +} + +static ssize_t cache_write_pipefs(struct file *filp, const char __user *buf, + size_t count, loff_t *ppos) +{ + struct cache_detail *cd = RPC_I(filp->f_path.dentry->d_inode)->private; + + return cache_write(filp, buf, count, ppos, cd); +} + +static unsigned int cache_poll_pipefs(struct file *filp, poll_table *wait) +{ + struct cache_detail *cd = RPC_I(filp->f_path.dentry->d_inode)->private; + + return cache_poll(filp, wait, cd); +} + +static int cache_ioctl_pipefs(struct inode *inode, struct file *filp, + unsigned int cmd, unsigned long arg) +{ + struct cache_detail *cd = RPC_I(inode)->private; + + return cache_ioctl(inode, filp, cmd, arg, cd); +} + +static int cache_open_pipefs(struct inode *inode, struct file *filp) +{ + struct cache_detail *cd = RPC_I(inode)->private; + + return cache_open(inode, filp, cd); +} + +static int cache_release_pipefs(struct inode *inode, struct file *filp) +{ + struct cache_detail *cd = RPC_I(inode)->private; + + return cache_release(inode, filp, cd); +} + +const struct file_operations cache_file_operations_pipefs = { + .owner = THIS_MODULE, + .llseek = no_llseek, + .read = cache_read_pipefs, + .write = cache_write_pipefs, + .poll = cache_poll_pipefs, + .ioctl = cache_ioctl_pipefs, /* for FIONREAD */ + .open = cache_open_pipefs, + .release = cache_release_pipefs, +}; + +static int content_open_pipefs(struct inode *inode, struct file *filp) +{ + struct cache_detail *cd = RPC_I(inode)->private; + + return content_open(inode, filp, cd); +} + +const struct file_operations content_file_operations_pipefs = { + .open = content_open_pipefs, + .read = seq_read, + .llseek = seq_lseek, + .release = seq_release_private, +}; + +static ssize_t read_flush_pipefs(struct file *filp, char __user *buf, + size_t count, loff_t *ppos) +{ + struct cache_detail *cd = RPC_I(filp->f_path.dentry->d_inode)->private; + + return read_flush(filp, buf, count, ppos, cd); +} + +static ssize_t write_flush_pipefs(struct file *filp, + const char __user *buf, + size_t count, loff_t *ppos) +{ + struct cache_detail *cd = RPC_I(filp->f_path.dentry->d_inode)->private; + + return write_flush(filp, buf, count, ppos, cd); +} + +const struct file_operations cache_flush_operations_pipefs = { + .open = nonseekable_open, + .read = read_flush_pipefs, + .write = write_flush_pipefs, +}; + +int sunrpc_cache_register_pipefs(struct dentry *parent, + const char *name, mode_t umode, + struct cache_detail *cd) +{ + struct qstr q; + struct dentry *dir; + int ret = 0; + + sunrpc_init_cache_detail(cd); + q.name = name; + q.len = strlen(name); + q.hash = full_name_hash(q.name, q.len); + dir = rpc_create_cache_dir(parent, &q, umode, cd); + if (!IS_ERR(dir)) + cd->u.pipefs.dir = dir; + else { + sunrpc_destroy_cache_detail(cd); + ret = PTR_ERR(dir); + } + return ret; +} +EXPORT_SYMBOL_GPL(sunrpc_cache_register_pipefs); + +void sunrpc_cache_unregister_pipefs(struct cache_detail *cd) +{ + rpc_remove_cache_dir(cd->u.pipefs.dir); + cd->u.pipefs.dir = NULL; + sunrpc_destroy_cache_detail(cd); +} +EXPORT_SYMBOL_GPL(sunrpc_cache_unregister_pipefs); + diff --git a/net/sunrpc/rpc_pipe.c b/net/sunrpc/rpc_pipe.c index 57e9cd3c49b6..8dd81535e08f 100644 --- a/net/sunrpc/rpc_pipe.c +++ b/net/sunrpc/rpc_pipe.c @@ -26,6 +26,7 @@ #include #include #include +#include static struct vfsmount *rpc_mount __read_mostly; static int rpc_mount_count; @@ -882,6 +883,48 @@ int rpc_remove_client_dir(struct dentry *dentry) return rpc_rmdir_depopulate(dentry, rpc_clntdir_depopulate); } +static const struct rpc_filelist cache_pipefs_files[3] = { + [0] = { + .name = "channel", + .i_fop = &cache_file_operations_pipefs, + .mode = S_IFIFO|S_IRUSR|S_IWUSR, + }, + [1] = { + .name = "content", + .i_fop = &content_file_operations_pipefs, + .mode = S_IFREG|S_IRUSR, + }, + [2] = { + .name = "flush", + .i_fop = &cache_flush_operations_pipefs, + .mode = S_IFREG|S_IRUSR|S_IWUSR, + }, +}; + +static int rpc_cachedir_populate(struct dentry *dentry, void *private) +{ + return rpc_populate(dentry, + cache_pipefs_files, 0, 3, + private); +} + +static void rpc_cachedir_depopulate(struct dentry *dentry) +{ + rpc_depopulate(dentry, cache_pipefs_files, 0, 3); +} + +struct dentry *rpc_create_cache_dir(struct dentry *parent, struct qstr *name, + mode_t umode, struct cache_detail *cd) +{ + return rpc_mkdir_populate(parent, name, umode, NULL, + rpc_cachedir_populate, cd); +} + +void rpc_remove_cache_dir(struct dentry *dentry) +{ + rpc_rmdir_depopulate(dentry, rpc_cachedir_depopulate); +} + /* * populate the filesystem */ -- cgit v1.2.3 From 47d439e9fb8a81a90022cfa785bf1c36c4e2aff6 Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Fri, 7 Aug 2009 14:53:57 -0400 Subject: security: define round_hint_to_min in !CONFIG_SECURITY Fix the header files to define round_hint_to_min() and to define mmap_min_addr_handler() in the !CONFIG_SECURITY case. Built and tested with !CONFIG_SECURITY Signed-off-by: Eric Paris Signed-off-by: James Morris --- include/linux/security.h | 30 +++++++++++++++--------------- 1 file changed, 15 insertions(+), 15 deletions(-) (limited to 'include') diff --git a/include/linux/security.h b/include/linux/security.h index 7b431155e392..57ead99d2593 100644 --- a/include/linux/security.h +++ b/include/linux/security.h @@ -121,6 +121,21 @@ struct request_sock; #define LSM_UNSAFE_PTRACE 2 #define LSM_UNSAFE_PTRACE_CAP 4 +/* + * If a hint addr is less than mmap_min_addr change hint to be as + * low as possible but still greater than mmap_min_addr + */ +static inline unsigned long round_hint_to_min(unsigned long hint) +{ + hint &= PAGE_MASK; + if (((void *)hint != NULL) && + (hint < mmap_min_addr)) + return PAGE_ALIGN(mmap_min_addr); + return hint; +} +extern int mmap_min_addr_handler(struct ctl_table *table, int write, struct file *filp, + void __user *buffer, size_t *lenp, loff_t *ppos); + #ifdef CONFIG_SECURITY struct security_mnt_opts { @@ -149,21 +164,6 @@ static inline void security_free_mnt_opts(struct security_mnt_opts *opts) opts->num_mnt_opts = 0; } -/* - * If a hint addr is less than mmap_min_addr change hint to be as - * low as possible but still greater than mmap_min_addr - */ -static inline unsigned long round_hint_to_min(unsigned long hint) -{ - hint &= PAGE_MASK; - if (((void *)hint != NULL) && - (hint < mmap_min_addr)) - return PAGE_ALIGN(mmap_min_addr); - return hint; -} - -extern int mmap_min_addr_handler(struct ctl_table *table, int write, struct file *filp, - void __user *buffer, size_t *lenp, loff_t *ppos); /** * struct security_operations - main security structure * -- cgit v1.2.3 From a8ad568dd8ca122aa8048ea067d3599820d1c1b4 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Mon, 10 Aug 2009 11:53:10 +0900 Subject: dma-ops: Remove flush_write_buffers() in dma-mapping-common.h This moves flush_write_buffers() in asm-generic/dma-mapping-common.h to arch/x86/kernel/pci-nommu.c. The purpose of this patch is that, we can avoid defining NULL flush_write_buffers() on IA64 and SPARC. dma-mapping-common.h is used by X86 and IA64 (and SPARC soon) but only X86 with CONFIG_X86_OOSTORE or CONFIG_X86_PPRO_FENCE actually uses flush_write_buffers(). CONFIG_X86_OOSTORE or CONFIG_X86_PPRO_FENCE is usable with only kernel/pci-nommu.c (that is, not usable with other X86 IOMMU implementations such as SWIOTLB, VT-d, etc) so we can safely move flush_write_buffers() in asm-generic/dma-mapping-common.h to arch/x86/kernel/pci-nommu.c. The further discussion is: http://lkml.org/lkml/2009/6/28/104 Signed-off-by: Arnd Bergmann Acked-by: FUJITA Tomonori Cc: davem@davemloft.net Cc: tony.luck@intel.com Cc: fenghua.yu@intel.com LKML-Reference: <1249872797-1314-2-git-send-email-fujita.tomonori@lab.ntt.co.jp> Signed-off-by: Ingo Molnar --- arch/x86/kernel/pci-nommu.c | 27 ++++++++++++++++++++++----- include/asm-generic/dma-mapping-common.h | 6 ------ 2 files changed, 22 insertions(+), 11 deletions(-) (limited to 'include') diff --git a/arch/x86/kernel/pci-nommu.c b/arch/x86/kernel/pci-nommu.c index c0a4222bf62b..a3933d4330cd 100644 --- a/arch/x86/kernel/pci-nommu.c +++ b/arch/x86/kernel/pci-nommu.c @@ -79,12 +79,29 @@ static void nommu_free_coherent(struct device *dev, size_t size, void *vaddr, free_pages((unsigned long)vaddr, get_order(size)); } +static void nommu_sync_single_for_device(struct device *dev, + dma_addr_t addr, size_t size, + enum dma_data_direction dir) +{ + flush_write_buffers(); +} + + +static void nommu_sync_sg_for_device(struct device *dev, + struct scatterlist *sg, int nelems, + enum dma_data_direction dir) +{ + flush_write_buffers(); +} + struct dma_map_ops nommu_dma_ops = { - .alloc_coherent = dma_generic_alloc_coherent, - .free_coherent = nommu_free_coherent, - .map_sg = nommu_map_sg, - .map_page = nommu_map_page, - .is_phys = 1, + .alloc_coherent = dma_generic_alloc_coherent, + .free_coherent = nommu_free_coherent, + .map_sg = nommu_map_sg, + .map_page = nommu_map_page, + .sync_single_for_device = nommu_sync_single_for_device, + .sync_sg_for_device = nommu_sync_sg_for_device, + .is_phys = 1, }; void __init no_iommu_init(void) diff --git a/include/asm-generic/dma-mapping-common.h b/include/asm-generic/dma-mapping-common.h index 5406a601185c..e694263445f7 100644 --- a/include/asm-generic/dma-mapping-common.h +++ b/include/asm-generic/dma-mapping-common.h @@ -103,7 +103,6 @@ static inline void dma_sync_single_for_cpu(struct device *dev, dma_addr_t addr, if (ops->sync_single_for_cpu) ops->sync_single_for_cpu(dev, addr, size, dir); debug_dma_sync_single_for_cpu(dev, addr, size, dir); - flush_write_buffers(); } static inline void dma_sync_single_for_device(struct device *dev, @@ -116,7 +115,6 @@ static inline void dma_sync_single_for_device(struct device *dev, if (ops->sync_single_for_device) ops->sync_single_for_device(dev, addr, size, dir); debug_dma_sync_single_for_device(dev, addr, size, dir); - flush_write_buffers(); } static inline void dma_sync_single_range_for_cpu(struct device *dev, @@ -132,7 +130,6 @@ static inline void dma_sync_single_range_for_cpu(struct device *dev, ops->sync_single_range_for_cpu(dev, addr, offset, size, dir); debug_dma_sync_single_range_for_cpu(dev, addr, offset, size, dir); - flush_write_buffers(); } else dma_sync_single_for_cpu(dev, addr, size, dir); } @@ -150,7 +147,6 @@ static inline void dma_sync_single_range_for_device(struct device *dev, ops->sync_single_range_for_device(dev, addr, offset, size, dir); debug_dma_sync_single_range_for_device(dev, addr, offset, size, dir); - flush_write_buffers(); } else dma_sync_single_for_device(dev, addr, size, dir); } @@ -165,7 +161,6 @@ dma_sync_sg_for_cpu(struct device *dev, struct scatterlist *sg, if (ops->sync_sg_for_cpu) ops->sync_sg_for_cpu(dev, sg, nelems, dir); debug_dma_sync_sg_for_cpu(dev, sg, nelems, dir); - flush_write_buffers(); } static inline void @@ -179,7 +174,6 @@ dma_sync_sg_for_device(struct device *dev, struct scatterlist *sg, ops->sync_sg_for_device(dev, sg, nelems, dir); debug_dma_sync_sg_for_device(dev, sg, nelems, dir); - flush_write_buffers(); } #define dma_map_single(d, a, s, r) dma_map_single_attrs(d, a, s, r, NULL) -- cgit v1.2.3 From a044560c3a1f0ad75ce685c1ed7604820b9ed319 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Mon, 10 Aug 2009 11:16:52 +0200 Subject: perf_counter: Correct PERF_SAMPLE_RAW output PERF_SAMPLE_* output switches should unconditionally output the correct format, as they are the only way to unambiguously parse the PERF_EVENT_SAMPLE data. Signed-off-by: Peter Zijlstra Acked-by: Frederic Weisbecker Cc: Arnaldo Carvalho de Melo Cc: Mike Galbraith Cc: Paul Mackerras LKML-Reference: <1249896447.17467.74.camel@twins> Signed-off-by: Ingo Molnar --- include/linux/perf_counter.h | 2 ++ include/trace/ftrace.h | 3 ++- kernel/perf_counter.c | 30 ++++++++++++++++++++++++------ 3 files changed, 28 insertions(+), 7 deletions(-) (limited to 'include') diff --git a/include/linux/perf_counter.h b/include/linux/perf_counter.h index 2aabe43c1d04..a9d823a93fe8 100644 --- a/include/linux/perf_counter.h +++ b/include/linux/perf_counter.h @@ -369,6 +369,8 @@ enum perf_event_type { * * { u64 nr, * u64 ips[nr]; } && PERF_SAMPLE_CALLCHAIN + * { u32 size; + * char data[size];}&& PERF_SAMPLE_RAW * }; */ PERF_EVENT_SAMPLE = 9, diff --git a/include/trace/ftrace.h b/include/trace/ftrace.h index 7fb16d90e7b1..7167b9b97da2 100644 --- a/include/trace/ftrace.h +++ b/include/trace/ftrace.h @@ -685,7 +685,8 @@ static void ftrace_profile_##call(proto) \ pc = preempt_count(); \ \ __data_size = ftrace_get_offsets_##call(&__data_offsets, args); \ - __entry_size = ALIGN(__data_size + sizeof(*entry), sizeof(u64));\ + __entry_size = ALIGN(__data_size + sizeof(*entry) + sizeof(u32),\ + sizeof(u64)); \ \ do { \ char raw_data[__entry_size]; \ diff --git a/kernel/perf_counter.c b/kernel/perf_counter.c index 546e62d62941..5229d1666fa5 100644 --- a/kernel/perf_counter.c +++ b/kernel/perf_counter.c @@ -2646,7 +2646,6 @@ static void perf_counter_output(struct perf_counter *counter, int nmi, u64 counter; } group_entry; struct perf_callchain_entry *callchain = NULL; - struct perf_raw_record *raw = NULL; int callchain_size = 0; u64 time; struct { @@ -2716,9 +2715,15 @@ static void perf_counter_output(struct perf_counter *counter, int nmi, } if (sample_type & PERF_SAMPLE_RAW) { - raw = data->raw; - if (raw) - header.size += raw->size; + int size = sizeof(u32); + + if (data->raw) + size += data->raw->size; + else + size += sizeof(u32); + + WARN_ON_ONCE(size & (sizeof(u64)-1)); + header.size += size; } ret = perf_output_begin(&handle, counter, header.size, nmi, 1); @@ -2784,8 +2789,21 @@ static void perf_counter_output(struct perf_counter *counter, int nmi, } } - if ((sample_type & PERF_SAMPLE_RAW) && raw) - perf_output_copy(&handle, raw->data, raw->size); + if (sample_type & PERF_SAMPLE_RAW) { + if (data->raw) { + perf_output_put(&handle, data->raw->size); + perf_output_copy(&handle, data->raw->data, data->raw->size); + } else { + struct { + u32 size; + u32 data; + } raw = { + .size = sizeof(u32), + .data = 0, + }; + perf_output_put(&handle, raw); + } + } perf_output_end(&handle); } -- cgit v1.2.3 From 7cd1837b5d24417eca667d674a97bea936849785 Mon Sep 17 00:00:00 2001 From: Jan Engelhardt Date: Fri, 12 Jun 2009 18:36:33 +0200 Subject: netfilter: xtables: remove xt_TOS v0 Superseded by xt_TOS v1 (v2.6.24-2396-g5c350e5). Signed-off-by: Jan Engelhardt --- Documentation/feature-removal-schedule.txt | 3 -- include/linux/netfilter_ipv4/Kbuild | 2 -- include/linux/netfilter_ipv4/ipt_TOS.h | 12 -------- include/linux/netfilter_ipv4/ipt_tos.h | 13 --------- net/netfilter/xt_DSCP.c | 46 ------------------------------ net/netfilter/xt_dscp.c | 17 ----------- 6 files changed, 93 deletions(-) delete mode 100644 include/linux/netfilter_ipv4/ipt_TOS.h delete mode 100644 include/linux/netfilter_ipv4/ipt_tos.h (limited to 'include') diff --git a/Documentation/feature-removal-schedule.txt b/Documentation/feature-removal-schedule.txt index f8cd450be9aa..3aa4a779092b 100644 --- a/Documentation/feature-removal-schedule.txt +++ b/Documentation/feature-removal-schedule.txt @@ -235,9 +235,6 @@ Who: Thomas Gleixner --------------------------- What (Why): - - include/linux/netfilter_ipv4/ipt_TOS.h ipt_tos.h header files - (superseded by xt_TOS/xt_tos target & match) - - "forwarding" header files like ipt_mac.h in include/linux/netfilter_ipv4/ and include/linux/netfilter_ipv6/ diff --git a/include/linux/netfilter_ipv4/Kbuild b/include/linux/netfilter_ipv4/Kbuild index 3a7105bb8f33..86d81a285f9c 100644 --- a/include/linux/netfilter_ipv4/Kbuild +++ b/include/linux/netfilter_ipv4/Kbuild @@ -9,7 +9,6 @@ header-y += ipt_NFQUEUE.h header-y += ipt_REJECT.h header-y += ipt_SAME.h header-y += ipt_TCPMSS.h -header-y += ipt_TOS.h header-y += ipt_TTL.h header-y += ipt_ULOG.h header-y += ipt_addrtype.h @@ -40,7 +39,6 @@ header-y += ipt_sctp.h header-y += ipt_state.h header-y += ipt_string.h header-y += ipt_tcpmss.h -header-y += ipt_tos.h header-y += ipt_ttl.h unifdef-y += ip_queue.h diff --git a/include/linux/netfilter_ipv4/ipt_TOS.h b/include/linux/netfilter_ipv4/ipt_TOS.h deleted file mode 100644 index 6bf9e1fdfd88..000000000000 --- a/include/linux/netfilter_ipv4/ipt_TOS.h +++ /dev/null @@ -1,12 +0,0 @@ -#ifndef _IPT_TOS_H_target -#define _IPT_TOS_H_target - -#ifndef IPTOS_NORMALSVC -#define IPTOS_NORMALSVC 0 -#endif - -struct ipt_tos_target_info { - u_int8_t tos; -}; - -#endif /*_IPT_TOS_H_target*/ diff --git a/include/linux/netfilter_ipv4/ipt_tos.h b/include/linux/netfilter_ipv4/ipt_tos.h deleted file mode 100644 index a21f5df23c50..000000000000 --- a/include/linux/netfilter_ipv4/ipt_tos.h +++ /dev/null @@ -1,13 +0,0 @@ -#ifndef _IPT_TOS_H -#define _IPT_TOS_H - -struct ipt_tos_info { - u_int8_t tos; - u_int8_t invert; -}; - -#ifndef IPTOS_NORMALSVC -#define IPTOS_NORMALSVC 0 -#endif - -#endif /*_IPT_TOS_H*/ diff --git a/net/netfilter/xt_DSCP.c b/net/netfilter/xt_DSCP.c index 6a347e768f86..74ce89260056 100644 --- a/net/netfilter/xt_DSCP.c +++ b/net/netfilter/xt_DSCP.c @@ -18,7 +18,6 @@ #include #include -#include MODULE_AUTHOR("Harald Welte "); MODULE_DESCRIPTION("Xtables: DSCP/TOS field modification"); @@ -72,41 +71,6 @@ static bool dscp_tg_check(const struct xt_tgchk_param *par) return true; } -static unsigned int -tos_tg_v0(struct sk_buff *skb, const struct xt_target_param *par) -{ - const struct ipt_tos_target_info *info = par->targinfo; - struct iphdr *iph = ip_hdr(skb); - u_int8_t oldtos; - - if ((iph->tos & IPTOS_TOS_MASK) != info->tos) { - if (!skb_make_writable(skb, sizeof(struct iphdr))) - return NF_DROP; - - iph = ip_hdr(skb); - oldtos = iph->tos; - iph->tos = (iph->tos & IPTOS_PREC_MASK) | info->tos; - csum_replace2(&iph->check, htons(oldtos), htons(iph->tos)); - } - - return XT_CONTINUE; -} - -static bool tos_tg_check_v0(const struct xt_tgchk_param *par) -{ - const struct ipt_tos_target_info *info = par->targinfo; - const uint8_t tos = info->tos; - - if (tos != IPTOS_LOWDELAY && tos != IPTOS_THROUGHPUT && - tos != IPTOS_RELIABILITY && tos != IPTOS_MINCOST && - tos != IPTOS_NORMALSVC) { - printk(KERN_WARNING "TOS: bad tos value %#x\n", tos); - return false; - } - - return true; -} - static unsigned int tos_tg(struct sk_buff *skb, const struct xt_target_param *par) { @@ -166,16 +130,6 @@ static struct xt_target dscp_tg_reg[] __read_mostly = { .table = "mangle", .me = THIS_MODULE, }, - { - .name = "TOS", - .revision = 0, - .family = NFPROTO_IPV4, - .table = "mangle", - .target = tos_tg_v0, - .targetsize = sizeof(struct ipt_tos_target_info), - .checkentry = tos_tg_check_v0, - .me = THIS_MODULE, - }, { .name = "TOS", .revision = 1, diff --git a/net/netfilter/xt_dscp.c b/net/netfilter/xt_dscp.c index c3f8085460d7..0280d3a8c161 100644 --- a/net/netfilter/xt_dscp.c +++ b/net/netfilter/xt_dscp.c @@ -15,7 +15,6 @@ #include #include -#include MODULE_AUTHOR("Harald Welte "); MODULE_DESCRIPTION("Xtables: DSCP/TOS field match"); @@ -55,14 +54,6 @@ static bool dscp_mt_check(const struct xt_mtchk_param *par) return true; } -static bool -tos_mt_v0(const struct sk_buff *skb, const struct xt_match_param *par) -{ - const struct ipt_tos_info *info = par->matchinfo; - - return (ip_hdr(skb)->tos == info->tos) ^ info->invert; -} - static bool tos_mt(const struct sk_buff *skb, const struct xt_match_param *par) { const struct xt_tos_match_info *info = par->matchinfo; @@ -92,14 +83,6 @@ static struct xt_match dscp_mt_reg[] __read_mostly = { .matchsize = sizeof(struct xt_dscp_info), .me = THIS_MODULE, }, - { - .name = "tos", - .revision = 0, - .family = NFPROTO_IPV4, - .match = tos_mt_v0, - .matchsize = sizeof(struct ipt_tos_info), - .me = THIS_MODULE, - }, { .name = "tos", .revision = 1, -- cgit v1.2.3 From e973a70ca033bfcd4d8b59d1f66bfc1e782e1276 Mon Sep 17 00:00:00 2001 From: Jan Engelhardt Date: Fri, 12 Jun 2009 18:42:12 +0200 Subject: netfilter: xtables: remove xt_CONNMARK v0 Superseded by xt_CONNMARK v1 (v2.6.24-2917-g0dc8c76). Signed-off-by: Jan Engelhardt --- Documentation/feature-removal-schedule.txt | 3 - include/linux/netfilter/xt_CONNMARK.h | 6 -- net/netfilter/xt_CONNMARK.c | 134 +++-------------------------- 3 files changed, 11 insertions(+), 132 deletions(-) (limited to 'include') diff --git a/Documentation/feature-removal-schedule.txt b/Documentation/feature-removal-schedule.txt index 3aa4a779092b..7eccf945d4e0 100644 --- a/Documentation/feature-removal-schedule.txt +++ b/Documentation/feature-removal-schedule.txt @@ -238,9 +238,6 @@ What (Why): - "forwarding" header files like ipt_mac.h in include/linux/netfilter_ipv4/ and include/linux/netfilter_ipv6/ - - xt_CONNMARK match revision 0 - (superseded by xt_CONNMARK match revision 1) - - xt_MARK target revisions 0 and 1 (superseded by xt_MARK match revision 2) diff --git a/include/linux/netfilter/xt_CONNMARK.h b/include/linux/netfilter/xt_CONNMARK.h index 7635c8ffdadb..0a8545866752 100644 --- a/include/linux/netfilter/xt_CONNMARK.h +++ b/include/linux/netfilter/xt_CONNMARK.h @@ -18,12 +18,6 @@ enum { XT_CONNMARK_RESTORE }; -struct xt_connmark_target_info { - unsigned long mark; - unsigned long mask; - __u8 mode; -}; - struct xt_connmark_tginfo1 { __u32 ctmark, ctmask, nfmask; __u8 mode; diff --git a/net/netfilter/xt_CONNMARK.c b/net/netfilter/xt_CONNMARK.c index d6e5ab463277..593457068ae1 100644 --- a/net/netfilter/xt_CONNMARK.c +++ b/net/netfilter/xt_CONNMARK.c @@ -35,45 +35,6 @@ MODULE_ALIAS("ip6t_CONNMARK"); #include #include -static unsigned int -connmark_tg_v0(struct sk_buff *skb, const struct xt_target_param *par) -{ - const struct xt_connmark_target_info *markinfo = par->targinfo; - struct nf_conn *ct; - enum ip_conntrack_info ctinfo; - u_int32_t diff; - u_int32_t mark; - u_int32_t newmark; - - ct = nf_ct_get(skb, &ctinfo); - if (ct) { - switch(markinfo->mode) { - case XT_CONNMARK_SET: - newmark = (ct->mark & ~markinfo->mask) | markinfo->mark; - if (newmark != ct->mark) { - ct->mark = newmark; - nf_conntrack_event_cache(IPCT_MARK, ct); - } - break; - case XT_CONNMARK_SAVE: - newmark = (ct->mark & ~markinfo->mask) | - (skb->mark & markinfo->mask); - if (ct->mark != newmark) { - ct->mark = newmark; - nf_conntrack_event_cache(IPCT_MARK, ct); - } - break; - case XT_CONNMARK_RESTORE: - mark = skb->mark; - diff = (ct->mark ^ mark) & markinfo->mask; - skb->mark = mark ^ diff; - break; - } - } - - return XT_CONTINUE; -} - static unsigned int connmark_tg(struct sk_buff *skb, const struct xt_target_param *par) { @@ -112,30 +73,6 @@ connmark_tg(struct sk_buff *skb, const struct xt_target_param *par) return XT_CONTINUE; } -static bool connmark_tg_check_v0(const struct xt_tgchk_param *par) -{ - const struct xt_connmark_target_info *matchinfo = par->targinfo; - - if (matchinfo->mode == XT_CONNMARK_RESTORE) { - if (strcmp(par->table, "mangle") != 0) { - printk(KERN_WARNING "CONNMARK: restore can only be " - "called from \"mangle\" table, not \"%s\"\n", - par->table); - return false; - } - } - if (matchinfo->mark > 0xffffffff || matchinfo->mask > 0xffffffff) { - printk(KERN_WARNING "CONNMARK: Only supports 32bit mark\n"); - return false; - } - if (nf_ct_l3proto_try_module_get(par->family) < 0) { - printk(KERN_WARNING "can't load conntrack support for " - "proto=%u\n", par->family); - return false; - } - return true; -} - static bool connmark_tg_check(const struct xt_tgchk_param *par) { if (nf_ct_l3proto_try_module_get(par->family) < 0) { @@ -151,74 +88,25 @@ static void connmark_tg_destroy(const struct xt_tgdtor_param *par) nf_ct_l3proto_module_put(par->family); } -#ifdef CONFIG_COMPAT -struct compat_xt_connmark_target_info { - compat_ulong_t mark, mask; - u_int8_t mode; - u_int8_t __pad1; - u_int16_t __pad2; -}; - -static void connmark_tg_compat_from_user_v0(void *dst, void *src) -{ - const struct compat_xt_connmark_target_info *cm = src; - struct xt_connmark_target_info m = { - .mark = cm->mark, - .mask = cm->mask, - .mode = cm->mode, - }; - memcpy(dst, &m, sizeof(m)); -} - -static int connmark_tg_compat_to_user_v0(void __user *dst, void *src) -{ - const struct xt_connmark_target_info *m = src; - struct compat_xt_connmark_target_info cm = { - .mark = m->mark, - .mask = m->mask, - .mode = m->mode, - }; - return copy_to_user(dst, &cm, sizeof(cm)) ? -EFAULT : 0; -} -#endif /* CONFIG_COMPAT */ - -static struct xt_target connmark_tg_reg[] __read_mostly = { - { - .name = "CONNMARK", - .revision = 0, - .family = NFPROTO_UNSPEC, - .checkentry = connmark_tg_check_v0, - .destroy = connmark_tg_destroy, - .target = connmark_tg_v0, - .targetsize = sizeof(struct xt_connmark_target_info), -#ifdef CONFIG_COMPAT - .compatsize = sizeof(struct compat_xt_connmark_target_info), - .compat_from_user = connmark_tg_compat_from_user_v0, - .compat_to_user = connmark_tg_compat_to_user_v0, -#endif - .me = THIS_MODULE - }, - { - .name = "CONNMARK", - .revision = 1, - .family = NFPROTO_UNSPEC, - .checkentry = connmark_tg_check, - .target = connmark_tg, - .targetsize = sizeof(struct xt_connmark_tginfo1), - .destroy = connmark_tg_destroy, - .me = THIS_MODULE, - }, +static struct xt_target connmark_tg_reg __read_mostly = { + .name = "CONNMARK", + .revision = 1, + .family = NFPROTO_UNSPEC, + .checkentry = connmark_tg_check, + .target = connmark_tg, + .targetsize = sizeof(struct xt_connmark_tginfo1), + .destroy = connmark_tg_destroy, + .me = THIS_MODULE, }; static int __init connmark_tg_init(void) { - return xt_register_targets(connmark_tg_reg, - ARRAY_SIZE(connmark_tg_reg)); + return xt_register_target(&connmark_tg_reg); } static void __exit connmark_tg_exit(void) { - xt_unregister_targets(connmark_tg_reg, ARRAY_SIZE(connmark_tg_reg)); + xt_unregister_target(&connmark_tg_reg); } module_init(connmark_tg_init); -- cgit v1.2.3 From c8001f7fd5a4684280fddceed9fae9ea2e4fb521 Mon Sep 17 00:00:00 2001 From: Jan Engelhardt Date: Fri, 12 Jun 2009 18:47:32 +0200 Subject: netfilter: xtables: remove xt_MARK v0, v1 Superseded by xt_MARK v2 (v2.6.24-2918-ge0a812a). Signed-off-by: Jan Engelhardt --- Documentation/feature-removal-schedule.txt | 3 - include/linux/netfilter/xt_MARK.h | 17 --- net/netfilter/xt_MARK.c | 163 ++--------------------------- 3 files changed, 9 insertions(+), 174 deletions(-) (limited to 'include') diff --git a/Documentation/feature-removal-schedule.txt b/Documentation/feature-removal-schedule.txt index 7eccf945d4e0..121e19c9eee6 100644 --- a/Documentation/feature-removal-schedule.txt +++ b/Documentation/feature-removal-schedule.txt @@ -238,9 +238,6 @@ What (Why): - "forwarding" header files like ipt_mac.h in include/linux/netfilter_ipv4/ and include/linux/netfilter_ipv6/ - - xt_MARK target revisions 0 and 1 - (superseded by xt_MARK match revision 2) - - xt_connmark match revision 0 (superseded by xt_connmark match revision 1) diff --git a/include/linux/netfilter/xt_MARK.h b/include/linux/netfilter/xt_MARK.h index 028304bcc0b1..bc9561bdef79 100644 --- a/include/linux/netfilter/xt_MARK.h +++ b/include/linux/netfilter/xt_MARK.h @@ -3,23 +3,6 @@ #include -/* Version 0 */ -struct xt_mark_target_info { - unsigned long mark; -}; - -/* Version 1 */ -enum { - XT_MARK_SET=0, - XT_MARK_AND, - XT_MARK_OR, -}; - -struct xt_mark_target_info_v1 { - unsigned long mark; - __u8 mode; -}; - struct xt_mark_tginfo2 { __u32 mark, mask; }; diff --git a/net/netfilter/xt_MARK.c b/net/netfilter/xt_MARK.c index 67574bcfb8ac..225f8d11e173 100644 --- a/net/netfilter/xt_MARK.c +++ b/net/netfilter/xt_MARK.c @@ -24,39 +24,6 @@ MODULE_DESCRIPTION("Xtables: packet mark modification"); MODULE_ALIAS("ipt_MARK"); MODULE_ALIAS("ip6t_MARK"); -static unsigned int -mark_tg_v0(struct sk_buff *skb, const struct xt_target_param *par) -{ - const struct xt_mark_target_info *markinfo = par->targinfo; - - skb->mark = markinfo->mark; - return XT_CONTINUE; -} - -static unsigned int -mark_tg_v1(struct sk_buff *skb, const struct xt_target_param *par) -{ - const struct xt_mark_target_info_v1 *markinfo = par->targinfo; - int mark = 0; - - switch (markinfo->mode) { - case XT_MARK_SET: - mark = markinfo->mark; - break; - - case XT_MARK_AND: - mark = skb->mark & markinfo->mark; - break; - - case XT_MARK_OR: - mark = skb->mark | markinfo->mark; - break; - } - - skb->mark = mark; - return XT_CONTINUE; -} - static unsigned int mark_tg(struct sk_buff *skb, const struct xt_target_param *par) { @@ -66,135 +33,23 @@ mark_tg(struct sk_buff *skb, const struct xt_target_param *par) return XT_CONTINUE; } -static bool mark_tg_check_v0(const struct xt_tgchk_param *par) -{ - const struct xt_mark_target_info *markinfo = par->targinfo; - - if (markinfo->mark > 0xffffffff) { - printk(KERN_WARNING "MARK: Only supports 32bit wide mark\n"); - return false; - } - return true; -} - -static bool mark_tg_check_v1(const struct xt_tgchk_param *par) -{ - const struct xt_mark_target_info_v1 *markinfo = par->targinfo; - - if (markinfo->mode != XT_MARK_SET - && markinfo->mode != XT_MARK_AND - && markinfo->mode != XT_MARK_OR) { - printk(KERN_WARNING "MARK: unknown mode %u\n", - markinfo->mode); - return false; - } - if (markinfo->mark > 0xffffffff) { - printk(KERN_WARNING "MARK: Only supports 32bit wide mark\n"); - return false; - } - return true; -} - -#ifdef CONFIG_COMPAT -struct compat_xt_mark_target_info { - compat_ulong_t mark; -}; - -static void mark_tg_compat_from_user_v0(void *dst, void *src) -{ - const struct compat_xt_mark_target_info *cm = src; - struct xt_mark_target_info m = { - .mark = cm->mark, - }; - memcpy(dst, &m, sizeof(m)); -} - -static int mark_tg_compat_to_user_v0(void __user *dst, void *src) -{ - const struct xt_mark_target_info *m = src; - struct compat_xt_mark_target_info cm = { - .mark = m->mark, - }; - return copy_to_user(dst, &cm, sizeof(cm)) ? -EFAULT : 0; -} - -struct compat_xt_mark_target_info_v1 { - compat_ulong_t mark; - u_int8_t mode; - u_int8_t __pad1; - u_int16_t __pad2; -}; - -static void mark_tg_compat_from_user_v1(void *dst, void *src) -{ - const struct compat_xt_mark_target_info_v1 *cm = src; - struct xt_mark_target_info_v1 m = { - .mark = cm->mark, - .mode = cm->mode, - }; - memcpy(dst, &m, sizeof(m)); -} - -static int mark_tg_compat_to_user_v1(void __user *dst, void *src) -{ - const struct xt_mark_target_info_v1 *m = src; - struct compat_xt_mark_target_info_v1 cm = { - .mark = m->mark, - .mode = m->mode, - }; - return copy_to_user(dst, &cm, sizeof(cm)) ? -EFAULT : 0; -} -#endif /* CONFIG_COMPAT */ - -static struct xt_target mark_tg_reg[] __read_mostly = { - { - .name = "MARK", - .family = NFPROTO_UNSPEC, - .revision = 0, - .checkentry = mark_tg_check_v0, - .target = mark_tg_v0, - .targetsize = sizeof(struct xt_mark_target_info), -#ifdef CONFIG_COMPAT - .compatsize = sizeof(struct compat_xt_mark_target_info), - .compat_from_user = mark_tg_compat_from_user_v0, - .compat_to_user = mark_tg_compat_to_user_v0, -#endif - .table = "mangle", - .me = THIS_MODULE, - }, - { - .name = "MARK", - .family = NFPROTO_UNSPEC, - .revision = 1, - .checkentry = mark_tg_check_v1, - .target = mark_tg_v1, - .targetsize = sizeof(struct xt_mark_target_info_v1), -#ifdef CONFIG_COMPAT - .compatsize = sizeof(struct compat_xt_mark_target_info_v1), - .compat_from_user = mark_tg_compat_from_user_v1, - .compat_to_user = mark_tg_compat_to_user_v1, -#endif - .table = "mangle", - .me = THIS_MODULE, - }, - { - .name = "MARK", - .revision = 2, - .family = NFPROTO_UNSPEC, - .target = mark_tg, - .targetsize = sizeof(struct xt_mark_tginfo2), - .me = THIS_MODULE, - }, +static struct xt_target mark_tg_reg __read_mostly = { + .name = "MARK", + .revision = 2, + .family = NFPROTO_UNSPEC, + .target = mark_tg, + .targetsize = sizeof(struct xt_mark_tginfo2), + .me = THIS_MODULE, }; static int __init mark_tg_init(void) { - return xt_register_targets(mark_tg_reg, ARRAY_SIZE(mark_tg_reg)); + return xt_register_target(&mark_tg_reg); } static void __exit mark_tg_exit(void) { - xt_unregister_targets(mark_tg_reg, ARRAY_SIZE(mark_tg_reg)); + xt_unregister_target(&mark_tg_reg); } module_init(mark_tg_init); -- cgit v1.2.3 From 84899a2b9adaf6c2e20d198d7c24562ce6b391d8 Mon Sep 17 00:00:00 2001 From: Jan Engelhardt Date: Fri, 12 Jun 2009 18:50:33 +0200 Subject: netfilter: xtables: remove xt_connmark v0 Superseded by xt_connmark v1 (v2.6.24-2919-g96e3227). Signed-off-by: Jan Engelhardt --- Documentation/feature-removal-schedule.txt | 3 - include/linux/netfilter/xt_connmark.h | 5 -- net/netfilter/xt_connmark.c | 101 ++++------------------------- 3 files changed, 11 insertions(+), 98 deletions(-) (limited to 'include') diff --git a/Documentation/feature-removal-schedule.txt b/Documentation/feature-removal-schedule.txt index 121e19c9eee6..54f935794922 100644 --- a/Documentation/feature-removal-schedule.txt +++ b/Documentation/feature-removal-schedule.txt @@ -238,9 +238,6 @@ What (Why): - "forwarding" header files like ipt_mac.h in include/linux/netfilter_ipv4/ and include/linux/netfilter_ipv6/ - - xt_connmark match revision 0 - (superseded by xt_connmark match revision 1) - - xt_conntrack match revision 0 (superseded by xt_conntrack match revision 1) diff --git a/include/linux/netfilter/xt_connmark.h b/include/linux/netfilter/xt_connmark.h index 571e266d004c..619e47cde01a 100644 --- a/include/linux/netfilter/xt_connmark.h +++ b/include/linux/netfilter/xt_connmark.h @@ -12,11 +12,6 @@ * (at your option) any later version. */ -struct xt_connmark_info { - unsigned long mark, mask; - __u8 invert; -}; - struct xt_connmark_mtinfo1 { __u32 mark, mask; __u8 invert; diff --git a/net/netfilter/xt_connmark.c b/net/netfilter/xt_connmark.c index 86cacab7a4a3..122aa8b0147b 100644 --- a/net/netfilter/xt_connmark.c +++ b/net/netfilter/xt_connmark.c @@ -47,36 +47,6 @@ connmark_mt(const struct sk_buff *skb, const struct xt_match_param *par) return ((ct->mark & info->mask) == info->mark) ^ info->invert; } -static bool -connmark_mt_v0(const struct sk_buff *skb, const struct xt_match_param *par) -{ - const struct xt_connmark_info *info = par->matchinfo; - const struct nf_conn *ct; - enum ip_conntrack_info ctinfo; - - ct = nf_ct_get(skb, &ctinfo); - if (!ct) - return false; - - return ((ct->mark & info->mask) == info->mark) ^ info->invert; -} - -static bool connmark_mt_check_v0(const struct xt_mtchk_param *par) -{ - const struct xt_connmark_info *cm = par->matchinfo; - - if (cm->mark > 0xffffffff || cm->mask > 0xffffffff) { - printk(KERN_WARNING "connmark: only support 32bit mark\n"); - return false; - } - if (nf_ct_l3proto_try_module_get(par->family) < 0) { - printk(KERN_WARNING "can't load conntrack support for " - "proto=%u\n", par->family); - return false; - } - return true; -} - static bool connmark_mt_check(const struct xt_mtchk_param *par) { if (nf_ct_l3proto_try_module_get(par->family) < 0) { @@ -92,74 +62,25 @@ static void connmark_mt_destroy(const struct xt_mtdtor_param *par) nf_ct_l3proto_module_put(par->family); } -#ifdef CONFIG_COMPAT -struct compat_xt_connmark_info { - compat_ulong_t mark, mask; - u_int8_t invert; - u_int8_t __pad1; - u_int16_t __pad2; -}; - -static void connmark_mt_compat_from_user_v0(void *dst, void *src) -{ - const struct compat_xt_connmark_info *cm = src; - struct xt_connmark_info m = { - .mark = cm->mark, - .mask = cm->mask, - .invert = cm->invert, - }; - memcpy(dst, &m, sizeof(m)); -} - -static int connmark_mt_compat_to_user_v0(void __user *dst, void *src) -{ - const struct xt_connmark_info *m = src; - struct compat_xt_connmark_info cm = { - .mark = m->mark, - .mask = m->mask, - .invert = m->invert, - }; - return copy_to_user(dst, &cm, sizeof(cm)) ? -EFAULT : 0; -} -#endif /* CONFIG_COMPAT */ - -static struct xt_match connmark_mt_reg[] __read_mostly = { - { - .name = "connmark", - .revision = 0, - .family = NFPROTO_UNSPEC, - .checkentry = connmark_mt_check_v0, - .match = connmark_mt_v0, - .destroy = connmark_mt_destroy, - .matchsize = sizeof(struct xt_connmark_info), -#ifdef CONFIG_COMPAT - .compatsize = sizeof(struct compat_xt_connmark_info), - .compat_from_user = connmark_mt_compat_from_user_v0, - .compat_to_user = connmark_mt_compat_to_user_v0, -#endif - .me = THIS_MODULE - }, - { - .name = "connmark", - .revision = 1, - .family = NFPROTO_UNSPEC, - .checkentry = connmark_mt_check, - .match = connmark_mt, - .matchsize = sizeof(struct xt_connmark_mtinfo1), - .destroy = connmark_mt_destroy, - .me = THIS_MODULE, - }, +static struct xt_match connmark_mt_reg __read_mostly = { + .name = "connmark", + .revision = 1, + .family = NFPROTO_UNSPEC, + .checkentry = connmark_mt_check, + .match = connmark_mt, + .matchsize = sizeof(struct xt_connmark_mtinfo1), + .destroy = connmark_mt_destroy, + .me = THIS_MODULE, }; static int __init connmark_mt_init(void) { - return xt_register_matches(connmark_mt_reg, - ARRAY_SIZE(connmark_mt_reg)); + return xt_register_match(&connmark_mt_reg); } static void __exit connmark_mt_exit(void) { - xt_unregister_matches(connmark_mt_reg, ARRAY_SIZE(connmark_mt_reg)); + xt_unregister_match(&connmark_mt_reg); } module_init(connmark_mt_init); -- cgit v1.2.3 From 9e05ec4b1804a1ba51f61fe169aef9b86edcd3f7 Mon Sep 17 00:00:00 2001 From: Jan Engelhardt Date: Fri, 12 Jun 2009 18:56:14 +0200 Subject: netfilter: xtables: remove xt_conntrack v0 Superseded by xt_conntrack v1 (v2.6.24-2921-g64eb12f). Signed-off-by: Jan Engelhardt --- Documentation/feature-removal-schedule.txt | 3 - include/linux/netfilter/xt_conntrack.h | 36 ------- net/netfilter/xt_conntrack.c | 155 +---------------------------- 3 files changed, 1 insertion(+), 193 deletions(-) (limited to 'include') diff --git a/Documentation/feature-removal-schedule.txt b/Documentation/feature-removal-schedule.txt index 54f935794922..6746473ef033 100644 --- a/Documentation/feature-removal-schedule.txt +++ b/Documentation/feature-removal-schedule.txt @@ -238,9 +238,6 @@ What (Why): - "forwarding" header files like ipt_mac.h in include/linux/netfilter_ipv4/ and include/linux/netfilter_ipv6/ - - xt_conntrack match revision 0 - (superseded by xt_conntrack match revision 1) - - xt_iprange match revision 0, include/linux/netfilter_ipv4/ipt_iprange.h (superseded by xt_iprange match revision 1) diff --git a/include/linux/netfilter/xt_conntrack.h b/include/linux/netfilter/xt_conntrack.h index 7ae05338e94c..54f47a2f6152 100644 --- a/include/linux/netfilter/xt_conntrack.h +++ b/include/linux/netfilter/xt_conntrack.h @@ -32,42 +32,6 @@ enum { XT_CONNTRACK_DIRECTION = 1 << 12, }; -/* This is exposed to userspace, so remains frozen in time. */ -struct ip_conntrack_old_tuple -{ - struct { - __be32 ip; - union { - __u16 all; - } u; - } src; - - struct { - __be32 ip; - union { - __u16 all; - } u; - - /* The protocol. */ - __u16 protonum; - } dst; -}; - -struct xt_conntrack_info -{ - unsigned int statemask, statusmask; - - struct ip_conntrack_old_tuple tuple[IP_CT_DIR_MAX]; - struct in_addr sipmsk[IP_CT_DIR_MAX], dipmsk[IP_CT_DIR_MAX]; - - unsigned long expires_min, expires_max; - - /* Flags word */ - __u8 flags; - /* Inverse flags */ - __u8 invflags; -}; - struct xt_conntrack_mtinfo1 { union nf_inet_addr origsrc_addr, origsrc_mask; union nf_inet_addr origdst_addr, origdst_mask; diff --git a/net/netfilter/xt_conntrack.c b/net/netfilter/xt_conntrack.c index fc581800698e..6dc4652f2fe8 100644 --- a/net/netfilter/xt_conntrack.c +++ b/net/netfilter/xt_conntrack.c @@ -19,100 +19,11 @@ MODULE_LICENSE("GPL"); MODULE_AUTHOR("Marc Boucher "); -MODULE_AUTHOR("Jan Engelhardt "); +MODULE_AUTHOR("Jan Engelhardt "); MODULE_DESCRIPTION("Xtables: connection tracking state match"); MODULE_ALIAS("ipt_conntrack"); MODULE_ALIAS("ip6t_conntrack"); -static bool -conntrack_mt_v0(const struct sk_buff *skb, const struct xt_match_param *par) -{ - const struct xt_conntrack_info *sinfo = par->matchinfo; - const struct nf_conn *ct; - enum ip_conntrack_info ctinfo; - unsigned int statebit; - - ct = nf_ct_get(skb, &ctinfo); - -#define FWINV(bool, invflg) ((bool) ^ !!(sinfo->invflags & (invflg))) - - if (ct == &nf_conntrack_untracked) - statebit = XT_CONNTRACK_STATE_UNTRACKED; - else if (ct) - statebit = XT_CONNTRACK_STATE_BIT(ctinfo); - else - statebit = XT_CONNTRACK_STATE_INVALID; - - if (sinfo->flags & XT_CONNTRACK_STATE) { - if (ct) { - if (test_bit(IPS_SRC_NAT_BIT, &ct->status)) - statebit |= XT_CONNTRACK_STATE_SNAT; - if (test_bit(IPS_DST_NAT_BIT, &ct->status)) - statebit |= XT_CONNTRACK_STATE_DNAT; - } - if (FWINV((statebit & sinfo->statemask) == 0, - XT_CONNTRACK_STATE)) - return false; - } - - if (ct == NULL) { - if (sinfo->flags & ~XT_CONNTRACK_STATE) - return false; - return true; - } - - if (sinfo->flags & XT_CONNTRACK_PROTO && - FWINV(nf_ct_protonum(ct) != - sinfo->tuple[IP_CT_DIR_ORIGINAL].dst.protonum, - XT_CONNTRACK_PROTO)) - return false; - - if (sinfo->flags & XT_CONNTRACK_ORIGSRC && - FWINV((ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple.src.u3.ip & - sinfo->sipmsk[IP_CT_DIR_ORIGINAL].s_addr) != - sinfo->tuple[IP_CT_DIR_ORIGINAL].src.ip, - XT_CONNTRACK_ORIGSRC)) - return false; - - if (sinfo->flags & XT_CONNTRACK_ORIGDST && - FWINV((ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple.dst.u3.ip & - sinfo->dipmsk[IP_CT_DIR_ORIGINAL].s_addr) != - sinfo->tuple[IP_CT_DIR_ORIGINAL].dst.ip, - XT_CONNTRACK_ORIGDST)) - return false; - - if (sinfo->flags & XT_CONNTRACK_REPLSRC && - FWINV((ct->tuplehash[IP_CT_DIR_REPLY].tuple.src.u3.ip & - sinfo->sipmsk[IP_CT_DIR_REPLY].s_addr) != - sinfo->tuple[IP_CT_DIR_REPLY].src.ip, - XT_CONNTRACK_REPLSRC)) - return false; - - if (sinfo->flags & XT_CONNTRACK_REPLDST && - FWINV((ct->tuplehash[IP_CT_DIR_REPLY].tuple.dst.u3.ip & - sinfo->dipmsk[IP_CT_DIR_REPLY].s_addr) != - sinfo->tuple[IP_CT_DIR_REPLY].dst.ip, - XT_CONNTRACK_REPLDST)) - return false; - - if (sinfo->flags & XT_CONNTRACK_STATUS && - FWINV((ct->status & sinfo->statusmask) == 0, - XT_CONNTRACK_STATUS)) - return false; - - if(sinfo->flags & XT_CONNTRACK_EXPIRES) { - unsigned long expires = timer_pending(&ct->timeout) ? - (ct->timeout.expires - jiffies)/HZ : 0; - - if (FWINV(!(expires >= sinfo->expires_min && - expires <= sinfo->expires_max), - XT_CONNTRACK_EXPIRES)) - return false; - } - return true; -#undef FWINV -} - static bool conntrack_addrcmp(const union nf_inet_addr *kaddr, const union nf_inet_addr *uaddr, @@ -337,71 +248,7 @@ static void conntrack_mt_destroy_v1(const struct xt_mtdtor_param *par) conntrack_mt_destroy(par); } -#ifdef CONFIG_COMPAT -struct compat_xt_conntrack_info -{ - compat_uint_t statemask; - compat_uint_t statusmask; - struct ip_conntrack_old_tuple tuple[IP_CT_DIR_MAX]; - struct in_addr sipmsk[IP_CT_DIR_MAX]; - struct in_addr dipmsk[IP_CT_DIR_MAX]; - compat_ulong_t expires_min; - compat_ulong_t expires_max; - u_int8_t flags; - u_int8_t invflags; -}; - -static void conntrack_mt_compat_from_user_v0(void *dst, void *src) -{ - const struct compat_xt_conntrack_info *cm = src; - struct xt_conntrack_info m = { - .statemask = cm->statemask, - .statusmask = cm->statusmask, - .expires_min = cm->expires_min, - .expires_max = cm->expires_max, - .flags = cm->flags, - .invflags = cm->invflags, - }; - memcpy(m.tuple, cm->tuple, sizeof(m.tuple)); - memcpy(m.sipmsk, cm->sipmsk, sizeof(m.sipmsk)); - memcpy(m.dipmsk, cm->dipmsk, sizeof(m.dipmsk)); - memcpy(dst, &m, sizeof(m)); -} - -static int conntrack_mt_compat_to_user_v0(void __user *dst, void *src) -{ - const struct xt_conntrack_info *m = src; - struct compat_xt_conntrack_info cm = { - .statemask = m->statemask, - .statusmask = m->statusmask, - .expires_min = m->expires_min, - .expires_max = m->expires_max, - .flags = m->flags, - .invflags = m->invflags, - }; - memcpy(cm.tuple, m->tuple, sizeof(cm.tuple)); - memcpy(cm.sipmsk, m->sipmsk, sizeof(cm.sipmsk)); - memcpy(cm.dipmsk, m->dipmsk, sizeof(cm.dipmsk)); - return copy_to_user(dst, &cm, sizeof(cm)) ? -EFAULT : 0; -} -#endif - static struct xt_match conntrack_mt_reg[] __read_mostly = { - { - .name = "conntrack", - .revision = 0, - .family = NFPROTO_IPV4, - .match = conntrack_mt_v0, - .checkentry = conntrack_mt_check, - .destroy = conntrack_mt_destroy, - .matchsize = sizeof(struct xt_conntrack_info), - .me = THIS_MODULE, -#ifdef CONFIG_COMPAT - .compatsize = sizeof(struct compat_xt_conntrack_info), - .compat_from_user = conntrack_mt_compat_from_user_v0, - .compat_to_user = conntrack_mt_compat_to_user_v0, -#endif - }, { .name = "conntrack", .revision = 1, -- cgit v1.2.3 From 36d4084dc8eb7a9a3655a2041097a46aff3061e9 Mon Sep 17 00:00:00 2001 From: Jan Engelhardt Date: Fri, 12 Jun 2009 18:58:19 +0200 Subject: netfilter: xtables: remove xt_iprange v0 Superseded by xt_iprange v1 (v2.6.24-2928-g1a50c5a1). Signed-off-by: Jan Engelhardt --- Documentation/feature-removal-schedule.txt | 4 --- include/linux/netfilter_ipv4/Kbuild | 1 - include/linux/netfilter_ipv4/ipt_iprange.h | 21 -------------- net/netfilter/xt_iprange.c | 45 ++---------------------------- 4 files changed, 2 insertions(+), 69 deletions(-) delete mode 100644 include/linux/netfilter_ipv4/ipt_iprange.h (limited to 'include') diff --git a/Documentation/feature-removal-schedule.txt b/Documentation/feature-removal-schedule.txt index 6746473ef033..8862b037f0ac 100644 --- a/Documentation/feature-removal-schedule.txt +++ b/Documentation/feature-removal-schedule.txt @@ -238,10 +238,6 @@ What (Why): - "forwarding" header files like ipt_mac.h in include/linux/netfilter_ipv4/ and include/linux/netfilter_ipv6/ - - xt_iprange match revision 0, - include/linux/netfilter_ipv4/ipt_iprange.h - (superseded by xt_iprange match revision 1) - - xt_mark match revision 0 (superseded by xt_mark match revision 1) diff --git a/include/linux/netfilter_ipv4/Kbuild b/include/linux/netfilter_ipv4/Kbuild index 86d81a285f9c..5e361ef44e87 100644 --- a/include/linux/netfilter_ipv4/Kbuild +++ b/include/linux/netfilter_ipv4/Kbuild @@ -23,7 +23,6 @@ header-y += ipt_ecn.h header-y += ipt_esp.h header-y += ipt_hashlimit.h header-y += ipt_helper.h -header-y += ipt_iprange.h header-y += ipt_length.h header-y += ipt_limit.h header-y += ipt_mac.h diff --git a/include/linux/netfilter_ipv4/ipt_iprange.h b/include/linux/netfilter_ipv4/ipt_iprange.h deleted file mode 100644 index 5f1aebde4d2f..000000000000 --- a/include/linux/netfilter_ipv4/ipt_iprange.h +++ /dev/null @@ -1,21 +0,0 @@ -#ifndef _IPT_IPRANGE_H -#define _IPT_IPRANGE_H - -#include -#include - -struct ipt_iprange { - /* Inclusive: network order. */ - __be32 min_ip, max_ip; -}; - -struct ipt_iprange_info -{ - struct ipt_iprange src; - struct ipt_iprange dst; - - /* Flags from above */ - u_int8_t flags; -}; - -#endif /* _IPT_IPRANGE_H */ diff --git a/net/netfilter/xt_iprange.c b/net/netfilter/xt_iprange.c index 501f9b623188..ffc96387d556 100644 --- a/net/netfilter/xt_iprange.c +++ b/net/netfilter/xt_iprange.c @@ -14,40 +14,6 @@ #include #include #include -#include - -static bool -iprange_mt_v0(const struct sk_buff *skb, const struct xt_match_param *par) -{ - const struct ipt_iprange_info *info = par->matchinfo; - const struct iphdr *iph = ip_hdr(skb); - - if (info->flags & IPRANGE_SRC) { - if ((ntohl(iph->saddr) < ntohl(info->src.min_ip) - || ntohl(iph->saddr) > ntohl(info->src.max_ip)) - ^ !!(info->flags & IPRANGE_SRC_INV)) { - pr_debug("src IP %pI4 NOT in range %s%pI4-%pI4\n", - &iph->saddr, - info->flags & IPRANGE_SRC_INV ? "(INV) " : "", - &info->src.min_ip, - &info->src.max_ip); - return false; - } - } - if (info->flags & IPRANGE_DST) { - if ((ntohl(iph->daddr) < ntohl(info->dst.min_ip) - || ntohl(iph->daddr) > ntohl(info->dst.max_ip)) - ^ !!(info->flags & IPRANGE_DST_INV)) { - pr_debug("dst IP %pI4 NOT in range %s%pI4-%pI4\n", - &iph->daddr, - info->flags & IPRANGE_DST_INV ? "(INV) " : "", - &info->dst.min_ip, - &info->dst.max_ip); - return false; - } - } - return true; -} static bool iprange_mt4(const struct sk_buff *skb, const struct xt_match_param *par) @@ -125,14 +91,6 @@ iprange_mt6(const struct sk_buff *skb, const struct xt_match_param *par) } static struct xt_match iprange_mt_reg[] __read_mostly = { - { - .name = "iprange", - .revision = 0, - .family = NFPROTO_IPV4, - .match = iprange_mt_v0, - .matchsize = sizeof(struct ipt_iprange_info), - .me = THIS_MODULE, - }, { .name = "iprange", .revision = 1, @@ -164,7 +122,8 @@ static void __exit iprange_mt_exit(void) module_init(iprange_mt_init); module_exit(iprange_mt_exit); MODULE_LICENSE("GPL"); -MODULE_AUTHOR("Jozsef Kadlecsik , Jan Engelhardt "); +MODULE_AUTHOR("Jozsef Kadlecsik "); +MODULE_AUTHOR("Jan Engelhardt "); MODULE_DESCRIPTION("Xtables: arbitrary IPv4 range matching"); MODULE_ALIAS("ipt_iprange"); MODULE_ALIAS("ip6t_iprange"); -- cgit v1.2.3 From 4725c7287ef2c4340cb433f59e40d143c1f43c22 Mon Sep 17 00:00:00 2001 From: Jan Engelhardt Date: Fri, 12 Jun 2009 19:02:27 +0200 Subject: netfilter: xtables: remove xt_mark v0 Superseded by xt_mark v1 (v2.6.24-2922-g17b0d7e). Signed-off-by: Jan Engelhardt --- Documentation/feature-removal-schedule.txt | 3 -- include/linux/netfilter/xt_mark.h | 5 -- net/netfilter/xt_mark.c | 86 ++++-------------------------- 3 files changed, 10 insertions(+), 84 deletions(-) (limited to 'include') diff --git a/Documentation/feature-removal-schedule.txt b/Documentation/feature-removal-schedule.txt index 8862b037f0ac..5556d2300bea 100644 --- a/Documentation/feature-removal-schedule.txt +++ b/Documentation/feature-removal-schedule.txt @@ -238,9 +238,6 @@ What (Why): - "forwarding" header files like ipt_mac.h in include/linux/netfilter_ipv4/ and include/linux/netfilter_ipv6/ - - xt_mark match revision 0 - (superseded by xt_mark match revision 1) - - xt_recent: the old ipt_recent proc dir (superseded by /proc/net/xt_recent) diff --git a/include/linux/netfilter/xt_mark.h b/include/linux/netfilter/xt_mark.h index 6fa460a3cc29..6607c8f38ea5 100644 --- a/include/linux/netfilter/xt_mark.h +++ b/include/linux/netfilter/xt_mark.h @@ -3,11 +3,6 @@ #include -struct xt_mark_info { - unsigned long mark, mask; - __u8 invert; -}; - struct xt_mark_mtinfo1 { __u32 mark, mask; __u8 invert; diff --git a/net/netfilter/xt_mark.c b/net/netfilter/xt_mark.c index 10b9e34bbc5b..1db07d8125f8 100644 --- a/net/netfilter/xt_mark.c +++ b/net/netfilter/xt_mark.c @@ -3,7 +3,7 @@ * * (C) 1999-2001 Marc Boucher * Copyright © CC Computer Consultants GmbH, 2007 - 2008 - * Jan Engelhardt + * Jan Engelhardt * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License version 2 as @@ -22,14 +22,6 @@ MODULE_DESCRIPTION("Xtables: packet mark match"); MODULE_ALIAS("ipt_mark"); MODULE_ALIAS("ip6t_mark"); -static bool -mark_mt_v0(const struct sk_buff *skb, const struct xt_match_param *par) -{ - const struct xt_mark_info *info = par->matchinfo; - - return ((skb->mark & info->mask) == info->mark) ^ info->invert; -} - static bool mark_mt(const struct sk_buff *skb, const struct xt_match_param *par) { @@ -38,81 +30,23 @@ mark_mt(const struct sk_buff *skb, const struct xt_match_param *par) return ((skb->mark & info->mask) == info->mark) ^ info->invert; } -static bool mark_mt_check_v0(const struct xt_mtchk_param *par) -{ - const struct xt_mark_info *minfo = par->matchinfo; - - if (minfo->mark > 0xffffffff || minfo->mask > 0xffffffff) { - printk(KERN_WARNING "mark: only supports 32bit mark\n"); - return false; - } - return true; -} - -#ifdef CONFIG_COMPAT -struct compat_xt_mark_info { - compat_ulong_t mark, mask; - u_int8_t invert; - u_int8_t __pad1; - u_int16_t __pad2; -}; - -static void mark_mt_compat_from_user_v0(void *dst, void *src) -{ - const struct compat_xt_mark_info *cm = src; - struct xt_mark_info m = { - .mark = cm->mark, - .mask = cm->mask, - .invert = cm->invert, - }; - memcpy(dst, &m, sizeof(m)); -} - -static int mark_mt_compat_to_user_v0(void __user *dst, void *src) -{ - const struct xt_mark_info *m = src; - struct compat_xt_mark_info cm = { - .mark = m->mark, - .mask = m->mask, - .invert = m->invert, - }; - return copy_to_user(dst, &cm, sizeof(cm)) ? -EFAULT : 0; -} -#endif /* CONFIG_COMPAT */ - -static struct xt_match mark_mt_reg[] __read_mostly = { - { - .name = "mark", - .revision = 0, - .family = NFPROTO_UNSPEC, - .checkentry = mark_mt_check_v0, - .match = mark_mt_v0, - .matchsize = sizeof(struct xt_mark_info), -#ifdef CONFIG_COMPAT - .compatsize = sizeof(struct compat_xt_mark_info), - .compat_from_user = mark_mt_compat_from_user_v0, - .compat_to_user = mark_mt_compat_to_user_v0, -#endif - .me = THIS_MODULE, - }, - { - .name = "mark", - .revision = 1, - .family = NFPROTO_UNSPEC, - .match = mark_mt, - .matchsize = sizeof(struct xt_mark_mtinfo1), - .me = THIS_MODULE, - }, +static struct xt_match mark_mt_reg __read_mostly = { + .name = "mark", + .revision = 1, + .family = NFPROTO_UNSPEC, + .match = mark_mt, + .matchsize = sizeof(struct xt_mark_mtinfo1), + .me = THIS_MODULE, }; static int __init mark_mt_init(void) { - return xt_register_matches(mark_mt_reg, ARRAY_SIZE(mark_mt_reg)); + return xt_register_match(&mark_mt_reg); } static void __exit mark_mt_exit(void) { - xt_unregister_matches(mark_mt_reg, ARRAY_SIZE(mark_mt_reg)); + xt_unregister_match(&mark_mt_reg); } module_init(mark_mt_init); -- cgit v1.2.3 From 6e2efaacb3579fd9643d0dc59963b58b801c03a1 Mon Sep 17 00:00:00 2001 From: Clemens Ladisch Date: Mon, 10 Aug 2009 10:06:53 +0200 Subject: sound: ymfpci: increase timer resolution to 96 kHz Allow the interval timer to be programmed with its full 96 kHz precision. Signed-off-by: Clemens Ladisch Signed-off-by: Takashi Iwai --- include/sound/ymfpci.h | 1 + sound/pci/ymfpci/ymfpci_main.c | 20 +++++++++++++++----- 2 files changed, 16 insertions(+), 5 deletions(-) (limited to 'include') diff --git a/include/sound/ymfpci.h b/include/sound/ymfpci.h index 05ead6698434..444cd6ba0ba7 100644 --- a/include/sound/ymfpci.h +++ b/include/sound/ymfpci.h @@ -331,6 +331,7 @@ struct snd_ymfpci { struct snd_ac97 *ac97; struct snd_rawmidi *rawmidi; struct snd_timer *timer; + unsigned int timer_ticks; struct pci_dev *pci; struct snd_card *card; diff --git a/sound/pci/ymfpci/ymfpci_main.c b/sound/pci/ymfpci/ymfpci_main.c index 2f0925236a1b..5518371db13f 100644 --- a/sound/pci/ymfpci/ymfpci_main.c +++ b/sound/pci/ymfpci/ymfpci_main.c @@ -834,7 +834,7 @@ static irqreturn_t snd_ymfpci_interrupt(int irq, void *dev_id) status = snd_ymfpci_readw(chip, YDSXGR_INTFLAG); if (status & 1) { if (chip->timer) - snd_timer_interrupt(chip->timer, chip->timer->sticks); + snd_timer_interrupt(chip->timer, chip->timer_ticks); } snd_ymfpci_writew(chip, YDSXGR_INTFLAG, status); @@ -1885,8 +1885,18 @@ static int snd_ymfpci_timer_start(struct snd_timer *timer) unsigned int count; chip = snd_timer_chip(timer); - count = (timer->sticks << 1) - 1; spin_lock_irqsave(&chip->reg_lock, flags); + if (timer->sticks > 1) { + chip->timer_ticks = timer->sticks; + count = timer->sticks - 1; + } else { + /* + * Divisor 1 is not allowed; fake it by using divisor 2 and + * counting two ticks for each interrupt. + */ + chip->timer_ticks = 2; + count = 2 - 1; + } snd_ymfpci_writew(chip, YDSXGR_TIMERCOUNT, count); snd_ymfpci_writeb(chip, YDSXGR_TIMERCTRL, 0x03); spin_unlock_irqrestore(&chip->reg_lock, flags); @@ -1909,14 +1919,14 @@ static int snd_ymfpci_timer_precise_resolution(struct snd_timer *timer, unsigned long *num, unsigned long *den) { *num = 1; - *den = 48000; + *den = 96000; return 0; } static struct snd_timer_hardware snd_ymfpci_timer_hw = { .flags = SNDRV_TIMER_HW_AUTO, - .resolution = 20833, /* 1/fs = 20.8333...us */ - .ticks = 0x8000, + .resolution = 10417, /* 1 / 96 kHz = 10.41666...us */ + .ticks = 0x10000, .start = snd_ymfpci_timer_start, .stop = snd_ymfpci_timer_stop, .precise_resolution = snd_ymfpci_timer_precise_resolution, -- cgit v1.2.3 From 6461caed83412ae3e9a16785ffa64396fb66c6a6 Mon Sep 17 00:00:00 2001 From: Jan Engelhardt Date: Fri, 12 Jun 2009 19:46:26 +0200 Subject: netfilter: xtables: remove xt_owner v0 Superseded by xt_owner v1 (v2.6.24-2388-g0265ab4). Signed-off-by: Jan Engelhardt --- include/linux/netfilter_ipv4/Kbuild | 1 - include/linux/netfilter_ipv4/ipt_owner.h | 20 ----- include/linux/netfilter_ipv6/Kbuild | 1 - include/linux/netfilter_ipv6/ip6t_owner.h | 18 ----- net/netfilter/xt_owner.c | 130 +++--------------------------- 5 files changed, 12 insertions(+), 158 deletions(-) delete mode 100644 include/linux/netfilter_ipv4/ipt_owner.h delete mode 100644 include/linux/netfilter_ipv6/ip6t_owner.h (limited to 'include') diff --git a/include/linux/netfilter_ipv4/Kbuild b/include/linux/netfilter_ipv4/Kbuild index 5e361ef44e87..541300531cb3 100644 --- a/include/linux/netfilter_ipv4/Kbuild +++ b/include/linux/netfilter_ipv4/Kbuild @@ -28,7 +28,6 @@ header-y += ipt_limit.h header-y += ipt_mac.h header-y += ipt_mark.h header-y += ipt_multiport.h -header-y += ipt_owner.h header-y += ipt_physdev.h header-y += ipt_pkttype.h header-y += ipt_policy.h diff --git a/include/linux/netfilter_ipv4/ipt_owner.h b/include/linux/netfilter_ipv4/ipt_owner.h deleted file mode 100644 index a78445be9992..000000000000 --- a/include/linux/netfilter_ipv4/ipt_owner.h +++ /dev/null @@ -1,20 +0,0 @@ -#ifndef _IPT_OWNER_H -#define _IPT_OWNER_H - -/* match and invert flags */ -#define IPT_OWNER_UID 0x01 -#define IPT_OWNER_GID 0x02 -#define IPT_OWNER_PID 0x04 -#define IPT_OWNER_SID 0x08 -#define IPT_OWNER_COMM 0x10 - -struct ipt_owner_info { - __kernel_uid32_t uid; - __kernel_gid32_t gid; - __kernel_pid_t pid; - __kernel_pid_t sid; - char comm[16]; - u_int8_t match, invert; /* flags */ -}; - -#endif /*_IPT_OWNER_H*/ diff --git a/include/linux/netfilter_ipv6/Kbuild b/include/linux/netfilter_ipv6/Kbuild index aca4bd1f6d7c..4610a16da0ab 100644 --- a/include/linux/netfilter_ipv6/Kbuild +++ b/include/linux/netfilter_ipv6/Kbuild @@ -14,7 +14,6 @@ header-y += ip6t_mark.h header-y += ip6t_mh.h header-y += ip6t_multiport.h header-y += ip6t_opts.h -header-y += ip6t_owner.h header-y += ip6t_physdev.h header-y += ip6t_policy.h header-y += ip6t_rt.h diff --git a/include/linux/netfilter_ipv6/ip6t_owner.h b/include/linux/netfilter_ipv6/ip6t_owner.h deleted file mode 100644 index ec5cc7a38c42..000000000000 --- a/include/linux/netfilter_ipv6/ip6t_owner.h +++ /dev/null @@ -1,18 +0,0 @@ -#ifndef _IP6T_OWNER_H -#define _IP6T_OWNER_H - -/* match and invert flags */ -#define IP6T_OWNER_UID 0x01 -#define IP6T_OWNER_GID 0x02 -#define IP6T_OWNER_PID 0x04 -#define IP6T_OWNER_SID 0x08 - -struct ip6t_owner_info { - __kernel_uid32_t uid; - __kernel_gid32_t gid; - __kernel_pid_t pid; - __kernel_pid_t sid; - u_int8_t match, invert; /* flags */ -}; - -#endif /*_IPT_OWNER_H*/ diff --git a/net/netfilter/xt_owner.c b/net/netfilter/xt_owner.c index 22b2a5e881ea..d24c76dffee2 100644 --- a/net/netfilter/xt_owner.c +++ b/net/netfilter/xt_owner.c @@ -5,7 +5,6 @@ * (C) 2000 Marc Boucher * * Copyright © CC Computer Consultants GmbH, 2007 - 2008 - * * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License version 2 as @@ -17,60 +16,6 @@ #include #include #include -#include -#include - -static bool -owner_mt_v0(const struct sk_buff *skb, const struct xt_match_param *par) -{ - const struct ipt_owner_info *info = par->matchinfo; - const struct file *filp; - - if (skb->sk == NULL || skb->sk->sk_socket == NULL) - return false; - - filp = skb->sk->sk_socket->file; - if (filp == NULL) - return false; - - if (info->match & IPT_OWNER_UID) - if ((filp->f_cred->fsuid != info->uid) ^ - !!(info->invert & IPT_OWNER_UID)) - return false; - - if (info->match & IPT_OWNER_GID) - if ((filp->f_cred->fsgid != info->gid) ^ - !!(info->invert & IPT_OWNER_GID)) - return false; - - return true; -} - -static bool -owner_mt6_v0(const struct sk_buff *skb, const struct xt_match_param *par) -{ - const struct ip6t_owner_info *info = par->matchinfo; - const struct file *filp; - - if (skb->sk == NULL || skb->sk->sk_socket == NULL) - return false; - - filp = skb->sk->sk_socket->file; - if (filp == NULL) - return false; - - if (info->match & IP6T_OWNER_UID) - if ((filp->f_cred->fsuid != info->uid) ^ - !!(info->invert & IP6T_OWNER_UID)) - return false; - - if (info->match & IP6T_OWNER_GID) - if ((filp->f_cred->fsgid != info->gid) ^ - !!(info->invert & IP6T_OWNER_GID)) - return false; - - return true; -} static bool owner_mt(const struct sk_buff *skb, const struct xt_match_param *par) @@ -107,81 +52,30 @@ owner_mt(const struct sk_buff *skb, const struct xt_match_param *par) return true; } -static bool owner_mt_check_v0(const struct xt_mtchk_param *par) -{ - const struct ipt_owner_info *info = par->matchinfo; - - if (info->match & (IPT_OWNER_PID | IPT_OWNER_SID | IPT_OWNER_COMM)) { - printk(KERN_WARNING KBUILD_MODNAME - ": PID, SID and command matching is not " - "supported anymore\n"); - return false; - } - - return true; -} - -static bool owner_mt6_check_v0(const struct xt_mtchk_param *par) -{ - const struct ip6t_owner_info *info = par->matchinfo; - - if (info->match & (IP6T_OWNER_PID | IP6T_OWNER_SID)) { - printk(KERN_WARNING KBUILD_MODNAME - ": PID and SID matching is not supported anymore\n"); - return false; - } - - return true; -} - -static struct xt_match owner_mt_reg[] __read_mostly = { - { - .name = "owner", - .revision = 0, - .family = NFPROTO_IPV4, - .match = owner_mt_v0, - .matchsize = sizeof(struct ipt_owner_info), - .checkentry = owner_mt_check_v0, - .hooks = (1 << NF_INET_LOCAL_OUT) | - (1 << NF_INET_POST_ROUTING), - .me = THIS_MODULE, - }, - { - .name = "owner", - .revision = 0, - .family = NFPROTO_IPV6, - .match = owner_mt6_v0, - .matchsize = sizeof(struct ip6t_owner_info), - .checkentry = owner_mt6_check_v0, - .hooks = (1 << NF_INET_LOCAL_OUT) | - (1 << NF_INET_POST_ROUTING), - .me = THIS_MODULE, - }, - { - .name = "owner", - .revision = 1, - .family = NFPROTO_UNSPEC, - .match = owner_mt, - .matchsize = sizeof(struct xt_owner_match_info), - .hooks = (1 << NF_INET_LOCAL_OUT) | - (1 << NF_INET_POST_ROUTING), - .me = THIS_MODULE, - }, +static struct xt_match owner_mt_reg __read_mostly = { + .name = "owner", + .revision = 1, + .family = NFPROTO_UNSPEC, + .match = owner_mt, + .matchsize = sizeof(struct xt_owner_match_info), + .hooks = (1 << NF_INET_LOCAL_OUT) | + (1 << NF_INET_POST_ROUTING), + .me = THIS_MODULE, }; static int __init owner_mt_init(void) { - return xt_register_matches(owner_mt_reg, ARRAY_SIZE(owner_mt_reg)); + return xt_register_match(&owner_mt_reg); } static void __exit owner_mt_exit(void) { - xt_unregister_matches(owner_mt_reg, ARRAY_SIZE(owner_mt_reg)); + xt_unregister_match(&owner_mt_reg); } module_init(owner_mt_init); module_exit(owner_mt_exit); -MODULE_AUTHOR("Jan Engelhardt "); +MODULE_AUTHOR("Jan Engelhardt "); MODULE_DESCRIPTION("Xtables: socket owner matching"); MODULE_LICENSE("GPL"); MODULE_ALIAS("ipt_owner"); -- cgit v1.2.3 From 93bb1e9d117bfc60306b2b8bd9e0fa7ba3c88636 Mon Sep 17 00:00:00 2001 From: Jan Engelhardt Date: Fri, 12 Jun 2009 19:47:21 +0200 Subject: netfilter: xtables: remove redirecting header files When IPv4 and IPv6 matches were unified approx. 3.5 years ago, they received new header filenames (e.g. xt_CLASSIFY.h). Let's remove the old ones now. Signed-off-by: Jan Engelhardt --- Documentation/feature-removal-schedule.txt | 3 - include/linux/netfilter_ipv4/Kbuild | 28 ------- include/linux/netfilter_ipv4/ipt_CLASSIFY.h | 7 -- include/linux/netfilter_ipv4/ipt_CONNMARK.h | 19 ----- include/linux/netfilter_ipv4/ipt_DSCP.h | 18 ----- include/linux/netfilter_ipv4/ipt_ECN.h | 4 +- include/linux/netfilter_ipv4/ipt_MARK.h | 18 ----- include/linux/netfilter_ipv4/ipt_NFQUEUE.h | 16 ---- include/linux/netfilter_ipv4/ipt_TCPMSS.h | 9 --- include/linux/netfilter_ipv4/ipt_comment.h | 10 --- include/linux/netfilter_ipv4/ipt_connbytes.h | 18 ----- include/linux/netfilter_ipv4/ipt_connmark.h | 7 -- include/linux/netfilter_ipv4/ipt_conntrack.h | 28 ------- include/linux/netfilter_ipv4/ipt_dccp.h | 15 ---- include/linux/netfilter_ipv4/ipt_dscp.h | 21 ------ include/linux/netfilter_ipv4/ipt_ecn.h | 4 +- include/linux/netfilter_ipv4/ipt_esp.h | 10 --- include/linux/netfilter_ipv4/ipt_hashlimit.h | 14 ---- include/linux/netfilter_ipv4/ipt_helper.h | 7 -- include/linux/netfilter_ipv4/ipt_length.h | 7 -- include/linux/netfilter_ipv4/ipt_limit.h | 8 -- include/linux/netfilter_ipv4/ipt_mac.h | 7 -- include/linux/netfilter_ipv4/ipt_mark.h | 9 --- include/linux/netfilter_ipv4/ipt_multiport.h | 15 ---- include/linux/netfilter_ipv4/ipt_physdev.h | 17 ----- include/linux/netfilter_ipv4/ipt_pkttype.h | 7 -- include/linux/netfilter_ipv4/ipt_policy.h | 23 ------ include/linux/netfilter_ipv4/ipt_recent.h | 21 ------ include/linux/netfilter_ipv4/ipt_sctp.h | 105 -------------------------- include/linux/netfilter_ipv4/ipt_state.h | 15 ---- include/linux/netfilter_ipv4/ipt_string.h | 10 --- include/linux/netfilter_ipv4/ipt_tcpmss.h | 7 -- include/linux/netfilter_ipv6/Kbuild | 11 +-- include/linux/netfilter_ipv6/ip6t_MARK.h | 9 --- include/linux/netfilter_ipv6/ip6t_esp.h | 10 --- include/linux/netfilter_ipv6/ip6t_length.h | 8 -- include/linux/netfilter_ipv6/ip6t_limit.h | 8 -- include/linux/netfilter_ipv6/ip6t_mac.h | 7 -- include/linux/netfilter_ipv6/ip6t_mark.h | 9 --- include/linux/netfilter_ipv6/ip6t_multiport.h | 14 ---- include/linux/netfilter_ipv6/ip6t_physdev.h | 17 ----- include/linux/netfilter_ipv6/ip6t_policy.h | 23 ------ 42 files changed, 5 insertions(+), 618 deletions(-) delete mode 100644 include/linux/netfilter_ipv4/ipt_CLASSIFY.h delete mode 100644 include/linux/netfilter_ipv4/ipt_CONNMARK.h delete mode 100644 include/linux/netfilter_ipv4/ipt_DSCP.h delete mode 100644 include/linux/netfilter_ipv4/ipt_MARK.h delete mode 100644 include/linux/netfilter_ipv4/ipt_NFQUEUE.h delete mode 100644 include/linux/netfilter_ipv4/ipt_TCPMSS.h delete mode 100644 include/linux/netfilter_ipv4/ipt_comment.h delete mode 100644 include/linux/netfilter_ipv4/ipt_connbytes.h delete mode 100644 include/linux/netfilter_ipv4/ipt_connmark.h delete mode 100644 include/linux/netfilter_ipv4/ipt_conntrack.h delete mode 100644 include/linux/netfilter_ipv4/ipt_dccp.h delete mode 100644 include/linux/netfilter_ipv4/ipt_dscp.h delete mode 100644 include/linux/netfilter_ipv4/ipt_esp.h delete mode 100644 include/linux/netfilter_ipv4/ipt_hashlimit.h delete mode 100644 include/linux/netfilter_ipv4/ipt_helper.h delete mode 100644 include/linux/netfilter_ipv4/ipt_length.h delete mode 100644 include/linux/netfilter_ipv4/ipt_limit.h delete mode 100644 include/linux/netfilter_ipv4/ipt_mac.h delete mode 100644 include/linux/netfilter_ipv4/ipt_mark.h delete mode 100644 include/linux/netfilter_ipv4/ipt_multiport.h delete mode 100644 include/linux/netfilter_ipv4/ipt_physdev.h delete mode 100644 include/linux/netfilter_ipv4/ipt_pkttype.h delete mode 100644 include/linux/netfilter_ipv4/ipt_policy.h delete mode 100644 include/linux/netfilter_ipv4/ipt_recent.h delete mode 100644 include/linux/netfilter_ipv4/ipt_sctp.h delete mode 100644 include/linux/netfilter_ipv4/ipt_state.h delete mode 100644 include/linux/netfilter_ipv4/ipt_string.h delete mode 100644 include/linux/netfilter_ipv4/ipt_tcpmss.h delete mode 100644 include/linux/netfilter_ipv6/ip6t_MARK.h delete mode 100644 include/linux/netfilter_ipv6/ip6t_esp.h delete mode 100644 include/linux/netfilter_ipv6/ip6t_length.h delete mode 100644 include/linux/netfilter_ipv6/ip6t_limit.h delete mode 100644 include/linux/netfilter_ipv6/ip6t_mac.h delete mode 100644 include/linux/netfilter_ipv6/ip6t_mark.h delete mode 100644 include/linux/netfilter_ipv6/ip6t_multiport.h delete mode 100644 include/linux/netfilter_ipv6/ip6t_physdev.h delete mode 100644 include/linux/netfilter_ipv6/ip6t_policy.h (limited to 'include') diff --git a/Documentation/feature-removal-schedule.txt b/Documentation/feature-removal-schedule.txt index 5556d2300bea..698e1e8b3042 100644 --- a/Documentation/feature-removal-schedule.txt +++ b/Documentation/feature-removal-schedule.txt @@ -235,9 +235,6 @@ Who: Thomas Gleixner --------------------------- What (Why): - - "forwarding" header files like ipt_mac.h in - include/linux/netfilter_ipv4/ and include/linux/netfilter_ipv6/ - - xt_recent: the old ipt_recent proc dir (superseded by /proc/net/xt_recent) diff --git a/include/linux/netfilter_ipv4/Kbuild b/include/linux/netfilter_ipv4/Kbuild index 541300531cb3..431b40761920 100644 --- a/include/linux/netfilter_ipv4/Kbuild +++ b/include/linux/netfilter_ipv4/Kbuild @@ -1,42 +1,14 @@ -header-y += ipt_CLASSIFY.h header-y += ipt_CLUSTERIP.h -header-y += ipt_CONNMARK.h -header-y += ipt_DSCP.h header-y += ipt_ECN.h header-y += ipt_LOG.h -header-y += ipt_MARK.h -header-y += ipt_NFQUEUE.h header-y += ipt_REJECT.h header-y += ipt_SAME.h -header-y += ipt_TCPMSS.h header-y += ipt_TTL.h header-y += ipt_ULOG.h header-y += ipt_addrtype.h header-y += ipt_ah.h -header-y += ipt_comment.h -header-y += ipt_connbytes.h -header-y += ipt_connmark.h -header-y += ipt_conntrack.h -header-y += ipt_dccp.h -header-y += ipt_dscp.h header-y += ipt_ecn.h -header-y += ipt_esp.h -header-y += ipt_hashlimit.h -header-y += ipt_helper.h -header-y += ipt_length.h -header-y += ipt_limit.h -header-y += ipt_mac.h -header-y += ipt_mark.h -header-y += ipt_multiport.h -header-y += ipt_physdev.h -header-y += ipt_pkttype.h -header-y += ipt_policy.h header-y += ipt_realm.h -header-y += ipt_recent.h -header-y += ipt_sctp.h -header-y += ipt_state.h -header-y += ipt_string.h -header-y += ipt_tcpmss.h header-y += ipt_ttl.h unifdef-y += ip_queue.h diff --git a/include/linux/netfilter_ipv4/ipt_CLASSIFY.h b/include/linux/netfilter_ipv4/ipt_CLASSIFY.h deleted file mode 100644 index a46d511b5c36..000000000000 --- a/include/linux/netfilter_ipv4/ipt_CLASSIFY.h +++ /dev/null @@ -1,7 +0,0 @@ -#ifndef _IPT_CLASSIFY_H -#define _IPT_CLASSIFY_H - -#include -#define ipt_classify_target_info xt_classify_target_info - -#endif /*_IPT_CLASSIFY_H */ diff --git a/include/linux/netfilter_ipv4/ipt_CONNMARK.h b/include/linux/netfilter_ipv4/ipt_CONNMARK.h deleted file mode 100644 index 9ecfee0a9e33..000000000000 --- a/include/linux/netfilter_ipv4/ipt_CONNMARK.h +++ /dev/null @@ -1,19 +0,0 @@ -#ifndef _IPT_CONNMARK_H_target -#define _IPT_CONNMARK_H_target - -/* Copyright (C) 2002,2004 MARA Systems AB - * by Henrik Nordstrom - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. - */ -#include -#define IPT_CONNMARK_SET XT_CONNMARK_SET -#define IPT_CONNMARK_SAVE XT_CONNMARK_SAVE -#define IPT_CONNMARK_RESTORE XT_CONNMARK_RESTORE - -#define ipt_connmark_target_info xt_connmark_target_info - -#endif /*_IPT_CONNMARK_H_target*/ diff --git a/include/linux/netfilter_ipv4/ipt_DSCP.h b/include/linux/netfilter_ipv4/ipt_DSCP.h deleted file mode 100644 index 3491e524d5ea..000000000000 --- a/include/linux/netfilter_ipv4/ipt_DSCP.h +++ /dev/null @@ -1,18 +0,0 @@ -/* iptables module for setting the IPv4 DSCP field - * - * (C) 2002 Harald Welte - * based on ipt_FTOS.c (C) 2000 by Matthew G. Marsh - * This software is distributed under GNU GPL v2, 1991 - * - * See RFC2474 for a description of the DSCP field within the IP Header. - * - * ipt_DSCP.h,v 1.7 2002/03/14 12:03:13 laforge Exp -*/ -#ifndef _IPT_DSCP_TARGET_H -#define _IPT_DSCP_TARGET_H -#include -#include - -#define ipt_DSCP_info xt_DSCP_info - -#endif /* _IPT_DSCP_TARGET_H */ diff --git a/include/linux/netfilter_ipv4/ipt_ECN.h b/include/linux/netfilter_ipv4/ipt_ECN.h index 94e0d9866469..7ca45918ab8e 100644 --- a/include/linux/netfilter_ipv4/ipt_ECN.h +++ b/include/linux/netfilter_ipv4/ipt_ECN.h @@ -8,9 +8,9 @@ */ #ifndef _IPT_ECN_TARGET_H #define _IPT_ECN_TARGET_H -#include +#include -#define IPT_ECN_IP_MASK (~IPT_DSCP_MASK) +#define IPT_ECN_IP_MASK (~XT_DSCP_MASK) #define IPT_ECN_OP_SET_IP 0x01 /* set ECN bits of IPv4 header */ #define IPT_ECN_OP_SET_ECE 0x10 /* set ECE bit of TCP header */ diff --git a/include/linux/netfilter_ipv4/ipt_MARK.h b/include/linux/netfilter_ipv4/ipt_MARK.h deleted file mode 100644 index 697a486a96d3..000000000000 --- a/include/linux/netfilter_ipv4/ipt_MARK.h +++ /dev/null @@ -1,18 +0,0 @@ -#ifndef _IPT_MARK_H_target -#define _IPT_MARK_H_target - -/* Backwards compatibility for old userspace */ - -#include - -/* Version 0 */ -#define ipt_mark_target_info xt_mark_target_info - -/* Version 1 */ -#define IPT_MARK_SET XT_MARK_SET -#define IPT_MARK_AND XT_MARK_AND -#define IPT_MARK_OR XT_MARK_OR - -#define ipt_mark_target_info_v1 xt_mark_target_info_v1 - -#endif /*_IPT_MARK_H_target*/ diff --git a/include/linux/netfilter_ipv4/ipt_NFQUEUE.h b/include/linux/netfilter_ipv4/ipt_NFQUEUE.h deleted file mode 100644 index 97a2a7557cb9..000000000000 --- a/include/linux/netfilter_ipv4/ipt_NFQUEUE.h +++ /dev/null @@ -1,16 +0,0 @@ -/* iptables module for using NFQUEUE mechanism - * - * (C) 2005 Harald Welte - * - * This software is distributed under GNU GPL v2, 1991 - * -*/ -#ifndef _IPT_NFQ_TARGET_H -#define _IPT_NFQ_TARGET_H - -/* Backwards compatibility for old userspace */ -#include - -#define ipt_NFQ_info xt_NFQ_info - -#endif /* _IPT_DSCP_TARGET_H */ diff --git a/include/linux/netfilter_ipv4/ipt_TCPMSS.h b/include/linux/netfilter_ipv4/ipt_TCPMSS.h deleted file mode 100644 index 7a850f945824..000000000000 --- a/include/linux/netfilter_ipv4/ipt_TCPMSS.h +++ /dev/null @@ -1,9 +0,0 @@ -#ifndef _IPT_TCPMSS_H -#define _IPT_TCPMSS_H - -#include - -#define ipt_tcpmss_info xt_tcpmss_info -#define IPT_TCPMSS_CLAMP_PMTU XT_TCPMSS_CLAMP_PMTU - -#endif /*_IPT_TCPMSS_H*/ diff --git a/include/linux/netfilter_ipv4/ipt_comment.h b/include/linux/netfilter_ipv4/ipt_comment.h deleted file mode 100644 index ae2afc2f7481..000000000000 --- a/include/linux/netfilter_ipv4/ipt_comment.h +++ /dev/null @@ -1,10 +0,0 @@ -#ifndef _IPT_COMMENT_H -#define _IPT_COMMENT_H - -#include - -#define IPT_MAX_COMMENT_LEN XT_MAX_COMMENT_LEN - -#define ipt_comment_info xt_comment_info - -#endif /* _IPT_COMMENT_H */ diff --git a/include/linux/netfilter_ipv4/ipt_connbytes.h b/include/linux/netfilter_ipv4/ipt_connbytes.h deleted file mode 100644 index f63e6ee91113..000000000000 --- a/include/linux/netfilter_ipv4/ipt_connbytes.h +++ /dev/null @@ -1,18 +0,0 @@ -#ifndef _IPT_CONNBYTES_H -#define _IPT_CONNBYTES_H - -#include -#define ipt_connbytes_what xt_connbytes_what - -#define IPT_CONNBYTES_PKTS XT_CONNBYTES_PKTS -#define IPT_CONNBYTES_BYTES XT_CONNBYTES_BYTES -#define IPT_CONNBYTES_AVGPKT XT_CONNBYTES_AVGPKT - -#define ipt_connbytes_direction xt_connbytes_direction -#define IPT_CONNBYTES_DIR_ORIGINAL XT_CONNBYTES_DIR_ORIGINAL -#define IPT_CONNBYTES_DIR_REPLY XT_CONNBYTES_DIR_REPLY -#define IPT_CONNBYTES_DIR_BOTH XT_CONNBYTES_DIR_BOTH - -#define ipt_connbytes_info xt_connbytes_info - -#endif diff --git a/include/linux/netfilter_ipv4/ipt_connmark.h b/include/linux/netfilter_ipv4/ipt_connmark.h deleted file mode 100644 index c7ba6560d44c..000000000000 --- a/include/linux/netfilter_ipv4/ipt_connmark.h +++ /dev/null @@ -1,7 +0,0 @@ -#ifndef _IPT_CONNMARK_H -#define _IPT_CONNMARK_H - -#include -#define ipt_connmark_info xt_connmark_info - -#endif /*_IPT_CONNMARK_H*/ diff --git a/include/linux/netfilter_ipv4/ipt_conntrack.h b/include/linux/netfilter_ipv4/ipt_conntrack.h deleted file mode 100644 index cde6762949c5..000000000000 --- a/include/linux/netfilter_ipv4/ipt_conntrack.h +++ /dev/null @@ -1,28 +0,0 @@ -/* Header file for kernel module to match connection tracking information. - * GPL (C) 2001 Marc Boucher (marc@mbsi.ca). - */ - -#ifndef _IPT_CONNTRACK_H -#define _IPT_CONNTRACK_H - -#include - -#define IPT_CONNTRACK_STATE_BIT(ctinfo) XT_CONNTRACK_STATE_BIT(ctinfo) -#define IPT_CONNTRACK_STATE_INVALID XT_CONNTRACK_STATE_INVALID - -#define IPT_CONNTRACK_STATE_SNAT XT_CONNTRACK_STATE_SNAT -#define IPT_CONNTRACK_STATE_DNAT XT_CONNTRACK_STATE_DNAT -#define IPT_CONNTRACK_STATE_UNTRACKED XT_CONNTRACK_STATE_UNTRACKED - -/* flags, invflags: */ -#define IPT_CONNTRACK_STATE XT_CONNTRACK_STATE -#define IPT_CONNTRACK_PROTO XT_CONNTRACK_PROTO -#define IPT_CONNTRACK_ORIGSRC XT_CONNTRACK_ORIGSRC -#define IPT_CONNTRACK_ORIGDST XT_CONNTRACK_ORIGDST -#define IPT_CONNTRACK_REPLSRC XT_CONNTRACK_REPLSRC -#define IPT_CONNTRACK_REPLDST XT_CONNTRACK_REPLDST -#define IPT_CONNTRACK_STATUS XT_CONNTRACK_STATUS -#define IPT_CONNTRACK_EXPIRES XT_CONNTRACK_EXPIRES - -#define ipt_conntrack_info xt_conntrack_info -#endif /*_IPT_CONNTRACK_H*/ diff --git a/include/linux/netfilter_ipv4/ipt_dccp.h b/include/linux/netfilter_ipv4/ipt_dccp.h deleted file mode 100644 index e70d11e1f53c..000000000000 --- a/include/linux/netfilter_ipv4/ipt_dccp.h +++ /dev/null @@ -1,15 +0,0 @@ -#ifndef _IPT_DCCP_H_ -#define _IPT_DCCP_H_ - -#include -#define IPT_DCCP_SRC_PORTS XT_DCCP_SRC_PORTS -#define IPT_DCCP_DEST_PORTS XT_DCCP_DEST_PORTS -#define IPT_DCCP_TYPE XT_DCCP_TYPE -#define IPT_DCCP_OPTION XT_DCCP_OPTION - -#define IPT_DCCP_VALID_FLAGS XT_DCCP_VALID_FLAGS - -#define ipt_dccp_info xt_dccp_info - -#endif /* _IPT_DCCP_H_ */ - diff --git a/include/linux/netfilter_ipv4/ipt_dscp.h b/include/linux/netfilter_ipv4/ipt_dscp.h deleted file mode 100644 index 4b82ca912b0e..000000000000 --- a/include/linux/netfilter_ipv4/ipt_dscp.h +++ /dev/null @@ -1,21 +0,0 @@ -/* iptables module for matching the IPv4 DSCP field - * - * (C) 2002 Harald Welte - * This software is distributed under GNU GPL v2, 1991 - * - * See RFC2474 for a description of the DSCP field within the IP Header. - * - * ipt_dscp.h,v 1.3 2002/08/05 19:00:21 laforge Exp -*/ -#ifndef _IPT_DSCP_H -#define _IPT_DSCP_H - -#include - -#define IPT_DSCP_MASK XT_DSCP_MASK -#define IPT_DSCP_SHIFT XT_DSCP_SHIFT -#define IPT_DSCP_MAX XT_DSCP_MAX - -#define ipt_dscp_info xt_dscp_info - -#endif /* _IPT_DSCP_H */ diff --git a/include/linux/netfilter_ipv4/ipt_ecn.h b/include/linux/netfilter_ipv4/ipt_ecn.h index 1f0d9a4d3378..9945baa4ccd7 100644 --- a/include/linux/netfilter_ipv4/ipt_ecn.h +++ b/include/linux/netfilter_ipv4/ipt_ecn.h @@ -8,9 +8,9 @@ */ #ifndef _IPT_ECN_H #define _IPT_ECN_H -#include +#include -#define IPT_ECN_IP_MASK (~IPT_DSCP_MASK) +#define IPT_ECN_IP_MASK (~XT_DSCP_MASK) #define IPT_ECN_OP_MATCH_IP 0x01 #define IPT_ECN_OP_MATCH_ECE 0x10 diff --git a/include/linux/netfilter_ipv4/ipt_esp.h b/include/linux/netfilter_ipv4/ipt_esp.h deleted file mode 100644 index 78296e7eeff9..000000000000 --- a/include/linux/netfilter_ipv4/ipt_esp.h +++ /dev/null @@ -1,10 +0,0 @@ -#ifndef _IPT_ESP_H -#define _IPT_ESP_H - -#include - -#define ipt_esp xt_esp -#define IPT_ESP_INV_SPI XT_ESP_INV_SPI -#define IPT_ESP_INV_MASK XT_ESP_INV_MASK - -#endif /*_IPT_ESP_H*/ diff --git a/include/linux/netfilter_ipv4/ipt_hashlimit.h b/include/linux/netfilter_ipv4/ipt_hashlimit.h deleted file mode 100644 index 5662120a3d7b..000000000000 --- a/include/linux/netfilter_ipv4/ipt_hashlimit.h +++ /dev/null @@ -1,14 +0,0 @@ -#ifndef _IPT_HASHLIMIT_H -#define _IPT_HASHLIMIT_H - -#include - -#define IPT_HASHLIMIT_SCALE XT_HASHLIMIT_SCALE -#define IPT_HASHLIMIT_HASH_DIP XT_HASHLIMIT_HASH_DIP -#define IPT_HASHLIMIT_HASH_DPT XT_HASHLIMIT_HASH_DPT -#define IPT_HASHLIMIT_HASH_SIP XT_HASHLIMIT_HASH_SIP -#define IPT_HASHLIMIT_HASH_SPT XT_HASHLIMIT_HASH_SPT - -#define ipt_hashlimit_info xt_hashlimit_info - -#endif /* _IPT_HASHLIMIT_H */ diff --git a/include/linux/netfilter_ipv4/ipt_helper.h b/include/linux/netfilter_ipv4/ipt_helper.h deleted file mode 100644 index 80452c218551..000000000000 --- a/include/linux/netfilter_ipv4/ipt_helper.h +++ /dev/null @@ -1,7 +0,0 @@ -#ifndef _IPT_HELPER_H -#define _IPT_HELPER_H - -#include -#define ipt_helper_info xt_helper_info - -#endif /* _IPT_HELPER_H */ diff --git a/include/linux/netfilter_ipv4/ipt_length.h b/include/linux/netfilter_ipv4/ipt_length.h deleted file mode 100644 index 9b45206ffcef..000000000000 --- a/include/linux/netfilter_ipv4/ipt_length.h +++ /dev/null @@ -1,7 +0,0 @@ -#ifndef _IPT_LENGTH_H -#define _IPT_LENGTH_H - -#include -#define ipt_length_info xt_length_info - -#endif /*_IPT_LENGTH_H*/ diff --git a/include/linux/netfilter_ipv4/ipt_limit.h b/include/linux/netfilter_ipv4/ipt_limit.h deleted file mode 100644 index 92f5cd07bbc4..000000000000 --- a/include/linux/netfilter_ipv4/ipt_limit.h +++ /dev/null @@ -1,8 +0,0 @@ -#ifndef _IPT_RATE_H -#define _IPT_RATE_H - -#include -#define IPT_LIMIT_SCALE XT_LIMIT_SCALE -#define ipt_rateinfo xt_rateinfo - -#endif /*_IPT_RATE_H*/ diff --git a/include/linux/netfilter_ipv4/ipt_mac.h b/include/linux/netfilter_ipv4/ipt_mac.h deleted file mode 100644 index b186008a3c47..000000000000 --- a/include/linux/netfilter_ipv4/ipt_mac.h +++ /dev/null @@ -1,7 +0,0 @@ -#ifndef _IPT_MAC_H -#define _IPT_MAC_H - -#include -#define ipt_mac_info xt_mac_info - -#endif /*_IPT_MAC_H*/ diff --git a/include/linux/netfilter_ipv4/ipt_mark.h b/include/linux/netfilter_ipv4/ipt_mark.h deleted file mode 100644 index bfde67c61224..000000000000 --- a/include/linux/netfilter_ipv4/ipt_mark.h +++ /dev/null @@ -1,9 +0,0 @@ -#ifndef _IPT_MARK_H -#define _IPT_MARK_H - -/* Backwards compatibility for old userspace */ -#include - -#define ipt_mark_info xt_mark_info - -#endif /*_IPT_MARK_H*/ diff --git a/include/linux/netfilter_ipv4/ipt_multiport.h b/include/linux/netfilter_ipv4/ipt_multiport.h deleted file mode 100644 index 55fe85eca88c..000000000000 --- a/include/linux/netfilter_ipv4/ipt_multiport.h +++ /dev/null @@ -1,15 +0,0 @@ -#ifndef _IPT_MULTIPORT_H -#define _IPT_MULTIPORT_H - -#include - -#define IPT_MULTIPORT_SOURCE XT_MULTIPORT_SOURCE -#define IPT_MULTIPORT_DESTINATION XT_MULTIPORT_DESTINATION -#define IPT_MULTIPORT_EITHER XT_MULTIPORT_EITHER - -#define IPT_MULTI_PORTS XT_MULTI_PORTS - -#define ipt_multiport xt_multiport -#define ipt_multiport_v1 xt_multiport_v1 - -#endif /*_IPT_MULTIPORT_H*/ diff --git a/include/linux/netfilter_ipv4/ipt_physdev.h b/include/linux/netfilter_ipv4/ipt_physdev.h deleted file mode 100644 index 2400e7140f26..000000000000 --- a/include/linux/netfilter_ipv4/ipt_physdev.h +++ /dev/null @@ -1,17 +0,0 @@ -#ifndef _IPT_PHYSDEV_H -#define _IPT_PHYSDEV_H - -/* Backwards compatibility for old userspace */ - -#include - -#define IPT_PHYSDEV_OP_IN XT_PHYSDEV_OP_IN -#define IPT_PHYSDEV_OP_OUT XT_PHYSDEV_OP_OUT -#define IPT_PHYSDEV_OP_BRIDGED XT_PHYSDEV_OP_BRIDGED -#define IPT_PHYSDEV_OP_ISIN XT_PHYSDEV_OP_ISIN -#define IPT_PHYSDEV_OP_ISOUT XT_PHYSDEV_OP_ISOUT -#define IPT_PHYSDEV_OP_MASK XT_PHYSDEV_OP_MASK - -#define ipt_physdev_info xt_physdev_info - -#endif /*_IPT_PHYSDEV_H*/ diff --git a/include/linux/netfilter_ipv4/ipt_pkttype.h b/include/linux/netfilter_ipv4/ipt_pkttype.h deleted file mode 100644 index ff1fbc949a0c..000000000000 --- a/include/linux/netfilter_ipv4/ipt_pkttype.h +++ /dev/null @@ -1,7 +0,0 @@ -#ifndef _IPT_PKTTYPE_H -#define _IPT_PKTTYPE_H - -#include -#define ipt_pkttype_info xt_pkttype_info - -#endif /*_IPT_PKTTYPE_H*/ diff --git a/include/linux/netfilter_ipv4/ipt_policy.h b/include/linux/netfilter_ipv4/ipt_policy.h deleted file mode 100644 index 1037fb2cd206..000000000000 --- a/include/linux/netfilter_ipv4/ipt_policy.h +++ /dev/null @@ -1,23 +0,0 @@ -#ifndef _IPT_POLICY_H -#define _IPT_POLICY_H - -#include - -#define IPT_POLICY_MAX_ELEM XT_POLICY_MAX_ELEM - -/* ipt_policy_flags */ -#define IPT_POLICY_MATCH_IN XT_POLICY_MATCH_IN -#define IPT_POLICY_MATCH_OUT XT_POLICY_MATCH_OUT -#define IPT_POLICY_MATCH_NONE XT_POLICY_MATCH_NONE -#define IPT_POLICY_MATCH_STRICT XT_POLICY_MATCH_STRICT - -/* ipt_policy_modes */ -#define IPT_POLICY_MODE_TRANSPORT XT_POLICY_MODE_TRANSPORT -#define IPT_POLICY_MODE_TUNNEL XT_POLICY_MODE_TUNNEL - -#define ipt_policy_spec xt_policy_spec -#define ipt_policy_addr xt_policy_addr -#define ipt_policy_elem xt_policy_elem -#define ipt_policy_info xt_policy_info - -#endif /* _IPT_POLICY_H */ diff --git a/include/linux/netfilter_ipv4/ipt_recent.h b/include/linux/netfilter_ipv4/ipt_recent.h deleted file mode 100644 index d636cca133c2..000000000000 --- a/include/linux/netfilter_ipv4/ipt_recent.h +++ /dev/null @@ -1,21 +0,0 @@ -#ifndef _IPT_RECENT_H -#define _IPT_RECENT_H - -#include - -#define ipt_recent_info xt_recent_mtinfo - -enum { - IPT_RECENT_CHECK = XT_RECENT_CHECK, - IPT_RECENT_SET = XT_RECENT_SET, - IPT_RECENT_UPDATE = XT_RECENT_UPDATE, - IPT_RECENT_REMOVE = XT_RECENT_REMOVE, - IPT_RECENT_TTL = XT_RECENT_TTL, - - IPT_RECENT_SOURCE = XT_RECENT_SOURCE, - IPT_RECENT_DEST = XT_RECENT_DEST, - - IPT_RECENT_NAME_LEN = XT_RECENT_NAME_LEN, -}; - -#endif /*_IPT_RECENT_H*/ diff --git a/include/linux/netfilter_ipv4/ipt_sctp.h b/include/linux/netfilter_ipv4/ipt_sctp.h deleted file mode 100644 index 80b3dbacd193..000000000000 --- a/include/linux/netfilter_ipv4/ipt_sctp.h +++ /dev/null @@ -1,105 +0,0 @@ -#ifndef _IPT_SCTP_H_ -#define _IPT_SCTP_H_ - -#define IPT_SCTP_SRC_PORTS 0x01 -#define IPT_SCTP_DEST_PORTS 0x02 -#define IPT_SCTP_CHUNK_TYPES 0x04 - -#define IPT_SCTP_VALID_FLAGS 0x07 - - -struct ipt_sctp_flag_info { - u_int8_t chunktype; - u_int8_t flag; - u_int8_t flag_mask; -}; - -#define IPT_NUM_SCTP_FLAGS 4 - -struct ipt_sctp_info { - u_int16_t dpts[2]; /* Min, Max */ - u_int16_t spts[2]; /* Min, Max */ - - u_int32_t chunkmap[256 / sizeof (u_int32_t)]; /* Bit mask of chunks to be matched according to RFC 2960 */ - -#define SCTP_CHUNK_MATCH_ANY 0x01 /* Match if any of the chunk types are present */ -#define SCTP_CHUNK_MATCH_ALL 0x02 /* Match if all of the chunk types are present */ -#define SCTP_CHUNK_MATCH_ONLY 0x04 /* Match if these are the only chunk types present */ - - u_int32_t chunk_match_type; - struct ipt_sctp_flag_info flag_info[IPT_NUM_SCTP_FLAGS]; - int flag_count; - - u_int32_t flags; - u_int32_t invflags; -}; - -#define bytes(type) (sizeof(type) * 8) - -#define SCTP_CHUNKMAP_SET(chunkmap, type) \ - do { \ - chunkmap[type / bytes(u_int32_t)] |= \ - 1 << (type % bytes(u_int32_t)); \ - } while (0) - -#define SCTP_CHUNKMAP_CLEAR(chunkmap, type) \ - do { \ - chunkmap[type / bytes(u_int32_t)] &= \ - ~(1 << (type % bytes(u_int32_t))); \ - } while (0) - -#define SCTP_CHUNKMAP_IS_SET(chunkmap, type) \ -({ \ - (chunkmap[type / bytes (u_int32_t)] & \ - (1 << (type % bytes (u_int32_t)))) ? 1: 0; \ -}) - -#define SCTP_CHUNKMAP_RESET(chunkmap) \ - do { \ - int i; \ - for (i = 0; i < ARRAY_SIZE(chunkmap); i++) \ - chunkmap[i] = 0; \ - } while (0) - -#define SCTP_CHUNKMAP_SET_ALL(chunkmap) \ - do { \ - int i; \ - for (i = 0; i < ARRAY_SIZE(chunkmap); i++) \ - chunkmap[i] = ~0; \ - } while (0) - -#define SCTP_CHUNKMAP_COPY(destmap, srcmap) \ - do { \ - int i; \ - for (i = 0; i < ARRAY_SIZE(chunkmap); i++) \ - destmap[i] = srcmap[i]; \ - } while (0) - -#define SCTP_CHUNKMAP_IS_CLEAR(chunkmap) \ -({ \ - int i; \ - int flag = 1; \ - for (i = 0; i < ARRAY_SIZE(chunkmap); i++) { \ - if (chunkmap[i]) { \ - flag = 0; \ - break; \ - } \ - } \ - flag; \ -}) - -#define SCTP_CHUNKMAP_IS_ALL_SET(chunkmap) \ -({ \ - int i; \ - int flag = 1; \ - for (i = 0; i < ARRAY_SIZE(chunkmap); i++) { \ - if (chunkmap[i] != ~0) { \ - flag = 0; \ - break; \ - } \ - } \ - flag; \ -}) - -#endif /* _IPT_SCTP_H_ */ - diff --git a/include/linux/netfilter_ipv4/ipt_state.h b/include/linux/netfilter_ipv4/ipt_state.h deleted file mode 100644 index a44a99cc28cc..000000000000 --- a/include/linux/netfilter_ipv4/ipt_state.h +++ /dev/null @@ -1,15 +0,0 @@ -#ifndef _IPT_STATE_H -#define _IPT_STATE_H - -/* Backwards compatibility for old userspace */ - -#include - -#define IPT_STATE_BIT XT_STATE_BIT -#define IPT_STATE_INVALID XT_STATE_INVALID - -#define IPT_STATE_UNTRACKED XT_STATE_UNTRACKED - -#define ipt_state_info xt_state_info - -#endif /*_IPT_STATE_H*/ diff --git a/include/linux/netfilter_ipv4/ipt_string.h b/include/linux/netfilter_ipv4/ipt_string.h deleted file mode 100644 index c26de3059903..000000000000 --- a/include/linux/netfilter_ipv4/ipt_string.h +++ /dev/null @@ -1,10 +0,0 @@ -#ifndef _IPT_STRING_H -#define _IPT_STRING_H - -#include - -#define IPT_STRING_MAX_PATTERN_SIZE XT_STRING_MAX_PATTERN_SIZE -#define IPT_STRING_MAX_ALGO_NAME_SIZE XT_STRING_MAX_ALGO_NAME_SIZE -#define ipt_string_info xt_string_info - -#endif /*_IPT_STRING_H*/ diff --git a/include/linux/netfilter_ipv4/ipt_tcpmss.h b/include/linux/netfilter_ipv4/ipt_tcpmss.h deleted file mode 100644 index 18bbc8e8e009..000000000000 --- a/include/linux/netfilter_ipv4/ipt_tcpmss.h +++ /dev/null @@ -1,7 +0,0 @@ -#ifndef _IPT_TCPMSS_MATCH_H -#define _IPT_TCPMSS_MATCH_H - -#include -#define ipt_tcpmss_match_info xt_tcpmss_match_info - -#endif /*_IPT_TCPMSS_MATCH_H*/ diff --git a/include/linux/netfilter_ipv6/Kbuild b/include/linux/netfilter_ipv6/Kbuild index 4610a16da0ab..e864eaee9e5e 100644 --- a/include/linux/netfilter_ipv6/Kbuild +++ b/include/linux/netfilter_ipv6/Kbuild @@ -1,21 +1,12 @@ header-y += ip6t_HL.h header-y += ip6t_LOG.h -header-y += ip6t_MARK.h header-y += ip6t_REJECT.h header-y += ip6t_ah.h -header-y += ip6t_esp.h header-y += ip6t_frag.h -header-y += ip6t_hl.h header-y += ip6t_ipv6header.h -header-y += ip6t_length.h -header-y += ip6t_limit.h -header-y += ip6t_mac.h -header-y += ip6t_mark.h +header-y += ip6t_hl.h header-y += ip6t_mh.h -header-y += ip6t_multiport.h header-y += ip6t_opts.h -header-y += ip6t_physdev.h -header-y += ip6t_policy.h header-y += ip6t_rt.h unifdef-y += ip6_tables.h diff --git a/include/linux/netfilter_ipv6/ip6t_MARK.h b/include/linux/netfilter_ipv6/ip6t_MARK.h deleted file mode 100644 index 7cf629a8ab92..000000000000 --- a/include/linux/netfilter_ipv6/ip6t_MARK.h +++ /dev/null @@ -1,9 +0,0 @@ -#ifndef _IP6T_MARK_H_target -#define _IP6T_MARK_H_target - -/* Backwards compatibility for old userspace */ -#include - -#define ip6t_mark_target_info xt_mark_target_info - -#endif /*_IP6T_MARK_H_target*/ diff --git a/include/linux/netfilter_ipv6/ip6t_esp.h b/include/linux/netfilter_ipv6/ip6t_esp.h deleted file mode 100644 index f62eaf53c16c..000000000000 --- a/include/linux/netfilter_ipv6/ip6t_esp.h +++ /dev/null @@ -1,10 +0,0 @@ -#ifndef _IP6T_ESP_H -#define _IP6T_ESP_H - -#include - -#define ip6t_esp xt_esp -#define IP6T_ESP_INV_SPI XT_ESP_INV_SPI -#define IP6T_ESP_INV_MASK XT_ESP_INV_MASK - -#endif /*_IP6T_ESP_H*/ diff --git a/include/linux/netfilter_ipv6/ip6t_length.h b/include/linux/netfilter_ipv6/ip6t_length.h deleted file mode 100644 index 9e9689d03ed7..000000000000 --- a/include/linux/netfilter_ipv6/ip6t_length.h +++ /dev/null @@ -1,8 +0,0 @@ -#ifndef _IP6T_LENGTH_H -#define _IP6T_LENGTH_H - -#include -#define ip6t_length_info xt_length_info - -#endif /*_IP6T_LENGTH_H*/ - diff --git a/include/linux/netfilter_ipv6/ip6t_limit.h b/include/linux/netfilter_ipv6/ip6t_limit.h deleted file mode 100644 index 487e5ea342c6..000000000000 --- a/include/linux/netfilter_ipv6/ip6t_limit.h +++ /dev/null @@ -1,8 +0,0 @@ -#ifndef _IP6T_RATE_H -#define _IP6T_RATE_H - -#include -#define IP6T_LIMIT_SCALE XT_LIMIT_SCALE -#define ip6t_rateinfo xt_rateinfo - -#endif /*_IP6T_RATE_H*/ diff --git a/include/linux/netfilter_ipv6/ip6t_mac.h b/include/linux/netfilter_ipv6/ip6t_mac.h deleted file mode 100644 index ac58e83e9423..000000000000 --- a/include/linux/netfilter_ipv6/ip6t_mac.h +++ /dev/null @@ -1,7 +0,0 @@ -#ifndef _IP6T_MAC_H -#define _IP6T_MAC_H - -#include -#define ip6t_mac_info xt_mac_info - -#endif /*_IP6T_MAC_H*/ diff --git a/include/linux/netfilter_ipv6/ip6t_mark.h b/include/linux/netfilter_ipv6/ip6t_mark.h deleted file mode 100644 index ff204951ddc3..000000000000 --- a/include/linux/netfilter_ipv6/ip6t_mark.h +++ /dev/null @@ -1,9 +0,0 @@ -#ifndef _IP6T_MARK_H -#define _IP6T_MARK_H - -/* Backwards compatibility for old userspace */ -#include - -#define ip6t_mark_info xt_mark_info - -#endif /*_IPT_MARK_H*/ diff --git a/include/linux/netfilter_ipv6/ip6t_multiport.h b/include/linux/netfilter_ipv6/ip6t_multiport.h deleted file mode 100644 index 042c92661cee..000000000000 --- a/include/linux/netfilter_ipv6/ip6t_multiport.h +++ /dev/null @@ -1,14 +0,0 @@ -#ifndef _IP6T_MULTIPORT_H -#define _IP6T_MULTIPORT_H - -#include - -#define IP6T_MULTIPORT_SOURCE XT_MULTIPORT_SOURCE -#define IP6T_MULTIPORT_DESTINATION XT_MULTIPORT_DESTINATION -#define IP6T_MULTIPORT_EITHER XT_MULTIPORT_EITHER - -#define IP6T_MULTI_PORTS XT_MULTI_PORTS - -#define ip6t_multiport xt_multiport - -#endif /*_IP6T_MULTIPORT_H*/ diff --git a/include/linux/netfilter_ipv6/ip6t_physdev.h b/include/linux/netfilter_ipv6/ip6t_physdev.h deleted file mode 100644 index c161c0a81b55..000000000000 --- a/include/linux/netfilter_ipv6/ip6t_physdev.h +++ /dev/null @@ -1,17 +0,0 @@ -#ifndef _IP6T_PHYSDEV_H -#define _IP6T_PHYSDEV_H - -/* Backwards compatibility for old userspace */ - -#include - -#define IP6T_PHYSDEV_OP_IN XT_PHYSDEV_OP_IN -#define IP6T_PHYSDEV_OP_OUT XT_PHYSDEV_OP_OUT -#define IP6T_PHYSDEV_OP_BRIDGED XT_PHYSDEV_OP_BRIDGED -#define IP6T_PHYSDEV_OP_ISIN XT_PHYSDEV_OP_ISIN -#define IP6T_PHYSDEV_OP_ISOUT XT_PHYSDEV_OP_ISOUT -#define IP6T_PHYSDEV_OP_MASK XT_PHYSDEV_OP_MASK - -#define ip6t_physdev_info xt_physdev_info - -#endif /*_IP6T_PHYSDEV_H*/ diff --git a/include/linux/netfilter_ipv6/ip6t_policy.h b/include/linux/netfilter_ipv6/ip6t_policy.h deleted file mode 100644 index b1c449d7ec89..000000000000 --- a/include/linux/netfilter_ipv6/ip6t_policy.h +++ /dev/null @@ -1,23 +0,0 @@ -#ifndef _IP6T_POLICY_H -#define _IP6T_POLICY_H - -#include - -#define IP6T_POLICY_MAX_ELEM XT_POLICY_MAX_ELEM - -/* ip6t_policy_flags */ -#define IP6T_POLICY_MATCH_IN XT_POLICY_MATCH_IN -#define IP6T_POLICY_MATCH_OUT XT_POLICY_MATCH_OUT -#define IP6T_POLICY_MATCH_NONE XT_POLICY_MATCH_NONE -#define IP6T_POLICY_MATCH_STRICT XT_POLICY_MATCH_STRICT - -/* ip6t_policy_modes */ -#define IP6T_POLICY_MODE_TRANSPORT XT_POLICY_MODE_TRANSPORT -#define IP6T_POLICY_MODE_TUNNEL XT_POLICY_MODE_TUNNEL - -#define ip6t_policy_spec xt_policy_spec -#define ip6t_policy_addr xt_policy_addr -#define ip6t_policy_elem xt_policy_elem -#define ip6t_policy_info xt_policy_info - -#endif /* _IP6T_POLICY_H */ -- cgit v1.2.3 From 98d89b4198cf7273968e9217a62ec7ccfd760171 Mon Sep 17 00:00:00 2001 From: Jan Engelhardt Date: Sun, 5 Jul 2009 15:55:06 +0200 Subject: netfilter: xtables: realign struct xt_target_param This commit gets rid of a padding hole as reported by pahole(1). Saves 8 bytes on x86_64. Signed-off-by: Jan Engelhardt --- include/linux/netfilter/x_tables.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/netfilter/x_tables.h b/include/linux/netfilter/x_tables.h index 1030b7593898..4fa6e4c263e0 100644 --- a/include/linux/netfilter/x_tables.h +++ b/include/linux/netfilter/x_tables.h @@ -238,9 +238,9 @@ struct xt_mtdtor_param { */ struct xt_target_param { const struct net_device *in, *out; - unsigned int hooknum; const struct xt_target *target; const void *targinfo; + unsigned int hooknum; u_int8_t family; }; -- cgit v1.2.3 From 1905b1bfc0de6f69a61dc03cac0d86a04b3216bd Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Thu, 6 Aug 2009 18:13:23 +0900 Subject: chrdev: implement __[un]register_chrdev() [un]register_chrdev() assume minor range 0-255. This patch adds __ prefixed versions which take @minorbase and @count explicitly. Signed-off-by: Tejun Heo Cc: Al Viro Cc: Greg Kroah-Hartman Signed-off-by: Takashi Iwai --- fs/char_dev.c | 39 ++++++++++++++++++++++++++------------- include/linux/fs.h | 19 ++++++++++++++++--- 2 files changed, 42 insertions(+), 16 deletions(-) (limited to 'include') diff --git a/fs/char_dev.c b/fs/char_dev.c index a173551e19d7..2f18c1e4e301 100644 --- a/fs/char_dev.c +++ b/fs/char_dev.c @@ -237,8 +237,10 @@ int alloc_chrdev_region(dev_t *dev, unsigned baseminor, unsigned count, } /** - * register_chrdev() - Register a major number for character devices. + * __register_chrdev() - create and register a cdev occupying a range of minors * @major: major device number or 0 for dynamic allocation + * @baseminor: first of the requested range of minor numbers + * @count: the number of minor numbers required * @name: name of this range of devices * @fops: file operations associated with this devices * @@ -254,19 +256,17 @@ int alloc_chrdev_region(dev_t *dev, unsigned baseminor, unsigned count, * /dev. It only helps to keep track of the different owners of devices. If * your module name has only one type of devices it's ok to use e.g. the name * of the module here. - * - * This function registers a range of 256 minor numbers. The first minor number - * is 0. */ -int register_chrdev(unsigned int major, const char *name, - const struct file_operations *fops) +int __register_chrdev(unsigned int major, unsigned int baseminor, + unsigned int count, const char *name, + const struct file_operations *fops) { struct char_device_struct *cd; struct cdev *cdev; char *s; int err = -ENOMEM; - cd = __register_chrdev_region(major, 0, 256, name); + cd = __register_chrdev_region(major, baseminor, count, name); if (IS_ERR(cd)) return PTR_ERR(cd); @@ -280,7 +280,7 @@ int register_chrdev(unsigned int major, const char *name, for (s = strchr(kobject_name(&cdev->kobj),'/'); s; s = strchr(s, '/')) *s = '!'; - err = cdev_add(cdev, MKDEV(cd->major, 0), 256); + err = cdev_add(cdev, MKDEV(cd->major, baseminor), count); if (err) goto out; @@ -290,7 +290,7 @@ int register_chrdev(unsigned int major, const char *name, out: kobject_put(&cdev->kobj); out2: - kfree(__unregister_chrdev_region(cd->major, 0, 256)); + kfree(__unregister_chrdev_region(cd->major, baseminor, count)); return err; } @@ -316,10 +316,23 @@ void unregister_chrdev_region(dev_t from, unsigned count) } } -void unregister_chrdev(unsigned int major, const char *name) +/** + * __unregister_chrdev - unregister and destroy a cdev + * @major: major device number + * @baseminor: first of the range of minor numbers + * @count: the number of minor numbers this cdev is occupying + * @name: name of this range of devices + * + * Unregister and destroy the cdev occupying the region described by + * @major, @baseminor and @count. This function undoes what + * __register_chrdev() did. + */ +void __unregister_chrdev(unsigned int major, unsigned int baseminor, + unsigned int count, const char *name) { struct char_device_struct *cd; - cd = __unregister_chrdev_region(major, 0, 256); + + cd = __unregister_chrdev_region(major, baseminor, count); if (cd && cd->cdev) cdev_del(cd->cdev); kfree(cd); @@ -568,6 +581,6 @@ EXPORT_SYMBOL(cdev_alloc); EXPORT_SYMBOL(cdev_del); EXPORT_SYMBOL(cdev_add); EXPORT_SYMBOL(cdev_index); -EXPORT_SYMBOL(register_chrdev); -EXPORT_SYMBOL(unregister_chrdev); +EXPORT_SYMBOL(__register_chrdev); +EXPORT_SYMBOL(__unregister_chrdev); EXPORT_SYMBOL(directly_mappable_cdev_bdi); diff --git a/include/linux/fs.h b/include/linux/fs.h index a36ffa5a77a4..6c36ab788854 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -1998,12 +1998,25 @@ extern void bd_release_from_disk(struct block_device *, struct gendisk *); #define CHRDEV_MAJOR_HASH_SIZE 255 extern int alloc_chrdev_region(dev_t *, unsigned, unsigned, const char *); extern int register_chrdev_region(dev_t, unsigned, const char *); -extern int register_chrdev(unsigned int, const char *, - const struct file_operations *); -extern void unregister_chrdev(unsigned int, const char *); +extern int __register_chrdev(unsigned int major, unsigned int baseminor, + unsigned int count, const char *name, + const struct file_operations *fops); +extern void __unregister_chrdev(unsigned int major, unsigned int baseminor, + unsigned int count, const char *name); extern void unregister_chrdev_region(dev_t, unsigned); extern void chrdev_show(struct seq_file *,off_t); +static inline int register_chrdev(unsigned int major, const char *name, + const struct file_operations *fops) +{ + return __register_chrdev(major, 0, 256, name, fops); +} + +static inline void unregister_chrdev(unsigned int major, const char *name) +{ + __unregister_chrdev(major, 0, 256, name); +} + /* fs/block_dev.c */ #define BDEVNAME_SIZE 32 /* Largest string for a blockdev identifier */ #define BDEVT_SIZE 10 /* Largest string for MAJ:MIN for blkdev */ -- cgit v1.2.3 From 2fc391112fb6f3424435a3aa2fda887497b5f807 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Mon, 10 Aug 2009 12:33:05 +0100 Subject: locking, sched: Give waitqueue spinlocks their own lockdep classes Give waitqueue spinlocks their own lockdep classes when they are initialised from init_waitqueue_head(). This means that struct wait_queue::func functions can operate other waitqueues. This is used by CacheFiles to catch the page from a backing fs being unlocked and to wake up another thread to take a copy of it. Signed-off-by: Peter Zijlstra Signed-off-by: David Howells Tested-by: Takashi Iwai Cc: linux-cachefs@redhat.com Cc: torvalds@osdl.org Cc: akpm@linux-foundation.org LKML-Reference: <20090810113305.17284.81508.stgit@warthog.procyon.org.uk> Signed-off-by: Ingo Molnar --- include/linux/wait.h | 9 ++++++++- kernel/wait.c | 5 +++-- 2 files changed, 11 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/include/linux/wait.h b/include/linux/wait.h index 6788e1a4d4ca..cf3c2f5dba51 100644 --- a/include/linux/wait.h +++ b/include/linux/wait.h @@ -77,7 +77,14 @@ struct task_struct; #define __WAIT_BIT_KEY_INITIALIZER(word, bit) \ { .flags = word, .bit_nr = bit, } -extern void init_waitqueue_head(wait_queue_head_t *q); +extern void __init_waitqueue_head(wait_queue_head_t *q, struct lock_class_key *); + +#define init_waitqueue_head(q) \ + do { \ + static struct lock_class_key __key; \ + \ + __init_waitqueue_head((q), &__key); \ + } while (0) #ifdef CONFIG_LOCKDEP # define __WAIT_QUEUE_HEAD_INIT_ONSTACK(name) \ diff --git a/kernel/wait.c b/kernel/wait.c index ea7c3b4275cf..c4bd3d825f35 100644 --- a/kernel/wait.c +++ b/kernel/wait.c @@ -10,13 +10,14 @@ #include #include -void init_waitqueue_head(wait_queue_head_t *q) +void __init_waitqueue_head(wait_queue_head_t *q, struct lock_class_key *key) { spin_lock_init(&q->lock); + lockdep_set_class(&q->lock, key); INIT_LIST_HEAD(&q->task_list); } -EXPORT_SYMBOL(init_waitqueue_head); +EXPORT_SYMBOL(__init_waitqueue_head); void add_wait_queue(wait_queue_head_t *q, wait_queue_t *wait) { -- cgit v1.2.3 From 304703aba31a87903b8c0db8f5e6890cac2d596d Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Mon, 10 Aug 2009 16:11:32 +0200 Subject: perf_counter: Subtract the buffer size field from the event record size We compute the perf raw sample size by aligning the raw ftrace event size plus the buffer size field itself. We do that instead of aligning only the perf raw sample size, so that we might economize some in some cases. But this buffer size field is not stored in the perf raw sample, we must then substract its size from the buffer once we computed the alignment unless we may get a useless u32 field in the buffer. Signed-off-by: Frederic Weisbecker Acked-by: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: Mike Galbraith Cc: Paul Mackerras LKML-Reference: <20090810141129.GA5124@nowhere> Signed-off-by: Ingo Molnar --- include/trace/ftrace.h | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/trace/ftrace.h b/include/trace/ftrace.h index 7167b9b97da2..a05524fa245d 100644 --- a/include/trace/ftrace.h +++ b/include/trace/ftrace.h @@ -637,7 +637,12 @@ __attribute__((section("_ftrace_events"))) event_##call = { \ * pc = preempt_count(); * * __data_size = ftrace_get_offsets_(&__data_offsets, args); - * __entry_size = __data_size + sizeof(*entry); + * + * // Below we want to get the aligned size by taking into account + * // the u32 field that will later store the buffer size + * __entry_size = ALIGN(__data_size + sizeof(*entry) + sizeof(u32), + * sizeof(u64)); + * __entry_size -= sizeof(u32); * * do { * char raw_data[__entry_size]; <- allocate our sample in the stack @@ -687,6 +692,7 @@ static void ftrace_profile_##call(proto) \ __data_size = ftrace_get_offsets_##call(&__data_offsets, args); \ __entry_size = ALIGN(__data_size + sizeof(*entry) + sizeof(u32),\ sizeof(u64)); \ + __entry_size -= sizeof(u32); \ \ do { \ char raw_data[__entry_size]; \ -- cgit v1.2.3 From 1853db0e02ae4088f102b0d8e59e83dc98f93f03 Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Mon, 10 Aug 2009 16:38:36 +0200 Subject: perf_counter: Zero dead bytes from ftrace raw samples size alignment After aligning the ftrace raw samples, there are dead bytes storing random data from the stack. We don't want to leak these to userspace, then zero these out. Before: 0x2de88 [0x50]: event: 9 . . ... raw event: size 80 bytes . 0000: 09 00 00 00 01 00 50 00 d0 c7 00 81 ff ff ff ff ......P........ . 0010: 68 01 00 00 68 01 00 00 2c 00 00 00 00 00 00 00 h...h...,...... . 0020: 2c 00 00 00 2b 00 01 02 68 01 00 00 68 01 00 00 ,...+...h...h.. . 0030: 6b 6f 6e 64 65 6d 61 6e 64 2f 30 00 00 00 00 00 kondemand/0.... . 0040: 68 01 00 00 40 7f 46 81 ff ff ff ff 00 10 1b 7f h...@.F........ ^ ^ ^ ^ Leak After: 0x2d318 [0x50]: event: 9 . . ... raw event: size 80 bytes . 0000: 09 00 00 00 01 00 50 00 d0 c7 00 81 ff ff ff ff ......P........ . 0010: 68 01 00 00 68 01 00 00 68 14 00 00 00 00 00 00 h...h...h...... . 0020: 2c 00 00 00 2b 00 01 02 68 01 00 00 68 01 00 00 ,...+...h...h.. . 0030: 6b 6f 6e 64 65 6d 61 6e 64 2f 30 00 00 00 00 00 kondemand/0.... . 0040: 68 01 00 00 a0 80 46 81 ff ff ff ff 00 00 00 00 h.....F........ ^ ^ ^ ^ Fixed Reported-by: Peter Zijlstra Signed-off-by: Frederic Weisbecker Cc: Frederic Weisbecker Cc: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: Mike Galbraith LKML-Reference: <1249915116-5210-1-git-send-email-fweisbec@gmail.com> Signed-off-by: Ingo Molnar Cc: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: Mike Galbraith --- include/trace/ftrace.h | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'include') diff --git a/include/trace/ftrace.h b/include/trace/ftrace.h index a05524fa245d..f64fbaae781a 100644 --- a/include/trace/ftrace.h +++ b/include/trace/ftrace.h @@ -648,6 +648,9 @@ __attribute__((section("_ftrace_events"))) event_##call = { \ * char raw_data[__entry_size]; <- allocate our sample in the stack * struct trace_entry *ent; * + * zero dead bytes from alignment to avoid stack leak to userspace: + * + * *(u64 *)(&raw_data[__entry_size - sizeof(u64)]) = 0ULL; * entry = (struct ftrace_raw_ *)raw_data; * ent = &entry->ent; * tracing_generic_entry_update(ent, irq_flags, pc); @@ -698,6 +701,7 @@ static void ftrace_profile_##call(proto) \ char raw_data[__entry_size]; \ struct trace_entry *ent; \ \ + *(u64 *)(&raw_data[__entry_size - sizeof(u64)]) = 0ULL; \ entry = (struct ftrace_raw_##call *)raw_data; \ ent = &entry->ent; \ tracing_generic_entry_update(ent, irq_flags, pc); \ -- cgit v1.2.3 From 066e0378c23f0a3db730893f6a041e4a3922a385 Mon Sep 17 00:00:00 2001 From: Jason Baron Date: Mon, 10 Aug 2009 16:52:23 -0400 Subject: tracing: Call arch_init_ftrace_syscalls at boot Call arch_init_ftrace_syscalls at boot, so we can determine early the set of syscalls for the syscall trace events. Signed-off-by: Jason Baron Cc: Lai Jiangshan Cc: Steven Rostedt Cc: Peter Zijlstra Cc: Mathieu Desnoyers Cc: Jiaying Zhang Cc: Martin Bligh Cc: Li Zefan Cc: Masami Hiramatsu Signed-off-by: Frederic Weisbecker --- arch/x86/kernel/ftrace.c | 15 ++++----------- include/trace/syscall.h | 1 - kernel/trace/trace_syscalls.c | 1 - 3 files changed, 4 insertions(+), 13 deletions(-) (limited to 'include') diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index afb31d72618d..0d93d409b8d2 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -516,31 +516,24 @@ int syscall_name_to_nr(char *name) return -1; } -void arch_init_ftrace_syscalls(void) +static int __init arch_init_ftrace_syscalls(void) { int i; struct syscall_metadata *meta; unsigned long **psys_syscall_table = &sys_call_table; - static atomic_t refs; - - if (atomic_inc_return(&refs) != 1) - goto end; syscalls_metadata = kzalloc(sizeof(*syscalls_metadata) * FTRACE_SYSCALL_MAX, GFP_KERNEL); if (!syscalls_metadata) { WARN_ON(1); - return; + return -ENOMEM; } for (i = 0; i < FTRACE_SYSCALL_MAX; i++) { meta = find_syscall_meta(psys_syscall_table[i]); syscalls_metadata[i] = meta; } - return; - - /* Paranoid: avoid overflow */ -end: - atomic_dec(&refs); + return 0; } +arch_initcall(arch_init_ftrace_syscalls); #endif diff --git a/include/trace/syscall.h b/include/trace/syscall.h index 8cfe515cbc47..c55fcce4fbb2 100644 --- a/include/trace/syscall.h +++ b/include/trace/syscall.h @@ -19,7 +19,6 @@ struct syscall_metadata { }; #ifdef CONFIG_FTRACE_SYSCALLS -extern void arch_init_ftrace_syscalls(void); extern struct syscall_metadata *syscall_nr_to_meta(int nr); extern void start_ftrace_syscalls(void); extern void stop_ftrace_syscalls(void); diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c index 5e579645ac86..08aed439feaf 100644 --- a/kernel/trace/trace_syscalls.c +++ b/kernel/trace/trace_syscalls.c @@ -106,7 +106,6 @@ void start_ftrace_syscalls(void) if (++refcount != 1) goto unlock; - arch_init_ftrace_syscalls(); read_lock_irqsave(&tasklist_lock, flags); do_each_thread(g, t) { -- cgit v1.2.3 From 63fbdab3157b72467013fe4dcf88c85e45280ef7 Mon Sep 17 00:00:00 2001 From: Jason Baron Date: Mon, 10 Aug 2009 16:52:27 -0400 Subject: tracing: Add DECLARE_TRACE_WITH_CALLBACK() macro Introduce a new 'DECLARE_TRACE_WITH_CALLBACK()' macro, so that tracepoints can associate an external register/unregister function. This prepares for the syscalls tracer conversion to trace events. We will need to perform arch level operations once a syscall event is turned on/off, such as TIF flags setting, hence the need of such specific callbacks. Signed-off-by: Jason Baron Cc: Lai Jiangshan Cc: Steven Rostedt Cc: Peter Zijlstra Cc: Mathieu Desnoyers Cc: Jiaying Zhang Cc: Martin Bligh Cc: Li Zefan Cc: Masami Hiramatsu Signed-off-by: Frederic Weisbecker --- include/linux/tracepoint.h | 31 +++++++++++++++++++++++++++---- 1 file changed, 27 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/include/linux/tracepoint.h b/include/linux/tracepoint.h index b9dc4ca0246f..5984ed04c03b 100644 --- a/include/linux/tracepoint.h +++ b/include/linux/tracepoint.h @@ -60,8 +60,10 @@ struct tracepoint { * Make sure the alignment of the structure in the __tracepoints section will * not add unwanted padding between the beginning of the section and the * structure. Force alignment to the same alignment as the section start. + * An optional set of (un)registration functions can be passed to perform any + * additional (un)registration work. */ -#define DECLARE_TRACE(name, proto, args) \ +#define DECLARE_TRACE_WITH_CALLBACK(name, proto, args, reg, unreg) \ extern struct tracepoint __tracepoint_##name; \ static inline void trace_##name(proto) \ { \ @@ -71,13 +73,30 @@ struct tracepoint { } \ static inline int register_trace_##name(void (*probe)(proto)) \ { \ - return tracepoint_probe_register(#name, (void *)probe); \ + int ret; \ + void (*func)(void) = reg; \ + \ + ret = tracepoint_probe_register(#name, (void *)probe); \ + if (func && !ret) \ + func(); \ + return ret; \ } \ static inline int unregister_trace_##name(void (*probe)(proto)) \ { \ - return tracepoint_probe_unregister(#name, (void *)probe);\ + int ret; \ + void (*func)(void) = unreg; \ + \ + ret = tracepoint_probe_unregister(#name, (void *)probe);\ + if (func && !ret) \ + func(); \ + return ret; \ } + +#define DECLARE_TRACE(name, proto, args) \ + DECLARE_TRACE_WITH_CALLBACK(name, TP_PROTO(proto), TP_ARGS(args),\ + NULL, NULL); + #define DEFINE_TRACE(name) \ static const char __tpstrtab_##name[] \ __attribute__((section("__tracepoints_strings"))) = #name; \ @@ -94,7 +113,7 @@ extern void tracepoint_update_probe_range(struct tracepoint *begin, struct tracepoint *end); #else /* !CONFIG_TRACEPOINTS */ -#define DECLARE_TRACE(name, proto, args) \ +#define DECLARE_TRACE_WITH_CALLBACK(name, proto, args, reg, unreg) \ static inline void _do_trace_##name(struct tracepoint *tp, proto) \ { } \ static inline void trace_##name(proto) \ @@ -108,6 +127,10 @@ extern void tracepoint_update_probe_range(struct tracepoint *begin, return -ENOSYS; \ } +#define DECLARE_TRACE(name, proto, args) \ + DECLARE_TRACE_WITH_CALLBACK(name, TP_PROTO(proto), TP_ARGS(args),\ + NULL, NULL); + #define DEFINE_TRACE(name) #define EXPORT_TRACEPOINT_SYMBOL_GPL(name) #define EXPORT_TRACEPOINT_SYMBOL(name) -- cgit v1.2.3 From a871bd33a6c0bc86fb47cd02ea2650dd43d3d95f Mon Sep 17 00:00:00 2001 From: Jason Baron Date: Mon, 10 Aug 2009 16:52:31 -0400 Subject: tracing: Add syscall tracepoints add two tracepoints in syscall exit and entry path, conditioned on TIF_SYSCALL_FTRACE. Supports the syscall trace event code. Signed-off-by: Jason Baron Cc: Lai Jiangshan Cc: Steven Rostedt Cc: Peter Zijlstra Cc: Mathieu Desnoyers Cc: Jiaying Zhang Cc: Martin Bligh Cc: Li Zefan Cc: Masami Hiramatsu Signed-off-by: Frederic Weisbecker --- arch/x86/kernel/ptrace.c | 7 +++++-- include/trace/syscall.h | 20 ++++++++++++++++++++ kernel/tracepoint.c | 38 ++++++++++++++++++++++++++++++++++++++ 3 files changed, 63 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/arch/x86/kernel/ptrace.c b/arch/x86/kernel/ptrace.c index 09ecbde91c13..34dd6f15185d 100644 --- a/arch/x86/kernel/ptrace.c +++ b/arch/x86/kernel/ptrace.c @@ -37,6 +37,9 @@ #include +DEFINE_TRACE(syscall_enter); +DEFINE_TRACE(syscall_exit); + #include "tls.h" enum x86_regset { @@ -1498,7 +1501,7 @@ asmregparm long syscall_trace_enter(struct pt_regs *regs) ret = -1L; if (unlikely(test_thread_flag(TIF_SYSCALL_FTRACE))) - ftrace_syscall_enter(regs); + trace_syscall_enter(regs, regs->orig_ax); if (unlikely(current->audit_context)) { if (IS_IA32) @@ -1524,7 +1527,7 @@ asmregparm void syscall_trace_leave(struct pt_regs *regs) audit_syscall_exit(AUDITSC_RESULT(regs->ax), regs->ax); if (unlikely(test_thread_flag(TIF_SYSCALL_FTRACE))) - ftrace_syscall_exit(regs); + trace_syscall_exit(regs, regs->ax); if (test_thread_flag(TIF_SYSCALL_TRACE)) tracehook_report_syscall_exit(regs, 0); diff --git a/include/trace/syscall.h b/include/trace/syscall.h index c55fcce4fbb2..3951d774de18 100644 --- a/include/trace/syscall.h +++ b/include/trace/syscall.h @@ -1,8 +1,28 @@ #ifndef _TRACE_SYSCALL_H #define _TRACE_SYSCALL_H +#include + #include + +extern void syscall_regfunc(void); +extern void syscall_unregfunc(void); + +DECLARE_TRACE_WITH_CALLBACK(syscall_enter, + TP_PROTO(struct pt_regs *regs, long id), + TP_ARGS(regs, id), + syscall_regfunc, + syscall_unregfunc +); + +DECLARE_TRACE_WITH_CALLBACK(syscall_exit, + TP_PROTO(struct pt_regs *regs, long ret), + TP_ARGS(regs, ret), + syscall_regfunc, + syscall_unregfunc +); + /* * A syscall entry in the ftrace syscalls array. * diff --git a/kernel/tracepoint.c b/kernel/tracepoint.c index 1ef5d3a601c7..070a42bb8920 100644 --- a/kernel/tracepoint.c +++ b/kernel/tracepoint.c @@ -24,6 +24,7 @@ #include #include #include +#include extern struct tracepoint __start___tracepoints[]; extern struct tracepoint __stop___tracepoints[]; @@ -577,3 +578,40 @@ static int init_tracepoints(void) __initcall(init_tracepoints); #endif /* CONFIG_MODULES */ + +static DEFINE_MUTEX(regfunc_mutex); +static int sys_tracepoint_refcount; + +void syscall_regfunc(void) +{ + unsigned long flags; + struct task_struct *g, *t; + + mutex_lock(®func_mutex); + if (!sys_tracepoint_refcount) { + read_lock_irqsave(&tasklist_lock, flags); + do_each_thread(g, t) { + set_tsk_thread_flag(t, TIF_SYSCALL_FTRACE); + } while_each_thread(g, t); + read_unlock_irqrestore(&tasklist_lock, flags); + } + sys_tracepoint_refcount++; + mutex_unlock(®func_mutex); +} + +void syscall_unregfunc(void) +{ + unsigned long flags; + struct task_struct *g, *t; + + mutex_lock(®func_mutex); + sys_tracepoint_refcount--; + if (!sys_tracepoint_refcount) { + read_lock_irqsave(&tasklist_lock, flags); + do_each_thread(g, t) { + clear_tsk_thread_flag(t, TIF_SYSCALL_FTRACE); + } while_each_thread(g, t); + read_unlock_irqrestore(&tasklist_lock, flags); + } + mutex_unlock(®func_mutex); +} -- cgit v1.2.3 From 69fd4f0eb2ececbf8ade55e31a933e174965745e Mon Sep 17 00:00:00 2001 From: Jason Baron Date: Mon, 10 Aug 2009 16:52:44 -0400 Subject: tracing: Add ftrace_event_call void * 'data' field add an optional void * pointer to 'ftrace_event_call' that is passed in for regfunc and unregfunc. This prepares for syscall tracepoints creation by passing the name of the syscall we want to trace and then retrieve its number through our arch syscall table. Signed-off-by: Jason Baron Cc: Lai Jiangshan Cc: Steven Rostedt Cc: Peter Zijlstra Cc: Mathieu Desnoyers Cc: Jiaying Zhang Cc: Martin Bligh Cc: Li Zefan Cc: Masami Hiramatsu Signed-off-by: Frederic Weisbecker --- include/linux/ftrace_event.h | 5 +++-- include/trace/ftrace.h | 4 ++-- kernel/trace/trace_events.c | 4 ++-- 3 files changed, 7 insertions(+), 6 deletions(-) (limited to 'include') diff --git a/include/linux/ftrace_event.h b/include/linux/ftrace_event.h index ac8c6f8cf242..8544f121d9f1 100644 --- a/include/linux/ftrace_event.h +++ b/include/linux/ftrace_event.h @@ -112,8 +112,8 @@ struct ftrace_event_call { struct dentry *dir; struct trace_event *event; int enabled; - int (*regfunc)(void); - void (*unregfunc)(void); + int (*regfunc)(void *); + void (*unregfunc)(void *); int id; int (*raw_init)(void); int (*show_format)(struct trace_seq *s); @@ -122,6 +122,7 @@ struct ftrace_event_call { int filter_active; struct event_filter *filter; void *mod; + void *data; atomic_t profile_count; int (*profile_enable)(struct ftrace_event_call *); diff --git a/include/trace/ftrace.h b/include/trace/ftrace.h index 25d3b02a06f8..46d81b5e8610 100644 --- a/include/trace/ftrace.h +++ b/include/trace/ftrace.h @@ -568,7 +568,7 @@ static void ftrace_raw_event_##call(proto) \ trace_nowake_buffer_unlock_commit(event, irq_flags, pc); \ } \ \ -static int ftrace_raw_reg_event_##call(void) \ +static int ftrace_raw_reg_event_##call(void *ptr) \ { \ int ret; \ \ @@ -579,7 +579,7 @@ static int ftrace_raw_reg_event_##call(void) \ return ret; \ } \ \ -static void ftrace_raw_unreg_event_##call(void) \ +static void ftrace_raw_unreg_event_##call(void *ptr) \ { \ unregister_trace_##call(ftrace_raw_event_##call); \ } \ diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c index f95f8470dd38..1d289e2d6693 100644 --- a/kernel/trace/trace_events.c +++ b/kernel/trace/trace_events.c @@ -86,14 +86,14 @@ static void ftrace_event_enable_disable(struct ftrace_event_call *call, if (call->enabled) { call->enabled = 0; tracing_stop_cmdline_record(); - call->unregfunc(); + call->unregfunc(call->data); } break; case 1: if (!call->enabled) { call->enabled = 1; tracing_start_cmdline_record(); - call->regfunc(); + call->regfunc(call->data); } break; } -- cgit v1.2.3 From fb34a08c3469b2be9eae626ccb96476b4687b810 Mon Sep 17 00:00:00 2001 From: Jason Baron Date: Mon, 10 Aug 2009 16:52:47 -0400 Subject: tracing: Add trace events for each syscall entry/exit Layer Frederic's syscall tracer on tracepoints. We create trace events via hooking into the SYSCALL_DEFINE macros. This allows us to individually toggle syscall entry and exit points on/off. Signed-off-by: Jason Baron Cc: Lai Jiangshan Cc: Steven Rostedt Cc: Peter Zijlstra Cc: Mathieu Desnoyers Cc: Jiaying Zhang Cc: Martin Bligh Cc: Li Zefan Cc: Masami Hiramatsu Signed-off-by: Frederic Weisbecker --- include/linux/syscalls.h | 61 +++++++++++++- include/trace/syscall.h | 18 ++--- kernel/trace/trace_syscalls.c | 183 +++++++++++++++++++++--------------------- 3 files changed, 159 insertions(+), 103 deletions(-) (limited to 'include') diff --git a/include/linux/syscalls.h b/include/linux/syscalls.h index 80de7003d8c2..5e5b4d33a31c 100644 --- a/include/linux/syscalls.h +++ b/include/linux/syscalls.h @@ -64,6 +64,7 @@ struct perf_counter_attr; #include #include #include +#include #include #include #include @@ -112,6 +113,59 @@ struct perf_counter_attr; #define __SC_STR_TDECL5(t, a, ...) #t, __SC_STR_TDECL4(__VA_ARGS__) #define __SC_STR_TDECL6(t, a, ...) #t, __SC_STR_TDECL5(__VA_ARGS__) + +#define SYSCALL_TRACE_ENTER_EVENT(sname) \ + static struct ftrace_event_call event_enter_##sname; \ + static int init_enter_##sname(void) \ + { \ + int num; \ + num = syscall_name_to_nr("sys"#sname); \ + if (num < 0) \ + return -ENOSYS; \ + register_ftrace_event(&event_syscall_enter); \ + INIT_LIST_HEAD(&event_enter_##sname.fields); \ + init_preds(&event_enter_##sname); \ + return 0; \ + } \ + static struct ftrace_event_call __used \ + __attribute__((__aligned__(4))) \ + __attribute__((section("_ftrace_events"))) \ + event_enter_##sname = { \ + .name = "sys_enter"#sname, \ + .system = "syscalls", \ + .event = &event_syscall_enter, \ + .raw_init = init_enter_##sname, \ + .regfunc = reg_event_syscall_enter, \ + .unregfunc = unreg_event_syscall_enter, \ + .data = "sys"#sname, \ + } + +#define SYSCALL_TRACE_EXIT_EVENT(sname) \ + static struct ftrace_event_call event_exit_##sname; \ + static int init_exit_##sname(void) \ + { \ + int num; \ + num = syscall_name_to_nr("sys"#sname); \ + if (num < 0) \ + return -ENOSYS; \ + register_ftrace_event(&event_syscall_exit); \ + INIT_LIST_HEAD(&event_exit_##sname.fields); \ + init_preds(&event_exit_##sname); \ + return 0; \ + } \ + static struct ftrace_event_call __used \ + __attribute__((__aligned__(4))) \ + __attribute__((section("_ftrace_events"))) \ + event_exit_##sname = { \ + .name = "sys_exit"#sname, \ + .system = "syscalls", \ + .event = &event_syscall_exit, \ + .raw_init = init_exit_##sname, \ + .regfunc = reg_event_syscall_exit, \ + .unregfunc = unreg_event_syscall_exit, \ + .data = "sys"#sname, \ + } + #define SYSCALL_METADATA(sname, nb) \ static const struct syscall_metadata __used \ __attribute__((__aligned__(4))) \ @@ -121,7 +175,9 @@ struct perf_counter_attr; .nb_args = nb, \ .types = types_##sname, \ .args = args_##sname, \ - } + }; \ + SYSCALL_TRACE_ENTER_EVENT(sname); \ + SYSCALL_TRACE_EXIT_EVENT(sname); #define SYSCALL_DEFINE0(sname) \ static const struct syscall_metadata __used \ @@ -131,8 +187,9 @@ struct perf_counter_attr; .name = "sys_"#sname, \ .nb_args = 0, \ }; \ + SYSCALL_TRACE_ENTER_EVENT(_##sname); \ + SYSCALL_TRACE_EXIT_EVENT(_##sname); \ asmlinkage long sys_##sname(void) - #else #define SYSCALL_DEFINE0(name) asmlinkage long sys_##name(void) #endif diff --git a/include/trace/syscall.h b/include/trace/syscall.h index 3951d774de18..73fb8b4a9955 100644 --- a/include/trace/syscall.h +++ b/include/trace/syscall.h @@ -2,6 +2,8 @@ #define _TRACE_SYSCALL_H #include +#include +#include #include @@ -40,15 +42,13 @@ struct syscall_metadata { #ifdef CONFIG_FTRACE_SYSCALLS extern struct syscall_metadata *syscall_nr_to_meta(int nr); -extern void start_ftrace_syscalls(void); -extern void stop_ftrace_syscalls(void); -extern void ftrace_syscall_enter(struct pt_regs *regs); -extern void ftrace_syscall_exit(struct pt_regs *regs); -#else -static inline void start_ftrace_syscalls(void) { } -static inline void stop_ftrace_syscalls(void) { } -static inline void ftrace_syscall_enter(struct pt_regs *regs) { } -static inline void ftrace_syscall_exit(struct pt_regs *regs) { } +extern int syscall_name_to_nr(char *name); +extern struct trace_event event_syscall_enter; +extern struct trace_event event_syscall_exit; +extern int reg_event_syscall_enter(void *ptr); +extern void unreg_event_syscall_enter(void *ptr); +extern int reg_event_syscall_exit(void *ptr); +extern void unreg_event_syscall_exit(void *ptr); #endif #endif /* _TRACE_SYSCALL_H */ diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c index 08aed439feaf..c7ae25ee95d8 100644 --- a/kernel/trace/trace_syscalls.c +++ b/kernel/trace/trace_syscalls.c @@ -1,15 +1,16 @@ #include #include +#include #include #include "trace_output.h" #include "trace.h" -/* Keep a counter of the syscall tracing users */ -static int refcount; - -/* Prevent from races on thread flags toggling */ static DEFINE_MUTEX(syscall_trace_lock); +static int sys_refcount_enter; +static int sys_refcount_exit; +static DECLARE_BITMAP(enabled_enter_syscalls, FTRACE_SYSCALL_MAX); +static DECLARE_BITMAP(enabled_exit_syscalls, FTRACE_SYSCALL_MAX); /* Option to display the parameters types */ enum { @@ -95,53 +96,7 @@ print_syscall_exit(struct trace_iterator *iter, int flags) return TRACE_TYPE_HANDLED; } -void start_ftrace_syscalls(void) -{ - unsigned long flags; - struct task_struct *g, *t; - - mutex_lock(&syscall_trace_lock); - - /* Don't enable the flag on the tasks twice */ - if (++refcount != 1) - goto unlock; - - read_lock_irqsave(&tasklist_lock, flags); - - do_each_thread(g, t) { - set_tsk_thread_flag(t, TIF_SYSCALL_FTRACE); - } while_each_thread(g, t); - - read_unlock_irqrestore(&tasklist_lock, flags); - -unlock: - mutex_unlock(&syscall_trace_lock); -} - -void stop_ftrace_syscalls(void) -{ - unsigned long flags; - struct task_struct *g, *t; - - mutex_lock(&syscall_trace_lock); - - /* There are perhaps still some users */ - if (--refcount) - goto unlock; - - read_lock_irqsave(&tasklist_lock, flags); - - do_each_thread(g, t) { - clear_tsk_thread_flag(t, TIF_SYSCALL_FTRACE); - } while_each_thread(g, t); - - read_unlock_irqrestore(&tasklist_lock, flags); - -unlock: - mutex_unlock(&syscall_trace_lock); -} - -void ftrace_syscall_enter(struct pt_regs *regs) +void ftrace_syscall_enter(struct pt_regs *regs, long id) { struct syscall_trace_enter *entry; struct syscall_metadata *sys_data; @@ -150,6 +105,8 @@ void ftrace_syscall_enter(struct pt_regs *regs) int syscall_nr; syscall_nr = syscall_get_nr(current, regs); + if (!test_bit(syscall_nr, enabled_enter_syscalls)) + return; sys_data = syscall_nr_to_meta(syscall_nr); if (!sys_data) @@ -170,7 +127,7 @@ void ftrace_syscall_enter(struct pt_regs *regs) trace_wake_up(); } -void ftrace_syscall_exit(struct pt_regs *regs) +void ftrace_syscall_exit(struct pt_regs *regs, long ret) { struct syscall_trace_exit *entry; struct syscall_metadata *sys_data; @@ -178,6 +135,8 @@ void ftrace_syscall_exit(struct pt_regs *regs) int syscall_nr; syscall_nr = syscall_get_nr(current, regs); + if (!test_bit(syscall_nr, enabled_exit_syscalls)) + return; sys_data = syscall_nr_to_meta(syscall_nr); if (!sys_data) @@ -196,54 +155,94 @@ void ftrace_syscall_exit(struct pt_regs *regs) trace_wake_up(); } -static int init_syscall_tracer(struct trace_array *tr) +int reg_event_syscall_enter(void *ptr) { - start_ftrace_syscalls(); - - return 0; + int ret = 0; + int num; + char *name; + + name = (char *)ptr; + num = syscall_name_to_nr(name); + if (num < 0 || num >= FTRACE_SYSCALL_MAX) + return -ENOSYS; + mutex_lock(&syscall_trace_lock); + if (!sys_refcount_enter) + ret = register_trace_syscall_enter(ftrace_syscall_enter); + if (ret) { + pr_info("event trace: Could not activate" + "syscall entry trace point"); + } else { + set_bit(num, enabled_enter_syscalls); + sys_refcount_enter++; + } + mutex_unlock(&syscall_trace_lock); + return ret; } -static void reset_syscall_tracer(struct trace_array *tr) +void unreg_event_syscall_enter(void *ptr) { - stop_ftrace_syscalls(); - tracing_reset_online_cpus(tr); -} - -static struct trace_event syscall_enter_event = { - .type = TRACE_SYSCALL_ENTER, - .trace = print_syscall_enter, -}; - -static struct trace_event syscall_exit_event = { - .type = TRACE_SYSCALL_EXIT, - .trace = print_syscall_exit, -}; + int num; + char *name; -static struct tracer syscall_tracer __read_mostly = { - .name = "syscall", - .init = init_syscall_tracer, - .reset = reset_syscall_tracer, - .flags = &syscalls_flags, -}; + name = (char *)ptr; + num = syscall_name_to_nr(name); + if (num < 0 || num >= FTRACE_SYSCALL_MAX) + return; + mutex_lock(&syscall_trace_lock); + sys_refcount_enter--; + clear_bit(num, enabled_enter_syscalls); + if (!sys_refcount_enter) + unregister_trace_syscall_enter(ftrace_syscall_enter); + mutex_unlock(&syscall_trace_lock); +} -__init int register_ftrace_syscalls(void) +int reg_event_syscall_exit(void *ptr) { - int ret; - - ret = register_ftrace_event(&syscall_enter_event); - if (!ret) { - printk(KERN_WARNING "event %d failed to register\n", - syscall_enter_event.type); - WARN_ON_ONCE(1); + int ret = 0; + int num; + char *name; + + name = (char *)ptr; + num = syscall_name_to_nr(name); + if (num < 0 || num >= FTRACE_SYSCALL_MAX) + return -ENOSYS; + mutex_lock(&syscall_trace_lock); + if (!sys_refcount_exit) + ret = register_trace_syscall_exit(ftrace_syscall_exit); + if (ret) { + pr_info("event trace: Could not activate" + "syscall exit trace point"); + } else { + set_bit(num, enabled_exit_syscalls); + sys_refcount_exit++; } + mutex_unlock(&syscall_trace_lock); + return ret; +} - ret = register_ftrace_event(&syscall_exit_event); - if (!ret) { - printk(KERN_WARNING "event %d failed to register\n", - syscall_exit_event.type); - WARN_ON_ONCE(1); - } +void unreg_event_syscall_exit(void *ptr) +{ + int num; + char *name; - return register_tracer(&syscall_tracer); + name = (char *)ptr; + num = syscall_name_to_nr(name); + if (num < 0 || num >= FTRACE_SYSCALL_MAX) + return; + mutex_lock(&syscall_trace_lock); + sys_refcount_exit--; + clear_bit(num, enabled_exit_syscalls); + if (!sys_refcount_exit) + unregister_trace_syscall_exit(ftrace_syscall_exit); + mutex_unlock(&syscall_trace_lock); } -device_initcall(register_ftrace_syscalls); + +struct trace_event event_syscall_enter = { + .trace = print_syscall_enter, + .type = TRACE_SYSCALL_ENTER +}; + +struct trace_event event_syscall_exit = { + .trace = print_syscall_exit, + .type = TRACE_SYSCALL_EXIT +}; -- cgit v1.2.3 From 64c12e0444fcc6b75eb49144ba46d43dbdc6bc8f Mon Sep 17 00:00:00 2001 From: Jason Baron Date: Mon, 10 Aug 2009 16:52:53 -0400 Subject: tracing: Add individual syscalls tracepoint id support The current state of syscalls tracepoints generates only one event id for every syscall events. This patch associates an id with each syscall trace event, so that we can identify each syscall trace event using the 'perf' tool. Signed-off-by: Jason Baron Cc: Lai Jiangshan Cc: Steven Rostedt Cc: Peter Zijlstra Cc: Mathieu Desnoyers Cc: Jiaying Zhang Cc: Martin Bligh Cc: Li Zefan Cc: Masami Hiramatsu Signed-off-by: Frederic Weisbecker --- arch/x86/kernel/ftrace.c | 10 ++++++++++ include/linux/syscalls.h | 22 ++++++++++++++++++---- include/trace/syscall.h | 8 ++++++++ kernel/trace/trace.h | 6 ------ kernel/trace/trace_syscalls.c | 26 ++++++++++++++++---------- 5 files changed, 52 insertions(+), 20 deletions(-) (limited to 'include') diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index 0d93d409b8d2..3cff1214e176 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -516,6 +516,16 @@ int syscall_name_to_nr(char *name) return -1; } +void set_syscall_enter_id(int num, int id) +{ + syscalls_metadata[num]->enter_id = id; +} + +void set_syscall_exit_id(int num, int id) +{ + syscalls_metadata[num]->exit_id = id; +} + static int __init arch_init_ftrace_syscalls(void) { int i; diff --git a/include/linux/syscalls.h b/include/linux/syscalls.h index 5e5b4d33a31c..ce4b01c658eb 100644 --- a/include/linux/syscalls.h +++ b/include/linux/syscalls.h @@ -116,13 +116,20 @@ struct perf_counter_attr; #define SYSCALL_TRACE_ENTER_EVENT(sname) \ static struct ftrace_event_call event_enter_##sname; \ + struct trace_event enter_syscall_print_##sname = { \ + .trace = print_syscall_enter, \ + }; \ static int init_enter_##sname(void) \ { \ - int num; \ + int num, id; \ num = syscall_name_to_nr("sys"#sname); \ if (num < 0) \ return -ENOSYS; \ - register_ftrace_event(&event_syscall_enter); \ + id = register_ftrace_event(&enter_syscall_print_##sname);\ + if (!id) \ + return -ENODEV; \ + event_enter_##sname.id = id; \ + set_syscall_enter_id(num, id); \ INIT_LIST_HEAD(&event_enter_##sname.fields); \ init_preds(&event_enter_##sname); \ return 0; \ @@ -142,13 +149,20 @@ struct perf_counter_attr; #define SYSCALL_TRACE_EXIT_EVENT(sname) \ static struct ftrace_event_call event_exit_##sname; \ + struct trace_event exit_syscall_print_##sname = { \ + .trace = print_syscall_exit, \ + }; \ static int init_exit_##sname(void) \ { \ - int num; \ + int num, id; \ num = syscall_name_to_nr("sys"#sname); \ if (num < 0) \ return -ENOSYS; \ - register_ftrace_event(&event_syscall_exit); \ + id = register_ftrace_event(&exit_syscall_print_##sname);\ + if (!id) \ + return -ENODEV; \ + event_exit_##sname.id = id; \ + set_syscall_exit_id(num, id); \ INIT_LIST_HEAD(&event_exit_##sname.fields); \ init_preds(&event_exit_##sname); \ return 0; \ diff --git a/include/trace/syscall.h b/include/trace/syscall.h index 73fb8b4a9955..df628404241a 100644 --- a/include/trace/syscall.h +++ b/include/trace/syscall.h @@ -32,23 +32,31 @@ DECLARE_TRACE_WITH_CALLBACK(syscall_exit, * @nb_args: number of parameters it takes * @types: list of types as strings * @args: list of args as strings (args[i] matches types[i]) + * @enter_id: associated ftrace enter event id + * @exit_id: associated ftrace exit event id */ struct syscall_metadata { const char *name; int nb_args; const char **types; const char **args; + int enter_id; + int exit_id; }; #ifdef CONFIG_FTRACE_SYSCALLS extern struct syscall_metadata *syscall_nr_to_meta(int nr); extern int syscall_name_to_nr(char *name); +void set_syscall_enter_id(int num, int id); +void set_syscall_exit_id(int num, int id); extern struct trace_event event_syscall_enter; extern struct trace_event event_syscall_exit; extern int reg_event_syscall_enter(void *ptr); extern void unreg_event_syscall_enter(void *ptr); extern int reg_event_syscall_exit(void *ptr); extern void unreg_event_syscall_exit(void *ptr); +enum print_line_t print_syscall_enter(struct trace_iterator *iter, int flags); +enum print_line_t print_syscall_exit(struct trace_iterator *iter, int flags); #endif #endif /* _TRACE_SYSCALL_H */ diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h index d682357e4b1f..300ef788c976 100644 --- a/kernel/trace/trace.h +++ b/kernel/trace/trace.h @@ -34,8 +34,6 @@ enum trace_type { TRACE_GRAPH_ENT, TRACE_USER_STACK, TRACE_HW_BRANCHES, - TRACE_SYSCALL_ENTER, - TRACE_SYSCALL_EXIT, TRACE_KMEM_ALLOC, TRACE_KMEM_FREE, TRACE_POWER, @@ -319,10 +317,6 @@ extern void __ftrace_bad_type(void); TRACE_KMEM_ALLOC); \ IF_ASSIGN(var, ent, struct kmemtrace_free_entry, \ TRACE_KMEM_FREE); \ - IF_ASSIGN(var, ent, struct syscall_trace_enter, \ - TRACE_SYSCALL_ENTER); \ - IF_ASSIGN(var, ent, struct syscall_trace_exit, \ - TRACE_SYSCALL_EXIT); \ __ftrace_bad_type(); \ } while (0) diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c index c7ae25ee95d8..e58a9c11ba85 100644 --- a/kernel/trace/trace_syscalls.c +++ b/kernel/trace/trace_syscalls.c @@ -36,14 +36,18 @@ print_syscall_enter(struct trace_iterator *iter, int flags) struct syscall_metadata *entry; int i, ret, syscall; - trace_assign_type(trace, ent); - + trace = (typeof(trace))ent; syscall = trace->nr; - entry = syscall_nr_to_meta(syscall); + if (!entry) goto end; + if (entry->enter_id != ent->type) { + WARN_ON_ONCE(1); + goto end; + } + ret = trace_seq_printf(s, "%s(", entry->name); if (!ret) return TRACE_TYPE_PARTIAL_LINE; @@ -78,16 +82,20 @@ print_syscall_exit(struct trace_iterator *iter, int flags) struct syscall_metadata *entry; int ret; - trace_assign_type(trace, ent); - + trace = (typeof(trace))ent; syscall = trace->nr; - entry = syscall_nr_to_meta(syscall); + if (!entry) { trace_seq_printf(s, "\n"); return TRACE_TYPE_HANDLED; } + if (entry->exit_id != ent->type) { + WARN_ON_ONCE(1); + return TRACE_TYPE_UNHANDLED; + } + ret = trace_seq_printf(s, "%s -> 0x%lx\n", entry->name, trace->ret); if (!ret) @@ -114,7 +122,7 @@ void ftrace_syscall_enter(struct pt_regs *regs, long id) size = sizeof(*entry) + sizeof(unsigned long) * sys_data->nb_args; - event = trace_current_buffer_lock_reserve(TRACE_SYSCALL_ENTER, size, + event = trace_current_buffer_lock_reserve(sys_data->enter_id, size, 0, 0); if (!event) return; @@ -142,7 +150,7 @@ void ftrace_syscall_exit(struct pt_regs *regs, long ret) if (!sys_data) return; - event = trace_current_buffer_lock_reserve(TRACE_SYSCALL_EXIT, + event = trace_current_buffer_lock_reserve(sys_data->exit_id, sizeof(*entry), 0, 0); if (!event) return; @@ -239,10 +247,8 @@ void unreg_event_syscall_exit(void *ptr) struct trace_event event_syscall_enter = { .trace = print_syscall_enter, - .type = TRACE_SYSCALL_ENTER }; struct trace_event event_syscall_exit = { .trace = print_syscall_exit, - .type = TRACE_SYSCALL_EXIT }; -- cgit v1.2.3 From f4b5ffccc83c82947f5d9f15d6f1b6edb1b71cd7 Mon Sep 17 00:00:00 2001 From: Jason Baron Date: Mon, 10 Aug 2009 16:53:02 -0400 Subject: tracing: Add perf counter support for syscalls tracing The perf counter support is automated for usual trace events. But we have to define specific callbacks for this to handle syscalls trace events Make 'perf stat -e syscalls:sys_enter_blah' work with syscall style tracepoints. Signed-off-by: Jason Baron Cc: Lai Jiangshan Cc: Steven Rostedt Cc: Peter Zijlstra Cc: Mathieu Desnoyers Cc: Jiaying Zhang Cc: Martin Bligh Cc: Li Zefan Cc: Masami Hiramatsu Signed-off-by: Frederic Weisbecker --- include/linux/perf_counter.h | 2 + include/linux/syscalls.h | 52 +++++++++++++++++- include/trace/syscall.h | 7 +++ kernel/trace/trace_syscalls.c | 121 ++++++++++++++++++++++++++++++++++++++++++ 4 files changed, 181 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/perf_counter.h b/include/linux/perf_counter.h index a9d823a93fe8..8e6460fb4c02 100644 --- a/include/linux/perf_counter.h +++ b/include/linux/perf_counter.h @@ -734,6 +734,8 @@ extern int sysctl_perf_counter_mlock; extern int sysctl_perf_counter_sample_rate; extern void perf_counter_init(void); +extern void perf_tpcounter_event(int event_id, u64 addr, u64 count, + void *record, int entry_size); #ifndef perf_misc_flags #define perf_misc_flags(regs) (user_mode(regs) ? PERF_EVENT_MISC_USER : \ diff --git a/include/linux/syscalls.h b/include/linux/syscalls.h index ce4b01c658eb..5541e75e140a 100644 --- a/include/linux/syscalls.h +++ b/include/linux/syscalls.h @@ -98,6 +98,53 @@ struct perf_counter_attr; #define __SC_TEST5(t5, a5, ...) __SC_TEST(t5); __SC_TEST4(__VA_ARGS__) #define __SC_TEST6(t6, a6, ...) __SC_TEST(t6); __SC_TEST5(__VA_ARGS__) +#ifdef CONFIG_EVENT_PROFILE +#define TRACE_SYS_ENTER_PROFILE(sname) \ +static int prof_sysenter_enable_##sname(struct ftrace_event_call *event_call) \ +{ \ + int ret = 0; \ + if (!atomic_inc_return(&event_enter_##sname.profile_count)) \ + ret = reg_prof_syscall_enter("sys"#sname); \ + return ret; \ +} \ + \ +static void prof_sysenter_disable_##sname(struct ftrace_event_call *event_call)\ +{ \ + if (atomic_add_negative(-1, &event_enter_##sname.profile_count)) \ + unreg_prof_syscall_enter("sys"#sname); \ +} + +#define TRACE_SYS_EXIT_PROFILE(sname) \ +static int prof_sysexit_enable_##sname(struct ftrace_event_call *event_call) \ +{ \ + int ret = 0; \ + if (!atomic_inc_return(&event_exit_##sname.profile_count)) \ + ret = reg_prof_syscall_exit("sys"#sname); \ + return ret; \ +} \ + \ +static void prof_sysexit_disable_##sname(struct ftrace_event_call *event_call) \ +{ \ + if (atomic_add_negative(-1, &event_exit_##sname.profile_count)) \ + unreg_prof_syscall_exit("sys"#sname); \ +} + +#define TRACE_SYS_ENTER_PROFILE_INIT(sname) \ + .profile_count = ATOMIC_INIT(-1), \ + .profile_enable = prof_sysenter_enable_##sname, \ + .profile_disable = prof_sysenter_disable_##sname, + +#define TRACE_SYS_EXIT_PROFILE_INIT(sname) \ + .profile_count = ATOMIC_INIT(-1), \ + .profile_enable = prof_sysexit_enable_##sname, \ + .profile_disable = prof_sysexit_disable_##sname, +#else +#define TRACE_SYS_ENTER_PROFILE(sname) +#define TRACE_SYS_ENTER_PROFILE_INIT(sname) +#define TRACE_SYS_EXIT_PROFILE(sname) +#define TRACE_SYS_EXIT_PROFILE_INIT(sname) +#endif + #ifdef CONFIG_FTRACE_SYSCALLS #define __SC_STR_ADECL1(t, a) #a #define __SC_STR_ADECL2(t, a, ...) #a, __SC_STR_ADECL1(__VA_ARGS__) @@ -113,7 +160,6 @@ struct perf_counter_attr; #define __SC_STR_TDECL5(t, a, ...) #t, __SC_STR_TDECL4(__VA_ARGS__) #define __SC_STR_TDECL6(t, a, ...) #t, __SC_STR_TDECL5(__VA_ARGS__) - #define SYSCALL_TRACE_ENTER_EVENT(sname) \ static struct ftrace_event_call event_enter_##sname; \ struct trace_event enter_syscall_print_##sname = { \ @@ -134,6 +180,7 @@ struct perf_counter_attr; init_preds(&event_enter_##sname); \ return 0; \ } \ + TRACE_SYS_ENTER_PROFILE(sname); \ static struct ftrace_event_call __used \ __attribute__((__aligned__(4))) \ __attribute__((section("_ftrace_events"))) \ @@ -145,6 +192,7 @@ struct perf_counter_attr; .regfunc = reg_event_syscall_enter, \ .unregfunc = unreg_event_syscall_enter, \ .data = "sys"#sname, \ + TRACE_SYS_ENTER_PROFILE_INIT(sname) \ } #define SYSCALL_TRACE_EXIT_EVENT(sname) \ @@ -167,6 +215,7 @@ struct perf_counter_attr; init_preds(&event_exit_##sname); \ return 0; \ } \ + TRACE_SYS_EXIT_PROFILE(sname); \ static struct ftrace_event_call __used \ __attribute__((__aligned__(4))) \ __attribute__((section("_ftrace_events"))) \ @@ -178,6 +227,7 @@ struct perf_counter_attr; .regfunc = reg_event_syscall_exit, \ .unregfunc = unreg_event_syscall_exit, \ .data = "sys"#sname, \ + TRACE_SYS_EXIT_PROFILE_INIT(sname) \ } #define SYSCALL_METADATA(sname, nb) \ diff --git a/include/trace/syscall.h b/include/trace/syscall.h index df628404241a..3ab6dd18fa3a 100644 --- a/include/trace/syscall.h +++ b/include/trace/syscall.h @@ -57,6 +57,13 @@ extern int reg_event_syscall_exit(void *ptr); extern void unreg_event_syscall_exit(void *ptr); enum print_line_t print_syscall_enter(struct trace_iterator *iter, int flags); enum print_line_t print_syscall_exit(struct trace_iterator *iter, int flags); +#endif +#ifdef CONFIG_EVENT_PROFILE +int reg_prof_syscall_enter(char *name); +void unreg_prof_syscall_enter(char *name); +int reg_prof_syscall_exit(char *name); +void unreg_prof_syscall_exit(char *name); + #endif #endif /* _TRACE_SYSCALL_H */ diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c index e58a9c11ba85..f4eaec3d559a 100644 --- a/kernel/trace/trace_syscalls.c +++ b/kernel/trace/trace_syscalls.c @@ -1,6 +1,7 @@ #include #include #include +#include #include #include "trace_output.h" @@ -252,3 +253,123 @@ struct trace_event event_syscall_enter = { struct trace_event event_syscall_exit = { .trace = print_syscall_exit, }; + +#ifdef CONFIG_EVENT_PROFILE +static DECLARE_BITMAP(enabled_prof_enter_syscalls, FTRACE_SYSCALL_MAX); +static DECLARE_BITMAP(enabled_prof_exit_syscalls, FTRACE_SYSCALL_MAX); +static int sys_prof_refcount_enter; +static int sys_prof_refcount_exit; + +static void prof_syscall_enter(struct pt_regs *regs, long id) +{ + struct syscall_metadata *sys_data; + int syscall_nr; + + syscall_nr = syscall_get_nr(current, regs); + if (!test_bit(syscall_nr, enabled_prof_enter_syscalls)) + return; + + sys_data = syscall_nr_to_meta(syscall_nr); + if (!sys_data) + return; + + perf_tpcounter_event(sys_data->enter_id, 0, 1, NULL, 0); +} + +int reg_prof_syscall_enter(char *name) +{ + int ret = 0; + int num; + + num = syscall_name_to_nr(name); + if (num < 0 || num >= FTRACE_SYSCALL_MAX) + return -ENOSYS; + + mutex_lock(&syscall_trace_lock); + if (!sys_prof_refcount_enter) + ret = register_trace_syscall_enter(prof_syscall_enter); + if (ret) { + pr_info("event trace: Could not activate" + "syscall entry trace point"); + } else { + set_bit(num, enabled_prof_enter_syscalls); + sys_prof_refcount_enter++; + } + mutex_unlock(&syscall_trace_lock); + return ret; +} + +void unreg_prof_syscall_enter(char *name) +{ + int num; + + num = syscall_name_to_nr(name); + if (num < 0 || num >= FTRACE_SYSCALL_MAX) + return; + + mutex_lock(&syscall_trace_lock); + sys_prof_refcount_enter--; + clear_bit(num, enabled_prof_enter_syscalls); + if (!sys_prof_refcount_enter) + unregister_trace_syscall_enter(prof_syscall_enter); + mutex_unlock(&syscall_trace_lock); +} + +static void prof_syscall_exit(struct pt_regs *regs, long ret) +{ + struct syscall_metadata *sys_data; + int syscall_nr; + + syscall_nr = syscall_get_nr(current, regs); + if (!test_bit(syscall_nr, enabled_prof_exit_syscalls)) + return; + + sys_data = syscall_nr_to_meta(syscall_nr); + if (!sys_data) + return; + + perf_tpcounter_event(sys_data->exit_id, 0, 1, NULL, 0); +} + +int reg_prof_syscall_exit(char *name) +{ + int ret = 0; + int num; + + num = syscall_name_to_nr(name); + if (num < 0 || num >= FTRACE_SYSCALL_MAX) + return -ENOSYS; + + mutex_lock(&syscall_trace_lock); + if (!sys_prof_refcount_exit) + ret = register_trace_syscall_exit(prof_syscall_exit); + if (ret) { + pr_info("event trace: Could not activate" + "syscall entry trace point"); + } else { + set_bit(num, enabled_prof_exit_syscalls); + sys_prof_refcount_exit++; + } + mutex_unlock(&syscall_trace_lock); + return ret; +} + +void unreg_prof_syscall_exit(char *name) +{ + int num; + + num = syscall_name_to_nr(name); + if (num < 0 || num >= FTRACE_SYSCALL_MAX) + return; + + mutex_lock(&syscall_trace_lock); + sys_prof_refcount_exit--; + clear_bit(num, enabled_prof_exit_syscalls); + if (!sys_prof_refcount_exit) + unregister_trace_syscall_exit(prof_syscall_exit); + mutex_unlock(&syscall_trace_lock); +} + +#endif + + -- cgit v1.2.3 From e8f9f4d79a677f55c8ec3acbe87b33a87e2df0de Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Tue, 11 Aug 2009 17:42:52 +0200 Subject: tracing: Add ftrace event call parameter to its field descriptor handler Add the struct ftrace_event_call as a parameter of its show_format() callback. This way we can use it from the syscall trace events to retrieve the syscall name from the ftrace event call parameter and describe its fields using the syscalls metadata. Signed-off-by: Frederic Weisbecker Cc: Lai Jiangshan Cc: Steven Rostedt Cc: Peter Zijlstra Cc: Mathieu Desnoyers Cc: Jiaying Zhang Cc: Martin Bligh Cc: Li Zefan Cc: Masami Hiramatsu Cc: Jason Baron --- include/linux/ftrace_event.h | 3 ++- include/trace/ftrace.h | 3 ++- kernel/trace/trace_events.c | 2 +- kernel/trace/trace_export.c | 6 ++++-- 4 files changed, 9 insertions(+), 5 deletions(-) (limited to 'include') diff --git a/include/linux/ftrace_event.h b/include/linux/ftrace_event.h index 8544f121d9f1..189806b6e69e 100644 --- a/include/linux/ftrace_event.h +++ b/include/linux/ftrace_event.h @@ -116,7 +116,8 @@ struct ftrace_event_call { void (*unregfunc)(void *); int id; int (*raw_init)(void); - int (*show_format)(struct trace_seq *s); + int (*show_format)(struct ftrace_event_call *call, + struct trace_seq *s); int (*define_fields)(void); struct list_head fields; int filter_active; diff --git a/include/trace/ftrace.h b/include/trace/ftrace.h index 46d81b5e8610..b250b0616571 100644 --- a/include/trace/ftrace.h +++ b/include/trace/ftrace.h @@ -151,7 +151,8 @@ #undef TRACE_EVENT #define TRACE_EVENT(call, proto, args, tstruct, func, print) \ static int \ -ftrace_format_##call(struct trace_seq *s) \ +ftrace_format_##call(struct ftrace_event_call *unused, \ + struct trace_seq *s) \ { \ struct ftrace_raw_##call field __attribute__((unused)); \ int ret = 0; \ diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c index 1d289e2d6693..b568ade8f453 100644 --- a/kernel/trace/trace_events.c +++ b/kernel/trace/trace_events.c @@ -576,7 +576,7 @@ event_format_read(struct file *filp, char __user *ubuf, size_t cnt, trace_seq_printf(s, "format:\n"); trace_write_header(s); - r = call->show_format(s); + r = call->show_format(call, s); if (!r) { /* * ug! The format output is bigger than a PAGE!! diff --git a/kernel/trace/trace_export.c b/kernel/trace/trace_export.c index d06cf898dc86..956d4bc675e5 100644 --- a/kernel/trace/trace_export.c +++ b/kernel/trace/trace_export.c @@ -60,7 +60,8 @@ extern void __bad_type_size(void); #undef TRACE_EVENT_FORMAT #define TRACE_EVENT_FORMAT(call, proto, args, fmt, tstruct, tpfmt) \ static int \ -ftrace_format_##call(struct trace_seq *s) \ +ftrace_format_##call(struct ftrace_event_call *unused, \ + struct trace_seq *s) \ { \ struct args field; \ int ret; \ @@ -76,7 +77,8 @@ ftrace_format_##call(struct trace_seq *s) \ #define TRACE_EVENT_FORMAT_NOFILTER(call, proto, args, fmt, tstruct, \ tpfmt) \ static int \ -ftrace_format_##call(struct trace_seq *s) \ +ftrace_format_##call(struct ftrace_event_call *unused, \ + struct trace_seq *s) \ { \ struct args field; \ int ret; \ -- cgit v1.2.3 From dc4ddb4c0b7348f1c9759ae8a9e7d734dc1cda82 Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Tue, 11 Aug 2009 19:03:54 +0200 Subject: tracing: Add fields format definition for syscall events Define the format of the syscall trace fields to parse the binary values from a raw trace using the syscall events "format" file. This is defined dynamically using the syscalls metadata. It prepares the export of syscall event raw records to perf counters. Example: $ cat /debug/tracing/events/syscalls/sys_enter_sched_getparam/format name: sys_enter_sched_getparam ID: 39 format: field:unsigned short common_type; offset:0; size:2; field:unsigned char common_flags; offset:2; size:1; field:unsigned char common_preempt_count; offset:3; size:1; field:int common_pid; offset:4; size:4; field:int common_tgid; offset:8; size:4; field:pid_t pid; offset:12; size:8; field:struct sched_param * param; offset:20; size:8; print fmt: "pid: 0x%08lx, param: 0x%08lx", ((unsigned long)(REC->pid)), ((unsigned long)(REC->param)) Signed-off-by: Frederic Weisbecker Cc: Lai Jiangshan Cc: Steven Rostedt Cc: Peter Zijlstra Cc: Mathieu Desnoyers Cc: Jiaying Zhang Cc: Martin Bligh Cc: Li Zefan Cc: Masami Hiramatsu Cc: Jason Baron --- include/linux/syscalls.h | 1 + include/trace/syscall.h | 2 ++ kernel/trace/trace_syscalls.c | 46 +++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 49 insertions(+) (limited to 'include') diff --git a/include/linux/syscalls.h b/include/linux/syscalls.h index 5541e75e140a..87d06c173ddc 100644 --- a/include/linux/syscalls.h +++ b/include/linux/syscalls.h @@ -189,6 +189,7 @@ static void prof_sysexit_disable_##sname(struct ftrace_event_call *event_call) \ .system = "syscalls", \ .event = &event_syscall_enter, \ .raw_init = init_enter_##sname, \ + .show_format = ftrace_format_syscall, \ .regfunc = reg_event_syscall_enter, \ .unregfunc = unreg_event_syscall_enter, \ .data = "sys"#sname, \ diff --git a/include/trace/syscall.h b/include/trace/syscall.h index 3ab6dd18fa3a..0cb03625edd9 100644 --- a/include/trace/syscall.h +++ b/include/trace/syscall.h @@ -55,6 +55,8 @@ extern int reg_event_syscall_enter(void *ptr); extern void unreg_event_syscall_enter(void *ptr); extern int reg_event_syscall_exit(void *ptr); extern void unreg_event_syscall_exit(void *ptr); +extern int +ftrace_format_syscall(struct ftrace_event_call *call, struct trace_seq *s); enum print_line_t print_syscall_enter(struct trace_iterator *iter, int flags); enum print_line_t print_syscall_exit(struct trace_iterator *iter, int flags); #endif diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c index f4eaec3d559a..9ee6386cf842 100644 --- a/kernel/trace/trace_syscalls.c +++ b/kernel/trace/trace_syscalls.c @@ -105,6 +105,52 @@ print_syscall_exit(struct trace_iterator *iter, int flags) return TRACE_TYPE_HANDLED; } +int ftrace_format_syscall(struct ftrace_event_call *call, struct trace_seq *s) +{ + int i; + int nr; + int ret = 0; + struct syscall_metadata *entry; + int offset = sizeof(struct trace_entry); + + nr = syscall_name_to_nr((char *)call->data); + entry = syscall_nr_to_meta(nr); + + if (!entry) + return ret; + + for (i = 0; i < entry->nb_args; i++) { + ret = trace_seq_printf(s, "\tfield:%s %s;", entry->types[i], + entry->args[i]); + if (!ret) + return 0; + ret = trace_seq_printf(s, "\toffset:%d;\tsize:%lu;\n", offset, + sizeof(unsigned long)); + if (!ret) + return 0; + offset += sizeof(unsigned long); + } + + trace_seq_printf(s, "\nprint fmt: \""); + for (i = 0; i < entry->nb_args; i++) { + ret = trace_seq_printf(s, "%s: 0x%%0%lulx%s", entry->args[i], + sizeof(unsigned long), + i == entry->nb_args - 1 ? "\", " : ", "); + if (!ret) + return 0; + } + + for (i = 0; i < entry->nb_args; i++) { + ret = trace_seq_printf(s, "((unsigned long)(REC->%s))%s", + entry->args[i], + i == entry->nb_args - 1 ? "\n" : ", "); + if (!ret) + return 0; + } + + return ret; +} + void ftrace_syscall_enter(struct pt_regs *regs, long id) { struct syscall_trace_enter *entry; -- cgit v1.2.3 From 25a70e38cd57952b09a013bf070e03705ee4f2ff Mon Sep 17 00:00:00 2001 From: Daniel Mack Date: Wed, 12 Aug 2009 00:50:09 -0700 Subject: Input: eeti_ts - allow active high irq lines This adds a struct eeti_ts_platform_data which currently holds only one value to specify the interrupt polarity. The driver has a fallback if no platform data is passed in via the i2c_board_info, so no regression is caused. Signed-off-by: Daniel Mack Signed-off-by: Dmitry Torokhov --- drivers/input/touchscreen/eeti_ts.c | 22 +++++++++++++++++++--- include/linux/input/eeti_ts.h | 9 +++++++++ 2 files changed, 28 insertions(+), 3 deletions(-) create mode 100644 include/linux/input/eeti_ts.h (limited to 'include') diff --git a/drivers/input/touchscreen/eeti_ts.c b/drivers/input/touchscreen/eeti_ts.c index 3ab92222a525..9029bd3f34e5 100644 --- a/drivers/input/touchscreen/eeti_ts.c +++ b/drivers/input/touchscreen/eeti_ts.c @@ -32,6 +32,7 @@ #include #include #include +#include static int flip_x; module_param(flip_x, bool, 0644); @@ -46,7 +47,7 @@ struct eeti_ts_priv { struct input_dev *input; struct work_struct work; struct mutex mutex; - int irq; + int irq, irq_active_high; }; #define EETI_TS_BITDEPTH (11) @@ -58,6 +59,11 @@ struct eeti_ts_priv { #define REPORT_BIT_HAS_PRESSURE (1 << 6) #define REPORT_RES_BITS(v) (((v) >> 1) + EETI_TS_BITDEPTH) +static inline int eeti_ts_irq_active(struct eeti_ts_priv *priv) +{ + return gpio_get_value(irq_to_gpio(priv->irq)) == priv->irq_active_high; +} + static void eeti_ts_read(struct work_struct *work) { char buf[6]; @@ -67,7 +73,7 @@ static void eeti_ts_read(struct work_struct *work) mutex_lock(&priv->mutex); - while (!gpio_get_value(irq_to_gpio(priv->irq)) && --to) + while (eeti_ts_irq_active(priv) && --to) i2c_master_recv(priv->client, buf, sizeof(buf)); if (!to) { @@ -140,8 +146,10 @@ static void eeti_ts_close(struct input_dev *dev) static int __devinit eeti_ts_probe(struct i2c_client *client, const struct i2c_device_id *idp) { + struct eeti_ts_platform_data *pdata; struct eeti_ts_priv *priv; struct input_dev *input; + unsigned int irq_flags; int err = -ENOMEM; /* In contrast to what's described in the datasheet, there seems @@ -180,6 +188,14 @@ static int __devinit eeti_ts_probe(struct i2c_client *client, priv->input = input; priv->irq = client->irq; + pdata = client->dev.platform_data; + + if (pdata) + priv->irq_active_high = pdata->irq_active_high; + + irq_flags = priv->irq_active_high ? + IRQF_TRIGGER_RISING : IRQF_TRIGGER_FALLING; + INIT_WORK(&priv->work, eeti_ts_read); i2c_set_clientdata(client, priv); input_set_drvdata(input, priv); @@ -188,7 +204,7 @@ static int __devinit eeti_ts_probe(struct i2c_client *client, if (err) goto err1; - err = request_irq(priv->irq, eeti_ts_isr, IRQF_TRIGGER_FALLING, + err = request_irq(priv->irq, eeti_ts_isr, irq_flags, client->name, priv); if (err) { dev_err(&client->dev, "Unable to request touchscreen IRQ.\n"); diff --git a/include/linux/input/eeti_ts.h b/include/linux/input/eeti_ts.h new file mode 100644 index 000000000000..f875b316249d --- /dev/null +++ b/include/linux/input/eeti_ts.h @@ -0,0 +1,9 @@ +#ifndef LINUX_INPUT_EETI_TS_H +#define LINUX_INPUT_EETI_TS_H + +struct eeti_ts_platform_data { + unsigned int irq_active_high; +}; + +#endif /* LINUX_INPUT_EETI_TS_H */ + -- cgit v1.2.3 From 1ae88b2e446261c038f2c0c3150ffae142b227a2 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Wed, 12 Aug 2009 09:12:30 -0400 Subject: NFS: Fix an O_DIRECT Oops... We can't call nfs_readdata_release()/nfs_writedata_release() without first initialising and referencing args.context. Doing so inside nfs_direct_read_schedule_segment()/nfs_direct_write_schedule_segment() causes an Oops. We should rather be calling nfs_readdata_free()/nfs_writedata_free() in those cases. Looking at the O_DIRECT code, the "struct nfs_direct_req" is already referencing the nfs_open_context for us. Since the readdata and writedata structures carry a reference to that, we can simplify things by getting rid of the extra nfs_open_context references, so that we can replace all instances of nfs_readdata_release()/nfs_writedata_release(). Reported-by: Catalin Marinas Signed-off-by: Trond Myklebust Tested-by: Catalin Marinas Cc: stable@kernel.org Signed-off-by: Linus Torvalds --- fs/nfs/direct.c | 20 ++++++++++---------- fs/nfs/read.c | 6 ++---- fs/nfs/write.c | 6 ++---- include/linux/nfs_fs.h | 5 ++--- 4 files changed, 16 insertions(+), 21 deletions(-) (limited to 'include') diff --git a/fs/nfs/direct.c b/fs/nfs/direct.c index 489fc01a3204..e4e089a8f294 100644 --- a/fs/nfs/direct.c +++ b/fs/nfs/direct.c @@ -255,7 +255,7 @@ static void nfs_direct_read_release(void *calldata) if (put_dreq(dreq)) nfs_direct_complete(dreq); - nfs_readdata_release(calldata); + nfs_readdata_free(data); } static const struct rpc_call_ops nfs_read_direct_ops = { @@ -314,14 +314,14 @@ static ssize_t nfs_direct_read_schedule_segment(struct nfs_direct_req *dreq, data->npages, 1, 0, data->pagevec, NULL); up_read(¤t->mm->mmap_sem); if (result < 0) { - nfs_readdata_release(data); + nfs_readdata_free(data); break; } if ((unsigned)result < data->npages) { bytes = result * PAGE_SIZE; if (bytes <= pgbase) { nfs_direct_release_pages(data->pagevec, result); - nfs_readdata_release(data); + nfs_readdata_free(data); break; } bytes -= pgbase; @@ -334,7 +334,7 @@ static ssize_t nfs_direct_read_schedule_segment(struct nfs_direct_req *dreq, data->inode = inode; data->cred = msg.rpc_cred; data->args.fh = NFS_FH(inode); - data->args.context = get_nfs_open_context(ctx); + data->args.context = ctx; data->args.offset = pos; data->args.pgbase = pgbase; data->args.pages = data->pagevec; @@ -441,7 +441,7 @@ static void nfs_direct_free_writedata(struct nfs_direct_req *dreq) struct nfs_write_data *data = list_entry(dreq->rewrite_list.next, struct nfs_write_data, pages); list_del(&data->pages); nfs_direct_release_pages(data->pagevec, data->npages); - nfs_writedata_release(data); + nfs_writedata_free(data); } } @@ -534,7 +534,7 @@ static void nfs_direct_commit_release(void *calldata) dprintk("NFS: %5u commit returned %d\n", data->task.tk_pid, status); nfs_direct_write_complete(dreq, data->inode); - nfs_commitdata_release(calldata); + nfs_commit_free(data); } static const struct rpc_call_ops nfs_commit_direct_ops = { @@ -570,7 +570,7 @@ static void nfs_direct_commit_schedule(struct nfs_direct_req *dreq) data->args.fh = NFS_FH(data->inode); data->args.offset = 0; data->args.count = 0; - data->args.context = get_nfs_open_context(dreq->ctx); + data->args.context = dreq->ctx; data->res.count = 0; data->res.fattr = &data->fattr; data->res.verf = &data->verf; @@ -734,14 +734,14 @@ static ssize_t nfs_direct_write_schedule_segment(struct nfs_direct_req *dreq, data->npages, 0, 0, data->pagevec, NULL); up_read(¤t->mm->mmap_sem); if (result < 0) { - nfs_writedata_release(data); + nfs_writedata_free(data); break; } if ((unsigned)result < data->npages) { bytes = result * PAGE_SIZE; if (bytes <= pgbase) { nfs_direct_release_pages(data->pagevec, result); - nfs_writedata_release(data); + nfs_writedata_free(data); break; } bytes -= pgbase; @@ -756,7 +756,7 @@ static ssize_t nfs_direct_write_schedule_segment(struct nfs_direct_req *dreq, data->inode = inode; data->cred = msg.rpc_cred; data->args.fh = NFS_FH(inode); - data->args.context = get_nfs_open_context(ctx); + data->args.context = ctx; data->args.offset = pos; data->args.pgbase = pgbase; data->args.pages = data->pagevec; diff --git a/fs/nfs/read.c b/fs/nfs/read.c index 73ea5e8d66ce..12c9e66d3f1d 100644 --- a/fs/nfs/read.c +++ b/fs/nfs/read.c @@ -60,17 +60,15 @@ struct nfs_read_data *nfs_readdata_alloc(unsigned int pagecount) return p; } -static void nfs_readdata_free(struct nfs_read_data *p) +void nfs_readdata_free(struct nfs_read_data *p) { if (p && (p->pagevec != &p->page_array[0])) kfree(p->pagevec); mempool_free(p, nfs_rdata_mempool); } -void nfs_readdata_release(void *data) +static void nfs_readdata_release(struct nfs_read_data *rdata) { - struct nfs_read_data *rdata = data; - put_nfs_open_context(rdata->args.context); nfs_readdata_free(rdata); } diff --git a/fs/nfs/write.c b/fs/nfs/write.c index 0a0a2ff767c3..a34fae21fe10 100644 --- a/fs/nfs/write.c +++ b/fs/nfs/write.c @@ -87,17 +87,15 @@ struct nfs_write_data *nfs_writedata_alloc(unsigned int pagecount) return p; } -static void nfs_writedata_free(struct nfs_write_data *p) +void nfs_writedata_free(struct nfs_write_data *p) { if (p && (p->pagevec != &p->page_array[0])) kfree(p->pagevec); mempool_free(p, nfs_wdata_mempool); } -void nfs_writedata_release(void *data) +static void nfs_writedata_release(struct nfs_write_data *wdata) { - struct nfs_write_data *wdata = data; - put_nfs_open_context(wdata->args.context); nfs_writedata_free(wdata); } diff --git a/include/linux/nfs_fs.h b/include/linux/nfs_fs.h index fdffb413b192..f6b90240dd41 100644 --- a/include/linux/nfs_fs.h +++ b/include/linux/nfs_fs.h @@ -473,7 +473,6 @@ extern int nfs_writepages(struct address_space *, struct writeback_control *); extern int nfs_flush_incompatible(struct file *file, struct page *page); extern int nfs_updatepage(struct file *, struct page *, unsigned int, unsigned int); extern int nfs_writeback_done(struct rpc_task *, struct nfs_write_data *); -extern void nfs_writedata_release(void *); /* * Try to write back everything synchronously (but check the @@ -488,7 +487,6 @@ extern int nfs_wb_page_cancel(struct inode *inode, struct page* page); extern int nfs_commit_inode(struct inode *, int); extern struct nfs_write_data *nfs_commitdata_alloc(void); extern void nfs_commit_free(struct nfs_write_data *wdata); -extern void nfs_commitdata_release(void *wdata); #else static inline int nfs_commit_inode(struct inode *inode, int how) @@ -507,6 +505,7 @@ nfs_have_writebacks(struct inode *inode) * Allocate nfs_write_data structures */ extern struct nfs_write_data *nfs_writedata_alloc(unsigned int npages); +extern void nfs_writedata_free(struct nfs_write_data *); /* * linux/fs/nfs/read.c @@ -515,7 +514,6 @@ extern int nfs_readpage(struct file *, struct page *); extern int nfs_readpages(struct file *, struct address_space *, struct list_head *, unsigned); extern int nfs_readpage_result(struct rpc_task *, struct nfs_read_data *); -extern void nfs_readdata_release(void *data); extern int nfs_readpage_async(struct nfs_open_context *, struct inode *, struct page *); @@ -523,6 +521,7 @@ extern int nfs_readpage_async(struct nfs_open_context *, struct inode *, * Allocate nfs_read_data structures */ extern struct nfs_read_data *nfs_readdata_alloc(unsigned int npages); +extern void nfs_readdata_free(struct nfs_read_data *); /* * linux/fs/nfs3proc.c -- cgit v1.2.3 From a75f0236292a5fca65f26efedca48bd07db1834d Mon Sep 17 00:00:00 2001 From: Francisco Jerez Date: Wed, 12 Aug 2009 02:30:10 +0200 Subject: drm: Add more standard TV properties. Overscan, saturation, hue. Used in the nouveau driver for GPUs with integrated TV encoders. Signed-off-by: Francisco Jerez Signed-off-by: Dave Airlie --- drivers/gpu/drm/drm_crtc.c | 18 ++++++++++++++++++ include/drm/drm_crtc.h | 3 +++ 2 files changed, 21 insertions(+) (limited to 'include') diff --git a/drivers/gpu/drm/drm_crtc.c b/drivers/gpu/drm/drm_crtc.c index a8c831134fc3..362a538cdedc 100644 --- a/drivers/gpu/drm/drm_crtc.c +++ b/drivers/gpu/drm/drm_crtc.c @@ -736,6 +736,24 @@ int drm_mode_create_tv_properties(struct drm_device *dev, int num_modes, dev->mode_config.tv_flicker_reduction_property->values[0] = 0; dev->mode_config.tv_flicker_reduction_property->values[1] = 100; + dev->mode_config.tv_overscan_property = + drm_property_create(dev, DRM_MODE_PROP_RANGE, + "overscan", 2); + dev->mode_config.tv_overscan_property->values[0] = 0; + dev->mode_config.tv_overscan_property->values[1] = 100; + + dev->mode_config.tv_saturation_property = + drm_property_create(dev, DRM_MODE_PROP_RANGE, + "saturation", 2); + dev->mode_config.tv_saturation_property->values[0] = 0; + dev->mode_config.tv_saturation_property->values[1] = 100; + + dev->mode_config.tv_hue_property = + drm_property_create(dev, DRM_MODE_PROP_RANGE, + "hue", 2); + dev->mode_config.tv_hue_property->values[0] = 0; + dev->mode_config.tv_hue_property->values[1] = 100; + return 0; } EXPORT_SYMBOL(drm_mode_create_tv_properties); diff --git a/include/drm/drm_crtc.h b/include/drm/drm_crtc.h index 5f2cc0ca4c7d..db92a83f8ca9 100644 --- a/include/drm/drm_crtc.h +++ b/include/drm/drm_crtc.h @@ -575,6 +575,9 @@ struct drm_mode_config { struct drm_property *tv_brightness_property; struct drm_property *tv_contrast_property; struct drm_property *tv_flicker_reduction_property; + struct drm_property *tv_overscan_property; + struct drm_property *tv_saturation_property; + struct drm_property *tv_hue_property; /* Optional properties */ struct drm_property *scaling_mode_property; -- cgit v1.2.3 From 8505091d2a067ad27d4a82df9cff8eae6ee52fca Mon Sep 17 00:00:00 2001 From: Dmitry Baryshkov Date: Fri, 7 Aug 2009 02:58:36 +0000 Subject: af_ieee802154: drop IEEE802154_SIOC_ADD_SLAVE declaration IEEE802154_SIOC_ADD_SLAVE was used to allocate 802.15.4 interfaces on the top of radio. It's not used anymore, drop it. Signed-off-by: Dmitry Eremin-Solenikov Signed-off-by: David S. Miller --- include/net/af_ieee802154.h | 3 --- 1 file changed, 3 deletions(-) (limited to 'include') diff --git a/include/net/af_ieee802154.h b/include/net/af_ieee802154.h index 0d78605fb1a6..e9c70ead23fc 100644 --- a/include/net/af_ieee802154.h +++ b/include/net/af_ieee802154.h @@ -54,7 +54,4 @@ struct sockaddr_ieee802154 { struct ieee802154_addr addr; }; -/* master device */ -#define IEEE802154_SIOC_ADD_SLAVE (SIOCDEVPRIVATE + 0) - #endif -- cgit v1.2.3 From 78090a58c49f2f6213d0bb1b3b4c4df73e26865f Mon Sep 17 00:00:00 2001 From: Dmitry Baryshkov Date: Fri, 7 Aug 2009 02:58:38 +0000 Subject: nl802154: make ieee802154_policy constant Signed-off-by: Dmitry Eremin-Solenikov Signed-off-by: David S. Miller --- include/linux/nl802154.h | 2 +- net/ieee802154/nl_policy.c | 3 ++- 2 files changed, 3 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/include/linux/nl802154.h b/include/linux/nl802154.h index 2cda00ccfcca..266dd96ced96 100644 --- a/include/linux/nl802154.h +++ b/include/linux/nl802154.h @@ -69,7 +69,7 @@ enum { #define IEEE802154_ATTR_MAX (__IEEE802154_ATTR_MAX - 1) -extern struct nla_policy ieee802154_policy[]; +extern const struct nla_policy ieee802154_policy[]; /* commands */ /* REQ should be responded with CONF diff --git a/net/ieee802154/nl_policy.c b/net/ieee802154/nl_policy.c index c7d71d1adcac..83cb4ccef90d 100644 --- a/net/ieee802154/nl_policy.c +++ b/net/ieee802154/nl_policy.c @@ -24,7 +24,7 @@ #define NLA_HW_ADDR NLA_U64 -struct nla_policy ieee802154_policy[IEEE802154_ATTR_MAX + 1] = { +const struct nla_policy ieee802154_policy[IEEE802154_ATTR_MAX + 1] = { [IEEE802154_ATTR_DEV_NAME] = { .type = NLA_STRING, }, [IEEE802154_ATTR_DEV_INDEX] = { .type = NLA_U32, }, @@ -50,3 +50,4 @@ struct nla_policy ieee802154_policy[IEEE802154_ATTR_MAX + 1] = { [IEEE802154_ATTR_DURATION] = { .type = NLA_U8, }, [IEEE802154_ATTR_ED_LIST] = { .len = 27 }, }; + -- cgit v1.2.3 From 8e753dd0a82bd266256c20a20b98dfa48f98d21e Mon Sep 17 00:00:00 2001 From: Dmitry Baryshkov Date: Fri, 7 Aug 2009 02:58:40 +0000 Subject: nl802154: add support for dumping WPAN interface information Signed-off-by: Dmitry Eremin-Solenikov Signed-off-by: David S. Miller --- include/linux/nl802154.h | 2 + net/ieee802154/netlink.c | 106 ++++++++++++++++++++++++++++++++++++++++++++++- 2 files changed, 106 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/include/linux/nl802154.h b/include/linux/nl802154.h index 266dd96ced96..9a1af5f871a3 100644 --- a/include/linux/nl802154.h +++ b/include/linux/nl802154.h @@ -111,6 +111,8 @@ enum { IEEE802154_RX_ENABLE_REQ, /* Not supported yet */ IEEE802154_RX_ENABLE_CONF, /* Not supported yet */ + IEEE802154_LIST_IFACE, + __IEEE802154_CMD_MAX, }; diff --git a/net/ieee802154/netlink.c b/net/ieee802154/netlink.c index a615b9d13212..3fe02b394ad6 100644 --- a/net/ieee802154/netlink.c +++ b/net/ieee802154/netlink.c @@ -19,6 +19,7 @@ * Written by: * Sergey Lapin * Dmitry Eremin-Solenikov + * Maxim Osipov */ #include @@ -26,6 +27,7 @@ #include #include #include +#include #include #include #include @@ -73,7 +75,7 @@ static int ieee802154_nl_finish(struct sk_buff *msg) /* XXX: nlh is right at the start of msg */ void *hdr = genlmsg_data(NLMSG_DATA(msg->data)); - if (!genlmsg_end(msg, hdr)) + if (genlmsg_end(msg, hdr) < 0) goto out; return genlmsg_multicast(msg, 0, ieee802154_coord_mcgrp.id, @@ -260,6 +262,35 @@ nla_put_failure: } EXPORT_SYMBOL(ieee802154_nl_scan_confirm); +static int ieee802154_nl_fill_iface(struct sk_buff *msg, u32 pid, + u32 seq, int flags, struct net_device *dev) +{ + void *hdr; + + pr_debug("%s\n", __func__); + + hdr = genlmsg_put(msg, 0, seq, &ieee802154_coordinator_family, flags, + IEEE802154_LIST_IFACE); + if (!hdr) + goto out; + + NLA_PUT_STRING(msg, IEEE802154_ATTR_DEV_NAME, dev->name); + NLA_PUT_U32(msg, IEEE802154_ATTR_DEV_INDEX, dev->ifindex); + + NLA_PUT(msg, IEEE802154_ATTR_HW_ADDR, IEEE802154_ADDR_LEN, + dev->dev_addr); + NLA_PUT_U16(msg, IEEE802154_ATTR_SHORT_ADDR, + ieee802154_mlme_ops(dev)->get_short_addr(dev)); + NLA_PUT_U16(msg, IEEE802154_ATTR_PAN_ID, + ieee802154_mlme_ops(dev)->get_pan_id(dev)); + return genlmsg_end(msg, hdr); + +nla_put_failure: + genlmsg_cancel(msg, hdr); +out: + return -EMSGSIZE; +} + /* Requests from userspace */ static struct net_device *ieee802154_nl_get_dev(struct genl_info *info) { @@ -272,7 +303,7 @@ static struct net_device *ieee802154_nl_get_dev(struct genl_info *info) dev = dev_get_by_name(&init_net, name); } else if (info->attrs[IEEE802154_ATTR_DEV_INDEX]) dev = dev_get_by_index(&init_net, - nla_get_u32(info->attrs[IEEE802154_ATTR_DEV_INDEX])); + nla_get_u32(info->attrs[IEEE802154_ATTR_DEV_INDEX])); else return NULL; @@ -466,6 +497,67 @@ static int ieee802154_scan_req(struct sk_buff *skb, struct genl_info *info) return ret; } +static int ieee802154_list_iface(struct sk_buff *skb, + struct genl_info *info) +{ + /* Request for interface name, index, type, IEEE address, + PAN Id, short address */ + struct sk_buff *msg; + struct net_device *dev = NULL; + int rc = -ENOBUFS; + + pr_debug("%s\n", __func__); + + dev = ieee802154_nl_get_dev(info); + if (!dev) + return -ENODEV; + + msg = nlmsg_new(NLMSG_GOODSIZE, GFP_KERNEL); + if (!msg) + goto out_dev; + + rc = ieee802154_nl_fill_iface(msg, info->snd_pid, info->snd_seq, + 0, dev); + if (rc < 0) + goto out_free; + + dev_put(dev); + + return genlmsg_unicast(&init_net, msg, info->snd_pid); +out_free: + nlmsg_free(msg); +out_dev: + dev_put(dev); + return rc; + +} + +static int ieee802154_dump_iface(struct sk_buff *skb, + struct netlink_callback *cb) +{ + struct net *net = sock_net(skb->sk); + struct net_device *dev; + int idx; + int s_idx = cb->args[0]; + + pr_debug("%s\n", __func__); + + idx = 0; + for_each_netdev(net, dev) { + if (idx < s_idx || (dev->type != ARPHRD_IEEE802154)) + goto cont; + + if (ieee802154_nl_fill_iface(skb, NETLINK_CB(cb->skb).pid, + cb->nlh->nlmsg_seq, NLM_F_MULTI, dev) < 0) + break; +cont: + idx++; + } + cb->args[0] = idx; + + return skb->len; +} + #define IEEE802154_OP(_cmd, _func) \ { \ .cmd = _cmd, \ @@ -475,12 +567,22 @@ static int ieee802154_scan_req(struct sk_buff *skb, struct genl_info *info) .flags = GENL_ADMIN_PERM, \ } +#define IEEE802154_DUMP(_cmd, _func, _dump) \ + { \ + .cmd = _cmd, \ + .policy = ieee802154_policy, \ + .doit = _func, \ + .dumpit = _dump, \ + } + static struct genl_ops ieee802154_coordinator_ops[] = { IEEE802154_OP(IEEE802154_ASSOCIATE_REQ, ieee802154_associate_req), IEEE802154_OP(IEEE802154_ASSOCIATE_RESP, ieee802154_associate_resp), IEEE802154_OP(IEEE802154_DISASSOCIATE_REQ, ieee802154_disassociate_req), IEEE802154_OP(IEEE802154_SCAN_REQ, ieee802154_scan_req), IEEE802154_OP(IEEE802154_START_REQ, ieee802154_start_req), + IEEE802154_DUMP(IEEE802154_LIST_IFACE, ieee802154_list_iface, + ieee802154_dump_iface), }; static int __init ieee802154_nl_init(void) -- cgit v1.2.3 From 99eb8558642b988055d2b8b16a334475550f78d3 Mon Sep 17 00:00:00 2001 From: Dmitry Baryshkov Date: Fri, 7 Aug 2009 02:58:43 +0000 Subject: af_ieee802154: add support for WANT_ACK socket option Signed-off-by: Dmitry Eremin-Solenikov Signed-off-by: David S. Miller --- include/net/af_ieee802154.h | 5 ++++ net/ieee802154/dgram.c | 58 +++++++++++++++++++++++++++++++++++++++++---- 2 files changed, 59 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/include/net/af_ieee802154.h b/include/net/af_ieee802154.h index e9c70ead23fc..75e64c7a2960 100644 --- a/include/net/af_ieee802154.h +++ b/include/net/af_ieee802154.h @@ -54,4 +54,9 @@ struct sockaddr_ieee802154 { struct ieee802154_addr addr; }; +/* get/setsockopt */ +#define SOL_IEEE802154 0 + +#define WPAN_WANTACK 0 + #endif diff --git a/net/ieee802154/dgram.c b/net/ieee802154/dgram.c index 25018a9a53aa..77ae6852b93d 100644 --- a/net/ieee802154/dgram.c +++ b/net/ieee802154/dgram.c @@ -44,6 +44,7 @@ struct dgram_sock { struct ieee802154_addr dst_addr; unsigned bound:1; + unsigned want_ack:1; }; static inline struct dgram_sock *dgram_sk(const struct sock *sk) @@ -51,7 +52,6 @@ static inline struct dgram_sock *dgram_sk(const struct sock *sk) return container_of(sk, struct dgram_sock, sk); } - static void dgram_hash(struct sock *sk) { write_lock_bh(&dgram_lock); @@ -74,6 +74,7 @@ static int dgram_init(struct sock *sk) ro->dst_addr.addr_type = IEEE802154_ADDR_LONG; ro->dst_addr.pan_id = 0xffff; + ro->want_ack = 1; memset(&ro->dst_addr.hwaddr, 0xff, sizeof(ro->dst_addr.hwaddr)); return 0; } @@ -237,7 +238,10 @@ static int dgram_sendmsg(struct kiocb *iocb, struct sock *sk, skb_reset_network_header(skb); - mac_cb(skb)->flags = IEEE802154_FC_TYPE_DATA | MAC_CB_FLAG_ACKREQ; + mac_cb(skb)->flags = IEEE802154_FC_TYPE_DATA; + if (ro->want_ack) + mac_cb(skb)->flags |= MAC_CB_FLAG_ACKREQ; + mac_cb(skb)->seq = ieee802154_mlme_ops(dev)->get_dsn(dev); err = dev_hard_header(skb, dev, ETH_P_IEEE802154, &ro->dst_addr, ro->bound ? &ro->src_addr : NULL, size); @@ -382,13 +386,59 @@ int ieee802154_dgram_deliver(struct net_device *dev, struct sk_buff *skb) static int dgram_getsockopt(struct sock *sk, int level, int optname, char __user *optval, int __user *optlen) { - return -EOPNOTSUPP; + struct dgram_sock *ro = dgram_sk(sk); + + int val, len; + + if (level != SOL_IEEE802154) + return -EOPNOTSUPP; + + if (get_user(len, optlen)) + return -EFAULT; + + len = min_t(unsigned int, len, sizeof(int)); + + switch (optname) { + case WPAN_WANTACK: + val = ro->want_ack; + break; + default: + return -ENOPROTOOPT; + } + + if (put_user(len, optlen)) + return -EFAULT; + if (copy_to_user(optval, &val, len)) + return -EFAULT; + return 0; } static int dgram_setsockopt(struct sock *sk, int level, int optname, char __user *optval, int __user optlen) { - return -EOPNOTSUPP; + struct dgram_sock *ro = dgram_sk(sk); + int val; + int err = 0; + + if (optlen < sizeof(int)) + return -EINVAL; + + if (get_user(val, (int __user *)optval)) + return -EFAULT; + + lock_sock(sk); + + switch (optname) { + case WPAN_WANTACK: + ro->want_ack = !!val; + break; + default: + err = -ENOPROTOOPT; + break; + } + + release_sock(sk); + return err; } struct proto ieee802154_dgram_prot = { -- cgit v1.2.3 From acb8aacda3f0bc3aeb652f4365c078a2b0adb0bf Mon Sep 17 00:00:00 2001 From: Dmitry Baryshkov Date: Fri, 7 Aug 2009 02:58:44 +0000 Subject: nl802154: support START-CONFIRM primitive Signed-off-by: Dmitry Eremin-Solenikov Signed-off-by: David S. Miller --- include/net/nl802154.h | 9 +++++++++ net/ieee802154/netlink.c | 32 ++++++++++++++++++++++++++++++++ 2 files changed, 41 insertions(+) (limited to 'include') diff --git a/include/net/nl802154.h b/include/net/nl802154.h index 6096096f6d7d..e554ecd3727a 100644 --- a/include/net/nl802154.h +++ b/include/net/nl802154.h @@ -114,4 +114,13 @@ int ieee802154_nl_scan_confirm(struct net_device *dev, int ieee802154_nl_beacon_indic(struct net_device *dev, u16 panid, u16 coord_addr); +/** + * ieee802154_nl_start_confirm - Notify userland of completion of start. + * @dev: The device which was instructed to scan. + * @status: The status of the scan operation. + * + * Note: This is in section 7.1.14 of the IEEE 802.15.4 document. + */ +int ieee802154_nl_start_confirm(struct net_device *dev, u8 status); + #endif diff --git a/net/ieee802154/netlink.c b/net/ieee802154/netlink.c index 3fe02b394ad6..cd0567f06716 100644 --- a/net/ieee802154/netlink.c +++ b/net/ieee802154/netlink.c @@ -31,6 +31,7 @@ #include #include #include +#include #include static unsigned int ieee802154_seq_num; @@ -262,6 +263,31 @@ nla_put_failure: } EXPORT_SYMBOL(ieee802154_nl_scan_confirm); +int ieee802154_nl_start_confirm(struct net_device *dev, u8 status) +{ + struct sk_buff *msg; + + pr_debug("%s\n", __func__); + + msg = ieee802154_nl_create(0, IEEE802154_START_CONF); + if (!msg) + return -ENOBUFS; + + NLA_PUT_STRING(msg, IEEE802154_ATTR_DEV_NAME, dev->name); + NLA_PUT_U32(msg, IEEE802154_ATTR_DEV_INDEX, dev->ifindex); + NLA_PUT(msg, IEEE802154_ATTR_HW_ADDR, IEEE802154_ADDR_LEN, + dev->dev_addr); + + NLA_PUT_U8(msg, IEEE802154_ATTR_STATUS, status); + + return ieee802154_nl_finish(msg); + +nla_put_failure: + nlmsg_free(msg); + return -ENOBUFS; +} +EXPORT_SYMBOL(ieee802154_nl_start_confirm); + static int ieee802154_nl_fill_iface(struct sk_buff *msg, u32 pid, u32 seq, int flags, struct net_device *dev) { @@ -462,6 +488,12 @@ static int ieee802154_start_req(struct sk_buff *skb, struct genl_info *info) blx = nla_get_u8(info->attrs[IEEE802154_ATTR_BAT_EXT]); coord_realign = nla_get_u8(info->attrs[IEEE802154_ATTR_COORD_REALIGN]); + if (addr.short_addr == IEEE802154_ADDR_BROADCAST) { + ieee802154_nl_start_confirm(dev, IEEE802154_NO_SHORT_ADDRESS); + dev_put(dev); + return -EINVAL; + } + ret = ieee802154_mlme_ops(dev)->start_req(dev, &addr, channel, bcn_ord, sf_ord, pan_coord, blx, coord_realign); -- cgit v1.2.3 From aed7df86983ead0af8364fd26d8a9609ef2ed584 Mon Sep 17 00:00:00 2001 From: Jaswinder Singh Rajput Date: Sun, 9 Aug 2009 03:27:18 +0000 Subject: net: include/linux/icmpv6: includecheck fix for icmpv6.h fix the following 'make includecheck' warning: include/linux/icmpv6.h: linux/skbuff.h is included more than once. Signed-off-by: Jaswinder Singh Rajput Signed-off-by: David S. Miller --- include/linux/icmpv6.h | 2 -- 1 file changed, 2 deletions(-) (limited to 'include') diff --git a/include/linux/icmpv6.h b/include/linux/icmpv6.h index b6a85183c333..c0d8357917e2 100644 --- a/include/linux/icmpv6.h +++ b/include/linux/icmpv6.h @@ -171,8 +171,6 @@ struct icmp6_filter { #ifdef __KERNEL__ #include -#include - extern void icmpv6_send(struct sk_buff *skb, u8 type, u8 code, -- cgit v1.2.3 From 2e955856ff1212bd63dbbf403940c72eca5b4a8f Mon Sep 17 00:00:00 2001 From: françois romieu Date: Mon, 10 Aug 2009 19:44:19 +0000 Subject: r8169: phy init for the 8169scd Synced with Realtek's 6.011.00 r8169 driver. Signed-off-by: Francois Romieu Cc: Edward Hsu Signed-off-by: David S. Miller --- drivers/net/r8169.c | 68 +++++++++++++++++++++++++++++++++++++++++++++++++ include/linux/pci_ids.h | 2 ++ 2 files changed, 70 insertions(+) (limited to 'include') diff --git a/drivers/net/r8169.c b/drivers/net/r8169.c index 86722886ae47..4470fefbe97e 100644 --- a/drivers/net/r8169.c +++ b/drivers/net/r8169.c @@ -1389,6 +1389,71 @@ static void rtl8169sb_hw_phy_config(void __iomem *ioaddr) rtl_phy_write(ioaddr, phy_reg_init, ARRAY_SIZE(phy_reg_init)); } +static void rtl8169scd_hw_phy_config_quirk(struct rtl8169_private *tp, + void __iomem *ioaddr) +{ + struct pci_dev *pdev = tp->pci_dev; + u16 vendor_id, device_id; + + pci_read_config_word(pdev, PCI_SUBSYSTEM_VENDOR_ID, &vendor_id); + pci_read_config_word(pdev, PCI_SUBSYSTEM_ID, &device_id); + + if ((vendor_id != PCI_VENDOR_ID_GIGABYTE) || (device_id != 0xe000)) + return; + + mdio_write(ioaddr, 0x1f, 0x0001); + mdio_write(ioaddr, 0x10, 0xf01b); + mdio_write(ioaddr, 0x1f, 0x0000); +} + +static void rtl8169scd_hw_phy_config(struct rtl8169_private *tp, + void __iomem *ioaddr) +{ + struct phy_reg phy_reg_init[] = { + { 0x1f, 0x0001 }, + { 0x04, 0x0000 }, + { 0x03, 0x00a1 }, + { 0x02, 0x0008 }, + { 0x01, 0x0120 }, + { 0x00, 0x1000 }, + { 0x04, 0x0800 }, + { 0x04, 0x9000 }, + { 0x03, 0x802f }, + { 0x02, 0x4f02 }, + { 0x01, 0x0409 }, + { 0x00, 0xf099 }, + { 0x04, 0x9800 }, + { 0x04, 0xa000 }, + { 0x03, 0xdf01 }, + { 0x02, 0xdf20 }, + { 0x01, 0xff95 }, + { 0x00, 0xba00 }, + { 0x04, 0xa800 }, + { 0x04, 0xf000 }, + { 0x03, 0xdf01 }, + { 0x02, 0xdf20 }, + { 0x01, 0x101a }, + { 0x00, 0xa0ff }, + { 0x04, 0xf800 }, + { 0x04, 0x0000 }, + { 0x1f, 0x0000 }, + + { 0x1f, 0x0001 }, + { 0x10, 0xf41b }, + { 0x14, 0xfb54 }, + { 0x18, 0xf5c7 }, + { 0x1f, 0x0000 }, + + { 0x1f, 0x0001 }, + { 0x17, 0x0cc0 }, + { 0x1f, 0x0000 } + }; + + rtl_phy_write(ioaddr, phy_reg_init, ARRAY_SIZE(phy_reg_init)); + + rtl8169scd_hw_phy_config_quirk(tp, ioaddr); +} + static void rtl8169sce_hw_phy_config(void __iomem *ioaddr) { struct phy_reg phy_reg_init[] = { @@ -1681,6 +1746,9 @@ static void rtl_hw_phy_config(struct net_device *dev) case RTL_GIGA_MAC_VER_04: rtl8169sb_hw_phy_config(ioaddr); break; + case RTL_GIGA_MAC_VER_05: + rtl8169scd_hw_phy_config(tp, ioaddr); + break; case RTL_GIGA_MAC_VER_06: rtl8169sce_hw_phy_config(ioaddr); break; diff --git a/include/linux/pci_ids.h b/include/linux/pci_ids.h index 73b46b6b904f..fdc3110c90c0 100644 --- a/include/linux/pci_ids.h +++ b/include/linux/pci_ids.h @@ -1984,6 +1984,8 @@ #define PCI_VENDOR_ID_SAMSUNG 0x144d +#define PCI_VENDOR_ID_GIGABYTE 0x1458 + #define PCI_VENDOR_ID_AMBIT 0x1468 #define PCI_VENDOR_ID_MYRICOM 0x14c1 -- cgit v1.2.3 From 28402971d869e26271b25301011f667d3a5640c3 Mon Sep 17 00:00:00 2001 From: Ingo Molnar Date: Thu, 13 Aug 2009 10:13:22 +0200 Subject: perf_counter: Provide hw_perf_counter_setup_online() APIs Provide weak aliases for hw_perf_counter_setup_online(). This is used by the BTS patches (for v2.6.32), but it interacts with fixes so propagate this upstream. (it has no effect as of yet) Also export perf_counter_output() to architecture code. Cc: Peter Zijlstra Cc: Paul Mackerras Cc: Frederic Weisbecker LKML-Reference: Signed-off-by: Ingo Molnar --- include/linux/perf_counter.h | 2 ++ kernel/perf_counter.c | 10 +++++++++- 2 files changed, 11 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/perf_counter.h b/include/linux/perf_counter.h index a9d823a93fe8..2b36afe6ae0f 100644 --- a/include/linux/perf_counter.h +++ b/include/linux/perf_counter.h @@ -694,6 +694,8 @@ struct perf_sample_data { extern int perf_counter_overflow(struct perf_counter *counter, int nmi, struct perf_sample_data *data); +extern void perf_counter_output(struct perf_counter *counter, int nmi, + struct perf_sample_data *data); /* * Return 1 for a software counter, 0 for a hardware counter diff --git a/kernel/perf_counter.c b/kernel/perf_counter.c index b0b20a07f394..e26d2fcfa320 100644 --- a/kernel/perf_counter.c +++ b/kernel/perf_counter.c @@ -88,6 +88,7 @@ void __weak hw_perf_disable(void) { barrier(); } void __weak hw_perf_enable(void) { barrier(); } void __weak hw_perf_counter_setup(int cpu) { barrier(); } +void __weak hw_perf_counter_setup_online(int cpu) { barrier(); } int __weak hw_perf_group_sched_in(struct perf_counter *group_leader, @@ -2630,7 +2631,7 @@ static u32 perf_counter_tid(struct perf_counter *counter, struct task_struct *p) return task_pid_nr_ns(p, counter->ns); } -static void perf_counter_output(struct perf_counter *counter, int nmi, +void perf_counter_output(struct perf_counter *counter, int nmi, struct perf_sample_data *data) { int ret; @@ -4592,6 +4593,11 @@ perf_cpu_notify(struct notifier_block *self, unsigned long action, void *hcpu) perf_counter_init_cpu(cpu); break; + case CPU_ONLINE: + case CPU_ONLINE_FROZEN: + hw_perf_counter_setup_online(cpu); + break; + case CPU_DOWN_PREPARE: case CPU_DOWN_PREPARE_FROZEN: perf_counter_exit_cpu(cpu); @@ -4616,6 +4622,8 @@ void __init perf_counter_init(void) { perf_cpu_notify(&perf_cpu_nb, (unsigned long)CPU_UP_PREPARE, (void *)(long)smp_processor_id()); + perf_cpu_notify(&perf_cpu_nb, (unsigned long)CPU_ONLINE, + (void *)(long)smp_processor_id()); register_cpu_notifier(&perf_cpu_nb); } -- cgit v1.2.3 From 3dab77fb1bf89664bb1c9544607159dcab6f7d57 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Thu, 13 Aug 2009 11:47:53 +0200 Subject: perf: Rework/fix the whole read vs group stuff Replace PERF_SAMPLE_GROUP with PERF_SAMPLE_READ and introduce PERF_FORMAT_GROUP to deal with group reads in a more generic way. This allows you to get group reads out of read() as well. Signed-off-by: Peter Zijlstra Cc: Corey J Ashford Cc: Paul Mackerras Cc: stephane eranian LKML-Reference: <20090813103655.117411814@chello.nl> Signed-off-by: Ingo Molnar --- include/linux/perf_counter.h | 47 ++++++-- kernel/perf_counter.c | 274 +++++++++++++++++++++++++++++++------------ 2 files changed, 238 insertions(+), 83 deletions(-) (limited to 'include') diff --git a/include/linux/perf_counter.h b/include/linux/perf_counter.h index 2b36afe6ae0f..b53f7006cc4e 100644 --- a/include/linux/perf_counter.h +++ b/include/linux/perf_counter.h @@ -115,7 +115,7 @@ enum perf_counter_sample_format { PERF_SAMPLE_TID = 1U << 1, PERF_SAMPLE_TIME = 1U << 2, PERF_SAMPLE_ADDR = 1U << 3, - PERF_SAMPLE_GROUP = 1U << 4, + PERF_SAMPLE_READ = 1U << 4, PERF_SAMPLE_CALLCHAIN = 1U << 5, PERF_SAMPLE_ID = 1U << 6, PERF_SAMPLE_CPU = 1U << 7, @@ -127,16 +127,32 @@ enum perf_counter_sample_format { }; /* - * Bits that can be set in attr.read_format to request that - * reads on the counter should return the indicated quantities, - * in increasing order of bit value, after the counter value. + * The format of the data returned by read() on a perf counter fd, + * as specified by attr.read_format: + * + * struct read_format { + * { u64 value; + * { u64 time_enabled; } && PERF_FORMAT_ENABLED + * { u64 time_running; } && PERF_FORMAT_RUNNING + * { u64 id; } && PERF_FORMAT_ID + * } && !PERF_FORMAT_GROUP + * + * { u64 nr; + * { u64 time_enabled; } && PERF_FORMAT_ENABLED + * { u64 time_running; } && PERF_FORMAT_RUNNING + * { u64 value; + * { u64 id; } && PERF_FORMAT_ID + * } cntr[nr]; + * } && PERF_FORMAT_GROUP + * }; */ enum perf_counter_read_format { PERF_FORMAT_TOTAL_TIME_ENABLED = 1U << 0, PERF_FORMAT_TOTAL_TIME_RUNNING = 1U << 1, PERF_FORMAT_ID = 1U << 2, + PERF_FORMAT_GROUP = 1U << 3, - PERF_FORMAT_MAX = 1U << 3, /* non-ABI */ + PERF_FORMAT_MAX = 1U << 4, /* non-ABI */ }; #define PERF_ATTR_SIZE_VER0 64 /* sizeof first published struct */ @@ -343,10 +359,8 @@ enum perf_event_type { * struct { * struct perf_event_header header; * u32 pid, tid; - * u64 value; - * { u64 time_enabled; } && PERF_FORMAT_ENABLED - * { u64 time_running; } && PERF_FORMAT_RUNNING - * { u64 parent_id; } && PERF_FORMAT_ID + * + * struct read_format values; * }; */ PERF_EVENT_READ = 8, @@ -364,11 +378,22 @@ enum perf_event_type { * { u32 cpu, res; } && PERF_SAMPLE_CPU * { u64 period; } && PERF_SAMPLE_PERIOD * - * { u64 nr; - * { u64 id, val; } cnt[nr]; } && PERF_SAMPLE_GROUP + * { struct read_format values; } && PERF_SAMPLE_READ * * { u64 nr, * u64 ips[nr]; } && PERF_SAMPLE_CALLCHAIN + * + * # + * # The RAW record below is opaque data wrt the ABI + * # + * # That is, the ABI doesn't make any promises wrt to + * # the stability of its content, it may vary depending + * # on event, hardware, kernel version and phase of + * # the moon. + * # + * # In other words, PERF_SAMPLE_RAW contents are not an ABI. + * # + * * { u32 size; * char data[size];}&& PERF_SAMPLE_RAW * }; diff --git a/kernel/perf_counter.c b/kernel/perf_counter.c index 3dd4339589a0..b8c6b97a20a3 100644 --- a/kernel/perf_counter.c +++ b/kernel/perf_counter.c @@ -1692,7 +1692,32 @@ static int perf_release(struct inode *inode, struct file *file) return 0; } -static u64 perf_counter_read_tree(struct perf_counter *counter) +static int perf_counter_read_size(struct perf_counter *counter) +{ + int entry = sizeof(u64); /* value */ + int size = 0; + int nr = 1; + + if (counter->attr.read_format & PERF_FORMAT_TOTAL_TIME_ENABLED) + size += sizeof(u64); + + if (counter->attr.read_format & PERF_FORMAT_TOTAL_TIME_RUNNING) + size += sizeof(u64); + + if (counter->attr.read_format & PERF_FORMAT_ID) + entry += sizeof(u64); + + if (counter->attr.read_format & PERF_FORMAT_GROUP) { + nr += counter->group_leader->nr_siblings; + size += sizeof(u64); + } + + size += entry * nr; + + return size; +} + +static u64 perf_counter_read_value(struct perf_counter *counter) { struct perf_counter *child; u64 total = 0; @@ -1704,14 +1729,96 @@ static u64 perf_counter_read_tree(struct perf_counter *counter) return total; } +static int perf_counter_read_entry(struct perf_counter *counter, + u64 read_format, char __user *buf) +{ + int n = 0, count = 0; + u64 values[2]; + + values[n++] = perf_counter_read_value(counter); + if (read_format & PERF_FORMAT_ID) + values[n++] = primary_counter_id(counter); + + count = n * sizeof(u64); + + if (copy_to_user(buf, values, count)) + return -EFAULT; + + return count; +} + +static int perf_counter_read_group(struct perf_counter *counter, + u64 read_format, char __user *buf) +{ + struct perf_counter *leader = counter->group_leader, *sub; + int n = 0, size = 0, err = -EFAULT; + u64 values[3]; + + values[n++] = 1 + leader->nr_siblings; + if (read_format & PERF_FORMAT_TOTAL_TIME_ENABLED) { + values[n++] = leader->total_time_enabled + + atomic64_read(&leader->child_total_time_enabled); + } + if (read_format & PERF_FORMAT_TOTAL_TIME_RUNNING) { + values[n++] = leader->total_time_running + + atomic64_read(&leader->child_total_time_running); + } + + size = n * sizeof(u64); + + if (copy_to_user(buf, values, size)) + return -EFAULT; + + err = perf_counter_read_entry(leader, read_format, buf + size); + if (err < 0) + return err; + + size += err; + + list_for_each_entry(sub, &leader->sibling_list, list_entry) { + err = perf_counter_read_entry(counter, read_format, + buf + size); + if (err < 0) + return err; + + size += err; + } + + return size; +} + +static int perf_counter_read_one(struct perf_counter *counter, + u64 read_format, char __user *buf) +{ + u64 values[4]; + int n = 0; + + values[n++] = perf_counter_read_value(counter); + if (read_format & PERF_FORMAT_TOTAL_TIME_ENABLED) { + values[n++] = counter->total_time_enabled + + atomic64_read(&counter->child_total_time_enabled); + } + if (read_format & PERF_FORMAT_TOTAL_TIME_RUNNING) { + values[n++] = counter->total_time_running + + atomic64_read(&counter->child_total_time_running); + } + if (read_format & PERF_FORMAT_ID) + values[n++] = primary_counter_id(counter); + + if (copy_to_user(buf, values, n * sizeof(u64))) + return -EFAULT; + + return n * sizeof(u64); +} + /* * Read the performance counter - simple non blocking version for now */ static ssize_t perf_read_hw(struct perf_counter *counter, char __user *buf, size_t count) { - u64 values[4]; - int n; + u64 read_format = counter->attr.read_format; + int ret; /* * Return end-of-file for a read on a counter that is in @@ -1721,28 +1828,18 @@ perf_read_hw(struct perf_counter *counter, char __user *buf, size_t count) if (counter->state == PERF_COUNTER_STATE_ERROR) return 0; + if (count < perf_counter_read_size(counter)) + return -ENOSPC; + WARN_ON_ONCE(counter->ctx->parent_ctx); mutex_lock(&counter->child_mutex); - values[0] = perf_counter_read_tree(counter); - n = 1; - if (counter->attr.read_format & PERF_FORMAT_TOTAL_TIME_ENABLED) - values[n++] = counter->total_time_enabled + - atomic64_read(&counter->child_total_time_enabled); - if (counter->attr.read_format & PERF_FORMAT_TOTAL_TIME_RUNNING) - values[n++] = counter->total_time_running + - atomic64_read(&counter->child_total_time_running); - if (counter->attr.read_format & PERF_FORMAT_ID) - values[n++] = primary_counter_id(counter); + if (read_format & PERF_FORMAT_GROUP) + ret = perf_counter_read_group(counter, read_format, buf); + else + ret = perf_counter_read_one(counter, read_format, buf); mutex_unlock(&counter->child_mutex); - if (count < n * sizeof(u64)) - return -EINVAL; - count = n * sizeof(u64); - - if (copy_to_user(buf, values, count)) - return -EFAULT; - - return count; + return ret; } static ssize_t @@ -2631,6 +2728,79 @@ static u32 perf_counter_tid(struct perf_counter *counter, struct task_struct *p) return task_pid_nr_ns(p, counter->ns); } +static void perf_output_read_one(struct perf_output_handle *handle, + struct perf_counter *counter) +{ + u64 read_format = counter->attr.read_format; + u64 values[4]; + int n = 0; + + values[n++] = atomic64_read(&counter->count); + if (read_format & PERF_FORMAT_TOTAL_TIME_ENABLED) { + values[n++] = counter->total_time_enabled + + atomic64_read(&counter->child_total_time_enabled); + } + if (read_format & PERF_FORMAT_TOTAL_TIME_RUNNING) { + values[n++] = counter->total_time_running + + atomic64_read(&counter->child_total_time_running); + } + if (read_format & PERF_FORMAT_ID) + values[n++] = primary_counter_id(counter); + + perf_output_copy(handle, values, n * sizeof(u64)); +} + +/* + * XXX PERF_FORMAT_GROUP vs inherited counters seems difficult. + */ +static void perf_output_read_group(struct perf_output_handle *handle, + struct perf_counter *counter) +{ + struct perf_counter *leader = counter->group_leader, *sub; + u64 read_format = counter->attr.read_format; + u64 values[5]; + int n = 0; + + values[n++] = 1 + leader->nr_siblings; + + if (read_format & PERF_FORMAT_TOTAL_TIME_ENABLED) + values[n++] = leader->total_time_enabled; + + if (read_format & PERF_FORMAT_TOTAL_TIME_RUNNING) + values[n++] = leader->total_time_running; + + if (leader != counter) + leader->pmu->read(leader); + + values[n++] = atomic64_read(&leader->count); + if (read_format & PERF_FORMAT_ID) + values[n++] = primary_counter_id(leader); + + perf_output_copy(handle, values, n * sizeof(u64)); + + list_for_each_entry(sub, &leader->sibling_list, list_entry) { + n = 0; + + if (sub != counter) + sub->pmu->read(sub); + + values[n++] = atomic64_read(&sub->count); + if (read_format & PERF_FORMAT_ID) + values[n++] = primary_counter_id(sub); + + perf_output_copy(handle, values, n * sizeof(u64)); + } +} + +static void perf_output_read(struct perf_output_handle *handle, + struct perf_counter *counter) +{ + if (counter->attr.read_format & PERF_FORMAT_GROUP) + perf_output_read_group(handle, counter); + else + perf_output_read_one(handle, counter); +} + void perf_counter_output(struct perf_counter *counter, int nmi, struct perf_sample_data *data) { @@ -2642,10 +2812,6 @@ void perf_counter_output(struct perf_counter *counter, int nmi, struct { u32 pid, tid; } tid_entry; - struct { - u64 id; - u64 counter; - } group_entry; struct perf_callchain_entry *callchain = NULL; int callchain_size = 0; u64 time; @@ -2700,10 +2866,8 @@ void perf_counter_output(struct perf_counter *counter, int nmi, if (sample_type & PERF_SAMPLE_PERIOD) header.size += sizeof(u64); - if (sample_type & PERF_SAMPLE_GROUP) { - header.size += sizeof(u64) + - counter->nr_siblings * sizeof(group_entry); - } + if (sample_type & PERF_SAMPLE_READ) + header.size += perf_counter_read_size(counter); if (sample_type & PERF_SAMPLE_CALLCHAIN) { callchain = perf_callchain(data->regs); @@ -2760,26 +2924,8 @@ void perf_counter_output(struct perf_counter *counter, int nmi, if (sample_type & PERF_SAMPLE_PERIOD) perf_output_put(&handle, data->period); - /* - * XXX PERF_SAMPLE_GROUP vs inherited counters seems difficult. - */ - if (sample_type & PERF_SAMPLE_GROUP) { - struct perf_counter *leader, *sub; - u64 nr = counter->nr_siblings; - - perf_output_put(&handle, nr); - - leader = counter->group_leader; - list_for_each_entry(sub, &leader->sibling_list, list_entry) { - if (sub != counter) - sub->pmu->read(sub); - - group_entry.id = primary_counter_id(sub); - group_entry.counter = atomic64_read(&sub->count); - - perf_output_put(&handle, group_entry); - } - } + if (sample_type & PERF_SAMPLE_READ) + perf_output_read(&handle, counter); if (sample_type & PERF_SAMPLE_CALLCHAIN) { if (callchain) @@ -2818,8 +2964,6 @@ struct perf_read_event { u32 pid; u32 tid; - u64 value; - u64 format[3]; }; static void @@ -2831,34 +2975,20 @@ perf_counter_read_event(struct perf_counter *counter, .header = { .type = PERF_EVENT_READ, .misc = 0, - .size = sizeof(event) - sizeof(event.format), + .size = sizeof(event) + perf_counter_read_size(counter), }, .pid = perf_counter_pid(counter, task), .tid = perf_counter_tid(counter, task), - .value = atomic64_read(&counter->count), }; - int ret, i = 0; - - if (counter->attr.read_format & PERF_FORMAT_TOTAL_TIME_ENABLED) { - event.header.size += sizeof(u64); - event.format[i++] = counter->total_time_enabled; - } - - if (counter->attr.read_format & PERF_FORMAT_TOTAL_TIME_RUNNING) { - event.header.size += sizeof(u64); - event.format[i++] = counter->total_time_running; - } - - if (counter->attr.read_format & PERF_FORMAT_ID) { - event.header.size += sizeof(u64); - event.format[i++] = primary_counter_id(counter); - } + int ret; ret = perf_output_begin(&handle, counter, event.header.size, 0, 0); if (ret) return; - perf_output_copy(&handle, &event, event.header.size); + perf_output_put(&handle, event); + perf_output_read(&handle, counter); + perf_output_end(&handle); } @@ -3921,9 +4051,9 @@ perf_counter_alloc(struct perf_counter_attr *attr, atomic64_set(&hwc->period_left, hwc->sample_period); /* - * we currently do not support PERF_SAMPLE_GROUP on inherited counters + * we currently do not support PERF_FORMAT_GROUP on inherited counters */ - if (attr->inherit && (attr->sample_type & PERF_SAMPLE_GROUP)) + if (attr->inherit && (attr->read_format & PERF_FORMAT_GROUP)) goto done; switch (attr->type) { -- cgit v1.2.3 From e9b3cc1b3779fe10a80de4c3e7404bd308d0eae3 Mon Sep 17 00:00:00 2001 From: Neil Horman Date: Thu, 13 Aug 2009 05:19:44 +0000 Subject: net: skb ftracer - add tracepoint to skb_copy_datagram_iovec (v3) skb allocation / cosumption tracer - Add consumption tracepoint This patch adds a tracepoint to skb_copy_datagram_iovec, which is called each time a userspace process copies a frame from a socket receive queue to a user space buffer. It allows us to hook in and examine each sk_buff that the system receives on a per-socket bases, and can be use to compile a list of which skb's were received by which processes. Signed-off-by: Neil Horman include/trace/events/skb.h | 20 ++++++++++++++++++++ net/core/datagram.c | 3 +++ 2 files changed, 23 insertions(+) Signed-off-by: David S. Miller --- include/trace/events/skb.h | 20 ++++++++++++++++++++ net/core/datagram.c | 3 +++ 2 files changed, 23 insertions(+) (limited to 'include') diff --git a/include/trace/events/skb.h b/include/trace/events/skb.h index e499863b9669..4b2be6dc76f0 100644 --- a/include/trace/events/skb.h +++ b/include/trace/events/skb.h @@ -5,6 +5,7 @@ #define _TRACE_SKB_H #include +#include #include /* @@ -34,6 +35,25 @@ TRACE_EVENT(kfree_skb, __entry->skbaddr, __entry->protocol, __entry->location) ); +TRACE_EVENT(skb_copy_datagram_iovec, + + TP_PROTO(const struct sk_buff *skb, int len), + + TP_ARGS(skb, len), + + TP_STRUCT__entry( + __field( const void *, skbaddr ) + __field( int, len ) + ), + + TP_fast_assign( + __entry->skbaddr = skb; + __entry->len = len; + ), + + TP_printk("skbaddr=%p len=%d", __entry->skbaddr, __entry->len) +); + #endif /* _TRACE_SKB_H */ /* This part must be outside protection */ diff --git a/net/core/datagram.c b/net/core/datagram.c index b0fe69211eef..1c6cf3a1a4f6 100644 --- a/net/core/datagram.c +++ b/net/core/datagram.c @@ -55,6 +55,7 @@ #include #include #include +#include /* * Is a socket 'connection oriented' ? @@ -284,6 +285,8 @@ int skb_copy_datagram_iovec(const struct sk_buff *skb, int offset, int i, copy = start - offset; struct sk_buff *frag_iter; + trace_skb_copy_datagram_iovec(skb, len); + /* Copy header. */ if (copy > 0) { if (copy > len) -- cgit v1.2.3 From 9188499cdb117d86a1ea6b04374095b098d56936 Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Thu, 13 Aug 2009 09:44:57 -0400 Subject: security: introducing security_request_module Calling request_module() will trigger a userspace upcall which will load a new module into the kernel. This can be a dangerous event if the process able to trigger request_module() is able to control either the modprobe binary or the module binary. This patch adds a new security hook to request_module() which can be used by an LSM to control a processes ability to call request_module(). Signed-off-by: Eric Paris Acked-by: Serge Hallyn Signed-off-by: James Morris --- include/linux/security.h | 10 ++++++++++ kernel/kmod.c | 4 ++++ security/capability.c | 6 ++++++ security/security.c | 5 +++++ 4 files changed, 25 insertions(+) (limited to 'include') diff --git a/include/linux/security.h b/include/linux/security.h index 57ead99d2593..1e3dd86eea99 100644 --- a/include/linux/security.h +++ b/include/linux/security.h @@ -678,6 +678,9 @@ static inline void security_free_mnt_opts(struct security_mnt_opts *opts) * @inode points to the inode to use as a reference. * The current task must be the one that nominated @inode. * Return 0 if successful. + * @kernel_module_request: + * Ability to trigger the kernel to automatically upcall to userspace for + * userspace to load a kernel module with the given name. * @task_setuid: * Check permission before setting one or more of the user identity * attributes of the current process. The @flags parameter indicates @@ -1489,6 +1492,7 @@ struct security_operations { void (*cred_commit)(struct cred *new, const struct cred *old); int (*kernel_act_as)(struct cred *new, u32 secid); int (*kernel_create_files_as)(struct cred *new, struct inode *inode); + int (*kernel_module_request)(void); int (*task_setuid) (uid_t id0, uid_t id1, uid_t id2, int flags); int (*task_fix_setuid) (struct cred *new, const struct cred *old, int flags); @@ -1741,6 +1745,7 @@ int security_prepare_creds(struct cred *new, const struct cred *old, gfp_t gfp); void security_commit_creds(struct cred *new, const struct cred *old); int security_kernel_act_as(struct cred *new, u32 secid); int security_kernel_create_files_as(struct cred *new, struct inode *inode); +int security_kernel_module_request(void); int security_task_setuid(uid_t id0, uid_t id1, uid_t id2, int flags); int security_task_fix_setuid(struct cred *new, const struct cred *old, int flags); @@ -2292,6 +2297,11 @@ static inline int security_kernel_create_files_as(struct cred *cred, return 0; } +static inline int security_kernel_module_request(void) +{ + return 0; +} + static inline int security_task_setuid(uid_t id0, uid_t id1, uid_t id2, int flags) { diff --git a/kernel/kmod.c b/kernel/kmod.c index 385c31a1bdbf..5a7ae57f983f 100644 --- a/kernel/kmod.c +++ b/kernel/kmod.c @@ -78,6 +78,10 @@ int __request_module(bool wait, const char *fmt, ...) #define MAX_KMOD_CONCURRENT 50 /* Completely arbitrary value - KAO */ static int kmod_loop_msg; + ret = security_kernel_module_request(); + if (ret) + return ret; + va_start(args, fmt); ret = vsnprintf(module_name, MODULE_NAME_LEN, fmt, args); va_end(args); diff --git a/security/capability.c b/security/capability.c index ec0573054024..1b943f54b2ea 100644 --- a/security/capability.c +++ b/security/capability.c @@ -396,6 +396,11 @@ static int cap_kernel_create_files_as(struct cred *new, struct inode *inode) return 0; } +static int cap_kernel_module_request(void) +{ + return 0; +} + static int cap_task_setuid(uid_t id0, uid_t id1, uid_t id2, int flags) { return 0; @@ -945,6 +950,7 @@ void security_fixup_ops(struct security_operations *ops) set_to_cap_if_null(ops, cred_commit); set_to_cap_if_null(ops, kernel_act_as); set_to_cap_if_null(ops, kernel_create_files_as); + set_to_cap_if_null(ops, kernel_module_request); set_to_cap_if_null(ops, task_setuid); set_to_cap_if_null(ops, task_fix_setuid); set_to_cap_if_null(ops, task_setgid); diff --git a/security/security.c b/security/security.c index 4501c5e1f988..0e993f42ce3d 100644 --- a/security/security.c +++ b/security/security.c @@ -709,6 +709,11 @@ int security_kernel_create_files_as(struct cred *new, struct inode *inode) return security_ops->kernel_create_files_as(new, inode); } +int security_kernel_module_request(void) +{ + return security_ops->kernel_module_request(); +} + int security_task_setuid(uid_t id0, uid_t id1, uid_t id2, int flags) { return security_ops->task_setuid(id0, id1, id2, flags); -- cgit v1.2.3 From f322abf83feddc3c37c3a91794e0c5aece4af18e Mon Sep 17 00:00:00 2001 From: James Morris Date: Fri, 14 Aug 2009 11:19:29 +1000 Subject: security: update documentation for security_request_module Update documentation for security_request_module to indicate return value, as suggested by Serge Hallyn. Signed-off-by: James Morris --- include/linux/security.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include') diff --git a/include/linux/security.h b/include/linux/security.h index 1e3dd86eea99..a16d6b7c4ebe 100644 --- a/include/linux/security.h +++ b/include/linux/security.h @@ -681,6 +681,7 @@ static inline void security_free_mnt_opts(struct security_mnt_opts *opts) * @kernel_module_request: * Ability to trigger the kernel to automatically upcall to userspace for * userspace to load a kernel module with the given name. + * Return 0 if successful. * @task_setuid: * Check permission before setting one or more of the user identity * attributes of the current process. The @flags parameter indicates -- cgit v1.2.3 From 00ae4064b1445524752575dd84df227c0687c99d Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Fri, 14 Aug 2009 15:00:49 +0900 Subject: percpu: rename 4k first chunk allocator to page Page size isn't always 4k depending on arch and configuration. Rename 4k first chunk allocator to page. Signed-off-by: Tejun Heo Cc: David Howells --- Documentation/kernel-parameters.txt | 2 +- arch/x86/kernel/setup_percpu.c | 23 ++++++++++++----------- include/linux/percpu.h | 2 +- mm/percpu.c | 25 ++++++++++++++----------- 4 files changed, 28 insertions(+), 24 deletions(-) (limited to 'include') diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt index 7936b801fe6a..12e9eb77ee0d 100644 --- a/Documentation/kernel-parameters.txt +++ b/Documentation/kernel-parameters.txt @@ -1920,7 +1920,7 @@ and is between 256 and 4096 characters. It is defined in the file See arch/parisc/kernel/pdc_chassis.c percpu_alloc= [X86] Select which percpu first chunk allocator to use. - Allowed values are one of "lpage", "embed" and "4k". + Allowed values are one of "lpage", "embed" and "page". See comments in arch/x86/kernel/setup_percpu.c for details on each allocator. This parameter is primarily for debugging and performance comparison. diff --git a/arch/x86/kernel/setup_percpu.c b/arch/x86/kernel/setup_percpu.c index a26ff61e2fb0..1e17711c29d6 100644 --- a/arch/x86/kernel/setup_percpu.c +++ b/arch/x86/kernel/setup_percpu.c @@ -249,21 +249,22 @@ static ssize_t __init setup_pcpu_embed(size_t static_size, bool chosen) } /* - * 4k allocator + * Page allocator * - * Boring fallback 4k allocator. This allocator puts more pressure on - * PTE TLBs but other than that behaves nicely on both UMA and NUMA. + * Boring fallback 4k page allocator. This allocator puts more + * pressure on PTE TLBs but other than that behaves nicely on both UMA + * and NUMA. */ -static void __init pcpu4k_populate_pte(unsigned long addr) +static void __init pcpup_populate_pte(unsigned long addr) { populate_extra_pte(addr); } -static ssize_t __init setup_pcpu_4k(size_t static_size) +static ssize_t __init setup_pcpu_page(size_t static_size) { - return pcpu_4k_first_chunk(static_size, PERCPU_FIRST_CHUNK_RESERVE, - pcpu_fc_alloc, pcpu_fc_free, - pcpu4k_populate_pte); + return pcpu_page_first_chunk(static_size, PERCPU_FIRST_CHUNK_RESERVE, + pcpu_fc_alloc, pcpu_fc_free, + pcpup_populate_pte); } /* for explicit first chunk allocator selection */ @@ -307,7 +308,7 @@ void __init setup_per_cpu_areas(void) */ ret = -EINVAL; if (strlen(pcpu_chosen_alloc)) { - if (strcmp(pcpu_chosen_alloc, "4k")) { + if (strcmp(pcpu_chosen_alloc, "page")) { if (!strcmp(pcpu_chosen_alloc, "lpage")) ret = setup_pcpu_lpage(static_size, true); else if (!strcmp(pcpu_chosen_alloc, "embed")) @@ -317,7 +318,7 @@ void __init setup_per_cpu_areas(void) "specified\n", pcpu_chosen_alloc); if (ret < 0) pr_warning("PERCPU: %s allocator failed (%zd), " - "falling back to 4k\n", + "falling back to page size\n", pcpu_chosen_alloc, ret); } } else { @@ -326,7 +327,7 @@ void __init setup_per_cpu_areas(void) ret = setup_pcpu_embed(static_size, false); } if (ret < 0) - ret = setup_pcpu_4k(static_size); + ret = setup_pcpu_page(static_size); if (ret < 0) panic("cannot allocate static percpu area (%zu bytes, err=%zd)", static_size, ret); diff --git a/include/linux/percpu.h b/include/linux/percpu.h index e134c8229631..7989f61b03f3 100644 --- a/include/linux/percpu.h +++ b/include/linux/percpu.h @@ -74,7 +74,7 @@ extern ssize_t __init pcpu_embed_first_chunk( size_t static_size, size_t reserved_size, ssize_t dyn_size); -extern ssize_t __init pcpu_4k_first_chunk( +extern ssize_t __init pcpu_page_first_chunk( size_t static_size, size_t reserved_size, pcpu_fc_alloc_fn_t alloc_fn, pcpu_fc_free_fn_t free_fn, diff --git a/mm/percpu.c b/mm/percpu.c index cbddcbdab681..6feac7934904 100644 --- a/mm/percpu.c +++ b/mm/percpu.c @@ -1497,15 +1497,15 @@ ssize_t __init pcpu_embed_first_chunk(size_t static_size, size_t reserved_size, } /** - * pcpu_4k_first_chunk - map the first chunk using PAGE_SIZE pages + * pcpu_page_first_chunk - map the first chunk using PAGE_SIZE pages * @static_size: the size of static percpu area in bytes * @reserved_size: the size of reserved percpu area in bytes * @alloc_fn: function to allocate percpu page, always called with PAGE_SIZE * @free_fn: funtion to free percpu page, always called with PAGE_SIZE * @populate_pte_fn: function to populate pte * - * This is a helper to ease setting up embedded first percpu chunk and - * can be called where pcpu_setup_first_chunk() is expected. + * This is a helper to ease setting up page-remapped first percpu + * chunk and can be called where pcpu_setup_first_chunk() is expected. * * This is the basic allocator. Static percpu area is allocated * page-by-page into vmalloc area. @@ -1514,12 +1514,13 @@ ssize_t __init pcpu_embed_first_chunk(size_t static_size, size_t reserved_size, * The determined pcpu_unit_size which can be used to initialize * percpu access on success, -errno on failure. */ -ssize_t __init pcpu_4k_first_chunk(size_t static_size, size_t reserved_size, - pcpu_fc_alloc_fn_t alloc_fn, - pcpu_fc_free_fn_t free_fn, - pcpu_fc_populate_pte_fn_t populate_pte_fn) +ssize_t __init pcpu_page_first_chunk(size_t static_size, size_t reserved_size, + pcpu_fc_alloc_fn_t alloc_fn, + pcpu_fc_free_fn_t free_fn, + pcpu_fc_populate_pte_fn_t populate_pte_fn) { static struct vm_struct vm; + char psize_str[16]; int unit_pages; size_t pages_size; struct page **pages; @@ -1527,6 +1528,8 @@ ssize_t __init pcpu_4k_first_chunk(size_t static_size, size_t reserved_size, int i, j; ssize_t ret; + snprintf(psize_str, sizeof(psize_str), "%luK", PAGE_SIZE >> 10); + unit_pages = PFN_UP(max_t(size_t, static_size + reserved_size, PCPU_MIN_UNIT_SIZE)); @@ -1542,8 +1545,8 @@ ssize_t __init pcpu_4k_first_chunk(size_t static_size, size_t reserved_size, ptr = alloc_fn(cpu, PAGE_SIZE); if (!ptr) { - pr_warning("PERCPU: failed to allocate " - "4k page for cpu%u\n", cpu); + pr_warning("PERCPU: failed to allocate %s page " + "for cpu%u\n", psize_str, cpu); goto enomem; } pages[j++] = virt_to_page(ptr); @@ -1580,8 +1583,8 @@ ssize_t __init pcpu_4k_first_chunk(size_t static_size, size_t reserved_size, } /* we're ready, commit */ - pr_info("PERCPU: %d 4k pages/cpu @%p s%zu r%zu\n", - unit_pages, vm.addr, static_size, reserved_size); + pr_info("PERCPU: %d %s pages/cpu @%p s%zu r%zu\n", + unit_pages, psize_str, vm.addr, static_size, reserved_size); ret = pcpu_setup_first_chunk(static_size, reserved_size, -1, unit_pages << PAGE_SHIFT, vm.addr, NULL); -- cgit v1.2.3 From 08fc45806103e59a37418e84719b878f9bb32540 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Fri, 14 Aug 2009 15:00:49 +0900 Subject: percpu: build first chunk allocators selectively There's no need to build unused first chunk allocators in. Define CONFIG_NEED_PER_CPU_*_FIRST_CHUNK and let archs enable them selectively. Signed-off-by: Tejun Heo --- arch/x86/Kconfig | 10 ++++++++++ include/linux/percpu.h | 27 +++++---------------------- mm/percpu.c | 19 +++++++++++-------- 3 files changed, 26 insertions(+), 30 deletions(-) (limited to 'include') diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index e06b2eeff9f2..f7ac27215512 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -150,6 +150,16 @@ config ARCH_HAS_CACHE_LINE_SIZE config HAVE_SETUP_PER_CPU_AREA def_bool y +config NEED_PER_CPU_EMBED_FIRST_CHUNK + def_bool y + +config NEED_PER_CPU_PAGE_FIRST_CHUNK + def_bool y + +config NEED_PER_CPU_LPAGE_FIRST_CHUNK + def_bool y + depends on NEED_MULTIPLE_NODES + config HAVE_CPUMASK_OF_CPU_MAP def_bool X86_64_SMP diff --git a/include/linux/percpu.h b/include/linux/percpu.h index 7989f61b03f3..e26788e0da4a 100644 --- a/include/linux/percpu.h +++ b/include/linux/percpu.h @@ -70,17 +70,21 @@ extern size_t __init pcpu_setup_first_chunk( ssize_t dyn_size, size_t unit_size, void *base_addr, const int *unit_map); +#ifdef CONFIG_NEED_PER_CPU_EMBED_FIRST_CHUNK extern ssize_t __init pcpu_embed_first_chunk( size_t static_size, size_t reserved_size, ssize_t dyn_size); +#endif +#ifdef CONFIG_NEED_PER_CPU_PAGE_FIRST_CHUNK extern ssize_t __init pcpu_page_first_chunk( size_t static_size, size_t reserved_size, pcpu_fc_alloc_fn_t alloc_fn, pcpu_fc_free_fn_t free_fn, pcpu_fc_populate_pte_fn_t populate_pte_fn); +#endif -#ifdef CONFIG_NEED_MULTIPLE_NODES +#ifdef CONFIG_NEED_PER_CPU_LPAGE_FIRST_CHUNK extern int __init pcpu_lpage_build_unit_map( size_t static_size, size_t reserved_size, ssize_t *dyn_sizep, size_t *unit_sizep, @@ -98,27 +102,6 @@ extern ssize_t __init pcpu_lpage_first_chunk( extern void *pcpu_lpage_remapped(void *kaddr); #else -static inline int pcpu_lpage_build_unit_map( - size_t static_size, size_t reserved_size, - ssize_t *dyn_sizep, size_t *unit_sizep, - size_t lpage_size, int *unit_map, - pcpu_fc_cpu_distance_fn_t cpu_distance_fn) -{ - return -EINVAL; -} - -static inline ssize_t __init pcpu_lpage_first_chunk( - size_t static_size, size_t reserved_size, - size_t dyn_size, size_t unit_size, - size_t lpage_size, const int *unit_map, - int nr_units, - pcpu_fc_alloc_fn_t alloc_fn, - pcpu_fc_free_fn_t free_fn, - pcpu_fc_map_fn_t map_fn) -{ - return -EINVAL; -} - static inline void *pcpu_lpage_remapped(void *kaddr) { return NULL; diff --git a/mm/percpu.c b/mm/percpu.c index 6feac7934904..7971997de310 100644 --- a/mm/percpu.c +++ b/mm/percpu.c @@ -1414,8 +1414,9 @@ size_t __init pcpu_setup_first_chunk(size_t static_size, size_t reserved_size, return pcpu_unit_size; } -static size_t pcpu_calc_fc_sizes(size_t static_size, size_t reserved_size, - ssize_t *dyn_sizep) +static inline size_t pcpu_calc_fc_sizes(size_t static_size, + size_t reserved_size, + ssize_t *dyn_sizep) { size_t size_sum; @@ -1427,6 +1428,8 @@ static size_t pcpu_calc_fc_sizes(size_t static_size, size_t reserved_size, return size_sum; } +#if defined(CONFIG_NEED_PER_CPU_EMBED_FIRST_CHUNK) || \ + !defined(CONFIG_HAVE_SETUP_PER_CPU_AREA) /** * pcpu_embed_first_chunk - embed the first percpu chunk into bootmem * @static_size: the size of static percpu area in bytes @@ -1495,7 +1498,10 @@ ssize_t __init pcpu_embed_first_chunk(size_t static_size, size_t reserved_size, return pcpu_setup_first_chunk(static_size, reserved_size, dyn_size, unit_size, base, NULL); } +#endif /* CONFIG_NEED_PER_CPU_EMBED_FIRST_CHUNK || + !CONFIG_HAVE_SETUP_PER_CPU_AREA */ +#ifdef CONFIG_NEED_PER_CPU_PAGE_FIRST_CHUNK /** * pcpu_page_first_chunk - map the first chunk using PAGE_SIZE pages * @static_size: the size of static percpu area in bytes @@ -1598,12 +1604,9 @@ out_free_ar: free_bootmem(__pa(pages), pages_size); return ret; } +#endif /* CONFIG_NEED_PER_CPU_PAGE_FIRST_CHUNK */ -/* - * Large page remapping first chunk setup helper - */ -#ifdef CONFIG_NEED_MULTIPLE_NODES - +#ifdef CONFIG_NEED_PER_CPU_LPAGE_FIRST_CHUNK /** * pcpu_lpage_build_unit_map - build unit_map for large page remapping * @static_size: the size of static percpu area in bytes @@ -1982,7 +1985,7 @@ void *pcpu_lpage_remapped(void *kaddr) return NULL; } -#endif +#endif /* CONFIG_NEED_PER_CPU_LPAGE_FIRST_CHUNK */ /* * Generic percpu area setup. -- cgit v1.2.3 From f58dc01ba2ca9fe3ab2ba4ca43d9c8a735cf62d8 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Fri, 14 Aug 2009 15:00:50 +0900 Subject: percpu: generalize first chunk allocator selection Now that all first chunk allocators are in mm/percpu.c, it makes sense to make generalize percpu_alloc kernel parameter. Define PCPU_FC_* and set pcpu_chosen_fc using early_param() in mm/percpu.c. Arch code can use the set value to determine which first chunk allocator to use. Signed-off-by: Tejun Heo --- Documentation/kernel-parameters.txt | 11 ++++++----- arch/x86/kernel/setup_percpu.c | 24 ++++++------------------ include/linux/percpu.h | 12 ++++++++++++ mm/percpu.c | 32 ++++++++++++++++++++++++++++++++ 4 files changed, 56 insertions(+), 23 deletions(-) (limited to 'include') diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt index 12e9eb77ee0d..dee9ce2e6cfa 100644 --- a/Documentation/kernel-parameters.txt +++ b/Documentation/kernel-parameters.txt @@ -1919,11 +1919,12 @@ and is between 256 and 4096 characters. It is defined in the file Format: { 0 | 1 } See arch/parisc/kernel/pdc_chassis.c - percpu_alloc= [X86] Select which percpu first chunk allocator to use. - Allowed values are one of "lpage", "embed" and "page". - See comments in arch/x86/kernel/setup_percpu.c for - details on each allocator. This parameter is primarily - for debugging and performance comparison. + percpu_alloc= Select which percpu first chunk allocator to use. + Currently supported values are "embed", "page" and + "lpage". Archs may support subset or none of the + selections. See comments in mm/percpu.c for details + on each allocator. This parameter is primarily for + debugging and performance comparison. pf. [PARIDE] See Documentation/blockdev/paride.txt. diff --git a/arch/x86/kernel/setup_percpu.c b/arch/x86/kernel/setup_percpu.c index 1e17711c29d6..b961d99e6416 100644 --- a/arch/x86/kernel/setup_percpu.c +++ b/arch/x86/kernel/setup_percpu.c @@ -267,16 +267,6 @@ static ssize_t __init setup_pcpu_page(size_t static_size) pcpup_populate_pte); } -/* for explicit first chunk allocator selection */ -static char pcpu_chosen_alloc[16] __initdata; - -static int __init percpu_alloc_setup(char *str) -{ - strncpy(pcpu_chosen_alloc, str, sizeof(pcpu_chosen_alloc) - 1); - return 0; -} -early_param("percpu_alloc", percpu_alloc_setup); - static inline void setup_percpu_segment(int cpu) { #ifdef CONFIG_X86_32 @@ -307,19 +297,17 @@ void __init setup_per_cpu_areas(void) * each allocator for details. */ ret = -EINVAL; - if (strlen(pcpu_chosen_alloc)) { - if (strcmp(pcpu_chosen_alloc, "page")) { - if (!strcmp(pcpu_chosen_alloc, "lpage")) + if (pcpu_chosen_fc != PCPU_FC_AUTO) { + if (pcpu_chosen_fc != PCPU_FC_PAGE) { + if (pcpu_chosen_fc == PCPU_FC_LPAGE) ret = setup_pcpu_lpage(static_size, true); - else if (!strcmp(pcpu_chosen_alloc, "embed")) - ret = setup_pcpu_embed(static_size, true); else - pr_warning("PERCPU: unknown allocator %s " - "specified\n", pcpu_chosen_alloc); + ret = setup_pcpu_embed(static_size, true); + if (ret < 0) pr_warning("PERCPU: %s allocator failed (%zd), " "falling back to page size\n", - pcpu_chosen_alloc, ret); + pcpu_fc_names[pcpu_chosen_fc], ret); } } else { ret = setup_pcpu_lpage(static_size, false); diff --git a/include/linux/percpu.h b/include/linux/percpu.h index e26788e0da4a..9be05cbe5ee0 100644 --- a/include/linux/percpu.h +++ b/include/linux/percpu.h @@ -59,6 +59,18 @@ extern void *pcpu_base_addr; extern const int *pcpu_unit_map; +enum pcpu_fc { + PCPU_FC_AUTO, + PCPU_FC_EMBED, + PCPU_FC_PAGE, + PCPU_FC_LPAGE, + + PCPU_FC_NR, +}; +extern const char *pcpu_fc_names[PCPU_FC_NR]; + +extern enum pcpu_fc pcpu_chosen_fc; + typedef void * (*pcpu_fc_alloc_fn_t)(unsigned int cpu, size_t size); typedef void (*pcpu_fc_free_fn_t)(void *ptr, size_t size); typedef void (*pcpu_fc_populate_pte_fn_t)(unsigned long addr); diff --git a/mm/percpu.c b/mm/percpu.c index 7971997de310..7fb40bb1555a 100644 --- a/mm/percpu.c +++ b/mm/percpu.c @@ -1414,6 +1414,38 @@ size_t __init pcpu_setup_first_chunk(size_t static_size, size_t reserved_size, return pcpu_unit_size; } +const char *pcpu_fc_names[PCPU_FC_NR] __initdata = { + [PCPU_FC_AUTO] = "auto", + [PCPU_FC_EMBED] = "embed", + [PCPU_FC_PAGE] = "page", + [PCPU_FC_LPAGE] = "lpage", +}; + +enum pcpu_fc pcpu_chosen_fc __initdata = PCPU_FC_AUTO; + +static int __init percpu_alloc_setup(char *str) +{ + if (0) + /* nada */; +#ifdef CONFIG_NEED_PER_CPU_EMBED_FIRST_CHUNK + else if (!strcmp(str, "embed")) + pcpu_chosen_fc = PCPU_FC_EMBED; +#endif +#ifdef CONFIG_NEED_PER_CPU_PAGE_FIRST_CHUNK + else if (!strcmp(str, "page")) + pcpu_chosen_fc = PCPU_FC_PAGE; +#endif +#ifdef CONFIG_NEED_PER_CPU_LPAGE_FIRST_CHUNK + else if (!strcmp(str, "lpage")) + pcpu_chosen_fc = PCPU_FC_LPAGE; +#endif + else + pr_warning("PERCPU: unknown allocator %s specified\n", str); + + return 0; +} +early_param("percpu_alloc", percpu_alloc_setup); + static inline size_t pcpu_calc_fc_sizes(size_t static_size, size_t reserved_size, ssize_t *dyn_sizep) -- cgit v1.2.3 From 9a7737691e90d3cce0e5248f91826c50e5aa3fcf Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Fri, 14 Aug 2009 15:00:50 +0900 Subject: percpu: drop @static_size from first chunk allocators First chunk allocators assume percpu areas have been linked using one of PERCPU_*() macros and depend on __per_cpu_load symbol defined by those macros, so there isn't much point in passing in static area size explicitly when it can be easily calculated from __per_cpu_start and __per_cpu_end. Drop @static_size from all percpu first chunk allocators and helpers. Signed-off-by: Tejun Heo --- arch/x86/kernel/setup_percpu.c | 34 +++++++++++++++------------------- include/linux/percpu.h | 18 ++++++++---------- mm/percpu.c | 29 +++++++++++++---------------- 3 files changed, 36 insertions(+), 45 deletions(-) (limited to 'include') diff --git a/arch/x86/kernel/setup_percpu.c b/arch/x86/kernel/setup_percpu.c index b961d99e6416..8aad486c688f 100644 --- a/arch/x86/kernel/setup_percpu.c +++ b/arch/x86/kernel/setup_percpu.c @@ -157,7 +157,7 @@ static int pcpu_lpage_cpu_distance(unsigned int from, unsigned int to) return REMOTE_DISTANCE; } -static ssize_t __init setup_pcpu_lpage(size_t static_size, bool chosen) +static ssize_t __init setup_pcpu_lpage(bool chosen) { size_t reserve = PERCPU_MODULE_RESERVE + PERCPU_DYNAMIC_RESERVE; size_t dyn_size = reserve - PERCPU_FIRST_CHUNK_RESERVE; @@ -184,8 +184,7 @@ static ssize_t __init setup_pcpu_lpage(size_t static_size, bool chosen) return -ENOMEM; } - ret = pcpu_lpage_build_unit_map(static_size, - PERCPU_FIRST_CHUNK_RESERVE, + ret = pcpu_lpage_build_unit_map(PERCPU_FIRST_CHUNK_RESERVE, &dyn_size, &unit_size, PMD_SIZE, unit_map, pcpu_lpage_cpu_distance); if (ret < 0) { @@ -208,9 +207,8 @@ static ssize_t __init setup_pcpu_lpage(size_t static_size, bool chosen) } } - ret = pcpu_lpage_first_chunk(static_size, PERCPU_FIRST_CHUNK_RESERVE, - dyn_size, unit_size, PMD_SIZE, - unit_map, nr_units, + ret = pcpu_lpage_first_chunk(PERCPU_FIRST_CHUNK_RESERVE, dyn_size, + unit_size, PMD_SIZE, unit_map, nr_units, pcpu_fc_alloc, pcpu_fc_free, pcpul_map); out_free: if (ret < 0) @@ -218,7 +216,7 @@ out_free: return ret; } #else -static ssize_t __init setup_pcpu_lpage(size_t static_size, bool chosen) +static ssize_t __init setup_pcpu_lpage(bool chosen) { return -EINVAL; } @@ -232,7 +230,7 @@ static ssize_t __init setup_pcpu_lpage(size_t static_size, bool chosen) * mapping so that it can use PMD mapping without additional TLB * pressure. */ -static ssize_t __init setup_pcpu_embed(size_t static_size, bool chosen) +static ssize_t __init setup_pcpu_embed(bool chosen) { size_t reserve = PERCPU_MODULE_RESERVE + PERCPU_DYNAMIC_RESERVE; @@ -244,7 +242,7 @@ static ssize_t __init setup_pcpu_embed(size_t static_size, bool chosen) if (!chosen && (!cpu_has_pse || pcpu_need_numa())) return -EINVAL; - return pcpu_embed_first_chunk(static_size, PERCPU_FIRST_CHUNK_RESERVE, + return pcpu_embed_first_chunk(PERCPU_FIRST_CHUNK_RESERVE, reserve - PERCPU_FIRST_CHUNK_RESERVE); } @@ -260,9 +258,9 @@ static void __init pcpup_populate_pte(unsigned long addr) populate_extra_pte(addr); } -static ssize_t __init setup_pcpu_page(size_t static_size) +static ssize_t __init setup_pcpu_page(void) { - return pcpu_page_first_chunk(static_size, PERCPU_FIRST_CHUNK_RESERVE, + return pcpu_page_first_chunk(PERCPU_FIRST_CHUNK_RESERVE, pcpu_fc_alloc, pcpu_fc_free, pcpup_populate_pte); } @@ -282,7 +280,6 @@ static inline void setup_percpu_segment(int cpu) void __init setup_per_cpu_areas(void) { - size_t static_size = __per_cpu_end - __per_cpu_start; unsigned int cpu; unsigned long delta; size_t pcpu_unit_size; @@ -300,9 +297,9 @@ void __init setup_per_cpu_areas(void) if (pcpu_chosen_fc != PCPU_FC_AUTO) { if (pcpu_chosen_fc != PCPU_FC_PAGE) { if (pcpu_chosen_fc == PCPU_FC_LPAGE) - ret = setup_pcpu_lpage(static_size, true); + ret = setup_pcpu_lpage(true); else - ret = setup_pcpu_embed(static_size, true); + ret = setup_pcpu_embed(true); if (ret < 0) pr_warning("PERCPU: %s allocator failed (%zd), " @@ -310,15 +307,14 @@ void __init setup_per_cpu_areas(void) pcpu_fc_names[pcpu_chosen_fc], ret); } } else { - ret = setup_pcpu_lpage(static_size, false); + ret = setup_pcpu_lpage(false); if (ret < 0) - ret = setup_pcpu_embed(static_size, false); + ret = setup_pcpu_embed(false); } if (ret < 0) - ret = setup_pcpu_page(static_size); + ret = setup_pcpu_page(); if (ret < 0) - panic("cannot allocate static percpu area (%zu bytes, err=%zd)", - static_size, ret); + panic("cannot initialize percpu area (err=%zd)", ret); pcpu_unit_size = ret; diff --git a/include/linux/percpu.h b/include/linux/percpu.h index 9be05cbe5ee0..be2fc8fb9b6f 100644 --- a/include/linux/percpu.h +++ b/include/linux/percpu.h @@ -84,13 +84,12 @@ extern size_t __init pcpu_setup_first_chunk( #ifdef CONFIG_NEED_PER_CPU_EMBED_FIRST_CHUNK extern ssize_t __init pcpu_embed_first_chunk( - size_t static_size, size_t reserved_size, - ssize_t dyn_size); + size_t reserved_size, ssize_t dyn_size); #endif #ifdef CONFIG_NEED_PER_CPU_PAGE_FIRST_CHUNK extern ssize_t __init pcpu_page_first_chunk( - size_t static_size, size_t reserved_size, + size_t reserved_size, pcpu_fc_alloc_fn_t alloc_fn, pcpu_fc_free_fn_t free_fn, pcpu_fc_populate_pte_fn_t populate_pte_fn); @@ -98,16 +97,15 @@ extern ssize_t __init pcpu_page_first_chunk( #ifdef CONFIG_NEED_PER_CPU_LPAGE_FIRST_CHUNK extern int __init pcpu_lpage_build_unit_map( - size_t static_size, size_t reserved_size, - ssize_t *dyn_sizep, size_t *unit_sizep, - size_t lpage_size, int *unit_map, + size_t reserved_size, ssize_t *dyn_sizep, + size_t *unit_sizep, size_t lpage_size, + int *unit_map, pcpu_fc_cpu_distance_fn_t cpu_distance_fn); extern ssize_t __init pcpu_lpage_first_chunk( - size_t static_size, size_t reserved_size, - size_t dyn_size, size_t unit_size, - size_t lpage_size, const int *unit_map, - int nr_units, + size_t reserved_size, size_t dyn_size, + size_t unit_size, size_t lpage_size, + const int *unit_map, int nr_units, pcpu_fc_alloc_fn_t alloc_fn, pcpu_fc_free_fn_t free_fn, pcpu_fc_map_fn_t map_fn); diff --git a/mm/percpu.c b/mm/percpu.c index 7fb40bb1555a..e2ac58a39bb2 100644 --- a/mm/percpu.c +++ b/mm/percpu.c @@ -1464,7 +1464,6 @@ static inline size_t pcpu_calc_fc_sizes(size_t static_size, !defined(CONFIG_HAVE_SETUP_PER_CPU_AREA) /** * pcpu_embed_first_chunk - embed the first percpu chunk into bootmem - * @static_size: the size of static percpu area in bytes * @reserved_size: the size of reserved percpu area in bytes * @dyn_size: free size for dynamic allocation in bytes, -1 for auto * @@ -1489,9 +1488,9 @@ static inline size_t pcpu_calc_fc_sizes(size_t static_size, * The determined pcpu_unit_size which can be used to initialize * percpu access on success, -errno on failure. */ -ssize_t __init pcpu_embed_first_chunk(size_t static_size, size_t reserved_size, - ssize_t dyn_size) +ssize_t __init pcpu_embed_first_chunk(size_t reserved_size, ssize_t dyn_size) { + const size_t static_size = __per_cpu_end - __per_cpu_start; size_t size_sum, unit_size, chunk_size; void *base; unsigned int cpu; @@ -1536,7 +1535,6 @@ ssize_t __init pcpu_embed_first_chunk(size_t static_size, size_t reserved_size, #ifdef CONFIG_NEED_PER_CPU_PAGE_FIRST_CHUNK /** * pcpu_page_first_chunk - map the first chunk using PAGE_SIZE pages - * @static_size: the size of static percpu area in bytes * @reserved_size: the size of reserved percpu area in bytes * @alloc_fn: function to allocate percpu page, always called with PAGE_SIZE * @free_fn: funtion to free percpu page, always called with PAGE_SIZE @@ -1552,12 +1550,13 @@ ssize_t __init pcpu_embed_first_chunk(size_t static_size, size_t reserved_size, * The determined pcpu_unit_size which can be used to initialize * percpu access on success, -errno on failure. */ -ssize_t __init pcpu_page_first_chunk(size_t static_size, size_t reserved_size, +ssize_t __init pcpu_page_first_chunk(size_t reserved_size, pcpu_fc_alloc_fn_t alloc_fn, pcpu_fc_free_fn_t free_fn, pcpu_fc_populate_pte_fn_t populate_pte_fn) { static struct vm_struct vm; + const size_t static_size = __per_cpu_end - __per_cpu_start; char psize_str[16]; int unit_pages; size_t pages_size; @@ -1641,7 +1640,6 @@ out_free_ar: #ifdef CONFIG_NEED_PER_CPU_LPAGE_FIRST_CHUNK /** * pcpu_lpage_build_unit_map - build unit_map for large page remapping - * @static_size: the size of static percpu area in bytes * @reserved_size: the size of reserved percpu area in bytes * @dyn_sizep: in/out parameter for dynamic size, -1 for auto * @unit_sizep: out parameter for unit size @@ -1661,13 +1659,14 @@ out_free_ar: * On success, fills in @unit_map, sets *@dyn_sizep, *@unit_sizep and * returns the number of units to be allocated. -errno on failure. */ -int __init pcpu_lpage_build_unit_map(size_t static_size, size_t reserved_size, - ssize_t *dyn_sizep, size_t *unit_sizep, - size_t lpage_size, int *unit_map, +int __init pcpu_lpage_build_unit_map(size_t reserved_size, ssize_t *dyn_sizep, + size_t *unit_sizep, size_t lpage_size, + int *unit_map, pcpu_fc_cpu_distance_fn_t cpu_distance_fn) { static int group_map[NR_CPUS] __initdata; static int group_cnt[NR_CPUS] __initdata; + const size_t static_size = __per_cpu_end - __per_cpu_start; int group_cnt_max = 0; size_t size_sum, min_unit_size, alloc_size; int upa, max_upa, uninitialized_var(best_upa); /* units_per_alloc */ @@ -1819,7 +1818,6 @@ static void __init pcpul_lpage_dump_cfg(const char *lvl, size_t static_size, /** * pcpu_lpage_first_chunk - remap the first percpu chunk using large page - * @static_size: the size of static percpu area in bytes * @reserved_size: the size of reserved percpu area in bytes * @dyn_size: free size for dynamic allocation in bytes * @unit_size: unit size in bytes @@ -1850,15 +1848,15 @@ static void __init pcpul_lpage_dump_cfg(const char *lvl, size_t static_size, * The determined pcpu_unit_size which can be used to initialize * percpu access on success, -errno on failure. */ -ssize_t __init pcpu_lpage_first_chunk(size_t static_size, size_t reserved_size, - size_t dyn_size, size_t unit_size, - size_t lpage_size, const int *unit_map, - int nr_units, +ssize_t __init pcpu_lpage_first_chunk(size_t reserved_size, size_t dyn_size, + size_t unit_size, size_t lpage_size, + const int *unit_map, int nr_units, pcpu_fc_alloc_fn_t alloc_fn, pcpu_fc_free_fn_t free_fn, pcpu_fc_map_fn_t map_fn) { static struct vm_struct vm; + const size_t static_size = __per_cpu_end - __per_cpu_start; size_t chunk_size = unit_size * nr_units; size_t map_size; unsigned int cpu; @@ -2037,7 +2035,6 @@ EXPORT_SYMBOL(__per_cpu_offset); void __init setup_per_cpu_areas(void) { - size_t static_size = __per_cpu_end - __per_cpu_start; ssize_t unit_size; unsigned long delta; unsigned int cpu; @@ -2046,7 +2043,7 @@ void __init setup_per_cpu_areas(void) * Always reserve area for module percpu variables. That's * what the legacy allocator did. */ - unit_size = pcpu_embed_first_chunk(static_size, PERCPU_MODULE_RESERVE, + unit_size = pcpu_embed_first_chunk(PERCPU_MODULE_RESERVE, PERCPU_DYNAMIC_RESERVE); if (unit_size < 0) panic("Failed to initialized percpu areas."); -- cgit v1.2.3 From 1d9d32572163b30be81dbe1409dfa7ea9763d0e8 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Fri, 14 Aug 2009 15:00:50 +0900 Subject: percpu: make @dyn_size mandatory for pcpu_setup_first_chunk() Now that all actual first chunk allocation and copying happen in the first chunk allocators and helpers, there's no reason for pcpu_setup_first_chunk() to try to determine @dyn_size automatically. The only left user is page first chunk allocator. Make it determine dyn_size like other allocators and make @dyn_size mandatory for pcpu_setup_first_chunk(). Signed-off-by: Tejun Heo --- include/linux/percpu.h | 2 +- mm/percpu.c | 39 +++++++++++++++++++-------------------- 2 files changed, 20 insertions(+), 21 deletions(-) (limited to 'include') diff --git a/include/linux/percpu.h b/include/linux/percpu.h index be2fc8fb9b6f..0cfdd14d096a 100644 --- a/include/linux/percpu.h +++ b/include/linux/percpu.h @@ -79,7 +79,7 @@ typedef void (*pcpu_fc_map_fn_t)(void *ptr, size_t size, void *addr); extern size_t __init pcpu_setup_first_chunk( size_t static_size, size_t reserved_size, - ssize_t dyn_size, size_t unit_size, + size_t dyn_size, size_t unit_size, void *base_addr, const int *unit_map); #ifdef CONFIG_NEED_PER_CPU_EMBED_FIRST_CHUNK diff --git a/mm/percpu.c b/mm/percpu.c index e2ac58a39bb2..287f59cc5fb9 100644 --- a/mm/percpu.c +++ b/mm/percpu.c @@ -1235,7 +1235,7 @@ EXPORT_SYMBOL_GPL(free_percpu); * pcpu_setup_first_chunk - initialize the first percpu chunk * @static_size: the size of static percpu area in bytes * @reserved_size: the size of reserved percpu area in bytes, 0 for none - * @dyn_size: free size for dynamic allocation in bytes, -1 for auto + * @dyn_size: free size for dynamic allocation in bytes * @unit_size: unit size in bytes, must be multiple of PAGE_SIZE * @base_addr: mapped address * @unit_map: cpu -> unit map, NULL for sequential mapping @@ -1252,10 +1252,9 @@ EXPORT_SYMBOL_GPL(free_percpu); * limited offset range for symbol relocations to guarantee module * percpu symbols fall inside the relocatable range. * - * @dyn_size, if non-negative, determines the number of bytes - * available for dynamic allocation in the first chunk. Specifying - * non-negative value makes percpu leave alone the area beyond - * @static_size + @reserved_size + @dyn_size. + * @dyn_size determines the number of bytes available for dynamic + * allocation in the first chunk. The area between @static_size + + * @reserved_size + @dyn_size and @unit_size is unused. * * @unit_size specifies unit size and must be aligned to PAGE_SIZE and * equal to or larger than @static_size + @reserved_size + if @@ -1276,13 +1275,12 @@ EXPORT_SYMBOL_GPL(free_percpu); * percpu access. */ size_t __init pcpu_setup_first_chunk(size_t static_size, size_t reserved_size, - ssize_t dyn_size, size_t unit_size, + size_t dyn_size, size_t unit_size, void *base_addr, const int *unit_map) { static struct vm_struct first_vm; static int smap[2], dmap[2]; - size_t size_sum = static_size + reserved_size + - (dyn_size >= 0 ? dyn_size : 0); + size_t size_sum = static_size + reserved_size + dyn_size; struct pcpu_chunk *schunk, *dchunk = NULL; unsigned int cpu, tcpu; int i; @@ -1345,9 +1343,6 @@ size_t __init pcpu_setup_first_chunk(size_t static_size, size_t reserved_size, pcpu_chunk_struct_size = sizeof(struct pcpu_chunk) + BITS_TO_LONGS(pcpu_unit_pages) * sizeof(unsigned long); - if (dyn_size < 0) - dyn_size = pcpu_unit_size - static_size - reserved_size; - first_vm.flags = VM_ALLOC; first_vm.size = pcpu_chunk_size; first_vm.addr = base_addr; @@ -1557,6 +1552,8 @@ ssize_t __init pcpu_page_first_chunk(size_t reserved_size, { static struct vm_struct vm; const size_t static_size = __per_cpu_end - __per_cpu_start; + ssize_t dyn_size = -1; + size_t size_sum, unit_size; char psize_str[16]; int unit_pages; size_t pages_size; @@ -1567,8 +1564,9 @@ ssize_t __init pcpu_page_first_chunk(size_t reserved_size, snprintf(psize_str, sizeof(psize_str), "%luK", PAGE_SIZE >> 10); - unit_pages = PFN_UP(max_t(size_t, static_size + reserved_size, - PCPU_MIN_UNIT_SIZE)); + size_sum = pcpu_calc_fc_sizes(static_size, reserved_size, &dyn_size); + unit_size = max_t(size_t, size_sum, PCPU_MIN_UNIT_SIZE); + unit_pages = unit_size >> PAGE_SHIFT; /* unaligned allocations can't be freed, round up to page size */ pages_size = PFN_ALIGN(unit_pages * nr_cpu_ids * sizeof(pages[0])); @@ -1591,12 +1589,12 @@ ssize_t __init pcpu_page_first_chunk(size_t reserved_size, /* allocate vm area, map the pages and copy static data */ vm.flags = VM_ALLOC; - vm.size = nr_cpu_ids * unit_pages << PAGE_SHIFT; + vm.size = nr_cpu_ids * unit_size; vm_area_register_early(&vm, PAGE_SIZE); for_each_possible_cpu(cpu) { - unsigned long unit_addr = (unsigned long)vm.addr + - (cpu * unit_pages << PAGE_SHIFT); + unsigned long unit_addr = + (unsigned long)vm.addr + cpu * unit_size; for (i = 0; i < unit_pages; i++) populate_pte_fn(unit_addr + (i << PAGE_SHIFT)); @@ -1620,11 +1618,12 @@ ssize_t __init pcpu_page_first_chunk(size_t reserved_size, } /* we're ready, commit */ - pr_info("PERCPU: %d %s pages/cpu @%p s%zu r%zu\n", - unit_pages, psize_str, vm.addr, static_size, reserved_size); + pr_info("PERCPU: %d %s pages/cpu @%p s%zu r%zu d%zu\n", + unit_pages, psize_str, vm.addr, static_size, reserved_size, + dyn_size); - ret = pcpu_setup_first_chunk(static_size, reserved_size, -1, - unit_pages << PAGE_SHIFT, vm.addr, NULL); + ret = pcpu_setup_first_chunk(static_size, reserved_size, dyn_size, + unit_size, vm.addr, NULL); goto out_free_ar; enomem: -- cgit v1.2.3 From 3cbc85652767c38b252c8de55f9fd180b29e4c0d Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Fri, 14 Aug 2009 15:00:50 +0900 Subject: percpu: add @align to pcpu_fc_alloc_fn_t pcpu_fc_alloc_fn_t is about to see more interesting usage, add @align parameter. Signed-off-by: Tejun Heo --- arch/x86/kernel/setup_percpu.c | 4 ++-- include/linux/percpu.h | 3 ++- mm/percpu.c | 4 ++-- 3 files changed, 6 insertions(+), 5 deletions(-) (limited to 'include') diff --git a/arch/x86/kernel/setup_percpu.c b/arch/x86/kernel/setup_percpu.c index 8aad486c688f..660cde133141 100644 --- a/arch/x86/kernel/setup_percpu.c +++ b/arch/x86/kernel/setup_percpu.c @@ -126,9 +126,9 @@ static void * __init pcpu_alloc_bootmem(unsigned int cpu, unsigned long size, /* * Helpers for first chunk memory allocation */ -static void * __init pcpu_fc_alloc(unsigned int cpu, size_t size) +static void * __init pcpu_fc_alloc(unsigned int cpu, size_t size, size_t align) { - return pcpu_alloc_bootmem(cpu, size, size); + return pcpu_alloc_bootmem(cpu, size, align); } static void __init pcpu_fc_free(void *ptr, size_t size) diff --git a/include/linux/percpu.h b/include/linux/percpu.h index 0cfdd14d096a..d385dbcf190b 100644 --- a/include/linux/percpu.h +++ b/include/linux/percpu.h @@ -71,7 +71,8 @@ extern const char *pcpu_fc_names[PCPU_FC_NR]; extern enum pcpu_fc pcpu_chosen_fc; -typedef void * (*pcpu_fc_alloc_fn_t)(unsigned int cpu, size_t size); +typedef void * (*pcpu_fc_alloc_fn_t)(unsigned int cpu, size_t size, + size_t align); typedef void (*pcpu_fc_free_fn_t)(void *ptr, size_t size); typedef void (*pcpu_fc_populate_pte_fn_t)(unsigned long addr); typedef int (pcpu_fc_cpu_distance_fn_t)(unsigned int from, unsigned int to); diff --git a/mm/percpu.c b/mm/percpu.c index 287f59cc5fb9..3316e3aac7ee 100644 --- a/mm/percpu.c +++ b/mm/percpu.c @@ -1578,7 +1578,7 @@ ssize_t __init pcpu_page_first_chunk(size_t reserved_size, for (i = 0; i < unit_pages; i++) { void *ptr; - ptr = alloc_fn(cpu, PAGE_SIZE); + ptr = alloc_fn(cpu, PAGE_SIZE, PAGE_SIZE); if (!ptr) { pr_warning("PERCPU: failed to allocate %s page " "for cpu%u\n", psize_str, cpu); @@ -1888,7 +1888,7 @@ ssize_t __init pcpu_lpage_first_chunk(size_t reserved_size, size_t dyn_size, goto found; continue; found: - ptr = alloc_fn(cpu, lpage_size); + ptr = alloc_fn(cpu, lpage_size, lpage_size); if (!ptr) { pr_warning("PERCPU: failed to allocate large page " "for cpu%u\n", cpu); -- cgit v1.2.3 From 033e48fb82958053113178264ddb9d5038d5e38b Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Fri, 14 Aug 2009 15:00:51 +0900 Subject: percpu: move pcpu_lpage_build_unit_map() and pcpul_lpage_dump_cfg() upward Unit map handling will be generalized and extended and used for embedding sparse first chunk and other purposes. Relocate two unit_map related functions upward in preparation. This patch just moves the code without any actual change. Signed-off-by: Tejun Heo --- include/linux/percpu.h | 14 +- mm/percpu.c | 339 +++++++++++++++++++++++++------------------------ 2 files changed, 180 insertions(+), 173 deletions(-) (limited to 'include') diff --git a/include/linux/percpu.h b/include/linux/percpu.h index d385dbcf190b..570fb18de2ba 100644 --- a/include/linux/percpu.h +++ b/include/linux/percpu.h @@ -78,6 +78,14 @@ typedef void (*pcpu_fc_populate_pte_fn_t)(unsigned long addr); typedef int (pcpu_fc_cpu_distance_fn_t)(unsigned int from, unsigned int to); typedef void (*pcpu_fc_map_fn_t)(void *ptr, size_t size, void *addr); +#ifdef CONFIG_NEED_PER_CPU_LPAGE_FIRST_CHUNK +extern int __init pcpu_lpage_build_unit_map( + size_t reserved_size, ssize_t *dyn_sizep, + size_t *unit_sizep, size_t lpage_size, + int *unit_map, + pcpu_fc_cpu_distance_fn_t cpu_distance_fn); +#endif + extern size_t __init pcpu_setup_first_chunk( size_t static_size, size_t reserved_size, size_t dyn_size, size_t unit_size, @@ -97,12 +105,6 @@ extern ssize_t __init pcpu_page_first_chunk( #endif #ifdef CONFIG_NEED_PER_CPU_LPAGE_FIRST_CHUNK -extern int __init pcpu_lpage_build_unit_map( - size_t reserved_size, ssize_t *dyn_sizep, - size_t *unit_sizep, size_t lpage_size, - int *unit_map, - pcpu_fc_cpu_distance_fn_t cpu_distance_fn); - extern ssize_t __init pcpu_lpage_first_chunk( size_t reserved_size, size_t dyn_size, size_t unit_size, size_t lpage_size, diff --git a/mm/percpu.c b/mm/percpu.c index 3316e3aac7ee..2b9c4b2a2fc0 100644 --- a/mm/percpu.c +++ b/mm/percpu.c @@ -1231,6 +1231,178 @@ void free_percpu(void *ptr) } EXPORT_SYMBOL_GPL(free_percpu); +static inline size_t pcpu_calc_fc_sizes(size_t static_size, + size_t reserved_size, + ssize_t *dyn_sizep) +{ + size_t size_sum; + + size_sum = PFN_ALIGN(static_size + reserved_size + + (*dyn_sizep >= 0 ? *dyn_sizep : 0)); + if (*dyn_sizep != 0) + *dyn_sizep = size_sum - static_size - reserved_size; + + return size_sum; +} + +#ifdef CONFIG_NEED_PER_CPU_LPAGE_FIRST_CHUNK +/** + * pcpu_lpage_build_unit_map - build unit_map for large page remapping + * @reserved_size: the size of reserved percpu area in bytes + * @dyn_sizep: in/out parameter for dynamic size, -1 for auto + * @unit_sizep: out parameter for unit size + * @unit_map: unit_map to be filled + * @cpu_distance_fn: callback to determine distance between cpus + * + * This function builds cpu -> unit map and determine other parameters + * considering needed percpu size, large page size and distances + * between CPUs in NUMA. + * + * CPUs which are of LOCAL_DISTANCE both ways are grouped together and + * may share units in the same large page. The returned configuration + * is guaranteed to have CPUs on different nodes on different large + * pages and >=75% usage of allocated virtual address space. + * + * RETURNS: + * On success, fills in @unit_map, sets *@dyn_sizep, *@unit_sizep and + * returns the number of units to be allocated. -errno on failure. + */ +int __init pcpu_lpage_build_unit_map(size_t reserved_size, ssize_t *dyn_sizep, + size_t *unit_sizep, size_t lpage_size, + int *unit_map, + pcpu_fc_cpu_distance_fn_t cpu_distance_fn) +{ + static int group_map[NR_CPUS] __initdata; + static int group_cnt[NR_CPUS] __initdata; + const size_t static_size = __per_cpu_end - __per_cpu_start; + int group_cnt_max = 0; + size_t size_sum, min_unit_size, alloc_size; + int upa, max_upa, uninitialized_var(best_upa); /* units_per_alloc */ + int last_allocs; + unsigned int cpu, tcpu; + int group, unit; + + /* + * Determine min_unit_size, alloc_size and max_upa such that + * alloc_size is multiple of lpage_size and is the smallest + * which can accomodate 4k aligned segments which are equal to + * or larger than min_unit_size. + */ + size_sum = pcpu_calc_fc_sizes(static_size, reserved_size, dyn_sizep); + min_unit_size = max_t(size_t, size_sum, PCPU_MIN_UNIT_SIZE); + + alloc_size = roundup(min_unit_size, lpage_size); + upa = alloc_size / min_unit_size; + while (alloc_size % upa || ((alloc_size / upa) & ~PAGE_MASK)) + upa--; + max_upa = upa; + + /* group cpus according to their proximity */ + for_each_possible_cpu(cpu) { + group = 0; + next_group: + for_each_possible_cpu(tcpu) { + if (cpu == tcpu) + break; + if (group_map[tcpu] == group && + (cpu_distance_fn(cpu, tcpu) > LOCAL_DISTANCE || + cpu_distance_fn(tcpu, cpu) > LOCAL_DISTANCE)) { + group++; + goto next_group; + } + } + group_map[cpu] = group; + group_cnt[group]++; + group_cnt_max = max(group_cnt_max, group_cnt[group]); + } + + /* + * Expand unit size until address space usage goes over 75% + * and then as much as possible without using more address + * space. + */ + last_allocs = INT_MAX; + for (upa = max_upa; upa; upa--) { + int allocs = 0, wasted = 0; + + if (alloc_size % upa || ((alloc_size / upa) & ~PAGE_MASK)) + continue; + + for (group = 0; group_cnt[group]; group++) { + int this_allocs = DIV_ROUND_UP(group_cnt[group], upa); + allocs += this_allocs; + wasted += this_allocs * upa - group_cnt[group]; + } + + /* + * Don't accept if wastage is over 25%. The + * greater-than comparison ensures upa==1 always + * passes the following check. + */ + if (wasted > num_possible_cpus() / 3) + continue; + + /* and then don't consume more memory */ + if (allocs > last_allocs) + break; + last_allocs = allocs; + best_upa = upa; + } + *unit_sizep = alloc_size / best_upa; + + /* assign units to cpus accordingly */ + unit = 0; + for (group = 0; group_cnt[group]; group++) { + for_each_possible_cpu(cpu) + if (group_map[cpu] == group) + unit_map[cpu] = unit++; + unit = roundup(unit, best_upa); + } + + return unit; /* unit contains aligned number of units */ +} + +static bool __init pcpul_unit_to_cpu(int unit, const int *unit_map, + unsigned int *cpup); + +static void __init pcpul_lpage_dump_cfg(const char *lvl, size_t static_size, + size_t reserved_size, size_t dyn_size, + size_t unit_size, size_t lpage_size, + const int *unit_map, int nr_units) +{ + int width = 1, v = nr_units; + char empty_str[] = "--------"; + int upl, lpl; /* units per lpage, lpage per line */ + unsigned int cpu; + int lpage, unit; + + while (v /= 10) + width++; + empty_str[min_t(int, width, sizeof(empty_str) - 1)] = '\0'; + + upl = max_t(int, lpage_size / unit_size, 1); + lpl = rounddown_pow_of_two(max_t(int, 60 / (upl * (width + 1) + 2), 1)); + + printk("%spcpu-lpage: sta/res/dyn=%zu/%zu/%zu unit=%zu lpage=%zu", lvl, + static_size, reserved_size, dyn_size, unit_size, lpage_size); + + for (lpage = 0, unit = 0; unit < nr_units; unit++) { + if (!(unit % upl)) { + if (!(lpage++ % lpl)) { + printk("\n"); + printk("%spcpu-lpage: ", lvl); + } else + printk("| "); + } + if (pcpul_unit_to_cpu(unit, unit_map, &cpu)) + printk("%0*d ", width, cpu); + else + printk("%s ", empty_str); + } + printk("\n"); +} +#endif + /** * pcpu_setup_first_chunk - initialize the first percpu chunk * @static_size: the size of static percpu area in bytes @@ -1441,20 +1613,6 @@ static int __init percpu_alloc_setup(char *str) } early_param("percpu_alloc", percpu_alloc_setup); -static inline size_t pcpu_calc_fc_sizes(size_t static_size, - size_t reserved_size, - ssize_t *dyn_sizep) -{ - size_t size_sum; - - size_sum = PFN_ALIGN(static_size + reserved_size + - (*dyn_sizep >= 0 ? *dyn_sizep : 0)); - if (*dyn_sizep != 0) - *dyn_sizep = size_sum - static_size - reserved_size; - - return size_sum; -} - #if defined(CONFIG_NEED_PER_CPU_EMBED_FIRST_CHUNK) || \ !defined(CONFIG_HAVE_SETUP_PER_CPU_AREA) /** @@ -1637,122 +1795,6 @@ out_free_ar: #endif /* CONFIG_NEED_PER_CPU_PAGE_FIRST_CHUNK */ #ifdef CONFIG_NEED_PER_CPU_LPAGE_FIRST_CHUNK -/** - * pcpu_lpage_build_unit_map - build unit_map for large page remapping - * @reserved_size: the size of reserved percpu area in bytes - * @dyn_sizep: in/out parameter for dynamic size, -1 for auto - * @unit_sizep: out parameter for unit size - * @unit_map: unit_map to be filled - * @cpu_distance_fn: callback to determine distance between cpus - * - * This function builds cpu -> unit map and determine other parameters - * considering needed percpu size, large page size and distances - * between CPUs in NUMA. - * - * CPUs which are of LOCAL_DISTANCE both ways are grouped together and - * may share units in the same large page. The returned configuration - * is guaranteed to have CPUs on different nodes on different large - * pages and >=75% usage of allocated virtual address space. - * - * RETURNS: - * On success, fills in @unit_map, sets *@dyn_sizep, *@unit_sizep and - * returns the number of units to be allocated. -errno on failure. - */ -int __init pcpu_lpage_build_unit_map(size_t reserved_size, ssize_t *dyn_sizep, - size_t *unit_sizep, size_t lpage_size, - int *unit_map, - pcpu_fc_cpu_distance_fn_t cpu_distance_fn) -{ - static int group_map[NR_CPUS] __initdata; - static int group_cnt[NR_CPUS] __initdata; - const size_t static_size = __per_cpu_end - __per_cpu_start; - int group_cnt_max = 0; - size_t size_sum, min_unit_size, alloc_size; - int upa, max_upa, uninitialized_var(best_upa); /* units_per_alloc */ - int last_allocs; - unsigned int cpu, tcpu; - int group, unit; - - /* - * Determine min_unit_size, alloc_size and max_upa such that - * alloc_size is multiple of lpage_size and is the smallest - * which can accomodate 4k aligned segments which are equal to - * or larger than min_unit_size. - */ - size_sum = pcpu_calc_fc_sizes(static_size, reserved_size, dyn_sizep); - min_unit_size = max_t(size_t, size_sum, PCPU_MIN_UNIT_SIZE); - - alloc_size = roundup(min_unit_size, lpage_size); - upa = alloc_size / min_unit_size; - while (alloc_size % upa || ((alloc_size / upa) & ~PAGE_MASK)) - upa--; - max_upa = upa; - - /* group cpus according to their proximity */ - for_each_possible_cpu(cpu) { - group = 0; - next_group: - for_each_possible_cpu(tcpu) { - if (cpu == tcpu) - break; - if (group_map[tcpu] == group && - (cpu_distance_fn(cpu, tcpu) > LOCAL_DISTANCE || - cpu_distance_fn(tcpu, cpu) > LOCAL_DISTANCE)) { - group++; - goto next_group; - } - } - group_map[cpu] = group; - group_cnt[group]++; - group_cnt_max = max(group_cnt_max, group_cnt[group]); - } - - /* - * Expand unit size until address space usage goes over 75% - * and then as much as possible without using more address - * space. - */ - last_allocs = INT_MAX; - for (upa = max_upa; upa; upa--) { - int allocs = 0, wasted = 0; - - if (alloc_size % upa || ((alloc_size / upa) & ~PAGE_MASK)) - continue; - - for (group = 0; group_cnt[group]; group++) { - int this_allocs = DIV_ROUND_UP(group_cnt[group], upa); - allocs += this_allocs; - wasted += this_allocs * upa - group_cnt[group]; - } - - /* - * Don't accept if wastage is over 25%. The - * greater-than comparison ensures upa==1 always - * passes the following check. - */ - if (wasted > num_possible_cpus() / 3) - continue; - - /* and then don't consume more memory */ - if (allocs > last_allocs) - break; - last_allocs = allocs; - best_upa = upa; - } - *unit_sizep = alloc_size / best_upa; - - /* assign units to cpus accordingly */ - unit = 0; - for (group = 0; group_cnt[group]; group++) { - for_each_possible_cpu(cpu) - if (group_map[cpu] == group) - unit_map[cpu] = unit++; - unit = roundup(unit, best_upa); - } - - return unit; /* unit contains aligned number of units */ -} - struct pcpul_ent { void *ptr; void *map_addr; @@ -1778,43 +1820,6 @@ static bool __init pcpul_unit_to_cpu(int unit, const int *unit_map, return false; } -static void __init pcpul_lpage_dump_cfg(const char *lvl, size_t static_size, - size_t reserved_size, size_t dyn_size, - size_t unit_size, size_t lpage_size, - const int *unit_map, int nr_units) -{ - int width = 1, v = nr_units; - char empty_str[] = "--------"; - int upl, lpl; /* units per lpage, lpage per line */ - unsigned int cpu; - int lpage, unit; - - while (v /= 10) - width++; - empty_str[min_t(int, width, sizeof(empty_str) - 1)] = '\0'; - - upl = max_t(int, lpage_size / unit_size, 1); - lpl = rounddown_pow_of_two(max_t(int, 60 / (upl * (width + 1) + 2), 1)); - - printk("%spcpu-lpage: sta/res/dyn=%zu/%zu/%zu unit=%zu lpage=%zu", lvl, - static_size, reserved_size, dyn_size, unit_size, lpage_size); - - for (lpage = 0, unit = 0; unit < nr_units; unit++) { - if (!(unit % upl)) { - if (!(lpage++ % lpl)) { - printk("\n"); - printk("%spcpu-lpage: ", lvl); - } else - printk("| "); - } - if (pcpul_unit_to_cpu(unit, unit_map, &cpu)) - printk("%0*d ", width, cpu); - else - printk("%s ", empty_str); - } - printk("\n"); -} - /** * pcpu_lpage_first_chunk - remap the first percpu chunk using large page * @reserved_size: the size of reserved percpu area in bytes -- cgit v1.2.3 From fd1e8a1fe2b54df6c185b4fa65f181f50b9c4d4e Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Fri, 14 Aug 2009 15:00:51 +0900 Subject: percpu: introduce pcpu_alloc_info and pcpu_group_info Till now, non-linear cpu->unit map was expressed using an integer array which maps each cpu to a unit and used only by lpage allocator. Although how many units have been placed in a single contiguos area (group) is known while building unit_map, the information is lost when the result is recorded into the unit_map array. For lpage allocator, as all allocations are done by lpages and whether two adjacent lpages are in the same group or not is irrelevant, this didn't cause any problem. Non-linear cpu->unit mapping will be used for sparse embedding and this grouping information is necessary for that. This patch introduces pcpu_alloc_info which contains all the information necessary for initializing percpu allocator. pcpu_alloc_info contains array of pcpu_group_info which describes how units are grouped and mapped to cpus. pcpu_group_info also has base_offset field to specify its offset from the chunk's base address. pcpu_build_alloc_info() initializes this field as if all groups are allocated back-to-back as is currently done but this will be used to sparsely place groups. pcpu_alloc_info is a rather complex data structure which contains a flexible array which in turn points to nested cpu_map arrays. * pcpu_alloc_alloc_info() and pcpu_free_alloc_info() are provided to help dealing with pcpu_alloc_info. * pcpu_lpage_build_unit_map() is updated to build pcpu_alloc_info, generalized and renamed to pcpu_build_alloc_info(). @cpu_distance_fn may be NULL indicating that all cpus are of LOCAL_DISTANCE. * pcpul_lpage_dump_cfg() is updated to process pcpu_alloc_info, generalized and renamed to pcpu_dump_alloc_info(). It now also prints which group each alloc unit belongs to. * pcpu_setup_first_chunk() now takes pcpu_alloc_info instead of the separate parameters. All first chunk allocators are updated to use pcpu_build_alloc_info() to build alloc_info and call pcpu_setup_first_chunk() with it. This has the side effect of packing units for sparse possible cpus. ie. if cpus 0, 2 and 4 are possible, they'll be assigned unit 0, 1 and 2 instead of 0, 2 and 4. * x86 setup_pcpu_lpage() is updated to deal with alloc_info. * sparc64 setup_per_cpu_areas() is updated to build alloc_info. Although the changes made by this patch are pretty pervasive, it doesn't cause any behavior difference other than packing of sparse cpus. It mostly changes how information is passed among initialization functions and makes room for more flexibility. Signed-off-by: Tejun Heo Cc: Ingo Molnar Cc: David Miller --- arch/sparc/kernel/smp_64.c | 24 +- arch/x86/kernel/setup_percpu.c | 38 ++- include/linux/percpu.h | 42 +++- mm/percpu.c | 529 +++++++++++++++++++++++++---------------- 4 files changed, 389 insertions(+), 244 deletions(-) (limited to 'include') diff --git a/arch/sparc/kernel/smp_64.c b/arch/sparc/kernel/smp_64.c index 9856d866b77b..a42a4a744d14 100644 --- a/arch/sparc/kernel/smp_64.c +++ b/arch/sparc/kernel/smp_64.c @@ -1475,17 +1475,29 @@ static void __init pcpu_map_range(unsigned long start, unsigned long end, void __init setup_per_cpu_areas(void) { - size_t dyn_size, static_size = __per_cpu_end - __per_cpu_start; static struct vm_struct vm; + struct pcpu_alloc_info *ai; unsigned long delta, cpu; size_t size_sum, pcpu_unit_size; size_t ptrs_size; void **ptrs; - size_sum = PFN_ALIGN(static_size + PERCPU_MODULE_RESERVE + + ai = pcpu_alloc_alloc_info(1, nr_cpu_ids); + + ai->static_size = __per_cpu_end - __per_cpu_start; + ai->reserved_size = PERCPU_MODULE_RESERVE; + + size_sum = PFN_ALIGN(ai->static_size + ai->reserved_size + PERCPU_DYNAMIC_RESERVE); - dyn_size = size_sum - static_size - PERCPU_MODULE_RESERVE; + ai->dyn_size = size_sum - ai->static_size - ai->reserved_size; + ai->unit_size = PCPU_CHUNK_SIZE; + ai->atom_size = PCPU_CHUNK_SIZE; + ai->alloc_size = PCPU_CHUNK_SIZE; + ai->groups[0].nr_units = nr_cpu_ids; + + for_each_possible_cpu(cpu) + ai->groups[0].cpu_map[cpu] = cpu; ptrs_size = PFN_ALIGN(nr_cpu_ids * sizeof(ptrs[0])); ptrs = alloc_bootmem(ptrs_size); @@ -1497,7 +1509,7 @@ void __init setup_per_cpu_areas(void) free_bootmem(__pa(ptrs[cpu] + size_sum), PCPU_CHUNK_SIZE - size_sum); - memcpy(ptrs[cpu], __per_cpu_load, static_size); + memcpy(ptrs[cpu], __per_cpu_load, ai->static_size); } /* allocate address and map */ @@ -1514,9 +1526,7 @@ void __init setup_per_cpu_areas(void) pcpu_map_range(start, end, virt_to_page(ptrs[cpu])); } - pcpu_unit_size = pcpu_setup_first_chunk(static_size, - PERCPU_MODULE_RESERVE, dyn_size, - PCPU_CHUNK_SIZE, vm.addr, NULL); + pcpu_unit_size = pcpu_setup_first_chunk(ai, vm.addr); free_bootmem(__pa(ptrs), ptrs_size); diff --git a/arch/x86/kernel/setup_percpu.c b/arch/x86/kernel/setup_percpu.c index 660cde133141..db5f9c49fec5 100644 --- a/arch/x86/kernel/setup_percpu.c +++ b/arch/x86/kernel/setup_percpu.c @@ -161,9 +161,7 @@ static ssize_t __init setup_pcpu_lpage(bool chosen) { size_t reserve = PERCPU_MODULE_RESERVE + PERCPU_DYNAMIC_RESERVE; size_t dyn_size = reserve - PERCPU_FIRST_CHUNK_RESERVE; - size_t unit_map_size, unit_size; - int *unit_map; - int nr_units; + struct pcpu_alloc_info *ai; ssize_t ret; /* on non-NUMA, embedding is better */ @@ -177,26 +175,22 @@ static ssize_t __init setup_pcpu_lpage(bool chosen) } /* allocate and build unit_map */ - unit_map_size = nr_cpu_ids * sizeof(int); - unit_map = alloc_bootmem_nopanic(unit_map_size); - if (!unit_map) { - pr_warning("PERCPU: failed to allocate unit_map\n"); - return -ENOMEM; + ai = pcpu_build_alloc_info(PERCPU_FIRST_CHUNK_RESERVE, dyn_size, + PMD_SIZE, pcpu_lpage_cpu_distance); + if (IS_ERR(ai)) { + pr_warning("PERCPU: failed to build unit_map (%ld)\n", + PTR_ERR(ai)); + return PTR_ERR(ai); } - ret = pcpu_lpage_build_unit_map(PERCPU_FIRST_CHUNK_RESERVE, - &dyn_size, &unit_size, PMD_SIZE, - unit_map, pcpu_lpage_cpu_distance); - if (ret < 0) { - pr_warning("PERCPU: failed to build unit_map\n"); - goto out_free; - } - nr_units = ret; - /* do the parameters look okay? */ if (!chosen) { size_t vm_size = VMALLOC_END - VMALLOC_START; - size_t tot_size = nr_units * unit_size; + size_t tot_size = 0; + int group; + + for (group = 0; group < ai->nr_groups; group++) + tot_size += ai->unit_size * ai->groups[group].nr_units; /* don't consume more than 20% of vmalloc area */ if (tot_size > vm_size / 5) { @@ -207,12 +201,10 @@ static ssize_t __init setup_pcpu_lpage(bool chosen) } } - ret = pcpu_lpage_first_chunk(PERCPU_FIRST_CHUNK_RESERVE, dyn_size, - unit_size, PMD_SIZE, unit_map, nr_units, - pcpu_fc_alloc, pcpu_fc_free, pcpul_map); + ret = pcpu_lpage_first_chunk(ai, pcpu_fc_alloc, pcpu_fc_free, + pcpul_map); out_free: - if (ret < 0) - free_bootmem(__pa(unit_map), unit_map_size); + pcpu_free_alloc_info(ai); return ret; } #else diff --git a/include/linux/percpu.h b/include/linux/percpu.h index 570fb18de2ba..77b86be8ce4f 100644 --- a/include/linux/percpu.h +++ b/include/linux/percpu.h @@ -59,6 +59,25 @@ extern void *pcpu_base_addr; extern const int *pcpu_unit_map; +struct pcpu_group_info { + int nr_units; /* aligned # of units */ + unsigned long base_offset; /* base address offset */ + unsigned int *cpu_map; /* unit->cpu map, empty + * entries contain NR_CPUS */ +}; + +struct pcpu_alloc_info { + size_t static_size; + size_t reserved_size; + size_t dyn_size; + size_t unit_size; + size_t atom_size; + size_t alloc_size; + size_t __ai_size; /* internal, don't use */ + int nr_groups; /* 0 if grouping unnecessary */ + struct pcpu_group_info groups[]; +}; + enum pcpu_fc { PCPU_FC_AUTO, PCPU_FC_EMBED, @@ -78,18 +97,17 @@ typedef void (*pcpu_fc_populate_pte_fn_t)(unsigned long addr); typedef int (pcpu_fc_cpu_distance_fn_t)(unsigned int from, unsigned int to); typedef void (*pcpu_fc_map_fn_t)(void *ptr, size_t size, void *addr); -#ifdef CONFIG_NEED_PER_CPU_LPAGE_FIRST_CHUNK -extern int __init pcpu_lpage_build_unit_map( - size_t reserved_size, ssize_t *dyn_sizep, - size_t *unit_sizep, size_t lpage_size, - int *unit_map, +extern struct pcpu_alloc_info * __init pcpu_alloc_alloc_info(int nr_groups, + int nr_units); +extern void __init pcpu_free_alloc_info(struct pcpu_alloc_info *ai); + +extern struct pcpu_alloc_info * __init pcpu_build_alloc_info( + size_t reserved_size, ssize_t dyn_size, + size_t atom_size, pcpu_fc_cpu_distance_fn_t cpu_distance_fn); -#endif -extern size_t __init pcpu_setup_first_chunk( - size_t static_size, size_t reserved_size, - size_t dyn_size, size_t unit_size, - void *base_addr, const int *unit_map); +extern size_t __init pcpu_setup_first_chunk(const struct pcpu_alloc_info *ai, + void *base_addr); #ifdef CONFIG_NEED_PER_CPU_EMBED_FIRST_CHUNK extern ssize_t __init pcpu_embed_first_chunk( @@ -106,9 +124,7 @@ extern ssize_t __init pcpu_page_first_chunk( #ifdef CONFIG_NEED_PER_CPU_LPAGE_FIRST_CHUNK extern ssize_t __init pcpu_lpage_first_chunk( - size_t reserved_size, size_t dyn_size, - size_t unit_size, size_t lpage_size, - const int *unit_map, int nr_units, + const struct pcpu_alloc_info *ai, pcpu_fc_alloc_fn_t alloc_fn, pcpu_fc_free_fn_t free_fn, pcpu_fc_map_fn_t map_fn); diff --git a/mm/percpu.c b/mm/percpu.c index 2b9c4b2a2fc0..99f7fa682722 100644 --- a/mm/percpu.c +++ b/mm/percpu.c @@ -58,6 +58,7 @@ #include #include +#include #include #include #include @@ -1245,53 +1246,108 @@ static inline size_t pcpu_calc_fc_sizes(size_t static_size, return size_sum; } -#ifdef CONFIG_NEED_PER_CPU_LPAGE_FIRST_CHUNK /** - * pcpu_lpage_build_unit_map - build unit_map for large page remapping + * pcpu_alloc_alloc_info - allocate percpu allocation info + * @nr_groups: the number of groups + * @nr_units: the number of units + * + * Allocate ai which is large enough for @nr_groups groups containing + * @nr_units units. The returned ai's groups[0].cpu_map points to the + * cpu_map array which is long enough for @nr_units and filled with + * NR_CPUS. It's the caller's responsibility to initialize cpu_map + * pointer of other groups. + * + * RETURNS: + * Pointer to the allocated pcpu_alloc_info on success, NULL on + * failure. + */ +struct pcpu_alloc_info * __init pcpu_alloc_alloc_info(int nr_groups, + int nr_units) +{ + struct pcpu_alloc_info *ai; + size_t base_size, ai_size; + void *ptr; + int unit; + + base_size = ALIGN(sizeof(*ai) + nr_groups * sizeof(ai->groups[0]), + __alignof__(ai->groups[0].cpu_map[0])); + ai_size = base_size + nr_units * sizeof(ai->groups[0].cpu_map[0]); + + ptr = alloc_bootmem_nopanic(PFN_ALIGN(ai_size)); + if (!ptr) + return NULL; + ai = ptr; + ptr += base_size; + + ai->groups[0].cpu_map = ptr; + + for (unit = 0; unit < nr_units; unit++) + ai->groups[0].cpu_map[unit] = NR_CPUS; + + ai->nr_groups = nr_groups; + ai->__ai_size = PFN_ALIGN(ai_size); + + return ai; +} + +/** + * pcpu_free_alloc_info - free percpu allocation info + * @ai: pcpu_alloc_info to free + * + * Free @ai which was allocated by pcpu_alloc_alloc_info(). + */ +void __init pcpu_free_alloc_info(struct pcpu_alloc_info *ai) +{ + free_bootmem(__pa(ai), ai->__ai_size); +} + +/** + * pcpu_build_alloc_info - build alloc_info considering distances between CPUs * @reserved_size: the size of reserved percpu area in bytes - * @dyn_sizep: in/out parameter for dynamic size, -1 for auto - * @unit_sizep: out parameter for unit size - * @unit_map: unit_map to be filled - * @cpu_distance_fn: callback to determine distance between cpus + * @dyn_size: free size for dynamic allocation in bytes, -1 for auto + * @atom_size: allocation atom size + * @cpu_distance_fn: callback to determine distance between cpus, optional * - * This function builds cpu -> unit map and determine other parameters - * considering needed percpu size, large page size and distances - * between CPUs in NUMA. + * This function determines grouping of units, their mappings to cpus + * and other parameters considering needed percpu size, allocation + * atom size and distances between CPUs. * - * CPUs which are of LOCAL_DISTANCE both ways are grouped together and - * may share units in the same large page. The returned configuration - * is guaranteed to have CPUs on different nodes on different large - * pages and >=75% usage of allocated virtual address space. + * Groups are always mutliples of atom size and CPUs which are of + * LOCAL_DISTANCE both ways are grouped together and share space for + * units in the same group. The returned configuration is guaranteed + * to have CPUs on different nodes on different groups and >=75% usage + * of allocated virtual address space. * * RETURNS: - * On success, fills in @unit_map, sets *@dyn_sizep, *@unit_sizep and - * returns the number of units to be allocated. -errno on failure. + * On success, pointer to the new allocation_info is returned. On + * failure, ERR_PTR value is returned. */ -int __init pcpu_lpage_build_unit_map(size_t reserved_size, ssize_t *dyn_sizep, - size_t *unit_sizep, size_t lpage_size, - int *unit_map, - pcpu_fc_cpu_distance_fn_t cpu_distance_fn) +struct pcpu_alloc_info * __init pcpu_build_alloc_info( + size_t reserved_size, ssize_t dyn_size, + size_t atom_size, + pcpu_fc_cpu_distance_fn_t cpu_distance_fn) { static int group_map[NR_CPUS] __initdata; static int group_cnt[NR_CPUS] __initdata; const size_t static_size = __per_cpu_end - __per_cpu_start; - int group_cnt_max = 0; + int group_cnt_max = 0, nr_groups = 1, nr_units = 0; size_t size_sum, min_unit_size, alloc_size; int upa, max_upa, uninitialized_var(best_upa); /* units_per_alloc */ - int last_allocs; + int last_allocs, group, unit; unsigned int cpu, tcpu; - int group, unit; + struct pcpu_alloc_info *ai; + unsigned int *cpu_map; /* * Determine min_unit_size, alloc_size and max_upa such that - * alloc_size is multiple of lpage_size and is the smallest + * alloc_size is multiple of atom_size and is the smallest * which can accomodate 4k aligned segments which are equal to * or larger than min_unit_size. */ - size_sum = pcpu_calc_fc_sizes(static_size, reserved_size, dyn_sizep); + size_sum = pcpu_calc_fc_sizes(static_size, reserved_size, &dyn_size); min_unit_size = max_t(size_t, size_sum, PCPU_MIN_UNIT_SIZE); - alloc_size = roundup(min_unit_size, lpage_size); + alloc_size = roundup(min_unit_size, atom_size); upa = alloc_size / min_unit_size; while (alloc_size % upa || ((alloc_size / upa) & ~PAGE_MASK)) upa--; @@ -1304,10 +1360,11 @@ int __init pcpu_lpage_build_unit_map(size_t reserved_size, ssize_t *dyn_sizep, for_each_possible_cpu(tcpu) { if (cpu == tcpu) break; - if (group_map[tcpu] == group && + if (group_map[tcpu] == group && cpu_distance_fn && (cpu_distance_fn(cpu, tcpu) > LOCAL_DISTANCE || cpu_distance_fn(tcpu, cpu) > LOCAL_DISTANCE)) { group++; + nr_groups = max(nr_groups, group + 1); goto next_group; } } @@ -1328,7 +1385,7 @@ int __init pcpu_lpage_build_unit_map(size_t reserved_size, ssize_t *dyn_sizep, if (alloc_size % upa || ((alloc_size / upa) & ~PAGE_MASK)) continue; - for (group = 0; group_cnt[group]; group++) { + for (group = 0; group < nr_groups; group++) { int this_allocs = DIV_ROUND_UP(group_cnt[group], upa); allocs += this_allocs; wasted += this_allocs * upa - group_cnt[group]; @@ -1348,75 +1405,122 @@ int __init pcpu_lpage_build_unit_map(size_t reserved_size, ssize_t *dyn_sizep, last_allocs = allocs; best_upa = upa; } - *unit_sizep = alloc_size / best_upa; + upa = best_upa; + + /* allocate and fill alloc_info */ + for (group = 0; group < nr_groups; group++) + nr_units += roundup(group_cnt[group], upa); + + ai = pcpu_alloc_alloc_info(nr_groups, nr_units); + if (!ai) + return ERR_PTR(-ENOMEM); + cpu_map = ai->groups[0].cpu_map; + + for (group = 0; group < nr_groups; group++) { + ai->groups[group].cpu_map = cpu_map; + cpu_map += roundup(group_cnt[group], upa); + } + + ai->static_size = static_size; + ai->reserved_size = reserved_size; + ai->dyn_size = dyn_size; + ai->unit_size = alloc_size / upa; + ai->atom_size = atom_size; + ai->alloc_size = alloc_size; + + for (group = 0, unit = 0; group_cnt[group]; group++) { + struct pcpu_group_info *gi = &ai->groups[group]; + + /* + * Initialize base_offset as if all groups are located + * back-to-back. The caller should update this to + * reflect actual allocation. + */ + gi->base_offset = unit * ai->unit_size; - /* assign units to cpus accordingly */ - unit = 0; - for (group = 0; group_cnt[group]; group++) { for_each_possible_cpu(cpu) if (group_map[cpu] == group) - unit_map[cpu] = unit++; - unit = roundup(unit, best_upa); + gi->cpu_map[gi->nr_units++] = cpu; + gi->nr_units = roundup(gi->nr_units, upa); + unit += gi->nr_units; } + BUG_ON(unit != nr_units); - return unit; /* unit contains aligned number of units */ + return ai; } -static bool __init pcpul_unit_to_cpu(int unit, const int *unit_map, - unsigned int *cpup); - -static void __init pcpul_lpage_dump_cfg(const char *lvl, size_t static_size, - size_t reserved_size, size_t dyn_size, - size_t unit_size, size_t lpage_size, - const int *unit_map, int nr_units) +/** + * pcpu_dump_alloc_info - print out information about pcpu_alloc_info + * @lvl: loglevel + * @ai: allocation info to dump + * + * Print out information about @ai using loglevel @lvl. + */ +static void pcpu_dump_alloc_info(const char *lvl, + const struct pcpu_alloc_info *ai) { - int width = 1, v = nr_units; + int group_width = 1, cpu_width = 1, width; char empty_str[] = "--------"; - int upl, lpl; /* units per lpage, lpage per line */ - unsigned int cpu; - int lpage, unit; + int alloc = 0, alloc_end = 0; + int group, v; + int upa, apl; /* units per alloc, allocs per line */ + + v = ai->nr_groups; + while (v /= 10) + group_width++; + v = num_possible_cpus(); while (v /= 10) - width++; - empty_str[min_t(int, width, sizeof(empty_str) - 1)] = '\0'; + cpu_width++; + empty_str[min_t(int, cpu_width, sizeof(empty_str) - 1)] = '\0'; - upl = max_t(int, lpage_size / unit_size, 1); - lpl = rounddown_pow_of_two(max_t(int, 60 / (upl * (width + 1) + 2), 1)); + upa = ai->alloc_size / ai->unit_size; + width = upa * (cpu_width + 1) + group_width + 3; + apl = rounddown_pow_of_two(max(60 / width, 1)); - printk("%spcpu-lpage: sta/res/dyn=%zu/%zu/%zu unit=%zu lpage=%zu", lvl, - static_size, reserved_size, dyn_size, unit_size, lpage_size); + printk("%spcpu-alloc: s%zu r%zu d%zu u%zu alloc=%zu*%zu", + lvl, ai->static_size, ai->reserved_size, ai->dyn_size, + ai->unit_size, ai->alloc_size / ai->atom_size, ai->atom_size); - for (lpage = 0, unit = 0; unit < nr_units; unit++) { - if (!(unit % upl)) { - if (!(lpage++ % lpl)) { + for (group = 0; group < ai->nr_groups; group++) { + const struct pcpu_group_info *gi = &ai->groups[group]; + int unit = 0, unit_end = 0; + + BUG_ON(gi->nr_units % upa); + for (alloc_end += gi->nr_units / upa; + alloc < alloc_end; alloc++) { + if (!(alloc % apl)) { printk("\n"); - printk("%spcpu-lpage: ", lvl); - } else - printk("| "); + printk("%spcpu-alloc: ", lvl); + } + printk("[%0*d] ", group_width, group); + + for (unit_end += upa; unit < unit_end; unit++) + if (gi->cpu_map[unit] != NR_CPUS) + printk("%0*d ", cpu_width, + gi->cpu_map[unit]); + else + printk("%s ", empty_str); } - if (pcpul_unit_to_cpu(unit, unit_map, &cpu)) - printk("%0*d ", width, cpu); - else - printk("%s ", empty_str); } printk("\n"); } -#endif /** * pcpu_setup_first_chunk - initialize the first percpu chunk - * @static_size: the size of static percpu area in bytes - * @reserved_size: the size of reserved percpu area in bytes, 0 for none - * @dyn_size: free size for dynamic allocation in bytes - * @unit_size: unit size in bytes, must be multiple of PAGE_SIZE + * @ai: pcpu_alloc_info describing how to percpu area is shaped * @base_addr: mapped address - * @unit_map: cpu -> unit map, NULL for sequential mapping * * Initialize the first percpu chunk which contains the kernel static * perpcu area. This function is to be called from arch percpu area * setup path. * - * @reserved_size, if non-zero, specifies the amount of bytes to + * @ai contains all information necessary to initialize the first + * chunk and prime the dynamic percpu allocator. + * + * @ai->static_size is the size of static percpu area. + * + * @ai->reserved_size, if non-zero, specifies the amount of bytes to * reserve after the static area in the first chunk. This reserves * the first chunk such that it's available only through reserved * percpu allocation. This is primarily used to serve module percpu @@ -1424,13 +1528,26 @@ static void __init pcpul_lpage_dump_cfg(const char *lvl, size_t static_size, * limited offset range for symbol relocations to guarantee module * percpu symbols fall inside the relocatable range. * - * @dyn_size determines the number of bytes available for dynamic - * allocation in the first chunk. The area between @static_size + - * @reserved_size + @dyn_size and @unit_size is unused. + * @ai->dyn_size determines the number of bytes available for dynamic + * allocation in the first chunk. The area between @ai->static_size + + * @ai->reserved_size + @ai->dyn_size and @ai->unit_size is unused. * - * @unit_size specifies unit size and must be aligned to PAGE_SIZE and - * equal to or larger than @static_size + @reserved_size + if - * non-negative, @dyn_size. + * @ai->unit_size specifies unit size and must be aligned to PAGE_SIZE + * and equal to or larger than @ai->static_size + @ai->reserved_size + + * @ai->dyn_size. + * + * @ai->atom_size is the allocation atom size and used as alignment + * for vm areas. + * + * @ai->alloc_size is the allocation size and always multiple of + * @ai->atom_size. This is larger than @ai->atom_size if + * @ai->unit_size is larger than @ai->atom_size. + * + * @ai->nr_groups and @ai->groups describe virtual memory layout of + * percpu areas. Units which should be colocated are put into the + * same group. Dynamic VM areas will be allocated according to these + * groupings. If @ai->nr_groups is zero, a single group containing + * all units is assumed. * * The caller should have mapped the first chunk at @base_addr and * copied static data to each unit. @@ -1446,70 +1563,63 @@ static void __init pcpul_lpage_dump_cfg(const char *lvl, size_t static_size, * The determined pcpu_unit_size which can be used to initialize * percpu access. */ -size_t __init pcpu_setup_first_chunk(size_t static_size, size_t reserved_size, - size_t dyn_size, size_t unit_size, - void *base_addr, const int *unit_map) +size_t __init pcpu_setup_first_chunk(const struct pcpu_alloc_info *ai, + void *base_addr) { static struct vm_struct first_vm; static int smap[2], dmap[2]; - size_t size_sum = static_size + reserved_size + dyn_size; + size_t dyn_size = ai->dyn_size; + size_t size_sum = ai->static_size + ai->reserved_size + dyn_size; struct pcpu_chunk *schunk, *dchunk = NULL; - unsigned int cpu, tcpu; - int i; + unsigned int cpu; + int *unit_map; + int group, unit, i; /* sanity checks */ BUILD_BUG_ON(ARRAY_SIZE(smap) >= PCPU_DFL_MAP_ALLOC || ARRAY_SIZE(dmap) >= PCPU_DFL_MAP_ALLOC); - BUG_ON(!static_size); + BUG_ON(ai->nr_groups <= 0); + BUG_ON(!ai->static_size); BUG_ON(!base_addr); - BUG_ON(unit_size < size_sum); - BUG_ON(unit_size & ~PAGE_MASK); - BUG_ON(unit_size < PCPU_MIN_UNIT_SIZE); + BUG_ON(ai->unit_size < size_sum); + BUG_ON(ai->unit_size & ~PAGE_MASK); + BUG_ON(ai->unit_size < PCPU_MIN_UNIT_SIZE); + + pcpu_dump_alloc_info(KERN_DEBUG, ai); /* determine number of units and verify and initialize pcpu_unit_map */ - if (unit_map) { - int first_unit = INT_MAX, last_unit = INT_MIN; - - for_each_possible_cpu(cpu) { - int unit = unit_map[cpu]; - - BUG_ON(unit < 0); - for_each_possible_cpu(tcpu) { - if (tcpu == cpu) - break; - /* the mapping should be one-to-one */ - BUG_ON(unit_map[tcpu] == unit); - } + unit_map = alloc_bootmem(nr_cpu_ids * sizeof(unit_map[0])); - if (unit < first_unit) { - pcpu_first_unit_cpu = cpu; - first_unit = unit; - } - if (unit > last_unit) { - pcpu_last_unit_cpu = cpu; - last_unit = unit; - } - } - pcpu_nr_units = last_unit + 1; - pcpu_unit_map = unit_map; - } else { - int *identity_map; + for (cpu = 0; cpu < nr_cpu_ids; cpu++) + unit_map[cpu] = NR_CPUS; + pcpu_first_unit_cpu = NR_CPUS; - /* #units == #cpus, identity mapped */ - identity_map = alloc_bootmem(nr_cpu_ids * - sizeof(identity_map[0])); + for (group = 0, unit = 0; group < ai->nr_groups; group++, unit += i) { + const struct pcpu_group_info *gi = &ai->groups[group]; - for_each_possible_cpu(cpu) - identity_map[cpu] = cpu; + for (i = 0; i < gi->nr_units; i++) { + cpu = gi->cpu_map[i]; + if (cpu == NR_CPUS) + continue; - pcpu_first_unit_cpu = 0; - pcpu_last_unit_cpu = pcpu_nr_units - 1; - pcpu_nr_units = nr_cpu_ids; - pcpu_unit_map = identity_map; + BUG_ON(cpu > nr_cpu_ids || !cpu_possible(cpu)); + BUG_ON(unit_map[cpu] != NR_CPUS); + + unit_map[cpu] = unit + i; + if (pcpu_first_unit_cpu == NR_CPUS) + pcpu_first_unit_cpu = cpu; + } } + pcpu_last_unit_cpu = cpu; + pcpu_nr_units = unit; + + for_each_possible_cpu(cpu) + BUG_ON(unit_map[cpu] == NR_CPUS); + + pcpu_unit_map = unit_map; /* determine basic parameters */ - pcpu_unit_pages = unit_size >> PAGE_SHIFT; + pcpu_unit_pages = ai->unit_size >> PAGE_SHIFT; pcpu_unit_size = pcpu_unit_pages << PAGE_SHIFT; pcpu_chunk_size = pcpu_nr_units * pcpu_unit_size; pcpu_chunk_struct_size = sizeof(struct pcpu_chunk) + @@ -1543,17 +1653,17 @@ size_t __init pcpu_setup_first_chunk(size_t static_size, size_t reserved_size, schunk->immutable = true; bitmap_fill(schunk->populated, pcpu_unit_pages); - if (reserved_size) { - schunk->free_size = reserved_size; + if (ai->reserved_size) { + schunk->free_size = ai->reserved_size; pcpu_reserved_chunk = schunk; - pcpu_reserved_chunk_limit = static_size + reserved_size; + pcpu_reserved_chunk_limit = ai->static_size + ai->reserved_size; } else { schunk->free_size = dyn_size; dyn_size = 0; /* dynamic area covered */ } schunk->contig_hint = schunk->free_size; - schunk->map[schunk->map_used++] = -static_size; + schunk->map[schunk->map_used++] = -ai->static_size; if (schunk->free_size) schunk->map[schunk->map_used++] = schunk->free_size; @@ -1643,44 +1753,47 @@ early_param("percpu_alloc", percpu_alloc_setup); */ ssize_t __init pcpu_embed_first_chunk(size_t reserved_size, ssize_t dyn_size) { - const size_t static_size = __per_cpu_end - __per_cpu_start; - size_t size_sum, unit_size, chunk_size; + struct pcpu_alloc_info *ai; + size_t size_sum, chunk_size; void *base; - unsigned int cpu; + int unit; + ssize_t ret; - /* determine parameters and allocate */ - size_sum = pcpu_calc_fc_sizes(static_size, reserved_size, &dyn_size); + ai = pcpu_build_alloc_info(reserved_size, dyn_size, PAGE_SIZE, NULL); + if (IS_ERR(ai)) + return PTR_ERR(ai); + BUG_ON(ai->nr_groups != 1); + BUG_ON(ai->groups[0].nr_units != num_possible_cpus()); - unit_size = max_t(size_t, size_sum, PCPU_MIN_UNIT_SIZE); - chunk_size = unit_size * nr_cpu_ids; + size_sum = ai->static_size + ai->reserved_size + ai->dyn_size; + chunk_size = ai->unit_size * num_possible_cpus(); base = __alloc_bootmem_nopanic(chunk_size, PAGE_SIZE, __pa(MAX_DMA_ADDRESS)); if (!base) { pr_warning("PERCPU: failed to allocate %zu bytes for " "embedding\n", chunk_size); - return -ENOMEM; + ret = -ENOMEM; + goto out_free_ai; } /* return the leftover and copy */ - for (cpu = 0; cpu < nr_cpu_ids; cpu++) { - void *ptr = base + cpu * unit_size; - - if (cpu_possible(cpu)) { - free_bootmem(__pa(ptr + size_sum), - unit_size - size_sum); - memcpy(ptr, __per_cpu_load, static_size); - } else - free_bootmem(__pa(ptr), unit_size); + for (unit = 0; unit < num_possible_cpus(); unit++) { + void *ptr = base + unit * ai->unit_size; + + free_bootmem(__pa(ptr + size_sum), ai->unit_size - size_sum); + memcpy(ptr, __per_cpu_load, ai->static_size); } /* we're ready, commit */ pr_info("PERCPU: Embedded %zu pages/cpu @%p s%zu r%zu d%zu u%zu\n", - PFN_DOWN(size_sum), base, static_size, reserved_size, dyn_size, - unit_size); + PFN_DOWN(size_sum), base, ai->static_size, ai->reserved_size, + ai->dyn_size, ai->unit_size); - return pcpu_setup_first_chunk(static_size, reserved_size, dyn_size, - unit_size, base, NULL); + ret = pcpu_setup_first_chunk(ai, base); +out_free_ai: + pcpu_free_alloc_info(ai); + return ret; } #endif /* CONFIG_NEED_PER_CPU_EMBED_FIRST_CHUNK || !CONFIG_HAVE_SETUP_PER_CPU_AREA */ @@ -1709,31 +1822,34 @@ ssize_t __init pcpu_page_first_chunk(size_t reserved_size, pcpu_fc_populate_pte_fn_t populate_pte_fn) { static struct vm_struct vm; - const size_t static_size = __per_cpu_end - __per_cpu_start; - ssize_t dyn_size = -1; - size_t size_sum, unit_size; + struct pcpu_alloc_info *ai; char psize_str[16]; int unit_pages; size_t pages_size; struct page **pages; - unsigned int cpu; - int i, j; + int unit, i, j; ssize_t ret; snprintf(psize_str, sizeof(psize_str), "%luK", PAGE_SIZE >> 10); - size_sum = pcpu_calc_fc_sizes(static_size, reserved_size, &dyn_size); - unit_size = max_t(size_t, size_sum, PCPU_MIN_UNIT_SIZE); - unit_pages = unit_size >> PAGE_SHIFT; + ai = pcpu_build_alloc_info(reserved_size, -1, PAGE_SIZE, NULL); + if (IS_ERR(ai)) + return PTR_ERR(ai); + BUG_ON(ai->nr_groups != 1); + BUG_ON(ai->groups[0].nr_units != num_possible_cpus()); + + unit_pages = ai->unit_size >> PAGE_SHIFT; /* unaligned allocations can't be freed, round up to page size */ - pages_size = PFN_ALIGN(unit_pages * nr_cpu_ids * sizeof(pages[0])); + pages_size = PFN_ALIGN(unit_pages * num_possible_cpus() * + sizeof(pages[0])); pages = alloc_bootmem(pages_size); /* allocate pages */ j = 0; - for_each_possible_cpu(cpu) + for (unit = 0; unit < num_possible_cpus(); unit++) for (i = 0; i < unit_pages; i++) { + unsigned int cpu = ai->groups[0].cpu_map[unit]; void *ptr; ptr = alloc_fn(cpu, PAGE_SIZE, PAGE_SIZE); @@ -1747,18 +1863,18 @@ ssize_t __init pcpu_page_first_chunk(size_t reserved_size, /* allocate vm area, map the pages and copy static data */ vm.flags = VM_ALLOC; - vm.size = nr_cpu_ids * unit_size; + vm.size = num_possible_cpus() * ai->unit_size; vm_area_register_early(&vm, PAGE_SIZE); - for_each_possible_cpu(cpu) { + for (unit = 0; unit < num_possible_cpus(); unit++) { unsigned long unit_addr = - (unsigned long)vm.addr + cpu * unit_size; + (unsigned long)vm.addr + unit * ai->unit_size; for (i = 0; i < unit_pages; i++) populate_pte_fn(unit_addr + (i << PAGE_SHIFT)); /* pte already populated, the following shouldn't fail */ - ret = __pcpu_map_pages(unit_addr, &pages[cpu * unit_pages], + ret = __pcpu_map_pages(unit_addr, &pages[unit * unit_pages], unit_pages); if (ret < 0) panic("failed to map percpu area, err=%zd\n", ret); @@ -1772,16 +1888,15 @@ ssize_t __init pcpu_page_first_chunk(size_t reserved_size, */ /* copy static data */ - memcpy((void *)unit_addr, __per_cpu_load, static_size); + memcpy((void *)unit_addr, __per_cpu_load, ai->static_size); } /* we're ready, commit */ pr_info("PERCPU: %d %s pages/cpu @%p s%zu r%zu d%zu\n", - unit_pages, psize_str, vm.addr, static_size, reserved_size, - dyn_size); + unit_pages, psize_str, vm.addr, ai->static_size, + ai->reserved_size, ai->dyn_size); - ret = pcpu_setup_first_chunk(static_size, reserved_size, dyn_size, - unit_size, vm.addr, NULL); + ret = pcpu_setup_first_chunk(ai, vm.addr); goto out_free_ar; enomem: @@ -1790,6 +1905,7 @@ enomem: ret = -ENOMEM; out_free_ar: free_bootmem(__pa(pages), pages_size); + pcpu_free_alloc_info(ai); return ret; } #endif /* CONFIG_NEED_PER_CPU_PAGE_FIRST_CHUNK */ @@ -1805,38 +1921,50 @@ static size_t pcpul_lpage_size; static int pcpul_nr_lpages; static struct pcpul_ent *pcpul_map; -static bool __init pcpul_unit_to_cpu(int unit, const int *unit_map, +static bool __init pcpul_unit_to_cpu(int unit, const struct pcpu_alloc_info *ai, unsigned int *cpup) { - unsigned int cpu; + int group, cunit; - for_each_possible_cpu(cpu) - if (unit_map[cpu] == unit) { + for (group = 0, cunit = 0; group < ai->nr_groups; group++) { + const struct pcpu_group_info *gi = &ai->groups[group]; + + if (unit < cunit + gi->nr_units) { if (cpup) - *cpup = cpu; + *cpup = gi->cpu_map[unit - cunit]; return true; } + cunit += gi->nr_units; + } return false; } +static int __init pcpul_cpu_to_unit(int cpu, const struct pcpu_alloc_info *ai) +{ + int group, unit, i; + + for (group = 0, unit = 0; group < ai->nr_groups; group++, unit += i) { + const struct pcpu_group_info *gi = &ai->groups[group]; + + for (i = 0; i < gi->nr_units; i++) + if (gi->cpu_map[i] == cpu) + return unit + i; + } + BUG(); +} + /** * pcpu_lpage_first_chunk - remap the first percpu chunk using large page - * @reserved_size: the size of reserved percpu area in bytes - * @dyn_size: free size for dynamic allocation in bytes - * @unit_size: unit size in bytes - * @lpage_size: the size of a large page - * @unit_map: cpu -> unit mapping - * @nr_units: the number of units + * @ai: pcpu_alloc_info * @alloc_fn: function to allocate percpu lpage, always called with lpage_size * @free_fn: function to free percpu memory, @size <= lpage_size * @map_fn: function to map percpu lpage, always called with lpage_size * * This allocator uses large page to build and map the first chunk. - * Unlike other helpers, the caller should always specify @dyn_size - * and @unit_size. These parameters along with @unit_map and - * @nr_units can be determined using pcpu_lpage_build_unit_map(). - * This two stage initialization is to allow arch code to evaluate the + * Unlike other helpers, the caller should provide fully initialized + * @ai. This can be done using pcpu_build_alloc_info(). This two + * stage initialization is to allow arch code to evaluate the * parameters before committing to it. * * Large pages are allocated as directed by @unit_map and other @@ -1852,27 +1980,26 @@ static bool __init pcpul_unit_to_cpu(int unit, const int *unit_map, * The determined pcpu_unit_size which can be used to initialize * percpu access on success, -errno on failure. */ -ssize_t __init pcpu_lpage_first_chunk(size_t reserved_size, size_t dyn_size, - size_t unit_size, size_t lpage_size, - const int *unit_map, int nr_units, +ssize_t __init pcpu_lpage_first_chunk(const struct pcpu_alloc_info *ai, pcpu_fc_alloc_fn_t alloc_fn, pcpu_fc_free_fn_t free_fn, pcpu_fc_map_fn_t map_fn) { static struct vm_struct vm; - const size_t static_size = __per_cpu_end - __per_cpu_start; - size_t chunk_size = unit_size * nr_units; - size_t map_size; + const size_t lpage_size = ai->atom_size; + size_t chunk_size, map_size; unsigned int cpu; ssize_t ret; - int i, j, unit; + int i, j, unit, nr_units; - pcpul_lpage_dump_cfg(KERN_DEBUG, static_size, reserved_size, dyn_size, - unit_size, lpage_size, unit_map, nr_units); + nr_units = 0; + for (i = 0; i < ai->nr_groups; i++) + nr_units += ai->groups[i].nr_units; + chunk_size = ai->unit_size * nr_units; BUG_ON(chunk_size % lpage_size); - pcpul_size = static_size + reserved_size + dyn_size; + pcpul_size = ai->static_size + ai->reserved_size + ai->dyn_size; pcpul_lpage_size = lpage_size; pcpul_nr_lpages = chunk_size / lpage_size; @@ -1883,13 +2010,13 @@ ssize_t __init pcpu_lpage_first_chunk(size_t reserved_size, size_t dyn_size, /* allocate all pages */ for (i = 0; i < pcpul_nr_lpages; i++) { size_t offset = i * lpage_size; - int first_unit = offset / unit_size; - int last_unit = (offset + lpage_size - 1) / unit_size; + int first_unit = offset / ai->unit_size; + int last_unit = (offset + lpage_size - 1) / ai->unit_size; void *ptr; /* find out which cpu is mapped to this unit */ for (unit = first_unit; unit <= last_unit; unit++) - if (pcpul_unit_to_cpu(unit, unit_map, &cpu)) + if (pcpul_unit_to_cpu(unit, ai, &cpu)) goto found; continue; found: @@ -1905,12 +2032,12 @@ ssize_t __init pcpu_lpage_first_chunk(size_t reserved_size, size_t dyn_size, /* return unused holes */ for (unit = 0; unit < nr_units; unit++) { - size_t start = unit * unit_size; - size_t end = start + unit_size; + size_t start = unit * ai->unit_size; + size_t end = start + ai->unit_size; size_t off, next; /* don't free used part of occupied unit */ - if (pcpul_unit_to_cpu(unit, unit_map, NULL)) + if (pcpul_unit_to_cpu(unit, ai, NULL)) start += pcpul_size; /* unit can span more than one page, punch the holes */ @@ -1925,7 +2052,7 @@ ssize_t __init pcpu_lpage_first_chunk(size_t reserved_size, size_t dyn_size, /* allocate address, map and copy */ vm.flags = VM_ALLOC; vm.size = chunk_size; - vm_area_register_early(&vm, unit_size); + vm_area_register_early(&vm, ai->unit_size); for (i = 0; i < pcpul_nr_lpages; i++) { if (!pcpul_map[i].ptr) @@ -1935,15 +2062,15 @@ ssize_t __init pcpu_lpage_first_chunk(size_t reserved_size, size_t dyn_size, } for_each_possible_cpu(cpu) - memcpy(vm.addr + unit_map[cpu] * unit_size, __per_cpu_load, - static_size); + memcpy(vm.addr + pcpul_cpu_to_unit(cpu, ai) * ai->unit_size, + __per_cpu_load, ai->static_size); /* we're ready, commit */ pr_info("PERCPU: large pages @%p s%zu r%zu d%zu u%zu\n", - vm.addr, static_size, reserved_size, dyn_size, unit_size); + vm.addr, ai->static_size, ai->reserved_size, ai->dyn_size, + ai->unit_size); - ret = pcpu_setup_first_chunk(static_size, reserved_size, dyn_size, - unit_size, vm.addr, unit_map); + ret = pcpu_setup_first_chunk(ai, vm.addr); /* * Sort pcpul_map array for pcpu_lpage_remapped(). Unmapped -- cgit v1.2.3 From fb435d5233f8b6f9b93c11d6304d8e98fed03234 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Fri, 14 Aug 2009 15:00:51 +0900 Subject: percpu: add pcpu_unit_offsets[] Currently units are mapped sequentially into address space. This patch adds pcpu_unit_offsets[] which allows units to be mapped to arbitrary offsets from the chunk base address. This is necessary to allow sparse embedding which might would need to allocate address ranges and memory areas which aren't aligned to unit size but allocation atom size (page or large page size). This also simplifies things a bit by removing the need to calculate offset from unit number. With this change, there's no need for the arch code to know pcpu_unit_size. Update pcpu_setup_first_chunk() and first chunk allocators to return regular 0 or -errno return code instead of unit size or -errno. Signed-off-by: Tejun Heo Cc: David S. Miller --- arch/sparc/kernel/smp_64.c | 12 +++--- arch/x86/kernel/setup_percpu.c | 51 ++++++++++------------- include/linux/percpu.h | 16 ++++--- mm/percpu.c | 95 +++++++++++++++++++++--------------------- 4 files changed, 84 insertions(+), 90 deletions(-) (limited to 'include') diff --git a/arch/sparc/kernel/smp_64.c b/arch/sparc/kernel/smp_64.c index a42a4a744d14..b03fd362c629 100644 --- a/arch/sparc/kernel/smp_64.c +++ b/arch/sparc/kernel/smp_64.c @@ -1478,9 +1478,10 @@ void __init setup_per_cpu_areas(void) static struct vm_struct vm; struct pcpu_alloc_info *ai; unsigned long delta, cpu; - size_t size_sum, pcpu_unit_size; + size_t size_sum; size_t ptrs_size; void **ptrs; + int rc; ai = pcpu_alloc_alloc_info(1, nr_cpu_ids); @@ -1526,14 +1527,15 @@ void __init setup_per_cpu_areas(void) pcpu_map_range(start, end, virt_to_page(ptrs[cpu])); } - pcpu_unit_size = pcpu_setup_first_chunk(ai, vm.addr); + rc = pcpu_setup_first_chunk(ai, vm.addr); + if (rc) + panic("failed to setup percpu first chunk (%d)", rc); free_bootmem(__pa(ptrs), ptrs_size); delta = (unsigned long)pcpu_base_addr - (unsigned long)__per_cpu_start; - for_each_possible_cpu(cpu) { - __per_cpu_offset(cpu) = delta + cpu * pcpu_unit_size; - } + for_each_possible_cpu(cpu) + __per_cpu_offset(cpu) = delta + pcpu_unit_offsets[cpu]; /* Setup %g5 for the boot cpu. */ __local_per_cpu_offset = __per_cpu_offset(smp_processor_id()); diff --git a/arch/x86/kernel/setup_percpu.c b/arch/x86/kernel/setup_percpu.c index db5f9c49fec5..9becc5d4b518 100644 --- a/arch/x86/kernel/setup_percpu.c +++ b/arch/x86/kernel/setup_percpu.c @@ -157,12 +157,12 @@ static int pcpu_lpage_cpu_distance(unsigned int from, unsigned int to) return REMOTE_DISTANCE; } -static ssize_t __init setup_pcpu_lpage(bool chosen) +static int __init setup_pcpu_lpage(bool chosen) { size_t reserve = PERCPU_MODULE_RESERVE + PERCPU_DYNAMIC_RESERVE; size_t dyn_size = reserve - PERCPU_FIRST_CHUNK_RESERVE; struct pcpu_alloc_info *ai; - ssize_t ret; + int rc; /* on non-NUMA, embedding is better */ if (!chosen && !pcpu_need_numa()) @@ -196,19 +196,18 @@ static ssize_t __init setup_pcpu_lpage(bool chosen) if (tot_size > vm_size / 5) { pr_info("PERCPU: too large chunk size %zuMB for " "large page remap\n", tot_size >> 20); - ret = -EINVAL; + rc = -EINVAL; goto out_free; } } - ret = pcpu_lpage_first_chunk(ai, pcpu_fc_alloc, pcpu_fc_free, - pcpul_map); + rc = pcpu_lpage_first_chunk(ai, pcpu_fc_alloc, pcpu_fc_free, pcpul_map); out_free: pcpu_free_alloc_info(ai); - return ret; + return rc; } #else -static ssize_t __init setup_pcpu_lpage(bool chosen) +static int __init setup_pcpu_lpage(bool chosen) { return -EINVAL; } @@ -222,7 +221,7 @@ static ssize_t __init setup_pcpu_lpage(bool chosen) * mapping so that it can use PMD mapping without additional TLB * pressure. */ -static ssize_t __init setup_pcpu_embed(bool chosen) +static int __init setup_pcpu_embed(bool chosen) { size_t reserve = PERCPU_MODULE_RESERVE + PERCPU_DYNAMIC_RESERVE; @@ -250,7 +249,7 @@ static void __init pcpup_populate_pte(unsigned long addr) populate_extra_pte(addr); } -static ssize_t __init setup_pcpu_page(void) +static int __init setup_pcpu_page(void) { return pcpu_page_first_chunk(PERCPU_FIRST_CHUNK_RESERVE, pcpu_fc_alloc, pcpu_fc_free, @@ -274,8 +273,7 @@ void __init setup_per_cpu_areas(void) { unsigned int cpu; unsigned long delta; - size_t pcpu_unit_size; - ssize_t ret; + int rc; pr_info("NR_CPUS:%d nr_cpumask_bits:%d nr_cpu_ids:%d nr_node_ids:%d\n", NR_CPUS, nr_cpumask_bits, nr_cpu_ids, nr_node_ids); @@ -285,36 +283,33 @@ void __init setup_per_cpu_areas(void) * of large page mappings. Please read comments on top of * each allocator for details. */ - ret = -EINVAL; + rc = -EINVAL; if (pcpu_chosen_fc != PCPU_FC_AUTO) { if (pcpu_chosen_fc != PCPU_FC_PAGE) { if (pcpu_chosen_fc == PCPU_FC_LPAGE) - ret = setup_pcpu_lpage(true); + rc = setup_pcpu_lpage(true); else - ret = setup_pcpu_embed(true); + rc = setup_pcpu_embed(true); - if (ret < 0) - pr_warning("PERCPU: %s allocator failed (%zd), " + if (rc < 0) + pr_warning("PERCPU: %s allocator failed (%d), " "falling back to page size\n", - pcpu_fc_names[pcpu_chosen_fc], ret); + pcpu_fc_names[pcpu_chosen_fc], rc); } } else { - ret = setup_pcpu_lpage(false); - if (ret < 0) - ret = setup_pcpu_embed(false); + rc = setup_pcpu_lpage(false); + if (rc < 0) + rc = setup_pcpu_embed(false); } - if (ret < 0) - ret = setup_pcpu_page(); - if (ret < 0) - panic("cannot initialize percpu area (err=%zd)", ret); - - pcpu_unit_size = ret; + if (rc < 0) + rc = setup_pcpu_page(); + if (rc < 0) + panic("cannot initialize percpu area (err=%d)", rc); /* alrighty, percpu areas up and running */ delta = (unsigned long)pcpu_base_addr - (unsigned long)__per_cpu_start; for_each_possible_cpu(cpu) { - per_cpu_offset(cpu) = - delta + pcpu_unit_map[cpu] * pcpu_unit_size; + per_cpu_offset(cpu) = delta + pcpu_unit_offsets[cpu]; per_cpu(this_cpu_off, cpu) = per_cpu_offset(cpu); per_cpu(cpu_number, cpu) = cpu; setup_percpu_segment(cpu); diff --git a/include/linux/percpu.h b/include/linux/percpu.h index 77b86be8ce4f..a7ec840f596c 100644 --- a/include/linux/percpu.h +++ b/include/linux/percpu.h @@ -57,7 +57,7 @@ #endif extern void *pcpu_base_addr; -extern const int *pcpu_unit_map; +extern const unsigned long *pcpu_unit_offsets; struct pcpu_group_info { int nr_units; /* aligned # of units */ @@ -106,25 +106,23 @@ extern struct pcpu_alloc_info * __init pcpu_build_alloc_info( size_t atom_size, pcpu_fc_cpu_distance_fn_t cpu_distance_fn); -extern size_t __init pcpu_setup_first_chunk(const struct pcpu_alloc_info *ai, - void *base_addr); +extern int __init pcpu_setup_first_chunk(const struct pcpu_alloc_info *ai, + void *base_addr); #ifdef CONFIG_NEED_PER_CPU_EMBED_FIRST_CHUNK -extern ssize_t __init pcpu_embed_first_chunk( - size_t reserved_size, ssize_t dyn_size); +extern int __init pcpu_embed_first_chunk(size_t reserved_size, + ssize_t dyn_size); #endif #ifdef CONFIG_NEED_PER_CPU_PAGE_FIRST_CHUNK -extern ssize_t __init pcpu_page_first_chunk( - size_t reserved_size, +extern int __init pcpu_page_first_chunk(size_t reserved_size, pcpu_fc_alloc_fn_t alloc_fn, pcpu_fc_free_fn_t free_fn, pcpu_fc_populate_pte_fn_t populate_pte_fn); #endif #ifdef CONFIG_NEED_PER_CPU_LPAGE_FIRST_CHUNK -extern ssize_t __init pcpu_lpage_first_chunk( - const struct pcpu_alloc_info *ai, +extern int __init pcpu_lpage_first_chunk(const struct pcpu_alloc_info *ai, pcpu_fc_alloc_fn_t alloc_fn, pcpu_fc_free_fn_t free_fn, pcpu_fc_map_fn_t map_fn); diff --git a/mm/percpu.c b/mm/percpu.c index 99f7fa682722..653b02c40200 100644 --- a/mm/percpu.c +++ b/mm/percpu.c @@ -117,8 +117,8 @@ static unsigned int pcpu_last_unit_cpu __read_mostly; void *pcpu_base_addr __read_mostly; EXPORT_SYMBOL_GPL(pcpu_base_addr); -/* cpu -> unit map */ -const int *pcpu_unit_map __read_mostly; +static const int *pcpu_unit_map __read_mostly; /* cpu -> unit */ +const unsigned long *pcpu_unit_offsets __read_mostly; /* cpu -> unit offset */ /* * The first chunk which always exists. Note that unlike other @@ -196,8 +196,8 @@ static int pcpu_page_idx(unsigned int cpu, int page_idx) static unsigned long pcpu_chunk_addr(struct pcpu_chunk *chunk, unsigned int cpu, int page_idx) { - return (unsigned long)chunk->vm->addr + - (pcpu_page_idx(cpu, page_idx) << PAGE_SHIFT); + return (unsigned long)chunk->vm->addr + pcpu_unit_offsets[cpu] + + (page_idx << PAGE_SHIFT); } static struct page *pcpu_chunk_page(struct pcpu_chunk *chunk, @@ -341,7 +341,7 @@ static struct pcpu_chunk *pcpu_chunk_addr_search(void *addr) * space. Note that any possible cpu id can be used here, so * there's no need to worry about preemption or cpu hotplug. */ - addr += pcpu_unit_map[smp_processor_id()] * pcpu_unit_size; + addr += pcpu_unit_offsets[smp_processor_id()]; return pcpu_get_page_chunk(vmalloc_to_page(addr)); } @@ -1560,17 +1560,17 @@ static void pcpu_dump_alloc_info(const char *lvl, * and available for dynamic allocation like any other chunks. * * RETURNS: - * The determined pcpu_unit_size which can be used to initialize - * percpu access. + * 0 on success, -errno on failure. */ -size_t __init pcpu_setup_first_chunk(const struct pcpu_alloc_info *ai, - void *base_addr) +int __init pcpu_setup_first_chunk(const struct pcpu_alloc_info *ai, + void *base_addr) { static struct vm_struct first_vm; static int smap[2], dmap[2]; size_t dyn_size = ai->dyn_size; size_t size_sum = ai->static_size + ai->reserved_size + dyn_size; struct pcpu_chunk *schunk, *dchunk = NULL; + unsigned long *unit_off; unsigned int cpu; int *unit_map; int group, unit, i; @@ -1587,8 +1587,9 @@ size_t __init pcpu_setup_first_chunk(const struct pcpu_alloc_info *ai, pcpu_dump_alloc_info(KERN_DEBUG, ai); - /* determine number of units and verify and initialize pcpu_unit_map */ + /* determine number of units and initialize unit_map and base */ unit_map = alloc_bootmem(nr_cpu_ids * sizeof(unit_map[0])); + unit_off = alloc_bootmem(nr_cpu_ids * sizeof(unit_off[0])); for (cpu = 0; cpu < nr_cpu_ids; cpu++) unit_map[cpu] = NR_CPUS; @@ -1606,6 +1607,8 @@ size_t __init pcpu_setup_first_chunk(const struct pcpu_alloc_info *ai, BUG_ON(unit_map[cpu] != NR_CPUS); unit_map[cpu] = unit + i; + unit_off[cpu] = gi->base_offset + i * ai->unit_size; + if (pcpu_first_unit_cpu == NR_CPUS) pcpu_first_unit_cpu = cpu; } @@ -1617,6 +1620,7 @@ size_t __init pcpu_setup_first_chunk(const struct pcpu_alloc_info *ai, BUG_ON(unit_map[cpu] == NR_CPUS); pcpu_unit_map = unit_map; + pcpu_unit_offsets = unit_off; /* determine basic parameters */ pcpu_unit_pages = ai->unit_size >> PAGE_SHIFT; @@ -1688,7 +1692,7 @@ size_t __init pcpu_setup_first_chunk(const struct pcpu_alloc_info *ai, /* we're done */ pcpu_base_addr = schunk->vm->addr; - return pcpu_unit_size; + return 0; } const char *pcpu_fc_names[PCPU_FC_NR] __initdata = { @@ -1748,16 +1752,15 @@ early_param("percpu_alloc", percpu_alloc_setup); * size, the leftover is returned to the bootmem allocator. * * RETURNS: - * The determined pcpu_unit_size which can be used to initialize - * percpu access on success, -errno on failure. + * 0 on success, -errno on failure. */ -ssize_t __init pcpu_embed_first_chunk(size_t reserved_size, ssize_t dyn_size) +int __init pcpu_embed_first_chunk(size_t reserved_size, ssize_t dyn_size) { struct pcpu_alloc_info *ai; size_t size_sum, chunk_size; void *base; int unit; - ssize_t ret; + int rc; ai = pcpu_build_alloc_info(reserved_size, dyn_size, PAGE_SIZE, NULL); if (IS_ERR(ai)) @@ -1773,7 +1776,7 @@ ssize_t __init pcpu_embed_first_chunk(size_t reserved_size, ssize_t dyn_size) if (!base) { pr_warning("PERCPU: failed to allocate %zu bytes for " "embedding\n", chunk_size); - ret = -ENOMEM; + rc = -ENOMEM; goto out_free_ai; } @@ -1790,10 +1793,10 @@ ssize_t __init pcpu_embed_first_chunk(size_t reserved_size, ssize_t dyn_size) PFN_DOWN(size_sum), base, ai->static_size, ai->reserved_size, ai->dyn_size, ai->unit_size); - ret = pcpu_setup_first_chunk(ai, base); + rc = pcpu_setup_first_chunk(ai, base); out_free_ai: pcpu_free_alloc_info(ai); - return ret; + return rc; } #endif /* CONFIG_NEED_PER_CPU_EMBED_FIRST_CHUNK || !CONFIG_HAVE_SETUP_PER_CPU_AREA */ @@ -1813,13 +1816,12 @@ out_free_ai: * page-by-page into vmalloc area. * * RETURNS: - * The determined pcpu_unit_size which can be used to initialize - * percpu access on success, -errno on failure. + * 0 on success, -errno on failure. */ -ssize_t __init pcpu_page_first_chunk(size_t reserved_size, - pcpu_fc_alloc_fn_t alloc_fn, - pcpu_fc_free_fn_t free_fn, - pcpu_fc_populate_pte_fn_t populate_pte_fn) +int __init pcpu_page_first_chunk(size_t reserved_size, + pcpu_fc_alloc_fn_t alloc_fn, + pcpu_fc_free_fn_t free_fn, + pcpu_fc_populate_pte_fn_t populate_pte_fn) { static struct vm_struct vm; struct pcpu_alloc_info *ai; @@ -1827,8 +1829,7 @@ ssize_t __init pcpu_page_first_chunk(size_t reserved_size, int unit_pages; size_t pages_size; struct page **pages; - int unit, i, j; - ssize_t ret; + int unit, i, j, rc; snprintf(psize_str, sizeof(psize_str), "%luK", PAGE_SIZE >> 10); @@ -1874,10 +1875,10 @@ ssize_t __init pcpu_page_first_chunk(size_t reserved_size, populate_pte_fn(unit_addr + (i << PAGE_SHIFT)); /* pte already populated, the following shouldn't fail */ - ret = __pcpu_map_pages(unit_addr, &pages[unit * unit_pages], - unit_pages); - if (ret < 0) - panic("failed to map percpu area, err=%zd\n", ret); + rc = __pcpu_map_pages(unit_addr, &pages[unit * unit_pages], + unit_pages); + if (rc < 0) + panic("failed to map percpu area, err=%d\n", rc); /* * FIXME: Archs with virtual cache should flush local @@ -1896,17 +1897,17 @@ ssize_t __init pcpu_page_first_chunk(size_t reserved_size, unit_pages, psize_str, vm.addr, ai->static_size, ai->reserved_size, ai->dyn_size); - ret = pcpu_setup_first_chunk(ai, vm.addr); + rc = pcpu_setup_first_chunk(ai, vm.addr); goto out_free_ar; enomem: while (--j >= 0) free_fn(page_address(pages[j]), PAGE_SIZE); - ret = -ENOMEM; + rc = -ENOMEM; out_free_ar: free_bootmem(__pa(pages), pages_size); pcpu_free_alloc_info(ai); - return ret; + return rc; } #endif /* CONFIG_NEED_PER_CPU_PAGE_FIRST_CHUNK */ @@ -1977,20 +1978,18 @@ static int __init pcpul_cpu_to_unit(int cpu, const struct pcpu_alloc_info *ai) * pcpu_lpage_remapped(). * * RETURNS: - * The determined pcpu_unit_size which can be used to initialize - * percpu access on success, -errno on failure. + * 0 on success, -errno on failure. */ -ssize_t __init pcpu_lpage_first_chunk(const struct pcpu_alloc_info *ai, - pcpu_fc_alloc_fn_t alloc_fn, - pcpu_fc_free_fn_t free_fn, - pcpu_fc_map_fn_t map_fn) +int __init pcpu_lpage_first_chunk(const struct pcpu_alloc_info *ai, + pcpu_fc_alloc_fn_t alloc_fn, + pcpu_fc_free_fn_t free_fn, + pcpu_fc_map_fn_t map_fn) { static struct vm_struct vm; const size_t lpage_size = ai->atom_size; size_t chunk_size, map_size; unsigned int cpu; - ssize_t ret; - int i, j, unit, nr_units; + int i, j, unit, nr_units, rc; nr_units = 0; for (i = 0; i < ai->nr_groups; i++) @@ -2070,7 +2069,7 @@ ssize_t __init pcpu_lpage_first_chunk(const struct pcpu_alloc_info *ai, vm.addr, ai->static_size, ai->reserved_size, ai->dyn_size, ai->unit_size); - ret = pcpu_setup_first_chunk(ai, vm.addr); + rc = pcpu_setup_first_chunk(ai, vm.addr); /* * Sort pcpul_map array for pcpu_lpage_remapped(). Unmapped @@ -2094,7 +2093,7 @@ ssize_t __init pcpu_lpage_first_chunk(const struct pcpu_alloc_info *ai, while (pcpul_nr_lpages && !pcpul_map[pcpul_nr_lpages - 1].ptr) pcpul_nr_lpages--; - return ret; + return rc; enomem: for (i = 0; i < pcpul_nr_lpages; i++) @@ -2166,21 +2165,21 @@ EXPORT_SYMBOL(__per_cpu_offset); void __init setup_per_cpu_areas(void) { - ssize_t unit_size; unsigned long delta; unsigned int cpu; + int rc; /* * Always reserve area for module percpu variables. That's * what the legacy allocator did. */ - unit_size = pcpu_embed_first_chunk(PERCPU_MODULE_RESERVE, - PERCPU_DYNAMIC_RESERVE); - if (unit_size < 0) + rc = pcpu_embed_first_chunk(PERCPU_MODULE_RESERVE, + PERCPU_DYNAMIC_RESERVE); + if (rc < 0) panic("Failed to initialized percpu areas."); delta = (unsigned long)pcpu_base_addr - (unsigned long)__per_cpu_start; for_each_possible_cpu(cpu) - __per_cpu_offset[cpu] = delta + cpu * unit_size; + __per_cpu_offset[cpu] = delta + pcpu_unit_offsets[cpu]; } #endif /* CONFIG_HAVE_SETUP_PER_CPU_AREA */ -- cgit v1.2.3 From ca23e405e06d5fffb005df004c72781f76062f51 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Fri, 14 Aug 2009 15:00:52 +0900 Subject: vmalloc: implement pcpu_get_vm_areas() To directly use spread NUMA memories for percpu units, percpu allocator will be updated to allow sparsely mapping units in a chunk. As the distances between units can be very large, this makes allocating single vmap area for each chunk undesirable. This patch implements pcpu_get_vm_areas() and pcpu_free_vm_areas() which allocates and frees sparse congruent vmap areas. pcpu_get_vm_areas() take @offsets and @sizes array which define distances and sizes of vmap areas. It scans down from the top of vmalloc area looking for the top-most address which can accomodate all the areas. The top-down scan is to avoid interacting with regular vmallocs which can push up these congruent areas up little by little ending up wasting address space and page table. To speed up top-down scan, the highest possible address hint is maintained. Although the scan is linear from the hint, given the usual large holes between memory addresses between NUMA nodes, the scanning is highly likely to finish after finding the first hole for the last unit which is scanned first. Signed-off-by: Tejun Heo Cc: Nick Piggin --- include/linux/vmalloc.h | 6 + mm/vmalloc.c | 293 ++++++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 299 insertions(+) (limited to 'include') diff --git a/include/linux/vmalloc.h b/include/linux/vmalloc.h index a43ebec3a7b9..227c2a585e4f 100644 --- a/include/linux/vmalloc.h +++ b/include/linux/vmalloc.h @@ -115,4 +115,10 @@ extern rwlock_t vmlist_lock; extern struct vm_struct *vmlist; extern __init void vm_area_register_early(struct vm_struct *vm, size_t align); +struct vm_struct **pcpu_get_vm_areas(const unsigned long *offsets, + const size_t *sizes, int nr_vms, + size_t align, gfp_t gfp_mask); + +void pcpu_free_vm_areas(struct vm_struct **vms, int nr_vms); + #endif /* _LINUX_VMALLOC_H */ diff --git a/mm/vmalloc.c b/mm/vmalloc.c index 2eb461c3a46e..204b8243d8ab 100644 --- a/mm/vmalloc.c +++ b/mm/vmalloc.c @@ -265,6 +265,7 @@ struct vmap_area { static DEFINE_SPINLOCK(vmap_area_lock); static struct rb_root vmap_area_root = RB_ROOT; static LIST_HEAD(vmap_area_list); +static unsigned long vmap_area_pcpu_hole; static struct vmap_area *__find_vmap_area(unsigned long addr) { @@ -431,6 +432,15 @@ static void __free_vmap_area(struct vmap_area *va) RB_CLEAR_NODE(&va->rb_node); list_del_rcu(&va->list); + /* + * Track the highest possible candidate for pcpu area + * allocation. Areas outside of vmalloc area can be returned + * here too, consider only end addresses which fall inside + * vmalloc area proper. + */ + if (va->va_end > VMALLOC_START && va->va_end <= VMALLOC_END) + vmap_area_pcpu_hole = max(vmap_area_pcpu_hole, va->va_end); + call_rcu(&va->rcu_head, rcu_free_va); } @@ -1038,6 +1048,9 @@ void __init vmalloc_init(void) va->va_end = va->va_start + tmp->size; __insert_vmap_area(va); } + + vmap_area_pcpu_hole = VMALLOC_END; + vmap_initialized = true; } @@ -1821,6 +1834,286 @@ void free_vm_area(struct vm_struct *area) } EXPORT_SYMBOL_GPL(free_vm_area); +static struct vmap_area *node_to_va(struct rb_node *n) +{ + return n ? rb_entry(n, struct vmap_area, rb_node) : NULL; +} + +/** + * pvm_find_next_prev - find the next and prev vmap_area surrounding @end + * @end: target address + * @pnext: out arg for the next vmap_area + * @pprev: out arg for the previous vmap_area + * + * Returns: %true if either or both of next and prev are found, + * %false if no vmap_area exists + * + * Find vmap_areas end addresses of which enclose @end. ie. if not + * NULL, *pnext->va_end > @end and *pprev->va_end <= @end. + */ +static bool pvm_find_next_prev(unsigned long end, + struct vmap_area **pnext, + struct vmap_area **pprev) +{ + struct rb_node *n = vmap_area_root.rb_node; + struct vmap_area *va = NULL; + + while (n) { + va = rb_entry(n, struct vmap_area, rb_node); + if (end < va->va_end) + n = n->rb_left; + else if (end > va->va_end) + n = n->rb_right; + else + break; + } + + if (!va) + return false; + + if (va->va_end > end) { + *pnext = va; + *pprev = node_to_va(rb_prev(&(*pnext)->rb_node)); + } else { + *pprev = va; + *pnext = node_to_va(rb_next(&(*pprev)->rb_node)); + } + return true; +} + +/** + * pvm_determine_end - find the highest aligned address between two vmap_areas + * @pnext: in/out arg for the next vmap_area + * @pprev: in/out arg for the previous vmap_area + * @align: alignment + * + * Returns: determined end address + * + * Find the highest aligned address between *@pnext and *@pprev below + * VMALLOC_END. *@pnext and *@pprev are adjusted so that the aligned + * down address is between the end addresses of the two vmap_areas. + * + * Please note that the address returned by this function may fall + * inside *@pnext vmap_area. The caller is responsible for checking + * that. + */ +static unsigned long pvm_determine_end(struct vmap_area **pnext, + struct vmap_area **pprev, + unsigned long align) +{ + const unsigned long vmalloc_end = VMALLOC_END & ~(align - 1); + unsigned long addr; + + if (*pnext) + addr = min((*pnext)->va_start & ~(align - 1), vmalloc_end); + else + addr = vmalloc_end; + + while (*pprev && (*pprev)->va_end > addr) { + *pnext = *pprev; + *pprev = node_to_va(rb_prev(&(*pnext)->rb_node)); + } + + return addr; +} + +/** + * pcpu_get_vm_areas - allocate vmalloc areas for percpu allocator + * @offsets: array containing offset of each area + * @sizes: array containing size of each area + * @nr_vms: the number of areas to allocate + * @align: alignment, all entries in @offsets and @sizes must be aligned to this + * @gfp_mask: allocation mask + * + * Returns: kmalloc'd vm_struct pointer array pointing to allocated + * vm_structs on success, %NULL on failure + * + * Percpu allocator wants to use congruent vm areas so that it can + * maintain the offsets among percpu areas. This function allocates + * congruent vmalloc areas for it. These areas tend to be scattered + * pretty far, distance between two areas easily going up to + * gigabytes. To avoid interacting with regular vmallocs, these areas + * are allocated from top. + * + * Despite its complicated look, this allocator is rather simple. It + * does everything top-down and scans areas from the end looking for + * matching slot. While scanning, if any of the areas overlaps with + * existing vmap_area, the base address is pulled down to fit the + * area. Scanning is repeated till all the areas fit and then all + * necessary data structres are inserted and the result is returned. + */ +struct vm_struct **pcpu_get_vm_areas(const unsigned long *offsets, + const size_t *sizes, int nr_vms, + size_t align, gfp_t gfp_mask) +{ + const unsigned long vmalloc_start = ALIGN(VMALLOC_START, align); + const unsigned long vmalloc_end = VMALLOC_END & ~(align - 1); + struct vmap_area **vas, *prev, *next; + struct vm_struct **vms; + int area, area2, last_area, term_area; + unsigned long base, start, end, last_end; + bool purged = false; + + gfp_mask &= GFP_RECLAIM_MASK; + + /* verify parameters and allocate data structures */ + BUG_ON(align & ~PAGE_MASK || !is_power_of_2(align)); + for (last_area = 0, area = 0; area < nr_vms; area++) { + start = offsets[area]; + end = start + sizes[area]; + + /* is everything aligned properly? */ + BUG_ON(!IS_ALIGNED(offsets[area], align)); + BUG_ON(!IS_ALIGNED(sizes[area], align)); + + /* detect the area with the highest address */ + if (start > offsets[last_area]) + last_area = area; + + for (area2 = 0; area2 < nr_vms; area2++) { + unsigned long start2 = offsets[area2]; + unsigned long end2 = start2 + sizes[area2]; + + if (area2 == area) + continue; + + BUG_ON(start2 >= start && start2 < end); + BUG_ON(end2 <= end && end2 > start); + } + } + last_end = offsets[last_area] + sizes[last_area]; + + if (vmalloc_end - vmalloc_start < last_end) { + WARN_ON(true); + return NULL; + } + + vms = kzalloc(sizeof(vms[0]) * nr_vms, gfp_mask); + vas = kzalloc(sizeof(vas[0]) * nr_vms, gfp_mask); + if (!vas || !vms) + goto err_free; + + for (area = 0; area < nr_vms; area++) { + vas[area] = kzalloc(sizeof(struct vmap_area), gfp_mask); + vms[area] = kzalloc(sizeof(struct vm_struct), gfp_mask); + if (!vas[area] || !vms[area]) + goto err_free; + } +retry: + spin_lock(&vmap_area_lock); + + /* start scanning - we scan from the top, begin with the last area */ + area = term_area = last_area; + start = offsets[area]; + end = start + sizes[area]; + + if (!pvm_find_next_prev(vmap_area_pcpu_hole, &next, &prev)) { + base = vmalloc_end - last_end; + goto found; + } + base = pvm_determine_end(&next, &prev, align) - end; + + while (true) { + BUG_ON(next && next->va_end <= base + end); + BUG_ON(prev && prev->va_end > base + end); + + /* + * base might have underflowed, add last_end before + * comparing. + */ + if (base + last_end < vmalloc_start + last_end) { + spin_unlock(&vmap_area_lock); + if (!purged) { + purge_vmap_area_lazy(); + purged = true; + goto retry; + } + goto err_free; + } + + /* + * If next overlaps, move base downwards so that it's + * right below next and then recheck. + */ + if (next && next->va_start < base + end) { + base = pvm_determine_end(&next, &prev, align) - end; + term_area = area; + continue; + } + + /* + * If prev overlaps, shift down next and prev and move + * base so that it's right below new next and then + * recheck. + */ + if (prev && prev->va_end > base + start) { + next = prev; + prev = node_to_va(rb_prev(&next->rb_node)); + base = pvm_determine_end(&next, &prev, align) - end; + term_area = area; + continue; + } + + /* + * This area fits, move on to the previous one. If + * the previous one is the terminal one, we're done. + */ + area = (area + nr_vms - 1) % nr_vms; + if (area == term_area) + break; + start = offsets[area]; + end = start + sizes[area]; + pvm_find_next_prev(base + end, &next, &prev); + } +found: + /* we've found a fitting base, insert all va's */ + for (area = 0; area < nr_vms; area++) { + struct vmap_area *va = vas[area]; + + va->va_start = base + offsets[area]; + va->va_end = va->va_start + sizes[area]; + __insert_vmap_area(va); + } + + vmap_area_pcpu_hole = base + offsets[last_area]; + + spin_unlock(&vmap_area_lock); + + /* insert all vm's */ + for (area = 0; area < nr_vms; area++) + insert_vmalloc_vm(vms[area], vas[area], VM_ALLOC, + pcpu_get_vm_areas); + + kfree(vas); + return vms; + +err_free: + for (area = 0; area < nr_vms; area++) { + if (vas) + kfree(vas[area]); + if (vms) + kfree(vms[area]); + } + kfree(vas); + kfree(vms); + return NULL; +} + +/** + * pcpu_free_vm_areas - free vmalloc areas for percpu allocator + * @vms: vm_struct pointer array returned by pcpu_get_vm_areas() + * @nr_vms: the number of allocated areas + * + * Free vm_structs and the array allocated by pcpu_get_vm_areas(). + */ +void pcpu_free_vm_areas(struct vm_struct **vms, int nr_vms) +{ + int i; + + for (i = 0; i < nr_vms; i++) + free_vm_area(vms[i]); + kfree(vms); +} #ifdef CONFIG_PROC_FS static void *s_start(struct seq_file *m, loff_t *pos) -- cgit v1.2.3 From c8826dd538602d730ed2c18c6753f1bbfa6c4933 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Fri, 14 Aug 2009 15:00:52 +0900 Subject: percpu: update embedding first chunk allocator to handle sparse units Now that percpu core can handle very sparse units, given that vmalloc space is large enough, embedding first chunk allocator can use any memory to build the first chunk. This patch teaches pcpu_embed_first_chunk() about distances between cpus and to use alloc/free callbacks to allocate node specific areas for each group and use them for the first chunk. This brings the benefits of embedding allocator to NUMA configurations - no extra TLB pressure with the flexibility of unified dynamic allocator and no need to restructure arch code to build memory layout suitable for percpu. With units put into atom_size aligned groups according to cpu distances, using large page for dynamic chunks is also easily possible with falling back to reuglar pages if large allocation fails. Embedding allocator users are converted to specify NULL cpu_distance_fn, so this patch doesn't cause any visible behavior difference. Following patches will convert them. Signed-off-by: Tejun Heo --- arch/x86/kernel/setup_percpu.c | 4 +- include/linux/percpu.h | 7 ++- mm/percpu.c | 113 +++++++++++++++++++++++++++++++---------- 3 files changed, 93 insertions(+), 31 deletions(-) (limited to 'include') diff --git a/arch/x86/kernel/setup_percpu.c b/arch/x86/kernel/setup_percpu.c index 9becc5d4b518..67f6314de9f1 100644 --- a/arch/x86/kernel/setup_percpu.c +++ b/arch/x86/kernel/setup_percpu.c @@ -234,7 +234,9 @@ static int __init setup_pcpu_embed(bool chosen) return -EINVAL; return pcpu_embed_first_chunk(PERCPU_FIRST_CHUNK_RESERVE, - reserve - PERCPU_FIRST_CHUNK_RESERVE); + reserve - PERCPU_FIRST_CHUNK_RESERVE, + PAGE_SIZE, NULL, pcpu_fc_alloc, + pcpu_fc_free); } /* diff --git a/include/linux/percpu.h b/include/linux/percpu.h index a7ec840f596c..25359932740e 100644 --- a/include/linux/percpu.h +++ b/include/linux/percpu.h @@ -110,8 +110,11 @@ extern int __init pcpu_setup_first_chunk(const struct pcpu_alloc_info *ai, void *base_addr); #ifdef CONFIG_NEED_PER_CPU_EMBED_FIRST_CHUNK -extern int __init pcpu_embed_first_chunk(size_t reserved_size, - ssize_t dyn_size); +extern int __init pcpu_embed_first_chunk(size_t reserved_size, ssize_t dyn_size, + size_t atom_size, + pcpu_fc_cpu_distance_fn_t cpu_distance_fn, + pcpu_fc_alloc_fn_t alloc_fn, + pcpu_fc_free_fn_t free_fn); #endif #ifdef CONFIG_NEED_PER_CPU_PAGE_FIRST_CHUNK diff --git a/mm/percpu.c b/mm/percpu.c index cc9c4c64606d..c2826d05505c 100644 --- a/mm/percpu.c +++ b/mm/percpu.c @@ -1747,15 +1747,25 @@ early_param("percpu_alloc", percpu_alloc_setup); * pcpu_embed_first_chunk - embed the first percpu chunk into bootmem * @reserved_size: the size of reserved percpu area in bytes * @dyn_size: free size for dynamic allocation in bytes, -1 for auto + * @atom_size: allocation atom size + * @cpu_distance_fn: callback to determine distance between cpus, optional + * @alloc_fn: function to allocate percpu page + * @free_fn: funtion to free percpu page * * This is a helper to ease setting up embedded first percpu chunk and * can be called where pcpu_setup_first_chunk() is expected. * * If this function is used to setup the first chunk, it is allocated - * as a contiguous area using bootmem allocator and used as-is without - * being mapped into vmalloc area. This enables the first chunk to - * piggy back on the linear physical mapping which often uses larger - * page size. + * by calling @alloc_fn and used as-is without being mapped into + * vmalloc area. Allocations are always whole multiples of @atom_size + * aligned to @atom_size. + * + * This enables the first chunk to piggy back on the linear physical + * mapping which often uses larger page size. Please note that this + * can result in very sparse cpu->unit mapping on NUMA machines thus + * requiring large vmalloc address space. Don't use this allocator if + * vmalloc space is not orders of magnitude larger than distances + * between node memory addresses (ie. 32bit NUMA machines). * * When @dyn_size is positive, dynamic area might be larger than * specified to fill page alignment. When @dyn_size is auto, @@ -1763,53 +1773,88 @@ early_param("percpu_alloc", percpu_alloc_setup); * and reserved areas. * * If the needed size is smaller than the minimum or specified unit - * size, the leftover is returned to the bootmem allocator. + * size, the leftover is returned using @free_fn. * * RETURNS: * 0 on success, -errno on failure. */ -int __init pcpu_embed_first_chunk(size_t reserved_size, ssize_t dyn_size) +int __init pcpu_embed_first_chunk(size_t reserved_size, ssize_t dyn_size, + size_t atom_size, + pcpu_fc_cpu_distance_fn_t cpu_distance_fn, + pcpu_fc_alloc_fn_t alloc_fn, + pcpu_fc_free_fn_t free_fn) { + void *base = (void *)ULONG_MAX; + void **areas = NULL; struct pcpu_alloc_info *ai; - size_t size_sum, chunk_size; - void *base; - int unit; - int rc; + size_t size_sum, areas_size; + int group, i, rc; - ai = pcpu_build_alloc_info(reserved_size, dyn_size, PAGE_SIZE, NULL); + ai = pcpu_build_alloc_info(reserved_size, dyn_size, atom_size, + cpu_distance_fn); if (IS_ERR(ai)) return PTR_ERR(ai); - BUG_ON(ai->nr_groups != 1); - BUG_ON(ai->groups[0].nr_units != num_possible_cpus()); size_sum = ai->static_size + ai->reserved_size + ai->dyn_size; - chunk_size = ai->unit_size * num_possible_cpus(); + areas_size = PFN_ALIGN(ai->nr_groups * sizeof(void *)); - base = __alloc_bootmem_nopanic(chunk_size, PAGE_SIZE, - __pa(MAX_DMA_ADDRESS)); - if (!base) { - pr_warning("PERCPU: failed to allocate %zu bytes for " - "embedding\n", chunk_size); + areas = alloc_bootmem_nopanic(areas_size); + if (!areas) { rc = -ENOMEM; - goto out_free_ai; + goto out_free; } - /* return the leftover and copy */ - for (unit = 0; unit < num_possible_cpus(); unit++) { - void *ptr = base + unit * ai->unit_size; + /* allocate, copy and determine base address */ + for (group = 0; group < ai->nr_groups; group++) { + struct pcpu_group_info *gi = &ai->groups[group]; + unsigned int cpu = NR_CPUS; + void *ptr; + + for (i = 0; i < gi->nr_units && cpu == NR_CPUS; i++) + cpu = gi->cpu_map[i]; + BUG_ON(cpu == NR_CPUS); + + /* allocate space for the whole group */ + ptr = alloc_fn(cpu, gi->nr_units * ai->unit_size, atom_size); + if (!ptr) { + rc = -ENOMEM; + goto out_free_areas; + } + areas[group] = ptr; - free_bootmem(__pa(ptr + size_sum), ai->unit_size - size_sum); - memcpy(ptr, __per_cpu_load, ai->static_size); + base = min(ptr, base); + + for (i = 0; i < gi->nr_units; i++, ptr += ai->unit_size) { + if (gi->cpu_map[i] == NR_CPUS) { + /* unused unit, free whole */ + free_fn(ptr, ai->unit_size); + continue; + } + /* copy and return the unused part */ + memcpy(ptr, __per_cpu_load, ai->static_size); + free_fn(ptr + size_sum, ai->unit_size - size_sum); + } } - /* we're ready, commit */ + /* base address is now known, determine group base offsets */ + for (group = 0; group < ai->nr_groups; group++) + ai->groups[group].base_offset = areas[group] - base; + pr_info("PERCPU: Embedded %zu pages/cpu @%p s%zu r%zu d%zu u%zu\n", PFN_DOWN(size_sum), base, ai->static_size, ai->reserved_size, ai->dyn_size, ai->unit_size); rc = pcpu_setup_first_chunk(ai, base); -out_free_ai: + goto out_free; + +out_free_areas: + for (group = 0; group < ai->nr_groups; group++) + free_fn(areas[group], + ai->groups[group].nr_units * ai->unit_size); +out_free: pcpu_free_alloc_info(ai); + if (areas) + free_bootmem(__pa(areas), areas_size); return rc; } #endif /* CONFIG_NEED_PER_CPU_EMBED_FIRST_CHUNK || @@ -2177,6 +2222,17 @@ void *pcpu_lpage_remapped(void *kaddr) unsigned long __per_cpu_offset[NR_CPUS] __read_mostly; EXPORT_SYMBOL(__per_cpu_offset); +static void * __init pcpu_dfl_fc_alloc(unsigned int cpu, size_t size, + size_t align) +{ + return __alloc_bootmem_nopanic(size, align, __pa(MAX_DMA_ADDRESS)); +} + +static void __init pcpu_dfl_fc_free(void *ptr, size_t size) +{ + free_bootmem(__pa(ptr), size); +} + void __init setup_per_cpu_areas(void) { unsigned long delta; @@ -2188,7 +2244,8 @@ void __init setup_per_cpu_areas(void) * what the legacy allocator did. */ rc = pcpu_embed_first_chunk(PERCPU_MODULE_RESERVE, - PERCPU_DYNAMIC_RESERVE); + PERCPU_DYNAMIC_RESERVE, PAGE_SIZE, NULL, + pcpu_dfl_fc_alloc, pcpu_dfl_fc_free); if (rc < 0) panic("Failed to initialized percpu areas."); -- cgit v1.2.3 From e933a73f48e3b2d40cfa56d81e2646f194b5a66a Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Fri, 14 Aug 2009 15:00:53 +0900 Subject: percpu: kill lpage first chunk allocator With x86 converted to embedding allocator, lpage doesn't have any user left. Kill it along with cpa handling code. Signed-off-by: Tejun Heo Cc: Jan Beulich --- Documentation/kernel-parameters.txt | 10 +- arch/x86/mm/pageattr.c | 20 +-- include/linux/percpu.h | 16 --- mm/percpu.c | 241 ------------------------------------ 4 files changed, 6 insertions(+), 281 deletions(-) (limited to 'include') diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt index dee9ce2e6cfa..e710093e3d32 100644 --- a/Documentation/kernel-parameters.txt +++ b/Documentation/kernel-parameters.txt @@ -1920,11 +1920,11 @@ and is between 256 and 4096 characters. It is defined in the file See arch/parisc/kernel/pdc_chassis.c percpu_alloc= Select which percpu first chunk allocator to use. - Currently supported values are "embed", "page" and - "lpage". Archs may support subset or none of the - selections. See comments in mm/percpu.c for details - on each allocator. This parameter is primarily for - debugging and performance comparison. + Currently supported values are "embed" and "page". + Archs may support subset or none of the selections. + See comments in mm/percpu.c for details on each + allocator. This parameter is primarily for debugging + and performance comparison. pf. [PARIDE] See Documentation/blockdev/paride.txt. diff --git a/arch/x86/mm/pageattr.c b/arch/x86/mm/pageattr.c index dce282f65700..f53cfc7f963d 100644 --- a/arch/x86/mm/pageattr.c +++ b/arch/x86/mm/pageattr.c @@ -687,7 +687,7 @@ static int cpa_process_alias(struct cpa_data *cpa) { struct cpa_data alias_cpa; unsigned long laddr = (unsigned long)__va(cpa->pfn << PAGE_SHIFT); - unsigned long vaddr, remapped; + unsigned long vaddr; int ret; if (cpa->pfn >= max_pfn_mapped) @@ -745,24 +745,6 @@ static int cpa_process_alias(struct cpa_data *cpa) } #endif - /* - * If the PMD page was partially used for per-cpu remapping, - * the recycled area needs to be split and modified. Because - * the area is always proper subset of a PMD page - * cpa->numpages is guaranteed to be 1 for these areas, so - * there's no need to loop over and check for further remaps. - */ - remapped = (unsigned long)pcpu_lpage_remapped((void *)laddr); - if (remapped) { - WARN_ON(cpa->numpages > 1); - alias_cpa = *cpa; - alias_cpa.vaddr = &remapped; - alias_cpa.flags &= ~(CPA_PAGES_ARRAY | CPA_ARRAY); - ret = __change_page_attr_set_clr(&alias_cpa, 0); - if (ret) - return ret; - } - return 0; } diff --git a/include/linux/percpu.h b/include/linux/percpu.h index 25359932740e..878836ca999c 100644 --- a/include/linux/percpu.h +++ b/include/linux/percpu.h @@ -82,7 +82,6 @@ enum pcpu_fc { PCPU_FC_AUTO, PCPU_FC_EMBED, PCPU_FC_PAGE, - PCPU_FC_LPAGE, PCPU_FC_NR, }; @@ -95,7 +94,6 @@ typedef void * (*pcpu_fc_alloc_fn_t)(unsigned int cpu, size_t size, typedef void (*pcpu_fc_free_fn_t)(void *ptr, size_t size); typedef void (*pcpu_fc_populate_pte_fn_t)(unsigned long addr); typedef int (pcpu_fc_cpu_distance_fn_t)(unsigned int from, unsigned int to); -typedef void (*pcpu_fc_map_fn_t)(void *ptr, size_t size, void *addr); extern struct pcpu_alloc_info * __init pcpu_alloc_alloc_info(int nr_groups, int nr_units); @@ -124,20 +122,6 @@ extern int __init pcpu_page_first_chunk(size_t reserved_size, pcpu_fc_populate_pte_fn_t populate_pte_fn); #endif -#ifdef CONFIG_NEED_PER_CPU_LPAGE_FIRST_CHUNK -extern int __init pcpu_lpage_first_chunk(const struct pcpu_alloc_info *ai, - pcpu_fc_alloc_fn_t alloc_fn, - pcpu_fc_free_fn_t free_fn, - pcpu_fc_map_fn_t map_fn); - -extern void *pcpu_lpage_remapped(void *kaddr); -#else -static inline void *pcpu_lpage_remapped(void *kaddr) -{ - return NULL; -} -#endif - /* * Use this to get to a cpu's version of the per-cpu object * dynamically allocated. Non-atomic access to the current CPU's diff --git a/mm/percpu.c b/mm/percpu.c index c2826d05505c..77933928107d 100644 --- a/mm/percpu.c +++ b/mm/percpu.c @@ -1713,7 +1713,6 @@ const char *pcpu_fc_names[PCPU_FC_NR] __initdata = { [PCPU_FC_AUTO] = "auto", [PCPU_FC_EMBED] = "embed", [PCPU_FC_PAGE] = "page", - [PCPU_FC_LPAGE] = "lpage", }; enum pcpu_fc pcpu_chosen_fc __initdata = PCPU_FC_AUTO; @@ -1729,10 +1728,6 @@ static int __init percpu_alloc_setup(char *str) #ifdef CONFIG_NEED_PER_CPU_PAGE_FIRST_CHUNK else if (!strcmp(str, "page")) pcpu_chosen_fc = PCPU_FC_PAGE; -#endif -#ifdef CONFIG_NEED_PER_CPU_LPAGE_FIRST_CHUNK - else if (!strcmp(str, "lpage")) - pcpu_chosen_fc = PCPU_FC_LPAGE; #endif else pr_warning("PERCPU: unknown allocator %s specified\n", str); @@ -1970,242 +1965,6 @@ out_free_ar: } #endif /* CONFIG_NEED_PER_CPU_PAGE_FIRST_CHUNK */ -#ifdef CONFIG_NEED_PER_CPU_LPAGE_FIRST_CHUNK -struct pcpul_ent { - void *ptr; - void *map_addr; -}; - -static size_t pcpul_size; -static size_t pcpul_lpage_size; -static int pcpul_nr_lpages; -static struct pcpul_ent *pcpul_map; - -static bool __init pcpul_unit_to_cpu(int unit, const struct pcpu_alloc_info *ai, - unsigned int *cpup) -{ - int group, cunit; - - for (group = 0, cunit = 0; group < ai->nr_groups; group++) { - const struct pcpu_group_info *gi = &ai->groups[group]; - - if (unit < cunit + gi->nr_units) { - if (cpup) - *cpup = gi->cpu_map[unit - cunit]; - return true; - } - cunit += gi->nr_units; - } - - return false; -} - -static int __init pcpul_cpu_to_unit(int cpu, const struct pcpu_alloc_info *ai) -{ - int group, unit, i; - - for (group = 0, unit = 0; group < ai->nr_groups; group++, unit += i) { - const struct pcpu_group_info *gi = &ai->groups[group]; - - for (i = 0; i < gi->nr_units; i++) - if (gi->cpu_map[i] == cpu) - return unit + i; - } - BUG(); -} - -/** - * pcpu_lpage_first_chunk - remap the first percpu chunk using large page - * @ai: pcpu_alloc_info - * @alloc_fn: function to allocate percpu lpage, always called with lpage_size - * @free_fn: function to free percpu memory, @size <= lpage_size - * @map_fn: function to map percpu lpage, always called with lpage_size - * - * This allocator uses large page to build and map the first chunk. - * Unlike other helpers, the caller should provide fully initialized - * @ai. This can be done using pcpu_build_alloc_info(). This two - * stage initialization is to allow arch code to evaluate the - * parameters before committing to it. - * - * Large pages are allocated as directed by @unit_map and other - * parameters and mapped to vmalloc space. Unused holes are returned - * to the page allocator. Note that these holes end up being actively - * mapped twice - once to the physical mapping and to the vmalloc area - * for the first percpu chunk. Depending on architecture, this might - * cause problem when changing page attributes of the returned area. - * These double mapped areas can be detected using - * pcpu_lpage_remapped(). - * - * RETURNS: - * 0 on success, -errno on failure. - */ -int __init pcpu_lpage_first_chunk(const struct pcpu_alloc_info *ai, - pcpu_fc_alloc_fn_t alloc_fn, - pcpu_fc_free_fn_t free_fn, - pcpu_fc_map_fn_t map_fn) -{ - static struct vm_struct vm; - const size_t lpage_size = ai->atom_size; - size_t chunk_size, map_size; - unsigned int cpu; - int i, j, unit, nr_units, rc; - - nr_units = 0; - for (i = 0; i < ai->nr_groups; i++) - nr_units += ai->groups[i].nr_units; - - chunk_size = ai->unit_size * nr_units; - BUG_ON(chunk_size % lpage_size); - - pcpul_size = ai->static_size + ai->reserved_size + ai->dyn_size; - pcpul_lpage_size = lpage_size; - pcpul_nr_lpages = chunk_size / lpage_size; - - /* allocate pointer array and alloc large pages */ - map_size = pcpul_nr_lpages * sizeof(pcpul_map[0]); - pcpul_map = alloc_bootmem(map_size); - - /* allocate all pages */ - for (i = 0; i < pcpul_nr_lpages; i++) { - size_t offset = i * lpage_size; - int first_unit = offset / ai->unit_size; - int last_unit = (offset + lpage_size - 1) / ai->unit_size; - void *ptr; - - /* find out which cpu is mapped to this unit */ - for (unit = first_unit; unit <= last_unit; unit++) - if (pcpul_unit_to_cpu(unit, ai, &cpu)) - goto found; - continue; - found: - ptr = alloc_fn(cpu, lpage_size, lpage_size); - if (!ptr) { - pr_warning("PERCPU: failed to allocate large page " - "for cpu%u\n", cpu); - goto enomem; - } - - pcpul_map[i].ptr = ptr; - } - - /* return unused holes */ - for (unit = 0; unit < nr_units; unit++) { - size_t start = unit * ai->unit_size; - size_t end = start + ai->unit_size; - size_t off, next; - - /* don't free used part of occupied unit */ - if (pcpul_unit_to_cpu(unit, ai, NULL)) - start += pcpul_size; - - /* unit can span more than one page, punch the holes */ - for (off = start; off < end; off = next) { - void *ptr = pcpul_map[off / lpage_size].ptr; - next = min(roundup(off + 1, lpage_size), end); - if (ptr) - free_fn(ptr + off % lpage_size, next - off); - } - } - - /* allocate address, map and copy */ - vm.flags = VM_ALLOC; - vm.size = chunk_size; - vm_area_register_early(&vm, ai->unit_size); - - for (i = 0; i < pcpul_nr_lpages; i++) { - if (!pcpul_map[i].ptr) - continue; - pcpul_map[i].map_addr = vm.addr + i * lpage_size; - map_fn(pcpul_map[i].ptr, lpage_size, pcpul_map[i].map_addr); - } - - for_each_possible_cpu(cpu) - memcpy(vm.addr + pcpul_cpu_to_unit(cpu, ai) * ai->unit_size, - __per_cpu_load, ai->static_size); - - /* we're ready, commit */ - pr_info("PERCPU: large pages @%p s%zu r%zu d%zu u%zu\n", - vm.addr, ai->static_size, ai->reserved_size, ai->dyn_size, - ai->unit_size); - - rc = pcpu_setup_first_chunk(ai, vm.addr); - - /* - * Sort pcpul_map array for pcpu_lpage_remapped(). Unmapped - * lpages are pushed to the end and trimmed. - */ - for (i = 0; i < pcpul_nr_lpages - 1; i++) - for (j = i + 1; j < pcpul_nr_lpages; j++) { - struct pcpul_ent tmp; - - if (!pcpul_map[j].ptr) - continue; - if (pcpul_map[i].ptr && - pcpul_map[i].ptr < pcpul_map[j].ptr) - continue; - - tmp = pcpul_map[i]; - pcpul_map[i] = pcpul_map[j]; - pcpul_map[j] = tmp; - } - - while (pcpul_nr_lpages && !pcpul_map[pcpul_nr_lpages - 1].ptr) - pcpul_nr_lpages--; - - return rc; - -enomem: - for (i = 0; i < pcpul_nr_lpages; i++) - if (pcpul_map[i].ptr) - free_fn(pcpul_map[i].ptr, lpage_size); - free_bootmem(__pa(pcpul_map), map_size); - return -ENOMEM; -} - -/** - * pcpu_lpage_remapped - determine whether a kaddr is in pcpul recycled area - * @kaddr: the kernel address in question - * - * Determine whether @kaddr falls in the pcpul recycled area. This is - * used by pageattr to detect VM aliases and break up the pcpu large - * page mapping such that the same physical page is not mapped under - * different attributes. - * - * The recycled area is always at the tail of a partially used large - * page. - * - * RETURNS: - * Address of corresponding remapped pcpu address if match is found; - * otherwise, NULL. - */ -void *pcpu_lpage_remapped(void *kaddr) -{ - unsigned long lpage_mask = pcpul_lpage_size - 1; - void *lpage_addr = (void *)((unsigned long)kaddr & ~lpage_mask); - unsigned long offset = (unsigned long)kaddr & lpage_mask; - int left = 0, right = pcpul_nr_lpages - 1; - int pos; - - /* pcpul in use at all? */ - if (!pcpul_map) - return NULL; - - /* okay, perform binary search */ - while (left <= right) { - pos = (left + right) / 2; - - if (pcpul_map[pos].ptr < lpage_addr) - left = pos + 1; - else if (pcpul_map[pos].ptr > lpage_addr) - right = pos - 1; - else - return pcpul_map[pos].map_addr + offset; - } - - return NULL; -} -#endif /* CONFIG_NEED_PER_CPU_LPAGE_FIRST_CHUNK */ - /* * Generic percpu area setup. * -- cgit v1.2.3 From f401a6f7ede753e56b84025e7d2db0d5ef560ce6 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Fri, 7 Aug 2009 14:51:05 +0200 Subject: cfg80211: use reassociation when possible With the move of everything related to the SME from mac80211 to cfg80211, we lost the ability to send reassociation frames. This adds them back, but only for wireless extensions. With the userspace SME, it shall control assoc vs. reassoc (it already can do so with the nl80211 interface). I haven't touched the connect() implementation, so it is not possible to reassociate with the nl80211 connect primitive. I think that should be done with the NL80211_CMD_ROAM command, but we'll have to see how that can be handled in the future, especially with fullmac chips. This patch addresses only the immediate regression we had in mac80211, which previously sent reassoc. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- include/net/cfg80211.h | 4 ++-- net/wireless/core.h | 4 +++- net/wireless/mlme.c | 11 ++++++++++ net/wireless/sme.c | 54 ++++++++++++++++++++++++++++++++++++++----------- net/wireless/wext-sme.c | 8 +++++++- 5 files changed, 65 insertions(+), 16 deletions(-) (limited to 'include') diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index 64df51d9a89f..1ee30fcd6fdc 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -1335,10 +1335,10 @@ struct wireless_dev { struct cfg80211_cached_keys *keys; u8 *ie; size_t ie_len; - u8 bssid[ETH_ALEN]; + u8 bssid[ETH_ALEN], prev_bssid[ETH_ALEN]; u8 ssid[IEEE80211_MAX_SSID_LEN]; s8 default_key, default_mgmt_key; - bool ps; + bool ps, prev_bssid_valid; int ps_timeout; } wext; #endif diff --git a/net/wireless/core.h b/net/wireless/core.h index 5696b95af9be..92e0492b0e4d 100644 --- a/net/wireless/core.h +++ b/net/wireless/core.h @@ -335,7 +335,8 @@ void __cfg80211_connect_result(struct net_device *dev, const u8 *bssid, int __cfg80211_connect(struct cfg80211_registered_device *rdev, struct net_device *dev, struct cfg80211_connect_params *connect, - struct cfg80211_cached_keys *connkeys); + struct cfg80211_cached_keys *connkeys, + const u8 *prev_bssid); int cfg80211_connect(struct cfg80211_registered_device *rdev, struct net_device *dev, struct cfg80211_connect_params *connect, @@ -353,6 +354,7 @@ int cfg80211_mgd_wext_connect(struct cfg80211_registered_device *rdev, struct wireless_dev *wdev); void cfg80211_conn_work(struct work_struct *work); +bool cfg80211_sme_failed_reassoc(struct wireless_dev *wdev); /* internal helpers */ int cfg80211_validate_key_settings(struct cfg80211_registered_device *rdev, diff --git a/net/wireless/mlme.c b/net/wireless/mlme.c index 51d5df67c632..da64071ceb84 100644 --- a/net/wireless/mlme.c +++ b/net/wireless/mlme.c @@ -67,6 +67,16 @@ void cfg80211_send_rx_assoc(struct net_device *dev, const u8 *buf, size_t len) status_code = le16_to_cpu(mgmt->u.assoc_resp.status_code); + /* + * This is a bit of a hack, we don't notify userspace of + * a (re-)association reply if we tried to send a reassoc + * and got a reject -- we only try again with an assoc + * frame instead of reassoc. + */ + if (status_code != WLAN_STATUS_SUCCESS && wdev->conn && + cfg80211_sme_failed_reassoc(wdev)) + goto out; + nl80211_send_rx_assoc(rdev, dev, buf, len, GFP_KERNEL); if (status_code == WLAN_STATUS_SUCCESS) { @@ -97,6 +107,7 @@ void cfg80211_send_rx_assoc(struct net_device *dev, const u8 *buf, size_t len) cfg80211_put_bss(&bss->pub); } + out: wdev_unlock(wdev); } EXPORT_SYMBOL(cfg80211_send_rx_assoc); diff --git a/net/wireless/sme.c b/net/wireless/sme.c index 219c3bc2c37d..104b33e34d22 100644 --- a/net/wireless/sme.c +++ b/net/wireless/sme.c @@ -27,10 +27,10 @@ struct cfg80211_conn { CFG80211_CONN_ASSOCIATE_NEXT, CFG80211_CONN_ASSOCIATING, } state; - u8 bssid[ETH_ALEN]; + u8 bssid[ETH_ALEN], prev_bssid[ETH_ALEN]; u8 *ie; size_t ie_len; - bool auto_auth; + bool auto_auth, prev_bssid_valid; }; @@ -110,6 +110,7 @@ static int cfg80211_conn_do_work(struct wireless_dev *wdev) { struct cfg80211_registered_device *rdev = wiphy_to_dev(wdev->wiphy); struct cfg80211_connect_params *params; + const u8 *prev_bssid = NULL; int err; ASSERT_WDEV_LOCK(wdev); @@ -135,15 +136,11 @@ static int cfg80211_conn_do_work(struct wireless_dev *wdev) case CFG80211_CONN_ASSOCIATE_NEXT: BUG_ON(!rdev->ops->assoc); wdev->conn->state = CFG80211_CONN_ASSOCIATING; - /* - * We could, later, implement roaming here and then actually - * set prev_bssid to non-NULL. But then we need to be aware - * that some APs don't like that -- so we'd need to retry - * the association. - */ + if (wdev->conn->prev_bssid_valid) + prev_bssid = wdev->conn->prev_bssid; err = __cfg80211_mlme_assoc(rdev, wdev->netdev, params->channel, params->bssid, - NULL, + prev_bssid, params->ssid, params->ssid_len, params->ie, params->ie_len, false, ¶ms->crypto); @@ -316,6 +313,28 @@ void cfg80211_sme_rx_auth(struct net_device *dev, } } +bool cfg80211_sme_failed_reassoc(struct wireless_dev *wdev) +{ + struct wiphy *wiphy = wdev->wiphy; + struct cfg80211_registered_device *rdev = wiphy_to_dev(wiphy); + + if (WARN_ON(!wdev->conn)) + return false; + + if (!wdev->conn->prev_bssid_valid) + return false; + + /* + * Some stupid APs don't accept reassoc, so we + * need to fall back to trying regular assoc. + */ + wdev->conn->prev_bssid_valid = false; + wdev->conn->state = CFG80211_CONN_ASSOCIATE_NEXT; + schedule_work(&rdev->conn_work); + + return true; +} + void __cfg80211_connect_result(struct net_device *dev, const u8 *bssid, const u8 *req_ie, size_t req_ie_len, const u8 *resp_ie, size_t resp_ie_len, @@ -359,8 +378,11 @@ void __cfg80211_connect_result(struct net_device *dev, const u8 *bssid, memset(&wrqu, 0, sizeof(wrqu)); wrqu.ap_addr.sa_family = ARPHRD_ETHER; - if (bssid && status == WLAN_STATUS_SUCCESS) + if (bssid && status == WLAN_STATUS_SUCCESS) { memcpy(wrqu.ap_addr.sa_data, bssid, ETH_ALEN); + memcpy(wdev->wext.prev_bssid, bssid, ETH_ALEN); + wdev->wext.prev_bssid_valid = true; + } wireless_send_event(dev, SIOCGIWAP, &wrqu, NULL); } #endif @@ -511,6 +533,8 @@ void __cfg80211_roamed(struct wireless_dev *wdev, const u8 *bssid, memset(&wrqu, 0, sizeof(wrqu)); wrqu.ap_addr.sa_family = ARPHRD_ETHER; memcpy(wrqu.ap_addr.sa_data, bssid, ETH_ALEN); + memcpy(wdev->wext.prev_bssid, bssid, ETH_ALEN); + wdev->wext.prev_bssid_valid = true; wireless_send_event(wdev->netdev, SIOCGIWAP, &wrqu, NULL); #endif } @@ -643,7 +667,8 @@ EXPORT_SYMBOL(cfg80211_disconnected); int __cfg80211_connect(struct cfg80211_registered_device *rdev, struct net_device *dev, struct cfg80211_connect_params *connect, - struct cfg80211_cached_keys *connkeys) + struct cfg80211_cached_keys *connkeys, + const u8 *prev_bssid) { struct wireless_dev *wdev = dev->ieee80211_ptr; struct ieee80211_channel *chan; @@ -742,6 +767,11 @@ int __cfg80211_connect(struct cfg80211_registered_device *rdev, wdev->sme_state = CFG80211_SME_CONNECTING; wdev->connect_keys = connkeys; + if (prev_bssid) { + memcpy(wdev->conn->prev_bssid, prev_bssid, ETH_ALEN); + wdev->conn->prev_bssid_valid = true; + } + /* we're good if we have both BSSID and channel */ if (wdev->conn->params.bssid && wdev->conn->params.channel) { wdev->conn->state = CFG80211_CONN_AUTHENTICATE_NEXT; @@ -794,7 +824,7 @@ int cfg80211_connect(struct cfg80211_registered_device *rdev, mutex_lock(&rdev->devlist_mtx); wdev_lock(dev->ieee80211_ptr); - err = __cfg80211_connect(rdev, dev, connect, connkeys); + err = __cfg80211_connect(rdev, dev, connect, connkeys, NULL); wdev_unlock(dev->ieee80211_ptr); mutex_unlock(&rdev->devlist_mtx); diff --git a/net/wireless/wext-sme.c b/net/wireless/wext-sme.c index fe1a53639122..907470063f22 100644 --- a/net/wireless/wext-sme.c +++ b/net/wireless/wext-sme.c @@ -15,6 +15,7 @@ int cfg80211_mgd_wext_connect(struct cfg80211_registered_device *rdev, struct wireless_dev *wdev) { struct cfg80211_cached_keys *ck = NULL; + const u8 *prev_bssid = NULL; int err, i; ASSERT_RDEV_LOCK(rdev); @@ -42,8 +43,12 @@ int cfg80211_mgd_wext_connect(struct cfg80211_registered_device *rdev, for (i = 0; i < 6; i++) ck->params[i].key = ck->data[i]; } + + if (wdev->wext.prev_bssid_valid) + prev_bssid = wdev->wext.prev_bssid; + err = __cfg80211_connect(rdev, wdev->netdev, - &wdev->wext.connect, ck); + &wdev->wext.connect, ck, prev_bssid); if (err) kfree(ck); @@ -184,6 +189,7 @@ int cfg80211_mgd_wext_siwessid(struct net_device *dev, goto out; } + wdev->wext.prev_bssid_valid = false; wdev->wext.connect.ssid = wdev->wext.ssid; memcpy(wdev->wext.ssid, ssid, len); wdev->wext.connect.ssid_len = len; -- cgit v1.2.3 From f5ea9120be2e5d5c846243416cfdce01d02f5836 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Fri, 7 Aug 2009 16:17:38 +0200 Subject: nl80211: add generation number to all dumps In order for userspace to be able to figure out whether it obtained a consistent snapshot of data or not when using netlink dumps, we need to have a generation number in each dump message that indicates whether the list has changed or not -- its value is arbitrary. This patch adds such a number to all dumps, this needs some mac80211 involvement to keep track of a generation number to start with when adding/removing mesh paths or stations. The wiphy and netdev lists can be fully handled within cfg80211, of course, but generation numbers need to be stored there as well. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- include/linux/nl80211.h | 17 ++++++++++++----- include/net/cfg80211.h | 12 ++++++++++++ net/mac80211/cfg.c | 4 ++++ net/mac80211/ieee80211_i.h | 1 + net/mac80211/mesh.h | 2 ++ net/mac80211/mesh_pathtbl.c | 5 +++++ net/mac80211/sta_info.c | 2 ++ net/wireless/core.c | 5 +++++ net/wireless/core.h | 2 ++ net/wireless/nl80211.c | 31 +++++++++++++++++++++---------- net/wireless/scan.c | 1 + 11 files changed, 67 insertions(+), 15 deletions(-) (limited to 'include') diff --git a/include/linux/nl80211.h b/include/linux/nl80211.h index cb3dc6027fd9..a8d71ed43a0e 100644 --- a/include/linux/nl80211.h +++ b/include/linux/nl80211.h @@ -480,10 +480,6 @@ enum nl80211_commands { * @NL80211_ATTR_SCAN_FREQUENCIES: nested attribute with frequencies (in MHz) * @NL80211_ATTR_SCAN_SSIDS: nested attribute with SSIDs, leave out for passive * scanning and include a zero-length SSID (wildcard) for wildcard scan - * @NL80211_ATTR_SCAN_GENERATION: the scan generation increases whenever the - * scan result list changes (BSS expired or added) so that applications - * can verify that they got a single, consistent snapshot (when all dump - * messages carried the same generation number) * @NL80211_ATTR_BSS: scan result BSS * * @NL80211_ATTR_REG_INITIATOR: indicates who requested the regulatory domain @@ -580,6 +576,14 @@ enum nl80211_commands { * * @NL80211_ATTR_PID: Process ID of a network namespace. * + * @NL80211_ATTR_GENERATION: Used to indicate consistent snapshots for + * dumps. This number increases whenever the object list being + * dumped changes, and as such userspace can verify that it has + * obtained a complete and consistent snapshot by verifying that + * all dump messages contain the same generation number. If it + * changed then the list changed and the dump should be repeated + * completely from scratch. + * * @NL80211_ATTR_MAX: highest attribute number currently defined * @__NL80211_ATTR_AFTER_LAST: internal use */ @@ -651,7 +655,7 @@ enum nl80211_attrs { NL80211_ATTR_SCAN_FREQUENCIES, NL80211_ATTR_SCAN_SSIDS, - NL80211_ATTR_SCAN_GENERATION, + NL80211_ATTR_GENERATION, /* replaces old SCAN_GENERATION */ NL80211_ATTR_BSS, NL80211_ATTR_REG_INITIATOR, @@ -716,6 +720,9 @@ enum nl80211_attrs { NL80211_ATTR_MAX = __NL80211_ATTR_AFTER_LAST - 1 }; +/* source-level API compatibility */ +#define NL80211_ATTR_SCAN_GENERATION NL80211_ATTR_GENERATION + /* * Allow user space programs to use #ifdef on new attributes by defining them * here diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index 1ee30fcd6fdc..de7d116acc3d 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -372,6 +372,10 @@ struct rate_info { * @txrate: current unicast bitrate to this station * @rx_packets: packets received from this station * @tx_packets: packets transmitted to this station + * @generation: generation number for nl80211 dumps. + * This number should increase every time the list of stations + * changes, i.e. when a station is added or removed, so that + * userspace can tell whether it got a consistent snapshot. */ struct station_info { u32 filled; @@ -385,6 +389,8 @@ struct station_info { struct rate_info txrate; u32 rx_packets; u32 tx_packets; + + int generation; }; /** @@ -444,6 +450,10 @@ enum mpath_info_flags { * @flags: mesh path flags * @discovery_timeout: total mesh path discovery timeout, in msecs * @discovery_retries: mesh path discovery retries + * @generation: generation number for nl80211 dumps. + * This number should increase every time the list of mesh paths + * changes, i.e. when a station is added or removed, so that + * userspace can tell whether it got a consistent snapshot. */ struct mpath_info { u32 filled; @@ -454,6 +464,8 @@ struct mpath_info { u32 discovery_timeout; u8 discovery_retries; u8 flags; + + int generation; }; /** diff --git a/net/mac80211/cfg.c b/net/mac80211/cfg.c index 4bbf5007799b..5608f6c68413 100644 --- a/net/mac80211/cfg.c +++ b/net/mac80211/cfg.c @@ -323,6 +323,8 @@ static void sta_set_sinfo(struct sta_info *sta, struct station_info *sinfo) { struct ieee80211_sub_if_data *sdata = sta->sdata; + sinfo->generation = sdata->local->sta_generation; + sinfo->filled = STATION_INFO_INACTIVE_TIME | STATION_INFO_RX_BYTES | STATION_INFO_TX_BYTES | @@ -909,6 +911,8 @@ static void mpath_set_pinfo(struct mesh_path *mpath, u8 *next_hop, else memset(next_hop, 0, ETH_ALEN); + pinfo->generation = mesh_paths_generation; + pinfo->filled = MPATH_INFO_FRAME_QLEN | MPATH_INFO_DSN | MPATH_INFO_METRIC | diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h index 989591787aee..99433222bc5c 100644 --- a/net/mac80211/ieee80211_i.h +++ b/net/mac80211/ieee80211_i.h @@ -678,6 +678,7 @@ struct ieee80211_local { struct list_head sta_list; struct sta_info *sta_hash[STA_HASH_SIZE]; struct timer_list sta_cleanup; + int sta_generation; struct sk_buff_head pending[IEEE80211_MAX_QUEUES]; struct tasklet_struct tx_pending_tasklet; diff --git a/net/mac80211/mesh.h b/net/mac80211/mesh.h index 2a2ed182cb7e..ce538814b9bd 100644 --- a/net/mac80211/mesh.h +++ b/net/mac80211/mesh.h @@ -265,6 +265,8 @@ void mesh_path_discard_frame(struct sk_buff *skb, void mesh_path_quiesce(struct ieee80211_sub_if_data *sdata); void mesh_path_restart(struct ieee80211_sub_if_data *sdata); +extern int mesh_paths_generation; + #ifdef CONFIG_MAC80211_MESH extern int mesh_allocated; diff --git a/net/mac80211/mesh_pathtbl.c b/net/mac80211/mesh_pathtbl.c index 04b9e4d61b8e..431865a58622 100644 --- a/net/mac80211/mesh_pathtbl.c +++ b/net/mac80211/mesh_pathtbl.c @@ -38,6 +38,8 @@ struct mpath_node { static struct mesh_table *mesh_paths; static struct mesh_table *mpp_paths; /* Store paths for MPP&MAP */ +int mesh_paths_generation; + /* This lock will have the grow table function as writer and add / delete nodes * as readers. When reading the table (i.e. doing lookups) we are well protected * by RCU @@ -243,6 +245,8 @@ int mesh_path_add(u8 *dst, struct ieee80211_sub_if_data *sdata) mesh_paths->mean_chain_len * (mesh_paths->hash_mask + 1)) grow = 1; + mesh_paths_generation++; + spin_unlock(&mesh_paths->hashwlock[hash_idx]); read_unlock(&pathtbl_resize_lock); if (grow) { @@ -484,6 +488,7 @@ int mesh_path_del(u8 *addr, struct ieee80211_sub_if_data *sdata) err = -ENXIO; enddel: + mesh_paths_generation++; spin_unlock(&mesh_paths->hashwlock[hash_idx]); read_unlock(&pathtbl_resize_lock); return err; diff --git a/net/mac80211/sta_info.c b/net/mac80211/sta_info.c index a360bceeba59..eec001491e66 100644 --- a/net/mac80211/sta_info.c +++ b/net/mac80211/sta_info.c @@ -349,6 +349,7 @@ int sta_info_insert(struct sta_info *sta) goto out_free; } list_add(&sta->list, &local->sta_list); + local->sta_generation++; local->num_sta++; sta_info_hash_add(local, sta); @@ -485,6 +486,7 @@ static void __sta_info_unlink(struct sta_info **sta) } local->num_sta--; + local->sta_generation++; if (local->ops->sta_notify) { if (sdata->vif.type == NL80211_IFTYPE_AP_VLAN) diff --git a/net/wireless/core.c b/net/wireless/core.c index 1e189306560d..62e1ac00879b 100644 --- a/net/wireless/core.c +++ b/net/wireless/core.c @@ -32,6 +32,7 @@ MODULE_DESCRIPTION("wireless configuration support"); * only read the list, and that can happen quite * often because we need to do it for each command */ LIST_HEAD(cfg80211_rdev_list); +int cfg80211_rdev_list_generation; /* * This is used to protect the cfg80211_rdev_list @@ -511,6 +512,7 @@ int wiphy_register(struct wiphy *wiphy) wiphy_update_regulatory(wiphy, NL80211_REGDOM_SET_BY_CORE); list_add(&rdev->list, &cfg80211_rdev_list); + cfg80211_rdev_list_generation++; mutex_unlock(&cfg80211_mutex); @@ -593,6 +595,7 @@ void wiphy_unregister(struct wiphy *wiphy) reg_device_remove(wiphy); list_del(&rdev->list); + cfg80211_rdev_list_generation++; device_del(&rdev->wiphy.dev); debugfs_remove(rdev->wiphy.debugfsdir); @@ -653,6 +656,7 @@ static int cfg80211_netdev_notifier_call(struct notifier_block * nb, spin_lock_init(&wdev->event_lock); mutex_lock(&rdev->devlist_mtx); list_add(&wdev->list, &rdev->netdev_list); + rdev->devlist_generation++; /* can only change netns with wiphy */ dev->features |= NETIF_F_NETNS_LOCAL; @@ -733,6 +737,7 @@ static int cfg80211_netdev_notifier_call(struct notifier_block * nb, if (!list_empty(&wdev->list)) { sysfs_remove_link(&dev->dev.kobj, "phy80211"); list_del_init(&wdev->list); + rdev->devlist_generation++; mutex_destroy(&wdev->mtx); #ifdef CONFIG_WIRELESS_EXT kfree(wdev->wext.keys); diff --git a/net/wireless/core.h b/net/wireless/core.h index 92e0492b0e4d..639db52eeff7 100644 --- a/net/wireless/core.h +++ b/net/wireless/core.h @@ -49,6 +49,7 @@ struct cfg80211_registered_device { /* associate netdev list */ struct mutex devlist_mtx; struct list_head netdev_list; + int devlist_generation; /* BSSes/scanning */ spinlock_t bss_lock; @@ -101,6 +102,7 @@ bool wiphy_idx_valid(int wiphy_idx) extern struct mutex cfg80211_mutex; extern struct list_head cfg80211_rdev_list; +extern int cfg80211_rdev_list_generation; #define assert_cfg80211_lock() WARN_ON(!mutex_is_locked(&cfg80211_mutex)) diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index 2ff7376f35a3..b3d5c1df08dd 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -408,6 +408,9 @@ static int nl80211_send_wiphy(struct sk_buff *msg, u32 pid, u32 seq, int flags, NLA_PUT_U32(msg, NL80211_ATTR_WIPHY, dev->wiphy_idx); NLA_PUT_STRING(msg, NL80211_ATTR_WIPHY_NAME, wiphy_name(&dev->wiphy)); + NLA_PUT_U32(msg, NL80211_ATTR_GENERATION, + cfg80211_rdev_list_generation); + NLA_PUT_U8(msg, NL80211_ATTR_WIPHY_RETRY_SHORT, dev->wiphy.retry_short); NLA_PUT_U8(msg, NL80211_ATTR_WIPHY_RETRY_LONG, @@ -825,6 +828,11 @@ static int nl80211_send_iface(struct sk_buff *msg, u32 pid, u32 seq, int flags, NLA_PUT_U32(msg, NL80211_ATTR_WIPHY, rdev->wiphy_idx); NLA_PUT_STRING(msg, NL80211_ATTR_IFNAME, dev->name); NLA_PUT_U32(msg, NL80211_ATTR_IFTYPE, dev->ieee80211_ptr->iftype); + + NLA_PUT_U32(msg, NL80211_ATTR_GENERATION, + rdev->devlist_generation ^ + (cfg80211_rdev_list_generation << 2)); + return genlmsg_end(msg, hdr); nla_put_failure: @@ -838,12 +846,12 @@ static int nl80211_dump_interface(struct sk_buff *skb, struct netlink_callback * int if_idx = 0; int wp_start = cb->args[0]; int if_start = cb->args[1]; - struct cfg80211_registered_device *dev; + struct cfg80211_registered_device *rdev; struct wireless_dev *wdev; mutex_lock(&cfg80211_mutex); - list_for_each_entry(dev, &cfg80211_rdev_list, list) { - if (!net_eq(wiphy_net(&dev->wiphy), sock_net(skb->sk))) + list_for_each_entry(rdev, &cfg80211_rdev_list, list) { + if (!net_eq(wiphy_net(&rdev->wiphy), sock_net(skb->sk))) continue; if (wp_idx < wp_start) { wp_idx++; @@ -851,21 +859,21 @@ static int nl80211_dump_interface(struct sk_buff *skb, struct netlink_callback * } if_idx = 0; - mutex_lock(&dev->devlist_mtx); - list_for_each_entry(wdev, &dev->netdev_list, list) { + mutex_lock(&rdev->devlist_mtx); + list_for_each_entry(wdev, &rdev->netdev_list, list) { if (if_idx < if_start) { if_idx++; continue; } if (nl80211_send_iface(skb, NETLINK_CB(cb->skb).pid, cb->nlh->nlmsg_seq, NLM_F_MULTI, - dev, wdev->netdev) < 0) { - mutex_unlock(&dev->devlist_mtx); + rdev, wdev->netdev) < 0) { + mutex_unlock(&rdev->devlist_mtx); goto out; } if_idx++; } - mutex_unlock(&dev->devlist_mtx); + mutex_unlock(&rdev->devlist_mtx); wp_idx++; } @@ -1616,6 +1624,8 @@ static int nl80211_send_station(struct sk_buff *msg, u32 pid, u32 seq, NLA_PUT_U32(msg, NL80211_ATTR_IFINDEX, dev->ifindex); NLA_PUT(msg, NL80211_ATTR_MAC, ETH_ALEN, mac_addr); + NLA_PUT_U32(msg, NL80211_ATTR_GENERATION, sinfo->generation); + sinfoattr = nla_nest_start(msg, NL80211_ATTR_STA_INFO); if (!sinfoattr) goto nla_put_failure; @@ -2101,6 +2111,8 @@ static int nl80211_send_mpath(struct sk_buff *msg, u32 pid, u32 seq, NLA_PUT(msg, NL80211_ATTR_MAC, ETH_ALEN, dst); NLA_PUT(msg, NL80211_ATTR_MPATH_NEXT_HOP, ETH_ALEN, next_hop); + NLA_PUT_U32(msg, NL80211_ATTR_GENERATION, pinfo->generation); + pinfoattr = nla_nest_start(msg, NL80211_ATTR_MPATH_INFO); if (!pinfoattr) goto nla_put_failure; @@ -3090,8 +3102,7 @@ static int nl80211_send_bss(struct sk_buff *msg, u32 pid, u32 seq, int flags, if (!hdr) return -1; - NLA_PUT_U32(msg, NL80211_ATTR_SCAN_GENERATION, - rdev->bss_generation); + NLA_PUT_U32(msg, NL80211_ATTR_GENERATION, rdev->bss_generation); NLA_PUT_U32(msg, NL80211_ATTR_IFINDEX, wdev->netdev->ifindex); bss = nla_nest_start(msg, NL80211_ATTR_BSS); diff --git a/net/wireless/scan.c b/net/wireless/scan.c index 0ccf3a07dc02..1bcb1312bd94 100644 --- a/net/wireless/scan.c +++ b/net/wireless/scan.c @@ -562,6 +562,7 @@ void cfg80211_unlink_bss(struct wiphy *wiphy, struct cfg80211_bss *pub) spin_lock_bh(&dev->bss_lock); list_del(&bss->list); + dev->bss_generation++; rb_erase(&bss->rbn, &dev->bss_tree); spin_unlock_bh(&dev->bss_lock); -- cgit v1.2.3 From c555b9b3713e05586fabe85f4e46f28859e72930 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Fri, 7 Aug 2009 16:23:43 +0200 Subject: mac80211: explain TX retry and status Add some more documentation including an example so that it's clearer what should be done for TX retries. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- include/net/mac80211.h | 15 +++++++++++++++ 1 file changed, 15 insertions(+) (limited to 'include') diff --git a/include/net/mac80211.h b/include/net/mac80211.h index e2fb5767e1fa..467eed71be22 100644 --- a/include/net/mac80211.h +++ b/include/net/mac80211.h @@ -338,6 +338,21 @@ enum mac80211_rate_control_flags { * * When used for transmit status reporting, the driver should * always report the rate along with the flags it used. + * + * &struct ieee80211_tx_info contains an array of these structs + * in the control information, and it will be filled by the rate + * control algorithm according to what should be sent. For example, + * if this array contains, in the format { , } the + * information + * { 3, 2 }, { 2, 2 }, { 1, 4 }, { -1, 0 }, { -1, 0 } + * then this means that the frame should be transmitted + * up to twice at rate 3, up to twice at rate 2, and up to four + * times at rate 1 if it doesn't get acknowledged. Say it gets + * acknowledged by the peer after the fifth attempt, the status + * information should then contain + * { 3, 2 }, { 2, 2 }, { 1, 1 }, { -1, 0 } ... + * since it was transmitted twice at rate 3, twice at rate 2 + * and once at rate 1 after which we received an acknowledgement. */ struct ieee80211_tx_rate { s8 idx; -- cgit v1.2.3 From ab5b5342fd0ba5b9a2f58a94c5d41dd074b7c48e Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Fri, 7 Aug 2009 16:28:09 +0200 Subject: mac80211: document TX powersave filter requirements This documents what's required to implement that TX powersave filter properly wrt. handling hardware queues. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- include/net/mac80211.h | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/net/mac80211.h b/include/net/mac80211.h index 467eed71be22..cd4eb20f98f5 100644 --- a/include/net/mac80211.h +++ b/include/net/mac80211.h @@ -239,7 +239,14 @@ struct ieee80211_bss_conf { * @IEEE80211_TX_CTL_AMPDU: this frame should be sent as part of an A-MPDU * @IEEE80211_TX_CTL_INJECTED: Frame was injected, internal to mac80211. * @IEEE80211_TX_STAT_TX_FILTERED: The frame was not transmitted - * because the destination STA was in powersave mode. + * because the destination STA was in powersave mode. Note that to + * avoid race conditions, the filter must be set by the hardware or + * firmware upon receiving a frame that indicates that the station + * went to sleep (must be done on device to filter frames already on + * the queue) and may only be unset after mac80211 gives the OK for + * that by setting the IEEE80211_TX_CTL_CLEAR_PS_FILT (see above), + * since only then is it guaranteed that no more frames are in the + * hardware queue. * @IEEE80211_TX_STAT_ACK: Frame was acknowledged * @IEEE80211_TX_STAT_AMPDU: The frame was aggregated, so status * is for the whole aggregation. -- cgit v1.2.3 From ad5351db89681515681c5d5659ddf4c69e3cc6f5 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Fri, 7 Aug 2009 16:42:15 +0200 Subject: mac80211: allow DMA optimisation If we have a lot of frames to transmit at once, for instance with fragmentation, it can be an optimisation to only tell the DMA engine about them on the last fragment/frame to avoid banging the IO too much. This patch allows implementation such an optimisation by telling the driver when more frames can be expected. Currently, this is used by mac80211 only on fragmented frames, but could also be used in the future on other frames when the queue was full and there are multiple frames pending. Note that drivers need to be careful when using this flag, they need to kick their DMA engines not just when this flag is clear, but also when the queue gets full so that progress can be made. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- include/net/mac80211.h | 5 +++++ net/mac80211/tx.c | 3 +++ 2 files changed, 8 insertions(+) (limited to 'include') diff --git a/include/net/mac80211.h b/include/net/mac80211.h index cd4eb20f98f5..76d43e12cc29 100644 --- a/include/net/mac80211.h +++ b/include/net/mac80211.h @@ -268,6 +268,10 @@ struct ieee80211_bss_conf { * @IEEE80211_TX_CTL_PSPOLL_RESPONSE: (internal?) * This frame is a response to a PS-poll frame and should be sent * although the station is in powersave mode. + * @IEEE80211_TX_CTL_MORE_FRAMES: More frames will be passed to the + * transmit function after the current frame, this can be used + * by drivers to kick the DMA queue only if unset or when the + * queue gets full. */ enum mac80211_tx_control_flags { IEEE80211_TX_CTL_REQ_TX_STATUS = BIT(0), @@ -288,6 +292,7 @@ enum mac80211_tx_control_flags { IEEE80211_TX_INTFL_RETRIED = BIT(15), IEEE80211_TX_INTFL_DONT_ENCRYPT = BIT(16), IEEE80211_TX_CTL_PSPOLL_RESPONSE = BIT(17), + IEEE80211_TX_CTL_MORE_FRAMES = BIT(18), }; /** diff --git a/net/mac80211/tx.c b/net/mac80211/tx.c index 7cffaa046b33..7f2e4cdb8904 100644 --- a/net/mac80211/tx.c +++ b/net/mac80211/tx.c @@ -1154,6 +1154,9 @@ static int __ieee80211_tx(struct ieee80211_local *local, next = skb->next; len = skb->len; + if (next) + info->flags |= IEEE80211_TX_CTL_MORE_FRAMES; + sdata = vif_to_sdata(info->control.vif); switch (sdata->vif.type) { -- cgit v1.2.3 From 5ba63533bbf653631faab60f6988506160ec6ba4 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Fri, 7 Aug 2009 17:54:07 +0200 Subject: cfg80211: fix alignment problem in scan request The memory layout for scan requests was rather wrong, we put the scan SSIDs before the channels which could lead to the channel pointers being unaligned in memory. It turns out that using a pointer to the channel array isn't necessary anyway since we can embed a zero-length array into the struct. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- include/net/cfg80211.h | 4 +++- net/mac80211/ieee80211_i.h | 2 +- net/mac80211/main.c | 16 ++++++++-------- net/mac80211/scan.c | 10 +++++----- net/wireless/nl80211.c | 3 +-- net/wireless/scan.c | 4 ++-- net/wireless/sme.c | 3 +-- 7 files changed, 21 insertions(+), 21 deletions(-) (limited to 'include') diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index de7d116acc3d..d5756c9fe3d3 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -559,7 +559,6 @@ struct cfg80211_ssid { struct cfg80211_scan_request { struct cfg80211_ssid *ssids; int n_ssids; - struct ieee80211_channel **channels; u32 n_channels; const u8 *ie; size_t ie_len; @@ -568,6 +567,9 @@ struct cfg80211_scan_request { struct wiphy *wiphy; struct net_device *dev; bool aborted; + + /* keep last */ + struct ieee80211_channel *channels[0]; }; /** diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h index 99433222bc5c..d6bd7dd77960 100644 --- a/net/mac80211/ieee80211_i.h +++ b/net/mac80211/ieee80211_i.h @@ -715,7 +715,7 @@ struct ieee80211_local { struct mutex scan_mtx; unsigned long scanning; struct cfg80211_ssid scan_ssid; - struct cfg80211_scan_request int_scan_req; + struct cfg80211_scan_request *int_scan_req; struct cfg80211_scan_request *scan_req; struct ieee80211_channel *scan_channel; const u8 *orig_ies; diff --git a/net/mac80211/main.c b/net/mac80211/main.c index 0c4f8e122ed6..b03fd84777fa 100644 --- a/net/mac80211/main.c +++ b/net/mac80211/main.c @@ -765,9 +765,9 @@ int ieee80211_register_hw(struct ieee80211_hw *hw) supp_ht = supp_ht || sband->ht_cap.ht_supported; } - local->int_scan_req.n_channels = channels; - local->int_scan_req.channels = kzalloc(sizeof(void *) * channels, GFP_KERNEL); - if (!local->int_scan_req.channels) + local->int_scan_req = kzalloc(sizeof(*local->int_scan_req) + + sizeof(void *) * channels, GFP_KERNEL); + if (!local->int_scan_req) return -ENOMEM; /* if low-level driver supports AP, we also support VLAN */ @@ -882,13 +882,13 @@ int ieee80211_register_hw(struct ieee80211_hw *hw) /* alloc internal scan request */ i = 0; - local->int_scan_req.ssids = &local->scan_ssid; - local->int_scan_req.n_ssids = 1; + local->int_scan_req->ssids = &local->scan_ssid; + local->int_scan_req->n_ssids = 1; for (band = 0; band < IEEE80211_NUM_BANDS; band++) { if (!hw->wiphy->bands[band]) continue; for (j = 0; j < hw->wiphy->bands[band]->n_channels; j++) { - local->int_scan_req.channels[i] = + local->int_scan_req->channels[i] = &hw->wiphy->bands[band]->channels[j]; i++; } @@ -920,7 +920,7 @@ int ieee80211_register_hw(struct ieee80211_hw *hw) fail_workqueue: wiphy_unregister(local->hw.wiphy); fail_wiphy_register: - kfree(local->int_scan_req.channels); + kfree(local->int_scan_req->channels); return result; } EXPORT_SYMBOL(ieee80211_register_hw); @@ -962,7 +962,7 @@ void ieee80211_unregister_hw(struct ieee80211_hw *hw) wiphy_unregister(local->hw.wiphy); ieee80211_wep_free(local); ieee80211_led_exit(local); - kfree(local->int_scan_req.channels); + kfree(local->int_scan_req); } EXPORT_SYMBOL(ieee80211_unregister_hw); diff --git a/net/mac80211/scan.c b/net/mac80211/scan.c index 244f53f3c8b4..e091cbc3434f 100644 --- a/net/mac80211/scan.c +++ b/net/mac80211/scan.c @@ -277,7 +277,7 @@ void ieee80211_scan_completed(struct ieee80211_hw *hw, bool aborted) if (test_bit(SCAN_HW_SCANNING, &local->scanning)) ieee80211_restore_scan_ies(local); - if (local->scan_req != &local->int_scan_req) + if (local->scan_req != local->int_scan_req) cfg80211_scan_done(local->scan_req, aborted); local->scan_req = NULL; @@ -423,7 +423,7 @@ static int __ieee80211_start_scan(struct ieee80211_sub_if_data *sdata, local->scan_req = req; local->scan_sdata = sdata; - if (req != &local->int_scan_req && + if (req != local->int_scan_req && sdata->vif.type == NL80211_IFTYPE_STATION && !list_empty(&ifmgd->work_list)) { /* actually wait for the work it's doing to finish/time out */ @@ -743,10 +743,10 @@ int ieee80211_request_internal_scan(struct ieee80211_sub_if_data *sdata, if (local->scan_req) goto unlock; - memcpy(local->int_scan_req.ssids[0].ssid, ssid, IEEE80211_MAX_SSID_LEN); - local->int_scan_req.ssids[0].ssid_len = ssid_len; + memcpy(local->int_scan_req->ssids[0].ssid, ssid, IEEE80211_MAX_SSID_LEN); + local->int_scan_req->ssids[0].ssid_len = ssid_len; - ret = __ieee80211_start_scan(sdata, &sdata->local->int_scan_req); + ret = __ieee80211_start_scan(sdata, sdata->local->int_scan_req); unlock: mutex_unlock(&local->scan_mtx); return ret; diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index b3d5c1df08dd..667a87d307da 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -3002,10 +3002,9 @@ static int nl80211_trigger_scan(struct sk_buff *skb, struct genl_info *info) goto out; } - request->channels = (void *)((char *)request + sizeof(*request)); request->n_channels = n_channels; if (n_ssids) - request->ssids = (void *)(request->channels + n_channels); + request->ssids = (void *)&request->channels[n_channels]; request->n_ssids = n_ssids; if (ie_len) { if (request->ssids) diff --git a/net/wireless/scan.c b/net/wireless/scan.c index 1bcb1312bd94..e6c1f11595da 100644 --- a/net/wireless/scan.c +++ b/net/wireless/scan.c @@ -612,8 +612,8 @@ int cfg80211_wext_siwscan(struct net_device *dev, creq->wiphy = wiphy; creq->dev = dev; - creq->ssids = (void *)(creq + 1); - creq->channels = (void *)(creq->ssids + 1); + /* SSIDs come after channels */ + creq->ssids = (void *)&creq->channels[n_channels]; creq->n_channels = n_channels; creq->n_ssids = 1; diff --git a/net/wireless/sme.c b/net/wireless/sme.c index 104b33e34d22..8e2ef54ea714 100644 --- a/net/wireless/sme.c +++ b/net/wireless/sme.c @@ -65,7 +65,6 @@ static int cfg80211_conn_scan(struct wireless_dev *wdev) if (!request) return -ENOMEM; - request->channels = (void *)((char *)request + sizeof(*request)); if (wdev->conn->params.channel) request->channels[0] = wdev->conn->params.channel; else { @@ -82,7 +81,7 @@ static int cfg80211_conn_scan(struct wireless_dev *wdev) } } request->n_channels = n_channels; - request->ssids = (void *)(request->channels + n_channels); + request->ssids = (void *)&request->channels[n_channels]; request->n_ssids = 1; memcpy(request->ssids[0].ssid, wdev->conn->params.ssid, -- cgit v1.2.3 From d5b96a6f39a8aaa7534069b3db71048df44f023b Mon Sep 17 00:00:00 2001 From: Pat Erley Date: Sat, 8 Aug 2009 17:53:19 -0400 Subject: mac80211: remove max_bandwidth This removes the max_bandwidth attribute. It is only ever written to, and is duplicated by max_bandwidth_khz in the regulatory code. Signed-off-by: Pat Erley Acked-by: Johannes Berg Signed-off-by: John W. Linville --- include/net/cfg80211.h | 2 -- net/wireless/reg.c | 3 --- 2 files changed, 5 deletions(-) (limited to 'include') diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index d5756c9fe3d3..223913434e51 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -80,7 +80,6 @@ enum ieee80211_channel_flags { * with cfg80211. * * @center_freq: center frequency in MHz - * @max_bandwidth: maximum allowed bandwidth for this channel, in MHz * @hw_value: hardware-specific value for the channel * @flags: channel flags from &enum ieee80211_channel_flags. * @orig_flags: channel flags at registration time, used by regulatory @@ -97,7 +96,6 @@ enum ieee80211_channel_flags { struct ieee80211_channel { enum ieee80211_band band; u16 center_freq; - u8 max_bandwidth; u16 hw_value; u32 flags; int max_antenna_gain; diff --git a/net/wireless/reg.c b/net/wireless/reg.c index 0f61ae613f3b..fc7a4849c990 100644 --- a/net/wireless/reg.c +++ b/net/wireless/reg.c @@ -1018,7 +1018,6 @@ static void handle_channel(struct wiphy *wiphy, enum ieee80211_band band, map_regdom_flags(reg_rule->flags) | bw_flags; chan->max_antenna_gain = chan->orig_mag = (int) MBI_TO_DBI(power_rule->max_antenna_gain); - chan->max_bandwidth = KHZ_TO_MHZ(desired_bw_khz); chan->max_power = chan->orig_mpwr = (int) MBM_TO_DBM(power_rule->max_eirp); return; @@ -1027,7 +1026,6 @@ static void handle_channel(struct wiphy *wiphy, enum ieee80211_band band, chan->flags = flags | bw_flags | map_regdom_flags(reg_rule->flags); chan->max_antenna_gain = min(chan->orig_mag, (int) MBI_TO_DBI(power_rule->max_antenna_gain)); - chan->max_bandwidth = KHZ_TO_MHZ(desired_bw_khz); if (chan->orig_mpwr) chan->max_power = min(chan->orig_mpwr, (int) MBM_TO_DBM(power_rule->max_eirp)); @@ -1329,7 +1327,6 @@ static void handle_channel_custom(struct wiphy *wiphy, chan->flags |= map_regdom_flags(reg_rule->flags) | bw_flags; chan->max_antenna_gain = (int) MBI_TO_DBI(power_rule->max_antenna_gain); - chan->max_bandwidth = KHZ_TO_MHZ(desired_bw_khz); chan->max_power = (int) MBM_TO_DBM(power_rule->max_eirp); } -- cgit v1.2.3 From f679056b2fdd4e9b7c8eb42ba447cd9646236305 Mon Sep 17 00:00:00 2001 From: Gábor Stefanik Date: Mon, 10 Aug 2009 21:23:08 +0200 Subject: ssb: Implement the remaining rev.8 SPROM vars needed for LP-PHY MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Also add a "SPEX32" macro for extracting 32-bit SPROM variables. Signed-off-by: Gábor Stefanik Signed-off-by: John W. Linville --- drivers/ssb/pci.c | 53 ++++++++++++++++++++++++++++++++++- include/linux/ssb/ssb.h | 44 ++++++++++++++++++++++++----- include/linux/ssb/ssb_regs.h | 66 +++++++++++++++++++++++++++++++++++++++----- 3 files changed, 148 insertions(+), 15 deletions(-) (limited to 'include') diff --git a/drivers/ssb/pci.c b/drivers/ssb/pci.c index 40ea41762247..593fc618a2ea 100644 --- a/drivers/ssb/pci.c +++ b/drivers/ssb/pci.c @@ -169,8 +169,14 @@ err_pci: /* Get the word-offset for a SSB_SPROM_XXX define. */ #define SPOFF(offset) (((offset) - SSB_SPROM_BASE) / sizeof(u16)) /* Helper to extract some _offset, which is one of the SSB_SPROM_XXX defines. */ -#define SPEX(_outvar, _offset, _mask, _shift) \ +#define SPEX16(_outvar, _offset, _mask, _shift) \ out->_outvar = ((in[SPOFF(_offset)] & (_mask)) >> (_shift)) +#define SPEX32(_outvar, _offset, _mask, _shift) \ + out->_outvar = ((((u32)in[SPOFF((_offset)+2)] << 16 | \ + in[SPOFF(_offset)]) & (_mask)) >> (_shift)) +#define SPEX(_outvar, _offset, _mask, _shift) \ + SPEX16(_outvar, _offset, _mask, _shift) + static inline u8 ssb_crc8(u8 crc, u8 data) { @@ -480,6 +486,8 @@ static void sprom_extract_r8(struct ssb_sprom *out, const u16 *in) SPEX(country_code, SSB_SPROM8_CCODE, 0xFFFF, 0); SPEX(boardflags_lo, SSB_SPROM8_BFLLO, 0xFFFF, 0); SPEX(boardflags_hi, SSB_SPROM8_BFLHI, 0xFFFF, 0); + SPEX(boardflags2_lo, SSB_SPROM8_BFL2LO, 0xFFFF, 0); + SPEX(boardflags2_hi, SSB_SPROM8_BFL2HI, 0xFFFF, 0); SPEX(ant_available_a, SSB_SPROM8_ANTAVAIL, SSB_SPROM8_ANTAVAIL_A, SSB_SPROM8_ANTAVAIL_A_SHIFT); SPEX(ant_available_bg, SSB_SPROM8_ANTAVAIL, SSB_SPROM8_ANTAVAIL_BG, @@ -490,12 +498,55 @@ static void sprom_extract_r8(struct ssb_sprom *out, const u16 *in) SPEX(maxpwr_a, SSB_SPROM8_MAXP_A, SSB_SPROM8_MAXP_A_MASK, 0); SPEX(itssi_a, SSB_SPROM8_MAXP_A, SSB_SPROM8_ITSSI_A, SSB_SPROM8_ITSSI_A_SHIFT); + SPEX(maxpwr_ah, SSB_SPROM8_MAXP_AHL, SSB_SPROM8_MAXP_AH_MASK, 0); + SPEX(maxpwr_al, SSB_SPROM8_MAXP_AHL, SSB_SPROM8_MAXP_AL_MASK, + SSB_SPROM8_MAXP_AL_SHIFT); SPEX(gpio0, SSB_SPROM8_GPIOA, SSB_SPROM8_GPIOA_P0, 0); SPEX(gpio1, SSB_SPROM8_GPIOA, SSB_SPROM8_GPIOA_P1, SSB_SPROM8_GPIOA_P1_SHIFT); SPEX(gpio2, SSB_SPROM8_GPIOB, SSB_SPROM8_GPIOB_P2, 0); SPEX(gpio3, SSB_SPROM8_GPIOB, SSB_SPROM8_GPIOB_P3, SSB_SPROM8_GPIOB_P3_SHIFT); + SPEX(tri2g, SSB_SPROM8_TRI25G, SSB_SPROM8_TRI2G, 0); + SPEX(tri5g, SSB_SPROM8_TRI25G, SSB_SPROM8_TRI5G, + SSB_SPROM8_TRI5G_SHIFT); + SPEX(tri5gl, SSB_SPROM8_TRI5GHL, SSB_SPROM8_TRI5GL, 0); + SPEX(tri5gh, SSB_SPROM8_TRI5GHL, SSB_SPROM8_TRI5GH, + SSB_SPROM8_TRI5GH_SHIFT); + SPEX(rxpo2g, SSB_SPROM8_RXPO, SSB_SPROM8_RXPO2G, 0); + SPEX(rxpo5g, SSB_SPROM8_RXPO, SSB_SPROM8_RXPO5G, + SSB_SPROM8_RXPO5G_SHIFT); + SPEX(rssismf2g, SSB_SPROM8_RSSIPARM2G, SSB_SPROM8_RSSISMF2G, 0); + SPEX(rssismc2g, SSB_SPROM8_RSSIPARM2G, SSB_SPROM8_RSSISMC2G, + SSB_SPROM8_RSSISMC2G_SHIFT); + SPEX(rssisav2g, SSB_SPROM8_RSSIPARM2G, SSB_SPROM8_RSSISAV2G, + SSB_SPROM8_RSSISAV2G_SHIFT); + SPEX(bxa2g, SSB_SPROM8_RSSIPARM2G, SSB_SPROM8_BXA2G, + SSB_SPROM8_BXA2G_SHIFT); + SPEX(rssismf5g, SSB_SPROM8_RSSIPARM5G, SSB_SPROM8_RSSISMF5G, 0); + SPEX(rssismc5g, SSB_SPROM8_RSSIPARM5G, SSB_SPROM8_RSSISMC5G, + SSB_SPROM8_RSSISMC5G_SHIFT); + SPEX(rssisav5g, SSB_SPROM8_RSSIPARM5G, SSB_SPROM8_RSSISAV5G, + SSB_SPROM8_RSSISAV5G_SHIFT); + SPEX(bxa5g, SSB_SPROM8_RSSIPARM5G, SSB_SPROM8_BXA5G, + SSB_SPROM8_BXA5G_SHIFT); + SPEX(pa0b0, SSB_SPROM8_PA0B0, 0xFFFF, 0); + SPEX(pa0b1, SSB_SPROM8_PA0B1, 0xFFFF, 0); + SPEX(pa0b2, SSB_SPROM8_PA0B2, 0xFFFF, 0); + SPEX(pa1b0, SSB_SPROM8_PA1B0, 0xFFFF, 0); + SPEX(pa1b1, SSB_SPROM8_PA1B1, 0xFFFF, 0); + SPEX(pa1b2, SSB_SPROM8_PA1B2, 0xFFFF, 0); + SPEX(pa1lob0, SSB_SPROM8_PA1LOB0, 0xFFFF, 0); + SPEX(pa1lob1, SSB_SPROM8_PA1LOB1, 0xFFFF, 0); + SPEX(pa1lob2, SSB_SPROM8_PA1LOB2, 0xFFFF, 0); + SPEX(pa1hib0, SSB_SPROM8_PA1HIB0, 0xFFFF, 0); + SPEX(pa1hib1, SSB_SPROM8_PA1HIB1, 0xFFFF, 0); + SPEX(pa1hib2, SSB_SPROM8_PA1HIB2, 0xFFFF, 0); + SPEX(cck2gpo, SSB_SPROM8_CCK2GPO, 0xFFFF, 0); + SPEX32(ofdm2gpo, SSB_SPROM8_OFDM2GPO, 0xFFFFFFFF, 0); + SPEX32(ofdm5glpo, SSB_SPROM8_OFDM5GLPO, 0xFFFFFFFF, 0); + SPEX32(ofdm5gpo, SSB_SPROM8_OFDM5GPO, 0xFFFFFFFF, 0); + SPEX32(ofdm5ghpo, SSB_SPROM8_OFDM5GHPO, 0xFFFFFFFF, 0); /* Extract the antenna gain values. */ SPEX(antenna_gain.ghz24.a0, SSB_SPROM8_AGAIN01, diff --git a/include/linux/ssb/ssb.h b/include/linux/ssb/ssb.h index 5ae8fa22d331..17ffc1f84d76 100644 --- a/include/linux/ssb/ssb.h +++ b/include/linux/ssb/ssb.h @@ -27,24 +27,54 @@ struct ssb_sprom { u8 et1mdcport; /* MDIO for enet1 */ u8 board_rev; /* Board revision number from SPROM. */ u8 country_code; /* Country Code */ - u8 ant_available_a; /* A-PHY antenna available bits (up to 4) */ - u8 ant_available_bg; /* B/G-PHY antenna available bits (up to 4) */ + u8 ant_available_a; /* 2GHz antenna available bits (up to 4) */ + u8 ant_available_bg; /* 5GHz antenna available bits (up to 4) */ u16 pa0b0; u16 pa0b1; u16 pa0b2; u16 pa1b0; u16 pa1b1; u16 pa1b2; + u16 pa1lob0; + u16 pa1lob1; + u16 pa1lob2; + u16 pa1hib0; + u16 pa1hib1; + u16 pa1hib2; u8 gpio0; /* GPIO pin 0 */ u8 gpio1; /* GPIO pin 1 */ u8 gpio2; /* GPIO pin 2 */ u8 gpio3; /* GPIO pin 3 */ - u16 maxpwr_a; /* A-PHY Amplifier Max Power (in dBm Q5.2) */ - u16 maxpwr_bg; /* B/G-PHY Amplifier Max Power (in dBm Q5.2) */ + u16 maxpwr_bg; /* 2.4GHz Amplifier Max Power (in dBm Q5.2) */ + u16 maxpwr_al; /* 5.2GHz Amplifier Max Power (in dBm Q5.2) */ + u16 maxpwr_a; /* 5.3GHz Amplifier Max Power (in dBm Q5.2) */ + u16 maxpwr_ah; /* 5.8GHz Amplifier Max Power (in dBm Q5.2) */ u8 itssi_a; /* Idle TSSI Target for A-PHY */ u8 itssi_bg; /* Idle TSSI Target for B/G-PHY */ - u16 boardflags_lo; /* Boardflags (low 16 bits) */ - u16 boardflags_hi; /* Boardflags (high 16 bits) */ + u8 tri2g; /* 2.4GHz TX isolation */ + u8 tri5gl; /* 5.2GHz TX isolation */ + u8 tri5g; /* 5.3GHz TX isolation */ + u8 tri5gh; /* 5.8GHz TX isolation */ + u8 rxpo2g; /* 2GHz RX power offset */ + u8 rxpo5g; /* 5GHz RX power offset */ + u8 rssisav2g; /* 2GHz RSSI params */ + u8 rssismc2g; + u8 rssismf2g; + u8 bxa2g; /* 2GHz BX arch */ + u8 rssisav5g; /* 5GHz RSSI params */ + u8 rssismc5g; + u8 rssismf5g; + u8 bxa5g; /* 5GHz BX arch */ + u16 cck2gpo; /* CCK power offset */ + u32 ofdm2gpo; /* 2.4GHz OFDM power offset */ + u32 ofdm5glpo; /* 5.2GHz OFDM power offset */ + u32 ofdm5gpo; /* 5.3GHz OFDM power offset */ + u32 ofdm5ghpo; /* 5.8GHz OFDM power offset */ + u16 boardflags_lo; /* Board flags (bits 0-15) */ + u16 boardflags_hi; /* Board flags (bits 16-31) */ + u16 boardflags2_lo; /* Board flags (bits 32-47) */ + u16 boardflags2_hi; /* Board flags (bits 48-63) */ + /* TODO store board flags in a single u64 */ /* Antenna gain values for up to 4 antennas * on each band. Values in dBm/4 (Q5.2). Negative gain means the @@ -58,7 +88,7 @@ struct ssb_sprom { } ghz5; /* 5GHz band */ } antenna_gain; - /* TODO - add any parameters needed from rev 2, 3, or 4 SPROMs */ + /* TODO - add any parameters needed from rev 2, 3, 4, 5 or 8 SPROMs */ }; /* Information about the PCB the circuitry is soldered on. */ diff --git a/include/linux/ssb/ssb_regs.h b/include/linux/ssb/ssb_regs.h index a01b982b5783..9ae9082eaeb4 100644 --- a/include/linux/ssb/ssb_regs.h +++ b/include/linux/ssb/ssb_regs.h @@ -162,7 +162,7 @@ /* SPROM shadow area. If not otherwise noted, fields are * two bytes wide. Note that the SPROM can _only_ be read - * in two-byte quantinies. + * in two-byte quantities. */ #define SSB_SPROMSIZE_WORDS 64 #define SSB_SPROMSIZE_BYTES (SSB_SPROMSIZE_WORDS * sizeof(u16)) @@ -327,8 +327,11 @@ #define SSB_SPROM5_GPIOB_P3_SHIFT 8 /* SPROM Revision 8 */ -#define SSB_SPROM8_BFLLO 0x1084 /* Boardflags (low 16 bits) */ -#define SSB_SPROM8_BFLHI 0x1086 /* Boardflags Hi */ +#define SSB_SPROM8_BOARDREV 0x1082 /* Board revision */ +#define SSB_SPROM8_BFLLO 0x1084 /* Board flags (bits 0-15) */ +#define SSB_SPROM8_BFLHI 0x1086 /* Board flags (bits 16-31) */ +#define SSB_SPROM8_BFL2LO 0x1088 /* Board flags (bits 32-47) */ +#define SSB_SPROM8_BFL2HI 0x108A /* Board flags (bits 48-63) */ #define SSB_SPROM8_IL0MAC 0x108C /* 6 byte MAC address */ #define SSB_SPROM8_CCODE 0x1092 /* 2 byte country code */ #define SSB_SPROM8_ANTAVAIL 0x109C /* Antenna available bitfields*/ @@ -354,14 +357,63 @@ #define SSB_SPROM8_GPIOB_P2 0x00FF /* Pin 2 */ #define SSB_SPROM8_GPIOB_P3 0xFF00 /* Pin 3 */ #define SSB_SPROM8_GPIOB_P3_SHIFT 8 -#define SSB_SPROM8_MAXP_BG 0x10C0 /* Max Power BG in path 1 */ -#define SSB_SPROM8_MAXP_BG_MASK 0x00FF /* Mask for Max Power BG */ +#define SSB_SPROM8_RSSIPARM2G 0x10A4 /* RSSI params for 2GHz */ +#define SSB_SPROM8_RSSISMF2G 0x000F +#define SSB_SPROM8_RSSISMC2G 0x00F0 +#define SSB_SPROM8_RSSISMC2G_SHIFT 4 +#define SSB_SPROM8_RSSISAV2G 0x0700 +#define SSB_SPROM8_RSSISAV2G_SHIFT 8 +#define SSB_SPROM8_BXA2G 0x1800 +#define SSB_SPROM8_BXA2G_SHIFT 11 +#define SSB_SPROM8_RSSIPARM5G 0x10A6 /* RSSI params for 5GHz */ +#define SSB_SPROM8_RSSISMF5G 0x000F +#define SSB_SPROM8_RSSISMC5G 0x00F0 +#define SSB_SPROM8_RSSISMC5G_SHIFT 4 +#define SSB_SPROM8_RSSISAV5G 0x0700 +#define SSB_SPROM8_RSSISAV5G_SHIFT 8 +#define SSB_SPROM8_BXA5G 0x1800 +#define SSB_SPROM8_BXA5G_SHIFT 11 +#define SSB_SPROM8_TRI25G 0x10A8 /* TX isolation 2.4&5.3GHz */ +#define SSB_SPROM8_TRI2G 0x00FF /* TX isolation 2.4GHz */ +#define SSB_SPROM8_TRI5G 0xFF00 /* TX isolation 5.3GHz */ +#define SSB_SPROM8_TRI5G_SHIFT 8 +#define SSB_SPROM8_TRI5GHL 0x10AA /* TX isolation 5.2/5.8GHz */ +#define SSB_SPROM8_TRI5GL 0x00FF /* TX isolation 5.2GHz */ +#define SSB_SPROM8_TRI5GH 0xFF00 /* TX isolation 5.8GHz */ +#define SSB_SPROM8_TRI5GH_SHIFT 8 +#define SSB_SPROM8_RXPO 0x10AC /* RX power offsets */ +#define SSB_SPROM8_RXPO2G 0x00FF /* 2GHz RX power offset */ +#define SSB_SPROM8_RXPO5G 0xFF00 /* 5GHz RX power offset */ +#define SSB_SPROM8_RXPO5G_SHIFT 8 +#define SSB_SPROM8_MAXP_BG 0x10C0 /* Max Power 2GHz in path 1 */ +#define SSB_SPROM8_MAXP_BG_MASK 0x00FF /* Mask for Max Power 2GHz */ #define SSB_SPROM8_ITSSI_BG 0xFF00 /* Mask for path 1 itssi_bg */ #define SSB_SPROM8_ITSSI_BG_SHIFT 8 -#define SSB_SPROM8_MAXP_A 0x10C8 /* Max Power A in path 1 */ -#define SSB_SPROM8_MAXP_A_MASK 0x00FF /* Mask for Max Power A */ +#define SSB_SPROM8_PA0B0 0x10C2 /* 2GHz power amp settings */ +#define SSB_SPROM8_PA0B1 0x10C4 +#define SSB_SPROM8_PA0B2 0x10C6 +#define SSB_SPROM8_MAXP_A 0x10C8 /* Max Power 5.3GHz */ +#define SSB_SPROM8_MAXP_A_MASK 0x00FF /* Mask for Max Power 5.3GHz */ #define SSB_SPROM8_ITSSI_A 0xFF00 /* Mask for path 1 itssi_a */ #define SSB_SPROM8_ITSSI_A_SHIFT 8 +#define SSB_SPROM8_MAXP_AHL 0x10CA /* Max Power 5.2/5.8GHz */ +#define SSB_SPROM8_MAXP_AH_MASK 0x00FF /* Mask for Max Power 5.8GHz */ +#define SSB_SPROM8_MAXP_AL_MASK 0xFF00 /* Mask for Max Power 5.2GHz */ +#define SSB_SPROM8_MAXP_AL_SHIFT 8 +#define SSB_SPROM8_PA1B0 0x10CC /* 5.3GHz power amp settings */ +#define SSB_SPROM8_PA1B1 0x10CE +#define SSB_SPROM8_PA1B2 0x10D0 +#define SSB_SPROM8_PA1LOB0 0x10D2 /* 5.2GHz power amp settings */ +#define SSB_SPROM8_PA1LOB1 0x10D4 +#define SSB_SPROM8_PA1LOB2 0x10D6 +#define SSB_SPROM8_PA1HIB0 0x10D8 /* 5.8GHz power amp settings */ +#define SSB_SPROM8_PA1HIB1 0x10DA +#define SSB_SPROM8_PA1HIB2 0x10DC +#define SSB_SPROM8_CCK2GPO 0x1140 /* CCK power offset */ +#define SSB_SPROM8_OFDM2GPO 0x1142 /* 2.4GHz OFDM power offset */ +#define SSB_SPROM8_OFDM5GPO 0x1146 /* 5.3GHz OFDM power offset */ +#define SSB_SPROM8_OFDM5GLPO 0x114A /* 5.2GHz OFDM power offset */ +#define SSB_SPROM8_OFDM5GHPO 0x114E /* 5.8GHz OFDM power offset */ /* Values for SSB_SPROM1_BINF_CCODE */ enum { -- cgit v1.2.3 From 7834ddbcc7a097443761b0722e8c9fb8511b95b1 Mon Sep 17 00:00:00 2001 From: Jussi Kivilinna Date: Tue, 11 Aug 2009 22:57:16 +0300 Subject: usbnet: add rx queue pausing Add rx queue pausing to usbnet. This is needed by rndis_wlan so that it can control rx queue and prevent received packets from being send forward before rndis_wlan receives and handles 'media connect'-indication. Without this establishing WPA connections is hard and fail often. [v2] - removed unneeded use of skb_clone Cc: David Brownell Signed-off-by: Jussi Kivilinna Signed-off-by: John W. Linville --- drivers/net/usb/usbnet.c | 44 ++++++++++++++++++++++++++++++++++++++- drivers/net/wireless/rndis_wlan.c | 13 +++++++++++- include/linux/usb/usbnet.h | 6 ++++++ 3 files changed, 61 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/drivers/net/usb/usbnet.c b/drivers/net/usb/usbnet.c index af1fe4696509..7d471fca2743 100644 --- a/drivers/net/usb/usbnet.c +++ b/drivers/net/usb/usbnet.c @@ -233,6 +233,11 @@ void usbnet_skb_return (struct usbnet *dev, struct sk_buff *skb) { int status; + if (test_bit(EVENT_RX_PAUSED, &dev->flags)) { + skb_queue_tail(&dev->rxq_pause, skb); + return; + } + skb->protocol = eth_type_trans (skb, dev->net); dev->net->stats.rx_packets++; dev->net->stats.rx_bytes += skb->len; @@ -525,6 +530,41 @@ static void intr_complete (struct urb *urb) deverr(dev, "intr resubmit --> %d", status); } +/*-------------------------------------------------------------------------*/ +void usbnet_pause_rx(struct usbnet *dev) +{ + set_bit(EVENT_RX_PAUSED, &dev->flags); + + if (netif_msg_rx_status(dev)) + devdbg(dev, "paused rx queue enabled"); +} +EXPORT_SYMBOL_GPL(usbnet_pause_rx); + +void usbnet_resume_rx(struct usbnet *dev) +{ + struct sk_buff *skb; + int num = 0; + + clear_bit(EVENT_RX_PAUSED, &dev->flags); + + while ((skb = skb_dequeue(&dev->rxq_pause)) != NULL) { + usbnet_skb_return(dev, skb); + num++; + } + + tasklet_schedule(&dev->bh); + + if (netif_msg_rx_status(dev)) + devdbg(dev, "paused rx queue disabled, %d skbs requeued", num); +} +EXPORT_SYMBOL_GPL(usbnet_resume_rx); + +void usbnet_purge_paused_rxq(struct usbnet *dev) +{ + skb_queue_purge(&dev->rxq_pause); +} +EXPORT_SYMBOL_GPL(usbnet_purge_paused_rxq); + /*-------------------------------------------------------------------------*/ // unlink pending rx/tx; completion handlers do all other cleanup @@ -623,6 +663,8 @@ int usbnet_stop (struct net_device *net) usb_kill_urb(dev->interrupt); + usbnet_purge_paused_rxq(dev); + /* deferred work (task, timer, softirq) must also stop. * can't flush_scheduled_work() until we drop rtnl (later), * else workers could deadlock; so make workers a NOP. @@ -1113,7 +1155,6 @@ static void usbnet_bh (unsigned long param) } - /*------------------------------------------------------------------------- * * USB Device Driver support @@ -1210,6 +1251,7 @@ usbnet_probe (struct usb_interface *udev, const struct usb_device_id *prod) skb_queue_head_init (&dev->rxq); skb_queue_head_init (&dev->txq); skb_queue_head_init (&dev->done); + skb_queue_head_init(&dev->rxq_pause); dev->bh.func = usbnet_bh; dev->bh.data = (unsigned long) dev; INIT_WORK (&dev->kevent, kevent); diff --git a/drivers/net/wireless/rndis_wlan.c b/drivers/net/wireless/rndis_wlan.c index 828dc1825bba..d42692dfbc67 100644 --- a/drivers/net/wireless/rndis_wlan.c +++ b/drivers/net/wireless/rndis_wlan.c @@ -1764,8 +1764,15 @@ static int rndis_iw_set_essid(struct net_device *dev, if (!wrqu->essid.flags || length == 0) return disassociate(usbdev, 1); - else + else { + /* Pause and purge rx queue, so we don't pass packets before + * 'media connect'-indication. + */ + usbnet_pause_rx(usbdev); + usbnet_purge_paused_rxq(usbdev); + return set_essid(usbdev, &ssid); + } } @@ -2328,6 +2335,8 @@ get_bssid: memcpy(evt.ap_addr.sa_data, bssid, ETH_ALEN); wireless_send_event(usbdev->net, SIOCGIWAP, &evt, NULL); } + + usbnet_resume_rx(usbdev); } if (test_and_clear_bit(WORK_LINK_DOWN, &priv->work_pending)) { @@ -2541,6 +2550,8 @@ static void rndis_wlan_indication(struct usbnet *usbdev, void *ind, int buflen) switch (msg->status) { case RNDIS_STATUS_MEDIA_CONNECT: + usbnet_pause_rx(usbdev); + devinfo(usbdev, "media connect"); /* queue work to avoid recursive calls into rndis_command */ diff --git a/include/linux/usb/usbnet.h b/include/linux/usb/usbnet.h index de8b4b18961b..09514252d84e 100644 --- a/include/linux/usb/usbnet.h +++ b/include/linux/usb/usbnet.h @@ -53,6 +53,7 @@ struct usbnet { struct sk_buff_head rxq; struct sk_buff_head txq; struct sk_buff_head done; + struct sk_buff_head rxq_pause; struct urb *interrupt; struct tasklet_struct bh; @@ -63,6 +64,7 @@ struct usbnet { # define EVENT_RX_MEMORY 2 # define EVENT_STS_SPLIT 3 # define EVENT_LINK_RESET 4 +# define EVENT_RX_PAUSED 5 }; static inline struct usb_driver *driver_of(struct usb_interface *intf) @@ -190,6 +192,10 @@ extern void usbnet_defer_kevent (struct usbnet *, int); extern void usbnet_skb_return (struct usbnet *, struct sk_buff *); extern void usbnet_unlink_rx_urbs(struct usbnet *); +extern void usbnet_pause_rx(struct usbnet *); +extern void usbnet_resume_rx(struct usbnet *); +extern void usbnet_purge_paused_rxq(struct usbnet *); + extern int usbnet_get_settings (struct net_device *net, struct ethtool_cmd *cmd); extern int usbnet_set_settings (struct net_device *net, struct ethtool_cmd *cmd); extern u32 usbnet_get_link (struct net_device *net); -- cgit v1.2.3 From 16cb9d42b68b339852e8914f2538ca9a2aec616c Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Wed, 12 Aug 2009 23:33:20 +0200 Subject: cfg80211: allow driver to override PS default Sometimes drivers might have a good reason to override the PS default, like iwlwifi right now where it affects RX performance significantly at this point. This will allow them to override the default, if desired, in a way that users can still change it according to their trade-off choices, not the driver's, like would happen if the driver just disabled PS completely then. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- include/net/cfg80211.h | 4 ++++ net/wireless/core.c | 4 +++- 2 files changed, 7 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index 223913434e51..0b146bb2dd14 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -1108,6 +1108,9 @@ struct cfg80211_ops { * @net: the network namespace this wiphy currently lives in * @netnsok: if set to false, do not allow changing the netns of this * wiphy at all + * @ps_default: default for powersave, will be set depending on the + * kernel's default on wiphy_new(), but can be changed by the + * driver if it has a good reason to override the default */ struct wiphy { /* assign these fields before you register the wiphy */ @@ -1123,6 +1126,7 @@ struct wiphy { bool disable_beacon_hints; bool netnsok; + bool ps_default; enum cfg80211_signal_type signal_type; diff --git a/net/wireless/core.c b/net/wireless/core.c index 35d83bedfe5b..bc99e4ec7463 100644 --- a/net/wireless/core.c +++ b/net/wireless/core.c @@ -412,6 +412,8 @@ struct wiphy *wiphy_new(const struct cfg80211_ops *ops, int sizeof_priv) rdev->wiphy.dev.class = &ieee80211_class; rdev->wiphy.dev.platform_data = rdev; + rdev->wiphy.ps_default = CONFIG_CFG80211_DEFAULT_PS_VALUE; + wiphy_net_set(&rdev->wiphy, &init_net); rdev->rfkill_ops.set_block = cfg80211_rfkill_set_block; @@ -674,7 +676,7 @@ static int cfg80211_netdev_notifier_call(struct notifier_block * nb, wdev->wext.default_key = -1; wdev->wext.default_mgmt_key = -1; wdev->wext.connect.auth_type = NL80211_AUTHTYPE_AUTOMATIC; - wdev->wext.ps = CONFIG_CFG80211_DEFAULT_PS_VALUE; + wdev->wext.ps = wdev->wiphy->ps_default; wdev->wext.ps_timeout = 100; if (rdev->ops->set_power_mgmt) if (rdev->ops->set_power_mgmt(wdev->wiphy, dev, -- cgit v1.2.3 From 9f162d2a810b4db48f7b8d7e734d0932c81ec2a1 Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Fri, 14 Aug 2009 17:18:44 +0300 Subject: sunrpc: hton -> cpu_to_be* htonl is already defined as cpu_to_be32. cpu_to_be64 has architecture specific optimized implementations. Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- include/linux/sunrpc/xdr.h | 5 ++--- net/sunrpc/xdr.c | 6 +++--- 2 files changed, 5 insertions(+), 6 deletions(-) (limited to 'include') diff --git a/include/linux/sunrpc/xdr.h b/include/linux/sunrpc/xdr.h index b99c625fddfe..f94bbdc75c4b 100644 --- a/include/linux/sunrpc/xdr.h +++ b/include/linux/sunrpc/xdr.h @@ -117,9 +117,8 @@ static inline __be32 *xdr_encode_array(__be32 *p, const void *s, unsigned int le static inline __be32 * xdr_encode_hyper(__be32 *p, __u64 val) { - *p++ = htonl(val >> 32); - *p++ = htonl(val & 0xFFFFFFFF); - return p; + *(__be64 *)p = cpu_to_be64(val); + return p + 2; } static inline __be32 * diff --git a/net/sunrpc/xdr.c b/net/sunrpc/xdr.c index 406e26de584e..0d05d2554b42 100644 --- a/net/sunrpc/xdr.c +++ b/net/sunrpc/xdr.c @@ -24,7 +24,7 @@ xdr_encode_netobj(__be32 *p, const struct xdr_netobj *obj) unsigned int quadlen = XDR_QUADLEN(obj->len); p[quadlen] = 0; /* zero trailing bytes */ - *p++ = htonl(obj->len); + *p++ = cpu_to_be32(obj->len); memcpy(p, obj->data, obj->len); return p + XDR_QUADLEN(obj->len); } @@ -83,7 +83,7 @@ EXPORT_SYMBOL_GPL(xdr_encode_opaque_fixed); */ __be32 *xdr_encode_opaque(__be32 *p, const void *ptr, unsigned int nbytes) { - *p++ = htonl(nbytes); + *p++ = cpu_to_be32(nbytes); return xdr_encode_opaque_fixed(p, ptr, nbytes); } EXPORT_SYMBOL_GPL(xdr_encode_opaque); @@ -779,7 +779,7 @@ EXPORT_SYMBOL_GPL(xdr_decode_word); int xdr_encode_word(struct xdr_buf *buf, unsigned int base, u32 obj) { - __be32 raw = htonl(obj); + __be32 raw = cpu_to_be32(obj); return write_bytes_to_xdr_buf(buf, base, &raw, sizeof(obj)); } -- cgit v1.2.3 From 98866b5abe1513cdacc011874ca045d40002eccd Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Fri, 14 Aug 2009 17:18:49 +0300 Subject: sunrpc: ntoh -> be*_to_cpu ntohl is already defined as be32_to_cpu. be64_to_cpu has architecture specific optimized implementations. Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- include/linux/sunrpc/xdr.h | 5 ++--- net/sunrpc/xdr.c | 6 +++--- 2 files changed, 5 insertions(+), 6 deletions(-) (limited to 'include') diff --git a/include/linux/sunrpc/xdr.h b/include/linux/sunrpc/xdr.h index f94bbdc75c4b..7da466ba4b0d 100644 --- a/include/linux/sunrpc/xdr.h +++ b/include/linux/sunrpc/xdr.h @@ -124,9 +124,8 @@ xdr_encode_hyper(__be32 *p, __u64 val) static inline __be32 * xdr_decode_hyper(__be32 *p, __u64 *valp) { - *valp = ((__u64) ntohl(*p++)) << 32; - *valp |= ntohl(*p++); - return p; + *valp = be64_to_cpup((__be64 *)p); + return p + 2; } /* diff --git a/net/sunrpc/xdr.c b/net/sunrpc/xdr.c index 0d05d2554b42..8bd690c48b69 100644 --- a/net/sunrpc/xdr.c +++ b/net/sunrpc/xdr.c @@ -35,7 +35,7 @@ xdr_decode_netobj(__be32 *p, struct xdr_netobj *obj) { unsigned int len; - if ((len = ntohl(*p++)) > XDR_MAX_NETOBJ) + if ((len = be32_to_cpu(*p++)) > XDR_MAX_NETOBJ) return NULL; obj->len = len; obj->data = (u8 *) p; @@ -101,7 +101,7 @@ xdr_decode_string_inplace(__be32 *p, char **sp, { u32 len; - len = ntohl(*p++); + len = be32_to_cpu(*p++); if (len > maxlen) return NULL; *lenp = len; @@ -771,7 +771,7 @@ xdr_decode_word(struct xdr_buf *buf, unsigned int base, u32 *obj) status = read_bytes_from_xdr_buf(buf, base, &raw, sizeof(*obj)); if (status) return status; - *obj = ntohl(raw); + *obj = be32_to_cpu(raw); return 0; } EXPORT_SYMBOL_GPL(xdr_decode_word); -- cgit v1.2.3 From bb2af4f54ffa8245d5ce278cae9c66198bc14d8b Mon Sep 17 00:00:00 2001 From: Yi Zou Date: Fri, 14 Aug 2009 12:41:57 +0000 Subject: net: Add NETIF_F_FCOE_MTU to indicate support for a different MTU for FCoE Add NETIF_F_FCOE_MTU to indicate that the NIC can support a secondary MTU for converged traffic of LAN and Fiber Channel over Ethernet (FCoE). The MTU for FCoE is 2158 = 14 (FCoE header) + 24 (FC header) + 2112 (FC max payload) + 4 (FC CRC) + 4 (FCoE trailer). Signed-off-by: Yi Zou Signed-off-by: Jeff Kirsher Signed-off-by: David S. Miller --- include/linux/netdevice.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include') diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index 9f25ab2899de..9192cdf5bd2c 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -701,6 +701,7 @@ struct net_device /* the GSO_MASK reserves bits 16 through 23 */ #define NETIF_F_FCOE_CRC (1 << 24) /* FCoE CRC32 */ #define NETIF_F_SCTP_CSUM (1 << 25) /* SCTP checksum offload */ +#define NETIF_F_FCOE_MTU (1 << 26) /* Supports max FCoE MTU, 2158 bytes*/ /* Segmentation offload features */ #define NETIF_F_GSO_SHIFT 16 -- cgit v1.2.3 From 31089c13bcb18d2cd2a3ddfbe3a28666346f237e Mon Sep 17 00:00:00 2001 From: John Stultz Date: Fri, 14 Aug 2009 15:47:18 +0200 Subject: timekeeping: Introduce timekeeping_leap_insert Move the adjustment of xtime, wall_to_monotonic and the update of the vsyscall variables to the timekeeping code. Signed-off-by: John Stultz Signed-off-by: Martin Schwidefsky LKML-Reference: <20090814134807.609730216@de.ibm.com> Signed-off-by: Thomas Gleixner --- include/linux/time.h | 1 + kernel/time/ntp.c | 7 ++----- kernel/time/timekeeping.c | 7 +++++++ 3 files changed, 10 insertions(+), 5 deletions(-) (limited to 'include') diff --git a/include/linux/time.h b/include/linux/time.h index ea16c1a01d51..e7c844558884 100644 --- a/include/linux/time.h +++ b/include/linux/time.h @@ -147,6 +147,7 @@ extern struct timespec timespec_trunc(struct timespec t, unsigned gran); extern int timekeeping_valid_for_hres(void); extern void update_wall_time(void); extern void update_xtime_cache(u64 nsec); +extern void timekeeping_leap_insert(int leapsecond); struct tms; extern void do_sys_times(struct tms *); diff --git a/kernel/time/ntp.c b/kernel/time/ntp.c index 7fc64375ff43..4800f933910e 100644 --- a/kernel/time/ntp.c +++ b/kernel/time/ntp.c @@ -194,8 +194,7 @@ static enum hrtimer_restart ntp_leap_second(struct hrtimer *timer) case TIME_OK: break; case TIME_INS: - xtime.tv_sec--; - wall_to_monotonic.tv_sec++; + timekeeping_leap_insert(-1); time_state = TIME_OOP; printk(KERN_NOTICE "Clock: inserting leap second 23:59:60 UTC\n"); @@ -203,9 +202,8 @@ static enum hrtimer_restart ntp_leap_second(struct hrtimer *timer) res = HRTIMER_RESTART; break; case TIME_DEL: - xtime.tv_sec++; + timekeeping_leap_insert(1); time_tai--; - wall_to_monotonic.tv_sec--; time_state = TIME_WAIT; printk(KERN_NOTICE "Clock: deleting leap second 23:59:59 UTC\n"); @@ -219,7 +217,6 @@ static enum hrtimer_restart ntp_leap_second(struct hrtimer *timer) time_state = TIME_OK; break; } - update_vsyscall(&xtime, clock); write_sequnlock(&xtime_lock); diff --git a/kernel/time/timekeeping.c b/kernel/time/timekeeping.c index 02c0b2c9c674..b8b70fb545fc 100644 --- a/kernel/time/timekeeping.c +++ b/kernel/time/timekeeping.c @@ -58,6 +58,13 @@ void update_xtime_cache(u64 nsec) struct clocksource *clock; +/* must hold xtime_lock */ +void timekeeping_leap_insert(int leapsecond) +{ + xtime.tv_sec += leapsecond; + wall_to_monotonic.tv_sec -= leapsecond; + update_vsyscall(&xtime, clock); +} #ifdef CONFIG_GENERIC_TIME /** -- cgit v1.2.3 From a0f7d48bfb95a4c5172a2756dbc4b82afc8e9ae4 Mon Sep 17 00:00:00 2001 From: Martin Schwidefsky Date: Fri, 14 Aug 2009 15:47:19 +0200 Subject: timekeeping: Remove clocksource inline functions The three inline functions clocksource_read, clocksource_enable and clocksource_disable are simple wrappers of an indirect call plus the copy from and to the mult_orig value. The functions are exclusively used by the timekeeping code which has intimate knowledge of the clocksource anyway. Therefore remove the inline functions. No functional change. Signed-off-by: Martin Schwidefsky Acked-by: John Stultz Cc: Daniel Walker LKML-Reference: <20090814134807.903108946@de.ibm.com> Signed-off-by: Thomas Gleixner --- include/linux/clocksource.h | 58 --------------------------------------------- kernel/time/timekeeping.c | 41 ++++++++++++++++++++++---------- 2 files changed, 28 insertions(+), 71 deletions(-) (limited to 'include') diff --git a/include/linux/clocksource.h b/include/linux/clocksource.h index 1219be4fb42e..a1ef46f61c81 100644 --- a/include/linux/clocksource.h +++ b/include/linux/clocksource.h @@ -267,64 +267,6 @@ static inline u32 clocksource_hz2mult(u32 hz, u32 shift_constant) return (u32)tmp; } -/** - * clocksource_read: - Access the clocksource's current cycle value - * @cs: pointer to clocksource being read - * - * Uses the clocksource to return the current cycle_t value - */ -static inline cycle_t clocksource_read(struct clocksource *cs) -{ - return cs->read(cs); -} - -/** - * clocksource_enable: - enable clocksource - * @cs: pointer to clocksource - * - * Enables the specified clocksource. The clocksource callback - * function should start up the hardware and setup mult and field - * members of struct clocksource to reflect hardware capabilities. - */ -static inline int clocksource_enable(struct clocksource *cs) -{ - int ret = 0; - - if (cs->enable) - ret = cs->enable(cs); - - /* - * The frequency may have changed while the clocksource - * was disabled. If so the code in ->enable() must update - * the mult value to reflect the new frequency. Make sure - * mult_orig follows this change. - */ - cs->mult_orig = cs->mult; - - return ret; -} - -/** - * clocksource_disable: - disable clocksource - * @cs: pointer to clocksource - * - * Disables the specified clocksource. The clocksource callback - * function should power down the now unused hardware block to - * save power. - */ -static inline void clocksource_disable(struct clocksource *cs) -{ - /* - * Save mult_orig in mult so clocksource_enable() can - * restore the value regardless if ->enable() updates - * the value of mult or not. - */ - cs->mult = cs->mult_orig; - - if (cs->disable) - cs->disable(cs); -} - /** * cyc2ns - converts clocksource cycles to nanoseconds * @cs: Pointer to clocksource diff --git a/kernel/time/timekeeping.c b/kernel/time/timekeeping.c index b8b70fb545fc..016a2591d719 100644 --- a/kernel/time/timekeeping.c +++ b/kernel/time/timekeeping.c @@ -79,7 +79,7 @@ static void clocksource_forward_now(void) cycle_t cycle_now, cycle_delta; s64 nsec; - cycle_now = clocksource_read(clock); + cycle_now = clock->read(clock); cycle_delta = (cycle_now - clock->cycle_last) & clock->mask; clock->cycle_last = cycle_now; @@ -114,7 +114,7 @@ void getnstimeofday(struct timespec *ts) *ts = xtime; /* read clocksource: */ - cycle_now = clocksource_read(clock); + cycle_now = clock->read(clock); /* calculate the delta since the last update_wall_time: */ cycle_delta = (cycle_now - clock->cycle_last) & clock->mask; @@ -146,7 +146,7 @@ ktime_t ktime_get(void) nsecs = xtime.tv_nsec + wall_to_monotonic.tv_nsec; /* read clocksource: */ - cycle_now = clocksource_read(clock); + cycle_now = clock->read(clock); /* calculate the delta since the last update_wall_time: */ cycle_delta = (cycle_now - clock->cycle_last) & clock->mask; @@ -186,7 +186,7 @@ void ktime_get_ts(struct timespec *ts) tomono = wall_to_monotonic; /* read clocksource: */ - cycle_now = clocksource_read(clock); + cycle_now = clock->read(clock); /* calculate the delta since the last update_wall_time: */ cycle_delta = (cycle_now - clock->cycle_last) & clock->mask; @@ -274,16 +274,29 @@ static void change_clocksource(void) clocksource_forward_now(); - if (clocksource_enable(new)) + if (new->enable && !new->enable(new)) return; + /* + * The frequency may have changed while the clocksource + * was disabled. If so the code in ->enable() must update + * the mult value to reflect the new frequency. Make sure + * mult_orig follows this change. + */ + new->mult_orig = new->mult; new->raw_time = clock->raw_time; old = clock; clock = new; - clocksource_disable(old); + /* + * Save mult_orig in mult so that the value can be restored + * regardless if ->enable() updates the value of mult or not. + */ + old->mult = old->mult_orig; + if (old->disable) + old->disable(old); clock->cycle_last = 0; - clock->cycle_last = clocksource_read(clock); + clock->cycle_last = clock->read(clock); clock->error = 0; clock->xtime_nsec = 0; clocksource_calculate_interval(clock, NTP_INTERVAL_LENGTH); @@ -373,7 +386,7 @@ void getrawmonotonic(struct timespec *ts) seq = read_seqbegin(&xtime_lock); /* read clocksource: */ - cycle_now = clocksource_read(clock); + cycle_now = clock->read(clock); /* calculate the delta since the last update_wall_time: */ cycle_delta = (cycle_now - clock->cycle_last) & clock->mask; @@ -435,9 +448,12 @@ void __init timekeeping_init(void) ntp_init(); clock = clocksource_get_next(); - clocksource_enable(clock); + if (clock->enable) + clock->enable(clock); + /* set mult_orig on enable */ + clock->mult_orig = clock->mult; clocksource_calculate_interval(clock, NTP_INTERVAL_LENGTH); - clock->cycle_last = clocksource_read(clock); + clock->cycle_last = clock->read(clock); xtime.tv_sec = sec; xtime.tv_nsec = 0; @@ -477,8 +493,7 @@ static int timekeeping_resume(struct sys_device *dev) } update_xtime_cache(0); /* re-base the last cycle value */ - clock->cycle_last = 0; - clock->cycle_last = clocksource_read(clock); + clock->cycle_last = clock->read(clock); clock->error = 0; timekeeping_suspended = 0; write_sequnlock_irqrestore(&xtime_lock, flags); @@ -630,7 +645,7 @@ void update_wall_time(void) return; #ifdef CONFIG_GENERIC_TIME - offset = (clocksource_read(clock) - clock->cycle_last) & clock->mask; + offset = (clock->read(clock) - clock->cycle_last) & clock->mask; #else offset = clock->cycle_interval; #endif -- cgit v1.2.3 From f1b82746c1e93daf24e1ab9bfbd39bcdb2e7018b Mon Sep 17 00:00:00 2001 From: Martin Schwidefsky Date: Fri, 14 Aug 2009 15:47:21 +0200 Subject: clocksource: Cleanup clocksource selection If a non high-resolution clocksource is first set as override clock and then registered it becomes active even if the system is in one-shot mode. Move the override check from sysfs_override_clocksource to the clocksource selection. That fixes the bug and simplifies the code. The check in clocksource_register for double registration of the same clocksource is removed without replacement. To find the initial clocksource a new weak function in jiffies.c is defined that returns the jiffies clocksource. The architecture code can then override the weak function with a more suitable clocksource, e.g. the TOD clock on s390. [ tglx: Folded in a fix from John Stultz ] Signed-off-by: Martin Schwidefsky Acked-by: John Stultz Cc: Daniel Walker LKML-Reference: <20090814134808.388024160@de.ibm.com> Signed-off-by: Thomas Gleixner --- arch/s390/kernel/time.c | 4 ++ include/linux/clocksource.h | 2 + kernel/time/clocksource.c | 134 +++++++++++++++++--------------------------- kernel/time/jiffies.c | 6 +- kernel/time/timekeeping.c | 4 +- 5 files changed, 64 insertions(+), 86 deletions(-) (limited to 'include') diff --git a/arch/s390/kernel/time.c b/arch/s390/kernel/time.c index d4c8e9c47c81..afefe514df0f 100644 --- a/arch/s390/kernel/time.c +++ b/arch/s390/kernel/time.c @@ -205,6 +205,10 @@ static struct clocksource clocksource_tod = { .flags = CLOCK_SOURCE_IS_CONTINUOUS, }; +struct clocksource * __init clocksource_default_clock(void) +{ + return &clocksource_tod; +} void update_vsyscall(struct timespec *wall_time, struct clocksource *clock) { diff --git a/include/linux/clocksource.h b/include/linux/clocksource.h index a1ef46f61c81..f263b3abf46e 100644 --- a/include/linux/clocksource.h +++ b/include/linux/clocksource.h @@ -14,6 +14,7 @@ #include #include #include +#include #include #include @@ -322,6 +323,7 @@ extern void clocksource_touch_watchdog(void); extern struct clocksource* clocksource_get_next(void); extern void clocksource_change_rating(struct clocksource *cs, int rating); extern void clocksource_resume(void); +extern struct clocksource * __init __weak clocksource_default_clock(void); #ifdef CONFIG_GENERIC_TIME_VSYSCALL extern void update_vsyscall(struct timespec *ts, struct clocksource *c); diff --git a/kernel/time/clocksource.c b/kernel/time/clocksource.c index 7466cb811251..e91662e87cde 100644 --- a/kernel/time/clocksource.c +++ b/kernel/time/clocksource.c @@ -21,7 +21,6 @@ * * TODO WishList: * o Allow clocksource drivers to be unregistered - * o get rid of clocksource_jiffies extern */ #include @@ -107,12 +106,9 @@ u64 timecounter_cyc2time(struct timecounter *tc, } EXPORT_SYMBOL(timecounter_cyc2time); -/* XXX - Would like a better way for initializing curr_clocksource */ -extern struct clocksource clocksource_jiffies; - /*[Clocksource internal variables]--------- * curr_clocksource: - * currently selected clocksource. Initialized to clocksource_jiffies. + * currently selected clocksource. * next_clocksource: * pending next selected clocksource. * clocksource_list: @@ -123,9 +119,8 @@ extern struct clocksource clocksource_jiffies; * override_name: * Name of the user-specified clocksource. */ -static struct clocksource *curr_clocksource = &clocksource_jiffies; +static struct clocksource *curr_clocksource; static struct clocksource *next_clocksource; -static struct clocksource *clocksource_override; static LIST_HEAD(clocksource_list); static DEFINE_SPINLOCK(clocksource_lock); static char override_name[32]; @@ -320,6 +315,7 @@ void clocksource_touch_watchdog(void) clocksource_resume_watchdog(); } +#ifdef CONFIG_GENERIC_TIME /** * clocksource_get_next - Returns the selected clocksource * @@ -339,56 +335,65 @@ struct clocksource *clocksource_get_next(void) } /** - * select_clocksource - Selects the best registered clocksource. + * clocksource_select - Select the best clocksource available * * Private function. Must hold clocksource_lock when called. * * Select the clocksource with the best rating, or the clocksource, * which is selected by userspace override. */ -static struct clocksource *select_clocksource(void) +static void clocksource_select(void) { - struct clocksource *next; + struct clocksource *best, *cs; if (list_empty(&clocksource_list)) - return NULL; + return; + /* First clocksource on the list has the best rating. */ + best = list_first_entry(&clocksource_list, struct clocksource, list); + /* Check for the override clocksource. */ + list_for_each_entry(cs, &clocksource_list, list) { + if (strcmp(cs->name, override_name) != 0) + continue; + /* + * Check to make sure we don't switch to a non-highres + * capable clocksource if the tick code is in oneshot + * mode (highres or nohz) + */ + if (!(cs->flags & CLOCK_SOURCE_VALID_FOR_HRES) && + tick_oneshot_mode_active()) { + /* Override clocksource cannot be used. */ + printk(KERN_WARNING "Override clocksource %s is not " + "HRT compatible. Cannot switch while in " + "HRT/NOHZ mode\n", cs->name); + override_name[0] = 0; + } else + /* Override clocksource can be used. */ + best = cs; + break; + } + if (curr_clocksource != best) + next_clocksource = best; +} - if (clocksource_override) - next = clocksource_override; - else - next = list_entry(clocksource_list.next, struct clocksource, - list); +#else /* CONFIG_GENERIC_TIME */ - if (next == curr_clocksource) - return NULL; +static void clocksource_select(void) { } - return next; -} +#endif /* * Enqueue the clocksource sorted by rating */ -static int clocksource_enqueue(struct clocksource *c) +static void clocksource_enqueue(struct clocksource *cs) { - struct list_head *tmp, *entry = &clocksource_list; - - list_for_each(tmp, &clocksource_list) { - struct clocksource *cs; + struct list_head *entry = &clocksource_list; + struct clocksource *tmp; - cs = list_entry(tmp, struct clocksource, list); - if (cs == c) - return -EBUSY; + list_for_each_entry(tmp, &clocksource_list, list) /* Keep track of the place, where to insert */ - if (cs->rating >= c->rating) - entry = tmp; - } - list_add(&c->list, entry); - - if (strlen(c->name) == strlen(override_name) && - !strcmp(c->name, override_name)) - clocksource_override = c; - - return 0; + if (tmp->rating >= cs->rating) + entry = &tmp->list; + list_add(&cs->list, entry); } /** @@ -397,19 +402,16 @@ static int clocksource_enqueue(struct clocksource *c) * * Returns -EBUSY if registration fails, zero otherwise. */ -int clocksource_register(struct clocksource *c) +int clocksource_register(struct clocksource *cs) { unsigned long flags; - int ret; spin_lock_irqsave(&clocksource_lock, flags); - ret = clocksource_enqueue(c); - if (!ret) - next_clocksource = select_clocksource(); + clocksource_enqueue(cs); + clocksource_select(); spin_unlock_irqrestore(&clocksource_lock, flags); - if (!ret) - clocksource_check_watchdog(c); - return ret; + clocksource_check_watchdog(cs); + return 0; } EXPORT_SYMBOL(clocksource_register); @@ -425,7 +427,7 @@ void clocksource_change_rating(struct clocksource *cs, int rating) list_del(&cs->list); cs->rating = rating; clocksource_enqueue(cs); - next_clocksource = select_clocksource(); + clocksource_select(); spin_unlock_irqrestore(&clocksource_lock, flags); } @@ -438,9 +440,7 @@ void clocksource_unregister(struct clocksource *cs) spin_lock_irqsave(&clocksource_lock, flags); list_del(&cs->list); - if (clocksource_override == cs) - clocksource_override = NULL; - next_clocksource = select_clocksource(); + clocksource_select(); spin_unlock_irqrestore(&clocksource_lock, flags); } @@ -478,9 +478,7 @@ static ssize_t sysfs_override_clocksource(struct sys_device *dev, struct sysdev_attribute *attr, const char *buf, size_t count) { - struct clocksource *ovr = NULL; size_t ret = count; - int len; /* strings from sysfs write are not 0 terminated! */ if (count >= sizeof(override_name)) @@ -495,37 +493,7 @@ static ssize_t sysfs_override_clocksource(struct sys_device *dev, if (count > 0) memcpy(override_name, buf, count); override_name[count] = 0; - - len = strlen(override_name); - if (len) { - struct clocksource *cs; - - ovr = clocksource_override; - /* try to select it: */ - list_for_each_entry(cs, &clocksource_list, list) { - if (strlen(cs->name) == len && - !strcmp(cs->name, override_name)) - ovr = cs; - } - } - - /* - * Check to make sure we don't switch to a non-highres capable - * clocksource if the tick code is in oneshot mode (highres or nohz) - */ - if (tick_oneshot_mode_active() && ovr && - !(ovr->flags & CLOCK_SOURCE_VALID_FOR_HRES)) { - printk(KERN_WARNING "%s clocksource is not HRT compatible. " - "Cannot switch while in HRT/NOHZ mode\n", ovr->name); - ovr = NULL; - override_name[0] = 0; - } - - /* Reselect, when the override name has changed */ - if (ovr != clocksource_override) { - clocksource_override = ovr; - next_clocksource = select_clocksource(); - } + clocksource_select(); spin_unlock_irq(&clocksource_lock); diff --git a/kernel/time/jiffies.c b/kernel/time/jiffies.c index c3f6c30816e3..5404a8456909 100644 --- a/kernel/time/jiffies.c +++ b/kernel/time/jiffies.c @@ -61,7 +61,6 @@ struct clocksource clocksource_jiffies = { .read = jiffies_read, .mask = 0xffffffff, /*32bits*/ .mult = NSEC_PER_JIFFY << JIFFIES_SHIFT, /* details above */ - .mult_orig = NSEC_PER_JIFFY << JIFFIES_SHIFT, .shift = JIFFIES_SHIFT, }; @@ -71,3 +70,8 @@ static int __init init_jiffies_clocksource(void) } core_initcall(init_jiffies_clocksource); + +struct clocksource * __init __weak clocksource_default_clock(void) +{ + return &clocksource_jiffies; +} diff --git a/kernel/time/timekeeping.c b/kernel/time/timekeeping.c index b5673016089f..325a9b63265a 100644 --- a/kernel/time/timekeeping.c +++ b/kernel/time/timekeeping.c @@ -269,7 +269,7 @@ static void change_clocksource(void) new = clocksource_get_next(); - if (clock == new) + if (!new || clock == new) return; clocksource_forward_now(); @@ -446,7 +446,7 @@ void __init timekeeping_init(void) ntp_init(); - clock = clocksource_get_next(); + clock = clocksource_default_clock(); if (clock->enable) clock->enable(clock); /* set mult_orig on enable */ -- cgit v1.2.3 From c55c87c892c1875deace0c8fc28787335277fdf2 Mon Sep 17 00:00:00 2001 From: Martin Schwidefsky Date: Fri, 14 Aug 2009 15:47:25 +0200 Subject: clocksource: Move watchdog downgrade to a work queue thread Move the downgrade of an unstable clocksource from the timer interrupt context into the process context of a work queue thread. This is needed to be able to do the clocksource switch with stop_machine. Signed-off-by: Martin Schwidefsky Cc: Ingo Molnar Acked-by: John Stultz Cc: Daniel Walker LKML-Reference: <20090814134809.354926067@de.ibm.com> Signed-off-by: Thomas Gleixner --- include/linux/clocksource.h | 1 + kernel/time/clocksource.c | 56 +++++++++++++++++++++++++++++++-------------- 2 files changed, 40 insertions(+), 17 deletions(-) (limited to 'include') diff --git a/include/linux/clocksource.h b/include/linux/clocksource.h index f263b3abf46e..19ad43af62d0 100644 --- a/include/linux/clocksource.h +++ b/include/linux/clocksource.h @@ -213,6 +213,7 @@ extern struct clocksource *clock; /* current clocksource */ #define CLOCK_SOURCE_WATCHDOG 0x10 #define CLOCK_SOURCE_VALID_FOR_HRES 0x20 +#define CLOCK_SOURCE_UNSTABLE 0x40 /* simplify initialization of mask field */ #define CLOCKSOURCE_MASK(bits) (cycle_t)((bits) < 64 ? ((1ULL<<(bits))-1) : -1) diff --git a/kernel/time/clocksource.c b/kernel/time/clocksource.c index 56aaa749645d..f1508019bfb4 100644 --- a/kernel/time/clocksource.c +++ b/kernel/time/clocksource.c @@ -143,10 +143,13 @@ fs_initcall(clocksource_done_booting); static LIST_HEAD(watchdog_list); static struct clocksource *watchdog; static struct timer_list watchdog_timer; +static struct work_struct watchdog_work; static DEFINE_SPINLOCK(watchdog_lock); static cycle_t watchdog_last; static int watchdog_running; +static void clocksource_watchdog_work(struct work_struct *work); + /* * Interval: 0.5sec Threshold: 0.0625s */ @@ -158,15 +161,16 @@ static void clocksource_unstable(struct clocksource *cs, int64_t delta) printk(KERN_WARNING "Clocksource %s unstable (delta = %Ld ns)\n", cs->name, delta); cs->flags &= ~(CLOCK_SOURCE_VALID_FOR_HRES | CLOCK_SOURCE_WATCHDOG); - clocksource_change_rating(cs, 0); - list_del(&cs->wd_list); + cs->flags |= CLOCK_SOURCE_UNSTABLE; + schedule_work(&watchdog_work); } static void clocksource_watchdog(unsigned long data) { - struct clocksource *cs, *tmp; + struct clocksource *cs; cycle_t csnow, wdnow; int64_t wd_nsec, cs_nsec; + int next_cpu; spin_lock(&watchdog_lock); if (!watchdog_running) @@ -176,7 +180,12 @@ static void clocksource_watchdog(unsigned long data) wd_nsec = cyc2ns(watchdog, (wdnow - watchdog_last) & watchdog->mask); watchdog_last = wdnow; - list_for_each_entry_safe(cs, tmp, &watchdog_list, wd_list) { + list_for_each_entry(cs, &watchdog_list, wd_list) { + + /* Clocksource already marked unstable? */ + if (cs->flags & CLOCK_SOURCE_UNSTABLE) + continue; + csnow = cs->read(cs); /* Clocksource initialized ? */ @@ -207,19 +216,15 @@ static void clocksource_watchdog(unsigned long data) } } - if (!list_empty(&watchdog_list)) { - /* - * Cycle through CPUs to check if the CPUs stay - * synchronized to each other. - */ - int next_cpu = cpumask_next(raw_smp_processor_id(), - cpu_online_mask); - - if (next_cpu >= nr_cpu_ids) - next_cpu = cpumask_first(cpu_online_mask); - watchdog_timer.expires += WATCHDOG_INTERVAL; - add_timer_on(&watchdog_timer, next_cpu); - } + /* + * Cycle through CPUs to check if the CPUs stay synchronized + * to each other. + */ + next_cpu = cpumask_next(raw_smp_processor_id(), cpu_online_mask); + if (next_cpu >= nr_cpu_ids) + next_cpu = cpumask_first(cpu_online_mask); + watchdog_timer.expires += WATCHDOG_INTERVAL; + add_timer_on(&watchdog_timer, next_cpu); out: spin_unlock(&watchdog_lock); } @@ -228,6 +233,7 @@ static inline void clocksource_start_watchdog(void) { if (watchdog_running || !watchdog || list_empty(&watchdog_list)) return; + INIT_WORK(&watchdog_work, clocksource_watchdog_work); init_timer(&watchdog_timer); watchdog_timer.function = clocksource_watchdog; watchdog_last = watchdog->read(watchdog); @@ -313,6 +319,22 @@ static void clocksource_dequeue_watchdog(struct clocksource *cs) spin_unlock_irqrestore(&watchdog_lock, flags); } +static void clocksource_watchdog_work(struct work_struct *work) +{ + struct clocksource *cs, *tmp; + unsigned long flags; + + spin_lock_irqsave(&watchdog_lock, flags); + list_for_each_entry_safe(cs, tmp, &watchdog_list, wd_list) + if (cs->flags & CLOCK_SOURCE_UNSTABLE) { + list_del_init(&cs->wd_list); + clocksource_change_rating(cs, 0); + } + /* Check if the watchdog timer needs to be stopped. */ + clocksource_stop_watchdog(); + spin_unlock(&watchdog_lock); +} + #else /* CONFIG_CLOCKSOURCE_WATCHDOG */ static void clocksource_enqueue_watchdog(struct clocksource *cs) -- cgit v1.2.3 From 155ec60226ae0ae2aadaa57c951a58a359331030 Mon Sep 17 00:00:00 2001 From: Martin Schwidefsky Date: Fri, 14 Aug 2009 15:47:26 +0200 Subject: timekeeping: Introduce struct timekeeper Add struct timekeeper to keep the internal values timekeeping.c needs in regard to the currently selected clock source. This moves the timekeeping intervals, xtime_nsec and the ntp error value from struct clocksource to struct timekeeper. The raw_time is removed from the clocksource as well. It gets treated like xtime as a global variable. Eventually xtime raw_time should be moved to struct timekeeper. [ tglx: minor cleanup ] Signed-off-by: Martin Schwidefsky Cc: Ingo Molnar Acked-by: John Stultz Cc: Daniel Walker LKML-Reference: <20090814134809.613209842@de.ibm.com> Signed-off-by: Thomas Gleixner --- arch/s390/kernel/time.c | 1 - include/linux/clocksource.h | 54 +--------- kernel/time/clocksource.c | 6 +- kernel/time/timekeeping.c | 235 +++++++++++++++++++++++++++++--------------- 4 files changed, 164 insertions(+), 132 deletions(-) (limited to 'include') diff --git a/arch/s390/kernel/time.c b/arch/s390/kernel/time.c index afefe514df0f..e76c2e7a8b9a 100644 --- a/arch/s390/kernel/time.c +++ b/arch/s390/kernel/time.c @@ -280,7 +280,6 @@ void __init time_init(void) now = get_clock(); tod_to_timeval(now - TOD_UNIX_EPOCH, &xtime); clocksource_tod.cycle_last = now; - clocksource_tod.raw_time = xtime; tod_to_timeval(sched_clock_base_cc - TOD_UNIX_EPOCH, &ts); set_normalized_timespec(&wall_to_monotonic, -ts.tv_sec, -ts.tv_nsec); write_sequnlock_irqrestore(&xtime_lock, flags); diff --git a/include/linux/clocksource.h b/include/linux/clocksource.h index 19ad43af62d0..e12e3095e2fb 100644 --- a/include/linux/clocksource.h +++ b/include/linux/clocksource.h @@ -155,8 +155,6 @@ extern u64 timecounter_cyc2time(struct timecounter *tc, * @flags: flags describing special properties * @vread: vsyscall based read * @resume: resume function for the clocksource, if necessary - * @cycle_interval: Used internally by timekeeping core, please ignore. - * @xtime_interval: Used internally by timekeeping core, please ignore. */ struct clocksource { /* @@ -182,19 +180,12 @@ struct clocksource { #define CLKSRC_FSYS_MMIO_SET(mmio, addr) do { } while (0) #endif - /* timekeeping specific data, ignore */ - cycle_t cycle_interval; - u64 xtime_interval; - u32 raw_interval; /* * Second part is written at each timer interrupt * Keep it in a different cache line to dirty no * more than one cache line. */ cycle_t cycle_last ____cacheline_aligned_in_smp; - u64 xtime_nsec; - s64 error; - struct timespec raw_time; #ifdef CONFIG_CLOCKSOURCE_WATCHDOG /* Watchdog related data, used by the framework */ @@ -203,8 +194,6 @@ struct clocksource { #endif }; -extern struct clocksource *clock; /* current clocksource */ - /* * Clock source flags bits:: */ @@ -270,50 +259,15 @@ static inline u32 clocksource_hz2mult(u32 hz, u32 shift_constant) } /** - * cyc2ns - converts clocksource cycles to nanoseconds - * @cs: Pointer to clocksource - * @cycles: Cycles + * clocksource_cyc2ns - converts clocksource cycles to nanoseconds * - * Uses the clocksource and ntp ajdustment to convert cycle_ts to nanoseconds. + * Converts cycles to nanoseconds, using the given mult and shift. * * XXX - This could use some mult_lxl_ll() asm optimization */ -static inline s64 cyc2ns(struct clocksource *cs, cycle_t cycles) +static inline s64 clocksource_cyc2ns(cycle_t cycles, u32 mult, u32 shift) { - u64 ret = (u64)cycles; - ret = (ret * cs->mult) >> cs->shift; - return ret; -} - -/** - * clocksource_calculate_interval - Calculates a clocksource interval struct - * - * @c: Pointer to clocksource. - * @length_nsec: Desired interval length in nanoseconds. - * - * Calculates a fixed cycle/nsec interval for a given clocksource/adjustment - * pair and interval request. - * - * Unless you're the timekeeping code, you should not be using this! - */ -static inline void clocksource_calculate_interval(struct clocksource *c, - unsigned long length_nsec) -{ - u64 tmp; - - /* Do the ns -> cycle conversion first, using original mult */ - tmp = length_nsec; - tmp <<= c->shift; - tmp += c->mult_orig/2; - do_div(tmp, c->mult_orig); - - c->cycle_interval = (cycle_t)tmp; - if (c->cycle_interval == 0) - c->cycle_interval = 1; - - /* Go back from cycles -> shifted ns, this time use ntp adjused mult */ - c->xtime_interval = (u64)c->cycle_interval * c->mult; - c->raw_interval = ((u64)c->cycle_interval * c->mult_orig) >> c->shift; + return ((u64) cycles * mult) >> shift; } diff --git a/kernel/time/clocksource.c b/kernel/time/clocksource.c index f1508019bfb4..f18c9a6bdcf4 100644 --- a/kernel/time/clocksource.c +++ b/kernel/time/clocksource.c @@ -177,7 +177,8 @@ static void clocksource_watchdog(unsigned long data) goto out; wdnow = watchdog->read(watchdog); - wd_nsec = cyc2ns(watchdog, (wdnow - watchdog_last) & watchdog->mask); + wd_nsec = clocksource_cyc2ns((wdnow - watchdog_last) & watchdog->mask, + watchdog->mult, watchdog->shift); watchdog_last = wdnow; list_for_each_entry(cs, &watchdog_list, wd_list) { @@ -196,7 +197,8 @@ static void clocksource_watchdog(unsigned long data) } /* Check the deviation from the watchdog clocksource. */ - cs_nsec = cyc2ns(cs, (csnow - cs->wd_last) & cs->mask); + cs_nsec = clocksource_cyc2ns((csnow - cs->wd_last) & + cs->mask, cs->mult, cs->shift); cs->wd_last = csnow; if (abs(cs_nsec - wd_nsec) > WATCHDOG_THRESHOLD) { clocksource_unstable(cs, cs_nsec - wd_nsec); diff --git a/kernel/time/timekeeping.c b/kernel/time/timekeeping.c index 325a9b63265a..7af45cbf6b13 100644 --- a/kernel/time/timekeeping.c +++ b/kernel/time/timekeeping.c @@ -19,6 +19,65 @@ #include #include +/* Structure holding internal timekeeping values. */ +struct timekeeper { + /* Current clocksource used for timekeeping. */ + struct clocksource *clock; + + /* Number of clock cycles in one NTP interval. */ + cycle_t cycle_interval; + /* Number of clock shifted nano seconds in one NTP interval. */ + u64 xtime_interval; + /* Raw nano seconds accumulated per NTP interval. */ + u32 raw_interval; + + /* Clock shifted nano seconds remainder not stored in xtime.tv_nsec. */ + u64 xtime_nsec; + /* Difference between accumulated time and NTP time in ntp + * shifted nano seconds. */ + s64 ntp_error; +}; + +struct timekeeper timekeeper; + +/** + * timekeeper_setup_internals - Set up internals to use clocksource clock. + * + * @clock: Pointer to clocksource. + * + * Calculates a fixed cycle/nsec interval for a given clocksource/adjustment + * pair and interval request. + * + * Unless you're the timekeeping code, you should not be using this! + */ +static void timekeeper_setup_internals(struct clocksource *clock) +{ + cycle_t interval; + u64 tmp; + + timekeeper.clock = clock; + clock->cycle_last = clock->read(clock); + + /* Do the ns -> cycle conversion first, using original mult */ + tmp = NTP_INTERVAL_LENGTH; + tmp <<= clock->shift; + tmp += clock->mult_orig/2; + do_div(tmp, clock->mult_orig); + if (tmp == 0) + tmp = 1; + + interval = (cycle_t) tmp; + timekeeper.cycle_interval = interval; + + /* Go back from cycles -> shifted ns */ + timekeeper.xtime_interval = (u64) interval * clock->mult; + timekeeper.raw_interval = + ((u64) interval * clock->mult_orig) >> clock->shift; + + timekeeper.xtime_nsec = 0; + + timekeeper.ntp_error = 0; +} /* * This read-write spinlock protects us from races in SMP while @@ -46,6 +105,11 @@ struct timespec xtime __attribute__ ((aligned (16))); struct timespec wall_to_monotonic __attribute__ ((aligned (16))); static unsigned long total_sleep_time; /* seconds */ +/* + * The raw monotonic time for the CLOCK_MONOTONIC_RAW posix clock. + */ +struct timespec raw_time; + /* flag for if timekeeping is suspended */ int __read_mostly timekeeping_suspended; @@ -56,42 +120,42 @@ void update_xtime_cache(u64 nsec) timespec_add_ns(&xtime_cache, nsec); } -struct clocksource *clock; - /* must hold xtime_lock */ void timekeeping_leap_insert(int leapsecond) { xtime.tv_sec += leapsecond; wall_to_monotonic.tv_sec -= leapsecond; - update_vsyscall(&xtime, clock); + update_vsyscall(&xtime, timekeeper.clock); } #ifdef CONFIG_GENERIC_TIME /** - * clocksource_forward_now - update clock to the current time + * timekeeping_forward_now - update clock to the current time * * Forward the current clock to update its state since the last call to * update_wall_time(). This is useful before significant clock changes, * as it avoids having to deal with this time offset explicitly. */ -static void clocksource_forward_now(void) +static void timekeeping_forward_now(void) { cycle_t cycle_now, cycle_delta; + struct clocksource *clock; s64 nsec; + clock = timekeeper.clock; cycle_now = clock->read(clock); cycle_delta = (cycle_now - clock->cycle_last) & clock->mask; clock->cycle_last = cycle_now; - nsec = cyc2ns(clock, cycle_delta); + nsec = clocksource_cyc2ns(cycle_delta, clock->mult, clock->shift); /* If arch requires, add in gettimeoffset() */ nsec += arch_gettimeoffset(); timespec_add_ns(&xtime, nsec); - nsec = ((s64)cycle_delta * clock->mult_orig) >> clock->shift; - clock->raw_time.tv_nsec += nsec; + nsec = clocksource_cyc2ns(cycle_delta, clock->mult_orig, clock->shift); + timespec_add_ns(&raw_time, nsec); } /** @@ -103,6 +167,7 @@ static void clocksource_forward_now(void) void getnstimeofday(struct timespec *ts) { cycle_t cycle_now, cycle_delta; + struct clocksource *clock; unsigned long seq; s64 nsecs; @@ -114,13 +179,15 @@ void getnstimeofday(struct timespec *ts) *ts = xtime; /* read clocksource: */ + clock = timekeeper.clock; cycle_now = clock->read(clock); /* calculate the delta since the last update_wall_time: */ cycle_delta = (cycle_now - clock->cycle_last) & clock->mask; /* convert to nanoseconds: */ - nsecs = cyc2ns(clock, cycle_delta); + nsecs = clocksource_cyc2ns(cycle_delta, clock->mult, + clock->shift); /* If arch requires, add in gettimeoffset() */ nsecs += arch_gettimeoffset(); @@ -135,6 +202,7 @@ EXPORT_SYMBOL(getnstimeofday); ktime_t ktime_get(void) { cycle_t cycle_now, cycle_delta; + struct clocksource *clock; unsigned int seq; s64 secs, nsecs; @@ -146,13 +214,15 @@ ktime_t ktime_get(void) nsecs = xtime.tv_nsec + wall_to_monotonic.tv_nsec; /* read clocksource: */ + clock = timekeeper.clock; cycle_now = clock->read(clock); /* calculate the delta since the last update_wall_time: */ cycle_delta = (cycle_now - clock->cycle_last) & clock->mask; /* convert to nanoseconds: */ - nsecs += cyc2ns(clock, cycle_delta); + nsecs += clocksource_cyc2ns(cycle_delta, clock->mult, + clock->shift); } while (read_seqretry(&xtime_lock, seq)); /* @@ -174,6 +244,7 @@ EXPORT_SYMBOL_GPL(ktime_get); void ktime_get_ts(struct timespec *ts) { cycle_t cycle_now, cycle_delta; + struct clocksource *clock; struct timespec tomono; unsigned int seq; s64 nsecs; @@ -186,13 +257,15 @@ void ktime_get_ts(struct timespec *ts) tomono = wall_to_monotonic; /* read clocksource: */ + clock = timekeeper.clock; cycle_now = clock->read(clock); /* calculate the delta since the last update_wall_time: */ cycle_delta = (cycle_now - clock->cycle_last) & clock->mask; /* convert to nanoseconds: */ - nsecs = cyc2ns(clock, cycle_delta); + nsecs = clocksource_cyc2ns(cycle_delta, clock->mult, + clock->shift); } while (read_seqretry(&xtime_lock, seq)); @@ -233,7 +306,7 @@ int do_settimeofday(struct timespec *tv) write_seqlock_irqsave(&xtime_lock, flags); - clocksource_forward_now(); + timekeeping_forward_now(); ts_delta.tv_sec = tv->tv_sec - xtime.tv_sec; ts_delta.tv_nsec = tv->tv_nsec - xtime.tv_nsec; @@ -243,10 +316,10 @@ int do_settimeofday(struct timespec *tv) update_xtime_cache(0); - clock->error = 0; + timekeeper.ntp_error = 0; ntp_clear(); - update_vsyscall(&xtime, clock); + update_vsyscall(&xtime, timekeeper.clock); write_sequnlock_irqrestore(&xtime_lock, flags); @@ -269,10 +342,10 @@ static void change_clocksource(void) new = clocksource_get_next(); - if (!new || clock == new) + if (!new || timekeeper.clock == new) return; - clocksource_forward_now(); + timekeeping_forward_now(); if (new->enable && !new->enable(new)) return; @@ -284,9 +357,9 @@ static void change_clocksource(void) */ new->mult_orig = new->mult; - new->raw_time = clock->raw_time; - old = clock; - clock = new; + old = timekeeper.clock; + timekeeper_setup_internals(new); + /* * Save mult_orig in mult so that the value can be restored * regardless if ->enable() updates the value of mult or not. @@ -295,22 +368,10 @@ static void change_clocksource(void) if (old->disable) old->disable(old); - clock->cycle_last = clock->read(clock); - clock->error = 0; - clock->xtime_nsec = 0; - clocksource_calculate_interval(clock, NTP_INTERVAL_LENGTH); - tick_clock_notify(); - - /* - * We're holding xtime lock and waking up klogd would deadlock - * us on enqueue. So no printing! - printk(KERN_INFO "Time: %s clocksource has been installed.\n", - clock->name); - */ } #else /* GENERIC_TIME */ -static inline void clocksource_forward_now(void) { } +static inline void timekeeping_forward_now(void) { } static inline void change_clocksource(void) { } /** @@ -380,20 +441,23 @@ void getrawmonotonic(struct timespec *ts) unsigned long seq; s64 nsecs; cycle_t cycle_now, cycle_delta; + struct clocksource *clock; do { seq = read_seqbegin(&xtime_lock); /* read clocksource: */ + clock = timekeeper.clock; cycle_now = clock->read(clock); /* calculate the delta since the last update_wall_time: */ cycle_delta = (cycle_now - clock->cycle_last) & clock->mask; /* convert to nanoseconds: */ - nsecs = ((s64)cycle_delta * clock->mult_orig) >> clock->shift; + nsecs = clocksource_cyc2ns(cycle_delta, clock->mult_orig, + clock->shift); - *ts = clock->raw_time; + *ts = raw_time; } while (read_seqretry(&xtime_lock, seq)); @@ -413,7 +477,7 @@ int timekeeping_valid_for_hres(void) do { seq = read_seqbegin(&xtime_lock); - ret = clock->flags & CLOCK_SOURCE_VALID_FOR_HRES; + ret = timekeeper.clock->flags & CLOCK_SOURCE_VALID_FOR_HRES; } while (read_seqretry(&xtime_lock, seq)); @@ -439,6 +503,7 @@ unsigned long __attribute__((weak)) read_persistent_clock(void) */ void __init timekeeping_init(void) { + struct clocksource *clock; unsigned long flags; unsigned long sec = read_persistent_clock(); @@ -451,11 +516,13 @@ void __init timekeeping_init(void) clock->enable(clock); /* set mult_orig on enable */ clock->mult_orig = clock->mult; - clocksource_calculate_interval(clock, NTP_INTERVAL_LENGTH); - clock->cycle_last = clock->read(clock); + + timekeeper_setup_internals(clock); xtime.tv_sec = sec; xtime.tv_nsec = 0; + raw_time.tv_sec = 0; + raw_time.tv_nsec = 0; set_normalized_timespec(&wall_to_monotonic, -xtime.tv_sec, -xtime.tv_nsec); update_xtime_cache(0); @@ -492,8 +559,8 @@ static int timekeeping_resume(struct sys_device *dev) } update_xtime_cache(0); /* re-base the last cycle value */ - clock->cycle_last = clock->read(clock); - clock->error = 0; + timekeeper.clock->cycle_last = timekeeper.clock->read(timekeeper.clock); + timekeeper.ntp_error = 0; timekeeping_suspended = 0; write_sequnlock_irqrestore(&xtime_lock, flags); @@ -514,7 +581,7 @@ static int timekeeping_suspend(struct sys_device *dev, pm_message_t state) timekeeping_suspend_time = read_persistent_clock(); write_seqlock_irqsave(&xtime_lock, flags); - clocksource_forward_now(); + timekeeping_forward_now(); timekeeping_suspended = 1; write_sequnlock_irqrestore(&xtime_lock, flags); @@ -549,7 +616,7 @@ device_initcall(timekeeping_init_device); * If the error is already larger, we look ahead even further * to compensate for late or lost adjustments. */ -static __always_inline int clocksource_bigadjust(s64 error, s64 *interval, +static __always_inline int timekeeping_bigadjust(s64 error, s64 *interval, s64 *offset) { s64 tick_error, i; @@ -565,7 +632,7 @@ static __always_inline int clocksource_bigadjust(s64 error, s64 *interval, * here. This is tuned so that an error of about 1 msec is adjusted * within about 1 sec (or 2^20 nsec in 2^SHIFT_HZ ticks). */ - error2 = clock->error >> (NTP_SCALE_SHIFT + 22 - 2 * SHIFT_HZ); + error2 = timekeeper.ntp_error >> (NTP_SCALE_SHIFT + 22 - 2 * SHIFT_HZ); error2 = abs(error2); for (look_ahead = 0; error2 > 0; look_ahead++) error2 >>= 2; @@ -574,8 +641,9 @@ static __always_inline int clocksource_bigadjust(s64 error, s64 *interval, * Now calculate the error in (1 << look_ahead) ticks, but first * remove the single look ahead already included in the error. */ - tick_error = tick_length >> (NTP_SCALE_SHIFT - clock->shift + 1); - tick_error -= clock->xtime_interval >> 1; + tick_error = tick_length >> + (NTP_SCALE_SHIFT - timekeeper.clock->shift + 1); + tick_error -= timekeeper.xtime_interval >> 1; error = ((error - tick_error) >> look_ahead) + tick_error; /* Finally calculate the adjustment shift value. */ @@ -600,18 +668,19 @@ static __always_inline int clocksource_bigadjust(s64 error, s64 *interval, * this is optimized for the most common adjustments of -1,0,1, * for other values we can do a bit more work. */ -static void clocksource_adjust(s64 offset) +static void timekeeping_adjust(s64 offset) { - s64 error, interval = clock->cycle_interval; + s64 error, interval = timekeeper.cycle_interval; int adj; - error = clock->error >> (NTP_SCALE_SHIFT - clock->shift - 1); + error = timekeeper.ntp_error >> + (NTP_SCALE_SHIFT - timekeeper.clock->shift - 1); if (error > interval) { error >>= 2; if (likely(error <= interval)) adj = 1; else - adj = clocksource_bigadjust(error, &interval, &offset); + adj = timekeeping_bigadjust(error, &interval, &offset); } else if (error < -interval) { error >>= 2; if (likely(error >= -interval)) { @@ -619,15 +688,15 @@ static void clocksource_adjust(s64 offset) interval = -interval; offset = -offset; } else - adj = clocksource_bigadjust(error, &interval, &offset); + adj = timekeeping_bigadjust(error, &interval, &offset); } else return; - clock->mult += adj; - clock->xtime_interval += interval; - clock->xtime_nsec -= offset; - clock->error -= (interval - offset) << - (NTP_SCALE_SHIFT - clock->shift); + timekeeper.clock->mult += adj; + timekeeper.xtime_interval += interval; + timekeeper.xtime_nsec -= offset; + timekeeper.ntp_error -= (interval - offset) << + (NTP_SCALE_SHIFT - timekeeper.clock->shift); } /** @@ -637,53 +706,59 @@ static void clocksource_adjust(s64 offset) */ void update_wall_time(void) { + struct clocksource *clock; cycle_t offset; + s64 nsecs; /* Make sure we're fully resumed: */ if (unlikely(timekeeping_suspended)) return; + clock = timekeeper.clock; #ifdef CONFIG_GENERIC_TIME offset = (clock->read(clock) - clock->cycle_last) & clock->mask; #else - offset = clock->cycle_interval; + offset = timekeeper.cycle_interval; #endif - clock->xtime_nsec = (s64)xtime.tv_nsec << clock->shift; + timekeeper.xtime_nsec = (s64)xtime.tv_nsec << clock->shift; /* normally this loop will run just once, however in the * case of lost or late ticks, it will accumulate correctly. */ - while (offset >= clock->cycle_interval) { + while (offset >= timekeeper.cycle_interval) { + u64 nsecps = (u64)NSEC_PER_SEC << clock->shift; + /* accumulate one interval */ - offset -= clock->cycle_interval; - clock->cycle_last += clock->cycle_interval; + offset -= timekeeper.cycle_interval; + clock->cycle_last += timekeeper.cycle_interval; - clock->xtime_nsec += clock->xtime_interval; - if (clock->xtime_nsec >= (u64)NSEC_PER_SEC << clock->shift) { - clock->xtime_nsec -= (u64)NSEC_PER_SEC << clock->shift; + timekeeper.xtime_nsec += timekeeper.xtime_interval; + if (timekeeper.xtime_nsec >= nsecps) { + timekeeper.xtime_nsec -= nsecps; xtime.tv_sec++; second_overflow(); } - clock->raw_time.tv_nsec += clock->raw_interval; - if (clock->raw_time.tv_nsec >= NSEC_PER_SEC) { - clock->raw_time.tv_nsec -= NSEC_PER_SEC; - clock->raw_time.tv_sec++; + raw_time.tv_nsec += timekeeper.raw_interval; + if (raw_time.tv_nsec >= NSEC_PER_SEC) { + raw_time.tv_nsec -= NSEC_PER_SEC; + raw_time.tv_sec++; } /* accumulate error between NTP and clock interval */ - clock->error += tick_length; - clock->error -= clock->xtime_interval << (NTP_SCALE_SHIFT - clock->shift); + timekeeper.ntp_error += tick_length; + timekeeper.ntp_error -= timekeeper.xtime_interval << + (NTP_SCALE_SHIFT - clock->shift); } /* correct the clock when NTP error is too big */ - clocksource_adjust(offset); + timekeeping_adjust(offset); /* * Since in the loop above, we accumulate any amount of time * in xtime_nsec over a second into xtime.tv_sec, its possible for * xtime_nsec to be fairly small after the loop. Further, if we're - * slightly speeding the clocksource up in clocksource_adjust(), + * slightly speeding the clocksource up in timekeeping_adjust(), * its possible the required corrective factor to xtime_nsec could * cause it to underflow. * @@ -695,24 +770,26 @@ void update_wall_time(void) * We'll correct this error next time through this function, when * xtime_nsec is not as small. */ - if (unlikely((s64)clock->xtime_nsec < 0)) { - s64 neg = -(s64)clock->xtime_nsec; - clock->xtime_nsec = 0; - clock->error += neg << (NTP_SCALE_SHIFT - clock->shift); + if (unlikely((s64)timekeeper.xtime_nsec < 0)) { + s64 neg = -(s64)timekeeper.xtime_nsec; + timekeeper.xtime_nsec = 0; + timekeeper.ntp_error += neg << (NTP_SCALE_SHIFT - clock->shift); } /* store full nanoseconds into xtime after rounding it up and * add the remainder to the error difference. */ - xtime.tv_nsec = ((s64)clock->xtime_nsec >> clock->shift) + 1; - clock->xtime_nsec -= (s64)xtime.tv_nsec << clock->shift; - clock->error += clock->xtime_nsec << (NTP_SCALE_SHIFT - clock->shift); + xtime.tv_nsec = ((s64)timekeeper.xtime_nsec >> clock->shift) + 1; + timekeeper.xtime_nsec -= (s64)xtime.tv_nsec << clock->shift; + timekeeper.ntp_error += timekeeper.xtime_nsec << + (NTP_SCALE_SHIFT - clock->shift); - update_xtime_cache(cyc2ns(clock, offset)); + nsecs = clocksource_cyc2ns(offset, clock->mult, clock->shift); + update_xtime_cache(nsecs); /* check to see if there is a new clocksource to use */ change_clocksource(); - update_vsyscall(&xtime, clock); + update_vsyscall(&xtime, timekeeper.clock); } /** -- cgit v1.2.3 From 0a54419836254a27baecd9037103171bcbabaf67 Mon Sep 17 00:00:00 2001 From: Martin Schwidefsky Date: Fri, 14 Aug 2009 15:47:28 +0200 Subject: timekeeping: Move NTP adjusted clock multiplier to struct timekeeper The clocksource structure has two multipliers, the unmodified multiplier clock->mult_orig and the NTP corrected multiplier clock->mult. The NTP multiplier is misplaced in the struct clocksource, this is private information of the timekeeping code. Add the mult field to the struct timekeeper to contain the NTP corrected value, keep the unmodifed multiplier in clock->mult and remove clock->mult_orig. Signed-off-by: Martin Schwidefsky Cc: Ingo Molnar Acked-by: John Stultz Cc: Daniel Walker LKML-Reference: <20090814134810.149047645@de.ibm.com> Signed-off-by: Thomas Gleixner --- arch/arm/plat-omap/common.c | 7 ++---- include/linux/clocksource.h | 4 +--- kernel/time/timekeeping.c | 53 ++++++++++++++++++++------------------------- 3 files changed, 27 insertions(+), 37 deletions(-) (limited to 'include') diff --git a/arch/arm/plat-omap/common.c b/arch/arm/plat-omap/common.c index ebcf006406f9..95587b6c0259 100644 --- a/arch/arm/plat-omap/common.c +++ b/arch/arm/plat-omap/common.c @@ -253,11 +253,8 @@ static struct clocksource clocksource_32k = { */ unsigned long long sched_clock(void) { - unsigned long long ret; - - ret = (unsigned long long)clocksource_32k.read(&clocksource_32k); - ret = (ret * clocksource_32k.mult_orig) >> clocksource_32k.shift; - return ret; + return clocksource_cyc2ns(clocksource_32k.read(&clocksource_32k), + clocksource_32k.mult, clocksource_32k.shift); } static int __init omap_init_clocksource_32k(void) diff --git a/include/linux/clocksource.h b/include/linux/clocksource.h index e12e3095e2fb..e34015effeb6 100644 --- a/include/linux/clocksource.h +++ b/include/linux/clocksource.h @@ -149,8 +149,7 @@ extern u64 timecounter_cyc2time(struct timecounter *tc, * @disable: optional function to disable the clocksource * @mask: bitmask for two's complement * subtraction of non 64 bit counters - * @mult: cycle to nanosecond multiplier (adjusted by NTP) - * @mult_orig: cycle to nanosecond multiplier (unadjusted by NTP) + * @mult: cycle to nanosecond multiplier * @shift: cycle to nanosecond divisor (power of two) * @flags: flags describing special properties * @vread: vsyscall based read @@ -168,7 +167,6 @@ struct clocksource { void (*disable)(struct clocksource *cs); cycle_t mask; u32 mult; - u32 mult_orig; u32 shift; unsigned long flags; cycle_t (*vread)(void); diff --git a/kernel/time/timekeeping.c b/kernel/time/timekeeping.c index dfdab1cefe1e..f4056f6c2632 100644 --- a/kernel/time/timekeeping.c +++ b/kernel/time/timekeeping.c @@ -41,6 +41,8 @@ struct timekeeper { /* Shift conversion between clock shifted nano seconds and * ntp shifted nano seconds. */ int ntp_error_shift; + /* NTP adjusted clock multiplier */ + u32 mult; }; struct timekeeper timekeeper; @@ -66,8 +68,8 @@ static void timekeeper_setup_internals(struct clocksource *clock) /* Do the ns -> cycle conversion first, using original mult */ tmp = NTP_INTERVAL_LENGTH; tmp <<= clock->shift; - tmp += clock->mult_orig/2; - do_div(tmp, clock->mult_orig); + tmp += clock->mult/2; + do_div(tmp, clock->mult); if (tmp == 0) tmp = 1; @@ -77,13 +79,20 @@ static void timekeeper_setup_internals(struct clocksource *clock) /* Go back from cycles -> shifted ns */ timekeeper.xtime_interval = (u64) interval * clock->mult; timekeeper.raw_interval = - ((u64) interval * clock->mult_orig) >> clock->shift; + ((u64) interval * clock->mult) >> clock->shift; timekeeper.xtime_nsec = 0; timekeeper.shift = clock->shift; timekeeper.ntp_error = 0; timekeeper.ntp_error_shift = NTP_SCALE_SHIFT - clock->shift; + + /* + * The timekeeper keeps its own mult values for the currently + * active clocksource. These value will be adjusted via NTP + * to counteract clock drifting. + */ + timekeeper.mult = clock->mult; } /* @@ -154,14 +163,15 @@ static void timekeeping_forward_now(void) cycle_delta = (cycle_now - clock->cycle_last) & clock->mask; clock->cycle_last = cycle_now; - nsec = clocksource_cyc2ns(cycle_delta, clock->mult, clock->shift); + nsec = clocksource_cyc2ns(cycle_delta, timekeeper.mult, + timekeeper.shift); /* If arch requires, add in gettimeoffset() */ nsec += arch_gettimeoffset(); timespec_add_ns(&xtime, nsec); - nsec = clocksource_cyc2ns(cycle_delta, clock->mult_orig, clock->shift); + nsec = clocksource_cyc2ns(cycle_delta, clock->mult, clock->shift); timespec_add_ns(&raw_time, nsec); } @@ -193,8 +203,8 @@ void getnstimeofday(struct timespec *ts) cycle_delta = (cycle_now - clock->cycle_last) & clock->mask; /* convert to nanoseconds: */ - nsecs = clocksource_cyc2ns(cycle_delta, clock->mult, - clock->shift); + nsecs = clocksource_cyc2ns(cycle_delta, timekeeper.mult, + timekeeper.shift); /* If arch requires, add in gettimeoffset() */ nsecs += arch_gettimeoffset(); @@ -228,8 +238,8 @@ ktime_t ktime_get(void) cycle_delta = (cycle_now - clock->cycle_last) & clock->mask; /* convert to nanoseconds: */ - nsecs += clocksource_cyc2ns(cycle_delta, clock->mult, - clock->shift); + nsecs += clocksource_cyc2ns(cycle_delta, timekeeper.mult, + timekeeper.shift); } while (read_seqretry(&xtime_lock, seq)); /* @@ -271,8 +281,8 @@ void ktime_get_ts(struct timespec *ts) cycle_delta = (cycle_now - clock->cycle_last) & clock->mask; /* convert to nanoseconds: */ - nsecs = clocksource_cyc2ns(cycle_delta, clock->mult, - clock->shift); + nsecs = clocksource_cyc2ns(cycle_delta, timekeeper.mult, + timekeeper.shift); } while (read_seqretry(&xtime_lock, seq)); @@ -356,22 +366,10 @@ static void change_clocksource(void) if (new->enable && !new->enable(new)) return; - /* - * The frequency may have changed while the clocksource - * was disabled. If so the code in ->enable() must update - * the mult value to reflect the new frequency. Make sure - * mult_orig follows this change. - */ - new->mult_orig = new->mult; old = timekeeper.clock; timekeeper_setup_internals(new); - /* - * Save mult_orig in mult so that the value can be restored - * regardless if ->enable() updates the value of mult or not. - */ - old->mult = old->mult_orig; if (old->disable) old->disable(old); @@ -461,7 +459,7 @@ void getrawmonotonic(struct timespec *ts) cycle_delta = (cycle_now - clock->cycle_last) & clock->mask; /* convert to nanoseconds: */ - nsecs = clocksource_cyc2ns(cycle_delta, clock->mult_orig, + nsecs = clocksource_cyc2ns(cycle_delta, clock->mult, clock->shift); *ts = raw_time; @@ -521,9 +519,6 @@ void __init timekeeping_init(void) clock = clocksource_default_clock(); if (clock->enable) clock->enable(clock); - /* set mult_orig on enable */ - clock->mult_orig = clock->mult; - timekeeper_setup_internals(clock); xtime.tv_sec = sec; @@ -697,7 +692,7 @@ static void timekeeping_adjust(s64 offset) } else return; - timekeeper.clock->mult += adj; + timekeeper.mult += adj; timekeeper.xtime_interval += interval; timekeeper.xtime_nsec -= offset; timekeeper.ntp_error -= (interval - offset) << @@ -789,7 +784,7 @@ void update_wall_time(void) timekeeper.ntp_error += timekeeper.xtime_nsec << timekeeper.ntp_error_shift; - nsecs = clocksource_cyc2ns(offset, clock->mult, clock->shift); + nsecs = clocksource_cyc2ns(offset, timekeeper.mult, timekeeper.shift); update_xtime_cache(nsecs); /* check to see if there is a new clocksource to use */ -- cgit v1.2.3 From 75c5158f70c065b9704b924503d96e8297838f79 Mon Sep 17 00:00:00 2001 From: Martin Schwidefsky Date: Fri, 14 Aug 2009 15:47:30 +0200 Subject: timekeeping: Update clocksource with stop_machine update_wall_time calls change_clocksource HZ times per second to check if a new clock source is available. In close to 100% of all calls there is no new clock. Replace the tick based check by an update done with stop_machine. Signed-off-by: Martin Schwidefsky Cc: Ingo Molnar Acked-by: John Stultz Cc: Daniel Walker LKML-Reference: <20090814134810.711836357@de.ibm.com> Signed-off-by: Thomas Gleixner --- include/linux/clocksource.h | 2 + kernel/time/clocksource.c | 112 +++++++++++++++++--------------------------- kernel/time/timekeeping.c | 41 ++++++++++------ 3 files changed, 72 insertions(+), 83 deletions(-) (limited to 'include') diff --git a/include/linux/clocksource.h b/include/linux/clocksource.h index e34015effeb6..9ea40ff26f0e 100644 --- a/include/linux/clocksource.h +++ b/include/linux/clocksource.h @@ -291,4 +291,6 @@ static inline void update_vsyscall_tz(void) } #endif +extern void timekeeping_notify(struct clocksource *clock); + #endif /* _LINUX_CLOCKSOURCE_H */ diff --git a/kernel/time/clocksource.c b/kernel/time/clocksource.c index f18c9a6bdcf4..a1657b5fdeb9 100644 --- a/kernel/time/clocksource.c +++ b/kernel/time/clocksource.c @@ -109,35 +109,17 @@ EXPORT_SYMBOL(timecounter_cyc2time); /*[Clocksource internal variables]--------- * curr_clocksource: * currently selected clocksource. - * next_clocksource: - * pending next selected clocksource. * clocksource_list: * linked list with the registered clocksources - * clocksource_lock: - * protects manipulations to curr_clocksource and next_clocksource - * and the clocksource_list + * clocksource_mutex: + * protects manipulations to curr_clocksource and the clocksource_list * override_name: * Name of the user-specified clocksource. */ static struct clocksource *curr_clocksource; -static struct clocksource *next_clocksource; static LIST_HEAD(clocksource_list); -static DEFINE_SPINLOCK(clocksource_lock); +static DEFINE_MUTEX(clocksource_mutex); static char override_name[32]; -static int finished_booting; - -/* clocksource_done_booting - Called near the end of core bootup - * - * Hack to avoid lots of clocksource churn at boot time. - * We use fs_initcall because we want this to start before - * device_initcall but after subsys_initcall. - */ -static int __init clocksource_done_booting(void) -{ - finished_booting = 1; - return 0; -} -fs_initcall(clocksource_done_booting); #ifdef CONFIG_CLOCKSOURCE_WATCHDOG static LIST_HEAD(watchdog_list); @@ -356,18 +338,16 @@ static inline void clocksource_resume_watchdog(void) { } void clocksource_resume(void) { struct clocksource *cs; - unsigned long flags; - spin_lock_irqsave(&clocksource_lock, flags); + mutex_lock(&clocksource_mutex); - list_for_each_entry(cs, &clocksource_list, list) { + list_for_each_entry(cs, &clocksource_list, list) if (cs->resume) cs->resume(); - } clocksource_resume_watchdog(); - spin_unlock_irqrestore(&clocksource_lock, flags); + mutex_unlock(&clocksource_mutex); } /** @@ -383,28 +363,13 @@ void clocksource_touch_watchdog(void) } #ifdef CONFIG_GENERIC_TIME -/** - * clocksource_get_next - Returns the selected clocksource - * - */ -struct clocksource *clocksource_get_next(void) -{ - unsigned long flags; - spin_lock_irqsave(&clocksource_lock, flags); - if (next_clocksource && finished_booting) { - curr_clocksource = next_clocksource; - next_clocksource = NULL; - } - spin_unlock_irqrestore(&clocksource_lock, flags); - - return curr_clocksource; -} +static int finished_booting; /** * clocksource_select - Select the best clocksource available * - * Private function. Must hold clocksource_lock when called. + * Private function. Must hold clocksource_mutex when called. * * Select the clocksource with the best rating, or the clocksource, * which is selected by userspace override. @@ -413,7 +378,7 @@ static void clocksource_select(void) { struct clocksource *best, *cs; - if (list_empty(&clocksource_list)) + if (!finished_booting || list_empty(&clocksource_list)) return; /* First clocksource on the list has the best rating. */ best = list_first_entry(&clocksource_list, struct clocksource, list); @@ -438,13 +403,31 @@ static void clocksource_select(void) best = cs; break; } - if (curr_clocksource != best) - next_clocksource = best; + if (curr_clocksource != best) { + printk(KERN_INFO "Switching to clocksource %s\n", best->name); + curr_clocksource = best; + timekeeping_notify(curr_clocksource); + } } +/* + * clocksource_done_booting - Called near the end of core bootup + * + * Hack to avoid lots of clocksource churn at boot time. + * We use fs_initcall because we want this to start before + * device_initcall but after subsys_initcall. + */ +static int __init clocksource_done_booting(void) +{ + finished_booting = 1; + clocksource_select(); + return 0; +} +fs_initcall(clocksource_done_booting); + #else /* CONFIG_GENERIC_TIME */ -static void clocksource_select(void) { } +static inline void clocksource_select(void) { } #endif @@ -471,13 +454,11 @@ static void clocksource_enqueue(struct clocksource *cs) */ int clocksource_register(struct clocksource *cs) { - unsigned long flags; - - spin_lock_irqsave(&clocksource_lock, flags); + mutex_lock(&clocksource_mutex); clocksource_enqueue(cs); clocksource_select(); - spin_unlock_irqrestore(&clocksource_lock, flags); clocksource_enqueue_watchdog(cs); + mutex_unlock(&clocksource_mutex); return 0; } EXPORT_SYMBOL(clocksource_register); @@ -487,14 +468,12 @@ EXPORT_SYMBOL(clocksource_register); */ void clocksource_change_rating(struct clocksource *cs, int rating) { - unsigned long flags; - - spin_lock_irqsave(&clocksource_lock, flags); + mutex_lock(&clocksource_mutex); list_del(&cs->list); cs->rating = rating; clocksource_enqueue(cs); clocksource_select(); - spin_unlock_irqrestore(&clocksource_lock, flags); + mutex_unlock(&clocksource_mutex); } EXPORT_SYMBOL(clocksource_change_rating); @@ -503,13 +482,11 @@ EXPORT_SYMBOL(clocksource_change_rating); */ void clocksource_unregister(struct clocksource *cs) { - unsigned long flags; - + mutex_lock(&clocksource_mutex); clocksource_dequeue_watchdog(cs); - spin_lock_irqsave(&clocksource_lock, flags); list_del(&cs->list); clocksource_select(); - spin_unlock_irqrestore(&clocksource_lock, flags); + mutex_unlock(&clocksource_mutex); } EXPORT_SYMBOL(clocksource_unregister); @@ -527,9 +504,9 @@ sysfs_show_current_clocksources(struct sys_device *dev, { ssize_t count = 0; - spin_lock_irq(&clocksource_lock); + mutex_lock(&clocksource_mutex); count = snprintf(buf, PAGE_SIZE, "%s\n", curr_clocksource->name); - spin_unlock_irq(&clocksource_lock); + mutex_unlock(&clocksource_mutex); return count; } @@ -557,14 +534,14 @@ static ssize_t sysfs_override_clocksource(struct sys_device *dev, if (buf[count-1] == '\n') count--; - spin_lock_irq(&clocksource_lock); + mutex_lock(&clocksource_mutex); if (count > 0) memcpy(override_name, buf, count); override_name[count] = 0; clocksource_select(); - spin_unlock_irq(&clocksource_lock); + mutex_unlock(&clocksource_mutex); return ret; } @@ -584,7 +561,7 @@ sysfs_show_available_clocksources(struct sys_device *dev, struct clocksource *src; ssize_t count = 0; - spin_lock_irq(&clocksource_lock); + mutex_lock(&clocksource_mutex); list_for_each_entry(src, &clocksource_list, list) { /* * Don't show non-HRES clocksource if the tick code is @@ -596,7 +573,7 @@ sysfs_show_available_clocksources(struct sys_device *dev, max((ssize_t)PAGE_SIZE - count, (ssize_t)0), "%s ", src->name); } - spin_unlock_irq(&clocksource_lock); + mutex_unlock(&clocksource_mutex); count += snprintf(buf + count, max((ssize_t)PAGE_SIZE - count, (ssize_t)0), "\n"); @@ -651,11 +628,10 @@ device_initcall(init_clocksource_sysfs); */ static int __init boot_override_clocksource(char* str) { - unsigned long flags; - spin_lock_irqsave(&clocksource_lock, flags); + mutex_lock(&clocksource_mutex); if (str) strlcpy(override_name, str, sizeof(override_name)); - spin_unlock_irqrestore(&clocksource_lock, flags); + mutex_unlock(&clocksource_mutex); return 1; } diff --git a/kernel/time/timekeeping.c b/kernel/time/timekeeping.c index 27ae01b596b7..41579e7fcf9d 100644 --- a/kernel/time/timekeeping.c +++ b/kernel/time/timekeeping.c @@ -18,6 +18,7 @@ #include #include #include +#include /* Structure holding internal timekeeping values. */ struct timekeeper { @@ -179,6 +180,7 @@ void timekeeping_leap_insert(int leapsecond) } #ifdef CONFIG_GENERIC_TIME + /** * timekeeping_forward_now - update clock to the current time * @@ -351,31 +353,40 @@ EXPORT_SYMBOL(do_settimeofday); * * Accumulates current time interval and initializes new clocksource */ -static void change_clocksource(void) +static int change_clocksource(void *data) { struct clocksource *new, *old; - new = clocksource_get_next(); - - if (!new || timekeeper.clock == new) - return; + new = (struct clocksource *) data; timekeeping_forward_now(); + if (!new->enable || new->enable(new) == 0) { + old = timekeeper.clock; + timekeeper_setup_internals(new); + if (old->disable) + old->disable(old); + } + return 0; +} - if (new->enable && !new->enable(new)) +/** + * timekeeping_notify - Install a new clock source + * @clock: pointer to the clock source + * + * This function is called from clocksource.c after a new, better clock + * source has been registered. The caller holds the clocksource_mutex. + */ +void timekeeping_notify(struct clocksource *clock) +{ + if (timekeeper.clock == clock) return; - - old = timekeeper.clock; - timekeeper_setup_internals(new); - - if (old->disable) - old->disable(old); - + stop_machine(change_clocksource, clock, NULL); tick_clock_notify(); } + #else /* GENERIC_TIME */ + static inline void timekeeping_forward_now(void) { } -static inline void change_clocksource(void) { } /** * ktime_get - get the monotonic time in ktime_t format @@ -416,6 +427,7 @@ void ktime_get_ts(struct timespec *ts) ts->tv_nsec + tomono.tv_nsec); } EXPORT_SYMBOL_GPL(ktime_get_ts); + #endif /* !GENERIC_TIME */ /** @@ -773,7 +785,6 @@ void update_wall_time(void) update_xtime_cache(nsecs); /* check to see if there is a new clocksource to use */ - change_clocksource(); update_vsyscall(&xtime, timekeeper.clock); } -- cgit v1.2.3 From d4f587c67fc39e0030ddd718675e252e208da4d7 Mon Sep 17 00:00:00 2001 From: Martin Schwidefsky Date: Fri, 14 Aug 2009 15:47:31 +0200 Subject: timekeeping: Increase granularity of read_persistent_clock() The persistent clock of some architectures (e.g. s390) have a better granularity than seconds. To reduce the delta between the host clock and the guest clock in a virtualized system change the read_persistent_clock function to return a struct timespec. Signed-off-by: Martin Schwidefsky Cc: Ingo Molnar Acked-by: John Stultz Cc: Daniel Walker LKML-Reference: <20090814134811.013873340@de.ibm.com> Signed-off-by: Thomas Gleixner --- arch/m68knommu/kernel/time.c | 5 ++-- arch/mips/dec/time.c | 5 ++-- arch/mips/lasat/ds1603.c | 5 ++-- arch/mips/lasat/sysctl.c | 8 ++++-- arch/mips/lemote/lm2e/setup.c | 5 ++-- arch/mips/mti-malta/malta-time.c | 5 ++-- arch/mips/pmc-sierra/yosemite/setup.c | 5 ++-- arch/mips/sibyte/swarm/setup.c | 15 +++++++--- arch/mips/sni/time.c | 5 ++-- arch/powerpc/kernel/time.c | 7 +++-- arch/s390/kernel/time.c | 22 +++------------ arch/sh/kernel/time.c | 6 ++-- arch/x86/kernel/rtc.c | 5 ++-- arch/xtensa/kernel/time.c | 5 ++-- include/linux/time.h | 2 +- kernel/time/timekeeping.c | 52 +++++++++++++++++++---------------- 16 files changed, 83 insertions(+), 74 deletions(-) (limited to 'include') diff --git a/arch/m68knommu/kernel/time.c b/arch/m68knommu/kernel/time.c index d182b2f72211..68432248515c 100644 --- a/arch/m68knommu/kernel/time.c +++ b/arch/m68knommu/kernel/time.c @@ -72,9 +72,10 @@ static unsigned long read_rtc_mmss(void) return mktime(year, mon, day, hour, min, sec);; } -unsigned long read_persistent_clock(void) +void read_persistent_clock(struct timespec *ts) { - return read_rtc_mmss(); + ts->tv_sec = read_rtc_mmss(); + ts->tv_nsec = 0; } int update_persistent_clock(struct timespec now) diff --git a/arch/mips/dec/time.c b/arch/mips/dec/time.c index 463136e6685a..02f505f23c32 100644 --- a/arch/mips/dec/time.c +++ b/arch/mips/dec/time.c @@ -18,7 +18,7 @@ #include #include -unsigned long read_persistent_clock(void) +void read_persistent_clock(struct timespec *ts) { unsigned int year, mon, day, hour, min, sec, real_year; unsigned long flags; @@ -53,7 +53,8 @@ unsigned long read_persistent_clock(void) year += real_year - 72 + 2000; - return mktime(year, mon, day, hour, min, sec); + ts->tv_sec = mktime(year, mon, day, hour, min, sec); + ts->tv_nsec = 0; } /* diff --git a/arch/mips/lasat/ds1603.c b/arch/mips/lasat/ds1603.c index 52cb1436a12a..c6fd96ff118d 100644 --- a/arch/mips/lasat/ds1603.c +++ b/arch/mips/lasat/ds1603.c @@ -135,7 +135,7 @@ static void rtc_end_op(void) lasat_ndelay(1000); } -unsigned long read_persistent_clock(void) +void read_persistent_clock(struct timespec *ts) { unsigned long word; unsigned long flags; @@ -147,7 +147,8 @@ unsigned long read_persistent_clock(void) rtc_end_op(); spin_unlock_irqrestore(&rtc_lock, flags); - return word; + ts->tv_sec = word; + ts->tv_nsec = 0; } int rtc_mips_set_mmss(unsigned long time) diff --git a/arch/mips/lasat/sysctl.c b/arch/mips/lasat/sysctl.c index 8f88886feb12..3f04d4c406b7 100644 --- a/arch/mips/lasat/sysctl.c +++ b/arch/mips/lasat/sysctl.c @@ -92,10 +92,12 @@ static int rtctmp; int proc_dolasatrtc(ctl_table *table, int write, struct file *filp, void *buffer, size_t *lenp, loff_t *ppos) { + struct timespec ts; int r; if (!write) { - rtctmp = read_persistent_clock(); + read_persistent_clock(&ts); + rtctmp = ts.tv_sec; /* check for time < 0 and set to 0 */ if (rtctmp < 0) rtctmp = 0; @@ -134,9 +136,11 @@ int sysctl_lasat_rtc(ctl_table *table, void *oldval, size_t *oldlenp, void *newval, size_t newlen) { + struct timespec ts; int r; - rtctmp = read_persistent_clock(); + read_persistent_clock(&ts); + rtctmp = ts.tv_sec; if (rtctmp < 0) rtctmp = 0; r = sysctl_intvec(table, oldval, oldlenp, newval, newlen); diff --git a/arch/mips/lemote/lm2e/setup.c b/arch/mips/lemote/lm2e/setup.c index ebd6ceaef2fd..24b355df6127 100644 --- a/arch/mips/lemote/lm2e/setup.c +++ b/arch/mips/lemote/lm2e/setup.c @@ -54,9 +54,10 @@ void __init plat_time_init(void) mips_hpt_frequency = cpu_clock_freq / 2; } -unsigned long read_persistent_clock(void) +void read_persistent_clock(struct timespec *ts) { - return mc146818_get_cmos_time(); + ts->tv_sec = mc146818_get_cmos_time(); + ts->tv_nsec = 0; } void (*__wbflush)(void); diff --git a/arch/mips/mti-malta/malta-time.c b/arch/mips/mti-malta/malta-time.c index 0b97d47691fc..3c6f190aa61c 100644 --- a/arch/mips/mti-malta/malta-time.c +++ b/arch/mips/mti-malta/malta-time.c @@ -100,9 +100,10 @@ static unsigned int __init estimate_cpu_frequency(void) return count; } -unsigned long read_persistent_clock(void) +void read_persistent_clock(struct timespec *ts) { - return mc146818_get_cmos_time(); + ts->tv_sec = mc146818_get_cmos_time(); + ts->tv_nsec = 0; } static void __init plat_perf_setup(void) diff --git a/arch/mips/pmc-sierra/yosemite/setup.c b/arch/mips/pmc-sierra/yosemite/setup.c index 2d3c0dca275d..3498ac9c35af 100644 --- a/arch/mips/pmc-sierra/yosemite/setup.c +++ b/arch/mips/pmc-sierra/yosemite/setup.c @@ -70,7 +70,7 @@ void __init bus_error_init(void) } -unsigned long read_persistent_clock(void) +void read_persistent_clock(struct timespec *ts) { unsigned int year, month, day, hour, min, sec; unsigned long flags; @@ -92,7 +92,8 @@ unsigned long read_persistent_clock(void) m48t37_base->control = 0x00; spin_unlock_irqrestore(&rtc_lock, flags); - return mktime(year, month, day, hour, min, sec); + ts->tv_sec = mktime(year, month, day, hour, min, sec); + ts->tv_nsec = 0; } int rtc_mips_set_time(unsigned long tim) diff --git a/arch/mips/sibyte/swarm/setup.c b/arch/mips/sibyte/swarm/setup.c index 672e45d495a9..623ffc933c4c 100644 --- a/arch/mips/sibyte/swarm/setup.c +++ b/arch/mips/sibyte/swarm/setup.c @@ -87,19 +87,26 @@ enum swarm_rtc_type { enum swarm_rtc_type swarm_rtc_type; -unsigned long read_persistent_clock(void) +void read_persistent_clock(struct timespec *ts) { + unsigned long sec; + switch (swarm_rtc_type) { case RTC_XICOR: - return xicor_get_time(); + sec = xicor_get_time(); + break; case RTC_M4LT81: - return m41t81_get_time(); + sec = m41t81_get_time(); + break; case RTC_NONE: default: - return mktime(2000, 1, 1, 0, 0, 0); + sec = mktime(2000, 1, 1, 0, 0, 0); + break; } + ts->tv_sec = sec; + tv->tv_nsec = 0; } int rtc_mips_set_time(unsigned long sec) diff --git a/arch/mips/sni/time.c b/arch/mips/sni/time.c index 0d9ec1a5c24a..62df6a598e0a 100644 --- a/arch/mips/sni/time.c +++ b/arch/mips/sni/time.c @@ -182,7 +182,8 @@ void __init plat_time_init(void) setup_pit_timer(); } -unsigned long read_persistent_clock(void) +void read_persistent_clock(struct timespec *ts) { - return -1; + ts->tv_sec = -1; + ts->tv_nsec = 0; } diff --git a/arch/powerpc/kernel/time.c b/arch/powerpc/kernel/time.c index eae4511ceeac..ad63f30fe3da 100644 --- a/arch/powerpc/kernel/time.c +++ b/arch/powerpc/kernel/time.c @@ -769,7 +769,7 @@ int update_persistent_clock(struct timespec now) return ppc_md.set_rtc_time(&tm); } -unsigned long read_persistent_clock(void) +void read_persistent_clock(struct timespec *ts) { struct rtc_time tm; static int first = 1; @@ -787,8 +787,9 @@ unsigned long read_persistent_clock(void) if (!ppc_md.get_rtc_time) return 0; ppc_md.get_rtc_time(&tm); - return mktime(tm.tm_year+1900, tm.tm_mon+1, tm.tm_mday, - tm.tm_hour, tm.tm_min, tm.tm_sec); + ts->tv_sec = mktime(tm.tm_year+1900, tm.tm_mon+1, tm.tm_mday, + tm.tm_hour, tm.tm_min, tm.tm_sec); + ts->tv_nsec = 0; } /* clocksource code */ diff --git a/arch/s390/kernel/time.c b/arch/s390/kernel/time.c index e76c2e7a8b9a..a94ec48587b4 100644 --- a/arch/s390/kernel/time.c +++ b/arch/s390/kernel/time.c @@ -182,12 +182,9 @@ static void timing_alert_interrupt(__u16 code) static void etr_reset(void); static void stp_reset(void); -unsigned long read_persistent_clock(void) +void read_persistent_clock(struct timespec *ts) { - struct timespec ts; - - tod_to_timeval(get_clock() - TOD_UNIX_EPOCH, &ts); - return ts.tv_sec; + tod_to_timeval(get_clock() - TOD_UNIX_EPOCH, ts); } static cycle_t read_tod_clock(struct clocksource *cs) @@ -248,7 +245,6 @@ void __init time_init(void) { struct timespec ts; unsigned long flags; - cycle_t now; /* Reset time synchronization interfaces. */ etr_reset(); @@ -266,20 +262,10 @@ void __init time_init(void) panic("Could not register TOD clock source"); /* - * The TOD clock is an accurate clock. The xtime should be - * initialized in a way that the difference between TOD and - * xtime is reasonably small. Too bad that timekeeping_init - * sets xtime.tv_nsec to zero. In addition the clock source - * change from the jiffies clock source to the TOD clock - * source add another error of up to 1/HZ second. The same - * function sets wall_to_monotonic to a value that is too - * small for /proc/uptime to be accurate. - * Reset xtime and wall_to_monotonic to sane values. + * Reset wall_to_monotonic to the initial timestamp created + * in head.S to get a precise value in /proc/uptime. */ write_seqlock_irqsave(&xtime_lock, flags); - now = get_clock(); - tod_to_timeval(now - TOD_UNIX_EPOCH, &xtime); - clocksource_tod.cycle_last = now; tod_to_timeval(sched_clock_base_cc - TOD_UNIX_EPOCH, &ts); set_normalized_timespec(&wall_to_monotonic, -ts.tv_sec, -ts.tv_nsec); write_sequnlock_irqrestore(&xtime_lock, flags); diff --git a/arch/sh/kernel/time.c b/arch/sh/kernel/time.c index 9b352a1e3fb4..3f4706aa975e 100644 --- a/arch/sh/kernel/time.c +++ b/arch/sh/kernel/time.c @@ -39,11 +39,9 @@ void (*rtc_sh_get_time)(struct timespec *) = null_rtc_get_time; int (*rtc_sh_set_time)(const time_t) = null_rtc_set_time; #ifdef CONFIG_GENERIC_CMOS_UPDATE -unsigned long read_persistent_clock(void) +void read_persistent_clock(struct timespec *ts) { - struct timespec tv; - rtc_sh_get_time(&tv); - return tv.tv_sec; + rtc_sh_get_time(&ts); } int update_persistent_clock(struct timespec now) diff --git a/arch/x86/kernel/rtc.c b/arch/x86/kernel/rtc.c index 5d465b207e72..bf67dcb4a44c 100644 --- a/arch/x86/kernel/rtc.c +++ b/arch/x86/kernel/rtc.c @@ -178,7 +178,7 @@ static int set_rtc_mmss(unsigned long nowtime) } /* not static: needed by APM */ -unsigned long read_persistent_clock(void) +void read_persistent_clock(struct timespec *ts) { unsigned long retval, flags; @@ -186,7 +186,8 @@ unsigned long read_persistent_clock(void) retval = get_wallclock(); spin_unlock_irqrestore(&rtc_lock, flags); - return retval; + ts->tv_sec = retval; + ts->tv_nsec = 0; } int update_persistent_clock(struct timespec now) diff --git a/arch/xtensa/kernel/time.c b/arch/xtensa/kernel/time.c index 8848120d291b..19085ff0484a 100644 --- a/arch/xtensa/kernel/time.c +++ b/arch/xtensa/kernel/time.c @@ -59,9 +59,8 @@ static struct irqaction timer_irqaction = { void __init time_init(void) { - xtime.tv_nsec = 0; - xtime.tv_sec = read_persistent_clock(); - + /* FIXME: xtime&wall_to_monotonic are set in timekeeping_init. */ + read_persistent_clock(&xtime); set_normalized_timespec(&wall_to_monotonic, -xtime.tv_sec, -xtime.tv_nsec); diff --git a/include/linux/time.h b/include/linux/time.h index e7c844558884..53a3216f0d1b 100644 --- a/include/linux/time.h +++ b/include/linux/time.h @@ -101,7 +101,7 @@ extern struct timespec xtime; extern struct timespec wall_to_monotonic; extern seqlock_t xtime_lock; -extern unsigned long read_persistent_clock(void); +extern void read_persistent_clock(struct timespec *ts); extern int update_persistent_clock(struct timespec now); extern int no_sync_cmos_clock __read_mostly; void timekeeping_init(void); diff --git a/kernel/time/timekeeping.c b/kernel/time/timekeeping.c index 41579e7fcf9d..f1a21ce491e6 100644 --- a/kernel/time/timekeeping.c +++ b/kernel/time/timekeeping.c @@ -154,7 +154,7 @@ __cacheline_aligned_in_smp DEFINE_SEQLOCK(xtime_lock); */ struct timespec xtime __attribute__ ((aligned (16))); struct timespec wall_to_monotonic __attribute__ ((aligned (16))); -static unsigned long total_sleep_time; /* seconds */ +static struct timespec total_sleep_time; /* * The raw monotonic time for the CLOCK_MONOTONIC_RAW posix clock. @@ -487,17 +487,18 @@ int timekeeping_valid_for_hres(void) } /** - * read_persistent_clock - Return time in seconds from the persistent clock. + * read_persistent_clock - Return time from the persistent clock. * * Weak dummy function for arches that do not yet support it. - * Returns seconds from epoch using the battery backed persistent clock. - * Returns zero if unsupported. + * Reads the time from the battery backed persistent clock. + * Returns a timespec with tv_sec=0 and tv_nsec=0 if unsupported. * * XXX - Do be sure to remove it once all arches implement it. */ -unsigned long __attribute__((weak)) read_persistent_clock(void) +void __attribute__((weak)) read_persistent_clock(struct timespec *ts) { - return 0; + ts->tv_sec = 0; + ts->tv_nsec = 0; } /* @@ -507,7 +508,9 @@ void __init timekeeping_init(void) { struct clocksource *clock; unsigned long flags; - unsigned long sec = read_persistent_clock(); + struct timespec now; + + read_persistent_clock(&now); write_seqlock_irqsave(&xtime_lock, flags); @@ -518,19 +521,20 @@ void __init timekeeping_init(void) clock->enable(clock); timekeeper_setup_internals(clock); - xtime.tv_sec = sec; - xtime.tv_nsec = 0; + xtime.tv_sec = now.tv_sec; + xtime.tv_nsec = now.tv_nsec; raw_time.tv_sec = 0; raw_time.tv_nsec = 0; set_normalized_timespec(&wall_to_monotonic, -xtime.tv_sec, -xtime.tv_nsec); update_xtime_cache(0); - total_sleep_time = 0; + total_sleep_time.tv_sec = 0; + total_sleep_time.tv_nsec = 0; write_sequnlock_irqrestore(&xtime_lock, flags); } /* time in seconds when suspend began */ -static unsigned long timekeeping_suspend_time; +static struct timespec timekeeping_suspend_time; /** * timekeeping_resume - Resumes the generic timekeeping subsystem. @@ -543,18 +547,19 @@ static unsigned long timekeeping_suspend_time; static int timekeeping_resume(struct sys_device *dev) { unsigned long flags; - unsigned long now = read_persistent_clock(); + struct timespec ts; + + read_persistent_clock(&ts); clocksource_resume(); write_seqlock_irqsave(&xtime_lock, flags); - if (now && (now > timekeeping_suspend_time)) { - unsigned long sleep_length = now - timekeeping_suspend_time; - - xtime.tv_sec += sleep_length; - wall_to_monotonic.tv_sec -= sleep_length; - total_sleep_time += sleep_length; + if (timespec_compare(&ts, &timekeeping_suspend_time) > 0) { + ts = timespec_sub(ts, timekeeping_suspend_time); + xtime = timespec_add_safe(xtime, ts); + wall_to_monotonic = timespec_sub(wall_to_monotonic, ts); + total_sleep_time = timespec_add_safe(total_sleep_time, ts); } update_xtime_cache(0); /* re-base the last cycle value */ @@ -577,7 +582,7 @@ static int timekeeping_suspend(struct sys_device *dev, pm_message_t state) { unsigned long flags; - timekeeping_suspend_time = read_persistent_clock(); + read_persistent_clock(&timekeeping_suspend_time); write_seqlock_irqsave(&xtime_lock, flags); timekeeping_forward_now(); @@ -801,9 +806,10 @@ void update_wall_time(void) */ void getboottime(struct timespec *ts) { - set_normalized_timespec(ts, - - (wall_to_monotonic.tv_sec + total_sleep_time), - - wall_to_monotonic.tv_nsec); + struct timespec boottime; + + boottime = timespec_add_safe(wall_to_monotonic, total_sleep_time); + set_normalized_timespec(ts, -boottime.tv_sec, -boottime.tv_nsec); } /** @@ -812,7 +818,7 @@ void getboottime(struct timespec *ts) */ void monotonic_to_bootbased(struct timespec *ts) { - ts->tv_sec += total_sleep_time; + *ts = timespec_add_safe(*ts, total_sleep_time); } unsigned long get_seconds(void) -- cgit v1.2.3 From 23970e389e9cee43c4b41023935e1417271708b2 Mon Sep 17 00:00:00 2001 From: Martin Schwidefsky Date: Fri, 14 Aug 2009 15:47:32 +0200 Subject: timekeeping: Introduce read_boot_clock Add the new function read_boot_clock to get the exact time the system has been started. For architectures without support for exact boot time a new weak function is added that returns 0. Use the exact boot time to initialize wall_to_monotonic, or xtime if the read_boot_clock returned 0. Signed-off-by: Martin Schwidefsky Cc: Ingo Molnar Acked-by: John Stultz Cc: Daniel Walker LKML-Reference: <20090814134811.296703241@de.ibm.com> Signed-off-by: Thomas Gleixner --- arch/s390/kernel/time.c | 17 +++++------------ include/linux/time.h | 1 + kernel/time/timekeeping.c | 24 ++++++++++++++++++++++-- 3 files changed, 28 insertions(+), 14 deletions(-) (limited to 'include') diff --git a/arch/s390/kernel/time.c b/arch/s390/kernel/time.c index a94ec48587b4..6bff1a1d9060 100644 --- a/arch/s390/kernel/time.c +++ b/arch/s390/kernel/time.c @@ -187,6 +187,11 @@ void read_persistent_clock(struct timespec *ts) tod_to_timeval(get_clock() - TOD_UNIX_EPOCH, ts); } +void read_boot_clock(struct timespec *ts) +{ + tod_to_timeval(sched_clock_base_cc - TOD_UNIX_EPOCH, ts); +} + static cycle_t read_tod_clock(struct clocksource *cs) { return get_clock(); @@ -243,9 +248,6 @@ void update_vsyscall_tz(void) */ void __init time_init(void) { - struct timespec ts; - unsigned long flags; - /* Reset time synchronization interfaces. */ etr_reset(); stp_reset(); @@ -261,15 +263,6 @@ void __init time_init(void) if (clocksource_register(&clocksource_tod) != 0) panic("Could not register TOD clock source"); - /* - * Reset wall_to_monotonic to the initial timestamp created - * in head.S to get a precise value in /proc/uptime. - */ - write_seqlock_irqsave(&xtime_lock, flags); - tod_to_timeval(sched_clock_base_cc - TOD_UNIX_EPOCH, &ts); - set_normalized_timespec(&wall_to_monotonic, -ts.tv_sec, -ts.tv_nsec); - write_sequnlock_irqrestore(&xtime_lock, flags); - /* Enable TOD clock interrupts on the boot cpu. */ init_cpu_timer(); diff --git a/include/linux/time.h b/include/linux/time.h index 53a3216f0d1b..f505988398e6 100644 --- a/include/linux/time.h +++ b/include/linux/time.h @@ -102,6 +102,7 @@ extern struct timespec wall_to_monotonic; extern seqlock_t xtime_lock; extern void read_persistent_clock(struct timespec *ts); +extern void read_boot_clock(struct timespec *ts); extern int update_persistent_clock(struct timespec now); extern int no_sync_cmos_clock __read_mostly; void timekeeping_init(void); diff --git a/kernel/time/timekeeping.c b/kernel/time/timekeeping.c index f1a21ce491e6..15e06defca55 100644 --- a/kernel/time/timekeeping.c +++ b/kernel/time/timekeeping.c @@ -501,6 +501,21 @@ void __attribute__((weak)) read_persistent_clock(struct timespec *ts) ts->tv_nsec = 0; } +/** + * read_boot_clock - Return time of the system start. + * + * Weak dummy function for arches that do not yet support it. + * Function to read the exact time the system has been started. + * Returns a timespec with tv_sec=0 and tv_nsec=0 if unsupported. + * + * XXX - Do be sure to remove it once all arches implement it. + */ +void __attribute__((weak)) read_boot_clock(struct timespec *ts) +{ + ts->tv_sec = 0; + ts->tv_nsec = 0; +} + /* * timekeeping_init - Initializes the clocksource and common timekeeping values */ @@ -508,9 +523,10 @@ void __init timekeeping_init(void) { struct clocksource *clock; unsigned long flags; - struct timespec now; + struct timespec now, boot; read_persistent_clock(&now); + read_boot_clock(&boot); write_seqlock_irqsave(&xtime_lock, flags); @@ -525,8 +541,12 @@ void __init timekeeping_init(void) xtime.tv_nsec = now.tv_nsec; raw_time.tv_sec = 0; raw_time.tv_nsec = 0; + if (boot.tv_sec == 0 && boot.tv_nsec == 0) { + boot.tv_sec = xtime.tv_sec; + boot.tv_nsec = xtime.tv_nsec; + } set_normalized_timespec(&wall_to_monotonic, - -xtime.tv_sec, -xtime.tv_nsec); + -boot.tv_sec, -boot.tv_nsec); update_xtime_cache(0); total_sleep_time.tv_sec = 0; total_sleep_time.tv_nsec = 0; -- cgit v1.2.3 From 799e64f05f4bfaad2bb3165cab95c8c992a1c296 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Sat, 15 Aug 2009 09:53:47 -0700 Subject: cpu hotplug: Introduce cpu_notifier() to handle !HOTPLUG_CPU case This patch introduces a new cpu_notifier() API that is similar to hotcpu_notifier(), but which also notifies of CPUs coming online during boot in the !HOTPLUG_CPU case. Reported-by: Ingo Molnar Reported-by: Hugh Dickins Tested-by: Hugh Dickins Signed-off-by: Paul E. McKenney Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: josht@linux.vnet.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org Cc: benh@kernel.crashing.org LKML-Reference: <12503552312611-git-send-email-> Signed-off-by: Ingo Molnar --- include/linux/cpu.h | 17 ++++++++++++----- 1 file changed, 12 insertions(+), 5 deletions(-) (limited to 'include') diff --git a/include/linux/cpu.h b/include/linux/cpu.h index 4d668e05d458..47536197ffdd 100644 --- a/include/linux/cpu.h +++ b/include/linux/cpu.h @@ -48,6 +48,15 @@ struct notifier_block; #ifdef CONFIG_SMP /* Need to know about CPUs going up/down? */ +#if defined(CONFIG_HOTPLUG_CPU) || !defined(MODULE) +#define cpu_notifier(fn, pri) { \ + static struct notifier_block fn##_nb __cpuinitdata = \ + { .notifier_call = fn, .priority = pri }; \ + register_cpu_notifier(&fn##_nb); \ +} +#else /* #if defined(CONFIG_HOTPLUG_CPU) || !defined(MODULE) */ +#define cpu_notifier(fn, pri) do { (void)(fn); } while (0) +#endif /* #else #if defined(CONFIG_HOTPLUG_CPU) || !defined(MODULE) */ #ifdef CONFIG_HOTPLUG_CPU extern int register_cpu_notifier(struct notifier_block *nb); extern void unregister_cpu_notifier(struct notifier_block *nb); @@ -74,6 +83,8 @@ extern void cpu_maps_update_done(void); #else /* CONFIG_SMP */ +#define cpu_notifier(fn, pri) do { (void)(fn); } while (0) + static inline int register_cpu_notifier(struct notifier_block *nb) { return 0; @@ -99,11 +110,7 @@ extern struct sysdev_class cpu_sysdev_class; extern void get_online_cpus(void); extern void put_online_cpus(void); -#define hotcpu_notifier(fn, pri) { \ - static struct notifier_block fn##_nb __cpuinitdata = \ - { .notifier_call = fn, .priority = pri }; \ - register_cpu_notifier(&fn##_nb); \ -} +#define hotcpu_notifier(fn, pri) cpu_notifier(fn, pri) #define register_hotcpu_notifier(nb) register_cpu_notifier(nb) #define unregister_hotcpu_notifier(nb) unregister_cpu_notifier(nb) int cpu_down(unsigned int cpu); -- cgit v1.2.3 From 2bf49690325b62480a42f7afed5e9f164173c570 Mon Sep 17 00:00:00 2001 From: Thomas Liu Date: Tue, 14 Jul 2009 12:14:09 -0400 Subject: SELinux: Convert avc_audit to use lsm_audit.h Convert avc_audit in security/selinux/avc.c to use lsm_audit.h, for better maintainability. - changed selinux to use common_audit_data instead of avc_audit_data - eliminated code in avc.c and used code from lsm_audit.h instead. Had to add a LSM_AUDIT_NO_AUDIT to lsm_audit.h so that avc_audit can call common_lsm_audit and do the pre and post callbacks without doing the actual dump. This makes it so that the patched version behaves the same way as the unpatched version. Also added a denied field to the selinux_audit_data private space, once again to make it so that the patched version behaves like the unpatched. I've tested and confirmed that AVCs look the same before and after this patch. Signed-off-by: Thomas Liu Acked-by: Stephen Smalley Signed-off-by: James Morris --- include/linux/lsm_audit.h | 2 + security/Makefile | 4 +- security/lsm_audit.c | 2 + security/selinux/avc.c | 197 ++++++++---------------------------- security/selinux/hooks.c | 142 +++++++++++++------------- security/selinux/include/avc.h | 49 +-------- security/selinux/include/netlabel.h | 4 +- security/selinux/include/xfrm.h | 8 +- security/selinux/netlabel.c | 2 +- security/selinux/xfrm.c | 4 +- 10 files changed, 131 insertions(+), 283 deletions(-) (limited to 'include') diff --git a/include/linux/lsm_audit.h b/include/linux/lsm_audit.h index a5514a3a4f17..190c37854870 100644 --- a/include/linux/lsm_audit.h +++ b/include/linux/lsm_audit.h @@ -33,6 +33,7 @@ struct common_audit_data { #define LSM_AUDIT_DATA_IPC 4 #define LSM_AUDIT_DATA_TASK 5 #define LSM_AUDIT_DATA_KEY 6 +#define LSM_AUDIT_NO_AUDIT 7 struct task_struct *tsk; union { struct { @@ -86,6 +87,7 @@ struct common_audit_data { u16 tclass; u32 requested; u32 audited; + u32 denied; struct av_decision *avd; int result; } selinux_audit_data; diff --git a/security/Makefile b/security/Makefile index b56e7f9ecbc2..95ecc06392d7 100644 --- a/security/Makefile +++ b/security/Makefile @@ -16,9 +16,7 @@ obj-$(CONFIG_SECURITYFS) += inode.o # Must precede capability.o in order to stack properly. obj-$(CONFIG_SECURITY_SELINUX) += selinux/built-in.o obj-$(CONFIG_SECURITY_SMACK) += smack/built-in.o -ifeq ($(CONFIG_AUDIT),y) -obj-$(CONFIG_SECURITY_SMACK) += lsm_audit.o -endif +obj-$(CONFIG_AUDIT) += lsm_audit.o obj-$(CONFIG_SECURITY_TOMOYO) += tomoyo/built-in.o obj-$(CONFIG_SECURITY_ROOTPLUG) += root_plug.o obj-$(CONFIG_CGROUP_DEVICE) += device_cgroup.o diff --git a/security/lsm_audit.c b/security/lsm_audit.c index 94b868494b31..500aad0ebd6a 100644 --- a/security/lsm_audit.c +++ b/security/lsm_audit.c @@ -220,6 +220,8 @@ static void dump_common_audit_data(struct audit_buffer *ab, } switch (a->type) { + case LSM_AUDIT_NO_AUDIT: + return; case LSM_AUDIT_DATA_IPC: audit_log_format(ab, " key=%d ", a->u.ipc_id); break; diff --git a/security/selinux/avc.c b/security/selinux/avc.c index 236aaa2ea86d..e3d19014259b 100644 --- a/security/selinux/avc.c +++ b/security/selinux/avc.c @@ -492,23 +492,35 @@ out: return node; } -static inline void avc_print_ipv6_addr(struct audit_buffer *ab, - struct in6_addr *addr, __be16 port, - char *name1, char *name2) +/** + * avc_audit_pre_callback - SELinux specific information + * will be called by generic audit code + * @ab: the audit buffer + * @a: audit_data + */ +static void avc_audit_pre_callback(struct audit_buffer *ab, void *a) { - if (!ipv6_addr_any(addr)) - audit_log_format(ab, " %s=%pI6", name1, addr); - if (port) - audit_log_format(ab, " %s=%d", name2, ntohs(port)); + struct common_audit_data *ad = a; + audit_log_format(ab, "avc: %s ", + ad->selinux_audit_data.denied ? "denied" : "granted"); + avc_dump_av(ab, ad->selinux_audit_data.tclass, + ad->selinux_audit_data.audited); + audit_log_format(ab, " for "); } -static inline void avc_print_ipv4_addr(struct audit_buffer *ab, __be32 addr, - __be16 port, char *name1, char *name2) +/** + * avc_audit_post_callback - SELinux specific information + * will be called by generic audit code + * @ab: the audit buffer + * @a: audit_data + */ +static void avc_audit_post_callback(struct audit_buffer *ab, void *a) { - if (addr) - audit_log_format(ab, " %s=%pI4", name1, &addr); - if (port) - audit_log_format(ab, " %s=%d", name2, ntohs(port)); + struct common_audit_data *ad = a; + audit_log_format(ab, " "); + avc_dump_query(ab, ad->selinux_audit_data.ssid, + ad->selinux_audit_data.tsid, + ad->selinux_audit_data.tclass); } /** @@ -532,13 +544,10 @@ static inline void avc_print_ipv4_addr(struct audit_buffer *ab, __be32 addr, */ void avc_audit(u32 ssid, u32 tsid, u16 tclass, u32 requested, - struct av_decision *avd, int result, struct avc_audit_data *a) + struct av_decision *avd, int result, struct common_audit_data *a) { - struct task_struct *tsk = current; - struct inode *inode = NULL; + struct common_audit_data stack_data; u32 denied, audited; - struct audit_buffer *ab; - denied = requested & ~avd->allowed; if (denied) { audited = denied; @@ -551,144 +560,20 @@ void avc_audit(u32 ssid, u32 tsid, if (!(audited & avd->auditallow)) return; } - - ab = audit_log_start(current->audit_context, GFP_ATOMIC, AUDIT_AVC); - if (!ab) - return; /* audit_panic has been called */ - audit_log_format(ab, "avc: %s ", denied ? "denied" : "granted"); - avc_dump_av(ab, tclass, audited); - audit_log_format(ab, " for "); - if (a && a->tsk) - tsk = a->tsk; - if (tsk && tsk->pid) { - audit_log_format(ab, " pid=%d comm=", tsk->pid); - audit_log_untrustedstring(ab, tsk->comm); - } - if (a) { - switch (a->type) { - case AVC_AUDIT_DATA_IPC: - audit_log_format(ab, " key=%d", a->u.ipc_id); - break; - case AVC_AUDIT_DATA_CAP: - audit_log_format(ab, " capability=%d", a->u.cap); - break; - case AVC_AUDIT_DATA_FS: - if (a->u.fs.path.dentry) { - struct dentry *dentry = a->u.fs.path.dentry; - if (a->u.fs.path.mnt) { - audit_log_d_path(ab, "path=", - &a->u.fs.path); - } else { - audit_log_format(ab, " name="); - audit_log_untrustedstring(ab, dentry->d_name.name); - } - inode = dentry->d_inode; - } else if (a->u.fs.inode) { - struct dentry *dentry; - inode = a->u.fs.inode; - dentry = d_find_alias(inode); - if (dentry) { - audit_log_format(ab, " name="); - audit_log_untrustedstring(ab, dentry->d_name.name); - dput(dentry); - } - } - if (inode) - audit_log_format(ab, " dev=%s ino=%lu", - inode->i_sb->s_id, - inode->i_ino); - break; - case AVC_AUDIT_DATA_NET: - if (a->u.net.sk) { - struct sock *sk = a->u.net.sk; - struct unix_sock *u; - int len = 0; - char *p = NULL; - - switch (sk->sk_family) { - case AF_INET: { - struct inet_sock *inet = inet_sk(sk); - - avc_print_ipv4_addr(ab, inet->rcv_saddr, - inet->sport, - "laddr", "lport"); - avc_print_ipv4_addr(ab, inet->daddr, - inet->dport, - "faddr", "fport"); - break; - } - case AF_INET6: { - struct inet_sock *inet = inet_sk(sk); - struct ipv6_pinfo *inet6 = inet6_sk(sk); - - avc_print_ipv6_addr(ab, &inet6->rcv_saddr, - inet->sport, - "laddr", "lport"); - avc_print_ipv6_addr(ab, &inet6->daddr, - inet->dport, - "faddr", "fport"); - break; - } - case AF_UNIX: - u = unix_sk(sk); - if (u->dentry) { - struct path path = { - .dentry = u->dentry, - .mnt = u->mnt - }; - audit_log_d_path(ab, "path=", - &path); - break; - } - if (!u->addr) - break; - len = u->addr->len-sizeof(short); - p = &u->addr->name->sun_path[0]; - audit_log_format(ab, " path="); - if (*p) - audit_log_untrustedstring(ab, p); - else - audit_log_n_hex(ab, p, len); - break; - } - } - - switch (a->u.net.family) { - case AF_INET: - avc_print_ipv4_addr(ab, a->u.net.v4info.saddr, - a->u.net.sport, - "saddr", "src"); - avc_print_ipv4_addr(ab, a->u.net.v4info.daddr, - a->u.net.dport, - "daddr", "dest"); - break; - case AF_INET6: - avc_print_ipv6_addr(ab, &a->u.net.v6info.saddr, - a->u.net.sport, - "saddr", "src"); - avc_print_ipv6_addr(ab, &a->u.net.v6info.daddr, - a->u.net.dport, - "daddr", "dest"); - break; - } - if (a->u.net.netif > 0) { - struct net_device *dev; - - /* NOTE: we always use init's namespace */ - dev = dev_get_by_index(&init_net, - a->u.net.netif); - if (dev) { - audit_log_format(ab, " netif=%s", - dev->name); - dev_put(dev); - } - } - break; - } + if (!a) { + a = &stack_data; + memset(a, 0, sizeof(*a)); + a->type = LSM_AUDIT_NO_AUDIT; } - audit_log_format(ab, " "); - avc_dump_query(ab, ssid, tsid, tclass); - audit_log_end(ab); + a->selinux_audit_data.tclass = tclass; + a->selinux_audit_data.requested = requested; + a->selinux_audit_data.ssid = ssid; + a->selinux_audit_data.tsid = tsid; + a->selinux_audit_data.audited = audited; + a->selinux_audit_data.denied = denied; + a->lsm_pre_audit = avc_audit_pre_callback; + a->lsm_post_audit = avc_audit_post_callback; + common_lsm_audit(a); } /** @@ -956,7 +841,7 @@ out: * another -errno upon other errors. */ int avc_has_perm(u32 ssid, u32 tsid, u16 tclass, - u32 requested, struct avc_audit_data *auditdata) + u32 requested, struct common_audit_data *auditdata) { struct av_decision avd; int rc; diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c index 5aa45b168122..254b7983657d 100644 --- a/security/selinux/hooks.c +++ b/security/selinux/hooks.c @@ -1478,14 +1478,14 @@ static int task_has_capability(struct task_struct *tsk, const struct cred *cred, int cap, int audit) { - struct avc_audit_data ad; + struct common_audit_data ad; struct av_decision avd; u16 sclass; u32 sid = cred_sid(cred); u32 av = CAP_TO_MASK(cap); int rc; - AVC_AUDIT_DATA_INIT(&ad, CAP); + COMMON_AUDIT_DATA_INIT(&ad, CAP); ad.tsk = tsk; ad.u.cap = cap; @@ -1524,10 +1524,10 @@ static int task_has_system(struct task_struct *tsk, static int inode_has_perm(const struct cred *cred, struct inode *inode, u32 perms, - struct avc_audit_data *adp) + struct common_audit_data *adp) { struct inode_security_struct *isec; - struct avc_audit_data ad; + struct common_audit_data ad; u32 sid; if (unlikely(IS_PRIVATE(inode))) @@ -1538,7 +1538,7 @@ static int inode_has_perm(const struct cred *cred, if (!adp) { adp = &ad; - AVC_AUDIT_DATA_INIT(&ad, FS); + COMMON_AUDIT_DATA_INIT(&ad, FS); ad.u.fs.inode = inode; } @@ -1554,9 +1554,9 @@ static inline int dentry_has_perm(const struct cred *cred, u32 av) { struct inode *inode = dentry->d_inode; - struct avc_audit_data ad; + struct common_audit_data ad; - AVC_AUDIT_DATA_INIT(&ad, FS); + COMMON_AUDIT_DATA_INIT(&ad, FS); ad.u.fs.path.mnt = mnt; ad.u.fs.path.dentry = dentry; return inode_has_perm(cred, inode, av, &ad); @@ -1576,11 +1576,11 @@ static int file_has_perm(const struct cred *cred, { struct file_security_struct *fsec = file->f_security; struct inode *inode = file->f_path.dentry->d_inode; - struct avc_audit_data ad; + struct common_audit_data ad; u32 sid = cred_sid(cred); int rc; - AVC_AUDIT_DATA_INIT(&ad, FS); + COMMON_AUDIT_DATA_INIT(&ad, FS); ad.u.fs.path = file->f_path; if (sid != fsec->sid) { @@ -1611,7 +1611,7 @@ static int may_create(struct inode *dir, struct inode_security_struct *dsec; struct superblock_security_struct *sbsec; u32 sid, newsid; - struct avc_audit_data ad; + struct common_audit_data ad; int rc; dsec = dir->i_security; @@ -1620,7 +1620,7 @@ static int may_create(struct inode *dir, sid = tsec->sid; newsid = tsec->create_sid; - AVC_AUDIT_DATA_INIT(&ad, FS); + COMMON_AUDIT_DATA_INIT(&ad, FS); ad.u.fs.path.dentry = dentry; rc = avc_has_perm(sid, dsec->sid, SECCLASS_DIR, @@ -1664,7 +1664,7 @@ static int may_link(struct inode *dir, { struct inode_security_struct *dsec, *isec; - struct avc_audit_data ad; + struct common_audit_data ad; u32 sid = current_sid(); u32 av; int rc; @@ -1672,7 +1672,7 @@ static int may_link(struct inode *dir, dsec = dir->i_security; isec = dentry->d_inode->i_security; - AVC_AUDIT_DATA_INIT(&ad, FS); + COMMON_AUDIT_DATA_INIT(&ad, FS); ad.u.fs.path.dentry = dentry; av = DIR__SEARCH; @@ -1707,7 +1707,7 @@ static inline int may_rename(struct inode *old_dir, struct dentry *new_dentry) { struct inode_security_struct *old_dsec, *new_dsec, *old_isec, *new_isec; - struct avc_audit_data ad; + struct common_audit_data ad; u32 sid = current_sid(); u32 av; int old_is_dir, new_is_dir; @@ -1718,7 +1718,7 @@ static inline int may_rename(struct inode *old_dir, old_is_dir = S_ISDIR(old_dentry->d_inode->i_mode); new_dsec = new_dir->i_security; - AVC_AUDIT_DATA_INIT(&ad, FS); + COMMON_AUDIT_DATA_INIT(&ad, FS); ad.u.fs.path.dentry = old_dentry; rc = avc_has_perm(sid, old_dsec->sid, SECCLASS_DIR, @@ -1760,7 +1760,7 @@ static inline int may_rename(struct inode *old_dir, static int superblock_has_perm(const struct cred *cred, struct super_block *sb, u32 perms, - struct avc_audit_data *ad) + struct common_audit_data *ad) { struct superblock_security_struct *sbsec; u32 sid = cred_sid(cred); @@ -2100,7 +2100,7 @@ static int selinux_bprm_set_creds(struct linux_binprm *bprm) const struct task_security_struct *old_tsec; struct task_security_struct *new_tsec; struct inode_security_struct *isec; - struct avc_audit_data ad; + struct common_audit_data ad; struct inode *inode = bprm->file->f_path.dentry->d_inode; int rc; @@ -2138,7 +2138,7 @@ static int selinux_bprm_set_creds(struct linux_binprm *bprm) return rc; } - AVC_AUDIT_DATA_INIT(&ad, FS); + COMMON_AUDIT_DATA_INIT(&ad, FS); ad.u.fs.path = bprm->file->f_path; if (bprm->file->f_path.mnt->mnt_flags & MNT_NOSUID) @@ -2231,7 +2231,7 @@ extern struct dentry *selinux_null; static inline void flush_unauthorized_files(const struct cred *cred, struct files_struct *files) { - struct avc_audit_data ad; + struct common_audit_data ad; struct file *file, *devnull = NULL; struct tty_struct *tty; struct fdtable *fdt; @@ -2265,7 +2265,7 @@ static inline void flush_unauthorized_files(const struct cred *cred, /* Revalidate access to inherited open files. */ - AVC_AUDIT_DATA_INIT(&ad, FS); + COMMON_AUDIT_DATA_INIT(&ad, FS); spin_lock(&files->file_lock); for (;;) { @@ -2514,7 +2514,7 @@ out: static int selinux_sb_kern_mount(struct super_block *sb, int flags, void *data) { const struct cred *cred = current_cred(); - struct avc_audit_data ad; + struct common_audit_data ad; int rc; rc = superblock_doinit(sb, data); @@ -2525,7 +2525,7 @@ static int selinux_sb_kern_mount(struct super_block *sb, int flags, void *data) if (flags & MS_KERNMOUNT) return 0; - AVC_AUDIT_DATA_INIT(&ad, FS); + COMMON_AUDIT_DATA_INIT(&ad, FS); ad.u.fs.path.dentry = sb->s_root; return superblock_has_perm(cred, sb, FILESYSTEM__MOUNT, &ad); } @@ -2533,9 +2533,9 @@ static int selinux_sb_kern_mount(struct super_block *sb, int flags, void *data) static int selinux_sb_statfs(struct dentry *dentry) { const struct cred *cred = current_cred(); - struct avc_audit_data ad; + struct common_audit_data ad; - AVC_AUDIT_DATA_INIT(&ad, FS); + COMMON_AUDIT_DATA_INIT(&ad, FS); ad.u.fs.path.dentry = dentry->d_sb->s_root; return superblock_has_perm(cred, dentry->d_sb, FILESYSTEM__GETATTR, &ad); } @@ -2755,7 +2755,7 @@ static int selinux_inode_setxattr(struct dentry *dentry, const char *name, struct inode *inode = dentry->d_inode; struct inode_security_struct *isec = inode->i_security; struct superblock_security_struct *sbsec; - struct avc_audit_data ad; + struct common_audit_data ad; u32 newsid, sid = current_sid(); int rc = 0; @@ -2769,7 +2769,7 @@ static int selinux_inode_setxattr(struct dentry *dentry, const char *name, if (!is_owner_or_cap(inode)) return -EPERM; - AVC_AUDIT_DATA_INIT(&ad, FS); + COMMON_AUDIT_DATA_INIT(&ad, FS); ad.u.fs.path.dentry = dentry; rc = avc_has_perm(sid, isec->sid, isec->sclass, @@ -3418,7 +3418,7 @@ static void selinux_task_to_inode(struct task_struct *p, /* Returns error only if unable to parse addresses */ static int selinux_parse_skb_ipv4(struct sk_buff *skb, - struct avc_audit_data *ad, u8 *proto) + struct common_audit_data *ad, u8 *proto) { int offset, ihlen, ret = -EINVAL; struct iphdr _iph, *ih; @@ -3499,7 +3499,7 @@ out: /* Returns error only if unable to parse addresses */ static int selinux_parse_skb_ipv6(struct sk_buff *skb, - struct avc_audit_data *ad, u8 *proto) + struct common_audit_data *ad, u8 *proto) { u8 nexthdr; int ret = -EINVAL, offset; @@ -3570,7 +3570,7 @@ out: #endif /* IPV6 */ -static int selinux_parse_skb(struct sk_buff *skb, struct avc_audit_data *ad, +static int selinux_parse_skb(struct sk_buff *skb, struct common_audit_data *ad, char **_addrp, int src, u8 *proto) { char *addrp; @@ -3652,7 +3652,7 @@ static int socket_has_perm(struct task_struct *task, struct socket *sock, u32 perms) { struct inode_security_struct *isec; - struct avc_audit_data ad; + struct common_audit_data ad; u32 sid; int err = 0; @@ -3662,7 +3662,7 @@ static int socket_has_perm(struct task_struct *task, struct socket *sock, goto out; sid = task_sid(task); - AVC_AUDIT_DATA_INIT(&ad, NET); + COMMON_AUDIT_DATA_INIT(&ad, NET); ad.u.net.sk = sock->sk; err = avc_has_perm(sid, isec->sid, isec->sclass, perms, &ad); @@ -3749,7 +3749,7 @@ static int selinux_socket_bind(struct socket *sock, struct sockaddr *address, in if (family == PF_INET || family == PF_INET6) { char *addrp; struct inode_security_struct *isec; - struct avc_audit_data ad; + struct common_audit_data ad; struct sockaddr_in *addr4 = NULL; struct sockaddr_in6 *addr6 = NULL; unsigned short snum; @@ -3778,7 +3778,7 @@ static int selinux_socket_bind(struct socket *sock, struct sockaddr *address, in snum, &sid); if (err) goto out; - AVC_AUDIT_DATA_INIT(&ad, NET); + COMMON_AUDIT_DATA_INIT(&ad, NET); ad.u.net.sport = htons(snum); ad.u.net.family = family; err = avc_has_perm(isec->sid, sid, @@ -3811,7 +3811,7 @@ static int selinux_socket_bind(struct socket *sock, struct sockaddr *address, in if (err) goto out; - AVC_AUDIT_DATA_INIT(&ad, NET); + COMMON_AUDIT_DATA_INIT(&ad, NET); ad.u.net.sport = htons(snum); ad.u.net.family = family; @@ -3845,7 +3845,7 @@ static int selinux_socket_connect(struct socket *sock, struct sockaddr *address, isec = SOCK_INODE(sock)->i_security; if (isec->sclass == SECCLASS_TCP_SOCKET || isec->sclass == SECCLASS_DCCP_SOCKET) { - struct avc_audit_data ad; + struct common_audit_data ad; struct sockaddr_in *addr4 = NULL; struct sockaddr_in6 *addr6 = NULL; unsigned short snum; @@ -3870,7 +3870,7 @@ static int selinux_socket_connect(struct socket *sock, struct sockaddr *address, perm = (isec->sclass == SECCLASS_TCP_SOCKET) ? TCP_SOCKET__NAME_CONNECT : DCCP_SOCKET__NAME_CONNECT; - AVC_AUDIT_DATA_INIT(&ad, NET); + COMMON_AUDIT_DATA_INIT(&ad, NET); ad.u.net.dport = htons(snum); ad.u.net.family = sk->sk_family; err = avc_has_perm(isec->sid, sid, isec->sclass, perm, &ad); @@ -3960,13 +3960,13 @@ static int selinux_socket_unix_stream_connect(struct socket *sock, struct sk_security_struct *ssec; struct inode_security_struct *isec; struct inode_security_struct *other_isec; - struct avc_audit_data ad; + struct common_audit_data ad; int err; isec = SOCK_INODE(sock)->i_security; other_isec = SOCK_INODE(other)->i_security; - AVC_AUDIT_DATA_INIT(&ad, NET); + COMMON_AUDIT_DATA_INIT(&ad, NET); ad.u.net.sk = other->sk; err = avc_has_perm(isec->sid, other_isec->sid, @@ -3992,13 +3992,13 @@ static int selinux_socket_unix_may_send(struct socket *sock, { struct inode_security_struct *isec; struct inode_security_struct *other_isec; - struct avc_audit_data ad; + struct common_audit_data ad; int err; isec = SOCK_INODE(sock)->i_security; other_isec = SOCK_INODE(other)->i_security; - AVC_AUDIT_DATA_INIT(&ad, NET); + COMMON_AUDIT_DATA_INIT(&ad, NET); ad.u.net.sk = other->sk; err = avc_has_perm(isec->sid, other_isec->sid, @@ -4011,7 +4011,7 @@ static int selinux_socket_unix_may_send(struct socket *sock, static int selinux_inet_sys_rcv_skb(int ifindex, char *addrp, u16 family, u32 peer_sid, - struct avc_audit_data *ad) + struct common_audit_data *ad) { int err; u32 if_sid; @@ -4039,10 +4039,10 @@ static int selinux_sock_rcv_skb_compat(struct sock *sk, struct sk_buff *skb, struct sk_security_struct *sksec = sk->sk_security; u32 peer_sid; u32 sk_sid = sksec->sid; - struct avc_audit_data ad; + struct common_audit_data ad; char *addrp; - AVC_AUDIT_DATA_INIT(&ad, NET); + COMMON_AUDIT_DATA_INIT(&ad, NET); ad.u.net.netif = skb->iif; ad.u.net.family = family; err = selinux_parse_skb(skb, &ad, &addrp, 1, NULL); @@ -4080,7 +4080,7 @@ static int selinux_socket_sock_rcv_skb(struct sock *sk, struct sk_buff *skb) struct sk_security_struct *sksec = sk->sk_security; u16 family = sk->sk_family; u32 sk_sid = sksec->sid; - struct avc_audit_data ad; + struct common_audit_data ad; char *addrp; u8 secmark_active; u8 peerlbl_active; @@ -4104,7 +4104,7 @@ static int selinux_socket_sock_rcv_skb(struct sock *sk, struct sk_buff *skb) if (!secmark_active && !peerlbl_active) return 0; - AVC_AUDIT_DATA_INIT(&ad, NET); + COMMON_AUDIT_DATA_INIT(&ad, NET); ad.u.net.netif = skb->iif; ad.u.net.family = family; err = selinux_parse_skb(skb, &ad, &addrp, 1, NULL); @@ -4362,7 +4362,7 @@ static unsigned int selinux_ip_forward(struct sk_buff *skb, int ifindex, int err; char *addrp; u32 peer_sid; - struct avc_audit_data ad; + struct common_audit_data ad; u8 secmark_active; u8 netlbl_active; u8 peerlbl_active; @@ -4379,7 +4379,7 @@ static unsigned int selinux_ip_forward(struct sk_buff *skb, int ifindex, if (selinux_skb_peerlbl_sid(skb, family, &peer_sid) != 0) return NF_DROP; - AVC_AUDIT_DATA_INIT(&ad, NET); + COMMON_AUDIT_DATA_INIT(&ad, NET); ad.u.net.netif = ifindex; ad.u.net.family = family; if (selinux_parse_skb(skb, &ad, &addrp, 1, NULL) != 0) @@ -4467,7 +4467,7 @@ static unsigned int selinux_ip_postroute_compat(struct sk_buff *skb, { struct sock *sk = skb->sk; struct sk_security_struct *sksec; - struct avc_audit_data ad; + struct common_audit_data ad; char *addrp; u8 proto; @@ -4475,7 +4475,7 @@ static unsigned int selinux_ip_postroute_compat(struct sk_buff *skb, return NF_ACCEPT; sksec = sk->sk_security; - AVC_AUDIT_DATA_INIT(&ad, NET); + COMMON_AUDIT_DATA_INIT(&ad, NET); ad.u.net.netif = ifindex; ad.u.net.family = family; if (selinux_parse_skb(skb, &ad, &addrp, 0, &proto)) @@ -4499,7 +4499,7 @@ static unsigned int selinux_ip_postroute(struct sk_buff *skb, int ifindex, u32 secmark_perm; u32 peer_sid; struct sock *sk; - struct avc_audit_data ad; + struct common_audit_data ad; char *addrp; u8 secmark_active; u8 peerlbl_active; @@ -4558,7 +4558,7 @@ static unsigned int selinux_ip_postroute(struct sk_buff *skb, int ifindex, secmark_perm = PACKET__SEND; } - AVC_AUDIT_DATA_INIT(&ad, NET); + COMMON_AUDIT_DATA_INIT(&ad, NET); ad.u.net.netif = ifindex; ad.u.net.family = family; if (selinux_parse_skb(skb, &ad, &addrp, 0, NULL)) @@ -4628,13 +4628,13 @@ static int selinux_netlink_send(struct sock *sk, struct sk_buff *skb) static int selinux_netlink_recv(struct sk_buff *skb, int capability) { int err; - struct avc_audit_data ad; + struct common_audit_data ad; err = cap_netlink_recv(skb, capability); if (err) return err; - AVC_AUDIT_DATA_INIT(&ad, CAP); + COMMON_AUDIT_DATA_INIT(&ad, CAP); ad.u.cap = capability; return avc_has_perm(NETLINK_CB(skb).sid, NETLINK_CB(skb).sid, @@ -4693,12 +4693,12 @@ static int ipc_has_perm(struct kern_ipc_perm *ipc_perms, u32 perms) { struct ipc_security_struct *isec; - struct avc_audit_data ad; + struct common_audit_data ad; u32 sid = current_sid(); isec = ipc_perms->security; - AVC_AUDIT_DATA_INIT(&ad, IPC); + COMMON_AUDIT_DATA_INIT(&ad, IPC); ad.u.ipc_id = ipc_perms->key; return avc_has_perm(sid, isec->sid, isec->sclass, perms, &ad); @@ -4718,7 +4718,7 @@ static void selinux_msg_msg_free_security(struct msg_msg *msg) static int selinux_msg_queue_alloc_security(struct msg_queue *msq) { struct ipc_security_struct *isec; - struct avc_audit_data ad; + struct common_audit_data ad; u32 sid = current_sid(); int rc; @@ -4728,7 +4728,7 @@ static int selinux_msg_queue_alloc_security(struct msg_queue *msq) isec = msq->q_perm.security; - AVC_AUDIT_DATA_INIT(&ad, IPC); + COMMON_AUDIT_DATA_INIT(&ad, IPC); ad.u.ipc_id = msq->q_perm.key; rc = avc_has_perm(sid, isec->sid, SECCLASS_MSGQ, @@ -4748,12 +4748,12 @@ static void selinux_msg_queue_free_security(struct msg_queue *msq) static int selinux_msg_queue_associate(struct msg_queue *msq, int msqflg) { struct ipc_security_struct *isec; - struct avc_audit_data ad; + struct common_audit_data ad; u32 sid = current_sid(); isec = msq->q_perm.security; - AVC_AUDIT_DATA_INIT(&ad, IPC); + COMMON_AUDIT_DATA_INIT(&ad, IPC); ad.u.ipc_id = msq->q_perm.key; return avc_has_perm(sid, isec->sid, SECCLASS_MSGQ, @@ -4792,7 +4792,7 @@ static int selinux_msg_queue_msgsnd(struct msg_queue *msq, struct msg_msg *msg, { struct ipc_security_struct *isec; struct msg_security_struct *msec; - struct avc_audit_data ad; + struct common_audit_data ad; u32 sid = current_sid(); int rc; @@ -4813,7 +4813,7 @@ static int selinux_msg_queue_msgsnd(struct msg_queue *msq, struct msg_msg *msg, return rc; } - AVC_AUDIT_DATA_INIT(&ad, IPC); + COMMON_AUDIT_DATA_INIT(&ad, IPC); ad.u.ipc_id = msq->q_perm.key; /* Can this process write to the queue? */ @@ -4837,14 +4837,14 @@ static int selinux_msg_queue_msgrcv(struct msg_queue *msq, struct msg_msg *msg, { struct ipc_security_struct *isec; struct msg_security_struct *msec; - struct avc_audit_data ad; + struct common_audit_data ad; u32 sid = task_sid(target); int rc; isec = msq->q_perm.security; msec = msg->security; - AVC_AUDIT_DATA_INIT(&ad, IPC); + COMMON_AUDIT_DATA_INIT(&ad, IPC); ad.u.ipc_id = msq->q_perm.key; rc = avc_has_perm(sid, isec->sid, @@ -4859,7 +4859,7 @@ static int selinux_msg_queue_msgrcv(struct msg_queue *msq, struct msg_msg *msg, static int selinux_shm_alloc_security(struct shmid_kernel *shp) { struct ipc_security_struct *isec; - struct avc_audit_data ad; + struct common_audit_data ad; u32 sid = current_sid(); int rc; @@ -4869,7 +4869,7 @@ static int selinux_shm_alloc_security(struct shmid_kernel *shp) isec = shp->shm_perm.security; - AVC_AUDIT_DATA_INIT(&ad, IPC); + COMMON_AUDIT_DATA_INIT(&ad, IPC); ad.u.ipc_id = shp->shm_perm.key; rc = avc_has_perm(sid, isec->sid, SECCLASS_SHM, @@ -4889,12 +4889,12 @@ static void selinux_shm_free_security(struct shmid_kernel *shp) static int selinux_shm_associate(struct shmid_kernel *shp, int shmflg) { struct ipc_security_struct *isec; - struct avc_audit_data ad; + struct common_audit_data ad; u32 sid = current_sid(); isec = shp->shm_perm.security; - AVC_AUDIT_DATA_INIT(&ad, IPC); + COMMON_AUDIT_DATA_INIT(&ad, IPC); ad.u.ipc_id = shp->shm_perm.key; return avc_has_perm(sid, isec->sid, SECCLASS_SHM, @@ -4951,7 +4951,7 @@ static int selinux_shm_shmat(struct shmid_kernel *shp, static int selinux_sem_alloc_security(struct sem_array *sma) { struct ipc_security_struct *isec; - struct avc_audit_data ad; + struct common_audit_data ad; u32 sid = current_sid(); int rc; @@ -4961,7 +4961,7 @@ static int selinux_sem_alloc_security(struct sem_array *sma) isec = sma->sem_perm.security; - AVC_AUDIT_DATA_INIT(&ad, IPC); + COMMON_AUDIT_DATA_INIT(&ad, IPC); ad.u.ipc_id = sma->sem_perm.key; rc = avc_has_perm(sid, isec->sid, SECCLASS_SEM, @@ -4981,12 +4981,12 @@ static void selinux_sem_free_security(struct sem_array *sma) static int selinux_sem_associate(struct sem_array *sma, int semflg) { struct ipc_security_struct *isec; - struct avc_audit_data ad; + struct common_audit_data ad; u32 sid = current_sid(); isec = sma->sem_perm.security; - AVC_AUDIT_DATA_INIT(&ad, IPC); + COMMON_AUDIT_DATA_INIT(&ad, IPC); ad.u.ipc_id = sma->sem_perm.key; return avc_has_perm(sid, isec->sid, SECCLASS_SEM, diff --git a/security/selinux/include/avc.h b/security/selinux/include/avc.h index ae4c3a0e2c1a..e94e82f73818 100644 --- a/security/selinux/include/avc.h +++ b/security/selinux/include/avc.h @@ -13,6 +13,7 @@ #include #include #include +#include #include #include #include @@ -36,48 +37,6 @@ struct inode; struct sock; struct sk_buff; -/* Auxiliary data to use in generating the audit record. */ -struct avc_audit_data { - char type; -#define AVC_AUDIT_DATA_FS 1 -#define AVC_AUDIT_DATA_NET 2 -#define AVC_AUDIT_DATA_CAP 3 -#define AVC_AUDIT_DATA_IPC 4 - struct task_struct *tsk; - union { - struct { - struct path path; - struct inode *inode; - } fs; - struct { - int netif; - struct sock *sk; - u16 family; - __be16 dport; - __be16 sport; - union { - struct { - __be32 daddr; - __be32 saddr; - } v4; - struct { - struct in6_addr daddr; - struct in6_addr saddr; - } v6; - } fam; - } net; - int cap; - int ipc_id; - } u; -}; - -#define v4info fam.v4 -#define v6info fam.v6 - -/* Initialize an AVC audit data structure. */ -#define AVC_AUDIT_DATA_INIT(_d,_t) \ - { memset((_d), 0, sizeof(struct avc_audit_data)); (_d)->type = AVC_AUDIT_DATA_##_t; } - /* * AVC statistics */ @@ -98,7 +57,9 @@ void __init avc_init(void); void avc_audit(u32 ssid, u32 tsid, u16 tclass, u32 requested, - struct av_decision *avd, int result, struct avc_audit_data *auditdata); + struct av_decision *avd, + int result, + struct common_audit_data *a); #define AVC_STRICT 1 /* Ignore permissive mode. */ int avc_has_perm_noaudit(u32 ssid, u32 tsid, @@ -108,7 +69,7 @@ int avc_has_perm_noaudit(u32 ssid, u32 tsid, int avc_has_perm(u32 ssid, u32 tsid, u16 tclass, u32 requested, - struct avc_audit_data *auditdata); + struct common_audit_data *auditdata); u32 avc_policy_seqno(void); diff --git a/security/selinux/include/netlabel.h b/security/selinux/include/netlabel.h index b4b5b9b2f0be..8d7384280a7a 100644 --- a/security/selinux/include/netlabel.h +++ b/security/selinux/include/netlabel.h @@ -59,7 +59,7 @@ int selinux_netlbl_socket_post_create(struct sock *sk, u16 family); int selinux_netlbl_sock_rcv_skb(struct sk_security_struct *sksec, struct sk_buff *skb, u16 family, - struct avc_audit_data *ad); + struct common_audit_data *ad); int selinux_netlbl_socket_setsockopt(struct socket *sock, int level, int optname); @@ -129,7 +129,7 @@ static inline int selinux_netlbl_socket_post_create(struct sock *sk, static inline int selinux_netlbl_sock_rcv_skb(struct sk_security_struct *sksec, struct sk_buff *skb, u16 family, - struct avc_audit_data *ad) + struct common_audit_data *ad) { return 0; } diff --git a/security/selinux/include/xfrm.h b/security/selinux/include/xfrm.h index 289e24b39e3e..13128f9a3e5a 100644 --- a/security/selinux/include/xfrm.h +++ b/security/selinux/include/xfrm.h @@ -41,9 +41,9 @@ static inline int selinux_xfrm_enabled(void) } int selinux_xfrm_sock_rcv_skb(u32 sid, struct sk_buff *skb, - struct avc_audit_data *ad); + struct common_audit_data *ad); int selinux_xfrm_postroute_last(u32 isec_sid, struct sk_buff *skb, - struct avc_audit_data *ad, u8 proto); + struct common_audit_data *ad, u8 proto); int selinux_xfrm_decode_session(struct sk_buff *skb, u32 *sid, int ckall); static inline void selinux_xfrm_notify_policyload(void) @@ -57,13 +57,13 @@ static inline int selinux_xfrm_enabled(void) } static inline int selinux_xfrm_sock_rcv_skb(u32 isec_sid, struct sk_buff *skb, - struct avc_audit_data *ad) + struct common_audit_data *ad) { return 0; } static inline int selinux_xfrm_postroute_last(u32 isec_sid, struct sk_buff *skb, - struct avc_audit_data *ad, u8 proto) + struct common_audit_data *ad, u8 proto) { return 0; } diff --git a/security/selinux/netlabel.c b/security/selinux/netlabel.c index 2e984413c7b2..e68823741ad5 100644 --- a/security/selinux/netlabel.c +++ b/security/selinux/netlabel.c @@ -342,7 +342,7 @@ int selinux_netlbl_socket_post_create(struct sock *sk, u16 family) int selinux_netlbl_sock_rcv_skb(struct sk_security_struct *sksec, struct sk_buff *skb, u16 family, - struct avc_audit_data *ad) + struct common_audit_data *ad) { int rc; u32 nlbl_sid; diff --git a/security/selinux/xfrm.c b/security/selinux/xfrm.c index 72b18452e1a1..f3cb9ed731a9 100644 --- a/security/selinux/xfrm.c +++ b/security/selinux/xfrm.c @@ -401,7 +401,7 @@ int selinux_xfrm_state_delete(struct xfrm_state *x) * gone thru the IPSec process. */ int selinux_xfrm_sock_rcv_skb(u32 isec_sid, struct sk_buff *skb, - struct avc_audit_data *ad) + struct common_audit_data *ad) { int i, rc = 0; struct sec_path *sp; @@ -442,7 +442,7 @@ int selinux_xfrm_sock_rcv_skb(u32 isec_sid, struct sk_buff *skb, * checked in the selinux_xfrm_state_pol_flow_match hook above. */ int selinux_xfrm_postroute_last(u32 isec_sid, struct sk_buff *skb, - struct avc_audit_data *ad, u8 proto) + struct common_audit_data *ad, u8 proto) { struct dst_entry *dst; int rc = 0; -- cgit v1.2.3 From cefb87efc9aa0288849149484870d5ab989fbd59 Mon Sep 17 00:00:00 2001 From: Dave Airlie Date: Sun, 16 Aug 2009 21:05:45 +1000 Subject: drm/radeon/kms: implement bo busy check + current domain This implements the busy ioctl along with a current domain check. returns 0 or -EBUSY puts the current domain no matter what the answer. Signed-off-by: Dave Airlie --- drivers/gpu/drm/radeon/radeon.h | 1 + drivers/gpu/drm/radeon/radeon_gem.c | 22 +++++++++++++++++++++- drivers/gpu/drm/radeon/radeon_object.c | 19 +++++++++++++++++++ include/drm/radeon_drm.h | 2 +- 4 files changed, 42 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h index 87170a56e37b..79ad98264e33 100644 --- a/drivers/gpu/drm/radeon/radeon.h +++ b/drivers/gpu/drm/radeon/radeon.h @@ -242,6 +242,7 @@ int radeon_object_pin(struct radeon_object *robj, uint32_t domain, uint64_t *gpu_addr); void radeon_object_unpin(struct radeon_object *robj); int radeon_object_wait(struct radeon_object *robj); +int radeon_object_busy_domain(struct radeon_object *robj, uint32_t *cur_placement); int radeon_object_evict_vram(struct radeon_device *rdev); int radeon_object_mmap(struct radeon_object *robj, uint64_t *offset); void radeon_object_force_delete(struct radeon_device *rdev); diff --git a/drivers/gpu/drm/radeon/radeon_gem.c b/drivers/gpu/drm/radeon/radeon_gem.c index cded5180c752..d4ceff13bbb1 100644 --- a/drivers/gpu/drm/radeon/radeon_gem.c +++ b/drivers/gpu/drm/radeon/radeon_gem.c @@ -262,7 +262,27 @@ int radeon_gem_mmap_ioctl(struct drm_device *dev, void *data, int radeon_gem_busy_ioctl(struct drm_device *dev, void *data, struct drm_file *filp) { - /* FIXME: implement */ + struct drm_radeon_gem_busy *args = data; + struct drm_gem_object *gobj; + struct radeon_object *robj; + int r; + uint32_t cur_placement; + + gobj = drm_gem_object_lookup(dev, filp, args->handle); + if (gobj == NULL) { + return -EINVAL; + } + robj = gobj->driver_private; + r = radeon_object_busy_domain(robj, &cur_placement); + if (cur_placement == TTM_PL_VRAM) + args->domain = RADEON_GEM_DOMAIN_VRAM; + if (cur_placement == TTM_PL_FLAG_TT) + args->domain = RADEON_GEM_DOMAIN_GTT; + if (cur_placement == TTM_PL_FLAG_SYSTEM) + args->domain = RADEON_GEM_DOMAIN_CPU; + mutex_lock(&dev->struct_mutex); + drm_gem_object_unreference(gobj); + mutex_unlock(&dev->struct_mutex); return 0; } diff --git a/drivers/gpu/drm/radeon/radeon_object.c b/drivers/gpu/drm/radeon/radeon_object.c index e98cae3bf4a6..b85fb83d7ae8 100644 --- a/drivers/gpu/drm/radeon/radeon_object.c +++ b/drivers/gpu/drm/radeon/radeon_object.c @@ -316,6 +316,25 @@ int radeon_object_wait(struct radeon_object *robj) return r; } +int radeon_object_busy_domain(struct radeon_object *robj, uint32_t *cur_placement) +{ + int r = 0; + + r = radeon_object_reserve(robj, true); + if (unlikely(r != 0)) { + DRM_ERROR("radeon: failed to reserve object for waiting.\n"); + return r; + } + spin_lock(&robj->tobj.lock); + *cur_placement = robj->tobj.mem.mem_type; + if (robj->tobj.sync_obj) { + r = ttm_bo_wait(&robj->tobj, true, true, true); + } + spin_unlock(&robj->tobj.lock); + radeon_object_unreserve(robj); + return r; +} + int radeon_object_evict_vram(struct radeon_device *rdev) { if (rdev->flags & RADEON_IS_IGP) { diff --git a/include/drm/radeon_drm.h b/include/drm/radeon_drm.h index af4b4826997e..f81c3232accd 100644 --- a/include/drm/radeon_drm.h +++ b/include/drm/radeon_drm.h @@ -838,7 +838,7 @@ struct drm_radeon_gem_wait_idle { struct drm_radeon_gem_busy { uint32_t handle; - uint32_t busy; + uint32_t domain; }; struct drm_radeon_gem_pread { -- cgit v1.2.3 From 9c0d90103c7e0eb6e638e5b649e9f6d8d9c1b4b3 Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Fri, 31 Jul 2009 12:53:58 -0400 Subject: Capabilities: move cap_file_mmap to commoncap.c Currently we duplicate the mmap_min_addr test in cap_file_mmap and in security_file_mmap if !CONFIG_SECURITY. This patch moves cap_file_mmap into commoncap.c and then calls that function directly from security_file_mmap ifndef CONFIG_SECURITY like all of the other capability checks are done. Signed-off-by: Eric Paris Acked-by: Serge Hallyn Signed-off-by: James Morris --- include/linux/security.h | 7 ++++--- security/capability.c | 9 --------- security/commoncap.c | 30 ++++++++++++++++++++++++++++++ 3 files changed, 34 insertions(+), 12 deletions(-) (limited to 'include') diff --git a/include/linux/security.h b/include/linux/security.h index 5eff459b3833..ac4bc3760b46 100644 --- a/include/linux/security.h +++ b/include/linux/security.h @@ -66,6 +66,9 @@ extern int cap_inode_setxattr(struct dentry *dentry, const char *name, extern int cap_inode_removexattr(struct dentry *dentry, const char *name); extern int cap_inode_need_killpriv(struct dentry *dentry); extern int cap_inode_killpriv(struct dentry *dentry); +extern int cap_file_mmap(struct file *file, unsigned long reqprot, + unsigned long prot, unsigned long flags, + unsigned long addr, unsigned long addr_only); extern int cap_task_fix_setuid(struct cred *new, const struct cred *old, int flags); extern int cap_task_prctl(int option, unsigned long arg2, unsigned long arg3, unsigned long arg4, unsigned long arg5); @@ -2197,9 +2200,7 @@ static inline int security_file_mmap(struct file *file, unsigned long reqprot, unsigned long addr, unsigned long addr_only) { - if ((addr < mmap_min_addr) && !capable(CAP_SYS_RAWIO)) - return -EACCES; - return 0; + return cap_file_mmap(file, reqprot, prot, flags, addr, addr_only); } static inline int security_file_mprotect(struct vm_area_struct *vma, diff --git a/security/capability.c b/security/capability.c index 21b6cead6a8e..88f752e8152c 100644 --- a/security/capability.c +++ b/security/capability.c @@ -330,15 +330,6 @@ static int cap_file_ioctl(struct file *file, unsigned int command, return 0; } -static int cap_file_mmap(struct file *file, unsigned long reqprot, - unsigned long prot, unsigned long flags, - unsigned long addr, unsigned long addr_only) -{ - if ((addr < mmap_min_addr) && !capable(CAP_SYS_RAWIO)) - return -EACCES; - return 0; -} - static int cap_file_mprotect(struct vm_area_struct *vma, unsigned long reqprot, unsigned long prot) { diff --git a/security/commoncap.c b/security/commoncap.c index 48b7e0228fa3..6bcf6e81e547 100644 --- a/security/commoncap.c +++ b/security/commoncap.c @@ -984,3 +984,33 @@ int cap_vm_enough_memory(struct mm_struct *mm, long pages) cap_sys_admin = 1; return __vm_enough_memory(mm, pages, cap_sys_admin); } + +/* + * cap_file_mmap - check if able to map given addr + * @file: unused + * @reqprot: unused + * @prot: unused + * @flags: unused + * @addr: address attempting to be mapped + * @addr_only: unused + * + * If the process is attempting to map memory below mmap_min_addr they need + * CAP_SYS_RAWIO. The other parameters to this function are unused by the + * capability security module. Returns 0 if this mapping should be allowed + * -EPERM if not. + */ +int cap_file_mmap(struct file *file, unsigned long reqprot, + unsigned long prot, unsigned long flags, + unsigned long addr, unsigned long addr_only) +{ + int ret = 0; + + if (addr < mmap_min_addr) { + ret = cap_capable(current, current_cred(), CAP_SYS_RAWIO, + SECURITY_CAP_AUDIT); + /* set PF_SUPERPRIV if it turns out we allow the low mmap */ + if (ret == 0) + current->flags |= PF_SUPERPRIV; + } + return ret; +} -- cgit v1.2.3 From 788084aba2ab7348257597496befcbccabdc98a3 Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Fri, 31 Jul 2009 12:54:11 -0400 Subject: Security/SELinux: seperate lsm specific mmap_min_addr Currently SELinux enforcement of controls on the ability to map low memory is determined by the mmap_min_addr tunable. This patch causes SELinux to ignore the tunable and instead use a seperate Kconfig option specific to how much space the LSM should protect. The tunable will now only control the need for CAP_SYS_RAWIO and SELinux permissions will always protect the amount of low memory designated by CONFIG_LSM_MMAP_MIN_ADDR. This allows users who need to disable the mmap_min_addr controls (usual reason being they run WINE as a non-root user) to do so and still have SELinux controls preventing confined domains (like a web server) from being able to map some area of low memory. Signed-off-by: Eric Paris Signed-off-by: James Morris --- include/linux/mm.h | 15 --------------- include/linux/security.h | 17 +++++++++++++++++ kernel/sysctl.c | 7 ++++--- mm/Kconfig | 6 +++--- mm/mmap.c | 3 --- mm/nommu.c | 3 --- security/Kconfig | 16 ++++++++++++++++ security/Makefile | 2 +- security/commoncap.c | 2 +- security/min_addr.c | 49 ++++++++++++++++++++++++++++++++++++++++++++++++ security/selinux/hooks.c | 2 +- 11 files changed, 92 insertions(+), 30 deletions(-) create mode 100644 security/min_addr.c (limited to 'include') diff --git a/include/linux/mm.h b/include/linux/mm.h index ba3a7cb1eaa0..9a72cc78e6b8 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -34,8 +34,6 @@ extern int sysctl_legacy_va_layout; #define sysctl_legacy_va_layout 0 #endif -extern unsigned long mmap_min_addr; - #include #include #include @@ -574,19 +572,6 @@ static inline void set_page_links(struct page *page, enum zone_type zone, set_page_section(page, pfn_to_section_nr(pfn)); } -/* - * If a hint addr is less than mmap_min_addr change hint to be as - * low as possible but still greater than mmap_min_addr - */ -static inline unsigned long round_hint_to_min(unsigned long hint) -{ - hint &= PAGE_MASK; - if (((void *)hint != NULL) && - (hint < mmap_min_addr)) - return PAGE_ALIGN(mmap_min_addr); - return hint; -} - /* * Some inline functions in vmstat.h depend on page_zone() */ diff --git a/include/linux/security.h b/include/linux/security.h index ac4bc3760b46..dc3472c1f781 100644 --- a/include/linux/security.h +++ b/include/linux/security.h @@ -28,6 +28,7 @@ #include #include #include +#include /* PAGE_ALIGN */ #include #include #include @@ -95,6 +96,7 @@ extern int cap_netlink_send(struct sock *sk, struct sk_buff *skb); extern int cap_netlink_recv(struct sk_buff *skb, int cap); extern unsigned long mmap_min_addr; +extern unsigned long dac_mmap_min_addr; /* * Values used in the task_security_ops calls */ @@ -147,6 +149,21 @@ static inline void security_free_mnt_opts(struct security_mnt_opts *opts) opts->num_mnt_opts = 0; } +/* + * If a hint addr is less than mmap_min_addr change hint to be as + * low as possible but still greater than mmap_min_addr + */ +static inline unsigned long round_hint_to_min(unsigned long hint) +{ + hint &= PAGE_MASK; + if (((void *)hint != NULL) && + (hint < mmap_min_addr)) + return PAGE_ALIGN(mmap_min_addr); + return hint; +} + +extern int mmap_min_addr_handler(struct ctl_table *table, int write, struct file *filp, + void __user *buffer, size_t *lenp, loff_t *ppos); /** * struct security_operations - main security structure * diff --git a/kernel/sysctl.c b/kernel/sysctl.c index 98e02328c67d..58be76017fd0 100644 --- a/kernel/sysctl.c +++ b/kernel/sysctl.c @@ -49,6 +49,7 @@ #include #include #include +#include #include #include @@ -1306,10 +1307,10 @@ static struct ctl_table vm_table[] = { { .ctl_name = CTL_UNNUMBERED, .procname = "mmap_min_addr", - .data = &mmap_min_addr, - .maxlen = sizeof(unsigned long), + .data = &dac_mmap_min_addr, + .maxlen = sizeof(unsigned long), .mode = 0644, - .proc_handler = &proc_doulongvec_minmax, + .proc_handler = &mmap_min_addr_handler, }, #ifdef CONFIG_NUMA { diff --git a/mm/Kconfig b/mm/Kconfig index c948d4ca8bde..fe5f674d7a7d 100644 --- a/mm/Kconfig +++ b/mm/Kconfig @@ -225,9 +225,9 @@ config DEFAULT_MMAP_MIN_ADDR For most ia64, ppc64 and x86 users with lots of address space a value of 65536 is reasonable and should cause no problems. On arm and other archs it should not be higher than 32768. - Programs which use vm86 functionality would either need additional - permissions from either the LSM or the capabilities module or have - this protection disabled. + Programs which use vm86 functionality or have some need to map + this low address space will need CAP_SYS_RAWIO or disable this + protection by setting the value to 0. This value can be changed after boot using the /proc/sys/vm/mmap_min_addr tunable. diff --git a/mm/mmap.c b/mm/mmap.c index 34579b23ebd5..8101de490c73 100644 --- a/mm/mmap.c +++ b/mm/mmap.c @@ -88,9 +88,6 @@ int sysctl_overcommit_ratio = 50; /* default is 50% */ int sysctl_max_map_count __read_mostly = DEFAULT_MAX_MAP_COUNT; struct percpu_counter vm_committed_as; -/* amount of vm to protect from userspace access */ -unsigned long mmap_min_addr = CONFIG_DEFAULT_MMAP_MIN_ADDR; - /* * Check that a process has enough memory to allocate a new virtual * mapping. 0 means there is enough memory for the allocation to diff --git a/mm/nommu.c b/mm/nommu.c index 53cab10fece4..28754c40be98 100644 --- a/mm/nommu.c +++ b/mm/nommu.c @@ -69,9 +69,6 @@ int sysctl_max_map_count = DEFAULT_MAX_MAP_COUNT; int sysctl_nr_trim_pages = CONFIG_NOMMU_INITIAL_TRIM_EXCESS; int heap_stack_gap = 0; -/* amount of vm to protect from userspace access */ -unsigned long mmap_min_addr = CONFIG_DEFAULT_MMAP_MIN_ADDR; - atomic_long_t mmap_pages_allocated; EXPORT_SYMBOL(mem_map); diff --git a/security/Kconfig b/security/Kconfig index d23c839038f0..9c60c346a91d 100644 --- a/security/Kconfig +++ b/security/Kconfig @@ -113,6 +113,22 @@ config SECURITY_ROOTPLUG If you are unsure how to answer this question, answer N. +config LSM_MMAP_MIN_ADDR + int "Low address space for LSM to from user allocation" + depends on SECURITY && SECURITY_SELINUX + default 65535 + help + This is the portion of low virtual memory which should be protected + from userspace allocation. Keeping a user from writing to low pages + can help reduce the impact of kernel NULL pointer bugs. + + For most ia64, ppc64 and x86 users with lots of address space + a value of 65536 is reasonable and should cause no problems. + On arm and other archs it should not be higher than 32768. + Programs which use vm86 functionality or have some need to map + this low address space will need the permission specific to the + systems running LSM. + source security/selinux/Kconfig source security/smack/Kconfig source security/tomoyo/Kconfig diff --git a/security/Makefile b/security/Makefile index c67557cdaa85..b56e7f9ecbc2 100644 --- a/security/Makefile +++ b/security/Makefile @@ -8,7 +8,7 @@ subdir-$(CONFIG_SECURITY_SMACK) += smack subdir-$(CONFIG_SECURITY_TOMOYO) += tomoyo # always enable default capabilities -obj-y += commoncap.o +obj-y += commoncap.o min_addr.o # Object file lists obj-$(CONFIG_SECURITY) += security.o capability.o diff --git a/security/commoncap.c b/security/commoncap.c index 6bcf6e81e547..e3097c0a1311 100644 --- a/security/commoncap.c +++ b/security/commoncap.c @@ -1005,7 +1005,7 @@ int cap_file_mmap(struct file *file, unsigned long reqprot, { int ret = 0; - if (addr < mmap_min_addr) { + if (addr < dac_mmap_min_addr) { ret = cap_capable(current, current_cred(), CAP_SYS_RAWIO, SECURITY_CAP_AUDIT); /* set PF_SUPERPRIV if it turns out we allow the low mmap */ diff --git a/security/min_addr.c b/security/min_addr.c new file mode 100644 index 000000000000..14cc7b3b8d03 --- /dev/null +++ b/security/min_addr.c @@ -0,0 +1,49 @@ +#include +#include +#include +#include + +/* amount of vm to protect from userspace access by both DAC and the LSM*/ +unsigned long mmap_min_addr; +/* amount of vm to protect from userspace using CAP_SYS_RAWIO (DAC) */ +unsigned long dac_mmap_min_addr = CONFIG_DEFAULT_MMAP_MIN_ADDR; +/* amount of vm to protect from userspace using the LSM = CONFIG_LSM_MMAP_MIN_ADDR */ + +/* + * Update mmap_min_addr = max(dac_mmap_min_addr, CONFIG_LSM_MMAP_MIN_ADDR) + */ +static void update_mmap_min_addr(void) +{ +#ifdef CONFIG_LSM_MMAP_MIN_ADDR + if (dac_mmap_min_addr > CONFIG_LSM_MMAP_MIN_ADDR) + mmap_min_addr = dac_mmap_min_addr; + else + mmap_min_addr = CONFIG_LSM_MMAP_MIN_ADDR; +#else + mmap_min_addr = dac_mmap_min_addr; +#endif +} + +/* + * sysctl handler which just sets dac_mmap_min_addr = the new value and then + * calls update_mmap_min_addr() so non MAP_FIXED hints get rounded properly + */ +int mmap_min_addr_handler(struct ctl_table *table, int write, struct file *filp, + void __user *buffer, size_t *lenp, loff_t *ppos) +{ + int ret; + + ret = proc_doulongvec_minmax(table, write, filp, buffer, lenp, ppos); + + update_mmap_min_addr(); + + return ret; +} + +int __init init_mmap_min_addr(void) +{ + update_mmap_min_addr(); + + return 0; +} +pure_initcall(init_mmap_min_addr); diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c index e6d1432b0800..8d8b69c5664e 100644 --- a/security/selinux/hooks.c +++ b/security/selinux/hooks.c @@ -3036,7 +3036,7 @@ static int selinux_file_mmap(struct file *file, unsigned long reqprot, * at bad behaviour/exploit that we always want to get the AVC, even * if DAC would have also denied the operation. */ - if (addr < mmap_min_addr) { + if (addr < CONFIG_LSM_MMAP_MIN_ADDR) { rc = avc_has_perm(sid, sid, SECCLASS_MEMPROTECT, MEMPROTECT__MMAP_ZERO, NULL); if (rc) -- cgit v1.2.3 From 1d9959734a1949ea4f2427bd2d8b21ede6b2441c Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Fri, 7 Aug 2009 14:53:57 -0400 Subject: security: define round_hint_to_min in !CONFIG_SECURITY Fix the header files to define round_hint_to_min() and to define mmap_min_addr_handler() in the !CONFIG_SECURITY case. Built and tested with !CONFIG_SECURITY Signed-off-by: Eric Paris Signed-off-by: James Morris --- include/linux/security.h | 30 +++++++++++++++--------------- 1 file changed, 15 insertions(+), 15 deletions(-) (limited to 'include') diff --git a/include/linux/security.h b/include/linux/security.h index dc3472c1f781..1f16eea2017b 100644 --- a/include/linux/security.h +++ b/include/linux/security.h @@ -121,6 +121,21 @@ struct request_sock; #define LSM_UNSAFE_PTRACE 2 #define LSM_UNSAFE_PTRACE_CAP 4 +/* + * If a hint addr is less than mmap_min_addr change hint to be as + * low as possible but still greater than mmap_min_addr + */ +static inline unsigned long round_hint_to_min(unsigned long hint) +{ + hint &= PAGE_MASK; + if (((void *)hint != NULL) && + (hint < mmap_min_addr)) + return PAGE_ALIGN(mmap_min_addr); + return hint; +} +extern int mmap_min_addr_handler(struct ctl_table *table, int write, struct file *filp, + void __user *buffer, size_t *lenp, loff_t *ppos); + #ifdef CONFIG_SECURITY struct security_mnt_opts { @@ -149,21 +164,6 @@ static inline void security_free_mnt_opts(struct security_mnt_opts *opts) opts->num_mnt_opts = 0; } -/* - * If a hint addr is less than mmap_min_addr change hint to be as - * low as possible but still greater than mmap_min_addr - */ -static inline unsigned long round_hint_to_min(unsigned long hint) -{ - hint &= PAGE_MASK; - if (((void *)hint != NULL) && - (hint < mmap_min_addr)) - return PAGE_ALIGN(mmap_min_addr); - return hint; -} - -extern int mmap_min_addr_handler(struct ctl_table *table, int write, struct file *filp, - void __user *buffer, size_t *lenp, loff_t *ppos); /** * struct security_operations - main security structure * -- cgit v1.2.3 From b25c340c195447afb1860da580fe2a85a6b652c5 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Thu, 13 Aug 2009 12:17:22 +0200 Subject: genirq: Add oneshot support For threaded interrupt handlers we expect the hard interrupt handler part to mask the interrupt on the originating device. The interrupt line itself is reenabled after the hard interrupt handler has executed. This requires access to the originating device from hard interrupt context which is not always possible. There are devices which can only be accessed via a bus (i2c, spi, ...). The bus access requires thread context. For such devices we need to keep the interrupt line masked until the threaded handler has executed. Add a new flag IRQF_ONESHOT which allows drivers to request that the interrupt is not unmasked after the hard interrupt context handler has been executed and the thread has been woken. The interrupt line is unmasked after the thread handler function has been executed. Note that for now IRQF_ONESHOT cannot be used with IRQF_SHARED to avoid complex accounting mechanisms. For oneshot interrupts the primary handler simply returns IRQ_WAKE_THREAD and does nothing else. A generic implementation irq_default_primary_handler() is provided to avoid useless copies all over the place. It is automatically installed when request_threaded_irq() is called with handler=NULL and thread_fn!=NULL. Signed-off-by: Thomas Gleixner Cc: Mark Brown Cc: Dmitry Torokhov Cc: Trilok Soni Cc: Pavel Machek Cc: Brian Swetland Cc: Joonyoung Shim Cc: m.szyprowski@samsung.com Cc: t.fujak@samsung.com Cc: kyungmin.park@samsung.com, Cc: David Brownell Cc: Daniel Ribeiro Cc: arve@android.com Cc: Barry Song <21cnbao@gmail.com> --- include/linux/interrupt.h | 4 ++++ include/linux/irq.h | 1 + kernel/irq/chip.c | 14 +++++++++++--- kernel/irq/manage.c | 49 +++++++++++++++++++++++++++++++++++++++++++---- 4 files changed, 61 insertions(+), 7 deletions(-) (limited to 'include') diff --git a/include/linux/interrupt.h b/include/linux/interrupt.h index 35e7df1e9f30..1ac57e522a1f 100644 --- a/include/linux/interrupt.h +++ b/include/linux/interrupt.h @@ -50,6 +50,9 @@ * IRQF_IRQPOLL - Interrupt is used for polling (only the interrupt that is * registered first in an shared interrupt is considered for * performance reasons) + * IRQF_ONESHOT - Interrupt is not reenabled after the hardirq handler finished. + * Used by threaded interrupts which need to keep the + * irq line disabled until the threaded handler has been run. */ #define IRQF_DISABLED 0x00000020 #define IRQF_SAMPLE_RANDOM 0x00000040 @@ -59,6 +62,7 @@ #define IRQF_PERCPU 0x00000400 #define IRQF_NOBALANCING 0x00000800 #define IRQF_IRQPOLL 0x00001000 +#define IRQF_ONESHOT 0x00002000 /* * Bits used by threaded handlers: diff --git a/include/linux/irq.h b/include/linux/irq.h index cb2e77a3f7f7..5e7c6ee8c35c 100644 --- a/include/linux/irq.h +++ b/include/linux/irq.h @@ -69,6 +69,7 @@ typedef void (*irq_flow_handler_t)(unsigned int irq, #define IRQ_MOVE_PCNTXT 0x01000000 /* IRQ migration from process context */ #define IRQ_AFFINITY_SET 0x02000000 /* IRQ affinity was set from userspace*/ #define IRQ_SUSPENDED 0x04000000 /* IRQ has gone through suspend sequence */ +#define IRQ_ONESHOT 0x08000000 /* IRQ is not unmasked after hardirq */ #ifdef CONFIG_IRQ_PER_CPU # define CHECK_IRQ_PER_CPU(var) ((var) & IRQ_PER_CPU) diff --git a/kernel/irq/chip.c b/kernel/irq/chip.c index 13c68e71b726..b08c0d24f202 100644 --- a/kernel/irq/chip.c +++ b/kernel/irq/chip.c @@ -382,7 +382,10 @@ handle_level_irq(unsigned int irq, struct irq_desc *desc) spin_lock(&desc->lock); desc->status &= ~IRQ_INPROGRESS; - if (!(desc->status & IRQ_DISABLED) && desc->chip->unmask) + + if (unlikely(desc->status & IRQ_ONESHOT)) + desc->status |= IRQ_MASKED; + else if (!(desc->status & IRQ_DISABLED) && desc->chip->unmask) desc->chip->unmask(irq); out_unlock: spin_unlock(&desc->lock); @@ -478,8 +481,13 @@ handle_edge_irq(unsigned int irq, struct irq_desc *desc) kstat_incr_irqs_this_cpu(irq, desc); /* Start handling the irq */ - if (desc->chip->ack) - desc->chip->ack(irq); + if (unlikely(desc->status & IRQ_ONESHOT)) { + desc->status |= IRQ_MASKED; + mask_ack_irq(desc, irq); + } else { + if (desc->chip->ack) + desc->chip->ack(irq); + } /* Mark the IRQ currently in progress.*/ desc->status |= IRQ_INPROGRESS; diff --git a/kernel/irq/manage.c b/kernel/irq/manage.c index d222515a5a06..d7f7b5fd2476 100644 --- a/kernel/irq/manage.c +++ b/kernel/irq/manage.c @@ -436,6 +436,16 @@ int __irq_set_trigger(struct irq_desc *desc, unsigned int irq, return ret; } +/* + * Default primary interrupt handler for threaded interrupts. Is + * assigned as primary handler when request_threaded_irq is called + * with handler == NULL. Useful for oneshot interrupts. + */ +static irqreturn_t irq_default_primary_handler(int irq, void *dev_id) +{ + return IRQ_WAKE_THREAD; +} + static int irq_wait_for_interrupt(struct irqaction *action) { while (!kthread_should_stop()) { @@ -451,6 +461,21 @@ static int irq_wait_for_interrupt(struct irqaction *action) return -1; } +/* + * Oneshot interrupts keep the irq line masked until the threaded + * handler finished. unmask if the interrupt has not been disabled and + * is marked MASKED. + */ +static void irq_finalize_oneshot(unsigned int irq, struct irq_desc *desc) +{ + spin_lock_irq(&desc->lock); + if (!(desc->status & IRQ_DISABLED) && (desc->status & IRQ_MASKED)) { + desc->status &= ~IRQ_MASKED; + desc->chip->unmask(irq); + } + spin_unlock_irq(&desc->lock); +} + #ifdef CONFIG_SMP /* * Check whether we need to change the affinity of the interrupt thread. @@ -492,7 +517,7 @@ static int irq_thread(void *data) struct sched_param param = { .sched_priority = MAX_USER_RT_PRIO/2, }; struct irqaction *action = data; struct irq_desc *desc = irq_to_desc(action->irq); - int wake; + int wake, oneshot = desc->status & IRQ_ONESHOT; sched_setscheduler(current, SCHED_FIFO, ¶m); current->irqaction = action; @@ -518,6 +543,9 @@ static int irq_thread(void *data) spin_unlock_irq(&desc->lock); action->thread_fn(action->irq, action->dev_id); + + if (oneshot) + irq_finalize_oneshot(action->irq, desc); } wake = atomic_dec_and_test(&desc->threads_active); @@ -590,6 +618,10 @@ __setup_irq(unsigned int irq, struct irq_desc *desc, struct irqaction *new) rand_initialize_irq(irq); } + /* Oneshot interrupts are not allowed with shared */ + if ((new->flags & IRQF_ONESHOT) && (new->flags & IRQF_SHARED)) + return -EINVAL; + /* * Threaded handler ? */ @@ -663,9 +695,12 @@ __setup_irq(unsigned int irq, struct irq_desc *desc, struct irqaction *new) desc->status |= IRQ_PER_CPU; #endif - desc->status &= ~(IRQ_AUTODETECT | IRQ_WAITING | + desc->status &= ~(IRQ_AUTODETECT | IRQ_WAITING | IRQ_ONESHOT | IRQ_INPROGRESS | IRQ_SPURIOUS_DISABLED); + if (new->flags & IRQF_ONESHOT) + desc->status |= IRQ_ONESHOT; + if (!(desc->status & IRQ_NOAUTOEN)) { desc->depth = 0; desc->status &= ~IRQ_DISABLED; @@ -878,6 +913,8 @@ EXPORT_SYMBOL(free_irq); * @irq: Interrupt line to allocate * @handler: Function to be called when the IRQ occurs. * Primary handler for threaded interrupts + * If NULL and thread_fn != NULL the default + * primary handler is installed * @thread_fn: Function called from the irq handler thread * If NULL, no irq thread is created * @irqflags: Interrupt type flags @@ -957,8 +994,12 @@ int request_threaded_irq(unsigned int irq, irq_handler_t handler, if (desc->status & IRQ_NOREQUEST) return -EINVAL; - if (!handler) - return -EINVAL; + + if (!handler) { + if (!thread_fn) + return -EINVAL; + handler = irq_default_primary_handler; + } action = kzalloc(sizeof(struct irqaction), GFP_KERNEL); if (!action) -- cgit v1.2.3 From 70aedd24d20e75198f5a0b11750faabbb56924e2 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Thu, 13 Aug 2009 12:17:48 +0200 Subject: genirq: Add buslock support Some interrupt chips are connected to a "slow" bus (i2c, spi ...). The bus access needs to sleep and therefor cannot be called in atomic contexts. Some of the generic interrupt management functions like disable_irq(), enable_irq() ... call interrupt chip functions with the irq_desc->lock held and interrupts disabled. This does not work for such devices. Provide a separate synchronization mechanism for such interrupt chips. The irq_chip structure is extended by two optional functions (bus_lock and bus_sync_and_unlock). The idea is to serialize the bus access for those operations in the core code so that drivers which are behind that bus operated interrupt controller do not have to worry about it and just can use the normal interfaces. To achieve this we add two function pointers to the irq_chip: bus_lock and bus_sync_unlock. bus_lock() is called to serialize access to the interrupt controller bus. Now the core code can issue chip->mask/unmask ... commands without changing the fast path code at all. The chip implementation merily stores that information in a chip private data structure and returns. No bus interaction as these functions are called from atomic context. After that bus_sync_unlock() is called outside the atomic context. Now the chip implementation issues the bus commands, waits for completion and unlocks the interrupt controller bus. The irq_chip implementation as pseudo code: struct irq_chip_data { struct mutex mutex; unsigned int irq_offset; unsigned long mask; unsigned long mask_status; } static void bus_lock(unsigned int irq) { struct irq_chip_data *data = get_irq_desc_chip_data(irq); mutex_lock(&data->mutex); } static void mask(unsigned int irq) { struct irq_chip_data *data = get_irq_desc_chip_data(irq); irq -= data->irq_offset; data->mask |= (1 << irq); } static void unmask(unsigned int irq) { struct irq_chip_data *data = get_irq_desc_chip_data(irq); irq -= data->irq_offset; data->mask &= ~(1 << irq); } static void bus_sync_unlock(unsigned int irq) { struct irq_chip_data *data = get_irq_desc_chip_data(irq); if (data->mask != data->mask_status) { do_bus_magic_to_set_mask(data->mask); data->mask_status = data->mask; } mutex_unlock(&data->mutex); } The device drivers can use request_threaded_irq, free_irq, disable_irq and enable_irq as usual with the only restriction that the calls need to come from non atomic context. Signed-off-by: Thomas Gleixner Cc: Mark Brown Cc: Dmitry Torokhov Cc: Trilok Soni Cc: Pavel Machek Cc: Brian Swetland Cc: Joonyoung Shim Cc: m.szyprowski@samsung.com Cc: t.fujak@samsung.com Cc: kyungmin.park@samsung.com, Cc: David Brownell Cc: Daniel Ribeiro Cc: arve@android.com Cc: Barry Song <21cnbao@gmail.com> --- include/linux/irq.h | 6 ++++++ kernel/irq/chip.c | 2 ++ kernel/irq/internals.h | 13 +++++++++++++ kernel/irq/manage.c | 19 ++++++++++++++++++- 4 files changed, 39 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/irq.h b/include/linux/irq.h index 5e7c6ee8c35c..ce8171bc6fac 100644 --- a/include/linux/irq.h +++ b/include/linux/irq.h @@ -101,6 +101,9 @@ struct msi_desc; * @set_type: set the flow type (IRQ_TYPE_LEVEL/etc.) of an IRQ * @set_wake: enable/disable power-management wake-on of an IRQ * + * @bus_lock: function to lock access to slow bus (i2c) chips + * @bus_sync_unlock: function to sync and unlock slow bus (i2c) chips + * * @release: release function solely used by UML * @typename: obsoleted by name, kept as migration helper */ @@ -124,6 +127,9 @@ struct irq_chip { int (*set_type)(unsigned int irq, unsigned int flow_type); int (*set_wake)(unsigned int irq, unsigned int on); + void (*bus_lock)(unsigned int irq); + void (*bus_sync_unlock)(unsigned int irq); + /* Currently used only by UML, might disappear one day.*/ #ifdef CONFIG_IRQ_RELEASE_METHOD void (*release)(unsigned int irq, void *dev_id); diff --git a/kernel/irq/chip.c b/kernel/irq/chip.c index b08c0d24f202..f856330e684a 100644 --- a/kernel/irq/chip.c +++ b/kernel/irq/chip.c @@ -580,6 +580,7 @@ __set_irq_handler(unsigned int irq, irq_flow_handler_t handle, int is_chained, desc->chip = &dummy_irq_chip; } + chip_bus_lock(irq, desc); spin_lock_irqsave(&desc->lock, flags); /* Uninstall? */ @@ -599,6 +600,7 @@ __set_irq_handler(unsigned int irq, irq_flow_handler_t handle, int is_chained, desc->chip->startup(irq); } spin_unlock_irqrestore(&desc->lock, flags); + chip_bus_sync_unlock(irq, desc); } EXPORT_SYMBOL_GPL(__set_irq_handler); diff --git a/kernel/irq/internals.h b/kernel/irq/internals.h index e70ed5592eb9..1b5d742c6a77 100644 --- a/kernel/irq/internals.h +++ b/kernel/irq/internals.h @@ -44,6 +44,19 @@ extern int irq_select_affinity_usr(unsigned int irq); extern void irq_set_thread_affinity(struct irq_desc *desc); +/* Inline functions for support of irq chips on slow busses */ +static inline void chip_bus_lock(unsigned int irq, struct irq_desc *desc) +{ + if (unlikely(desc->chip->bus_lock)) + desc->chip->bus_lock(irq); +} + +static inline void chip_bus_sync_unlock(unsigned int irq, struct irq_desc *desc) +{ + if (unlikely(desc->chip->bus_sync_unlock)) + desc->chip->bus_sync_unlock(irq); +} + /* * Debugging printout: */ diff --git a/kernel/irq/manage.c b/kernel/irq/manage.c index d7f7b5fd2476..0a3fd5b524c9 100644 --- a/kernel/irq/manage.c +++ b/kernel/irq/manage.c @@ -230,9 +230,11 @@ void disable_irq_nosync(unsigned int irq) if (!desc) return; + chip_bus_lock(irq, desc); spin_lock_irqsave(&desc->lock, flags); __disable_irq(desc, irq, false); spin_unlock_irqrestore(&desc->lock, flags); + chip_bus_sync_unlock(irq, desc); } EXPORT_SYMBOL(disable_irq_nosync); @@ -294,7 +296,8 @@ void __enable_irq(struct irq_desc *desc, unsigned int irq, bool resume) * matches the last disable, processing of interrupts on this * IRQ line is re-enabled. * - * This function may be called from IRQ context. + * This function may be called from IRQ context only when + * desc->chip->bus_lock and desc->chip->bus_sync_unlock are NULL ! */ void enable_irq(unsigned int irq) { @@ -304,9 +307,11 @@ void enable_irq(unsigned int irq) if (!desc) return; + chip_bus_lock(irq, desc); spin_lock_irqsave(&desc->lock, flags); __enable_irq(desc, irq, false); spin_unlock_irqrestore(&desc->lock, flags); + chip_bus_sync_unlock(irq, desc); } EXPORT_SYMBOL(enable_irq); @@ -468,12 +473,14 @@ static int irq_wait_for_interrupt(struct irqaction *action) */ static void irq_finalize_oneshot(unsigned int irq, struct irq_desc *desc) { + chip_bus_lock(irq, desc); spin_lock_irq(&desc->lock); if (!(desc->status & IRQ_DISABLED) && (desc->status & IRQ_MASKED)) { desc->status &= ~IRQ_MASKED; desc->chip->unmask(irq); } spin_unlock_irq(&desc->lock); + chip_bus_sync_unlock(irq, desc); } #ifdef CONFIG_SMP @@ -904,7 +911,14 @@ EXPORT_SYMBOL_GPL(remove_irq); */ void free_irq(unsigned int irq, void *dev_id) { + struct irq_desc *desc = irq_to_desc(irq); + + if (!desc) + return; + + chip_bus_lock(irq, desc); kfree(__free_irq(irq, dev_id)); + chip_bus_sync_unlock(irq, desc); } EXPORT_SYMBOL(free_irq); @@ -1011,7 +1025,10 @@ int request_threaded_irq(unsigned int irq, irq_handler_t handler, action->name = devname; action->dev_id = dev_id; + chip_bus_lock(irq, desc); retval = __setup_irq(irq, desc, action); + chip_bus_sync_unlock(irq, desc); + if (retval) kfree(action); -- cgit v1.2.3 From 399b5da29b9f851eb7b96e2882097127f003e87c Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Thu, 13 Aug 2009 13:21:38 +0200 Subject: genirq: Support nested threaded irq handling Interrupt chips which are behind a slow bus (i2c, spi ...) and demultiplex other interrupt sources need to run their interrupt handler in a thread. The demultiplexed interrupt handlers need to run in thread context as well and need to finish before the demux handler thread can reenable the interrupt line. So the easiest way is to run the sub device handlers in the context of the demultiplexing handler thread. To avoid that a separate thread is created for the subdevices the function set_nested_irq_thread() is provided which sets the IRQ_NESTED_THREAD flag in the interrupt descriptor. A driver which calls request_threaded_irq() must not be aware of the fact that the threaded handler is called in the context of the demultiplexing handler thread. The setup code checks the IRQ_NESTED_THREAD flag which was set from the irq chip setup code and does not setup a separate thread for the interrupt. The primary function which is provided by the device driver is replaced by an internal dummy function which warns when it is called. For the demultiplexing handler a helper function handle_nested_irq() is provided which calls the demux interrupt thread function in the context of the caller and does the proper interrupt accounting and takes the interrupt disabled status of the demultiplexed subdevice into account. Signed-off-by: Thomas Gleixner Cc: Mark Brown Cc: Dmitry Torokhov Cc: Trilok Soni Cc: Pavel Machek Cc: Brian Swetland Cc: Joonyoung Shim Cc: m.szyprowski@samsung.com Cc: t.fujak@samsung.com Cc: kyungmin.park@samsung.com, Cc: David Brownell Cc: Daniel Ribeiro Cc: arve@android.com Cc: Barry Song <21cnbao@gmail.com> --- include/linux/irq.h | 3 +++ kernel/irq/chip.c | 67 +++++++++++++++++++++++++++++++++++++++++++++++++++++ kernel/irq/manage.c | 34 ++++++++++++++++++++++++--- 3 files changed, 101 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/include/linux/irq.h b/include/linux/irq.h index ce8171bc6fac..8778ee993937 100644 --- a/include/linux/irq.h +++ b/include/linux/irq.h @@ -70,6 +70,7 @@ typedef void (*irq_flow_handler_t)(unsigned int irq, #define IRQ_AFFINITY_SET 0x02000000 /* IRQ affinity was set from userspace*/ #define IRQ_SUSPENDED 0x04000000 /* IRQ has gone through suspend sequence */ #define IRQ_ONESHOT 0x08000000 /* IRQ is not unmasked after hardirq */ +#define IRQ_NESTED_THREAD 0x10000000 /* IRQ is nested into another, no own handler thread */ #ifdef CONFIG_IRQ_PER_CPU # define CHECK_IRQ_PER_CPU(var) ((var) & IRQ_PER_CPU) @@ -386,6 +387,8 @@ set_irq_chained_handler(unsigned int irq, __set_irq_handler(irq, handle, 1, NULL); } +extern void set_irq_nested_thread(unsigned int irq, int nest); + extern void set_irq_noprobe(unsigned int irq); extern void set_irq_probe(unsigned int irq); diff --git a/kernel/irq/chip.c b/kernel/irq/chip.c index f856330e684a..5765aad94998 100644 --- a/kernel/irq/chip.c +++ b/kernel/irq/chip.c @@ -222,6 +222,34 @@ int set_irq_chip_data(unsigned int irq, void *data) } EXPORT_SYMBOL(set_irq_chip_data); +/** + * set_irq_nested_thread - Set/Reset the IRQ_NESTED_THREAD flag of an irq + * + * @irq: Interrupt number + * @nest: 0 to clear / 1 to set the IRQ_NESTED_THREAD flag + * + * The IRQ_NESTED_THREAD flag indicates that on + * request_threaded_irq() no separate interrupt thread should be + * created for the irq as the handler are called nested in the + * context of a demultiplexing interrupt handler thread. + */ +void set_irq_nested_thread(unsigned int irq, int nest) +{ + struct irq_desc *desc = irq_to_desc(irq); + unsigned long flags; + + if (!desc) + return; + + spin_lock_irqsave(&desc->lock, flags); + if (nest) + desc->status |= IRQ_NESTED_THREAD; + else + desc->status &= ~IRQ_NESTED_THREAD; + spin_unlock_irqrestore(&desc->lock, flags); +} +EXPORT_SYMBOL_GPL(set_irq_nested_thread); + /* * default enable function */ @@ -299,6 +327,45 @@ static inline void mask_ack_irq(struct irq_desc *desc, int irq) } } +/* + * handle_nested_irq - Handle a nested irq from a irq thread + * @irq: the interrupt number + * + * Handle interrupts which are nested into a threaded interrupt + * handler. The handler function is called inside the calling + * threads context. + */ +void handle_nested_irq(unsigned int irq) +{ + struct irq_desc *desc = irq_to_desc(irq); + struct irqaction *action; + irqreturn_t action_ret; + + might_sleep(); + + spin_lock_irq(&desc->lock); + + kstat_incr_irqs_this_cpu(irq, desc); + + action = desc->action; + if (unlikely(!action || (desc->status & IRQ_DISABLED))) + goto out_unlock; + + desc->status |= IRQ_INPROGRESS; + spin_unlock_irq(&desc->lock); + + action_ret = action->thread_fn(action->irq, action->dev_id); + if (!noirqdebug) + note_interrupt(irq, desc, action_ret); + + spin_lock_irq(&desc->lock); + desc->status &= ~IRQ_INPROGRESS; + +out_unlock: + spin_unlock_irq(&desc->lock); +} +EXPORT_SYMBOL_GPL(handle_nested_irq); + /** * handle_simple_irq - Simple and software-decoded IRQs. * @irq: the interrupt number diff --git a/kernel/irq/manage.c b/kernel/irq/manage.c index 0a3fd5b524c9..e485cea04bf1 100644 --- a/kernel/irq/manage.c +++ b/kernel/irq/manage.c @@ -451,6 +451,16 @@ static irqreturn_t irq_default_primary_handler(int irq, void *dev_id) return IRQ_WAKE_THREAD; } +/* + * Primary handler for nested threaded interrupts. Should never be + * called. + */ +static irqreturn_t irq_nested_primary_handler(int irq, void *dev_id) +{ + WARN(1, "Primary handler called for nested irq %d\n", irq); + return IRQ_NONE; +} + static int irq_wait_for_interrupt(struct irqaction *action) { while (!kthread_should_stop()) { @@ -600,7 +610,7 @@ __setup_irq(unsigned int irq, struct irq_desc *desc, struct irqaction *new) struct irqaction *old, **old_ptr; const char *old_name = NULL; unsigned long flags; - int shared = 0; + int nested, shared = 0; int ret; if (!desc) @@ -630,9 +640,27 @@ __setup_irq(unsigned int irq, struct irq_desc *desc, struct irqaction *new) return -EINVAL; /* - * Threaded handler ? + * Check whether the interrupt nests into another interrupt + * thread. + */ + nested = desc->status & IRQ_NESTED_THREAD; + if (nested) { + if (!new->thread_fn) + return -EINVAL; + /* + * Replace the primary handler which was provided from + * the driver for non nested interrupt handling by the + * dummy function which warns when called. + */ + new->handler = irq_nested_primary_handler; + } + + /* + * Create a handler thread when a thread function is supplied + * and the interrupt does not nest into another interrupt + * thread. */ - if (new->thread_fn) { + if (new->thread_fn && !nested) { struct task_struct *t; t = kthread_create(irq_thread, new, "irq/%d-%s", irq, -- cgit v1.2.3 From 7ead8b8313d92b3a69a1a61b0dcbc4cd66c960dc Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Mon, 17 Aug 2009 16:56:28 +0800 Subject: tracing/events: Add module tracepoints Add trace points to trace module_load, module_free, module_get, module_put and module_request, and use trace_event facility to get the trace output. Here's the sample output: TASK-PID CPU# TIMESTAMP FUNCTION | | | | | <...>-42 [000] 1.758380: module_request: fb0 wait=1 call_site=fb_open ... <...>-60 [000] 3.269403: module_load: scsi_wait_scan <...>-60 [000] 3.269432: module_put: scsi_wait_scan call_site=sys_init_module refcnt=0 <...>-61 [001] 3.273168: module_free: scsi_wait_scan ... <...>-1021 [000] 13.836081: module_load: sunrpc <...>-1021 [000] 13.840589: module_put: sunrpc call_site=sys_init_module refcnt=-1 <...>-1027 [000] 13.848098: module_get: sunrpc call_site=try_module_get refcnt=0 <...>-1027 [000] 13.848308: module_get: sunrpc call_site=get_filesystem refcnt=1 <...>-1027 [000] 13.848692: module_put: sunrpc call_site=put_filesystem refcnt=0 ... modprobe-2587 [001] 1088.437213: module_load: trace_events_sample F modprobe-2587 [001] 1088.437786: module_put: trace_events_sample call_site=sys_init_module refcnt=0 Note: - the taints flag can be 'F', 'C' and/or 'P' if mod->taints != 0 - the module refcnt is percpu, so it can be negative in a specific cpu Signed-off-by: Li Zefan Acked-by: Rusty Russell Cc: Steven Rostedt Cc: Frederic Weisbecker Cc: Rusty Russell LKML-Reference: <4A891B3C.5030608@cn.fujitsu.com> Signed-off-by: Ingo Molnar --- include/linux/module.h | 14 ++++- include/trace/events/module.h | 126 ++++++++++++++++++++++++++++++++++++++++++ kernel/kmod.c | 4 ++ kernel/module.c | 11 ++++ 4 files changed, 152 insertions(+), 3 deletions(-) create mode 100644 include/trace/events/module.h (limited to 'include') diff --git a/include/linux/module.h b/include/linux/module.h index 098bdb7bfacf..f8f92d015efe 100644 --- a/include/linux/module.h +++ b/include/linux/module.h @@ -17,10 +17,12 @@ #include #include #include -#include +#include #include +#include + /* Not Yet Implemented */ #define MODULE_SUPPORTED_DEVICE(name) @@ -462,7 +464,10 @@ static inline local_t *__module_ref_addr(struct module *mod, int cpu) static inline void __module_get(struct module *module) { if (module) { - local_inc(__module_ref_addr(module, get_cpu())); + unsigned int cpu = get_cpu(); + local_inc(__module_ref_addr(module, cpu)); + trace_module_get(module, _THIS_IP_, + local_read(__module_ref_addr(module, cpu))); put_cpu(); } } @@ -473,8 +478,11 @@ static inline int try_module_get(struct module *module) if (module) { unsigned int cpu = get_cpu(); - if (likely(module_is_live(module))) + if (likely(module_is_live(module))) { local_inc(__module_ref_addr(module, cpu)); + trace_module_get(module, _THIS_IP_, + local_read(__module_ref_addr(module, cpu))); + } else ret = 0; put_cpu(); diff --git a/include/trace/events/module.h b/include/trace/events/module.h new file mode 100644 index 000000000000..84160fb18478 --- /dev/null +++ b/include/trace/events/module.h @@ -0,0 +1,126 @@ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM module + +#if !defined(_TRACE_MODULE_H) || defined(TRACE_HEADER_MULTI_READ) +#define _TRACE_MODULE_H + +#include + +#ifdef CONFIG_MODULES + +struct module; + +#define show_module_flags(flags) __print_flags(flags, "", \ + { (1UL << TAINT_PROPRIETARY_MODULE), "P" }, \ + { (1UL << TAINT_FORCED_MODULE), "F" }, \ + { (1UL << TAINT_CRAP), "C" }) + +TRACE_EVENT(module_load, + + TP_PROTO(struct module *mod), + + TP_ARGS(mod), + + TP_STRUCT__entry( + __field( unsigned int, taints ) + __string( name, mod->name ) + ), + + TP_fast_assign( + __entry->taints = mod->taints; + __assign_str(name, mod->name); + ), + + TP_printk("%s %s", __get_str(name), show_module_flags(__entry->taints)) +); + +TRACE_EVENT(module_free, + + TP_PROTO(struct module *mod), + + TP_ARGS(mod), + + TP_STRUCT__entry( + __string( name, mod->name ) + ), + + TP_fast_assign( + __assign_str(name, mod->name); + ), + + TP_printk("%s", __get_str(name)) +); + +TRACE_EVENT(module_get, + + TP_PROTO(struct module *mod, unsigned long ip, int refcnt), + + TP_ARGS(mod, ip, refcnt), + + TP_STRUCT__entry( + __field( unsigned long, ip ) + __field( int, refcnt ) + __string( name, mod->name ) + ), + + TP_fast_assign( + __entry->ip = ip; + __entry->refcnt = refcnt; + __assign_str(name, mod->name); + ), + + TP_printk("%s call_site=%pf refcnt=%d", + __get_str(name), (void *)__entry->ip, __entry->refcnt) +); + +TRACE_EVENT(module_put, + + TP_PROTO(struct module *mod, unsigned long ip, int refcnt), + + TP_ARGS(mod, ip, refcnt), + + TP_STRUCT__entry( + __field( unsigned long, ip ) + __field( int, refcnt ) + __string( name, mod->name ) + ), + + TP_fast_assign( + __entry->ip = ip; + __entry->refcnt = refcnt; + __assign_str(name, mod->name); + ), + + TP_printk("%s call_site=%pf refcnt=%d", + __get_str(name), (void *)__entry->ip, __entry->refcnt) +); + +TRACE_EVENT(module_request, + + TP_PROTO(char *name, bool wait, unsigned long ip), + + TP_ARGS(name, wait, ip), + + TP_STRUCT__entry( + __field( bool, wait ) + __field( unsigned long, ip ) + __string( name, name ) + ), + + TP_fast_assign( + __entry->wait = wait; + __entry->ip = ip; + __assign_str(name, name); + ), + + TP_printk("%s wait=%d call_site=%pf", + __get_str(name), (int)__entry->wait, (void *)__entry->ip) +); + +#endif /* CONFIG_MODULES */ + +#endif /* _TRACE_MODULE_H */ + +/* This part must be outside protection */ +#include + diff --git a/kernel/kmod.c b/kernel/kmod.c index 385c31a1bdbf..a92280870e30 100644 --- a/kernel/kmod.c +++ b/kernel/kmod.c @@ -37,6 +37,8 @@ #include #include +#include + extern int max_threads; static struct workqueue_struct *khelper_wq; @@ -108,6 +110,8 @@ int __request_module(bool wait, const char *fmt, ...) return -ENOMEM; } + trace_module_request(module_name, wait, _RET_IP_); + ret = call_usermodehelper(modprobe_path, argv, envp, wait ? UMH_WAIT_PROC : UMH_WAIT_EXEC); atomic_dec(&kmod_concurrent); diff --git a/kernel/module.c b/kernel/module.c index fd1411403558..b1821438694e 100644 --- a/kernel/module.c +++ b/kernel/module.c @@ -55,6 +55,11 @@ #include #include +#define CREATE_TRACE_POINTS +#include + +EXPORT_TRACEPOINT_SYMBOL(module_get); + #if 0 #define DEBUGP printk #else @@ -940,6 +945,8 @@ void module_put(struct module *module) if (module) { unsigned int cpu = get_cpu(); local_dec(__module_ref_addr(module, cpu)); + trace_module_put(module, _RET_IP_, + local_read(__module_ref_addr(module, cpu))); /* Maybe they're waiting for us to drop reference? */ if (unlikely(!module_is_live(module))) wake_up_process(module->waiter); @@ -1491,6 +1498,8 @@ static int __unlink_module(void *_mod) /* Free a module, remove from lists, etc (must hold module_mutex). */ static void free_module(struct module *mod) { + trace_module_free(mod); + /* Delete from various lists */ stop_machine(__unlink_module, mod, NULL); remove_notes_attrs(mod); @@ -2358,6 +2367,8 @@ static noinline struct module *load_module(void __user *umod, /* Get rid of temporary copy */ vfree(hdr); + trace_module_load(mod); + /* Done! */ return mod; -- cgit v1.2.3 From 0ccff1a49def92d6b838a6da166c89004b3a4d0c Mon Sep 17 00:00:00 2001 From: H Hartley Sweeten Date: Mon, 17 Aug 2009 22:38:04 -0400 Subject: jbd2: bitfields should be unsigned This fixes sparse noise: error: dubious one-bit signed bitfield Signed-off-by: H Hartley Sweeten Signed-off-by: Andrew Morton Signed-off-by: "Theodore Ts'o" Cc: Jan Kara --- include/linux/jbd2.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/jbd2.h b/include/linux/jbd2.h index d97eb652d6ca..52695d3dfd0b 100644 --- a/include/linux/jbd2.h +++ b/include/linux/jbd2.h @@ -652,7 +652,7 @@ struct transaction_s * This transaction is being forced and some process is * waiting for it to finish. */ - int t_synchronous_commit:1; + unsigned int t_synchronous_commit:1; /* * For use by the filesystem to store fs-specific data -- cgit v1.2.3 From c1a8f1f1c8e01eab5862c8db39b49ace814e6c66 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Sun, 16 Aug 2009 09:36:49 +0000 Subject: net: restore gnet_stats_basic to previous definition In 5e140dfc1fe87eae27846f193086724806b33c7d "net: reorder struct Qdisc for better SMP performance" the definition of struct gnet_stats_basic changed incompatibly, as copies of this struct are shipped to userland via netlink. Restoring old behavior is not welcome, for performance reason. Fix is to use a private structure for kernel, and teach gnet_stats_copy_basic() to convert from kernel to user land, using legacy structure (struct gnet_stats_basic) Based on a report and initial patch from Michael Spang. Reported-by: Michael Spang Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- include/linux/gen_stats.h | 5 +++++ include/net/act_api.h | 2 +- include/net/gen_stats.h | 10 +++++----- include/net/netfilter/xt_rateest.h | 2 +- include/net/sch_generic.h | 2 +- net/core/gen_estimator.c | 12 ++++++------ net/core/gen_stats.c | 11 ++++++++--- net/netfilter/xt_RATEEST.c | 2 +- net/sched/sch_atm.c | 2 +- net/sched/sch_cbq.c | 2 +- net/sched/sch_drr.c | 2 +- net/sched/sch_hfsc.c | 2 +- net/sched/sch_htb.c | 2 +- 13 files changed, 33 insertions(+), 23 deletions(-) (limited to 'include') diff --git a/include/linux/gen_stats.h b/include/linux/gen_stats.h index 0ffa41df0ee8..710e901085d0 100644 --- a/include/linux/gen_stats.h +++ b/include/linux/gen_stats.h @@ -19,6 +19,11 @@ enum { * @packets: number of seen packets */ struct gnet_stats_basic +{ + __u64 bytes; + __u32 packets; +}; +struct gnet_stats_basic_packed { __u64 bytes; __u32 packets; diff --git a/include/net/act_api.h b/include/net/act_api.h index 565eed8fe496..c05fd717c588 100644 --- a/include/net/act_api.h +++ b/include/net/act_api.h @@ -16,7 +16,7 @@ struct tcf_common { u32 tcfc_capab; int tcfc_action; struct tcf_t tcfc_tm; - struct gnet_stats_basic tcfc_bstats; + struct gnet_stats_basic_packed tcfc_bstats; struct gnet_stats_queue tcfc_qstats; struct gnet_stats_rate_est tcfc_rate_est; spinlock_t tcfc_lock; diff --git a/include/net/gen_stats.h b/include/net/gen_stats.h index d136b5240ef2..c1488553e349 100644 --- a/include/net/gen_stats.h +++ b/include/net/gen_stats.h @@ -28,7 +28,7 @@ extern int gnet_stats_start_copy_compat(struct sk_buff *skb, int type, spinlock_t *lock, struct gnet_dump *d); extern int gnet_stats_copy_basic(struct gnet_dump *d, - struct gnet_stats_basic *b); + struct gnet_stats_basic_packed *b); extern int gnet_stats_copy_rate_est(struct gnet_dump *d, struct gnet_stats_rate_est *r); extern int gnet_stats_copy_queue(struct gnet_dump *d, @@ -37,14 +37,14 @@ extern int gnet_stats_copy_app(struct gnet_dump *d, void *st, int len); extern int gnet_stats_finish_copy(struct gnet_dump *d); -extern int gen_new_estimator(struct gnet_stats_basic *bstats, +extern int gen_new_estimator(struct gnet_stats_basic_packed *bstats, struct gnet_stats_rate_est *rate_est, spinlock_t *stats_lock, struct nlattr *opt); -extern void gen_kill_estimator(struct gnet_stats_basic *bstats, +extern void gen_kill_estimator(struct gnet_stats_basic_packed *bstats, struct gnet_stats_rate_est *rate_est); -extern int gen_replace_estimator(struct gnet_stats_basic *bstats, +extern int gen_replace_estimator(struct gnet_stats_basic_packed *bstats, struct gnet_stats_rate_est *rate_est, spinlock_t *stats_lock, struct nlattr *opt); -extern bool gen_estimator_active(const struct gnet_stats_basic *bstats, +extern bool gen_estimator_active(const struct gnet_stats_basic_packed *bstats, const struct gnet_stats_rate_est *rate_est); #endif diff --git a/include/net/netfilter/xt_rateest.h b/include/net/netfilter/xt_rateest.h index 65d594dffbff..ddbf37e19616 100644 --- a/include/net/netfilter/xt_rateest.h +++ b/include/net/netfilter/xt_rateest.h @@ -8,7 +8,7 @@ struct xt_rateest { spinlock_t lock; struct gnet_estimator params; struct gnet_stats_rate_est rstats; - struct gnet_stats_basic bstats; + struct gnet_stats_basic_packed bstats; }; extern struct xt_rateest *xt_rateest_lookup(const char *name); diff --git a/include/net/sch_generic.h b/include/net/sch_generic.h index 964ffa0d8815..5482e9582f55 100644 --- a/include/net/sch_generic.h +++ b/include/net/sch_generic.h @@ -72,7 +72,7 @@ struct Qdisc */ unsigned long state; struct sk_buff_head q; - struct gnet_stats_basic bstats; + struct gnet_stats_basic_packed bstats; struct gnet_stats_queue qstats; }; diff --git a/net/core/gen_estimator.c b/net/core/gen_estimator.c index 78e5bfc454ae..493775f4f2f1 100644 --- a/net/core/gen_estimator.c +++ b/net/core/gen_estimator.c @@ -81,7 +81,7 @@ struct gen_estimator { struct list_head list; - struct gnet_stats_basic *bstats; + struct gnet_stats_basic_packed *bstats; struct gnet_stats_rate_est *rate_est; spinlock_t *stats_lock; int ewma_log; @@ -165,7 +165,7 @@ static void gen_add_node(struct gen_estimator *est) } static -struct gen_estimator *gen_find_node(const struct gnet_stats_basic *bstats, +struct gen_estimator *gen_find_node(const struct gnet_stats_basic_packed *bstats, const struct gnet_stats_rate_est *rate_est) { struct rb_node *p = est_root.rb_node; @@ -202,7 +202,7 @@ struct gen_estimator *gen_find_node(const struct gnet_stats_basic *bstats, * * NOTE: Called under rtnl_mutex */ -int gen_new_estimator(struct gnet_stats_basic *bstats, +int gen_new_estimator(struct gnet_stats_basic_packed *bstats, struct gnet_stats_rate_est *rate_est, spinlock_t *stats_lock, struct nlattr *opt) @@ -262,7 +262,7 @@ static void __gen_kill_estimator(struct rcu_head *head) * * NOTE: Called under rtnl_mutex */ -void gen_kill_estimator(struct gnet_stats_basic *bstats, +void gen_kill_estimator(struct gnet_stats_basic_packed *bstats, struct gnet_stats_rate_est *rate_est) { struct gen_estimator *e; @@ -292,7 +292,7 @@ EXPORT_SYMBOL(gen_kill_estimator); * * Returns 0 on success or a negative error code. */ -int gen_replace_estimator(struct gnet_stats_basic *bstats, +int gen_replace_estimator(struct gnet_stats_basic_packed *bstats, struct gnet_stats_rate_est *rate_est, spinlock_t *stats_lock, struct nlattr *opt) { @@ -308,7 +308,7 @@ EXPORT_SYMBOL(gen_replace_estimator); * * Returns true if estimator is active, and false if not. */ -bool gen_estimator_active(const struct gnet_stats_basic *bstats, +bool gen_estimator_active(const struct gnet_stats_basic_packed *bstats, const struct gnet_stats_rate_est *rate_est) { ASSERT_RTNL(); diff --git a/net/core/gen_stats.c b/net/core/gen_stats.c index c3d0ffeac243..8569310268ab 100644 --- a/net/core/gen_stats.c +++ b/net/core/gen_stats.c @@ -106,16 +106,21 @@ gnet_stats_start_copy(struct sk_buff *skb, int type, spinlock_t *lock, * if the room in the socket buffer was not sufficient. */ int -gnet_stats_copy_basic(struct gnet_dump *d, struct gnet_stats_basic *b) +gnet_stats_copy_basic(struct gnet_dump *d, struct gnet_stats_basic_packed *b) { if (d->compat_tc_stats) { d->tc_stats.bytes = b->bytes; d->tc_stats.packets = b->packets; } - if (d->tail) - return gnet_stats_copy(d, TCA_STATS_BASIC, b, sizeof(*b)); + if (d->tail) { + struct gnet_stats_basic sb; + memset(&sb, 0, sizeof(sb)); + sb.bytes = b->bytes; + sb.packets = b->packets; + return gnet_stats_copy(d, TCA_STATS_BASIC, &sb, sizeof(sb)); + } return 0; } diff --git a/net/netfilter/xt_RATEEST.c b/net/netfilter/xt_RATEEST.c index 43f5676b1af4..d80b8192e0d4 100644 --- a/net/netfilter/xt_RATEEST.c +++ b/net/netfilter/xt_RATEEST.c @@ -74,7 +74,7 @@ static unsigned int xt_rateest_tg(struct sk_buff *skb, const struct xt_target_param *par) { const struct xt_rateest_target_info *info = par->targinfo; - struct gnet_stats_basic *stats = &info->est->bstats; + struct gnet_stats_basic_packed *stats = &info->est->bstats; spin_lock_bh(&info->est->lock); stats->bytes += skb->len; diff --git a/net/sched/sch_atm.c b/net/sched/sch_atm.c index 2a8b83af7c47..ab82f145f689 100644 --- a/net/sched/sch_atm.c +++ b/net/sched/sch_atm.c @@ -49,7 +49,7 @@ struct atm_flow_data { struct socket *sock; /* for closing */ u32 classid; /* x:y type ID */ int ref; /* reference count */ - struct gnet_stats_basic bstats; + struct gnet_stats_basic_packed bstats; struct gnet_stats_queue qstats; struct atm_flow_data *next; struct atm_flow_data *excess; /* flow for excess traffic; diff --git a/net/sched/sch_cbq.c b/net/sched/sch_cbq.c index 23a167670fd5..d5798e17a832 100644 --- a/net/sched/sch_cbq.c +++ b/net/sched/sch_cbq.c @@ -128,7 +128,7 @@ struct cbq_class long avgidle; long deficit; /* Saved deficit for WRR */ psched_time_t penalized; - struct gnet_stats_basic bstats; + struct gnet_stats_basic_packed bstats; struct gnet_stats_queue qstats; struct gnet_stats_rate_est rate_est; struct tc_cbq_xstats xstats; diff --git a/net/sched/sch_drr.c b/net/sched/sch_drr.c index 7597fe146866..12b2fb04b29b 100644 --- a/net/sched/sch_drr.c +++ b/net/sched/sch_drr.c @@ -22,7 +22,7 @@ struct drr_class { unsigned int refcnt; unsigned int filter_cnt; - struct gnet_stats_basic bstats; + struct gnet_stats_basic_packed bstats; struct gnet_stats_queue qstats; struct gnet_stats_rate_est rate_est; struct list_head alist; diff --git a/net/sched/sch_hfsc.c b/net/sched/sch_hfsc.c index 362c2811b2df..dad0144423da 100644 --- a/net/sched/sch_hfsc.c +++ b/net/sched/sch_hfsc.c @@ -116,7 +116,7 @@ struct hfsc_class struct Qdisc_class_common cl_common; unsigned int refcnt; /* usage count */ - struct gnet_stats_basic bstats; + struct gnet_stats_basic_packed bstats; struct gnet_stats_queue qstats; struct gnet_stats_rate_est rate_est; unsigned int level; /* class level in hierarchy */ diff --git a/net/sched/sch_htb.c b/net/sched/sch_htb.c index 88cd02626621..ec4d46399d59 100644 --- a/net/sched/sch_htb.c +++ b/net/sched/sch_htb.c @@ -74,7 +74,7 @@ enum htb_cmode { struct htb_class { struct Qdisc_class_common common; /* general class parameters */ - struct gnet_stats_basic bstats; + struct gnet_stats_basic_packed bstats; struct gnet_stats_queue qstats; struct gnet_stats_rate_est rate_est; struct tc_htb_xstats xstats; /* our special stats */ -- cgit v1.2.3 From 1314562a9ae5f39f6f595656023c1baf970831ef Mon Sep 17 00:00:00 2001 From: Hiroshi Shimamoto Date: Tue, 18 Aug 2009 15:06:02 +0900 Subject: sched, task_struct: stack_canary is not needed without CC_STACKPROTECTOR The field stack_canary is only used with CC_STACKPROTECTOR. This patch reduces task_struct size without CC_STACKPROTECTOR. Signed-off-by: Hiroshi Shimamoto LKML-Reference: <4A8A44CA.2020701@ct.jp.nec.com> Signed-off-by: Ingo Molnar --- include/linux/sched.h | 2 ++ 1 file changed, 2 insertions(+) (limited to 'include') diff --git a/include/linux/sched.h b/include/linux/sched.h index 195d72d5c102..7bc2d9290837 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -1237,8 +1237,10 @@ struct task_struct { pid_t pid; pid_t tgid; +#ifdef CONFIG_CC_STACKPROTECTOR /* Canary value for the -fstack-protector gcc feature */ unsigned long stack_canary; +#endif /* * pointers to (original) parent process, youngest child, younger sibling, -- cgit v1.2.3 From 0753ba01e126020bf0f8150934903b48935b697d Mon Sep 17 00:00:00 2001 From: KOSAKI Motohiro Date: Tue, 18 Aug 2009 14:11:10 -0700 Subject: mm: revert "oom: move oom_adj value" The commit 2ff05b2b (oom: move oom_adj value) moveed the oom_adj value to the mm_struct. It was a very good first step for sanitize OOM. However Paul Menage reported the commit makes regression to his job scheduler. Current OOM logic can kill OOM_DISABLED process. Why? His program has the code of similar to the following. ... set_oom_adj(OOM_DISABLE); /* The job scheduler never killed by oom */ ... if (vfork() == 0) { set_oom_adj(0); /* Invoked child can be killed */ execve("foo-bar-cmd"); } .... vfork() parent and child are shared the same mm_struct. then above set_oom_adj(0) doesn't only change oom_adj for vfork() child, it's also change oom_adj for vfork() parent. Then, vfork() parent (job scheduler) lost OOM immune and it was killed. Actually, fork-setting-exec idiom is very frequently used in userland program. We must not break this assumption. Then, this patch revert commit 2ff05b2b and related commit. Reverted commit list --------------------- - commit 2ff05b2b4e (oom: move oom_adj value from task_struct to mm_struct) - commit 4d8b9135c3 (oom: avoid unnecessary mm locking and scanning for OOM_DISABLE) - commit 8123681022 (oom: only oom kill exiting tasks with attached memory) - commit 933b787b57 (mm: copy over oom_adj value at fork time) Signed-off-by: KOSAKI Motohiro Cc: Paul Menage Cc: David Rientjes Cc: KAMEZAWA Hiroyuki Cc: Rik van Riel Cc: Linus Torvalds Cc: Oleg Nesterov Cc: Nick Piggin Cc: Mel Gorman Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- Documentation/filesystems/proc.txt | 15 +++------ fs/proc/base.c | 19 ++--------- include/linux/mm_types.h | 2 -- include/linux/sched.h | 1 + kernel/fork.c | 1 - mm/oom_kill.c | 64 +++++++++++++++++++++++--------------- 6 files changed, 48 insertions(+), 54 deletions(-) (limited to 'include') diff --git a/Documentation/filesystems/proc.txt b/Documentation/filesystems/proc.txt index fad18f9456e4..ffead13f9443 100644 --- a/Documentation/filesystems/proc.txt +++ b/Documentation/filesystems/proc.txt @@ -1167,13 +1167,11 @@ CHAPTER 3: PER-PROCESS PARAMETERS 3.1 /proc//oom_adj - Adjust the oom-killer score ------------------------------------------------------ -This file can be used to adjust the score used to select which processes should -be killed in an out-of-memory situation. The oom_adj value is a characteristic -of the task's mm, so all threads that share an mm with pid will have the same -oom_adj value. A high value will increase the likelihood of this process being -killed by the oom-killer. Valid values are in the range -16 to +15 as -explained below and a special value of -17, which disables oom-killing -altogether for threads sharing pid's mm. +This file can be used to adjust the score used to select which processes +should be killed in an out-of-memory situation. Giving it a high score will +increase the likelihood of this process being killed by the oom-killer. Valid +values are in the range -16 to +15, plus the special value -17, which disables +oom-killing altogether for this process. The process to be killed in an out-of-memory situation is selected among all others based on its badness score. This value equals the original memory size of the process @@ -1187,9 +1185,6 @@ the parent's score if they do not share the same memory. Thus forking servers are the prime candidates to be killed. Having only one 'hungry' child will make parent less preferable than the child. -/proc//oom_adj cannot be changed for kthreads since they are immune from -oom-killing already. - /proc//oom_score shows process' current badness score. The following heuristics are then applied: diff --git a/fs/proc/base.c b/fs/proc/base.c index 175db258942f..6f742f6658a9 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -1003,12 +1003,7 @@ static ssize_t oom_adjust_read(struct file *file, char __user *buf, if (!task) return -ESRCH; - task_lock(task); - if (task->mm) - oom_adjust = task->mm->oom_adj; - else - oom_adjust = OOM_DISABLE; - task_unlock(task); + oom_adjust = task->oomkilladj; put_task_struct(task); len = snprintf(buffer, sizeof(buffer), "%i\n", oom_adjust); @@ -1037,19 +1032,11 @@ static ssize_t oom_adjust_write(struct file *file, const char __user *buf, task = get_proc_task(file->f_path.dentry->d_inode); if (!task) return -ESRCH; - task_lock(task); - if (!task->mm) { - task_unlock(task); - put_task_struct(task); - return -EINVAL; - } - if (oom_adjust < task->mm->oom_adj && !capable(CAP_SYS_RESOURCE)) { - task_unlock(task); + if (oom_adjust < task->oomkilladj && !capable(CAP_SYS_RESOURCE)) { put_task_struct(task); return -EACCES; } - task->mm->oom_adj = oom_adjust; - task_unlock(task); + task->oomkilladj = oom_adjust; put_task_struct(task); if (end - buffer == 0) return -EIO; diff --git a/include/linux/mm_types.h b/include/linux/mm_types.h index 7acc8439d9b3..0042090a4d70 100644 --- a/include/linux/mm_types.h +++ b/include/linux/mm_types.h @@ -240,8 +240,6 @@ struct mm_struct { unsigned long saved_auxv[AT_VECTOR_SIZE]; /* for /proc/PID/auxv */ - s8 oom_adj; /* OOM kill score adjustment (bit shift) */ - cpumask_t cpu_vm_mask; /* Architecture-specific MM context */ diff --git a/include/linux/sched.h b/include/linux/sched.h index 3ab08e4bb6b8..0f1ea4a66957 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -1198,6 +1198,7 @@ struct task_struct { * a short time */ unsigned char fpu_counter; + s8 oomkilladj; /* OOM kill score adjustment (bit shift). */ #ifdef CONFIG_BLK_DEV_IO_TRACE unsigned int btrace_seq; #endif diff --git a/kernel/fork.c b/kernel/fork.c index 021e1138556e..144326b7af50 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -426,7 +426,6 @@ static struct mm_struct * mm_init(struct mm_struct * mm, struct task_struct *p) init_rwsem(&mm->mmap_sem); INIT_LIST_HEAD(&mm->mmlist); mm->flags = (current->mm) ? current->mm->flags : default_dump_filter; - mm->oom_adj = (current->mm) ? current->mm->oom_adj : 0; mm->core_state = NULL; mm->nr_ptes = 0; set_mm_counter(mm, file_rss, 0); diff --git a/mm/oom_kill.c b/mm/oom_kill.c index 175a67a78a99..a7b2460e922b 100644 --- a/mm/oom_kill.c +++ b/mm/oom_kill.c @@ -58,7 +58,6 @@ unsigned long badness(struct task_struct *p, unsigned long uptime) unsigned long points, cpu_time, run_time; struct mm_struct *mm; struct task_struct *child; - int oom_adj; task_lock(p); mm = p->mm; @@ -66,11 +65,6 @@ unsigned long badness(struct task_struct *p, unsigned long uptime) task_unlock(p); return 0; } - oom_adj = mm->oom_adj; - if (oom_adj == OOM_DISABLE) { - task_unlock(p); - return 0; - } /* * The memory size of the process is the basis for the badness. @@ -154,15 +148,15 @@ unsigned long badness(struct task_struct *p, unsigned long uptime) points /= 8; /* - * Adjust the score by oom_adj. + * Adjust the score by oomkilladj. */ - if (oom_adj) { - if (oom_adj > 0) { + if (p->oomkilladj) { + if (p->oomkilladj > 0) { if (!points) points = 1; - points <<= oom_adj; + points <<= p->oomkilladj; } else - points >>= -(oom_adj); + points >>= -(p->oomkilladj); } #ifdef DEBUG @@ -257,8 +251,11 @@ static struct task_struct *select_bad_process(unsigned long *ppoints, *ppoints = ULONG_MAX; } + if (p->oomkilladj == OOM_DISABLE) + continue; + points = badness(p, uptime.tv_sec); - if (points > *ppoints) { + if (points > *ppoints || !chosen) { chosen = p; *ppoints = points; } @@ -307,7 +304,8 @@ static void dump_tasks(const struct mem_cgroup *mem) } printk(KERN_INFO "[%5d] %5d %5d %8lu %8lu %3d %3d %s\n", p->pid, __task_cred(p)->uid, p->tgid, mm->total_vm, - get_mm_rss(mm), (int)task_cpu(p), mm->oom_adj, p->comm); + get_mm_rss(mm), (int)task_cpu(p), p->oomkilladj, + p->comm); task_unlock(p); } while_each_thread(g, p); } @@ -325,8 +323,11 @@ static void __oom_kill_task(struct task_struct *p, int verbose) return; } - if (!p->mm) + if (!p->mm) { + WARN_ON(1); + printk(KERN_WARNING "tried to kill an mm-less task!\n"); return; + } if (verbose) printk(KERN_ERR "Killed process %d (%s)\n", @@ -348,13 +349,28 @@ static int oom_kill_task(struct task_struct *p) struct mm_struct *mm; struct task_struct *g, *q; - task_lock(p); mm = p->mm; - if (!mm || mm->oom_adj == OOM_DISABLE) { - task_unlock(p); + + /* WARNING: mm may not be dereferenced since we did not obtain its + * value from get_task_mm(p). This is OK since all we need to do is + * compare mm to q->mm below. + * + * Furthermore, even if mm contains a non-NULL value, p->mm may + * change to NULL at any time since we do not hold task_lock(p). + * However, this is of no concern to us. + */ + + if (mm == NULL) return 1; - } - task_unlock(p); + + /* + * Don't kill the process if any threads are set to OOM_DISABLE + */ + do_each_thread(g, q) { + if (q->mm == mm && q->oomkilladj == OOM_DISABLE) + return 1; + } while_each_thread(g, q); + __oom_kill_task(p, 1); /* @@ -377,11 +393,10 @@ static int oom_kill_process(struct task_struct *p, gfp_t gfp_mask, int order, struct task_struct *c; if (printk_ratelimit()) { - task_lock(current); printk(KERN_WARNING "%s invoked oom-killer: " - "gfp_mask=0x%x, order=%d, oom_adj=%d\n", - current->comm, gfp_mask, order, - current->mm ? current->mm->oom_adj : OOM_DISABLE); + "gfp_mask=0x%x, order=%d, oomkilladj=%d\n", + current->comm, gfp_mask, order, current->oomkilladj); + task_lock(current); cpuset_print_task_mems_allowed(current); task_unlock(current); dump_stack(); @@ -394,9 +409,8 @@ static int oom_kill_process(struct task_struct *p, gfp_t gfp_mask, int order, /* * If the task is already exiting, don't alarm the sysadmin or kill * its children or threads, just set TIF_MEMDIE so it can die quickly - * if its mm is still attached. */ - if (p->mm && (p->flags & PF_EXITING)) { + if (p->flags & PF_EXITING) { __oom_kill_task(p, 0); return 0; } -- cgit v1.2.3 From 23428e6b4649adfbdaa6a0c93fc6d652bf5f9d44 Mon Sep 17 00:00:00 2001 From: Ben Hutchings Date: Tue, 18 Aug 2009 20:13:03 -0700 Subject: mdio: mdio_if_info::mmds should not be __bitwise I misunderstood the meaning of __bitwise. In practice it makes sparse warn about every use of mmds which is certainly not what we want. Signed-off-by: Ben Hutchings Signed-off-by: David S. Miller --- include/linux/mdio.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/mdio.h b/include/linux/mdio.h index cfdf1df2875e..c779b49a1fda 100644 --- a/include/linux/mdio.h +++ b/include/linux/mdio.h @@ -304,7 +304,7 @@ static inline __u16 mdio_phy_id_devad(int phy_id) */ struct mdio_if_info { int prtad; - u32 __bitwise mmds; + u32 mmds; unsigned mode_support; struct net_device *dev; -- cgit v1.2.3 From 776f3360de6ed246e973577828f725681120fd7a Mon Sep 17 00:00:00 2001 From: Dave Airlie Date: Wed, 19 Aug 2009 15:56:37 +1000 Subject: drm: fixup includes in encoder slave header files. Signed-off-by: Dave Airlie --- include/drm/drm_encoder_slave.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/include/drm/drm_encoder_slave.h b/include/drm/drm_encoder_slave.h index 821ec40c17d8..e5e5c94ca92c 100644 --- a/include/drm/drm_encoder_slave.h +++ b/include/drm/drm_encoder_slave.h @@ -27,8 +27,8 @@ #ifndef __DRM_ENCODER_SLAVE_H__ #define __DRM_ENCODER_SLAVE_H__ -#include -#include +#include "drmP.h" +#include "drm_crtc.h" /** * struct drm_encoder_slave_funcs - Entry points exposed by a slave encoder driver -- cgit v1.2.3 From 53bd83899f5ba6b0da8f5ef976129273854a72d4 Mon Sep 17 00:00:00 2001 From: Jesse Barnes Date: Wed, 1 Jul 2009 10:04:40 -0700 Subject: drm: clarify scaling property names Now that we're using the scaling property in the Intel driver I noticed that the names were a bit confusing. I've corrected them according to our discussion on IRC and the mailing list, though I've left out potential new additions for a new scaling property with an integer (or two) for the scaling factor. None of the drivers implement that today, but if someone wants to do it, I think it could be done with the addition of a single new type and a new property to describe the scaling factor in the X and Y directions. Signed-off-by: Jesse Barnes Acked-by: Alex Deucher Signed-off-by: Dave Airlie --- drivers/gpu/drm/drm_crtc.c | 8 ++++---- drivers/gpu/drm/i915/intel_lvds.c | 14 +++----------- include/drm/drm_mode.h | 9 +++++---- 3 files changed, 12 insertions(+), 19 deletions(-) (limited to 'include') diff --git a/drivers/gpu/drm/drm_crtc.c b/drivers/gpu/drm/drm_crtc.c index 362a538cdedc..39a6bc69d223 100644 --- a/drivers/gpu/drm/drm_crtc.c +++ b/drivers/gpu/drm/drm_crtc.c @@ -68,10 +68,10 @@ DRM_ENUM_NAME_FN(drm_get_dpms_name, drm_dpms_enum_list) */ static struct drm_prop_enum_list drm_scaling_mode_enum_list[] = { - { DRM_MODE_SCALE_NON_GPU, "Non-GPU" }, - { DRM_MODE_SCALE_FULLSCREEN, "Fullscreen" }, - { DRM_MODE_SCALE_NO_SCALE, "No scale" }, - { DRM_MODE_SCALE_ASPECT, "Aspect" }, + { DRM_MODE_SCALE_NONE, "None" }, + { DRM_MODE_SCALE_FULLSCREEN, "Full" }, + { DRM_MODE_SCALE_CENTER, "Center" }, + { DRM_MODE_SCALE_ASPECT, "Full aspect" }, }; static struct drm_prop_enum_list drm_dithering_mode_enum_list[] = diff --git a/drivers/gpu/drm/i915/intel_lvds.c b/drivers/gpu/drm/i915/intel_lvds.c index b59c65d19d81..5df486fbe056 100644 --- a/drivers/gpu/drm/i915/intel_lvds.c +++ b/drivers/gpu/drm/i915/intel_lvds.c @@ -38,14 +38,6 @@ #include "i915_drv.h" #include -/* - * the following four scaling options are defined. - * #define DRM_MODE_SCALE_NON_GPU 0 - * #define DRM_MODE_SCALE_FULLSCREEN 1 - * #define DRM_MODE_SCALE_NO_SCALE 2 - * #define DRM_MODE_SCALE_ASPECT 3 - */ - /* Private structure for the integrated LVDS support */ struct intel_lvds_priv { int fitting_mode; @@ -334,7 +326,7 @@ static bool intel_lvds_mode_fixup(struct drm_encoder *encoder, I915_WRITE(BCLRPAT_B, 0); switch (lvds_priv->fitting_mode) { - case DRM_MODE_SCALE_NO_SCALE: + case DRM_MODE_SCALE_CENTER: /* * For centered modes, we have to calculate border widths & * heights and modify the values programmed into the CRTC. @@ -670,8 +662,8 @@ static int intel_lvds_set_property(struct drm_connector *connector, connector->encoder) { struct drm_crtc *crtc = connector->encoder->crtc; struct intel_lvds_priv *lvds_priv = intel_output->dev_priv; - if (value == DRM_MODE_SCALE_NON_GPU) { - DRM_DEBUG_KMS("non_GPU property is unsupported\n"); + if (value == DRM_MODE_SCALE_NONE) { + DRM_DEBUG_KMS("no scaling not supported\n"); return 0; } if (lvds_priv->fitting_mode == value) { diff --git a/include/drm/drm_mode.h b/include/drm/drm_mode.h index 616aeb42b773..1f908416aedb 100644 --- a/include/drm/drm_mode.h +++ b/include/drm/drm_mode.h @@ -68,10 +68,11 @@ #define DRM_MODE_DPMS_OFF 3 /* Scaling mode options */ -#define DRM_MODE_SCALE_NON_GPU 0 -#define DRM_MODE_SCALE_FULLSCREEN 1 -#define DRM_MODE_SCALE_NO_SCALE 2 -#define DRM_MODE_SCALE_ASPECT 3 +#define DRM_MODE_SCALE_NONE 0 /* Unmodified timing (display or + software can still scale) */ +#define DRM_MODE_SCALE_FULLSCREEN 1 /* Full screen, ignore aspect */ +#define DRM_MODE_SCALE_CENTER 2 /* Centered, no scaling */ +#define DRM_MODE_SCALE_ASPECT 3 /* Full screen, preserve aspect */ /* Dithering mode options */ #define DRM_MODE_DITHERING_OFF 0 -- cgit v1.2.3 From 949ef70e2d1a5c12178875f513df34fc85d91a38 Mon Sep 17 00:00:00 2001 From: Pekka Paalanen Date: Mon, 17 Aug 2009 19:49:19 +0300 Subject: drm/kms: no need to return void value (encoder) Cc: Francisco Jerez Signed-off-by: Pekka Paalanen Signed-off-by: Dave Airlie --- include/drm/drm_encoder_slave.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/drm/drm_encoder_slave.h b/include/drm/drm_encoder_slave.h index e5e5c94ca92c..2f65633d28a7 100644 --- a/include/drm/drm_encoder_slave.h +++ b/include/drm/drm_encoder_slave.h @@ -154,7 +154,7 @@ static inline int drm_i2c_encoder_register(struct module *owner, */ static inline void drm_i2c_encoder_unregister(struct drm_i2c_encoder_driver *driver) { - return i2c_del_driver(&driver->i2c_driver); + i2c_del_driver(&driver->i2c_driver); } void drm_i2c_encoder_destroy(struct drm_encoder *encoder); -- cgit v1.2.3 From a0724fcf829e5afb66159ef68cb16a805ea11b42 Mon Sep 17 00:00:00 2001 From: Pekka Paalanen Date: Mon, 17 Aug 2009 01:18:38 +0300 Subject: drm/ttm: optimize bo_kmap_type values A micro-optimization on the function ttm_kmap_obj_virtual(). By defining the values of enum ttm_bo_kmap_obj::bo_kmap_type to have a bit indicating iomem, size of the function ttm_kmap_obj_virtual() will be reduced by 16 bytes on x86_64 (gcc 4.1.2). ttm_kmap_obj_virtual() may be heavily used, when buffer objects are accessed via wrappers, that work for both kinds of memory addresses: iomem cookies and kernel virtual. Signed-off-by: Pekka Paalanen Signed-off-by: Dave Airlie --- include/drm/ttm/ttm_bo_api.h | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) (limited to 'include') diff --git a/include/drm/ttm/ttm_bo_api.h b/include/drm/ttm/ttm_bo_api.h index cd22ab4b495c..99dc521aa1a9 100644 --- a/include/drm/ttm/ttm_bo_api.h +++ b/include/drm/ttm/ttm_bo_api.h @@ -245,14 +245,15 @@ struct ttm_buffer_object { * premapped region. */ +#define TTM_BO_MAP_IOMEM_MASK 0x80 struct ttm_bo_kmap_obj { void *virtual; struct page *page; enum { - ttm_bo_map_iomap, - ttm_bo_map_vmap, - ttm_bo_map_kmap, - ttm_bo_map_premapped, + ttm_bo_map_iomap = 1 | TTM_BO_MAP_IOMEM_MASK, + ttm_bo_map_vmap = 2, + ttm_bo_map_kmap = 3, + ttm_bo_map_premapped = 4 | TTM_BO_MAP_IOMEM_MASK, } bo_kmap_type; }; @@ -522,8 +523,7 @@ extern int ttm_bo_evict_mm(struct ttm_bo_device *bdev, unsigned mem_type); static inline void *ttm_kmap_obj_virtual(struct ttm_bo_kmap_obj *map, bool *is_iomem) { - *is_iomem = (map->bo_kmap_type == ttm_bo_map_iomap || - map->bo_kmap_type == ttm_bo_map_premapped); + *is_iomem = !!(map->bo_kmap_type & TTM_BO_MAP_IOMEM_MASK); return map->virtual; } -- cgit v1.2.3 From 327c225bd548bf7871f116a0baa5ebdac884e452 Mon Sep 17 00:00:00 2001 From: Thomas Hellstrom Date: Mon, 17 Aug 2009 16:28:37 +0200 Subject: drm: Enable drm drivers to add drm sysfs devices. Export utility functions for drivers to add specialized devices in the sysfs drm class subdirectory. Initially this will be needed form TTM to add a virtual device that handles power management. Signed-off-by: Thomas Hellstrom Signed-off-by: Dave Airlie --- drivers/gpu/drm/drm_sysfs.c | 25 +++++++++++++++++++++++++ include/drm/drm_sysfs.h | 12 ++++++++++++ 2 files changed, 37 insertions(+) create mode 100644 include/drm/drm_sysfs.h (limited to 'include') diff --git a/drivers/gpu/drm/drm_sysfs.c b/drivers/gpu/drm/drm_sysfs.c index adc179459c25..de154556c405 100644 --- a/drivers/gpu/drm/drm_sysfs.c +++ b/drivers/gpu/drm/drm_sysfs.c @@ -16,6 +16,7 @@ #include #include +#include "drm_sysfs.h" #include "drm_core.h" #include "drmP.h" @@ -515,3 +516,27 @@ void drm_sysfs_device_remove(struct drm_minor *minor) { device_unregister(&minor->kdev); } + + +/** + * drm_class_device_register - Register a struct device in the drm class. + * + * @dev: pointer to struct device to register. + * + * @dev should have all relevant members pre-filled with the exception + * of the class member. In particular, the device_type member must + * be set. + */ + +int drm_class_device_register(struct device *dev) +{ + dev->class = drm_class; + return device_register(dev); +} +EXPORT_SYMBOL_GPL(drm_class_device_register); + +void drm_class_device_unregister(struct device *dev) +{ + return device_unregister(dev); +} +EXPORT_SYMBOL_GPL(drm_class_device_unregister); diff --git a/include/drm/drm_sysfs.h b/include/drm/drm_sysfs.h new file mode 100644 index 000000000000..1d8e033fde67 --- /dev/null +++ b/include/drm/drm_sysfs.h @@ -0,0 +1,12 @@ +#ifndef _DRM_SYSFS_H_ +#define _DRM_SYSFS_H_ + +/** + * This minimalistic include file is intended for users (read TTM) that + * don't want to include the full drmP.h file. + */ + +extern int drm_class_device_register(struct device *dev); +extern void drm_class_device_unregister(struct device *dev); + +#endif -- cgit v1.2.3 From 5fd9cbad3a4ae82c83c55b9c621d156c326724ef Mon Sep 17 00:00:00 2001 From: Thomas Hellstrom Date: Mon, 17 Aug 2009 16:28:39 +0200 Subject: drm/ttm: Memory accounting rework. Use inclusive zones to simplify accounting and its sysfs representation. Use DMA32 accounting where applicable. Add a sysfs interface to make the heuristically determined limits readable and configurable. Signed-off-by: Thomas Hellstrom Signed-off-by: Dave Airlie --- drivers/gpu/drm/ttm/ttm_bo.c | 6 +- drivers/gpu/drm/ttm/ttm_global.c | 4 +- drivers/gpu/drm/ttm/ttm_memory.c | 488 +++++++++++++++++++++++++++++++++------ drivers/gpu/drm/ttm/ttm_tt.c | 29 +-- include/drm/ttm/ttm_memory.h | 43 ++-- include/drm/ttm/ttm_module.h | 2 + 6 files changed, 453 insertions(+), 119 deletions(-) (limited to 'include') diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c index c1c407f7cca3..f16909ceec93 100644 --- a/drivers/gpu/drm/ttm/ttm_bo.c +++ b/drivers/gpu/drm/ttm/ttm_bo.c @@ -70,7 +70,7 @@ static void ttm_bo_release_list(struct kref *list_kref) if (bo->destroy) bo->destroy(bo); else { - ttm_mem_global_free(bdev->mem_glob, bo->acc_size, false); + ttm_mem_global_free(bdev->mem_glob, bo->acc_size); kfree(bo); } } @@ -1065,14 +1065,14 @@ int ttm_buffer_object_create(struct ttm_bo_device *bdev, size_t acc_size = ttm_bo_size(bdev, (size + PAGE_SIZE - 1) >> PAGE_SHIFT); - ret = ttm_mem_global_alloc(mem_glob, acc_size, false, false, false); + ret = ttm_mem_global_alloc(mem_glob, acc_size, false, false); if (unlikely(ret != 0)) return ret; bo = kzalloc(sizeof(*bo), GFP_KERNEL); if (unlikely(bo == NULL)) { - ttm_mem_global_free(mem_glob, acc_size, false); + ttm_mem_global_free(mem_glob, acc_size); return -ENOMEM; } diff --git a/drivers/gpu/drm/ttm/ttm_global.c b/drivers/gpu/drm/ttm/ttm_global.c index 0b14eb1972b8..541744d00d3e 100644 --- a/drivers/gpu/drm/ttm/ttm_global.c +++ b/drivers/gpu/drm/ttm/ttm_global.c @@ -71,7 +71,7 @@ int ttm_global_item_ref(struct ttm_global_reference *ref) mutex_lock(&item->mutex); if (item->refcount == 0) { - item->object = kmalloc(ref->size, GFP_KERNEL); + item->object = kzalloc(ref->size, GFP_KERNEL); if (unlikely(item->object == NULL)) { ret = -ENOMEM; goto out_err; @@ -89,7 +89,6 @@ int ttm_global_item_ref(struct ttm_global_reference *ref) mutex_unlock(&item->mutex); return 0; out_err: - kfree(item->object); mutex_unlock(&item->mutex); item->object = NULL; return ret; @@ -105,7 +104,6 @@ void ttm_global_item_unref(struct ttm_global_reference *ref) BUG_ON(ref->object != item->object); if (--item->refcount == 0) { ref->release(ref); - kfree(item->object); item->object = NULL; } mutex_unlock(&item->mutex); diff --git a/drivers/gpu/drm/ttm/ttm_memory.c b/drivers/gpu/drm/ttm/ttm_memory.c index 87323d4ff68d..62fb5cf0899e 100644 --- a/drivers/gpu/drm/ttm/ttm_memory.c +++ b/drivers/gpu/drm/ttm/ttm_memory.c @@ -26,15 +26,180 @@ **************************************************************************/ #include "ttm/ttm_memory.h" +#include "ttm/ttm_module.h" #include #include #include #include #include -#define TTM_PFX "[TTM] " #define TTM_MEMORY_ALLOC_RETRIES 4 +struct ttm_mem_zone { + struct kobject kobj; + struct ttm_mem_global *glob; + const char *name; + uint64_t zone_mem; + uint64_t emer_mem; + uint64_t max_mem; + uint64_t swap_limit; + uint64_t used_mem; +}; + +static struct attribute ttm_mem_sys = { + .name = "zone_memory", + .mode = S_IRUGO +}; +static struct attribute ttm_mem_emer = { + .name = "emergency_memory", + .mode = S_IRUGO | S_IWUSR +}; +static struct attribute ttm_mem_max = { + .name = "available_memory", + .mode = S_IRUGO | S_IWUSR +}; +static struct attribute ttm_mem_swap = { + .name = "swap_limit", + .mode = S_IRUGO | S_IWUSR +}; +static struct attribute ttm_mem_used = { + .name = "used_memory", + .mode = S_IRUGO +}; + +static void ttm_mem_zone_kobj_release(struct kobject *kobj) +{ + struct ttm_mem_zone *zone = + container_of(kobj, struct ttm_mem_zone, kobj); + + printk(KERN_INFO TTM_PFX + "Zone %7s: Used memory at exit: %llu kiB.\n", + zone->name, (unsigned long long) zone->used_mem >> 10); + kfree(zone); +} + +static ssize_t ttm_mem_zone_show(struct kobject *kobj, + struct attribute *attr, + char *buffer) +{ + struct ttm_mem_zone *zone = + container_of(kobj, struct ttm_mem_zone, kobj); + uint64_t val = 0; + + spin_lock(&zone->glob->lock); + if (attr == &ttm_mem_sys) + val = zone->zone_mem; + else if (attr == &ttm_mem_emer) + val = zone->emer_mem; + else if (attr == &ttm_mem_max) + val = zone->max_mem; + else if (attr == &ttm_mem_swap) + val = zone->swap_limit; + else if (attr == &ttm_mem_used) + val = zone->used_mem; + spin_unlock(&zone->glob->lock); + + return snprintf(buffer, PAGE_SIZE, "%llu\n", + (unsigned long long) val >> 10); +} + +static void ttm_check_swapping(struct ttm_mem_global *glob); + +static ssize_t ttm_mem_zone_store(struct kobject *kobj, + struct attribute *attr, + const char *buffer, + size_t size) +{ + struct ttm_mem_zone *zone = + container_of(kobj, struct ttm_mem_zone, kobj); + int chars; + unsigned long val; + uint64_t val64; + + chars = sscanf(buffer, "%lu", &val); + if (chars == 0) + return size; + + val64 = val; + val64 <<= 10; + + spin_lock(&zone->glob->lock); + if (val64 > zone->zone_mem) + val64 = zone->zone_mem; + if (attr == &ttm_mem_emer) { + zone->emer_mem = val64; + if (zone->max_mem > val64) + zone->max_mem = val64; + } else if (attr == &ttm_mem_max) { + zone->max_mem = val64; + if (zone->emer_mem < val64) + zone->emer_mem = val64; + } else if (attr == &ttm_mem_swap) + zone->swap_limit = val64; + spin_unlock(&zone->glob->lock); + + ttm_check_swapping(zone->glob); + + return size; +} + +static struct attribute *ttm_mem_zone_attrs[] = { + &ttm_mem_sys, + &ttm_mem_emer, + &ttm_mem_max, + &ttm_mem_swap, + &ttm_mem_used, + NULL +}; + +static struct sysfs_ops ttm_mem_zone_ops = { + .show = &ttm_mem_zone_show, + .store = &ttm_mem_zone_store +}; + +static struct kobj_type ttm_mem_zone_kobj_type = { + .release = &ttm_mem_zone_kobj_release, + .sysfs_ops = &ttm_mem_zone_ops, + .default_attrs = ttm_mem_zone_attrs, +}; + +static void ttm_mem_global_kobj_release(struct kobject *kobj) +{ + struct ttm_mem_global *glob = + container_of(kobj, struct ttm_mem_global, kobj); + + kfree(glob); +} + +static struct kobj_type ttm_mem_glob_kobj_type = { + .release = &ttm_mem_global_kobj_release, +}; + +static bool ttm_zones_above_swap_target(struct ttm_mem_global *glob, + bool from_wq, uint64_t extra) +{ + unsigned int i; + struct ttm_mem_zone *zone; + uint64_t target; + + for (i = 0; i < glob->num_zones; ++i) { + zone = glob->zones[i]; + + if (from_wq) + target = zone->swap_limit; + else if (capable(CAP_SYS_ADMIN)) + target = zone->emer_mem; + else + target = zone->max_mem; + + target = (extra > target) ? 0ULL : target; + + if (zone->used_mem > target) + return true; + } + return false; +} + /** * At this point we only support a single shrink callback. * Extend this if needed, perhaps using a linked list of callbacks. @@ -42,34 +207,17 @@ * many threads may try to swap out at any given time. */ -static void ttm_shrink(struct ttm_mem_global *glob, bool from_workqueue, +static void ttm_shrink(struct ttm_mem_global *glob, bool from_wq, uint64_t extra) { int ret; struct ttm_mem_shrink *shrink; - uint64_t target; - uint64_t total_target; spin_lock(&glob->lock); if (glob->shrink == NULL) goto out; - if (from_workqueue) { - target = glob->swap_limit; - total_target = glob->total_memory_swap_limit; - } else if (capable(CAP_SYS_ADMIN)) { - total_target = glob->emer_total_memory; - target = glob->emer_memory; - } else { - total_target = glob->max_total_memory; - target = glob->max_memory; - } - - total_target = (extra >= total_target) ? 0 : total_target - extra; - target = (extra >= target) ? 0 : target - extra; - - while (glob->used_memory > target || - glob->used_total_memory > total_target) { + while (ttm_zones_above_swap_target(glob, from_wq, extra)) { shrink = glob->shrink; spin_unlock(&glob->lock); ret = shrink->do_shrink(shrink); @@ -81,6 +229,8 @@ out: spin_unlock(&glob->lock); } + + static void ttm_shrink_work(struct work_struct *work) { struct ttm_mem_global *glob = @@ -89,63 +239,178 @@ static void ttm_shrink_work(struct work_struct *work) ttm_shrink(glob, true, 0ULL); } +static int ttm_mem_init_kernel_zone(struct ttm_mem_global *glob, + const struct sysinfo *si) +{ + struct ttm_mem_zone *zone = kzalloc(sizeof(*zone), GFP_KERNEL); + uint64_t mem; + + if (unlikely(!zone)) + return -ENOMEM; + + mem = si->totalram - si->totalhigh; + mem *= si->mem_unit; + + zone->name = "kernel"; + zone->zone_mem = mem; + zone->max_mem = mem >> 1; + zone->emer_mem = (mem >> 1) + (mem >> 2); + zone->swap_limit = zone->max_mem - (mem >> 3); + zone->used_mem = 0; + zone->glob = glob; + glob->zone_kernel = zone; + glob->zones[glob->num_zones++] = zone; + kobject_init(&zone->kobj, &ttm_mem_zone_kobj_type); + return kobject_add(&zone->kobj, &glob->kobj, zone->name); +} + +#ifdef CONFIG_HIGHMEM +static int ttm_mem_init_highmem_zone(struct ttm_mem_global *glob, + const struct sysinfo *si) +{ + struct ttm_mem_zone *zone = kzalloc(sizeof(*zone), GFP_KERNEL); + uint64_t mem; + + if (unlikely(!zone)) + return -ENOMEM; + + if (si->totalhigh == 0) + return 0; + + mem = si->totalram; + mem *= si->mem_unit; + + zone->name = "highmem"; + zone->zone_mem = mem; + zone->max_mem = mem >> 1; + zone->emer_mem = (mem >> 1) + (mem >> 2); + zone->swap_limit = zone->max_mem - (mem >> 3); + zone->used_mem = 0; + zone->glob = glob; + glob->zone_highmem = zone; + glob->zones[glob->num_zones++] = zone; + kobject_init(&zone->kobj, &ttm_mem_zone_kobj_type); + return kobject_add(&zone->kobj, &glob->kobj, zone->name); +} +#else +static int ttm_mem_init_dma32_zone(struct ttm_mem_global *glob, + const struct sysinfo *si) +{ + struct ttm_mem_zone *zone = kzalloc(sizeof(*zone), GFP_KERNEL); + uint64_t mem; + + if (unlikely(!zone)) + return -ENOMEM; + + mem = si->totalram; + mem *= si->mem_unit; + + /** + * No special dma32 zone needed. + */ + + if (mem <= ((uint64_t) 1ULL << 32)) + return 0; + + /* + * Limit max dma32 memory to 4GB for now + * until we can figure out how big this + * zone really is. + */ + + mem = ((uint64_t) 1ULL << 32); + zone->name = "dma32"; + zone->zone_mem = mem; + zone->max_mem = mem >> 1; + zone->emer_mem = (mem >> 1) + (mem >> 2); + zone->swap_limit = zone->max_mem - (mem >> 3); + zone->used_mem = 0; + zone->glob = glob; + glob->zone_dma32 = zone; + glob->zones[glob->num_zones++] = zone; + kobject_init(&zone->kobj, &ttm_mem_zone_kobj_type); + return kobject_add(&zone->kobj, &glob->kobj, zone->name); +} +#endif + int ttm_mem_global_init(struct ttm_mem_global *glob) { struct sysinfo si; - uint64_t mem; + int ret; + int i; + struct ttm_mem_zone *zone; spin_lock_init(&glob->lock); glob->swap_queue = create_singlethread_workqueue("ttm_swap"); INIT_WORK(&glob->work, ttm_shrink_work); init_waitqueue_head(&glob->queue); + kobject_init(&glob->kobj, &ttm_mem_glob_kobj_type); + ret = kobject_add(&glob->kobj, + ttm_get_kobj(), + "memory_accounting"); + if (unlikely(ret != 0)) + goto out_no_zone; si_meminfo(&si); - mem = si.totalram - si.totalhigh; - mem *= si.mem_unit; - - glob->max_memory = mem >> 1; - glob->emer_memory = (mem >> 1) + (mem >> 2); - glob->swap_limit = glob->max_memory - (mem >> 3); - glob->used_memory = 0; - glob->used_total_memory = 0; - glob->shrink = NULL; - - mem = si.totalram; - mem *= si.mem_unit; - - glob->max_total_memory = mem >> 1; - glob->emer_total_memory = (mem >> 1) + (mem >> 2); - - glob->total_memory_swap_limit = glob->max_total_memory - (mem >> 3); - - printk(KERN_INFO TTM_PFX "TTM available graphics memory: %llu MiB\n", - glob->max_total_memory >> 20); - printk(KERN_INFO TTM_PFX "TTM available object memory: %llu MiB\n", - glob->max_memory >> 20); - + ret = ttm_mem_init_kernel_zone(glob, &si); + if (unlikely(ret != 0)) + goto out_no_zone; +#ifdef CONFIG_HIGHMEM + ret = ttm_mem_init_highmem_zone(glob, &si); + if (unlikely(ret != 0)) + goto out_no_zone; +#else + ret = ttm_mem_init_dma32_zone(glob, &si); + if (unlikely(ret != 0)) + goto out_no_zone; +#endif + for (i = 0; i < glob->num_zones; ++i) { + zone = glob->zones[i]; + printk(KERN_INFO TTM_PFX + "Zone %7s: Available graphics memory: %llu kiB.\n", + zone->name, (unsigned long long) zone->max_mem >> 10); + } return 0; +out_no_zone: + ttm_mem_global_release(glob); + return ret; } EXPORT_SYMBOL(ttm_mem_global_init); void ttm_mem_global_release(struct ttm_mem_global *glob) { - printk(KERN_INFO TTM_PFX "Used total memory is %llu bytes.\n", - (unsigned long long)glob->used_total_memory); + unsigned int i; + struct ttm_mem_zone *zone; + flush_workqueue(glob->swap_queue); destroy_workqueue(glob->swap_queue); glob->swap_queue = NULL; + for (i = 0; i < glob->num_zones; ++i) { + zone = glob->zones[i]; + kobject_del(&zone->kobj); + kobject_put(&zone->kobj); + } + kobject_del(&glob->kobj); + kobject_put(&glob->kobj); } EXPORT_SYMBOL(ttm_mem_global_release); -static inline void ttm_check_swapping(struct ttm_mem_global *glob) +static void ttm_check_swapping(struct ttm_mem_global *glob) { - bool needs_swapping; + bool needs_swapping = false; + unsigned int i; + struct ttm_mem_zone *zone; spin_lock(&glob->lock); - needs_swapping = (glob->used_memory > glob->swap_limit || - glob->used_total_memory > - glob->total_memory_swap_limit); + for (i = 0; i < glob->num_zones; ++i) { + zone = glob->zones[i]; + if (zone->used_mem > zone->swap_limit) { + needs_swapping = true; + break; + } + } + spin_unlock(&glob->lock); if (unlikely(needs_swapping)) @@ -153,44 +418,60 @@ static inline void ttm_check_swapping(struct ttm_mem_global *glob) } -void ttm_mem_global_free(struct ttm_mem_global *glob, - uint64_t amount, bool himem) +static void ttm_mem_global_free_zone(struct ttm_mem_global *glob, + struct ttm_mem_zone *single_zone, + uint64_t amount) { + unsigned int i; + struct ttm_mem_zone *zone; + spin_lock(&glob->lock); - glob->used_total_memory -= amount; - if (!himem) - glob->used_memory -= amount; - wake_up_all(&glob->queue); + for (i = 0; i < glob->num_zones; ++i) { + zone = glob->zones[i]; + if (single_zone && zone != single_zone) + continue; + zone->used_mem -= amount; + } spin_unlock(&glob->lock); } +void ttm_mem_global_free(struct ttm_mem_global *glob, + uint64_t amount) +{ + return ttm_mem_global_free_zone(glob, NULL, amount); +} + static int ttm_mem_global_reserve(struct ttm_mem_global *glob, - uint64_t amount, bool himem, bool reserve) + struct ttm_mem_zone *single_zone, + uint64_t amount, bool reserve) { uint64_t limit; - uint64_t lomem_limit; int ret = -ENOMEM; + unsigned int i; + struct ttm_mem_zone *zone; spin_lock(&glob->lock); + for (i = 0; i < glob->num_zones; ++i) { + zone = glob->zones[i]; + if (single_zone && zone != single_zone) + continue; - if (capable(CAP_SYS_ADMIN)) { - limit = glob->emer_total_memory; - lomem_limit = glob->emer_memory; - } else { - limit = glob->max_total_memory; - lomem_limit = glob->max_memory; - } + limit = (capable(CAP_SYS_ADMIN)) ? + zone->emer_mem : zone->max_mem; - if (unlikely(glob->used_total_memory + amount > limit)) - goto out_unlock; - if (unlikely(!himem && glob->used_memory + amount > lomem_limit)) - goto out_unlock; + if (zone->used_mem > limit) + goto out_unlock; + } if (reserve) { - glob->used_total_memory += amount; - if (!himem) - glob->used_memory += amount; + for (i = 0; i < glob->num_zones; ++i) { + zone = glob->zones[i]; + if (single_zone && zone != single_zone) + continue; + zone->used_mem += amount; + } } + ret = 0; out_unlock: spin_unlock(&glob->lock); @@ -199,12 +480,17 @@ out_unlock: return ret; } -int ttm_mem_global_alloc(struct ttm_mem_global *glob, uint64_t memory, - bool no_wait, bool interruptible, bool himem) + +static int ttm_mem_global_alloc_zone(struct ttm_mem_global *glob, + struct ttm_mem_zone *single_zone, + uint64_t memory, + bool no_wait, bool interruptible) { int count = TTM_MEMORY_ALLOC_RETRIES; - while (unlikely(ttm_mem_global_reserve(glob, memory, himem, true) + while (unlikely(ttm_mem_global_reserve(glob, + single_zone, + memory, true) != 0)) { if (no_wait) return -ENOMEM; @@ -216,6 +502,56 @@ int ttm_mem_global_alloc(struct ttm_mem_global *glob, uint64_t memory, return 0; } +int ttm_mem_global_alloc(struct ttm_mem_global *glob, uint64_t memory, + bool no_wait, bool interruptible) +{ + /** + * Normal allocations of kernel memory are registered in + * all zones. + */ + + return ttm_mem_global_alloc_zone(glob, NULL, memory, no_wait, + interruptible); +} + +int ttm_mem_global_alloc_page(struct ttm_mem_global *glob, + struct page *page, + bool no_wait, bool interruptible) +{ + + struct ttm_mem_zone *zone = NULL; + + /** + * Page allocations may be registed in a single zone + * only if highmem or !dma32. + */ + +#ifdef CONFIG_HIGHMEM + if (PageHighMem(page) && glob->zone_highmem != NULL) + zone = glob->zone_highmem; +#else + if (glob->zone_dma32 && page_to_pfn(page) > 0x00100000UL) + zone = glob->zone_kernel; +#endif + return ttm_mem_global_alloc_zone(glob, zone, PAGE_SIZE, no_wait, + interruptible); +} + +void ttm_mem_global_free_page(struct ttm_mem_global *glob, struct page *page) +{ + struct ttm_mem_zone *zone = NULL; + +#ifdef CONFIG_HIGHMEM + if (PageHighMem(page) && glob->zone_highmem != NULL) + zone = glob->zone_highmem; +#else + if (glob->zone_dma32 && page_to_pfn(page) > 0x00100000UL) + zone = glob->zone_kernel; +#endif + ttm_mem_global_free_zone(glob, zone, PAGE_SIZE); +} + + size_t ttm_round_pot(size_t size) { if ((size & (size - 1)) == 0) diff --git a/drivers/gpu/drm/ttm/ttm_tt.c b/drivers/gpu/drm/ttm/ttm_tt.c index 75dc8bd24592..4e1e2566d519 100644 --- a/drivers/gpu/drm/ttm/ttm_tt.c +++ b/drivers/gpu/drm/ttm/ttm_tt.c @@ -166,7 +166,7 @@ static void ttm_tt_free_user_pages(struct ttm_tt *ttm) set_page_dirty_lock(page); ttm->pages[i] = NULL; - ttm_mem_global_free(ttm->bdev->mem_glob, PAGE_SIZE, false); + ttm_mem_global_free(ttm->bdev->mem_glob, PAGE_SIZE); put_page(page); } ttm->state = tt_unpopulated; @@ -187,21 +187,14 @@ static struct page *__ttm_tt_get_page(struct ttm_tt *ttm, int index) if (!p) return NULL; - if (PageHighMem(p)) { - ret = - ttm_mem_global_alloc(mem_glob, PAGE_SIZE, - false, false, true); - if (unlikely(ret != 0)) - goto out_err; + ret = ttm_mem_global_alloc_page(mem_glob, p, false, false); + if (unlikely(ret != 0)) + goto out_err; + + if (PageHighMem(p)) ttm->pages[--ttm->first_himem_page] = p; - } else { - ret = - ttm_mem_global_alloc(mem_glob, PAGE_SIZE, - false, false, false); - if (unlikely(ret != 0)) - goto out_err; + else ttm->pages[++ttm->last_lomem_page] = p; - } } return p; out_err: @@ -355,8 +348,8 @@ static void ttm_tt_free_alloced_pages(struct ttm_tt *ttm) printk(KERN_ERR TTM_PFX "Erroneous page count. " "Leaking pages.\n"); - ttm_mem_global_free(ttm->bdev->mem_glob, PAGE_SIZE, - PageHighMem(cur_page)); + ttm_mem_global_free_page(ttm->bdev->mem_glob, + cur_page); __free_page(cur_page); } } @@ -411,7 +404,7 @@ int ttm_tt_set_user(struct ttm_tt *ttm, */ ret = ttm_mem_global_alloc(mem_glob, num_pages * PAGE_SIZE, - false, false, false); + false, false); if (unlikely(ret != 0)) return ret; @@ -422,7 +415,7 @@ int ttm_tt_set_user(struct ttm_tt *ttm, if (ret != num_pages && write) { ttm_tt_free_user_pages(ttm); - ttm_mem_global_free(mem_glob, num_pages * PAGE_SIZE, false); + ttm_mem_global_free(mem_glob, num_pages * PAGE_SIZE); return -ENOMEM; } diff --git a/include/drm/ttm/ttm_memory.h b/include/drm/ttm/ttm_memory.h index d8b8f042c4f1..6983a7cf4da4 100644 --- a/include/drm/ttm/ttm_memory.h +++ b/include/drm/ttm/ttm_memory.h @@ -32,6 +32,7 @@ #include #include #include +#include /** * struct ttm_mem_shrink - callback to shrink TTM memory usage. @@ -60,34 +61,33 @@ struct ttm_mem_shrink { * @queue: Wait queue for processes suspended waiting for memory. * @lock: Lock to protect the @shrink - and the memory accounting members, * that is, essentially the whole structure with some exceptions. - * @emer_memory: Lowmem memory limit available for root. - * @max_memory: Lowmem memory limit available for non-root. - * @swap_limit: Lowmem memory limit where the shrink workqueue kicks in. - * @used_memory: Currently used lowmem memory. - * @used_total_memory: Currently used total (lowmem + highmem) memory. - * @total_memory_swap_limit: Total memory limit where the shrink workqueue - * kicks in. - * @max_total_memory: Total memory available to non-root processes. - * @emer_total_memory: Total memory available to root processes. + * @zones: Array of pointers to accounting zones. + * @num_zones: Number of populated entries in the @zones array. + * @zone_kernel: Pointer to the kernel zone. + * @zone_highmem: Pointer to the highmem zone if there is one. + * @zone_dma32: Pointer to the dma32 zone if there is one. * * Note that this structure is not per device. It should be global for all * graphics devices. */ +#define TTM_MEM_MAX_ZONES 2 +struct ttm_mem_zone; struct ttm_mem_global { + struct kobject kobj; struct ttm_mem_shrink *shrink; struct workqueue_struct *swap_queue; struct work_struct work; wait_queue_head_t queue; spinlock_t lock; - uint64_t emer_memory; - uint64_t max_memory; - uint64_t swap_limit; - uint64_t used_memory; - uint64_t used_total_memory; - uint64_t total_memory_swap_limit; - uint64_t max_total_memory; - uint64_t emer_total_memory; + struct ttm_mem_zone *zones[TTM_MEM_MAX_ZONES]; + unsigned int num_zones; + struct ttm_mem_zone *zone_kernel; +#ifdef CONFIG_HIGHMEM + struct ttm_mem_zone *zone_highmem; +#else + struct ttm_mem_zone *zone_dma32; +#endif }; /** @@ -146,8 +146,13 @@ static inline void ttm_mem_unregister_shrink(struct ttm_mem_global *glob, extern int ttm_mem_global_init(struct ttm_mem_global *glob); extern void ttm_mem_global_release(struct ttm_mem_global *glob); extern int ttm_mem_global_alloc(struct ttm_mem_global *glob, uint64_t memory, - bool no_wait, bool interruptible, bool himem); + bool no_wait, bool interruptible); extern void ttm_mem_global_free(struct ttm_mem_global *glob, - uint64_t amount, bool himem); + uint64_t amount); +extern int ttm_mem_global_alloc_page(struct ttm_mem_global *glob, + struct page *page, + bool no_wait, bool interruptible); +extern void ttm_mem_global_free_page(struct ttm_mem_global *glob, + struct page *page); extern size_t ttm_round_pot(size_t size); #endif diff --git a/include/drm/ttm/ttm_module.h b/include/drm/ttm/ttm_module.h index 889a4c7958ae..0a72ac7c7e58 100644 --- a/include/drm/ttm/ttm_module.h +++ b/include/drm/ttm/ttm_module.h @@ -32,6 +32,7 @@ #define _TTM_MODULE_H_ #include +struct kobject; #define TTM_PFX "[TTM]" @@ -54,5 +55,6 @@ extern void ttm_global_init(void); extern void ttm_global_release(void); extern int ttm_global_item_ref(struct ttm_global_reference *ref); extern void ttm_global_item_unref(struct ttm_global_reference *ref); +extern struct kobject *ttm_get_kobj(void); #endif /* _TTM_MODULE_H_ */ -- cgit v1.2.3 From a987fcaa805fcb24ba885c2e29fd4fdb6816f08f Mon Sep 17 00:00:00 2001 From: Thomas Hellstrom Date: Tue, 18 Aug 2009 16:51:56 +0200 Subject: ttm: Make parts of a struct ttm_bo_device global. Common resources, like memory accounting and swap lists should be global and not per device. Introduce a struct ttm_bo_global to accomodate this, and register it with sysfs. Add a small sysfs interface to return the number of active buffer objects. Signed-off-by: Thomas Hellstrom Signed-off-by: Dave Airlie --- drivers/gpu/drm/radeon/radeon_object.h | 1 + drivers/gpu/drm/radeon/radeon_ttm.c | 33 +++- drivers/gpu/drm/ttm/ttm_bo.c | 292 ++++++++++++++++++++++----------- drivers/gpu/drm/ttm/ttm_bo_util.c | 4 +- drivers/gpu/drm/ttm/ttm_tt.c | 12 +- include/drm/ttm/ttm_bo_api.h | 1 + include/drm/ttm/ttm_bo_driver.h | 94 ++++++++--- 7 files changed, 296 insertions(+), 141 deletions(-) (limited to 'include') diff --git a/drivers/gpu/drm/radeon/radeon_object.h b/drivers/gpu/drm/radeon/radeon_object.h index 473e4775dc5a..10e8af6bb456 100644 --- a/drivers/gpu/drm/radeon/radeon_object.h +++ b/drivers/gpu/drm/radeon/radeon_object.h @@ -37,6 +37,7 @@ * TTM. */ struct radeon_mman { + struct ttm_bo_global_ref bo_global_ref; struct ttm_global_reference mem_global_ref; bool mem_global_referenced; struct ttm_bo_device bdev; diff --git a/drivers/gpu/drm/radeon/radeon_ttm.c b/drivers/gpu/drm/radeon/radeon_ttm.c index 1227a97f5169..343b6d6b99c6 100644 --- a/drivers/gpu/drm/radeon/radeon_ttm.c +++ b/drivers/gpu/drm/radeon/radeon_ttm.c @@ -77,9 +77,25 @@ static int radeon_ttm_global_init(struct radeon_device *rdev) global_ref->release = &radeon_ttm_mem_global_release; r = ttm_global_item_ref(global_ref); if (r != 0) { - DRM_ERROR("Failed referencing a global TTM memory object.\n"); + DRM_ERROR("Failed setting up TTM memory accounting " + "subsystem.\n"); return r; } + + rdev->mman.bo_global_ref.mem_glob = + rdev->mman.mem_global_ref.object; + global_ref = &rdev->mman.bo_global_ref.ref; + global_ref->global_type = TTM_GLOBAL_TTM_BO; + global_ref->size = sizeof(struct ttm_mem_global); + global_ref->init = &ttm_bo_global_init; + global_ref->release = &ttm_bo_global_release; + r = ttm_global_item_ref(global_ref); + if (r != 0) { + DRM_ERROR("Failed setting up TTM BO subsystem.\n"); + ttm_global_item_unref(&rdev->mman.mem_global_ref); + return r; + } + rdev->mman.mem_global_referenced = true; return 0; } @@ -87,6 +103,7 @@ static int radeon_ttm_global_init(struct radeon_device *rdev) static void radeon_ttm_global_fini(struct radeon_device *rdev) { if (rdev->mman.mem_global_referenced) { + ttm_global_item_unref(&rdev->mman.bo_global_ref.ref); ttm_global_item_unref(&rdev->mman.mem_global_ref); rdev->mman.mem_global_referenced = false; } @@ -286,9 +303,11 @@ static int radeon_move_vram_ram(struct ttm_buffer_object *bo, r = ttm_bo_move_ttm(bo, true, no_wait, new_mem); out_cleanup: if (tmp_mem.mm_node) { - spin_lock(&rdev->mman.bdev.lru_lock); + struct ttm_bo_global *glob = rdev->mman.bdev.glob; + + spin_lock(&glob->lru_lock); drm_mm_put_block(tmp_mem.mm_node); - spin_unlock(&rdev->mman.bdev.lru_lock); + spin_unlock(&glob->lru_lock); return r; } return r; @@ -323,9 +342,11 @@ static int radeon_move_ram_vram(struct ttm_buffer_object *bo, } out_cleanup: if (tmp_mem.mm_node) { - spin_lock(&rdev->mman.bdev.lru_lock); + struct ttm_bo_global *glob = rdev->mman.bdev.glob; + + spin_lock(&glob->lru_lock); drm_mm_put_block(tmp_mem.mm_node); - spin_unlock(&rdev->mman.bdev.lru_lock); + spin_unlock(&glob->lru_lock); return r; } return r; @@ -441,7 +462,7 @@ int radeon_ttm_init(struct radeon_device *rdev) } /* No others user of address space so set it to 0 */ r = ttm_bo_device_init(&rdev->mman.bdev, - rdev->mman.mem_global_ref.object, + rdev->mman.bo_global_ref.ref.object, &radeon_bo_driver, DRM_FILE_PAGE_OFFSET); if (r) { DRM_ERROR("failed initializing buffer object driver(%d).\n", r); diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c index f16909ceec93..0d0b1b7afbcf 100644 --- a/drivers/gpu/drm/ttm/ttm_bo.c +++ b/drivers/gpu/drm/ttm/ttm_bo.c @@ -45,6 +45,39 @@ static int ttm_bo_setup_vm(struct ttm_buffer_object *bo); static void ttm_bo_unmap_virtual(struct ttm_buffer_object *bo); static int ttm_bo_swapout(struct ttm_mem_shrink *shrink); +static void ttm_bo_global_kobj_release(struct kobject *kobj); + +static struct attribute ttm_bo_count = { + .name = "bo_count", + .mode = S_IRUGO +}; + +static ssize_t ttm_bo_global_show(struct kobject *kobj, + struct attribute *attr, + char *buffer) +{ + struct ttm_bo_global *glob = + container_of(kobj, struct ttm_bo_global, kobj); + + return snprintf(buffer, PAGE_SIZE, "%lu\n", + (unsigned long) atomic_read(&glob->bo_count)); +} + +static struct attribute *ttm_bo_global_attrs[] = { + &ttm_bo_count, + NULL +}; + +static struct sysfs_ops ttm_bo_global_ops = { + .show = &ttm_bo_global_show +}; + +static struct kobj_type ttm_bo_glob_kobj_type = { + .release = &ttm_bo_global_kobj_release, + .sysfs_ops = &ttm_bo_global_ops, + .default_attrs = ttm_bo_global_attrs +}; + static inline uint32_t ttm_bo_type_flags(unsigned type) { @@ -67,10 +100,11 @@ static void ttm_bo_release_list(struct kref *list_kref) if (bo->ttm) ttm_tt_destroy(bo->ttm); + atomic_dec(&bo->glob->bo_count); if (bo->destroy) bo->destroy(bo); else { - ttm_mem_global_free(bdev->mem_glob, bo->acc_size); + ttm_mem_global_free(bdev->glob->mem_glob, bo->acc_size); kfree(bo); } } @@ -107,7 +141,7 @@ static void ttm_bo_add_to_lru(struct ttm_buffer_object *bo) kref_get(&bo->list_kref); if (bo->ttm != NULL) { - list_add_tail(&bo->swap, &bdev->swap_lru); + list_add_tail(&bo->swap, &bo->glob->swap_lru); kref_get(&bo->list_kref); } } @@ -142,7 +176,7 @@ int ttm_bo_reserve_locked(struct ttm_buffer_object *bo, bool interruptible, bool no_wait, bool use_sequence, uint32_t sequence) { - struct ttm_bo_device *bdev = bo->bdev; + struct ttm_bo_global *glob = bo->glob; int ret; while (unlikely(atomic_cmpxchg(&bo->reserved, 0, 1) != 0)) { @@ -154,9 +188,9 @@ int ttm_bo_reserve_locked(struct ttm_buffer_object *bo, if (no_wait) return -EBUSY; - spin_unlock(&bdev->lru_lock); + spin_unlock(&glob->lru_lock); ret = ttm_bo_wait_unreserved(bo, interruptible); - spin_lock(&bdev->lru_lock); + spin_lock(&glob->lru_lock); if (unlikely(ret)) return ret; @@ -182,16 +216,16 @@ int ttm_bo_reserve(struct ttm_buffer_object *bo, bool interruptible, bool no_wait, bool use_sequence, uint32_t sequence) { - struct ttm_bo_device *bdev = bo->bdev; + struct ttm_bo_global *glob = bo->glob; int put_count = 0; int ret; - spin_lock(&bdev->lru_lock); + spin_lock(&glob->lru_lock); ret = ttm_bo_reserve_locked(bo, interruptible, no_wait, use_sequence, sequence); if (likely(ret == 0)) put_count = ttm_bo_del_from_lru(bo); - spin_unlock(&bdev->lru_lock); + spin_unlock(&glob->lru_lock); while (put_count--) kref_put(&bo->list_kref, ttm_bo_ref_bug); @@ -201,13 +235,13 @@ int ttm_bo_reserve(struct ttm_buffer_object *bo, void ttm_bo_unreserve(struct ttm_buffer_object *bo) { - struct ttm_bo_device *bdev = bo->bdev; + struct ttm_bo_global *glob = bo->glob; - spin_lock(&bdev->lru_lock); + spin_lock(&glob->lru_lock); ttm_bo_add_to_lru(bo); atomic_set(&bo->reserved, 0); wake_up_all(&bo->event_queue); - spin_unlock(&bdev->lru_lock); + spin_unlock(&glob->lru_lock); } EXPORT_SYMBOL(ttm_bo_unreserve); @@ -218,6 +252,7 @@ EXPORT_SYMBOL(ttm_bo_unreserve); static int ttm_bo_add_ttm(struct ttm_buffer_object *bo, bool zero_alloc) { struct ttm_bo_device *bdev = bo->bdev; + struct ttm_bo_global *glob = bo->glob; int ret = 0; uint32_t page_flags = 0; @@ -230,14 +265,14 @@ static int ttm_bo_add_ttm(struct ttm_buffer_object *bo, bool zero_alloc) page_flags |= TTM_PAGE_FLAG_ZERO_ALLOC; case ttm_bo_type_kernel: bo->ttm = ttm_tt_create(bdev, bo->num_pages << PAGE_SHIFT, - page_flags, bdev->dummy_read_page); + page_flags, glob->dummy_read_page); if (unlikely(bo->ttm == NULL)) ret = -ENOMEM; break; case ttm_bo_type_user: bo->ttm = ttm_tt_create(bdev, bo->num_pages << PAGE_SHIFT, page_flags | TTM_PAGE_FLAG_USER, - bdev->dummy_read_page); + glob->dummy_read_page); if (unlikely(bo->ttm == NULL)) ret = -ENOMEM; break; @@ -355,6 +390,7 @@ out_err: static int ttm_bo_cleanup_refs(struct ttm_buffer_object *bo, bool remove_all) { struct ttm_bo_device *bdev = bo->bdev; + struct ttm_bo_global *glob = bo->glob; struct ttm_bo_driver *driver = bdev->driver; int ret; @@ -366,7 +402,7 @@ static int ttm_bo_cleanup_refs(struct ttm_buffer_object *bo, bool remove_all) spin_unlock(&bo->lock); - spin_lock(&bdev->lru_lock); + spin_lock(&glob->lru_lock); ret = ttm_bo_reserve_locked(bo, false, false, false, 0); BUG_ON(ret); if (bo->ttm) @@ -381,7 +417,7 @@ static int ttm_bo_cleanup_refs(struct ttm_buffer_object *bo, bool remove_all) bo->mem.mm_node = NULL; } put_count = ttm_bo_del_from_lru(bo); - spin_unlock(&bdev->lru_lock); + spin_unlock(&glob->lru_lock); atomic_set(&bo->reserved, 0); @@ -391,14 +427,14 @@ static int ttm_bo_cleanup_refs(struct ttm_buffer_object *bo, bool remove_all) return 0; } - spin_lock(&bdev->lru_lock); + spin_lock(&glob->lru_lock); if (list_empty(&bo->ddestroy)) { void *sync_obj = bo->sync_obj; void *sync_obj_arg = bo->sync_obj_arg; kref_get(&bo->list_kref); list_add_tail(&bo->ddestroy, &bdev->ddestroy); - spin_unlock(&bdev->lru_lock); + spin_unlock(&glob->lru_lock); spin_unlock(&bo->lock); if (sync_obj) @@ -408,7 +444,7 @@ static int ttm_bo_cleanup_refs(struct ttm_buffer_object *bo, bool remove_all) ret = 0; } else { - spin_unlock(&bdev->lru_lock); + spin_unlock(&glob->lru_lock); spin_unlock(&bo->lock); ret = -EBUSY; } @@ -423,11 +459,12 @@ static int ttm_bo_cleanup_refs(struct ttm_buffer_object *bo, bool remove_all) static int ttm_bo_delayed_delete(struct ttm_bo_device *bdev, bool remove_all) { + struct ttm_bo_global *glob = bdev->glob; struct ttm_buffer_object *entry, *nentry; struct list_head *list, *next; int ret; - spin_lock(&bdev->lru_lock); + spin_lock(&glob->lru_lock); list_for_each_safe(list, next, &bdev->ddestroy) { entry = list_entry(list, struct ttm_buffer_object, ddestroy); nentry = NULL; @@ -444,16 +481,16 @@ static int ttm_bo_delayed_delete(struct ttm_bo_device *bdev, bool remove_all) } kref_get(&entry->list_kref); - spin_unlock(&bdev->lru_lock); + spin_unlock(&glob->lru_lock); ret = ttm_bo_cleanup_refs(entry, remove_all); kref_put(&entry->list_kref, ttm_bo_release_list); - spin_lock(&bdev->lru_lock); + spin_lock(&glob->lru_lock); if (nentry) { bool next_onlist = !list_empty(next); - spin_unlock(&bdev->lru_lock); + spin_unlock(&glob->lru_lock); kref_put(&nentry->list_kref, ttm_bo_release_list); - spin_lock(&bdev->lru_lock); + spin_lock(&glob->lru_lock); /* * Someone might have raced us and removed the * next entry from the list. We don't bother restarting @@ -467,7 +504,7 @@ static int ttm_bo_delayed_delete(struct ttm_bo_device *bdev, bool remove_all) break; } ret = !list_empty(&bdev->ddestroy); - spin_unlock(&bdev->lru_lock); + spin_unlock(&glob->lru_lock); return ret; } @@ -517,6 +554,7 @@ static int ttm_bo_evict(struct ttm_buffer_object *bo, unsigned mem_type, { int ret = 0; struct ttm_bo_device *bdev = bo->bdev; + struct ttm_bo_global *glob = bo->glob; struct ttm_mem_reg evict_mem; uint32_t proposed_placement; @@ -565,12 +603,12 @@ static int ttm_bo_evict(struct ttm_buffer_object *bo, unsigned mem_type, goto out; } - spin_lock(&bdev->lru_lock); + spin_lock(&glob->lru_lock); if (evict_mem.mm_node) { drm_mm_put_block(evict_mem.mm_node); evict_mem.mm_node = NULL; } - spin_unlock(&bdev->lru_lock); + spin_unlock(&glob->lru_lock); bo->evicted = true; out: return ret; @@ -585,6 +623,7 @@ static int ttm_bo_mem_force_space(struct ttm_bo_device *bdev, uint32_t mem_type, bool interruptible, bool no_wait) { + struct ttm_bo_global *glob = bdev->glob; struct drm_mm_node *node; struct ttm_buffer_object *entry; struct ttm_mem_type_manager *man = &bdev->man[mem_type]; @@ -598,7 +637,7 @@ retry_pre_get: if (unlikely(ret != 0)) return ret; - spin_lock(&bdev->lru_lock); + spin_lock(&glob->lru_lock); do { node = drm_mm_search_free(&man->manager, num_pages, mem->page_alignment, 1); @@ -619,7 +658,7 @@ retry_pre_get: if (likely(ret == 0)) put_count = ttm_bo_del_from_lru(entry); - spin_unlock(&bdev->lru_lock); + spin_unlock(&glob->lru_lock); if (unlikely(ret != 0)) return ret; @@ -635,21 +674,21 @@ retry_pre_get: if (ret) return ret; - spin_lock(&bdev->lru_lock); + spin_lock(&glob->lru_lock); } while (1); if (!node) { - spin_unlock(&bdev->lru_lock); + spin_unlock(&glob->lru_lock); return -ENOMEM; } node = drm_mm_get_block_atomic(node, num_pages, mem->page_alignment); if (unlikely(!node)) { - spin_unlock(&bdev->lru_lock); + spin_unlock(&glob->lru_lock); goto retry_pre_get; } - spin_unlock(&bdev->lru_lock); + spin_unlock(&glob->lru_lock); mem->mm_node = node; mem->mem_type = mem_type; return 0; @@ -697,6 +736,7 @@ int ttm_bo_mem_space(struct ttm_buffer_object *bo, bool interruptible, bool no_wait) { struct ttm_bo_device *bdev = bo->bdev; + struct ttm_bo_global *glob = bo->glob; struct ttm_mem_type_manager *man; uint32_t num_prios = bdev->driver->num_mem_type_prio; @@ -733,20 +773,20 @@ int ttm_bo_mem_space(struct ttm_buffer_object *bo, if (unlikely(ret)) return ret; - spin_lock(&bdev->lru_lock); + spin_lock(&glob->lru_lock); node = drm_mm_search_free(&man->manager, mem->num_pages, mem->page_alignment, 1); if (unlikely(!node)) { - spin_unlock(&bdev->lru_lock); + spin_unlock(&glob->lru_lock); break; } node = drm_mm_get_block_atomic(node, mem->num_pages, mem-> page_alignment); - spin_unlock(&bdev->lru_lock); + spin_unlock(&glob->lru_lock); } while (!node); } if (node) @@ -816,7 +856,7 @@ int ttm_bo_move_buffer(struct ttm_buffer_object *bo, uint32_t proposed_placement, bool interruptible, bool no_wait) { - struct ttm_bo_device *bdev = bo->bdev; + struct ttm_bo_global *glob = bo->glob; int ret = 0; struct ttm_mem_reg mem; @@ -852,9 +892,9 @@ int ttm_bo_move_buffer(struct ttm_buffer_object *bo, out_unlock: if (ret && mem.mm_node) { - spin_lock(&bdev->lru_lock); + spin_lock(&glob->lru_lock); drm_mm_put_block(mem.mm_node); - spin_unlock(&bdev->lru_lock); + spin_unlock(&glob->lru_lock); } return ret; } @@ -990,6 +1030,7 @@ int ttm_buffer_object_init(struct ttm_bo_device *bdev, INIT_LIST_HEAD(&bo->ddestroy); INIT_LIST_HEAD(&bo->swap); bo->bdev = bdev; + bo->glob = bdev->glob; bo->type = type; bo->num_pages = num_pages; bo->mem.mem_type = TTM_PL_SYSTEM; @@ -1002,6 +1043,7 @@ int ttm_buffer_object_init(struct ttm_bo_device *bdev, bo->seq_valid = false; bo->persistant_swap_storage = persistant_swap_storage; bo->acc_size = acc_size; + atomic_inc(&bo->glob->bo_count); ret = ttm_bo_check_placement(bo, flags, 0ULL); if (unlikely(ret != 0)) @@ -1040,13 +1082,13 @@ out_err: } EXPORT_SYMBOL(ttm_buffer_object_init); -static inline size_t ttm_bo_size(struct ttm_bo_device *bdev, +static inline size_t ttm_bo_size(struct ttm_bo_global *glob, unsigned long num_pages) { size_t page_array_size = (num_pages * sizeof(void *) + PAGE_SIZE - 1) & PAGE_MASK; - return bdev->ttm_bo_size + 2 * page_array_size; + return glob->ttm_bo_size + 2 * page_array_size; } int ttm_buffer_object_create(struct ttm_bo_device *bdev, @@ -1061,10 +1103,10 @@ int ttm_buffer_object_create(struct ttm_bo_device *bdev, { struct ttm_buffer_object *bo; int ret; - struct ttm_mem_global *mem_glob = bdev->mem_glob; + struct ttm_mem_global *mem_glob = bdev->glob->mem_glob; size_t acc_size = - ttm_bo_size(bdev, (size + PAGE_SIZE - 1) >> PAGE_SHIFT); + ttm_bo_size(bdev->glob, (size + PAGE_SIZE - 1) >> PAGE_SHIFT); ret = ttm_mem_global_alloc(mem_glob, acc_size, false, false); if (unlikely(ret != 0)) return ret; @@ -1118,6 +1160,7 @@ static int ttm_bo_force_list_clean(struct ttm_bo_device *bdev, struct list_head *head, unsigned mem_type, bool allow_errors) { + struct ttm_bo_global *glob = bdev->glob; struct ttm_buffer_object *entry; int ret; int put_count; @@ -1126,30 +1169,31 @@ static int ttm_bo_force_list_clean(struct ttm_bo_device *bdev, * Can't use standard list traversal since we're unlocking. */ - spin_lock(&bdev->lru_lock); + spin_lock(&glob->lru_lock); while (!list_empty(head)) { entry = list_first_entry(head, struct ttm_buffer_object, lru); kref_get(&entry->list_kref); ret = ttm_bo_reserve_locked(entry, false, false, false, 0); put_count = ttm_bo_del_from_lru(entry); - spin_unlock(&bdev->lru_lock); + spin_unlock(&glob->lru_lock); while (put_count--) kref_put(&entry->list_kref, ttm_bo_ref_bug); BUG_ON(ret); ret = ttm_bo_leave_list(entry, mem_type, allow_errors); ttm_bo_unreserve(entry); kref_put(&entry->list_kref, ttm_bo_release_list); - spin_lock(&bdev->lru_lock); + spin_lock(&glob->lru_lock); } - spin_unlock(&bdev->lru_lock); + spin_unlock(&glob->lru_lock); return 0; } int ttm_bo_clean_mm(struct ttm_bo_device *bdev, unsigned mem_type) { + struct ttm_bo_global *glob = bdev->glob; struct ttm_mem_type_manager *man = &bdev->man[mem_type]; int ret = -EINVAL; @@ -1171,13 +1215,13 @@ int ttm_bo_clean_mm(struct ttm_bo_device *bdev, unsigned mem_type) if (mem_type > 0) { ttm_bo_force_list_clean(bdev, &man->lru, mem_type, false); - spin_lock(&bdev->lru_lock); + spin_lock(&glob->lru_lock); if (drm_mm_clean(&man->manager)) drm_mm_takedown(&man->manager); else ret = -EBUSY; - spin_unlock(&bdev->lru_lock); + spin_unlock(&glob->lru_lock); } return ret; @@ -1251,11 +1295,83 @@ int ttm_bo_init_mm(struct ttm_bo_device *bdev, unsigned type, } EXPORT_SYMBOL(ttm_bo_init_mm); +static void ttm_bo_global_kobj_release(struct kobject *kobj) +{ + struct ttm_bo_global *glob = + container_of(kobj, struct ttm_bo_global, kobj); + + printk(KERN_INFO TTM_PFX "Freeing bo global.\n"); + ttm_mem_unregister_shrink(glob->mem_glob, &glob->shrink); + __free_page(glob->dummy_read_page); + kfree(glob); +} + +void ttm_bo_global_release(struct ttm_global_reference *ref) +{ + struct ttm_bo_global *glob = ref->object; + + kobject_del(&glob->kobj); + kobject_put(&glob->kobj); +} +EXPORT_SYMBOL(ttm_bo_global_release); + +int ttm_bo_global_init(struct ttm_global_reference *ref) +{ + struct ttm_bo_global_ref *bo_ref = + container_of(ref, struct ttm_bo_global_ref, ref); + struct ttm_bo_global *glob = ref->object; + int ret; + + mutex_init(&glob->device_list_mutex); + spin_lock_init(&glob->lru_lock); + glob->mem_glob = bo_ref->mem_glob; + glob->dummy_read_page = alloc_page(__GFP_ZERO | GFP_DMA32); + + if (unlikely(glob->dummy_read_page == NULL)) { + ret = -ENOMEM; + goto out_no_drp; + } + + INIT_LIST_HEAD(&glob->swap_lru); + INIT_LIST_HEAD(&glob->device_list); + + ttm_mem_init_shrink(&glob->shrink, ttm_bo_swapout); + ret = ttm_mem_register_shrink(glob->mem_glob, &glob->shrink); + if (unlikely(ret != 0)) { + printk(KERN_ERR TTM_PFX + "Could not register buffer object swapout.\n"); + goto out_no_shrink; + } + + glob->ttm_bo_extra_size = + ttm_round_pot(sizeof(struct ttm_tt)) + + ttm_round_pot(sizeof(struct ttm_backend)); + + glob->ttm_bo_size = glob->ttm_bo_extra_size + + ttm_round_pot(sizeof(struct ttm_buffer_object)); + + atomic_set(&glob->bo_count, 0); + + kobject_init(&glob->kobj, &ttm_bo_glob_kobj_type); + ret = kobject_add(&glob->kobj, ttm_get_kobj(), "buffer_objects"); + if (unlikely(ret != 0)) + kobject_put(&glob->kobj); + return ret; +out_no_shrink: + __free_page(glob->dummy_read_page); +out_no_drp: + kfree(glob); + return ret; +} +EXPORT_SYMBOL(ttm_bo_global_init); + + int ttm_bo_device_release(struct ttm_bo_device *bdev) { int ret = 0; unsigned i = TTM_NUM_MEM_TYPES; struct ttm_mem_type_manager *man; + struct ttm_bo_global *glob = bdev->glob; while (i--) { man = &bdev->man[i]; @@ -1271,98 +1387,74 @@ int ttm_bo_device_release(struct ttm_bo_device *bdev) } } + mutex_lock(&glob->device_list_mutex); + list_del(&bdev->device_list); + mutex_unlock(&glob->device_list_mutex); + if (!cancel_delayed_work(&bdev->wq)) flush_scheduled_work(); while (ttm_bo_delayed_delete(bdev, true)) ; - spin_lock(&bdev->lru_lock); + spin_lock(&glob->lru_lock); if (list_empty(&bdev->ddestroy)) TTM_DEBUG("Delayed destroy list was clean\n"); if (list_empty(&bdev->man[0].lru)) TTM_DEBUG("Swap list was clean\n"); - spin_unlock(&bdev->lru_lock); + spin_unlock(&glob->lru_lock); - ttm_mem_unregister_shrink(bdev->mem_glob, &bdev->shrink); BUG_ON(!drm_mm_clean(&bdev->addr_space_mm)); write_lock(&bdev->vm_lock); drm_mm_takedown(&bdev->addr_space_mm); write_unlock(&bdev->vm_lock); - __free_page(bdev->dummy_read_page); return ret; } EXPORT_SYMBOL(ttm_bo_device_release); -/* - * This function is intended to be called on drm driver load. - * If you decide to call it from firstopen, you must protect the call - * from a potentially racing ttm_bo_driver_finish in lastclose. - * (This may happen on X server restart). - */ - int ttm_bo_device_init(struct ttm_bo_device *bdev, - struct ttm_mem_global *mem_glob, - struct ttm_bo_driver *driver, uint64_t file_page_offset) + struct ttm_bo_global *glob, + struct ttm_bo_driver *driver, + uint64_t file_page_offset) { int ret = -EINVAL; - bdev->dummy_read_page = NULL; rwlock_init(&bdev->vm_lock); - spin_lock_init(&bdev->lru_lock); + spin_lock_init(&glob->lru_lock); bdev->driver = driver; - bdev->mem_glob = mem_glob; memset(bdev->man, 0, sizeof(bdev->man)); - bdev->dummy_read_page = alloc_page(__GFP_ZERO | GFP_DMA32); - if (unlikely(bdev->dummy_read_page == NULL)) { - ret = -ENOMEM; - goto out_err0; - } - /* * Initialize the system memory buffer type. * Other types need to be driver / IOCTL initialized. */ ret = ttm_bo_init_mm(bdev, TTM_PL_SYSTEM, 0, 0); if (unlikely(ret != 0)) - goto out_err1; + goto out_no_sys; bdev->addr_space_rb = RB_ROOT; ret = drm_mm_init(&bdev->addr_space_mm, file_page_offset, 0x10000000); if (unlikely(ret != 0)) - goto out_err2; + goto out_no_addr_mm; INIT_DELAYED_WORK(&bdev->wq, ttm_bo_delayed_workqueue); bdev->nice_mode = true; INIT_LIST_HEAD(&bdev->ddestroy); - INIT_LIST_HEAD(&bdev->swap_lru); bdev->dev_mapping = NULL; - ttm_mem_init_shrink(&bdev->shrink, ttm_bo_swapout); - ret = ttm_mem_register_shrink(mem_glob, &bdev->shrink); - if (unlikely(ret != 0)) { - printk(KERN_ERR TTM_PFX - "Could not register buffer object swapout.\n"); - goto out_err2; - } + bdev->glob = glob; - bdev->ttm_bo_extra_size = - ttm_round_pot(sizeof(struct ttm_tt)) + - ttm_round_pot(sizeof(struct ttm_backend)); - - bdev->ttm_bo_size = bdev->ttm_bo_extra_size + - ttm_round_pot(sizeof(struct ttm_buffer_object)); + mutex_lock(&glob->device_list_mutex); + list_add_tail(&bdev->device_list, &glob->device_list); + mutex_unlock(&glob->device_list_mutex); return 0; -out_err2: +out_no_addr_mm: ttm_bo_clean_mm(bdev, 0); -out_err1: - __free_page(bdev->dummy_read_page); -out_err0: +out_no_sys: return ret; } EXPORT_SYMBOL(ttm_bo_device_init); @@ -1607,21 +1699,21 @@ void ttm_bo_synccpu_write_release(struct ttm_buffer_object *bo) static int ttm_bo_swapout(struct ttm_mem_shrink *shrink) { - struct ttm_bo_device *bdev = - container_of(shrink, struct ttm_bo_device, shrink); + struct ttm_bo_global *glob = + container_of(shrink, struct ttm_bo_global, shrink); struct ttm_buffer_object *bo; int ret = -EBUSY; int put_count; uint32_t swap_placement = (TTM_PL_FLAG_CACHED | TTM_PL_FLAG_SYSTEM); - spin_lock(&bdev->lru_lock); + spin_lock(&glob->lru_lock); while (ret == -EBUSY) { - if (unlikely(list_empty(&bdev->swap_lru))) { - spin_unlock(&bdev->lru_lock); + if (unlikely(list_empty(&glob->swap_lru))) { + spin_unlock(&glob->lru_lock); return -EBUSY; } - bo = list_first_entry(&bdev->swap_lru, + bo = list_first_entry(&glob->swap_lru, struct ttm_buffer_object, swap); kref_get(&bo->list_kref); @@ -1633,16 +1725,16 @@ static int ttm_bo_swapout(struct ttm_mem_shrink *shrink) ret = ttm_bo_reserve_locked(bo, false, true, false, 0); if (unlikely(ret == -EBUSY)) { - spin_unlock(&bdev->lru_lock); + spin_unlock(&glob->lru_lock); ttm_bo_wait_unreserved(bo, false); kref_put(&bo->list_kref, ttm_bo_release_list); - spin_lock(&bdev->lru_lock); + spin_lock(&glob->lru_lock); } } BUG_ON(ret != 0); put_count = ttm_bo_del_from_lru(bo); - spin_unlock(&bdev->lru_lock); + spin_unlock(&glob->lru_lock); while (put_count--) kref_put(&bo->list_kref, ttm_bo_ref_bug); @@ -1696,6 +1788,6 @@ out: void ttm_bo_swapout_all(struct ttm_bo_device *bdev) { - while (ttm_bo_swapout(&bdev->shrink) == 0) + while (ttm_bo_swapout(&bdev->glob->shrink) == 0) ; } diff --git a/drivers/gpu/drm/ttm/ttm_bo_util.c b/drivers/gpu/drm/ttm/ttm_bo_util.c index bdec583901eb..12cd47aa18ce 100644 --- a/drivers/gpu/drm/ttm/ttm_bo_util.c +++ b/drivers/gpu/drm/ttm/ttm_bo_util.c @@ -41,9 +41,9 @@ void ttm_bo_free_old_node(struct ttm_buffer_object *bo) struct ttm_mem_reg *old_mem = &bo->mem; if (old_mem->mm_node) { - spin_lock(&bo->bdev->lru_lock); + spin_lock(&bo->glob->lru_lock); drm_mm_put_block(old_mem->mm_node); - spin_unlock(&bo->bdev->lru_lock); + spin_unlock(&bo->glob->lru_lock); } old_mem->mm_node = NULL; } diff --git a/drivers/gpu/drm/ttm/ttm_tt.c b/drivers/gpu/drm/ttm/ttm_tt.c index 4e1e2566d519..b0f73096d372 100644 --- a/drivers/gpu/drm/ttm/ttm_tt.c +++ b/drivers/gpu/drm/ttm/ttm_tt.c @@ -166,7 +166,7 @@ static void ttm_tt_free_user_pages(struct ttm_tt *ttm) set_page_dirty_lock(page); ttm->pages[i] = NULL; - ttm_mem_global_free(ttm->bdev->mem_glob, PAGE_SIZE); + ttm_mem_global_free(ttm->glob->mem_glob, PAGE_SIZE); put_page(page); } ttm->state = tt_unpopulated; @@ -177,8 +177,7 @@ static void ttm_tt_free_user_pages(struct ttm_tt *ttm) static struct page *__ttm_tt_get_page(struct ttm_tt *ttm, int index) { struct page *p; - struct ttm_bo_device *bdev = ttm->bdev; - struct ttm_mem_global *mem_glob = bdev->mem_glob; + struct ttm_mem_global *mem_glob = ttm->glob->mem_glob; int ret; while (NULL == (p = ttm->pages[index])) { @@ -348,7 +347,7 @@ static void ttm_tt_free_alloced_pages(struct ttm_tt *ttm) printk(KERN_ERR TTM_PFX "Erroneous page count. " "Leaking pages.\n"); - ttm_mem_global_free_page(ttm->bdev->mem_glob, + ttm_mem_global_free_page(ttm->glob->mem_glob, cur_page); __free_page(cur_page); } @@ -394,7 +393,7 @@ int ttm_tt_set_user(struct ttm_tt *ttm, struct mm_struct *mm = tsk->mm; int ret; int write = (ttm->page_flags & TTM_PAGE_FLAG_WRITE) != 0; - struct ttm_mem_global *mem_glob = ttm->bdev->mem_glob; + struct ttm_mem_global *mem_glob = ttm->glob->mem_glob; BUG_ON(num_pages != ttm->num_pages); BUG_ON((ttm->page_flags & TTM_PAGE_FLAG_USER) == 0); @@ -439,8 +438,7 @@ struct ttm_tt *ttm_tt_create(struct ttm_bo_device *bdev, unsigned long size, if (!ttm) return NULL; - ttm->bdev = bdev; - + ttm->glob = bdev->glob; ttm->num_pages = (size + PAGE_SIZE - 1) >> PAGE_SHIFT; ttm->first_himem_page = ttm->num_pages; ttm->last_lomem_page = -1; diff --git a/include/drm/ttm/ttm_bo_api.h b/include/drm/ttm/ttm_bo_api.h index 99dc521aa1a9..491146170522 100644 --- a/include/drm/ttm/ttm_bo_api.h +++ b/include/drm/ttm/ttm_bo_api.h @@ -155,6 +155,7 @@ struct ttm_buffer_object { * Members constant at init. */ + struct ttm_bo_global *glob; struct ttm_bo_device *bdev; unsigned long buffer_start; enum ttm_bo_type type; diff --git a/include/drm/ttm/ttm_bo_driver.h b/include/drm/ttm/ttm_bo_driver.h index 62ed733c52a2..9dc32f70b9a2 100644 --- a/include/drm/ttm/ttm_bo_driver.h +++ b/include/drm/ttm/ttm_bo_driver.h @@ -32,6 +32,7 @@ #include "ttm/ttm_bo_api.h" #include "ttm/ttm_memory.h" +#include "ttm/ttm_module.h" #include "drm_mm.h" #include "linux/workqueue.h" #include "linux/fs.h" @@ -160,7 +161,7 @@ struct ttm_tt { long last_lomem_page; uint32_t page_flags; unsigned long num_pages; - struct ttm_bo_device *bdev; + struct ttm_bo_global *glob; struct ttm_backend *be; struct task_struct *tsk; unsigned long start; @@ -355,24 +356,73 @@ struct ttm_bo_driver { void *(*sync_obj_ref) (void *sync_obj); }; -#define TTM_NUM_MEM_TYPES 8 +/** + * struct ttm_bo_global_ref - Argument to initialize a struct ttm_bo_global. + */ + +struct ttm_bo_global_ref { + struct ttm_global_reference ref; + struct ttm_mem_global *mem_glob; +}; -#define TTM_BO_PRIV_FLAG_MOVING 0 /* Buffer object is moving and needs - idling before CPU mapping */ -#define TTM_BO_PRIV_FLAG_MAX 1 /** - * struct ttm_bo_device - Buffer object driver device-specific data. + * struct ttm_bo_global - Buffer object driver global data. * * @mem_glob: Pointer to a struct ttm_mem_global object for accounting. - * @driver: Pointer to a struct ttm_bo_driver struct setup by the driver. - * @count: Current number of buffer object. - * @pages: Current number of pinned pages. * @dummy_read_page: Pointer to a dummy page used for mapping requests * of unpopulated pages. - * @shrink: A shrink callback object used for buffre object swap. + * @shrink: A shrink callback object used for buffer object swap. * @ttm_bo_extra_size: Extra size (sizeof(struct ttm_buffer_object) excluded) * used by a buffer object. This is excluding page arrays and backing pages. * @ttm_bo_size: This is @ttm_bo_extra_size + sizeof(struct ttm_buffer_object). + * @device_list_mutex: Mutex protecting the device list. + * This mutex is held while traversing the device list for pm options. + * @lru_lock: Spinlock protecting the bo subsystem lru lists. + * @device_list: List of buffer object devices. + * @swap_lru: Lru list of buffer objects used for swapping. + */ + +struct ttm_bo_global { + + /** + * Constant after init. + */ + + struct kobject kobj; + struct ttm_mem_global *mem_glob; + struct page *dummy_read_page; + struct ttm_mem_shrink shrink; + size_t ttm_bo_extra_size; + size_t ttm_bo_size; + struct mutex device_list_mutex; + spinlock_t lru_lock; + + /** + * Protected by device_list_mutex. + */ + struct list_head device_list; + + /** + * Protected by the lru_lock. + */ + struct list_head swap_lru; + + /** + * Internal protection. + */ + atomic_t bo_count; +}; + + +#define TTM_NUM_MEM_TYPES 8 + +#define TTM_BO_PRIV_FLAG_MOVING 0 /* Buffer object is moving and needs + idling before CPU mapping */ +#define TTM_BO_PRIV_FLAG_MAX 1 +/** + * struct ttm_bo_device - Buffer object driver device-specific data. + * + * @driver: Pointer to a struct ttm_bo_driver struct setup by the driver. * @man: An array of mem_type_managers. * @addr_space_mm: Range manager for the device address space. * lru_lock: Spinlock that protects the buffer+device lru lists and @@ -390,32 +440,21 @@ struct ttm_bo_device { /* * Constant after bo device init / atomic. */ - - struct ttm_mem_global *mem_glob; + struct list_head device_list; + struct ttm_bo_global *glob; struct ttm_bo_driver *driver; - struct page *dummy_read_page; - struct ttm_mem_shrink shrink; - - size_t ttm_bo_extra_size; - size_t ttm_bo_size; - rwlock_t vm_lock; + struct ttm_mem_type_manager man[TTM_NUM_MEM_TYPES]; /* * Protected by the vm lock. */ - struct ttm_mem_type_manager man[TTM_NUM_MEM_TYPES]; struct rb_root addr_space_rb; struct drm_mm addr_space_mm; /* - * Might want to change this to one lock per manager. - */ - spinlock_t lru_lock; - /* - * Protected by the lru lock. + * Protected by the global:lru lock. */ struct list_head ddestroy; - struct list_head swap_lru; /* * Protected by load / firstopen / lastclose /unload sync. @@ -629,6 +668,9 @@ extern int ttm_bo_pci_offset(struct ttm_bo_device *bdev, unsigned long *bus_offset, unsigned long *bus_size); +extern void ttm_bo_global_release(struct ttm_global_reference *ref); +extern int ttm_bo_global_init(struct ttm_global_reference *ref); + extern int ttm_bo_device_release(struct ttm_bo_device *bdev); /** @@ -646,7 +688,7 @@ extern int ttm_bo_device_release(struct ttm_bo_device *bdev); * !0: Failure. */ extern int ttm_bo_device_init(struct ttm_bo_device *bdev, - struct ttm_mem_global *mem_glob, + struct ttm_bo_global *glob, struct ttm_bo_driver *driver, uint64_t file_page_offset); -- cgit v1.2.3 From 10a5b66f625904ad5a2867cf7a28073e1236ff32 Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Wed, 19 Aug 2009 15:53:05 +0800 Subject: tracing/syscalls: Add fields format for exit events Add "format" file for syscall exit events: # cat events/syscalls/sys_exit_open/format name: sys_exit_open ID: 344 format: field:unsigned short common_type; offset:0; size:2; field:unsigned char common_flags; offset:2; size:1; field:unsigned char common_preempt_count; offset:3; size:1; field:int common_pid; offset:4; size:4; field:int common_tgid; offset:8; size:4; field:int nr; offset:12; size:4; field:unsigned long ret; offset:16; size:4; Signed-off-by: Li Zefan Cc: Jason Baron Cc: Steven Rostedt Cc: Frederic Weisbecker LKML-Reference: <4A8BAF61.3060307@cn.fujitsu.com> Signed-off-by: Ingo Molnar --- include/linux/syscalls.h | 3 ++- include/trace/syscall.h | 6 ++++-- kernel/trace/trace_syscalls.c | 18 +++++++++++++++++- 3 files changed, 23 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/include/linux/syscalls.h b/include/linux/syscalls.h index 87d06c173ddc..8d57f77794ee 100644 --- a/include/linux/syscalls.h +++ b/include/linux/syscalls.h @@ -189,7 +189,7 @@ static void prof_sysexit_disable_##sname(struct ftrace_event_call *event_call) \ .system = "syscalls", \ .event = &event_syscall_enter, \ .raw_init = init_enter_##sname, \ - .show_format = ftrace_format_syscall, \ + .show_format = syscall_enter_format, \ .regfunc = reg_event_syscall_enter, \ .unregfunc = unreg_event_syscall_enter, \ .data = "sys"#sname, \ @@ -225,6 +225,7 @@ static void prof_sysexit_disable_##sname(struct ftrace_event_call *event_call) \ .system = "syscalls", \ .event = &event_syscall_exit, \ .raw_init = init_exit_##sname, \ + .show_format = syscall_exit_format, \ .regfunc = reg_event_syscall_exit, \ .unregfunc = unreg_event_syscall_exit, \ .data = "sys"#sname, \ diff --git a/include/trace/syscall.h b/include/trace/syscall.h index 0cb03625edd9..5ce85d75d31b 100644 --- a/include/trace/syscall.h +++ b/include/trace/syscall.h @@ -55,8 +55,10 @@ extern int reg_event_syscall_enter(void *ptr); extern void unreg_event_syscall_enter(void *ptr); extern int reg_event_syscall_exit(void *ptr); extern void unreg_event_syscall_exit(void *ptr); -extern int -ftrace_format_syscall(struct ftrace_event_call *call, struct trace_seq *s); +extern int syscall_enter_format(struct ftrace_event_call *call, + struct trace_seq *s); +extern int syscall_exit_format(struct ftrace_event_call *call, + struct trace_seq *s); enum print_line_t print_syscall_enter(struct trace_iterator *iter, int flags); enum print_line_t print_syscall_exit(struct trace_iterator *iter, int flags); #endif diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c index d10daf0922b5..7336b6c265d7 100644 --- a/kernel/trace/trace_syscalls.c +++ b/kernel/trace/trace_syscalls.c @@ -97,7 +97,7 @@ extern char *__bad_type_size(void); __bad_type_size() : \ #type, #name, offsetof(typeof(trace), name), sizeof(trace.name) -int ftrace_format_syscall(struct ftrace_event_call *call, struct trace_seq *s) +int syscall_enter_format(struct ftrace_event_call *call, struct trace_seq *s) { int i; int nr; @@ -149,6 +149,22 @@ int ftrace_format_syscall(struct ftrace_event_call *call, struct trace_seq *s) return ret; } +int syscall_exit_format(struct ftrace_event_call *call, struct trace_seq *s) +{ + int ret; + struct syscall_trace_exit trace; + + ret = trace_seq_printf(s, + "\tfield:%s %s;\toffset:%zu;\tsize:%zu;\n" + "\tfield:%s %s;\toffset:%zu;\tsize:%zu;\n", + SYSCALL_FIELD(int, nr), + SYSCALL_FIELD(unsigned long, ret)); + if (!ret) + return 0; + + return trace_seq_printf(s, "\nprint fmt: \"0x%%lx\", REC->ret\n"); +} + void ftrace_syscall_enter(struct pt_regs *regs, long id) { struct syscall_trace_enter *entry; -- cgit v1.2.3 From 14be96c9716cb8c46dca94bd890defd7856e0734 Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Wed, 19 Aug 2009 15:53:52 +0800 Subject: tracing/events: Add ftrace_event_call param to define_fields() This parameter is needed by syscall events to add define_fields() handler. Signed-off-by: Li Zefan Cc: Jason Baron Cc: Steven Rostedt Cc: Frederic Weisbecker LKML-Reference: <4A8BAF90.6060801@cn.fujitsu.com> Signed-off-by: Ingo Molnar --- include/linux/ftrace_event.h | 2 +- include/trace/ftrace.h | 3 +-- kernel/trace/trace_events.c | 2 +- kernel/trace/trace_export.c | 5 ++--- 4 files changed, 5 insertions(+), 7 deletions(-) (limited to 'include') diff --git a/include/linux/ftrace_event.h b/include/linux/ftrace_event.h index 189806b6e69e..35b3a4a5ba86 100644 --- a/include/linux/ftrace_event.h +++ b/include/linux/ftrace_event.h @@ -118,7 +118,7 @@ struct ftrace_event_call { int (*raw_init)(void); int (*show_format)(struct ftrace_event_call *call, struct trace_seq *s); - int (*define_fields)(void); + int (*define_fields)(struct ftrace_event_call *); struct list_head fields; int filter_active; struct event_filter *filter; diff --git a/include/trace/ftrace.h b/include/trace/ftrace.h index b250b0616571..4e81c9b37515 100644 --- a/include/trace/ftrace.h +++ b/include/trace/ftrace.h @@ -294,10 +294,9 @@ ftrace_raw_output_##call(struct trace_iterator *iter, int flags) \ #undef TRACE_EVENT #define TRACE_EVENT(call, proto, args, tstruct, func, print) \ int \ -ftrace_define_fields_##call(void) \ +ftrace_define_fields_##call(struct ftrace_event_call *event_call) \ { \ struct ftrace_raw_##call field; \ - struct ftrace_event_call *event_call = &event_##call; \ int ret; \ \ __common_field(int, type, 1); \ diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c index b568ade8f453..af8fb8ebef0b 100644 --- a/kernel/trace/trace_events.c +++ b/kernel/trace/trace_events.c @@ -941,7 +941,7 @@ event_create_dir(struct ftrace_event_call *call, struct dentry *d_events, id); if (call->define_fields) { - ret = call->define_fields(); + ret = call->define_fields(call); if (ret < 0) { pr_warning("Could not initialize trace point" " events/%s\n", call->name); diff --git a/kernel/trace/trace_export.c b/kernel/trace/trace_export.c index 956d4bc675e5..cf2c752a25bf 100644 --- a/kernel/trace/trace_export.c +++ b/kernel/trace/trace_export.c @@ -119,7 +119,7 @@ ftrace_format_##call(struct ftrace_event_call *unused, \ #undef TRACE_EVENT_FORMAT #define TRACE_EVENT_FORMAT(call, proto, args, fmt, tstruct, tpfmt) \ -int ftrace_define_fields_##call(void); \ +int ftrace_define_fields_##call(struct ftrace_event_call *event_call); \ static int ftrace_raw_init_event_##call(void); \ \ struct ftrace_event_call __used \ @@ -184,9 +184,8 @@ __attribute__((section("_ftrace_events"))) event_##call = { \ #undef TRACE_EVENT_FORMAT #define TRACE_EVENT_FORMAT(call, proto, args, fmt, tstruct, tpfmt) \ int \ -ftrace_define_fields_##call(void) \ +ftrace_define_fields_##call(struct ftrace_event_call *event_call) \ { \ - struct ftrace_event_call *event_call = &event_##call; \ struct args field; \ int ret; \ \ -- cgit v1.2.3 From e647d6b314266adb904d4b84973eda0afa856946 Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Wed, 19 Aug 2009 15:54:32 +0800 Subject: tracing/events: Add trace_define_common_fields() Extract duplicate code. Also prepare for the later patch. Signed-off-by: Li Zefan Cc: Jason Baron Cc: Steven Rostedt Cc: Frederic Weisbecker LKML-Reference: <4A8BAFB8.1010304@cn.fujitsu.com> Signed-off-by: Ingo Molnar --- include/linux/ftrace_event.h | 8 +------- include/trace/ftrace.h | 8 +++----- kernel/trace/trace_events.c | 22 ++++++++++++++++++++++ kernel/trace/trace_export.c | 8 +++----- 4 files changed, 29 insertions(+), 17 deletions(-) (limited to 'include') diff --git a/include/linux/ftrace_event.h b/include/linux/ftrace_event.h index 35b3a4a5ba86..427cbae47f84 100644 --- a/include/linux/ftrace_event.h +++ b/include/linux/ftrace_event.h @@ -142,6 +142,7 @@ extern int filter_current_check_discard(struct ftrace_event_call *call, extern int trace_define_field(struct ftrace_event_call *call, char *type, char *name, int offset, int size, int is_signed); +extern int trace_define_common_fields(struct ftrace_event_call *call); #define is_signed_type(type) (((type)(-1)) < 0) @@ -166,11 +167,4 @@ do { \ __trace_printk(ip, fmt, ##args); \ } while (0) -#define __common_field(type, item, is_signed) \ - ret = trace_define_field(event_call, #type, "common_" #item, \ - offsetof(typeof(field.ent), item), \ - sizeof(field.ent.item), is_signed); \ - if (ret) \ - return ret; - #endif /* _LINUX_FTRACE_EVENT_H */ diff --git a/include/trace/ftrace.h b/include/trace/ftrace.h index 4e81c9b37515..127400255e4c 100644 --- a/include/trace/ftrace.h +++ b/include/trace/ftrace.h @@ -299,11 +299,9 @@ ftrace_define_fields_##call(struct ftrace_event_call *event_call) \ struct ftrace_raw_##call field; \ int ret; \ \ - __common_field(int, type, 1); \ - __common_field(unsigned char, flags, 0); \ - __common_field(unsigned char, preempt_count, 0); \ - __common_field(int, pid, 1); \ - __common_field(int, tgid, 1); \ + ret = trace_define_common_fields(event_call); \ + if (ret) \ + return ret; \ \ tstruct; \ \ diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c index af8fb8ebef0b..9c7ecfb3416f 100644 --- a/kernel/trace/trace_events.c +++ b/kernel/trace/trace_events.c @@ -62,6 +62,28 @@ err: } EXPORT_SYMBOL_GPL(trace_define_field); +#define __common_field(type, item) \ + ret = trace_define_field(call, #type, "common_" #item, \ + offsetof(typeof(ent), item), \ + sizeof(ent.item), \ + is_signed_type(type)); \ + if (ret) \ + return ret; + +int trace_define_common_fields(struct ftrace_event_call *call) +{ + int ret; + struct trace_entry ent; + + __common_field(unsigned short, type); + __common_field(unsigned char, flags); + __common_field(unsigned char, preempt_count); + __common_field(int, pid); + __common_field(int, tgid); + + return ret; +} + #ifdef CONFIG_MODULES static void trace_destroy_fields(struct ftrace_event_call *call) diff --git a/kernel/trace/trace_export.c b/kernel/trace/trace_export.c index cf2c752a25bf..70875303ae46 100644 --- a/kernel/trace/trace_export.c +++ b/kernel/trace/trace_export.c @@ -189,11 +189,9 @@ ftrace_define_fields_##call(struct ftrace_event_call *event_call) \ struct args field; \ int ret; \ \ - __common_field(unsigned char, type, 0); \ - __common_field(unsigned char, flags, 0); \ - __common_field(unsigned char, preempt_count, 0); \ - __common_field(int, pid, 1); \ - __common_field(int, tgid, 1); \ + ret = trace_define_common_fields(event_call); \ + if (ret) \ + return ret; \ \ tstruct; \ \ -- cgit v1.2.3 From 540b7b8d65575c80162f2a0f38e1d313c92a6042 Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Wed, 19 Aug 2009 15:54:51 +0800 Subject: tracing/syscalls: Add filtering support Add filtering support for syscall events: # echo 'mode == 0666' > events/syscalls/sys_enter_open # echo 'ret == 0' > events/syscalls/sys_exit_open # echo 1 > events/syscalls/sys_enter_open # echo 1 > events/syscalls/sys_exit_open # cat trace ... modprobe-3084 [001] 117.463140: sys_open(filename: 917d3e8, flags: 0, mode: 1b6) modprobe-3084 [001] 117.463176: sys_open -> 0x0 less-3086 [001] 117.510455: sys_open(filename: 9c6bdb8, flags: 8000, mode: 1b6) sendmail-2574 [001] 122.145840: sys_open(filename: b807a365, flags: 0, mode: 1b6) ... Signed-off-by: Li Zefan Cc: Jason Baron Cc: Steven Rostedt Cc: Frederic Weisbecker LKML-Reference: <4A8BAFCB.1040006@cn.fujitsu.com> Signed-off-by: Ingo Molnar --- include/linux/ftrace_event.h | 5 +++-- include/linux/syscalls.h | 16 +++++++++----- include/trace/syscall.h | 7 ++++++ kernel/trace/trace_events.c | 5 +++-- kernel/trace/trace_syscalls.c | 51 +++++++++++++++++++++++++++++++++++++++---- 5 files changed, 71 insertions(+), 13 deletions(-) (limited to 'include') diff --git a/include/linux/ftrace_event.h b/include/linux/ftrace_event.h index 427cbae47f84..df5b085c4150 100644 --- a/include/linux/ftrace_event.h +++ b/include/linux/ftrace_event.h @@ -140,8 +140,9 @@ extern int filter_current_check_discard(struct ftrace_event_call *call, void *rec, struct ring_buffer_event *event); -extern int trace_define_field(struct ftrace_event_call *call, char *type, - char *name, int offset, int size, int is_signed); +extern int trace_define_field(struct ftrace_event_call *call, + const char *type, const char *name, + int offset, int size, int is_signed); extern int trace_define_common_fields(struct ftrace_event_call *call); #define is_signed_type(type) (((type)(-1)) < 0) diff --git a/include/linux/syscalls.h b/include/linux/syscalls.h index 8d57f77794ee..f124c8995555 100644 --- a/include/linux/syscalls.h +++ b/include/linux/syscalls.h @@ -190,6 +190,7 @@ static void prof_sysexit_disable_##sname(struct ftrace_event_call *event_call) \ .event = &event_syscall_enter, \ .raw_init = init_enter_##sname, \ .show_format = syscall_enter_format, \ + .define_fields = syscall_enter_define_fields, \ .regfunc = reg_event_syscall_enter, \ .unregfunc = unreg_event_syscall_enter, \ .data = "sys"#sname, \ @@ -226,6 +227,7 @@ static void prof_sysexit_disable_##sname(struct ftrace_event_call *event_call) \ .event = &event_syscall_exit, \ .raw_init = init_exit_##sname, \ .show_format = syscall_exit_format, \ + .define_fields = syscall_exit_define_fields, \ .regfunc = reg_event_syscall_exit, \ .unregfunc = unreg_event_syscall_exit, \ .data = "sys"#sname, \ @@ -233,6 +235,8 @@ static void prof_sysexit_disable_##sname(struct ftrace_event_call *event_call) \ } #define SYSCALL_METADATA(sname, nb) \ + SYSCALL_TRACE_ENTER_EVENT(sname); \ + SYSCALL_TRACE_EXIT_EVENT(sname); \ static const struct syscall_metadata __used \ __attribute__((__aligned__(4))) \ __attribute__((section("__syscalls_metadata"))) \ @@ -241,20 +245,22 @@ static void prof_sysexit_disable_##sname(struct ftrace_event_call *event_call) \ .nb_args = nb, \ .types = types_##sname, \ .args = args_##sname, \ - }; \ - SYSCALL_TRACE_ENTER_EVENT(sname); \ - SYSCALL_TRACE_EXIT_EVENT(sname); + .enter_event = &event_enter_##sname, \ + .exit_event = &event_exit_##sname, \ + }; #define SYSCALL_DEFINE0(sname) \ + SYSCALL_TRACE_ENTER_EVENT(_##sname); \ + SYSCALL_TRACE_EXIT_EVENT(_##sname); \ static const struct syscall_metadata __used \ __attribute__((__aligned__(4))) \ __attribute__((section("__syscalls_metadata"))) \ __syscall_meta_##sname = { \ .name = "sys_"#sname, \ .nb_args = 0, \ + .enter_event = &event_enter__##sname, \ + .exit_event = &event_exit__##sname, \ }; \ - SYSCALL_TRACE_ENTER_EVENT(_##sname); \ - SYSCALL_TRACE_EXIT_EVENT(_##sname); \ asmlinkage long sys_##sname(void) #else #define SYSCALL_DEFINE0(name) asmlinkage long sys_##name(void) diff --git a/include/trace/syscall.h b/include/trace/syscall.h index 5ce85d75d31b..9661dd406b93 100644 --- a/include/trace/syscall.h +++ b/include/trace/syscall.h @@ -34,6 +34,8 @@ DECLARE_TRACE_WITH_CALLBACK(syscall_exit, * @args: list of args as strings (args[i] matches types[i]) * @enter_id: associated ftrace enter event id * @exit_id: associated ftrace exit event id + * @enter_event: associated syscall_enter trace event + * @exit_event: associated syscall_exit trace event */ struct syscall_metadata { const char *name; @@ -42,6 +44,9 @@ struct syscall_metadata { const char **args; int enter_id; int exit_id; + + struct ftrace_event_call *enter_event; + struct ftrace_event_call *exit_event; }; #ifdef CONFIG_FTRACE_SYSCALLS @@ -59,6 +64,8 @@ extern int syscall_enter_format(struct ftrace_event_call *call, struct trace_seq *s); extern int syscall_exit_format(struct ftrace_event_call *call, struct trace_seq *s); +extern int syscall_enter_define_fields(struct ftrace_event_call *call); +extern int syscall_exit_define_fields(struct ftrace_event_call *call); enum print_line_t print_syscall_enter(struct trace_iterator *iter, int flags); enum print_line_t print_syscall_exit(struct trace_iterator *iter, int flags); #endif diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c index 9c7ecfb3416f..79d352027a61 100644 --- a/kernel/trace/trace_events.c +++ b/kernel/trace/trace_events.c @@ -27,8 +27,8 @@ DEFINE_MUTEX(event_mutex); LIST_HEAD(ftrace_events); -int trace_define_field(struct ftrace_event_call *call, char *type, - char *name, int offset, int size, int is_signed) +int trace_define_field(struct ftrace_event_call *call, const char *type, + const char *name, int offset, int size, int is_signed) { struct ftrace_event_field *field; @@ -83,6 +83,7 @@ int trace_define_common_fields(struct ftrace_event_call *call) return ret; } +EXPORT_SYMBOL_GPL(trace_define_common_fields); #ifdef CONFIG_MODULES diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c index 7336b6c265d7..28e4dae4af21 100644 --- a/kernel/trace/trace_syscalls.c +++ b/kernel/trace/trace_syscalls.c @@ -165,6 +165,49 @@ int syscall_exit_format(struct ftrace_event_call *call, struct trace_seq *s) return trace_seq_printf(s, "\nprint fmt: \"0x%%lx\", REC->ret\n"); } +int syscall_enter_define_fields(struct ftrace_event_call *call) +{ + struct syscall_trace_enter trace; + struct syscall_metadata *meta; + int ret; + int nr; + int i; + int offset = offsetof(typeof(trace), args); + + nr = syscall_name_to_nr(call->data); + meta = syscall_nr_to_meta(nr); + + if (!meta) + return 0; + + ret = trace_define_common_fields(call); + if (ret) + return ret; + + for (i = 0; i < meta->nb_args; i++) { + ret = trace_define_field(call, meta->types[i], + meta->args[i], offset, + sizeof(unsigned long), 0); + offset += sizeof(unsigned long); + } + + return ret; +} + +int syscall_exit_define_fields(struct ftrace_event_call *call) +{ + struct syscall_trace_exit trace; + int ret; + + ret = trace_define_common_fields(call); + if (ret) + return ret; + + ret = trace_define_field(call, SYSCALL_FIELD(unsigned long, ret), 0); + + return ret; +} + void ftrace_syscall_enter(struct pt_regs *regs, long id) { struct syscall_trace_enter *entry; @@ -192,8 +235,8 @@ void ftrace_syscall_enter(struct pt_regs *regs, long id) entry->nr = syscall_nr; syscall_get_arguments(current, regs, 0, sys_data->nb_args, entry->args); - trace_current_buffer_unlock_commit(event, 0, 0); - trace_wake_up(); + if (!filter_current_check_discard(sys_data->enter_event, entry, event)) + trace_current_buffer_unlock_commit(event, 0, 0); } void ftrace_syscall_exit(struct pt_regs *regs, long ret) @@ -220,8 +263,8 @@ void ftrace_syscall_exit(struct pt_regs *regs, long ret) entry->nr = syscall_nr; entry->ret = syscall_get_return_value(current, regs); - trace_current_buffer_unlock_commit(event, 0, 0); - trace_wake_up(); + if (!filter_current_check_discard(sys_data->exit_event, entry, event)) + trace_current_buffer_unlock_commit(event, 0, 0); } int reg_event_syscall_enter(void *ptr) -- cgit v1.2.3 From 2bfb1070ba1fdb8cbc2b0b9ff61a3b0701ab40de Mon Sep 17 00:00:00 2001 From: Dmitry Eremin-Solenikov Date: Fri, 14 Aug 2009 16:13:12 +0400 Subject: ieee802154: add a sysfs representation of WPAN master devices Add a sysfs/in-kernel representation of LR-WPAN master devices. Signed-off-by: Dmitry Eremin-Solenikov --- include/net/wpan-phy.h | 63 ++++++++++++++++++ net/ieee802154/Makefile | 2 +- net/ieee802154/wpan-class.c | 159 ++++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 223 insertions(+), 1 deletion(-) create mode 100644 include/net/wpan-phy.h create mode 100644 net/ieee802154/wpan-class.c (limited to 'include') diff --git a/include/net/wpan-phy.h b/include/net/wpan-phy.h new file mode 100644 index 000000000000..547b1e271ac9 --- /dev/null +++ b/include/net/wpan-phy.h @@ -0,0 +1,63 @@ +/* + * Copyright (C) 2007, 2008, 2009 Siemens AG + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA. + * + * Written by: + * Dmitry Eremin-Solenikov + */ + +#ifndef WPAN_PHY_H +#define WPAN_PHY_H + +#include +#include + +struct wpan_phy { + struct mutex pib_lock; + + /* + * This is a PIB acording to 802.15.4-2006. + * We do not provide timing-related variables, as they + * aren't used outside of driver + */ + u8 current_channel; + u8 current_page; + u32 channels_supported; + u8 transmit_power; + u8 cca_mode; + + struct device dev; + int idx; + + char priv[0] __attribute__((__aligned__(NETDEV_ALIGN))); +}; + +struct wpan_phy *wpan_phy_alloc(size_t priv_size); +int wpan_phy_register(struct device *parent, struct wpan_phy *phy); +void wpan_phy_unregister(struct wpan_phy *phy); +void wpan_phy_free(struct wpan_phy *phy); + +static inline void *wpan_phy_priv(struct wpan_phy *phy) +{ + BUG_ON(!phy); + return &phy->priv; +} + +struct wpan_phy *wpan_phy_find(const char *str); +static inline const char *wpan_phy_name(struct wpan_phy *phy) +{ + return dev_name(&phy->dev); +} +#endif diff --git a/net/ieee802154/Makefile b/net/ieee802154/Makefile index f99338a26100..4068a9f5113e 100644 --- a/net/ieee802154/Makefile +++ b/net/ieee802154/Makefile @@ -1,4 +1,4 @@ -obj-$(CONFIG_IEEE802154) += nl802154.o af_802154.o +obj-$(CONFIG_IEEE802154) += nl802154.o af_802154.o wpan-class.o nl802154-y := netlink.o nl_policy.o af_802154-y := af_ieee802154.o raw.o dgram.o diff --git a/net/ieee802154/wpan-class.c b/net/ieee802154/wpan-class.c new file mode 100644 index 000000000000..f306604da67a --- /dev/null +++ b/net/ieee802154/wpan-class.c @@ -0,0 +1,159 @@ +/* + * Copyright (C) 2007, 2008, 2009 Siemens AG + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA. + * + */ + +#include +#include +#include + +#include + +#define MASTER_SHOW_COMPLEX(name, format_string, args...) \ +static ssize_t name ## _show(struct device *dev, \ + struct device_attribute *attr, char *buf) \ +{ \ + struct wpan_phy *phy = container_of(dev, struct wpan_phy, dev); \ + int ret; \ + \ + mutex_lock(&phy->pib_lock); \ + ret = sprintf(buf, format_string "\n", args); \ + mutex_unlock(&phy->pib_lock); \ + return ret; \ +} + +#define MASTER_SHOW(field, format_string) \ + MASTER_SHOW_COMPLEX(field, format_string, phy->field) + +MASTER_SHOW(current_channel, "%d"); +MASTER_SHOW(current_page, "%d"); +MASTER_SHOW(channels_supported, "%#x"); +MASTER_SHOW_COMPLEX(transmit_power, "%d +- %d dB", + ((signed char) (phy->transmit_power << 2)) >> 2, + (phy->transmit_power >> 6) ? (phy->transmit_power >> 6) * 3 : 1 ); +MASTER_SHOW(cca_mode, "%d"); + +static struct device_attribute pmib_attrs[] = { + __ATTR_RO(current_channel), + __ATTR_RO(current_page), + __ATTR_RO(channels_supported), + __ATTR_RO(transmit_power), + __ATTR_RO(cca_mode), + {}, +}; + +static void wpan_phy_release(struct device *d) +{ + struct wpan_phy *phy = container_of(d, struct wpan_phy, dev); + kfree(phy); +} + +static struct class wpan_phy_class = { + .name = "ieee802154", + .dev_release = wpan_phy_release, + .dev_attrs = pmib_attrs, +}; + +static DEFINE_MUTEX(wpan_phy_mutex); +static int wpan_phy_idx; + +static int wpan_phy_match(struct device *dev, void *data) +{ + return !strcmp(dev_name(dev), (const char *)data); +} + +struct wpan_phy *wpan_phy_find(const char *str) +{ + struct device *dev; + + if (WARN_ON(!str)) + return NULL; + + dev = class_find_device(&wpan_phy_class, NULL, + (void *)str, wpan_phy_match); + if (!dev) + return NULL; + + return container_of(dev, struct wpan_phy, dev); +} +EXPORT_SYMBOL(wpan_phy_find); + +static int wpan_phy_idx_valid(int idx) +{ + return idx >= 0; +} + +struct wpan_phy *wpan_phy_alloc(size_t priv_size) +{ + struct wpan_phy *phy = kzalloc(sizeof(*phy) + priv_size, + GFP_KERNEL); + + mutex_lock(&wpan_phy_mutex); + phy->idx = wpan_phy_idx++; + if (unlikely(!wpan_phy_idx_valid(phy->idx))) { + wpan_phy_idx--; + mutex_unlock(&wpan_phy_mutex); + kfree(phy); + return NULL; + } + mutex_unlock(&wpan_phy_mutex); + + mutex_init(&phy->pib_lock); + + device_initialize(&phy->dev); + dev_set_name(&phy->dev, "wpan-phy%d", phy->idx); + + phy->dev.class = &wpan_phy_class; + + return phy; +} +EXPORT_SYMBOL(wpan_phy_alloc); + +int wpan_phy_register(struct device *parent, struct wpan_phy *phy) +{ + phy->dev.parent = parent; + + return device_add(&phy->dev); +} +EXPORT_SYMBOL(wpan_phy_register); + +void wpan_phy_unregister(struct wpan_phy *phy) +{ + device_del(&phy->dev); +} +EXPORT_SYMBOL(wpan_phy_unregister); + +void wpan_phy_free(struct wpan_phy *phy) +{ + put_device(&phy->dev); +} +EXPORT_SYMBOL(wpan_phy_free); + +static int __init wpan_phy_class_init(void) +{ + return class_register(&wpan_phy_class); +} +subsys_initcall(wpan_phy_class_init); + +static void __exit wpan_phy_class_exit(void) +{ + class_unregister(&wpan_phy_class); +} +module_exit(wpan_phy_class_exit); + +MODULE_DESCRIPTION("IEEE 802.15.4 device class"); +MODULE_LICENSE("GPL v2"); + -- cgit v1.2.3 From 16eea493da563b5a3356a77c6d8776dffc29d3b6 Mon Sep 17 00:00:00 2001 From: Dmitry Eremin-Solenikov Date: Wed, 19 Aug 2009 19:32:24 +0400 Subject: ieee802154: add support for channel pages from IEEE 802.15.4-2006 IEEE 802.15.4-2006 adds new concept: channel pages, which can contain several channels. Add support for channel pages in the API and in the fakehard driver. Signed-off-by: Dmitry Eremin-Solenikov --- drivers/ieee802154/fakehard.c | 10 ++++++---- include/linux/nl802154.h | 2 ++ include/net/ieee802154_netdev.h | 6 +++--- include/net/nl802154.h | 2 +- net/ieee802154/netlink.c | 28 +++++++++++++++++++++++++--- net/ieee802154/nl_policy.c | 1 + 6 files changed, 38 insertions(+), 11 deletions(-) (limited to 'include') diff --git a/drivers/ieee802154/fakehard.c b/drivers/ieee802154/fakehard.c index 22a93bc764c5..c1c9697f9fde 100644 --- a/drivers/ieee802154/fakehard.c +++ b/drivers/ieee802154/fakehard.c @@ -119,12 +119,13 @@ static u8 fake_get_bsn(struct net_device *dev) * 802.15.4-2006 document. */ static int fake_assoc_req(struct net_device *dev, - struct ieee802154_addr *addr, u8 channel, u8 cap) + struct ieee802154_addr *addr, u8 channel, u8 page, u8 cap) { struct wpan_phy *phy = net_to_phy(dev); mutex_lock(&phy->pib_lock); phy->current_channel = channel; + phy->current_page = page; mutex_unlock(&phy->pib_lock); /* We simply emulate it here */ @@ -191,7 +192,7 @@ static int fake_disassoc_req(struct net_device *dev, * document, with 7.3.8 describing coordinator realignment. */ static int fake_start_req(struct net_device *dev, struct ieee802154_addr *addr, - u8 channel, + u8 channel, u8 page, u8 bcn_ord, u8 sf_ord, u8 pan_coord, u8 blx, u8 coord_realign) { @@ -199,6 +200,7 @@ static int fake_start_req(struct net_device *dev, struct ieee802154_addr *addr, mutex_lock(&phy->pib_lock); phy->current_channel = channel; + phy->current_page = page; mutex_unlock(&phy->pib_lock); /* We don't emulate beacons here at all, so START should fail */ @@ -222,11 +224,11 @@ static int fake_start_req(struct net_device *dev, struct ieee802154_addr *addr, * Note: This is in section 7.5.2.1 of the IEEE 802.15.4-2006 document. */ static int fake_scan_req(struct net_device *dev, u8 type, u32 channels, - u8 duration) + u8 page, u8 duration) { u8 edl[27] = {}; return ieee802154_nl_scan_confirm(dev, IEEE802154_SUCCESS, type, - channels, + channels, page, type == IEEE802154_MAC_SCAN_ED ? edl : NULL); } diff --git a/include/linux/nl802154.h b/include/linux/nl802154.h index 9a1af5f871a3..b7d9435d5a9f 100644 --- a/include/linux/nl802154.h +++ b/include/linux/nl802154.h @@ -64,6 +64,8 @@ enum { IEEE802154_ATTR_COORD_REALIGN, IEEE802154_ATTR_SEC, + IEEE802154_ATTR_PAGE, + __IEEE802154_ATTR_MAX, }; diff --git a/include/net/ieee802154_netdev.h b/include/net/ieee802154_netdev.h index e2506af3e7c8..5dc6a61952de 100644 --- a/include/net/ieee802154_netdev.h +++ b/include/net/ieee802154_netdev.h @@ -80,7 +80,7 @@ static inline int mac_cb_type(struct sk_buff *skb) struct ieee802154_mlme_ops { int (*assoc_req)(struct net_device *dev, struct ieee802154_addr *addr, - u8 channel, u8 cap); + u8 channel, u8 page, u8 cap); int (*assoc_resp)(struct net_device *dev, struct ieee802154_addr *addr, u16 short_addr, u8 status); @@ -89,10 +89,10 @@ struct ieee802154_mlme_ops { u8 reason); int (*start_req)(struct net_device *dev, struct ieee802154_addr *addr, - u8 channel, u8 bcn_ord, u8 sf_ord, + u8 channel, u8 page, u8 bcn_ord, u8 sf_ord, u8 pan_coord, u8 blx, u8 coord_realign); int (*scan_req)(struct net_device *dev, - u8 type, u32 channels, u8 duration); + u8 type, u32 channels, u8 page, u8 duration); /* * FIXME: these should become the part of PIB/MIB interface. diff --git a/include/net/nl802154.h b/include/net/nl802154.h index e554ecd3727a..99d2ba1c7e03 100644 --- a/include/net/nl802154.h +++ b/include/net/nl802154.h @@ -95,7 +95,7 @@ int ieee802154_nl_disassoc_confirm(struct net_device *dev, * Note: This API does not permit the return of an active scan result. */ int ieee802154_nl_scan_confirm(struct net_device *dev, - u8 status, u8 scan_type, u32 unscanned, + u8 status, u8 scan_type, u32 unscanned, u8 page, u8 *edl/*, struct list_head *pan_desc_list */); /** diff --git a/net/ieee802154/netlink.c b/net/ieee802154/netlink.c index cd0567f06716..2106ecbf0308 100644 --- a/net/ieee802154/netlink.c +++ b/net/ieee802154/netlink.c @@ -232,7 +232,7 @@ nla_put_failure: EXPORT_SYMBOL(ieee802154_nl_beacon_indic); int ieee802154_nl_scan_confirm(struct net_device *dev, - u8 status, u8 scan_type, u32 unscanned, + u8 status, u8 scan_type, u32 unscanned, u8 page, u8 *edl/* , struct list_head *pan_desc_list */) { struct sk_buff *msg; @@ -251,6 +251,7 @@ int ieee802154_nl_scan_confirm(struct net_device *dev, NLA_PUT_U8(msg, IEEE802154_ATTR_STATUS, status); NLA_PUT_U8(msg, IEEE802154_ATTR_SCAN_TYPE, scan_type); NLA_PUT_U32(msg, IEEE802154_ATTR_CHANNELS, unscanned); + NLA_PUT_U8(msg, IEEE802154_ATTR_PAGE, page); if (edl) NLA_PUT(msg, IEEE802154_ATTR_ED_LIST, 27, edl); @@ -349,6 +350,7 @@ static int ieee802154_associate_req(struct sk_buff *skb, { struct net_device *dev; struct ieee802154_addr addr; + u8 page; int ret = -EINVAL; if (!info->attrs[IEEE802154_ATTR_CHANNEL] || @@ -374,8 +376,14 @@ static int ieee802154_associate_req(struct sk_buff *skb, } addr.pan_id = nla_get_u16(info->attrs[IEEE802154_ATTR_COORD_PAN_ID]); + if (info->attrs[IEEE802154_ATTR_PAGE]) + page = nla_get_u8(info->attrs[IEEE802154_ATTR_PAGE]); + else + page = 0; + ret = ieee802154_mlme_ops(dev)->assoc_req(dev, &addr, nla_get_u8(info->attrs[IEEE802154_ATTR_CHANNEL]), + page, nla_get_u8(info->attrs[IEEE802154_ATTR_CAPABILITY])); dev_put(dev); @@ -458,6 +466,7 @@ static int ieee802154_start_req(struct sk_buff *skb, struct genl_info *info) struct ieee802154_addr addr; u8 channel, bcn_ord, sf_ord; + u8 page; int pan_coord, blx, coord_realign; int ret; @@ -488,13 +497,19 @@ static int ieee802154_start_req(struct sk_buff *skb, struct genl_info *info) blx = nla_get_u8(info->attrs[IEEE802154_ATTR_BAT_EXT]); coord_realign = nla_get_u8(info->attrs[IEEE802154_ATTR_COORD_REALIGN]); + if (info->attrs[IEEE802154_ATTR_PAGE]) + page = nla_get_u8(info->attrs[IEEE802154_ATTR_PAGE]); + else + page = 0; + + if (addr.short_addr == IEEE802154_ADDR_BROADCAST) { ieee802154_nl_start_confirm(dev, IEEE802154_NO_SHORT_ADDRESS); dev_put(dev); return -EINVAL; } - ret = ieee802154_mlme_ops(dev)->start_req(dev, &addr, channel, + ret = ieee802154_mlme_ops(dev)->start_req(dev, &addr, channel, page, bcn_ord, sf_ord, pan_coord, blx, coord_realign); dev_put(dev); @@ -508,6 +523,7 @@ static int ieee802154_scan_req(struct sk_buff *skb, struct genl_info *info) u8 type; u32 channels; u8 duration; + u8 page; if (!info->attrs[IEEE802154_ATTR_SCAN_TYPE] || !info->attrs[IEEE802154_ATTR_CHANNELS] || @@ -522,7 +538,13 @@ static int ieee802154_scan_req(struct sk_buff *skb, struct genl_info *info) channels = nla_get_u32(info->attrs[IEEE802154_ATTR_CHANNELS]); duration = nla_get_u8(info->attrs[IEEE802154_ATTR_DURATION]); - ret = ieee802154_mlme_ops(dev)->scan_req(dev, type, channels, + if (info->attrs[IEEE802154_ATTR_PAGE]) + page = nla_get_u8(info->attrs[IEEE802154_ATTR_PAGE]); + else + page = 0; + + + ret = ieee802154_mlme_ops(dev)->scan_req(dev, type, channels, page, duration); dev_put(dev); diff --git a/net/ieee802154/nl_policy.c b/net/ieee802154/nl_policy.c index 83cb4ccef90d..2363ebee02e7 100644 --- a/net/ieee802154/nl_policy.c +++ b/net/ieee802154/nl_policy.c @@ -33,6 +33,7 @@ const struct nla_policy ieee802154_policy[IEEE802154_ATTR_MAX + 1] = { [IEEE802154_ATTR_HW_ADDR] = { .type = NLA_HW_ADDR, }, [IEEE802154_ATTR_PAN_ID] = { .type = NLA_U16, }, [IEEE802154_ATTR_CHANNEL] = { .type = NLA_U8, }, + [IEEE802154_ATTR_PAGE] = { .type = NLA_U8, }, [IEEE802154_ATTR_COORD_SHORT_ADDR] = { .type = NLA_U16, }, [IEEE802154_ATTR_COORD_HW_ADDR] = { .type = NLA_HW_ADDR, }, [IEEE802154_ATTR_COORD_PAN_ID] = { .type = NLA_U16, }, -- cgit v1.2.3 From 929122cdd5d4c344e59f9b55f870a8fcf7aa0d27 Mon Sep 17 00:00:00 2001 From: Dmitry Eremin-Solenikov Date: Fri, 14 Aug 2009 20:00:20 +0400 Subject: Drop ARPHRD_IEEE802154_PHY There are not maste devices in mac802154 anymore, so drop ARPHRD_IEEE802154_PHY definition. Signed-off-by: Dmitry Eremin-Solenikov --- include/linux/if_arp.h | 1 - net/core/dev.c | 4 ++-- net/ieee802154/af_ieee802154.c | 4 +--- net/ieee802154/raw.c | 3 +-- 4 files changed, 4 insertions(+), 8 deletions(-) (limited to 'include') diff --git a/include/linux/if_arp.h b/include/linux/if_arp.h index b554300ef8bf..282eb37e2dec 100644 --- a/include/linux/if_arp.h +++ b/include/linux/if_arp.h @@ -87,7 +87,6 @@ #define ARPHRD_IEEE80211_PRISM 802 /* IEEE 802.11 + Prism2 header */ #define ARPHRD_IEEE80211_RADIOTAP 803 /* IEEE 802.11 + radiotap header */ #define ARPHRD_IEEE802154 804 -#define ARPHRD_IEEE802154_PHY 805 #define ARPHRD_PHONET 820 /* PhoNet media type */ #define ARPHRD_PHONET_PIPE 821 /* PhoNet pipe header */ diff --git a/net/core/dev.c b/net/core/dev.c index 09fb03fa1ae6..4b837896ebd2 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -269,7 +269,7 @@ static const unsigned short netdev_lock_type[] = ARPHRD_IRDA, ARPHRD_FCPP, ARPHRD_FCAL, ARPHRD_FCPL, ARPHRD_FCFABRIC, ARPHRD_IEEE802_TR, ARPHRD_IEEE80211, ARPHRD_IEEE80211_PRISM, ARPHRD_IEEE80211_RADIOTAP, ARPHRD_PHONET, - ARPHRD_PHONET_PIPE, ARPHRD_IEEE802154, ARPHRD_IEEE802154_PHY, + ARPHRD_PHONET_PIPE, ARPHRD_IEEE802154, ARPHRD_VOID, ARPHRD_NONE}; static const char *const netdev_lock_name[] = @@ -287,7 +287,7 @@ static const char *const netdev_lock_name[] = "_xmit_IRDA", "_xmit_FCPP", "_xmit_FCAL", "_xmit_FCPL", "_xmit_FCFABRIC", "_xmit_IEEE802_TR", "_xmit_IEEE80211", "_xmit_IEEE80211_PRISM", "_xmit_IEEE80211_RADIOTAP", "_xmit_PHONET", - "_xmit_PHONET_PIPE", "_xmit_IEEE802154", "_xmit_IEEE802154_PHY", + "_xmit_PHONET_PIPE", "_xmit_IEEE802154", "_xmit_VOID", "_xmit_NONE"}; static struct lock_class_key netdev_xmit_lock_key[ARRAY_SIZE(netdev_lock_type)]; diff --git a/net/ieee802154/af_ieee802154.c b/net/ieee802154/af_ieee802154.c index d504c349cb0c..cd949d5e451b 100644 --- a/net/ieee802154/af_ieee802154.c +++ b/net/ieee802154/af_ieee802154.c @@ -147,9 +147,7 @@ static int ieee802154_dev_ioctl(struct sock *sk, struct ifreq __user *arg, dev_load(sock_net(sk), ifr.ifr_name); dev = dev_get_by_name(sock_net(sk), ifr.ifr_name); - if ((dev->type == ARPHRD_IEEE802154 || - dev->type == ARPHRD_IEEE802154_PHY) && - dev->netdev_ops->ndo_do_ioctl) + if (dev->type == ARPHRD_IEEE802154 && dev->netdev_ops->ndo_do_ioctl) ret = dev->netdev_ops->ndo_do_ioctl(dev, &ifr, cmd); if (!ret && copy_to_user(arg, &ifr, sizeof(struct ifreq))) diff --git a/net/ieee802154/raw.c b/net/ieee802154/raw.c index 60dee69a1d04..4681501aae93 100644 --- a/net/ieee802154/raw.c +++ b/net/ieee802154/raw.c @@ -74,8 +74,7 @@ static int raw_bind(struct sock *sk, struct sockaddr *uaddr, int len) goto out; } - if (dev->type != ARPHRD_IEEE802154_PHY && - dev->type != ARPHRD_IEEE802154) { + if (dev->type != ARPHRD_IEEE802154) { err = -ENODEV; goto out_put; } -- cgit v1.2.3 From fc69f4a6af49ee69475dc4217924d9edf77760e0 Mon Sep 17 00:00:00 2001 From: Tai-hwa Liang Date: Sun, 10 May 2009 18:15:39 -0700 Subject: Input: add new driver for Sentelic Finger Sensing Pad This is the driver for Sentelic Finger Sensing Pad which can be found on MSI WIND Netbook. Signed-off-by: Tai-hwa Liang Signed-off-by: Dmitry Torokhov --- Documentation/input/sentelic.txt | 475 ++++++++++++++++++++ drivers/input/mouse/Kconfig | 8 + drivers/input/mouse/Makefile | 1 + drivers/input/mouse/psmouse-base.c | 26 +- drivers/input/mouse/psmouse.h | 1 + drivers/input/mouse/sentelic.c | 867 +++++++++++++++++++++++++++++++++++++ drivers/input/mouse/sentelic.h | 98 +++++ drivers/input/serio/libps2.c | 15 +- include/linux/libps2.h | 1 + 9 files changed, 1488 insertions(+), 4 deletions(-) create mode 100644 Documentation/input/sentelic.txt create mode 100644 drivers/input/mouse/sentelic.c create mode 100644 drivers/input/mouse/sentelic.h (limited to 'include') diff --git a/Documentation/input/sentelic.txt b/Documentation/input/sentelic.txt new file mode 100644 index 000000000000..f7160a2fb6a2 --- /dev/null +++ b/Documentation/input/sentelic.txt @@ -0,0 +1,475 @@ +Copyright (C) 2002-2008 Sentelic Corporation. +Last update: Oct-31-2008 + +============================================================================== +* Finger Sensing Pad Intellimouse Mode(scrolling wheel, 4th and 5th buttons) +============================================================================== +A) MSID 4: Scrolling wheel mode plus Forward page(4th button) and Backward + page (5th button) +@1. Set sample rate to 200; +@2. Set sample rate to 200; +@3. Set sample rate to 80; +@4. Issuing the "Get device ID" command (0xF2) and waits for the response; +@5. FSP will respond 0x04. + +Packet 1 + Bit 7 6 5 4 3 2 1 0 7 6 5 4 3 2 1 0 7 6 5 4 3 2 1 0 7 6 5 4 3 2 1 0 +BYTE |---------------|BYTE |---------------|BYTE|---------------|BYTE|---------------| + 1 |Y|X|y|x|1|M|R|L| 2 |X|X|X|X|X|X|X|X| 3 |Y|Y|Y|Y|Y|Y|Y|Y| 4 | | |B|F|W|W|W|W| + |---------------| |---------------| |---------------| |---------------| + +Byte 1: Bit7 => Y overflow + Bit6 => X overflow + Bit5 => Y sign bit + Bit4 => X sign bit + Bit3 => 1 + Bit2 => Middle Button, 1 is pressed, 0 is not pressed. + Bit1 => Right Button, 1 is pressed, 0 is not pressed. + Bit0 => Left Button, 1 is pressed, 0 is not pressed. +Byte 2: X Movement(9-bit 2's complement integers) +Byte 3: Y Movement(9-bit 2's complement integers) +Byte 4: Bit3~Bit0 => the scrolling wheel's movement since the last data report. + valid values, -8 ~ +7 + Bit4 => 1 = 4th mouse button is pressed, Forward one page. + 0 = 4th mouse button is not pressed. + Bit5 => 1 = 5th mouse button is pressed, Backward one page. + 0 = 5th mouse button is not pressed. + +B) MSID 6: Horizontal and Vertical scrolling. +@ Set bit 1 in register 0x40 to 1 + +# FSP replaces scrolling wheel's movement as 4 bits to show horizontal and + vertical scrolling. + +Packet 1 + Bit 7 6 5 4 3 2 1 0 7 6 5 4 3 2 1 0 7 6 5 4 3 2 1 0 7 6 5 4 3 2 1 0 +BYTE |---------------|BYTE |---------------|BYTE|---------------|BYTE|---------------| + 1 |Y|X|y|x|1|M|R|L| 2 |X|X|X|X|X|X|X|X| 3 |Y|Y|Y|Y|Y|Y|Y|Y| 4 | | |B|F|l|r|u|d| + |---------------| |---------------| |---------------| |---------------| + +Byte 1: Bit7 => Y overflow + Bit6 => X overflow + Bit5 => Y sign bit + Bit4 => X sign bit + Bit3 => 1 + Bit2 => Middle Button, 1 is pressed, 0 is not pressed. + Bit1 => Right Button, 1 is pressed, 0 is not pressed. + Bit0 => Left Button, 1 is pressed, 0 is not pressed. +Byte 2: X Movement(9-bit 2's complement integers) +Byte 3: Y Movement(9-bit 2's complement integers) +Byte 4: Bit0 => the Vertical scrolling movement downward. + Bit1 => the Vertical scrolling movement upward. + Bit2 => the Vertical scrolling movement rightward. + Bit3 => the Vertical scrolling movement leftward. + Bit4 => 1 = 4th mouse button is pressed, Forward one page. + 0 = 4th mouse button is not pressed. + Bit5 => 1 = 5th mouse button is pressed, Backward one page. + 0 = 5th mouse button is not pressed. + +C) MSID 7: +# FSP uses 2 packets(8 Bytes) data to represent Absolute Position + so we have PACKET NUMBER to identify packets. + If PACKET NUMBER is 0, the packet is Packet 1. + If PACKET NUMBER is 1, the packet is Packet 2. + Please count this number in program. + +# MSID6 special packet will be enable at the same time when enable MSID 7. + +============================================================================== +* Absolute position for STL3886-G0. +============================================================================== +@ Set bit 2 or 3 in register 0x40 to 1 +@ Set bit 6 in register 0x40 to 1 + +Packet 1 (ABSOLUTE POSITION) + Bit 7 6 5 4 3 2 1 0 7 6 5 4 3 2 1 0 7 6 5 4 3 2 1 0 7 6 5 4 3 2 1 0 +BYTE |---------------|BYTE |---------------|BYTE|---------------|BYTE|---------------| + 1 |0|1|V|1|1|M|R|L| 2 |X|X|X|X|X|X|X|X| 3 |Y|Y|Y|Y|Y|Y|Y|Y| 4 |r|l|d|u|X|X|Y|Y| + |---------------| |---------------| |---------------| |---------------| + +Byte 1: Bit7~Bit6 => 00, Normal data packet + => 01, Absolute coordination packet + => 10, Notify packet + Bit5 => valid bit + Bit4 => 1 + Bit3 => 1 + Bit2 => Middle Button, 1 is pressed, 0 is not pressed. + Bit1 => Right Button, 1 is pressed, 0 is not pressed. + Bit0 => Left Button, 1 is pressed, 0 is not pressed. +Byte 2: X coordinate (xpos[9:2]) +Byte 3: Y coordinate (ypos[9:2]) +Byte 4: Bit1~Bit0 => Y coordinate (xpos[1:0]) + Bit3~Bit2 => X coordinate (ypos[1:0]) + Bit4 => scroll up + Bit5 => scroll down + Bit6 => scroll left + Bit7 => scroll right + +Notify Packet for G0 + Bit 7 6 5 4 3 2 1 0 7 6 5 4 3 2 1 0 7 6 5 4 3 2 1 0 7 6 5 4 3 2 1 0 +BYTE |---------------|BYTE |---------------|BYTE|---------------|BYTE|---------------| + 1 |1|0|0|1|1|M|R|L| 2 |C|C|C|C|C|C|C|C| 3 |M|M|M|M|M|M|M|M| 4 |0|0|0|0|0|0|0|0| + |---------------| |---------------| |---------------| |---------------| + +Byte 1: Bit7~Bit6 => 00, Normal data packet + => 01, Absolute coordination packet + => 10, Notify packet + Bit5 => 0 + Bit4 => 1 + Bit3 => 1 + Bit2 => Middle Button, 1 is pressed, 0 is not pressed. + Bit1 => Right Button, 1 is pressed, 0 is not pressed. + Bit0 => Left Button, 1 is pressed, 0 is not pressed. +Byte 2: Message Type => 0x5A (Enable/Disable status packet) + Mode Type => 0xA5 (Normal/Icon mode status) +Byte 3: Message Type => 0x00 (Disabled) + => 0x01 (Enabled) + Mode Type => 0x00 (Normal) + => 0x01 (Icon) +Byte 4: Bit7~Bit0 => Don't Care + +============================================================================== +* Absolute position for STL3888-A0. +============================================================================== +Packet 1 (ABSOLUTE POSITION) + Bit 7 6 5 4 3 2 1 0 7 6 5 4 3 2 1 0 7 6 5 4 3 2 1 0 7 6 5 4 3 2 1 0 +BYTE |---------------|BYTE |---------------|BYTE|---------------|BYTE|---------------| + 1 |0|1|V|A|1|L|0|1| 2 |X|X|X|X|X|X|X|X| 3 |Y|Y|Y|Y|Y|Y|Y|Y| 4 |x|x|y|y|X|X|Y|Y| + |---------------| |---------------| |---------------| |---------------| + +Byte 1: Bit7~Bit6 => 00, Normal data packet + => 01, Absolute coordination packet + => 10, Notify packet + Bit5 => Valid bit, 0 means that the coordinate is invalid or finger up. + When both fingers are up, the last two reports have zero valid + bit. + Bit4 => arc + Bit3 => 1 + Bit2 => Left Button, 1 is pressed, 0 is released. + Bit1 => 0 + Bit0 => 1 +Byte 2: X coordinate (xpos[9:2]) +Byte 3: Y coordinate (ypos[9:2]) +Byte 4: Bit1~Bit0 => Y coordinate (xpos[1:0]) + Bit3~Bit2 => X coordinate (ypos[1:0]) + Bit5~Bit4 => y1_g + Bit7~Bit6 => x1_g + +Packet 2 (ABSOLUTE POSITION) + Bit 7 6 5 4 3 2 1 0 7 6 5 4 3 2 1 0 7 6 5 4 3 2 1 0 7 6 5 4 3 2 1 0 +BYTE |---------------|BYTE |---------------|BYTE|---------------|BYTE|---------------| + 1 |0|1|V|A|1|R|1|0| 2 |X|X|X|X|X|X|X|X| 3 |Y|Y|Y|Y|Y|Y|Y|Y| 4 |x|x|y|y|X|X|Y|Y| + |---------------| |---------------| |---------------| |---------------| + +Byte 1: Bit7~Bit6 => 00, Normal data packet + => 01, Absolute coordinates packet + => 10, Notify packet + Bit5 => Valid bit, 0 means that the coordinate is invalid or finger up. + When both fingers are up, the last two reports have zero valid + bit. + Bit4 => arc + Bit3 => 1 + Bit2 => Right Button, 1 is pressed, 0 is released. + Bit1 => 1 + Bit0 => 0 +Byte 2: X coordinate (xpos[9:2]) +Byte 3: Y coordinate (ypos[9:2]) +Byte 4: Bit1~Bit0 => Y coordinate (xpos[1:0]) + Bit3~Bit2 => X coordinate (ypos[1:0]) + Bit5~Bit4 => y2_g + Bit7~Bit6 => x2_g + +Notify Packet for STL3888-A0 + Bit 7 6 5 4 3 2 1 0 7 6 5 4 3 2 1 0 7 6 5 4 3 2 1 0 7 6 5 4 3 2 1 0 +BYTE |---------------|BYTE |---------------|BYTE|---------------|BYTE|---------------| + 1 |1|0|1|P|1|M|R|L| 2 |C|C|C|C|C|C|C|C| 3 |0|0|F|F|0|0|0|i| 4 |r|l|d|u|0|0|0|0| + |---------------| |---------------| |---------------| |---------------| + +Byte 1: Bit7~Bit6 => 00, Normal data packet + => 01, Absolute coordination packet + => 10, Notify packet + Bit5 => 1 + Bit4 => when in absolute coordinates mode (valid when EN_PKT_GO is 1): + 0: left button is generated by the on-pad command + 1: left button is generated by the external button + Bit3 => 1 + Bit2 => Middle Button, 1 is pressed, 0 is not pressed. + Bit1 => Right Button, 1 is pressed, 0 is not pressed. + Bit0 => Left Button, 1 is pressed, 0 is not pressed. +Byte 2: Message Type => 0xB7 (Multi Finger, Multi Coordinate mode) +Byte 3: Bit7~Bit6 => Don't care + Bit5~Bit4 => Number of fingers + Bit3~Bit1 => Reserved + Bit0 => 1: enter gesture mode; 0: leaving gesture mode +Byte 4: Bit7 => scroll right button + Bit6 => scroll left button + Bit5 => scroll down button + Bit4 => scroll up button + * Note that if gesture and additional button (Bit4~Bit7) + happen at the same time, the button information will not + be sent. + Bit3~Bit0 => Reserved + +Sample sequence of Multi-finger, Multi-coordinate mode: + + notify packet (valid bit == 1), abs pkt 1, abs pkt 2, abs pkt 1, + abs pkt 2, ..., notify packet(valid bit == 0) + +============================================================================== +* FSP Enable/Disable packet +============================================================================== + Bit 7 6 5 4 3 2 1 0 7 6 5 4 3 2 1 0 7 6 5 4 3 2 1 0 7 6 5 4 3 2 1 0 +BYTE |---------------|BYTE |---------------|BYTE|---------------|BYTE|---------------| + 1 |Y|X|0|0|1|M|R|L| 2 |0|1|0|1|1|0|1|E| 3 | | | | | | | | | 4 | | | | | | | | | + |---------------| |---------------| |---------------| |---------------| + +FSP will send out enable/disable packet when FSP receive PS/2 enable/disable +command. Host will receive the packet which Middle, Right, Left button will +be set. The packet only use byte 0 and byte 1 as a pattern of original packet. +Ignore the other bytes of the packet. + +Byte 1: Bit7 => 0, Y overflow + Bit6 => 0, X overflow + Bit5 => 0, Y sign bit + Bit4 => 0, X sign bit + Bit3 => 1 + Bit2 => 1, Middle Button + Bit1 => 1, Right Button + Bit0 => 1, Left Button +Byte 2: Bit7~1 => (0101101b) + Bit0 => 1 = Enable + 0 = Disable +Byte 3: Don't care +Byte 4: Don't care (MOUSE ID 3, 4) +Byte 5~8: Don't care (Absolute packet) + +============================================================================== +* PS/2 Command Set +============================================================================== + +FSP supports basic PS/2 commanding set and modes, refer to following URL for +details about PS/2 commands: + +http://www.computer-engineering.org/index.php?title=PS/2_Mouse_Interface + +============================================================================== +* Programming Sequence for Determining Packet Parsing Flow +============================================================================== +1. Identify FSP by reading device ID(0x00) and version(0x01) register + +2. Determine number of buttons by reading status2 (0x0b) register + + buttons = reg[0x0b] & 0x30 + + if buttons == 0x30 or buttons == 0x20: + # two/four buttons + Refer to 'Finger Sensing Pad PS/2 Mouse Intellimouse' + section A for packet parsing detail(ignore byte 4, bit ~ 7) + elif buttons == 0x10: + # 6 buttons + Refer to 'Finger Sensing Pad PS/2 Mouse Intellimouse' + section B for packet parsing detail + elif buttons == 0x00: + # 6 buttons + Refer to 'Finger Sensing Pad PS/2 Mouse Intellimouse' + section A for packet parsing detail + +============================================================================== +* Programming Sequence for Register Reading/Writing +============================================================================== + +Register inversion requirement: + + Following values needed to be inverted(the '~' operator in C) before being +sent to FSP: + + 0xe9, 0xee, 0xf2 and 0xff. + +Register swapping requirement: + + Following values needed to have their higher 4 bits and lower 4 bits being +swapped before being sent to FSP: + + 10, 20, 40, 60, 80, 100 and 200. + +Register reading sequence: + + 1. send 0xf3 PS/2 command to FSP; + + 2. send 0x66 PS/2 command to FSP; + + 3. send 0x88 PS/2 command to FSP; + + 4. send 0xf3 PS/2 command to FSP; + + 5. if the register address being to read is not required to be + inverted(refer to the 'Register inversion requirement' section), + goto step 6 + + 5a. send 0x68 PS/2 command to FSP; + + 5b. send the inverted register address to FSP and goto step 8; + + 6. if the register address being to read is not required to be + swapped(refer to the 'Register swapping requirement' section), + goto step 7 + + 6a. send 0xcc PS/2 command to FSP; + + 6b. send the swapped register address to FSP and goto step 8; + + 7. send 0x66 PS/2 command to FSP; + + 7a. send the original register address to FSP and goto step 8; + + 8. send 0xe9(status request) PS/2 command to FSP; + + 9. the response read from FSP should be the requested register value. + +Register writing sequence: + + 1. send 0xf3 PS/2 command to FSP; + + 2. if the register address being to write is not required to be + inverted(refer to the 'Register inversion requirement' section), + goto step 3 + + 2a. send 0x74 PS/2 command to FSP; + + 2b. send the inverted register address to FSP and goto step 5; + + 3. if the register address being to write is not required to be + swapped(refer to the 'Register swapping requirement' section), + goto step 4 + + 3a. send 0x77 PS/2 command to FSP; + + 3b. send the swapped register address to FSP and goto step 5; + + 4. send 0x55 PS/2 command to FSP; + + 4a. send the register address to FSP and goto step 5; + + 5. send 0xf3 PS/2 command to FSP; + + 6. if the register value being to write is not required to be + inverted(refer to the 'Register inversion requirement' section), + goto step 7 + + 6a. send 0x47 PS/2 command to FSP; + + 6b. send the inverted register value to FSP and goto step 9; + + 7. if the register value being to write is not required to be + swapped(refer to the 'Register swapping requirement' section), + goto step 8 + + 7a. send 0x44 PS/2 command to FSP; + + 7b. send the swapped register value to FSP and goto step 9; + + 8. send 0x33 PS/2 command to FSP; + + 8a. send the register value to FSP; + + 9. the register writing sequence is completed. + +============================================================================== +* Register Listing +============================================================================== + +offset width default r/w name +0x00 bit7~bit0 0x01 RO device ID + +0x01 bit7~bit0 0xc0 RW version ID + +0x02 bit7~bit0 0x01 RO vendor ID + +0x03 bit7~bit0 0x01 RO product ID + +0x04 bit3~bit0 0x01 RW revision ID + +0x0b RO test mode status 1 + bit3 1 RO 0: rotate 180 degree, 1: no rotation + + bit5~bit4 RO number of buttons + 11 => 2, lbtn/rbtn + 10 => 4, lbtn/rbtn/scru/scrd + 01 => 6, lbtn/rbtn/scru/scrd/scrl/scrr + 00 => 6, lbtn/rbtn/scru/scrd/fbtn/bbtn + +0x0f RW register file page control + bit0 0 RW 1 to enable page 1 register files + +0x10 RW system control 1 + bit0 1 RW Reserved, must be 1 + bit1 0 RW Reserved, must be 0 + bit4 1 RW Reserved, must be 0 + bit5 0 RW register clock gating enable + 0: read only, 1: read/write enable + (Note that following registers does not require clock gating being + enabled prior to write: 05 06 07 08 09 0c 0f 10 11 12 16 17 18 23 2e + 40 41 42 43.) + +0x31 RW on-pad command detection + bit7 0 RW on-pad command left button down tag + enable + 0: disable, 1: enable + +0x34 RW on-pad command control 5 + bit4~bit0 0x05 RW XLO in 0s/4/1, so 03h = 0010.1b = 2.5 + (Note that position unit is in 0.5 scanline) + + bit7 0 RW on-pad tap zone enable + 0: disable, 1: enable + +0x35 RW on-pad command control 6 + bit4~bit0 0x1d RW XHI in 0s/4/1, so 19h = 1100.1b = 12.5 + (Note that position unit is in 0.5 scanline) + +0x36 RW on-pad command control 7 + bit4~bit0 0x04 RW YLO in 0s/4/1, so 03h = 0010.1b = 2.5 + (Note that position unit is in 0.5 scanline) + +0x37 RW on-pad command control 8 + bit4~bit0 0x13 RW YHI in 0s/4/1, so 11h = 1000.1b = 8.5 + (Note that position unit is in 0.5 scanline) + +0x40 RW system control 5 + bit1 0 RW FSP Intellimouse mode enable + 0: disable, 1: enable + + bit2 0 RW movement + abs. coordinate mode enable + 0: disable, 1: enable + (Note that this function has the functionality of bit 1 even when + bit 1 is not set. However, the format is different from that of bit 1. + In addition, when bit 1 and bit 2 are set at the same time, bit 2 will + override bit 1.) + + bit3 0 RW abs. coordinate only mode enable + 0: disable, 1: enable + (Note that this function has the functionality of bit 1 even when + bit 1 is not set. However, the format is different from that of bit 1. + In addition, when bit 1, bit 2 and bit 3 are set at the same time, + bit 3 will override bit 1 and 2.) + + bit5 0 RW auto switch enable + 0: disable, 1: enable + + bit6 0 RW G0 abs. + notify packet format enable + 0: disable, 1: enable + (Note that the absolute/relative coordinate output still depends on + bit 2 and 3. That is, if any of those bit is 1, host will receive + absolute coordinates; otherwise, host only receives packets with + relative coordinate.) + +0x43 RW on-pad control + bit0 0 RW on-pad control enable + 0: disable, 1: enable + (Note that if this bit is cleared, bit 3/5 will be ineffective) + + bit3 0 RW on-pad fix vertical scrolling enable + 0: disable, 1: enable + + bit5 0 RW on-pad fix horizontal scrolling enable + 0: disable, 1: enable diff --git a/drivers/input/mouse/Kconfig b/drivers/input/mouse/Kconfig index 90bef5d498f0..3feeb3af8abd 100644 --- a/drivers/input/mouse/Kconfig +++ b/drivers/input/mouse/Kconfig @@ -107,6 +107,14 @@ config MOUSE_PS2_ELANTECH entries. For further information, see . +config MOUSE_PS2_SENTELIC + bool "Sentelic Finger Sensing Pad PS/2 protocol extension" + depends on MOUSE_PS2 + help + Say Y here if you have a laptop (such as MSI WIND Netbook) + with Sentelic Finger Sensing Pad touchpad. + + If unsure, say N. config MOUSE_PS2_TOUCHKIT bool "eGalax TouchKit PS/2 protocol extension" diff --git a/drivers/input/mouse/Makefile b/drivers/input/mouse/Makefile index ea58c9a372b6..570c84a4a654 100644 --- a/drivers/input/mouse/Makefile +++ b/drivers/input/mouse/Makefile @@ -27,5 +27,6 @@ psmouse-$(CONFIG_MOUSE_PS2_ELANTECH) += elantech.o psmouse-$(CONFIG_MOUSE_PS2_OLPC) += hgpk.o psmouse-$(CONFIG_MOUSE_PS2_LOGIPS2PP) += logips2pp.o psmouse-$(CONFIG_MOUSE_PS2_LIFEBOOK) += lifebook.o +psmouse-$(CONFIG_MOUSE_PS2_SENTELIC) += sentelic.o psmouse-$(CONFIG_MOUSE_PS2_TRACKPOINT) += trackpoint.o psmouse-$(CONFIG_MOUSE_PS2_TOUCHKIT) += touchkit_ps2.o diff --git a/drivers/input/mouse/psmouse-base.c b/drivers/input/mouse/psmouse-base.c index b407b355dceb..df318887ca09 100644 --- a/drivers/input/mouse/psmouse-base.c +++ b/drivers/input/mouse/psmouse-base.c @@ -30,6 +30,7 @@ #include "trackpoint.h" #include "touchkit_ps2.h" #include "elantech.h" +#include "sentelic.h" #define DRIVER_DESC "PS/2 mouse driver" @@ -666,6 +667,20 @@ static int psmouse_extensions(struct psmouse *psmouse, max_proto = PSMOUSE_IMEX; } +/* + * Try Finger Sensing Pad + */ + if (max_proto > PSMOUSE_IMEX) { + if (fsp_detect(psmouse, set_properties) == 0) { + if (!set_properties || fsp_init(psmouse) == 0) + return PSMOUSE_FSP; +/* + * Init failed, try basic relative protocols + */ + max_proto = PSMOUSE_IMEX; + } + } + if (max_proto > PSMOUSE_IMEX) { if (genius_detect(psmouse, set_properties) == 0) return PSMOUSE_GENPS; @@ -813,7 +828,16 @@ static const struct psmouse_protocol psmouse_protocols[] = { .detect = elantech_detect, .init = elantech_init, }, - #endif +#endif +#ifdef CONFIG_MOUSE_PS2_SENTELIC + { + .type = PSMOUSE_FSP, + .name = "FSPPS/2", + .alias = "fsp", + .detect = fsp_detect, + .init = fsp_init, + }, +#endif { .type = PSMOUSE_CORTRON, .name = "CortronPS/2", diff --git a/drivers/input/mouse/psmouse.h b/drivers/input/mouse/psmouse.h index e3562daeb2ed..cca1744c2a08 100644 --- a/drivers/input/mouse/psmouse.h +++ b/drivers/input/mouse/psmouse.h @@ -91,6 +91,7 @@ enum psmouse_type { PSMOUSE_CORTRON, PSMOUSE_HGPK, PSMOUSE_ELANTECH, + PSMOUSE_FSP, PSMOUSE_AUTO /* This one should always be last */ }; diff --git a/drivers/input/mouse/sentelic.c b/drivers/input/mouse/sentelic.c new file mode 100644 index 000000000000..97b1e72855a0 --- /dev/null +++ b/drivers/input/mouse/sentelic.c @@ -0,0 +1,867 @@ +/*- + * Finger Sensing Pad PS/2 mouse driver. + * + * Copyright (C) 2005-2007 Asia Vital Components Co., Ltd. + * Copyright (C) 2005-2009 Tai-hwa Liang, Sentelic Corporation. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version 2 + * of the License, or (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. + */ + +#include +#include +#include +#include +#include +#include +#include + +#include "psmouse.h" +#include "sentelic.h" + +/* + * Timeout for FSP PS/2 command only (in milliseconds). + */ +#define FSP_CMD_TIMEOUT 200 +#define FSP_CMD_TIMEOUT2 30 + +/** Driver version. */ +static const char fsp_drv_ver[] = "1.0.0-K"; + +/* + * Make sure that the value being sent to FSP will not conflict with + * possible sample rate values. + */ +static unsigned char fsp_test_swap_cmd(unsigned char reg_val) +{ + switch (reg_val) { + case 10: case 20: case 40: case 60: case 80: case 100: case 200: + /* + * The requested value being sent to FSP matched to possible + * sample rates, swap the given value such that the hardware + * wouldn't get confused. + */ + return (reg_val >> 4) | (reg_val << 4); + default: + return reg_val; /* swap isn't necessary */ + } +} + +/* + * Make sure that the value being sent to FSP will not conflict with certain + * commands. + */ +static unsigned char fsp_test_invert_cmd(unsigned char reg_val) +{ + switch (reg_val) { + case 0xe9: case 0xee: case 0xf2: case 0xff: + /* + * The requested value being sent to FSP matched to certain + * commands, inverse the given value such that the hardware + * wouldn't get confused. + */ + return ~reg_val; + default: + return reg_val; /* inversion isn't necessary */ + } +} + +static int fsp_reg_read(struct psmouse *psmouse, int reg_addr, int *reg_val) +{ + struct ps2dev *ps2dev = &psmouse->ps2dev; + unsigned char param[3]; + unsigned char addr; + int rc = -1; + + /* + * We need to shut off the device and switch it into command + * mode so we don't confuse our protocol handler. We don't need + * to do that for writes because sysfs set helper does this for + * us. + */ + ps2_command(ps2dev, NULL, PSMOUSE_CMD_DISABLE); + psmouse_set_state(psmouse, PSMOUSE_CMD_MODE); + mutex_lock(&ps2dev->cmd_mutex); + + if (ps2_sendbyte(ps2dev, 0xf3, FSP_CMD_TIMEOUT) < 0) + goto out; + + /* should return 0xfe(request for resending) */ + ps2_sendbyte(ps2dev, 0x66, FSP_CMD_TIMEOUT2); + /* should return 0xfc(failed) */ + ps2_sendbyte(ps2dev, 0x88, FSP_CMD_TIMEOUT2); + + if (ps2_sendbyte(ps2dev, 0xf3, FSP_CMD_TIMEOUT) < 0) + goto out; + + if ((addr = fsp_test_invert_cmd(reg_addr)) != reg_addr) { + ps2_sendbyte(ps2dev, 0x68, FSP_CMD_TIMEOUT2); + } else if ((addr = fsp_test_swap_cmd(reg_addr)) != reg_addr) { + /* swapping is required */ + ps2_sendbyte(ps2dev, 0xcc, FSP_CMD_TIMEOUT2); + /* expect 0xfe */ + } else { + /* swapping isn't necessary */ + ps2_sendbyte(ps2dev, 0x66, FSP_CMD_TIMEOUT2); + /* expect 0xfe */ + } + /* should return 0xfc(failed) */ + ps2_sendbyte(ps2dev, addr, FSP_CMD_TIMEOUT); + + if (__ps2_command(ps2dev, param, PSMOUSE_CMD_GETINFO) < 0) + goto out; + + *reg_val = param[2]; + rc = 0; + + out: + mutex_unlock(&ps2dev->cmd_mutex); + ps2_command(ps2dev, NULL, PSMOUSE_CMD_ENABLE); + psmouse_set_state(psmouse, PSMOUSE_ACTIVATED); + dev_dbg(&ps2dev->serio->dev, "READ REG: 0x%02x is 0x%02x (rc = %d)\n", + reg_addr, *reg_val, rc); + return rc; +} + +static int fsp_reg_write(struct psmouse *psmouse, int reg_addr, int reg_val) +{ + struct ps2dev *ps2dev = &psmouse->ps2dev; + unsigned char v; + int rc = -1; + + mutex_lock(&ps2dev->cmd_mutex); + + if (ps2_sendbyte(ps2dev, 0xf3, FSP_CMD_TIMEOUT) < 0) + goto out; + + if ((v = fsp_test_invert_cmd(reg_addr)) != reg_addr) { + /* inversion is required */ + ps2_sendbyte(ps2dev, 0x74, FSP_CMD_TIMEOUT2); + } else { + if ((v = fsp_test_swap_cmd(reg_addr)) != reg_addr) { + /* swapping is required */ + ps2_sendbyte(ps2dev, 0x77, FSP_CMD_TIMEOUT2); + } else { + /* swapping isn't necessary */ + ps2_sendbyte(ps2dev, 0x55, FSP_CMD_TIMEOUT2); + } + } + /* write the register address in correct order */ + ps2_sendbyte(ps2dev, v, FSP_CMD_TIMEOUT2); + + if (ps2_sendbyte(ps2dev, 0xf3, FSP_CMD_TIMEOUT) < 0) + return -1; + + if ((v = fsp_test_invert_cmd(reg_val)) != reg_val) { + /* inversion is required */ + ps2_sendbyte(ps2dev, 0x47, FSP_CMD_TIMEOUT2); + } else if ((v = fsp_test_swap_cmd(reg_val)) != reg_val) { + /* swapping is required */ + ps2_sendbyte(ps2dev, 0x44, FSP_CMD_TIMEOUT2); + } else { + /* swapping isn't necessary */ + ps2_sendbyte(ps2dev, 0x33, FSP_CMD_TIMEOUT2); + } + + /* write the register value in correct order */ + ps2_sendbyte(ps2dev, v, FSP_CMD_TIMEOUT2); + rc = 0; + + out: + mutex_unlock(&ps2dev->cmd_mutex); + dev_dbg(&ps2dev->serio->dev, "WRITE REG: 0x%02x to 0x%02x (rc = %d)\n", + reg_addr, reg_val, rc); + return rc; +} + +/* Enable register clock gating for writing certain registers */ +static int fsp_reg_write_enable(struct psmouse *psmouse, bool enable) +{ + int v, nv; + + if (fsp_reg_read(psmouse, FSP_REG_SYSCTL1, &v) == -1) + return -1; + + if (enable) + nv = v | FSP_BIT_EN_REG_CLK; + else + nv = v & ~FSP_BIT_EN_REG_CLK; + + /* only write if necessary */ + if (nv != v) + if (fsp_reg_write(psmouse, FSP_REG_SYSCTL1, nv) == -1) + return -1; + + return 0; +} + +static int fsp_page_reg_read(struct psmouse *psmouse, int *reg_val) +{ + struct ps2dev *ps2dev = &psmouse->ps2dev; + unsigned char param[3]; + int rc = -1; + + ps2_command(ps2dev, NULL, PSMOUSE_CMD_DISABLE); + psmouse_set_state(psmouse, PSMOUSE_CMD_MODE); + mutex_lock(&ps2dev->cmd_mutex); + + if (ps2_sendbyte(ps2dev, 0xf3, FSP_CMD_TIMEOUT) < 0) + goto out; + + ps2_sendbyte(ps2dev, 0x66, FSP_CMD_TIMEOUT2); + ps2_sendbyte(ps2dev, 0x88, FSP_CMD_TIMEOUT2); + + if (ps2_sendbyte(ps2dev, 0xf3, FSP_CMD_TIMEOUT) < 0) + goto out; + + ps2_sendbyte(ps2dev, 0x83, FSP_CMD_TIMEOUT2); + ps2_sendbyte(ps2dev, 0x88, FSP_CMD_TIMEOUT2); + + /* get the returned result */ + if (__ps2_command(ps2dev, param, PSMOUSE_CMD_GETINFO)) + goto out; + + *reg_val = param[2]; + rc = 0; + + out: + mutex_unlock(&ps2dev->cmd_mutex); + ps2_command(ps2dev, NULL, PSMOUSE_CMD_ENABLE); + psmouse_set_state(psmouse, PSMOUSE_ACTIVATED); + dev_dbg(&ps2dev->serio->dev, "READ PAGE REG: 0x%02x (rc = %d)\n", + *reg_val, rc); + return rc; +} + +static int fsp_page_reg_write(struct psmouse *psmouse, int reg_val) +{ + struct ps2dev *ps2dev = &psmouse->ps2dev; + unsigned char v; + int rc = -1; + + mutex_lock(&ps2dev->cmd_mutex); + + if (ps2_sendbyte(ps2dev, 0xf3, FSP_CMD_TIMEOUT) < 0) + goto out; + + ps2_sendbyte(ps2dev, 0x38, FSP_CMD_TIMEOUT2); + ps2_sendbyte(ps2dev, 0x88, FSP_CMD_TIMEOUT2); + + if (ps2_sendbyte(ps2dev, 0xf3, FSP_CMD_TIMEOUT) < 0) + return -1; + + if ((v = fsp_test_invert_cmd(reg_val)) != reg_val) { + ps2_sendbyte(ps2dev, 0x47, FSP_CMD_TIMEOUT2); + } else if ((v = fsp_test_swap_cmd(reg_val)) != reg_val) { + /* swapping is required */ + ps2_sendbyte(ps2dev, 0x44, FSP_CMD_TIMEOUT2); + } else { + /* swapping isn't necessary */ + ps2_sendbyte(ps2dev, 0x33, FSP_CMD_TIMEOUT2); + } + + ps2_sendbyte(ps2dev, v, FSP_CMD_TIMEOUT2); + rc = 0; + + out: + mutex_unlock(&ps2dev->cmd_mutex); + dev_dbg(&ps2dev->serio->dev, "WRITE PAGE REG: to 0x%02x (rc = %d)\n", + reg_val, rc); + return rc; +} + +static int fsp_get_version(struct psmouse *psmouse, int *version) +{ + if (fsp_reg_read(psmouse, FSP_REG_VERSION, version)) + return -EIO; + + return 0; +} + +static int fsp_get_revision(struct psmouse *psmouse, int *rev) +{ + if (fsp_reg_read(psmouse, FSP_REG_REVISION, rev)) + return -EIO; + + return 0; +} + +static int fsp_get_buttons(struct psmouse *psmouse, int *btn) +{ + static const int buttons[] = { + 0x16, /* Left/Middle/Right/Forward/Backward & Scroll Up/Down */ + 0x06, /* Left/Middle/Right & Scroll Up/Down/Right/Left */ + 0x04, /* Left/Middle/Right & Scroll Up/Down */ + 0x02, /* Left/Middle/Right */ + }; + int val; + + if (fsp_reg_read(psmouse, FSP_REG_TMOD_STATUS1, &val) == -1) + return -EIO; + + *btn = buttons[(val & 0x30) >> 4]; + return 0; +} + +/* Enable on-pad command tag output */ +static int fsp_opc_tag_enable(struct psmouse *psmouse, bool enable) +{ + int v, nv; + int res = 0; + + if (fsp_reg_read(psmouse, FSP_REG_OPC_QDOWN, &v) == -1) { + dev_err(&psmouse->ps2dev.serio->dev, "Unable get OPC state.\n"); + return -EIO; + } + + if (enable) + nv = v | FSP_BIT_EN_OPC_TAG; + else + nv = v & ~FSP_BIT_EN_OPC_TAG; + + /* only write if necessary */ + if (nv != v) { + fsp_reg_write_enable(psmouse, true); + res = fsp_reg_write(psmouse, FSP_REG_OPC_QDOWN, nv); + fsp_reg_write_enable(psmouse, false); + } + + if (res != 0) { + dev_err(&psmouse->ps2dev.serio->dev, + "Unable to enable OPC tag.\n"); + res = -EIO; + } + + return res; +} + +static int fsp_onpad_vscr(struct psmouse *psmouse, bool enable) +{ + struct fsp_data *pad = psmouse->private; + int val; + + if (fsp_reg_read(psmouse, FSP_REG_ONPAD_CTL, &val)) + return -EIO; + + pad->vscroll = enable; + + if (enable) + val |= (FSP_BIT_FIX_VSCR | FSP_BIT_ONPAD_ENABLE); + else + val &= ~FSP_BIT_FIX_VSCR; + + if (fsp_reg_write(psmouse, FSP_REG_ONPAD_CTL, val)) + return -EIO; + + return 0; +} + +static int fsp_onpad_hscr(struct psmouse *psmouse, bool enable) +{ + struct fsp_data *pad = psmouse->private; + int val, v2; + + if (fsp_reg_read(psmouse, FSP_REG_ONPAD_CTL, &val)) + return -EIO; + + if (fsp_reg_read(psmouse, FSP_REG_SYSCTL5, &v2)) + return -EIO; + + pad->hscroll = enable; + + if (enable) { + val |= (FSP_BIT_FIX_HSCR | FSP_BIT_ONPAD_ENABLE); + v2 |= FSP_BIT_EN_MSID6; + } else { + val &= ~FSP_BIT_FIX_HSCR; + v2 &= ~(FSP_BIT_EN_MSID6 | FSP_BIT_EN_MSID7 | FSP_BIT_EN_MSID8); + } + + if (fsp_reg_write(psmouse, FSP_REG_ONPAD_CTL, val)) + return -EIO; + + /* reconfigure horizontal scrolling packet output */ + if (fsp_reg_write(psmouse, FSP_REG_SYSCTL5, v2)) + return -EIO; + + return 0; +} + +/* + * Write device specific initial parameters. + * + * ex: 0xab 0xcd - write oxcd into register 0xab + */ +static ssize_t fsp_attr_set_setreg(struct psmouse *psmouse, void *data, + const char *buf, size_t count) +{ + unsigned long reg, val; + char *rest; + ssize_t retval; + + reg = simple_strtoul(buf, &rest, 16); + if (rest == buf || *rest != ' ' || reg > 0xff) + return -EINVAL; + + if (strict_strtoul(rest + 1, 16, &val) || val > 0xff) + return -EINVAL; + + if (fsp_reg_write_enable(psmouse, true)) + return -EIO; + + retval = fsp_reg_write(psmouse, reg, val) < 0 ? -EIO : count; + + fsp_reg_write_enable(psmouse, false); + + return count; +} + +PSMOUSE_DEFINE_WO_ATTR(setreg, S_IWUSR, NULL, fsp_attr_set_setreg); + +static ssize_t fsp_attr_show_getreg(struct psmouse *psmouse, + void *data, char *buf) +{ + struct fsp_data *pad = psmouse->private; + + return sprintf(buf, "%02x%02x\n", pad->last_reg, pad->last_val); +} + +/* + * Read a register from device. + * + * ex: 0xab -- read content from register 0xab + */ +static ssize_t fsp_attr_set_getreg(struct psmouse *psmouse, void *data, + const char *buf, size_t count) +{ + struct fsp_data *pad = psmouse->private; + unsigned long reg; + int val; + + if (strict_strtoul(buf, 16, ®) || reg > 0xff) + return -EINVAL; + + if (fsp_reg_read(psmouse, reg, &val)) + return -EIO; + + pad->last_reg = reg; + pad->last_val = val; + + return count; +} + +PSMOUSE_DEFINE_ATTR(getreg, S_IWUSR | S_IRUGO, NULL, + fsp_attr_show_getreg, fsp_attr_set_getreg); + +static ssize_t fsp_attr_show_pagereg(struct psmouse *psmouse, + void *data, char *buf) +{ + int val = 0; + + if (fsp_page_reg_read(psmouse, &val)) + return -EIO; + + return sprintf(buf, "%02x\n", val); +} + +static ssize_t fsp_attr_set_pagereg(struct psmouse *psmouse, void *data, + const char *buf, size_t count) +{ + unsigned long val; + + if (strict_strtoul(buf, 16, &val) || val > 0xff) + return -EINVAL; + + if (fsp_page_reg_write(psmouse, val)) + return -EIO; + + return count; +} + +PSMOUSE_DEFINE_ATTR(page, S_IWUSR | S_IRUGO, NULL, + fsp_attr_show_pagereg, fsp_attr_set_pagereg); + +static ssize_t fsp_attr_show_vscroll(struct psmouse *psmouse, + void *data, char *buf) +{ + struct fsp_data *pad = psmouse->private; + + return sprintf(buf, "%d\n", pad->vscroll); +} + +static ssize_t fsp_attr_set_vscroll(struct psmouse *psmouse, void *data, + const char *buf, size_t count) +{ + unsigned long val; + + if (strict_strtoul(buf, 10, &val) || val > 1) + return -EINVAL; + + fsp_onpad_vscr(psmouse, val); + + return count; +} + +PSMOUSE_DEFINE_ATTR(vscroll, S_IWUSR | S_IRUGO, NULL, + fsp_attr_show_vscroll, fsp_attr_set_vscroll); + +static ssize_t fsp_attr_show_hscroll(struct psmouse *psmouse, + void *data, char *buf) +{ + struct fsp_data *pad = psmouse->private; + + return sprintf(buf, "%d\n", pad->hscroll); +} + +static ssize_t fsp_attr_set_hscroll(struct psmouse *psmouse, void *data, + const char *buf, size_t count) +{ + unsigned long val; + + if (strict_strtoul(buf, 10, &val) || val > 1) + return -EINVAL; + + fsp_onpad_hscr(psmouse, val); + + return count; +} + +PSMOUSE_DEFINE_ATTR(hscroll, S_IWUSR | S_IRUGO, NULL, + fsp_attr_show_hscroll, fsp_attr_set_hscroll); + +static ssize_t fsp_attr_show_flags(struct psmouse *psmouse, + void *data, char *buf) +{ + struct fsp_data *pad = psmouse->private; + + return sprintf(buf, "%c\n", + pad->flags & FSPDRV_FLAG_EN_OPC ? 'C' : 'c'); +} + +static ssize_t fsp_attr_set_flags(struct psmouse *psmouse, void *data, + const char *buf, size_t count) +{ + struct fsp_data *pad = psmouse->private; + size_t i; + + for (i = 0; i < count; i++) { + switch (buf[i]) { + case 'C': + pad->flags |= FSPDRV_FLAG_EN_OPC; + break; + case 'c': + pad->flags &= ~FSPDRV_FLAG_EN_OPC; + break; + default: + return -EINVAL; + } + } + return count; +} + +PSMOUSE_DEFINE_ATTR(flags, S_IWUSR | S_IRUGO, NULL, + fsp_attr_show_flags, fsp_attr_set_flags); + +static ssize_t fsp_attr_show_ver(struct psmouse *psmouse, + void *data, char *buf) +{ + return sprintf(buf, "Sentelic FSP kernel module %s\n", fsp_drv_ver); +} + +PSMOUSE_DEFINE_RO_ATTR(ver, S_IRUGO, NULL, fsp_attr_show_ver); + +static struct attribute *fsp_attributes[] = { + &psmouse_attr_setreg.dattr.attr, + &psmouse_attr_getreg.dattr.attr, + &psmouse_attr_page.dattr.attr, + &psmouse_attr_vscroll.dattr.attr, + &psmouse_attr_hscroll.dattr.attr, + &psmouse_attr_flags.dattr.attr, + &psmouse_attr_ver.dattr.attr, + NULL +}; + +static struct attribute_group fsp_attribute_group = { + .attrs = fsp_attributes, +}; + +#ifdef FSP_DEBUG +static void fsp_packet_debug(unsigned char packet[]) +{ + static unsigned int ps2_packet_cnt; + static unsigned int ps2_last_second; + unsigned int jiffies_msec; + + ps2_packet_cnt++; + jiffies_msec = jiffies_to_msecs(jiffies); + printk(KERN_DEBUG "%08dms PS/2 packets: %02x, %02x, %02x, %02x\n", + jiffies_msec, packet[0], packet[1], packet[2], packet[3]); + + if (jiffies_msec - ps2_last_second > 1000) { + printk(KERN_DEBUG "PS/2 packets/sec = %d\n", ps2_packet_cnt); + ps2_packet_cnt = 0; + ps2_last_second = jiffies_msec; + } +} +#else +static void fsp_packet_debug(unsigned char packet[]) +{ +} +#endif + +static psmouse_ret_t fsp_process_byte(struct psmouse *psmouse) +{ + struct input_dev *dev = psmouse->dev; + struct fsp_data *ad = psmouse->private; + unsigned char *packet = psmouse->packet; + unsigned char button_status = 0, lscroll = 0, rscroll = 0; + int rel_x, rel_y; + + if (psmouse->pktcnt < 4) + return PSMOUSE_GOOD_DATA; + + /* + * Full packet accumulated, process it + */ + + switch (psmouse->packet[0] >> FSP_PKT_TYPE_SHIFT) { + case FSP_PKT_TYPE_ABS: + dev_warn(&psmouse->ps2dev.serio->dev, + "Unexpected absolute mode packet, ignored.\n"); + break; + + case FSP_PKT_TYPE_NORMAL_OPC: + /* on-pad click, filter it if necessary */ + if ((ad->flags & FSPDRV_FLAG_EN_OPC) != FSPDRV_FLAG_EN_OPC) + packet[0] &= ~BIT(0); + /* fall through */ + + case FSP_PKT_TYPE_NORMAL: + /* normal packet */ + /* special packet data translation from on-pad packets */ + if (packet[3] != 0) { + if (packet[3] & BIT(0)) + button_status |= 0x01; /* wheel down */ + if (packet[3] & BIT(1)) + button_status |= 0x0f; /* wheel up */ + if (packet[3] & BIT(2)) + button_status |= BIT(5);/* horizontal left */ + if (packet[3] & BIT(3)) + button_status |= BIT(4);/* horizontal right */ + /* push back to packet queue */ + if (button_status != 0) + packet[3] = button_status; + rscroll = (packet[3] >> 4) & 1; + lscroll = (packet[3] >> 5) & 1; + } + /* + * Processing wheel up/down and extra button events + */ + input_report_rel(dev, REL_WHEEL, + (int)(packet[3] & 8) - (int)(packet[3] & 7)); + input_report_rel(dev, REL_HWHEEL, lscroll - rscroll); + input_report_key(dev, BTN_BACK, lscroll); + input_report_key(dev, BTN_FORWARD, rscroll); + + /* + * Standard PS/2 Mouse + */ + input_report_key(dev, BTN_LEFT, packet[0] & 1); + input_report_key(dev, BTN_MIDDLE, (packet[0] >> 2) & 1); + input_report_key(dev, BTN_RIGHT, (packet[0] >> 1) & 1); + + rel_x = packet[1] ? (int)packet[1] - (int)((packet[0] << 4) & 0x100) : 0; + rel_y = packet[2] ? (int)((packet[0] << 3) & 0x100) - (int)packet[2] : 0; + + input_report_rel(dev, REL_X, rel_x); + input_report_rel(dev, REL_Y, rel_y); + break; + } + + input_sync(dev); + + fsp_packet_debug(packet); + + return PSMOUSE_FULL_PACKET; +} + +static int fsp_activate_protocol(struct psmouse *psmouse) +{ + struct fsp_data *pad = psmouse->private; + struct ps2dev *ps2dev = &psmouse->ps2dev; + unsigned char param[2]; + int val; + + /* + * Standard procedure to enter FSP Intellimouse mode + * (scrolling wheel, 4th and 5th buttons) + */ + param[0] = 200; + ps2_command(ps2dev, param, PSMOUSE_CMD_SETRATE); + param[0] = 200; + ps2_command(ps2dev, param, PSMOUSE_CMD_SETRATE); + param[0] = 80; + ps2_command(ps2dev, param, PSMOUSE_CMD_SETRATE); + + ps2_command(ps2dev, param, PSMOUSE_CMD_GETID); + if (param[0] != 0x04) { + dev_err(&psmouse->ps2dev.serio->dev, + "Unable to enable 4 bytes packet format.\n"); + return -EIO; + } + + if (fsp_reg_read(psmouse, FSP_REG_SYSCTL5, &val)) { + dev_err(&psmouse->ps2dev.serio->dev, + "Unable to read SYSCTL5 register.\n"); + return -EIO; + } + + val &= ~(FSP_BIT_EN_MSID7 | FSP_BIT_EN_MSID8 | FSP_BIT_EN_AUTO_MSID8); + /* Ensure we are not in absolute mode */ + val &= ~FSP_BIT_EN_PKT_G0; + if (pad->buttons == 0x06) { + /* Left/Middle/Right & Scroll Up/Down/Right/Left */ + val |= FSP_BIT_EN_MSID6; + } + + if (fsp_reg_write(psmouse, FSP_REG_SYSCTL5, val)) { + dev_err(&psmouse->ps2dev.serio->dev, + "Unable to set up required mode bits.\n"); + return -EIO; + } + + /* + * Enable OPC tags such that driver can tell the difference between + * on-pad and real button click + */ + if (fsp_opc_tag_enable(psmouse, true)) + dev_warn(&psmouse->ps2dev.serio->dev, + "Failed to enable OPC tag mode.\n"); + + /* Enable on-pad vertical and horizontal scrolling */ + fsp_onpad_vscr(psmouse, true); + fsp_onpad_hscr(psmouse, true); + + return 0; +} + +int fsp_detect(struct psmouse *psmouse, int set_properties) +{ + int id; + + if (fsp_reg_read(psmouse, FSP_REG_DEVICE_ID, &id)) + return -EIO; + + if (id != 0x01) + return -ENODEV; + + if (set_properties) { + psmouse->vendor = "Sentelic"; + psmouse->name = "FingerSensingPad"; + } + + return 0; +} + +static void fsp_reset(struct psmouse *psmouse) +{ + fsp_opc_tag_enable(psmouse, false); + fsp_onpad_vscr(psmouse, false); + fsp_onpad_hscr(psmouse, false); +} + +static void fsp_disconnect(struct psmouse *psmouse) +{ + sysfs_remove_group(&psmouse->ps2dev.serio->dev.kobj, + &fsp_attribute_group); + + fsp_reset(psmouse); + kfree(psmouse->private); +} + +static int fsp_reconnect(struct psmouse *psmouse) +{ + int version; + + if (fsp_detect(psmouse, 0)) + return -ENODEV; + + if (fsp_get_version(psmouse, &version)) + return -ENODEV; + + if (fsp_activate_protocol(psmouse)) + return -EIO; + + return 0; +} + +int fsp_init(struct psmouse *psmouse) +{ + struct fsp_data *priv; + int ver, rev, buttons; + int error; + + if (fsp_get_version(psmouse, &ver) || + fsp_get_revision(psmouse, &rev) || + fsp_get_buttons(psmouse, &buttons)) { + return -ENODEV; + } + + printk(KERN_INFO + "Finger Sensing Pad, hw: %d.%d.%d, sw: %s, buttons: %d\n", + ver >> 4, ver & 0x0F, rev, fsp_drv_ver, buttons & 7); + + psmouse->private = priv = kzalloc(sizeof(struct fsp_data), GFP_KERNEL); + if (!priv) + return -ENOMEM; + + priv->ver = ver; + priv->rev = rev; + priv->buttons = buttons; + + /* enable on-pad click by default */ + priv->flags |= FSPDRV_FLAG_EN_OPC; + + /* Set up various supported input event bits */ + __set_bit(BTN_BACK, psmouse->dev->keybit); + __set_bit(BTN_FORWARD, psmouse->dev->keybit); + __set_bit(REL_WHEEL, psmouse->dev->relbit); + __set_bit(REL_HWHEEL, psmouse->dev->relbit); + + psmouse->protocol_handler = fsp_process_byte; + psmouse->disconnect = fsp_disconnect; + psmouse->reconnect = fsp_reconnect; + psmouse->cleanup = fsp_reset; + psmouse->pktsize = 4; + + /* set default packet output based on number of buttons we found */ + error = fsp_activate_protocol(psmouse); + if (error) + goto err_out; + + error = sysfs_create_group(&psmouse->ps2dev.serio->dev.kobj, + &fsp_attribute_group); + if (error) { + dev_err(&psmouse->ps2dev.serio->dev, + "Failed to create sysfs attributes (%d)", error); + goto err_out; + } + + return 0; + + err_out: + kfree(psmouse->private); + psmouse->private = NULL; + return error; +} diff --git a/drivers/input/mouse/sentelic.h b/drivers/input/mouse/sentelic.h new file mode 100644 index 000000000000..083559c7282b --- /dev/null +++ b/drivers/input/mouse/sentelic.h @@ -0,0 +1,98 @@ +/*- + * Finger Sensing Pad PS/2 mouse driver. + * + * Copyright (C) 2005-2007 Asia Vital Components Co., Ltd. + * Copyright (C) 2005-2009 Tai-hwa Liang, Sentelic Corporation. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version 2 + * of the License, or (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. + */ + +#ifndef __SENTELIC_H +#define __SENTELIC_H + +/* Finger-sensing Pad information registers */ +#define FSP_REG_DEVICE_ID 0x00 +#define FSP_REG_VERSION 0x01 +#define FSP_REG_REVISION 0x04 +#define FSP_REG_TMOD_STATUS1 0x0B +#define FSP_BIT_NO_ROTATION BIT(3) +#define FSP_REG_PAGE_CTRL 0x0F + +/* Finger-sensing Pad control registers */ +#define FSP_REG_SYSCTL1 0x10 +#define FSP_BIT_EN_REG_CLK BIT(5) +#define FSP_REG_OPC_QDOWN 0x31 +#define FSP_BIT_EN_OPC_TAG BIT(7) +#define FSP_REG_OPTZ_XLO 0x34 +#define FSP_REG_OPTZ_XHI 0x35 +#define FSP_REG_OPTZ_YLO 0x36 +#define FSP_REG_OPTZ_YHI 0x37 +#define FSP_REG_SYSCTL5 0x40 +#define FSP_BIT_90_DEGREE BIT(0) +#define FSP_BIT_EN_MSID6 BIT(1) +#define FSP_BIT_EN_MSID7 BIT(2) +#define FSP_BIT_EN_MSID8 BIT(3) +#define FSP_BIT_EN_AUTO_MSID8 BIT(5) +#define FSP_BIT_EN_PKT_G0 BIT(6) + +#define FSP_REG_ONPAD_CTL 0x43 +#define FSP_BIT_ONPAD_ENABLE BIT(0) +#define FSP_BIT_ONPAD_FBBB BIT(1) +#define FSP_BIT_FIX_VSCR BIT(3) +#define FSP_BIT_FIX_HSCR BIT(5) +#define FSP_BIT_DRAG_LOCK BIT(6) + +/* Finger-sensing Pad packet formating related definitions */ + +/* absolute packet type */ +#define FSP_PKT_TYPE_NORMAL (0x00) +#define FSP_PKT_TYPE_ABS (0x01) +#define FSP_PKT_TYPE_NOTIFY (0x02) +#define FSP_PKT_TYPE_NORMAL_OPC (0x03) +#define FSP_PKT_TYPE_SHIFT (6) + +#ifdef __KERNEL__ + +struct fsp_data { + unsigned char ver; /* hardware version */ + unsigned char rev; /* hardware revison */ + unsigned char buttons; /* Number of buttons */ + unsigned int flags; +#define FSPDRV_FLAG_EN_OPC (0x001) /* enable on-pad clicking */ + + bool vscroll; /* Vertical scroll zone enabled */ + bool hscroll; /* Horizontal scroll zone enabled */ + + unsigned char last_reg; /* Last register we requested read from */ + unsigned char last_val; +}; + +#ifdef CONFIG_MOUSE_PS2_SENTELIC +extern int fsp_detect(struct psmouse *psmouse, int set_properties); +extern int fsp_init(struct psmouse *psmouse); +#else +inline int fsp_detect(struct psmouse *psmouse, int set_properties) +{ + return -ENOSYS; +} +inline int fsp_init(struct psmouse *psmouse) +{ + return -ENOSYS; +} +#endif + +#endif /* __KERNEL__ */ + +#endif /* !__SENTELIC_H */ diff --git a/drivers/input/serio/libps2.c b/drivers/input/serio/libps2.c index be5bbbb8ae4e..3a95b508bf27 100644 --- a/drivers/input/serio/libps2.c +++ b/drivers/input/serio/libps2.c @@ -161,7 +161,7 @@ static int ps2_adjust_timeout(struct ps2dev *ps2dev, int command, int timeout) * ps2_command() can only be called from a process context */ -int ps2_command(struct ps2dev *ps2dev, unsigned char *param, int command) +int __ps2_command(struct ps2dev *ps2dev, unsigned char *param, int command) { int timeout; int send = (command >> 12) & 0xf; @@ -179,8 +179,6 @@ int ps2_command(struct ps2dev *ps2dev, unsigned char *param, int command) return -1; } - mutex_lock(&ps2dev->cmd_mutex); - serio_pause_rx(ps2dev->serio); ps2dev->flags = command == PS2_CMD_GETID ? PS2_FLAG_WAITID : 0; ps2dev->cmdcnt = receive; @@ -231,7 +229,18 @@ int ps2_command(struct ps2dev *ps2dev, unsigned char *param, int command) ps2dev->flags = 0; serio_continue_rx(ps2dev->serio); + return rc; +} +EXPORT_SYMBOL(__ps2_command); + +int ps2_command(struct ps2dev *ps2dev, unsigned char *param, int command) +{ + int rc; + + mutex_lock(&ps2dev->cmd_mutex); + rc = __ps2_command(ps2dev, param, command); mutex_unlock(&ps2dev->cmd_mutex); + return rc; } EXPORT_SYMBOL(ps2_command); diff --git a/include/linux/libps2.h b/include/linux/libps2.h index b94534b7e266..fcf5fbe6a50c 100644 --- a/include/linux/libps2.h +++ b/include/linux/libps2.h @@ -44,6 +44,7 @@ struct ps2dev { void ps2_init(struct ps2dev *ps2dev, struct serio *serio); int ps2_sendbyte(struct ps2dev *ps2dev, unsigned char byte, int timeout); void ps2_drain(struct ps2dev *ps2dev, int maxbytes, int timeout); +int __ps2_command(struct ps2dev *ps2dev, unsigned char *param, int command); int ps2_command(struct ps2dev *ps2dev, unsigned char *param, int command); int ps2_handle_ack(struct ps2dev *ps2dev, unsigned char data); int ps2_handle_response(struct ps2dev *ps2dev, unsigned char data); -- cgit v1.2.3 From 3ac64beecd27400d12cc7afb4108eef26c499f6a Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Mon, 17 Aug 2009 16:16:53 +0200 Subject: mac80211: allow configure_filter callback to sleep Over time, a whole bunch of drivers have come up with their own scheme to delay the configure_filter operation to a workqueue. To be able to simplify things, allow configure_filter to sleep, and add a new prepare_multicast callback that drivers that need the multicast address list implement. This new callback must be atomic, but most drivers either don't care or just calculate a hash which can be done atomically and then uploaded to the hardware non-atomically. A cursory look suggests that at76c50x-usb, ar9170, mwl8k (which is actually very broken now), rt2x00, wl1251, wl1271 and zd1211 should make use of this new capability. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- drivers/net/wireless/adm8211.c | 42 +++++++++++++------- drivers/net/wireless/at76c50x-usb.c | 7 ++-- drivers/net/wireless/ath/ar9170/main.c | 43 +++++++++++--------- drivers/net/wireless/ath/ath5k/base.c | 64 ++++++++++++++++++------------ drivers/net/wireless/ath/ath9k/main.c | 3 +- drivers/net/wireless/b43/main.c | 2 +- drivers/net/wireless/b43legacy/main.c | 4 +- drivers/net/wireless/iwlwifi/iwl-core.c | 2 +- drivers/net/wireless/iwlwifi/iwl-core.h | 3 +- drivers/net/wireless/libertas_tf/main.c | 37 +++++++++++------ drivers/net/wireless/mac80211_hwsim.c | 4 +- drivers/net/wireless/mwl8k.c | 34 ++++++++++++---- drivers/net/wireless/p54/main.c | 2 +- drivers/net/wireless/rt2x00/rt2x00.h | 2 +- drivers/net/wireless/rt2x00/rt2x00mac.c | 2 +- drivers/net/wireless/rtl818x/rtl8180_dev.c | 11 ++++- drivers/net/wireless/rtl818x/rtl8187_dev.c | 11 ++++- drivers/net/wireless/wl12xx/wl1251_main.c | 4 +- drivers/net/wireless/wl12xx/wl1271_main.c | 4 +- drivers/net/wireless/zd1211rw/zd_mac.c | 44 +++++++++++++------- include/net/mac80211.h | 21 +++++++--- net/mac80211/driver-ops.h | 24 +++++++++-- net/mac80211/driver-trace.h | 36 ++++++++++++++--- net/mac80211/ieee80211_i.h | 3 ++ net/mac80211/iface.c | 15 ++----- net/mac80211/main.c | 24 ++++++++--- net/mac80211/scan.c | 16 +------- net/mac80211/util.c | 2 - 28 files changed, 297 insertions(+), 169 deletions(-) (limited to 'include') diff --git a/drivers/net/wireless/adm8211.c b/drivers/net/wireless/adm8211.c index 5695911bc602..b80f514877d8 100644 --- a/drivers/net/wireless/adm8211.c +++ b/drivers/net/wireless/adm8211.c @@ -1328,16 +1328,39 @@ static void adm8211_bss_info_changed(struct ieee80211_hw *dev, } } +static u64 adm8211_prepare_multicast(struct ieee80211_hw *hw, + int mc_count, struct dev_addr_list *mclist) +{ + unsigned int bit_nr, i; + u32 mc_filter[2]; + + mc_filter[1] = mc_filter[0] = 0; + + for (i = 0; i < mc_count; i++) { + if (!mclist) + break; + bit_nr = ether_crc(ETH_ALEN, mclist->dmi_addr) >> 26; + + bit_nr &= 0x3F; + mc_filter[bit_nr >> 5] |= 1 << (bit_nr & 31); + mclist = mclist->next; + } + + return mc_filter[0] | ((u64)(mc_filter[1]) << 32); +} + static void adm8211_configure_filter(struct ieee80211_hw *dev, unsigned int changed_flags, unsigned int *total_flags, - int mc_count, struct dev_mc_list *mclist) + u64 multicast) { static const u8 bcast[ETH_ALEN] = { 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF }; struct adm8211_priv *priv = dev->priv; - unsigned int bit_nr, new_flags; + unsigned int new_flags; u32 mc_filter[2]; - int i; + + mc_filter[0] = multicast; + mc_filter[1] = multicast >> 32; new_flags = 0; @@ -1346,23 +1369,13 @@ static void adm8211_configure_filter(struct ieee80211_hw *dev, priv->nar |= ADM8211_NAR_PR; priv->nar &= ~ADM8211_NAR_MM; mc_filter[1] = mc_filter[0] = ~0; - } else if ((*total_flags & FIF_ALLMULTI) || (mc_count > 32)) { + } else if (*total_flags & FIF_ALLMULTI || multicast == ~(0ULL)) { new_flags |= FIF_ALLMULTI; priv->nar &= ~ADM8211_NAR_PR; priv->nar |= ADM8211_NAR_MM; mc_filter[1] = mc_filter[0] = ~0; } else { priv->nar &= ~(ADM8211_NAR_MM | ADM8211_NAR_PR); - mc_filter[1] = mc_filter[0] = 0; - for (i = 0; i < mc_count; i++) { - if (!mclist) - break; - bit_nr = ether_crc(ETH_ALEN, mclist->dmi_addr) >> 26; - - bit_nr &= 0x3F; - mc_filter[bit_nr >> 5] |= 1 << (bit_nr & 31); - mclist = mclist->next; - } } ADM8211_IDLE_RX(); @@ -1757,6 +1770,7 @@ static const struct ieee80211_ops adm8211_ops = { .remove_interface = adm8211_remove_interface, .config = adm8211_config, .bss_info_changed = adm8211_bss_info_changed, + .prepare_multicast = adm8211_prepare_multicast, .configure_filter = adm8211_configure_filter, .get_stats = adm8211_get_stats, .get_tx_stats = adm8211_get_tx_stats, diff --git a/drivers/net/wireless/at76c50x-usb.c b/drivers/net/wireless/at76c50x-usb.c index 7218dbabad3e..a6e19545ac6a 100644 --- a/drivers/net/wireless/at76c50x-usb.c +++ b/drivers/net/wireless/at76c50x-usb.c @@ -1997,15 +1997,14 @@ static void at76_bss_info_changed(struct ieee80211_hw *hw, /* must be atomic */ static void at76_configure_filter(struct ieee80211_hw *hw, unsigned int changed_flags, - unsigned int *total_flags, int mc_count, - struct dev_addr_list *mc_list) + unsigned int *total_flags, u64 multicast) { struct at76_priv *priv = hw->priv; int flags; at76_dbg(DBG_MAC80211, "%s(): changed_flags=0x%08x " - "total_flags=0x%08x mc_count=%d", - __func__, changed_flags, *total_flags, mc_count); + "total_flags=0x%08x", + __func__, changed_flags, *total_flags); flags = changed_flags & AT76_SUPPORTED_FILTERS; *total_flags = AT76_SUPPORTED_FILTERS; diff --git a/drivers/net/wireless/ath/ar9170/main.c b/drivers/net/wireless/ath/ar9170/main.c index ea8c9419336d..6a9462e4fd87 100644 --- a/drivers/net/wireless/ath/ar9170/main.c +++ b/drivers/net/wireless/ath/ar9170/main.c @@ -2100,10 +2100,29 @@ unlock: mutex_unlock(&ar->mutex); } +static u64 ar9170_op_prepare_multicast(struct ieee80211_hw *hw, int mc_count, + struct dev_addr_list *mclist) +{ + u64 mchash; + int i; + + /* always get broadcast frames */ + mchash = 1ULL << (0xff >> 2); + + for (i = 0; i < mc_count; i++) { + if (WARN_ON(!mclist)) + break; + mchash |= 1ULL << (mclist->dmi_addr[5] >> 2); + mclist = mclist->next; + } + + return mchash; +} + static void ar9170_op_configure_filter(struct ieee80211_hw *hw, unsigned int changed_flags, unsigned int *new_flags, - int mc_count, struct dev_mc_list *mclist) + u64 multicast) { struct ar9170 *ar = hw->priv; @@ -2116,24 +2135,11 @@ static void ar9170_op_configure_filter(struct ieee80211_hw *hw, * then checking the error flags, later. */ - if (changed_flags & FIF_ALLMULTI) { - if (*new_flags & FIF_ALLMULTI) { - ar->want_mc_hash = ~0ULL; - } else { - u64 mchash; - int i; - - /* always get broadcast frames */ - mchash = 1ULL << (0xff >> 2); + if (changed_flags & FIF_ALLMULTI && *new_flags & FIF_ALLMULTI) + multicast = ~0ULL; - for (i = 0; i < mc_count; i++) { - if (WARN_ON(!mclist)) - break; - mchash |= 1ULL << (mclist->dmi_addr[5] >> 2); - mclist = mclist->next; - } - ar->want_mc_hash = mchash; - } + if (multicast != ar->want_mc_hash) { + ar->want_mc_hash = multicast; set_bit(AR9170_FILTER_CHANGED_MULTICAST, &ar->filter_changed); } @@ -2543,6 +2549,7 @@ static const struct ieee80211_ops ar9170_ops = { .add_interface = ar9170_op_add_interface, .remove_interface = ar9170_op_remove_interface, .config = ar9170_op_config, + .prepare_multicast = ar9170_op_prepare_multicast, .configure_filter = ar9170_op_configure_filter, .conf_tx = ar9170_conf_tx, .bss_info_changed = ar9170_op_bss_info_changed, diff --git a/drivers/net/wireless/ath/ath5k/base.c b/drivers/net/wireless/ath/ath5k/base.c index 2b3cf39dd4b1..3951b5b13424 100644 --- a/drivers/net/wireless/ath/ath5k/base.c +++ b/drivers/net/wireless/ath/ath5k/base.c @@ -229,10 +229,12 @@ static int ath5k_add_interface(struct ieee80211_hw *hw, static void ath5k_remove_interface(struct ieee80211_hw *hw, struct ieee80211_if_init_conf *conf); static int ath5k_config(struct ieee80211_hw *hw, u32 changed); +static u64 ath5k_prepare_multicast(struct ieee80211_hw *hw, + int mc_count, struct dev_addr_list *mc_list); static void ath5k_configure_filter(struct ieee80211_hw *hw, unsigned int changed_flags, unsigned int *new_flags, - int mc_count, struct dev_mc_list *mclist); + u64 multicast); static int ath5k_set_key(struct ieee80211_hw *hw, enum set_key_cmd cmd, struct ieee80211_vif *vif, struct ieee80211_sta *sta, @@ -260,6 +262,7 @@ static const struct ieee80211_ops ath5k_hw_ops = { .add_interface = ath5k_add_interface, .remove_interface = ath5k_remove_interface, .config = ath5k_config, + .prepare_multicast = ath5k_prepare_multicast, .configure_filter = ath5k_configure_filter, .set_key = ath5k_set_key, .get_stats = ath5k_get_stats, @@ -2853,6 +2856,37 @@ unlock: return ret; } +static u64 ath5k_prepare_multicast(struct ieee80211_hw *hw, + int mc_count, struct dev_addr_list *mclist) +{ + u32 mfilt[2], val; + int i; + u8 pos; + + mfilt[0] = 0; + mfilt[1] = 1; + + for (i = 0; i < mc_count; i++) { + if (!mclist) + break; + /* calculate XOR of eight 6-bit values */ + val = get_unaligned_le32(mclist->dmi_addr + 0); + pos = (val >> 18) ^ (val >> 12) ^ (val >> 6) ^ val; + val = get_unaligned_le32(mclist->dmi_addr + 3); + pos ^= (val >> 18) ^ (val >> 12) ^ (val >> 6) ^ val; + pos &= 0x3f; + mfilt[pos / 32] |= (1 << (pos % 32)); + /* XXX: we might be able to just do this instead, + * but not sure, needs testing, if we do use this we'd + * neet to inform below to not reset the mcast */ + /* ath5k_hw_set_mcast_filterindex(ah, + * mclist->dmi_addr[5]); */ + mclist = mclist->next; + } + + return ((u64)(mfilt[1]) << 32) | mfilt[0]; +} + #define SUPPORTED_FIF_FLAGS \ FIF_PROMISC_IN_BSS | FIF_ALLMULTI | FIF_FCSFAIL | \ FIF_PLCPFAIL | FIF_CONTROL | FIF_OTHER_BSS | \ @@ -2878,16 +2912,14 @@ unlock: static void ath5k_configure_filter(struct ieee80211_hw *hw, unsigned int changed_flags, unsigned int *new_flags, - int mc_count, struct dev_mc_list *mclist) + u64 multicast) { struct ath5k_softc *sc = hw->priv; struct ath5k_hw *ah = sc->ah; - u32 mfilt[2], val, rfilt; - u8 pos; - int i; + u32 mfilt[2], rfilt; - mfilt[0] = 0; - mfilt[1] = 0; + mfilt[0] = multicast; + mfilt[1] = multicast >> 32; /* Only deal with supported flags */ changed_flags &= SUPPORTED_FIF_FLAGS; @@ -2913,24 +2945,6 @@ static void ath5k_configure_filter(struct ieee80211_hw *hw, if (*new_flags & FIF_ALLMULTI) { mfilt[0] = ~0; mfilt[1] = ~0; - } else { - for (i = 0; i < mc_count; i++) { - if (!mclist) - break; - /* calculate XOR of eight 6-bit values */ - val = get_unaligned_le32(mclist->dmi_addr + 0); - pos = (val >> 18) ^ (val >> 12) ^ (val >> 6) ^ val; - val = get_unaligned_le32(mclist->dmi_addr + 3); - pos ^= (val >> 18) ^ (val >> 12) ^ (val >> 6) ^ val; - pos &= 0x3f; - mfilt[pos / 32] |= (1 << (pos % 32)); - /* XXX: we might be able to just do this instead, - * but not sure, needs testing, if we do use this we'd - * neet to inform below to not reset the mcast */ - /* ath5k_hw_set_mcast_filterindex(ah, - * mclist->dmi_addr[5]); */ - mclist = mclist->next; - } } /* This is the best we can do */ diff --git a/drivers/net/wireless/ath/ath9k/main.c b/drivers/net/wireless/ath/ath9k/main.c index 3e09b9ac165b..2f9c149fd481 100644 --- a/drivers/net/wireless/ath/ath9k/main.c +++ b/drivers/net/wireless/ath/ath9k/main.c @@ -2394,8 +2394,7 @@ skip_chan_change: static void ath9k_configure_filter(struct ieee80211_hw *hw, unsigned int changed_flags, unsigned int *total_flags, - int mc_count, - struct dev_mc_list *mclist) + u64 multicast) { struct ath_wiphy *aphy = hw->priv; struct ath_softc *sc = aphy->sc; diff --git a/drivers/net/wireless/b43/main.c b/drivers/net/wireless/b43/main.c index c5bece090420..78ddbc7f836b 100644 --- a/drivers/net/wireless/b43/main.c +++ b/drivers/net/wireless/b43/main.c @@ -3679,7 +3679,7 @@ out_unlock: static void b43_op_configure_filter(struct ieee80211_hw *hw, unsigned int changed, unsigned int *fflags, - int mc_count, struct dev_addr_list *mc_list) + u64 multicast) { struct b43_wl *wl = hw_to_b43_wl(hw); struct b43_wldev *dev = wl->current_dev; diff --git a/drivers/net/wireless/b43legacy/main.c b/drivers/net/wireless/b43legacy/main.c index b1435594921a..b166a6f9f055 100644 --- a/drivers/net/wireless/b43legacy/main.c +++ b/drivers/net/wireless/b43legacy/main.c @@ -2836,9 +2836,7 @@ static void b43legacy_op_bss_info_changed(struct ieee80211_hw *hw, static void b43legacy_op_configure_filter(struct ieee80211_hw *hw, unsigned int changed, - unsigned int *fflags, - int mc_count, - struct dev_addr_list *mc_list) + unsigned int *fflags,u64 multicast) { struct b43legacy_wl *wl = hw_to_b43legacy_wl(hw); struct b43legacy_wldev *dev = wl->current_dev; diff --git a/drivers/net/wireless/iwlwifi/iwl-core.c b/drivers/net/wireless/iwlwifi/iwl-core.c index c0efa6623c09..f1f6dabd8fbd 100644 --- a/drivers/net/wireless/iwlwifi/iwl-core.c +++ b/drivers/net/wireless/iwlwifi/iwl-core.c @@ -1514,7 +1514,7 @@ EXPORT_SYMBOL(iwl_irq_handle_error); void iwl_configure_filter(struct ieee80211_hw *hw, unsigned int changed_flags, unsigned int *total_flags, - int mc_count, struct dev_addr_list *mc_list) + u64 multicast) { struct iwl_priv *priv = hw->priv; __le32 *filter_flags = &priv->staging_rxon.filter_flags; diff --git a/drivers/net/wireless/iwlwifi/iwl-core.h b/drivers/net/wireless/iwlwifi/iwl-core.h index 4ca025a34daf..62d90364b61d 100644 --- a/drivers/net/wireless/iwlwifi/iwl-core.h +++ b/drivers/net/wireless/iwlwifi/iwl-core.h @@ -282,8 +282,7 @@ int iwl_set_decrypted_flag(struct iwl_priv *priv, void iwl_irq_handle_error(struct iwl_priv *priv); void iwl_configure_filter(struct ieee80211_hw *hw, unsigned int changed_flags, - unsigned int *total_flags, - int mc_count, struct dev_addr_list *mc_list); + unsigned int *total_flags, u64 multicast); int iwl_hw_nic_init(struct iwl_priv *priv); int iwl_setup_mac(struct iwl_priv *priv); int iwl_set_hw_params(struct iwl_priv *priv); diff --git a/drivers/net/wireless/libertas_tf/main.c b/drivers/net/wireless/libertas_tf/main.c index 4872345a2f61..019431d2f8a9 100644 --- a/drivers/net/wireless/libertas_tf/main.c +++ b/drivers/net/wireless/libertas_tf/main.c @@ -366,15 +366,35 @@ static int lbtf_op_config(struct ieee80211_hw *hw, u32 changed) return 0; } +static u64 lbtf_op_prepare_multicast(struct ieee80211_hw *hw, + int mc_count, struct dev_addr_list *mclist) +{ + struct lbtf_private *priv = hw->priv; + int i; + + if (!mc_count || mc_count > MRVDRV_MAX_MULTICAST_LIST_SIZE) + return mc_count; + + priv->nr_of_multicastmacaddr = mc_count; + for (i = 0; i < mc_count; i++) { + if (!mclist) + break; + memcpy(&priv->multicastlist[i], mclist->da_addr, + ETH_ALEN); + mclist = mclist->next; + } + + return mc_count; +} + #define SUPPORTED_FIF_FLAGS (FIF_PROMISC_IN_BSS | FIF_ALLMULTI) static void lbtf_op_configure_filter(struct ieee80211_hw *hw, unsigned int changed_flags, unsigned int *new_flags, - int mc_count, struct dev_mc_list *mclist) + u64 multicast) { struct lbtf_private *priv = hw->priv; int old_mac_control = priv->mac_control; - int i; changed_flags &= SUPPORTED_FIF_FLAGS; *new_flags &= SUPPORTED_FIF_FLAGS; @@ -386,20 +406,12 @@ static void lbtf_op_configure_filter(struct ieee80211_hw *hw, else priv->mac_control &= ~CMD_ACT_MAC_PROMISCUOUS_ENABLE; if (*new_flags & (FIF_ALLMULTI) || - mc_count > MRVDRV_MAX_MULTICAST_LIST_SIZE) { + multicast > MRVDRV_MAX_MULTICAST_LIST_SIZE) { priv->mac_control |= CMD_ACT_MAC_ALL_MULTICAST_ENABLE; priv->mac_control &= ~CMD_ACT_MAC_MULTICAST_ENABLE; - } else if (mc_count) { + } else if (multicast) { priv->mac_control |= CMD_ACT_MAC_MULTICAST_ENABLE; priv->mac_control &= ~CMD_ACT_MAC_ALL_MULTICAST_ENABLE; - priv->nr_of_multicastmacaddr = mc_count; - for (i = 0; i < mc_count; i++) { - if (!mclist) - break; - memcpy(&priv->multicastlist[i], mclist->da_addr, - ETH_ALEN); - mclist = mclist->next; - } lbtf_cmd_set_mac_multicast_addr(priv); } else { priv->mac_control &= ~(CMD_ACT_MAC_MULTICAST_ENABLE | @@ -461,6 +473,7 @@ static const struct ieee80211_ops lbtf_ops = { .add_interface = lbtf_op_add_interface, .remove_interface = lbtf_op_remove_interface, .config = lbtf_op_config, + .prepare_multicast = lbtf_op_prepare_multicast, .configure_filter = lbtf_op_configure_filter, .bss_info_changed = lbtf_op_bss_info_changed, }; diff --git a/drivers/net/wireless/mac80211_hwsim.c b/drivers/net/wireless/mac80211_hwsim.c index 930f5c7da4a6..6f6cd43592c8 100644 --- a/drivers/net/wireless/mac80211_hwsim.c +++ b/drivers/net/wireless/mac80211_hwsim.c @@ -582,9 +582,7 @@ static int mac80211_hwsim_config(struct ieee80211_hw *hw, u32 changed) static void mac80211_hwsim_configure_filter(struct ieee80211_hw *hw, unsigned int changed_flags, - unsigned int *total_flags, - int mc_count, - struct dev_addr_list *mc_list) + unsigned int *total_flags,u64 multicast) { struct mac80211_hwsim_data *data = hw->priv; diff --git a/drivers/net/wireless/mwl8k.c b/drivers/net/wireless/mwl8k.c index 8a6d3afe4122..f84387083e73 100644 --- a/drivers/net/wireless/mwl8k.c +++ b/drivers/net/wireless/mwl8k.c @@ -3251,31 +3251,50 @@ mwl8k_configure_filter_exit: return rc; } +static u64 mwl8k_prepare_multicast(struct ieee80211_hw *hw, + int mc_count, struct dev_addr_list *mclist) +{ + struct mwl8k_configure_filter_worker *worker; + + worker = kzalloc(sizeof(*worker), GFP_ATOMIC); + + if (!worker) + return 0; + + /* + * XXX: This is _HORRIBLY_ broken!! + * + * No locking, the mclist pointer might be invalid as soon as this + * function returns, something in the list might be invalidated + * once we get to the worker, etc... + */ + worker->mc_count = mc_count; + worker->mclist = mclist; + + return (u64)worker; +} + static void mwl8k_configure_filter(struct ieee80211_hw *hw, unsigned int changed_flags, unsigned int *total_flags, - int mc_count, - struct dev_addr_list *mclist) + u64 multicast) { - struct mwl8k_configure_filter_worker *worker; + struct mwl8k_configure_filter_worker *worker = (void *)multicast; struct mwl8k_priv *priv = hw->priv; /* Clear unsupported feature flags */ *total_flags &= MWL8K_SUPPORTED_IF_FLAGS; - if (!(changed_flags & MWL8K_SUPPORTED_IF_FLAGS) && !mc_count) + if (!(changed_flags & MWL8K_SUPPORTED_IF_FLAGS)) return; - worker = kzalloc(sizeof(*worker), GFP_ATOMIC); if (worker == NULL) return; worker->header.options = MWL8K_WQ_QUEUE_ONLY | MWL8K_WQ_TX_WAIT_EMPTY; worker->changed_flags = changed_flags; worker->total_flags = total_flags; - worker->mc_count = mc_count; - worker->mclist = mclist; mwl8k_queue_work(hw, &worker->header, priv->config_wq, mwl8k_configure_filter_wt); @@ -3441,6 +3460,7 @@ static const struct ieee80211_ops mwl8k_ops = { .remove_interface = mwl8k_remove_interface, .config = mwl8k_config, .bss_info_changed = mwl8k_bss_info_changed, + .prepare_multicast = mwl8k_prepare_multicast, .configure_filter = mwl8k_configure_filter, .set_rts_threshold = mwl8k_set_rts_threshold, .conf_tx = mwl8k_conf_tx, diff --git a/drivers/net/wireless/p54/main.c b/drivers/net/wireless/p54/main.c index 77203e346cda..4d486bf9f725 100644 --- a/drivers/net/wireless/p54/main.c +++ b/drivers/net/wireless/p54/main.c @@ -302,7 +302,7 @@ out: static void p54_configure_filter(struct ieee80211_hw *dev, unsigned int changed_flags, unsigned int *total_flags, - int mc_count, struct dev_mc_list *mclist) + u64 multicast) { struct p54_common *priv = dev->priv; diff --git a/drivers/net/wireless/rt2x00/rt2x00.h b/drivers/net/wireless/rt2x00/rt2x00.h index 99e89596cef6..39d7d9baafdd 100644 --- a/drivers/net/wireless/rt2x00/rt2x00.h +++ b/drivers/net/wireless/rt2x00/rt2x00.h @@ -978,7 +978,7 @@ int rt2x00mac_config(struct ieee80211_hw *hw, u32 changed); void rt2x00mac_configure_filter(struct ieee80211_hw *hw, unsigned int changed_flags, unsigned int *total_flags, - int mc_count, struct dev_addr_list *mc_list); + u64 multicast); int rt2x00mac_set_tim(struct ieee80211_hw *hw, struct ieee80211_sta *sta, bool set); #ifdef CONFIG_RT2X00_LIB_CRYPTO diff --git a/drivers/net/wireless/rt2x00/rt2x00mac.c b/drivers/net/wireless/rt2x00/rt2x00mac.c index cb7b6d459331..602f12699718 100644 --- a/drivers/net/wireless/rt2x00/rt2x00mac.c +++ b/drivers/net/wireless/rt2x00/rt2x00mac.c @@ -379,7 +379,7 @@ EXPORT_SYMBOL_GPL(rt2x00mac_config); void rt2x00mac_configure_filter(struct ieee80211_hw *hw, unsigned int changed_flags, unsigned int *total_flags, - int mc_count, struct dev_addr_list *mc_list) + u64 multicast) { struct rt2x00_dev *rt2x00dev = hw->priv; diff --git a/drivers/net/wireless/rtl818x/rtl8180_dev.c b/drivers/net/wireless/rtl818x/rtl8180_dev.c index 09f46abc730a..16429c49139c 100644 --- a/drivers/net/wireless/rtl818x/rtl8180_dev.c +++ b/drivers/net/wireless/rtl818x/rtl8180_dev.c @@ -728,10 +728,16 @@ static void rtl8180_bss_info_changed(struct ieee80211_hw *dev, priv->rf->conf_erp(dev, info); } +static u64 rtl8180_prepare_multicast(struct ieee80211_hw *dev, int mc_count, + struct dev_addr_list *mc_list) +{ + return mc_count; +} + static void rtl8180_configure_filter(struct ieee80211_hw *dev, unsigned int changed_flags, unsigned int *total_flags, - int mc_count, struct dev_addr_list *mclist) + u64 multicast) { struct rtl8180_priv *priv = dev->priv; @@ -741,7 +747,7 @@ static void rtl8180_configure_filter(struct ieee80211_hw *dev, priv->rx_conf ^= RTL818X_RX_CONF_CTRL; if (changed_flags & FIF_OTHER_BSS) priv->rx_conf ^= RTL818X_RX_CONF_MONITOR; - if (*total_flags & FIF_ALLMULTI || mc_count > 0) + if (*total_flags & FIF_ALLMULTI || multicast > 0) priv->rx_conf |= RTL818X_RX_CONF_MULTICAST; else priv->rx_conf &= ~RTL818X_RX_CONF_MULTICAST; @@ -768,6 +774,7 @@ static const struct ieee80211_ops rtl8180_ops = { .remove_interface = rtl8180_remove_interface, .config = rtl8180_config, .bss_info_changed = rtl8180_bss_info_changed, + .prepare_multicast = rtl8180_prepare_multicast, .configure_filter = rtl8180_configure_filter, }; diff --git a/drivers/net/wireless/rtl818x/rtl8187_dev.c b/drivers/net/wireless/rtl818x/rtl8187_dev.c index 53f57dc52226..90f38357393c 100644 --- a/drivers/net/wireless/rtl818x/rtl8187_dev.c +++ b/drivers/net/wireless/rtl818x/rtl8187_dev.c @@ -1192,10 +1192,16 @@ static void rtl8187_bss_info_changed(struct ieee80211_hw *dev, info->use_short_preamble); } +static u64 rtl8187_prepare_multicast(struct ieee80211_hw *dev, + int mc_count, struct dev_addr_list *mc_list) +{ + return mc_count; +} + static void rtl8187_configure_filter(struct ieee80211_hw *dev, unsigned int changed_flags, unsigned int *total_flags, - int mc_count, struct dev_addr_list *mclist) + u64 multicast) { struct rtl8187_priv *priv = dev->priv; @@ -1205,7 +1211,7 @@ static void rtl8187_configure_filter(struct ieee80211_hw *dev, priv->rx_conf ^= RTL818X_RX_CONF_CTRL; if (changed_flags & FIF_OTHER_BSS) priv->rx_conf ^= RTL818X_RX_CONF_MONITOR; - if (*total_flags & FIF_ALLMULTI || mc_count > 0) + if (*total_flags & FIF_ALLMULTI || multicast > 0) priv->rx_conf |= RTL818X_RX_CONF_MULTICAST; else priv->rx_conf &= ~RTL818X_RX_CONF_MULTICAST; @@ -1268,6 +1274,7 @@ static const struct ieee80211_ops rtl8187_ops = { .remove_interface = rtl8187_remove_interface, .config = rtl8187_config, .bss_info_changed = rtl8187_bss_info_changed, + .prepare_multicast = rtl8187_prepare_multicast, .configure_filter = rtl8187_configure_filter, .conf_tx = rtl8187_conf_tx }; diff --git a/drivers/net/wireless/wl12xx/wl1251_main.c b/drivers/net/wireless/wl12xx/wl1251_main.c index 7148934e464d..5809ef5b18f8 100644 --- a/drivers/net/wireless/wl12xx/wl1251_main.c +++ b/drivers/net/wireless/wl12xx/wl1251_main.c @@ -652,9 +652,7 @@ out: static void wl1251_op_configure_filter(struct ieee80211_hw *hw, unsigned int changed, - unsigned int *total, - int mc_count, - struct dev_addr_list *mc_list) + unsigned int *total,u64 multicast) { struct wl1251 *wl = hw->priv; diff --git a/drivers/net/wireless/wl12xx/wl1271_main.c b/drivers/net/wireless/wl12xx/wl1271_main.c index 4102d590b798..754be8179307 100644 --- a/drivers/net/wireless/wl12xx/wl1271_main.c +++ b/drivers/net/wireless/wl12xx/wl1271_main.c @@ -793,9 +793,7 @@ out: static void wl1271_op_configure_filter(struct ieee80211_hw *hw, unsigned int changed, - unsigned int *total, - int mc_count, - struct dev_addr_list *mc_list) + unsigned int *total,u64 multicast) { struct wl1271 *wl = hw->priv; diff --git a/drivers/net/wireless/zd1211rw/zd_mac.c b/drivers/net/wireless/zd1211rw/zd_mac.c index 55b7fbdc85dc..6d666359a42f 100644 --- a/drivers/net/wireless/zd1211rw/zd_mac.c +++ b/drivers/net/wireless/zd1211rw/zd_mac.c @@ -796,18 +796,40 @@ static void set_rx_filter_handler(struct work_struct *work) dev_err(zd_mac_dev(mac), "set_rx_filter_handler error %d\n", r); } +static u64 zd_op_prepare_multicast(struct ieee80211_hw *hw, + int mc_count, struct dev_addr_list *mclist) +{ + struct zd_mac *mac = zd_hw_mac(hw); + struct zd_mc_hash hash; + int i; + + zd_mc_clear(&hash); + + for (i = 0; i < mc_count; i++) { + if (!mclist) + break; + dev_dbg_f(zd_mac_dev(mac), "mc addr %pM\n", mclist->dmi_addr); + zd_mc_add_addr(&hash, mclist->dmi_addr); + mclist = mclist->next; + } + + return hash.low | ((u64)hash.high << 32); +} + #define SUPPORTED_FIF_FLAGS \ (FIF_PROMISC_IN_BSS | FIF_ALLMULTI | FIF_FCSFAIL | FIF_CONTROL | \ FIF_OTHER_BSS | FIF_BCN_PRBRESP_PROMISC) static void zd_op_configure_filter(struct ieee80211_hw *hw, unsigned int changed_flags, unsigned int *new_flags, - int mc_count, struct dev_mc_list *mclist) + u64 multicast) { - struct zd_mc_hash hash; + struct zd_mc_hash hash = { + .low = multicast, + .high = multicast >> 32, + }; struct zd_mac *mac = zd_hw_mac(hw); unsigned long flags; - int i; /* Only deal with supported flags */ changed_flags &= SUPPORTED_FIF_FLAGS; @@ -819,25 +841,16 @@ static void zd_op_configure_filter(struct ieee80211_hw *hw, if (!changed_flags) return; - if (*new_flags & (FIF_PROMISC_IN_BSS | FIF_ALLMULTI)) { + if (*new_flags & (FIF_PROMISC_IN_BSS | FIF_ALLMULTI)) zd_mc_add_all(&hash); - } else { - zd_mc_clear(&hash); - for (i = 0; i < mc_count; i++) { - if (!mclist) - break; - dev_dbg_f(zd_mac_dev(mac), "mc addr %pM\n", - mclist->dmi_addr); - zd_mc_add_addr(&hash, mclist->dmi_addr); - mclist = mclist->next; - } - } spin_lock_irqsave(&mac->lock, flags); mac->pass_failed_fcs = !!(*new_flags & FIF_FCSFAIL); mac->pass_ctrl = !!(*new_flags & FIF_CONTROL); mac->multicast_hash = hash; spin_unlock_irqrestore(&mac->lock, flags); + + /* XXX: these can be called here now, can sleep now! */ queue_work(zd_workqueue, &mac->set_multicast_hash_work); if (changed_flags & FIF_CONTROL) @@ -940,6 +953,7 @@ static const struct ieee80211_ops zd_ops = { .add_interface = zd_op_add_interface, .remove_interface = zd_op_remove_interface, .config = zd_op_config, + .prepare_multicast = zd_op_prepare_multicast, .configure_filter = zd_op_configure_filter, .bss_info_changed = zd_op_bss_info_changed, .get_tsf = zd_op_get_tsf, diff --git a/include/net/mac80211.h b/include/net/mac80211.h index 76d43e12cc29..bc865206e5f1 100644 --- a/include/net/mac80211.h +++ b/include/net/mac80211.h @@ -1219,10 +1219,13 @@ ieee80211_get_alt_retry_rate(const struct ieee80211_hw *hw, * the driver's configure_filter() function which frames should be * passed to mac80211 and which should be filtered out. * - * The configure_filter() callback is invoked with the parameters - * @mc_count and @mc_list for the combined multicast address list - * of all virtual interfaces, @changed_flags telling which flags - * were changed and @total_flags with the new flag states. + * Before configure_filter() is invoked, the prepare_multicast() + * callback is invoked with the parameters @mc_count and @mc_list + * for the combined multicast address list of all virtual interfaces. + * It's use is optional, and it returns a u64 that is passed to + * configure_filter(). Additionally, configure_filter() has the + * arguments @changed_flags telling which flags were changed and + * @total_flags with the new flag states. * * If your device has no multicast address filters your driver will * need to check both the %FIF_ALLMULTI flag and the @mc_count @@ -1375,9 +1378,13 @@ enum ieee80211_ampdu_mlme_action { * for association indication. The @changed parameter indicates which * of the bss parameters has changed when a call is made. * + * @prepare_multicast: Prepare for multicast filter configuration. + * This callback is optional, and its return value is passed + * to configure_filter(). This callback must be atomic. + * * @configure_filter: Configure the device's RX filter. * See the section "Frame filtering" for more information. - * This callback must be implemented and atomic. + * This callback must be implemented. * * @set_tim: Set TIM bit. mac80211 calls this function when a TIM bit * must be set or cleared for a given STA. Must be atomic. @@ -1479,10 +1486,12 @@ struct ieee80211_ops { struct ieee80211_vif *vif, struct ieee80211_bss_conf *info, u32 changed); + u64 (*prepare_multicast)(struct ieee80211_hw *hw, + int mc_count, struct dev_addr_list *mc_list); void (*configure_filter)(struct ieee80211_hw *hw, unsigned int changed_flags, unsigned int *total_flags, - int mc_count, struct dev_addr_list *mc_list); + u64 multicast); int (*set_tim)(struct ieee80211_hw *hw, struct ieee80211_sta *sta, bool set); int (*set_key)(struct ieee80211_hw *hw, enum set_key_cmd cmd, diff --git a/net/mac80211/driver-ops.h b/net/mac80211/driver-ops.h index 4100c361a99d..d231c9323ad1 100644 --- a/net/mac80211/driver-ops.h +++ b/net/mac80211/driver-ops.h @@ -55,16 +55,32 @@ static inline void drv_bss_info_changed(struct ieee80211_local *local, trace_drv_bss_info_changed(local, vif, info, changed); } +static inline u64 drv_prepare_multicast(struct ieee80211_local *local, + int mc_count, + struct dev_addr_list *mc_list) +{ + u64 ret = 0; + + if (local->ops->prepare_multicast) + ret = local->ops->prepare_multicast(&local->hw, mc_count, + mc_list); + + trace_drv_prepare_multicast(local, mc_count, ret); + + return ret; +} + static inline void drv_configure_filter(struct ieee80211_local *local, unsigned int changed_flags, unsigned int *total_flags, - int mc_count, - struct dev_addr_list *mc_list) + u64 multicast) { + might_sleep(); + local->ops->configure_filter(&local->hw, changed_flags, total_flags, - mc_count, mc_list); + multicast); trace_drv_configure_filter(local, changed_flags, total_flags, - mc_count); + multicast); } static inline int drv_set_tim(struct ieee80211_local *local, diff --git a/net/mac80211/driver-trace.h b/net/mac80211/driver-trace.h index 5a10da2d70fd..37b9051afcf3 100644 --- a/net/mac80211/driver-trace.h +++ b/net/mac80211/driver-trace.h @@ -191,31 +191,55 @@ TRACE_EVENT(drv_bss_info_changed, ) ); +TRACE_EVENT(drv_prepare_multicast, + TP_PROTO(struct ieee80211_local *local, int mc_count, u64 ret), + + TP_ARGS(local, mc_count, ret), + + TP_STRUCT__entry( + LOCAL_ENTRY + __field(int, mc_count) + __field(u64, ret) + ), + + TP_fast_assign( + LOCAL_ASSIGN; + __entry->mc_count = mc_count; + __entry->ret = ret; + ), + + TP_printk( + LOCAL_PR_FMT " prepare mc (%d): %llx", + LOCAL_PR_ARG, __entry->mc_count, + (unsigned long long) __entry->ret + ) +); + TRACE_EVENT(drv_configure_filter, TP_PROTO(struct ieee80211_local *local, unsigned int changed_flags, unsigned int *total_flags, - int mc_count), + u64 multicast), - TP_ARGS(local, changed_flags, total_flags, mc_count), + TP_ARGS(local, changed_flags, total_flags, multicast), TP_STRUCT__entry( LOCAL_ENTRY __field(unsigned int, changed) __field(unsigned int, total) - __field(int, mc) + __field(u64, multicast) ), TP_fast_assign( LOCAL_ASSIGN; __entry->changed = changed_flags; __entry->total = *total_flags; - __entry->mc = mc_count; + __entry->multicast = multicast; ), TP_printk( - LOCAL_PR_FMT " changed:%#x total:%#x mc:%d", - LOCAL_PR_ARG, __entry->changed, __entry->total, __entry->mc + LOCAL_PR_FMT " changed:%#x total:%#x", + LOCAL_PR_ARG, __entry->changed, __entry->total ) ); diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h index a6abc7dfd903..a07f01736a91 100644 --- a/net/mac80211/ieee80211_i.h +++ b/net/mac80211/ieee80211_i.h @@ -636,6 +636,9 @@ struct ieee80211_local { /* protects the aggregated multicast list and filter calls */ spinlock_t filter_lock; + /* used for uploading changed mc list */ + struct work_struct reconfig_filter; + /* aggregated multicast list */ struct dev_addr_list *mc_list; int mc_count; diff --git a/net/mac80211/iface.c b/net/mac80211/iface.c index e8fb03b91a44..b161301056df 100644 --- a/net/mac80211/iface.c +++ b/net/mac80211/iface.c @@ -227,9 +227,7 @@ static int ieee80211_open(struct net_device *dev) if (sdata->u.mntr_flags & MONITOR_FLAG_OTHER_BSS) local->fif_other_bss++; - spin_lock_bh(&local->filter_lock); ieee80211_configure_filter(local); - spin_unlock_bh(&local->filter_lock); break; default: conf.vif = &sdata->vif; @@ -241,17 +239,13 @@ static int ieee80211_open(struct net_device *dev) if (ieee80211_vif_is_mesh(&sdata->vif)) { local->fif_other_bss++; - spin_lock_bh(&local->filter_lock); ieee80211_configure_filter(local); - spin_unlock_bh(&local->filter_lock); ieee80211_start_mesh(sdata); } else if (sdata->vif.type == NL80211_IFTYPE_AP) { local->fif_pspoll++; - spin_lock_bh(&local->filter_lock); ieee80211_configure_filter(local); - spin_unlock_bh(&local->filter_lock); } changed |= ieee80211_reset_erp_info(sdata); @@ -404,10 +398,11 @@ static int ieee80211_stop(struct net_device *dev) spin_lock_bh(&local->filter_lock); __dev_addr_unsync(&local->mc_list, &local->mc_count, &dev->mc_list, &dev->mc_count); - ieee80211_configure_filter(local); spin_unlock_bh(&local->filter_lock); netif_addr_unlock_bh(dev); + ieee80211_configure_filter(local); + del_timer_sync(&local->dynamic_ps_timer); cancel_work_sync(&local->dynamic_ps_enable_work); @@ -458,9 +453,7 @@ static int ieee80211_stop(struct net_device *dev) if (sdata->u.mntr_flags & MONITOR_FLAG_OTHER_BSS) local->fif_other_bss--; - spin_lock_bh(&local->filter_lock); ieee80211_configure_filter(local); - spin_unlock_bh(&local->filter_lock); break; case NL80211_IFTYPE_STATION: del_timer_sync(&sdata->u.mgd.chswitch_timer); @@ -503,9 +496,7 @@ static int ieee80211_stop(struct net_device *dev) local->fif_other_bss--; atomic_dec(&local->iff_allmultis); - spin_lock_bh(&local->filter_lock); ieee80211_configure_filter(local); - spin_unlock_bh(&local->filter_lock); ieee80211_stop_mesh(sdata); } @@ -622,8 +613,8 @@ static void ieee80211_set_multicast_list(struct net_device *dev) spin_lock_bh(&local->filter_lock); __dev_addr_sync(&local->mc_list, &local->mc_count, &dev->mc_list, &dev->mc_count); - ieee80211_configure_filter(local); spin_unlock_bh(&local->filter_lock); + ieee80211_queue_work(&local->hw, &local->reconfig_filter); } /* diff --git a/net/mac80211/main.c b/net/mac80211/main.c index b03fd84777fa..05f923575fee 100644 --- a/net/mac80211/main.c +++ b/net/mac80211/main.c @@ -50,9 +50,9 @@ struct ieee80211_tx_status_rtap_hdr { } __attribute__ ((packed)); -/* must be called under mdev tx lock */ void ieee80211_configure_filter(struct ieee80211_local *local) { + u64 mc; unsigned int changed_flags; unsigned int new_flags = 0; @@ -62,7 +62,7 @@ void ieee80211_configure_filter(struct ieee80211_local *local) if (atomic_read(&local->iff_allmultis)) new_flags |= FIF_ALLMULTI; - if (local->monitors) + if (local->monitors || local->scanning) new_flags |= FIF_BCN_PRBRESP_PROMISC; if (local->fif_fcsfail) @@ -80,20 +80,30 @@ void ieee80211_configure_filter(struct ieee80211_local *local) if (local->fif_pspoll) new_flags |= FIF_PSPOLL; + spin_lock_bh(&local->filter_lock); changed_flags = local->filter_flags ^ new_flags; + mc = drv_prepare_multicast(local, local->mc_count, local->mc_list); + spin_unlock_bh(&local->filter_lock); + /* be a bit nasty */ new_flags |= (1<<31); - drv_configure_filter(local, changed_flags, &new_flags, - local->mc_count, - local->mc_list); + drv_configure_filter(local, changed_flags, &new_flags, mc); WARN_ON(new_flags & (1<<31)); local->filter_flags = new_flags & ~(1<<31); } +static void ieee80211_reconfig_filter(struct work_struct *work) +{ + struct ieee80211_local *local = + container_of(work, struct ieee80211_local, reconfig_filter); + + ieee80211_configure_filter(local); +} + int ieee80211_hw_config(struct ieee80211_local *local, u32 changed) { struct ieee80211_channel *chan, *scan_chan; @@ -692,6 +702,8 @@ struct ieee80211_hw *ieee80211_alloc_hw(size_t priv_data_len, INIT_WORK(&local->restart_work, ieee80211_restart_work); + INIT_WORK(&local->reconfig_filter, ieee80211_reconfig_filter); + INIT_WORK(&local->dynamic_ps_enable_work, ieee80211_dynamic_ps_enable_work); INIT_WORK(&local->dynamic_ps_disable_work, @@ -946,6 +958,8 @@ void ieee80211_unregister_hw(struct ieee80211_hw *hw) rtnl_unlock(); + cancel_work_sync(&local->reconfig_filter); + ieee80211_clear_tx_pending(local); sta_info_stop(local); rate_control_deinitialize(local); diff --git a/net/mac80211/scan.c b/net/mac80211/scan.c index e091cbc3434f..1e04be6b9129 100644 --- a/net/mac80211/scan.c +++ b/net/mac80211/scan.c @@ -292,13 +292,7 @@ void ieee80211_scan_completed(struct ieee80211_hw *hw, bool aborted) if (was_hw_scan) goto done; - spin_lock_bh(&local->filter_lock); - local->filter_flags &= ~FIF_BCN_PRBRESP_PROMISC; - drv_configure_filter(local, FIF_BCN_PRBRESP_PROMISC, - &local->filter_flags, - local->mc_count, - local->mc_list); - spin_unlock_bh(&local->filter_lock); + ieee80211_configure_filter(local); drv_sw_scan_complete(local); @@ -376,13 +370,7 @@ static int ieee80211_start_sw_scan(struct ieee80211_local *local) local->next_scan_state = SCAN_DECISION; local->scan_channel_idx = 0; - spin_lock_bh(&local->filter_lock); - local->filter_flags |= FIF_BCN_PRBRESP_PROMISC; - drv_configure_filter(local, FIF_BCN_PRBRESP_PROMISC, - &local->filter_flags, - local->mc_count, - local->mc_list); - spin_unlock_bh(&local->filter_lock); + ieee80211_configure_filter(local); /* TODO: start scan as soon as all nullfunc frames are ACKed */ ieee80211_queue_delayed_work(&local->hw, diff --git a/net/mac80211/util.c b/net/mac80211/util.c index e55d57f559ec..5eb306377c63 100644 --- a/net/mac80211/util.c +++ b/net/mac80211/util.c @@ -1076,9 +1076,7 @@ int ieee80211_reconfig(struct ieee80211_local *local) /* reconfigure hardware */ ieee80211_hw_config(local, ~0); - spin_lock_bh(&local->filter_lock); ieee80211_configure_filter(local); - spin_unlock_bh(&local->filter_lock); /* Finally also reconfigure all the BSS information */ list_for_each_entry(sdata, &local->interfaces, list) { -- cgit v1.2.3 From f424afa17899408cbd267a4c4534ca6fc9d8f71c Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Mon, 17 Aug 2009 16:18:07 +0200 Subject: mac80211: remove deprecated API All but two drivers have now stopped using the two deprecated members radio_enabled and beacon_int, so it's about time to remove them for good. Signed-off-by: Johannes Berg Acked-by: Kalle Valo Signed-off-by: John W. Linville --- drivers/net/wireless/at76c50x-usb.c | 5 ++--- drivers/net/wireless/mwl8k.c | 7 ------- include/net/mac80211.h | 15 --------------- net/mac80211/main.c | 4 ---- 4 files changed, 2 insertions(+), 29 deletions(-) (limited to 'include') diff --git a/drivers/net/wireless/at76c50x-usb.c b/drivers/net/wireless/at76c50x-usb.c index a6e19545ac6a..8e1a55dec351 100644 --- a/drivers/net/wireless/at76c50x-usb.c +++ b/drivers/net/wireless/at76c50x-usb.c @@ -1950,9 +1950,8 @@ static int at76_config(struct ieee80211_hw *hw, u32 changed) { struct at76_priv *priv = hw->priv; - at76_dbg(DBG_MAC80211, "%s(): channel %d radio %d", - __func__, hw->conf.channel->hw_value, - hw->conf.radio_enabled); + at76_dbg(DBG_MAC80211, "%s(): channel %d", + __func__, hw->conf.channel->hw_value); at76_dbg_dump(DBG_MAC80211, priv->bssid, ETH_ALEN, "bssid:"); mutex_lock(&priv->mtx); diff --git a/drivers/net/wireless/mwl8k.c b/drivers/net/wireless/mwl8k.c index f84387083e73..c32e93c8c410 100644 --- a/drivers/net/wireless/mwl8k.c +++ b/drivers/net/wireless/mwl8k.c @@ -3031,13 +3031,6 @@ static int mwl8k_config_wt(struct work_struct *wt) struct mwl8k_priv *priv = hw->priv; int rc = 0; - if (!conf->radio_enabled) { - mwl8k_cmd_802_11_radio_control(hw, MWL8K_RADIO_DISABLE); - priv->current_channel = NULL; - rc = 0; - goto mwl8k_config_exit; - } - if (mwl8k_cmd_802_11_radio_control(hw, MWL8K_RADIO_ENABLE)) { rc = -EINVAL; goto mwl8k_config_exit; diff --git a/include/net/mac80211.h b/include/net/mac80211.h index bc865206e5f1..aac84d7bd46e 100644 --- a/include/net/mac80211.h +++ b/include/net/mac80211.h @@ -583,7 +583,6 @@ enum ieee80211_conf_flags { /** * enum ieee80211_conf_changed - denotes which configuration changed * - * @_IEEE80211_CONF_CHANGE_RADIO_ENABLED: DEPRECATED * @IEEE80211_CONF_CHANGE_LISTEN_INTERVAL: the listen interval changed * @IEEE80211_CONF_CHANGE_RADIOTAP: the radiotap flag changed * @IEEE80211_CONF_CHANGE_PS: the PS flag or dynamic PS timeout changed @@ -593,7 +592,6 @@ enum ieee80211_conf_flags { * @IEEE80211_CONF_CHANGE_IDLE: Idle flag changed */ enum ieee80211_conf_changed { - _IEEE80211_CONF_CHANGE_RADIO_ENABLED = BIT(0), IEEE80211_CONF_CHANGE_LISTEN_INTERVAL = BIT(2), IEEE80211_CONF_CHANGE_RADIOTAP = BIT(3), IEEE80211_CONF_CHANGE_PS = BIT(4), @@ -603,14 +601,6 @@ enum ieee80211_conf_changed { IEEE80211_CONF_CHANGE_IDLE = BIT(8), }; -static inline __deprecated enum ieee80211_conf_changed -__IEEE80211_CONF_CHANGE_RADIO_ENABLED(void) -{ - return _IEEE80211_CONF_CHANGE_RADIO_ENABLED; -} -#define IEEE80211_CONF_CHANGE_RADIO_ENABLED \ - __IEEE80211_CONF_CHANGE_RADIO_ENABLED() - /** * struct ieee80211_conf - configuration of the device * @@ -618,9 +608,6 @@ __IEEE80211_CONF_CHANGE_RADIO_ENABLED(void) * * @flags: configuration flags defined above * - * @radio_enabled: when zero, driver is required to switch off the radio. - * @beacon_int: DEPRECATED, DO NOT USE - * * @listen_interval: listen interval in units of beacon interval * @max_sleep_period: the maximum number of beacon intervals to sleep for * before checking the beacon for a TIM bit (managed mode only); this @@ -644,13 +631,11 @@ __IEEE80211_CONF_CHANGE_RADIO_ENABLED(void) * number of transmissions not the number of retries */ struct ieee80211_conf { - int __deprecated beacon_int; u32 flags; int power_level, dynamic_ps_timeout; int max_sleep_period; u16 listen_interval; - bool __deprecated radio_enabled; u8 long_frame_max_tx_count, short_frame_max_tx_count; diff --git a/net/mac80211/main.c b/net/mac80211/main.c index 05f923575fee..3302df96f8d4 100644 --- a/net/mac80211/main.c +++ b/net/mac80211/main.c @@ -241,9 +241,6 @@ void ieee80211_bss_info_change_notify(struct ieee80211_sub_if_data *sdata, drv_bss_info_changed(local, &sdata->vif, &sdata->vif.bss_conf, changed); - - /* DEPRECATED */ - local->hw.conf.beacon_int = sdata->vif.bss_conf.beacon_int; } u32 ieee80211_reset_erp_info(struct ieee80211_sub_if_data *sdata) @@ -687,7 +684,6 @@ struct ieee80211_hw *ieee80211_alloc_hw(size_t priv_data_len, local->hw.max_rates = 1; local->hw.conf.long_frame_max_tx_count = wiphy->retry_long; local->hw.conf.short_frame_max_tx_count = wiphy->retry_short; - local->hw.conf.radio_enabled = true; local->user_power_level = -1; INIT_LIST_HEAD(&local->interfaces); -- cgit v1.2.3 From ad002395fd230528281083f4be71855ed7e35b04 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Tue, 18 Aug 2009 19:51:57 +0200 Subject: cfg80211: fix dangling scan request checking My patch "cfg80211: fix deadlock" broke the code it was supposed to fix, the scan request checking. But it's not trivial to put it back the way it was, since the original patch had a deadlock. Now do it in a completely new way: queue the check off to a work struct, where we can freely lock. But that has some more complications, like needing to wait for it to be done before the wiphy/rdev can be destroyed, so some code is required to handle that. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- include/net/cfg80211.h | 2 ++ net/wireless/core.c | 76 +++++++++++++++++++++++++++++++++++++++++--------- net/wireless/core.h | 2 ++ 3 files changed, 67 insertions(+), 13 deletions(-) (limited to 'include') diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index 0b146bb2dd14..3d874c620219 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -1325,6 +1325,8 @@ struct wireless_dev { struct mutex mtx; + struct work_struct cleanup_work; + /* currently used for IBSS and SME - might be rearranged later */ u8 ssid[IEEE80211_MAX_SSID_LEN]; u8 ssid_len; diff --git a/net/wireless/core.c b/net/wireless/core.c index 69a185ba9ff1..c150071b6f29 100644 --- a/net/wireless/core.c +++ b/net/wireless/core.c @@ -430,6 +430,8 @@ struct wiphy *wiphy_new(const struct cfg80211_ops *ops, int sizeof_priv) INIT_WORK(&rdev->conn_work, cfg80211_conn_work); INIT_WORK(&rdev->event_work, cfg80211_event_work); + init_waitqueue_head(&rdev->dev_wait); + /* * Initialize wiphy parameters to IEEE 802.11 MIB default values. * Fragmentation and RTS threshold are disabled by default with the @@ -574,7 +576,23 @@ void wiphy_unregister(struct wiphy *wiphy) /* protect the device list */ mutex_lock(&cfg80211_mutex); + wait_event(rdev->dev_wait, ({ + int __count; + mutex_lock(&rdev->devlist_mtx); + __count = rdev->opencount; + mutex_unlock(&rdev->devlist_mtx); + __count == 0;})); + + mutex_lock(&rdev->devlist_mtx); BUG_ON(!list_empty(&rdev->netdev_list)); + mutex_unlock(&rdev->devlist_mtx); + + /* + * First remove the hardware from everywhere, this makes + * it impossible to find from userspace. + */ + cfg80211_debugfs_rdev_del(rdev); + list_del(&rdev->list); /* * Try to grab rdev->mtx. If a command is still in progress, @@ -582,26 +600,18 @@ void wiphy_unregister(struct wiphy *wiphy) * down the device already. We wait for this command to complete * before unlinking the item from the list. * Note: as codified by the BUG_ON above we cannot get here if - * a virtual interface is still associated. Hence, we can only - * get to lock contention here if userspace issues a command - * that identified the hardware by wiphy index. + * a virtual interface is still present. Hence, we can only get + * to lock contention here if userspace issues a command that + * identified the hardware by wiphy index. */ cfg80211_lock_rdev(rdev); - - if (WARN_ON(rdev->scan_req)) { - rdev->scan_req->aborted = true; - ___cfg80211_scan_done(rdev); - } - + /* nothing */ cfg80211_unlock_rdev(rdev); - cfg80211_debugfs_rdev_del(rdev); - /* If this device got a regulatory hint tell core its * free to listen now to a new shiny device regulatory hint */ reg_device_remove(wiphy); - list_del(&rdev->list); cfg80211_rdev_list_generation++; device_del(&rdev->wiphy.dev); debugfs_remove(rdev->wiphy.debugfsdir); @@ -640,6 +650,31 @@ void wiphy_rfkill_set_hw_state(struct wiphy *wiphy, bool blocked) } EXPORT_SYMBOL(wiphy_rfkill_set_hw_state); +static void wdev_cleanup_work(struct work_struct *work) +{ + struct wireless_dev *wdev; + struct cfg80211_registered_device *rdev; + + wdev = container_of(work, struct wireless_dev, cleanup_work); + rdev = wiphy_to_dev(wdev->wiphy); + + cfg80211_lock_rdev(rdev); + + if (WARN_ON(rdev->scan_req && rdev->scan_req->dev == wdev->netdev)) { + rdev->scan_req->aborted = true; + ___cfg80211_scan_done(rdev); + } + + cfg80211_unlock_rdev(rdev); + + mutex_lock(&rdev->devlist_mtx); + rdev->opencount--; + mutex_unlock(&rdev->devlist_mtx); + wake_up(&rdev->dev_wait); + + dev_put(wdev->netdev); +} + static int cfg80211_netdev_notifier_call(struct notifier_block * nb, unsigned long state, void *ndev) @@ -663,6 +698,7 @@ static int cfg80211_netdev_notifier_call(struct notifier_block * nb, * are added with nl80211. */ mutex_init(&wdev->mtx); + INIT_WORK(&wdev->cleanup_work, wdev_cleanup_work); INIT_LIST_HEAD(&wdev->event_list); spin_lock_init(&wdev->event_lock); mutex_lock(&rdev->devlist_mtx); @@ -717,8 +753,22 @@ static int cfg80211_netdev_notifier_call(struct notifier_block * nb, default: break; } + dev_hold(dev); + schedule_work(&wdev->cleanup_work); break; case NETDEV_UP: + /* + * If we have a really quick DOWN/UP succession we may + * have this work still pending ... cancel it and see + * if it was pending, in which case we need to account + * for some of the work it would have done. + */ + if (cancel_work_sync(&wdev->cleanup_work)) { + mutex_lock(&rdev->devlist_mtx); + rdev->opencount--; + mutex_unlock(&rdev->devlist_mtx); + dev_put(dev); + } #ifdef CONFIG_WIRELESS_EXT cfg80211_lock_rdev(rdev); mutex_lock(&rdev->devlist_mtx); @@ -734,6 +784,7 @@ static int cfg80211_netdev_notifier_call(struct notifier_block * nb, break; } wdev_unlock(wdev); + rdev->opencount++; mutex_unlock(&rdev->devlist_mtx); cfg80211_unlock_rdev(rdev); #endif @@ -756,7 +807,6 @@ static int cfg80211_netdev_notifier_call(struct notifier_block * nb, sysfs_remove_link(&dev->dev.kobj, "phy80211"); list_del_init(&wdev->list); rdev->devlist_generation++; - mutex_destroy(&wdev->mtx); #ifdef CONFIG_WIRELESS_EXT kfree(wdev->wext.keys); #endif diff --git a/net/wireless/core.h b/net/wireless/core.h index c603f5286326..f565432ae22f 100644 --- a/net/wireless/core.h +++ b/net/wireless/core.h @@ -50,6 +50,8 @@ struct cfg80211_registered_device { struct mutex devlist_mtx; struct list_head netdev_list; int devlist_generation; + int opencount; /* also protected by devlist_mtx */ + wait_queue_head_t dev_wait; /* BSSes/scanning */ spinlock_t bss_lock; -- cgit v1.2.3 From 4dceef96756b667360741712a8e37490f8458516 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Thu, 20 Aug 2009 17:08:39 -0400 Subject: nfs: fix compile error in rpc_pipefs.h This include is needed for the definition of delayed_work. Signed-off-by: J. Bruce Fields Signed-off-by: Trond Myklebust --- include/linux/sunrpc/rpc_pipe_fs.h | 2 ++ 1 file changed, 2 insertions(+) (limited to 'include') diff --git a/include/linux/sunrpc/rpc_pipe_fs.h b/include/linux/sunrpc/rpc_pipe_fs.h index a92571a34556..cf14db975da0 100644 --- a/include/linux/sunrpc/rpc_pipe_fs.h +++ b/include/linux/sunrpc/rpc_pipe_fs.h @@ -3,6 +3,8 @@ #ifdef __KERNEL__ +#include + struct rpc_pipe_msg { struct list_head list; void *data; -- cgit v1.2.3 From e3b2415e281a97ade36d88404094a90cfea838c0 Mon Sep 17 00:00:00 2001 From: Dave Airlie Date: Fri, 21 Aug 2009 09:47:45 +1000 Subject: drm/radeon/kms: implement the bo busy ioctl properly. The previous patch assumes the ioctl already existed, when it actually didn't. It also didn't return the correct error code. Signed-off-by: Dave Airlie --- drivers/gpu/drm/radeon/radeon_gem.c | 2 +- drivers/gpu/drm/radeon/radeon_kms.c | 1 + include/drm/radeon_drm.h | 2 ++ 3 files changed, 4 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/drivers/gpu/drm/radeon/radeon_gem.c b/drivers/gpu/drm/radeon/radeon_gem.c index d4ceff13bbb1..14c199802920 100644 --- a/drivers/gpu/drm/radeon/radeon_gem.c +++ b/drivers/gpu/drm/radeon/radeon_gem.c @@ -283,7 +283,7 @@ int radeon_gem_busy_ioctl(struct drm_device *dev, void *data, mutex_lock(&dev->struct_mutex); drm_gem_object_unreference(gobj); mutex_unlock(&dev->struct_mutex); - return 0; + return r; } int radeon_gem_wait_idle_ioctl(struct drm_device *dev, void *data, diff --git a/drivers/gpu/drm/radeon/radeon_kms.c b/drivers/gpu/drm/radeon/radeon_kms.c index d2764bf6b2a2..11ed672543b1 100644 --- a/drivers/gpu/drm/radeon/radeon_kms.c +++ b/drivers/gpu/drm/radeon/radeon_kms.c @@ -318,5 +318,6 @@ struct drm_ioctl_desc radeon_ioctls_kms[] = { DRM_IOCTL_DEF(DRM_RADEON_INFO, radeon_info_ioctl, DRM_AUTH), DRM_IOCTL_DEF(DRM_RADEON_GEM_SET_TILING, radeon_gem_set_tiling_ioctl, DRM_AUTH), DRM_IOCTL_DEF(DRM_RADEON_GEM_GET_TILING, radeon_gem_get_tiling_ioctl, DRM_AUTH), + DRM_IOCTL_DEF(DRM_RADEON_GEM_BUSY, radeon_gem_busy_ioctl, DRM_AUTH), }; int radeon_max_kms_ioctl = DRM_ARRAY_SIZE(radeon_ioctls_kms); diff --git a/include/drm/radeon_drm.h b/include/drm/radeon_drm.h index f81c3232accd..b43925586b29 100644 --- a/include/drm/radeon_drm.h +++ b/include/drm/radeon_drm.h @@ -508,6 +508,7 @@ typedef struct { #define DRM_RADEON_INFO 0x27 #define DRM_RADEON_GEM_SET_TILING 0x28 #define DRM_RADEON_GEM_GET_TILING 0x29 +#define DRM_RADEON_GEM_BUSY 0x2a #define DRM_IOCTL_RADEON_CP_INIT DRM_IOW( DRM_COMMAND_BASE + DRM_RADEON_CP_INIT, drm_radeon_init_t) #define DRM_IOCTL_RADEON_CP_START DRM_IO( DRM_COMMAND_BASE + DRM_RADEON_CP_START) @@ -548,6 +549,7 @@ typedef struct { #define DRM_IOCTL_RADEON_INFO DRM_IOWR(DRM_COMMAND_BASE + DRM_RADEON_INFO, struct drm_radeon_info) #define DRM_IOCTL_RADEON_SET_TILING DRM_IOWR(DRM_COMMAND_BASE + DRM_RADEON_GEM_SET_TILING, struct drm_radeon_gem_set_tiling) #define DRM_IOCTL_RADEON_GET_TILING DRM_IOWR(DRM_COMMAND_BASE + DRM_RADEON_GEM_GET_TILING, struct drm_radeon_gem_get_tiling) +#define DRM_IOCTL_RADEON_GEM_BUSY DRM_IOWR(DRM_COMMAND_BASE + DRM_RADEON_GEM_BUSY, struct drm_radeon_gem_busy) typedef struct drm_radeon_init { enum { -- cgit v1.2.3 From 1700f5fde88f9a251037bc86bde538ee32c59905 Mon Sep 17 00:00:00 2001 From: Marek Vasut Date: Thu, 20 Aug 2009 22:05:53 -0700 Subject: Input: ucb1400_ts - enable ADC Filter This patch enables ADC filtering on UCB1400 codec by default. The benefit from this change is mostly on some Colibri boards where the ADCSYNC pin of the UCB1400 codec isn't connected causing the touchscreen to jitter very badly. This change has no visible effect on boards where the ADCSYNC pin is connected. Signed-off-by: Marek Vasut Tested-by: Palo Revak Signed-off-by: Dmitry Torokhov --- drivers/input/touchscreen/ucb1400_ts.c | 9 +++++++++ include/linux/ucb1400.h | 4 ++++ 2 files changed, 13 insertions(+) (limited to 'include') diff --git a/drivers/input/touchscreen/ucb1400_ts.c b/drivers/input/touchscreen/ucb1400_ts.c index 6954f5500108..3b345f9cf0c7 100644 --- a/drivers/input/touchscreen/ucb1400_ts.c +++ b/drivers/input/touchscreen/ucb1400_ts.c @@ -345,6 +345,7 @@ static int ucb1400_ts_detect_irq(struct ucb1400_ts *ucb) static int ucb1400_ts_probe(struct platform_device *dev) { int error, x_res, y_res; + u16 fcsr; struct ucb1400_ts *ucb = dev->dev.platform_data; ucb->ts_idev = input_allocate_device(); @@ -382,6 +383,14 @@ static int ucb1400_ts_probe(struct platform_device *dev) ucb->ts_idev->evbit[0] = BIT_MASK(EV_ABS) | BIT_MASK(EV_KEY); ucb->ts_idev->keybit[BIT_WORD(BTN_TOUCH)] = BIT_MASK(BTN_TOUCH); + /* + * Enable ADC filter to prevent horrible jitter on Colibri. + * This also further reduces jitter on boards where ADCSYNC + * pin is connected. + */ + fcsr = ucb1400_reg_read(ucb->ac97, UCB_FCSR); + ucb1400_reg_write(ucb->ac97, UCB_FCSR, fcsr | UCB_FCSR_AVE); + ucb1400_adc_enable(ucb->ac97); x_res = ucb1400_ts_read_xres(ucb); y_res = ucb1400_ts_read_yres(ucb); diff --git a/include/linux/ucb1400.h b/include/linux/ucb1400.h index ed889f4168f3..ae779bb8cc0f 100644 --- a/include/linux/ucb1400.h +++ b/include/linux/ucb1400.h @@ -73,6 +73,10 @@ #define UCB_ADC_DATA 0x68 #define UCB_ADC_DAT_VALID (1 << 15) + +#define UCB_FCSR 0x6c +#define UCB_FCSR_AVE (1 << 12) + #define UCB_ADC_DAT_MASK 0x3ff #define UCB_ID 0x7e -- cgit v1.2.3 From f779b3e513478218cbaaaa0a506d7801cab6fd14 Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Wed, 19 Aug 2009 19:11:39 -0400 Subject: drm/radeon: add GET_PARAM/INFO support for Z pipes Needed for occlusion queries on rv530 chips. Signed-off-by: Alex Deucher Signed-off-by: Dave Airlie --- drivers/gpu/drm/radeon/r300.c | 4 +++- drivers/gpu/drm/radeon/r420.c | 13 ++++++++++++- drivers/gpu/drm/radeon/r520.c | 1 - drivers/gpu/drm/radeon/radeon.h | 1 + drivers/gpu/drm/radeon/radeon_cp.c | 9 +++++++++ drivers/gpu/drm/radeon/radeon_drv.h | 5 ++++- drivers/gpu/drm/radeon/radeon_kms.c | 3 +++ drivers/gpu/drm/radeon/radeon_reg.h | 2 ++ drivers/gpu/drm/radeon/radeon_state.c | 3 +++ include/drm/radeon_drm.h | 2 ++ 10 files changed, 39 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/drivers/gpu/drm/radeon/r300.c b/drivers/gpu/drm/radeon/r300.c index c47579dcafa1..053f4ec397f7 100644 --- a/drivers/gpu/drm/radeon/r300.c +++ b/drivers/gpu/drm/radeon/r300.c @@ -448,6 +448,7 @@ void r300_gpu_init(struct radeon_device *rdev) /* rv350,rv370,rv380 */ rdev->num_gb_pipes = 1; } + rdev->num_z_pipes = 1; gb_tile_config = (R300_ENABLE_TILING | R300_TILE_SIZE_16); switch (rdev->num_gb_pipes) { case 2: @@ -486,7 +487,8 @@ void r300_gpu_init(struct radeon_device *rdev) printk(KERN_WARNING "Failed to wait MC idle while " "programming pipes. Bad things might happen.\n"); } - DRM_INFO("radeon: %d pipes initialized.\n", rdev->num_gb_pipes); + DRM_INFO("radeon: %d quad pipes, %d Z pipes initialized.\n", + rdev->num_gb_pipes, rdev->num_z_pipes); } int r300_ga_reset(struct radeon_device *rdev) diff --git a/drivers/gpu/drm/radeon/r420.c b/drivers/gpu/drm/radeon/r420.c index dea497a979f2..97426a6f370f 100644 --- a/drivers/gpu/drm/radeon/r420.c +++ b/drivers/gpu/drm/radeon/r420.c @@ -165,7 +165,18 @@ void r420_pipes_init(struct radeon_device *rdev) printk(KERN_WARNING "Failed to wait GUI idle while " "programming pipes. Bad things might happen.\n"); } - DRM_INFO("radeon: %d pipes initialized.\n", rdev->num_gb_pipes); + + if (rdev->family == CHIP_RV530) { + tmp = RREG32(RV530_GB_PIPE_SELECT2); + if ((tmp & 3) == 3) + rdev->num_z_pipes = 2; + else + rdev->num_z_pipes = 1; + } else + rdev->num_z_pipes = 1; + + DRM_INFO("radeon: %d quad pipes, %d z pipes initialized.\n", + rdev->num_gb_pipes, rdev->num_z_pipes); } void r420_gpu_init(struct radeon_device *rdev) diff --git a/drivers/gpu/drm/radeon/r520.c b/drivers/gpu/drm/radeon/r520.c index 09fb0b6ec7dd..ebd6b0f7bdff 100644 --- a/drivers/gpu/drm/radeon/r520.c +++ b/drivers/gpu/drm/radeon/r520.c @@ -177,7 +177,6 @@ void r520_gpu_init(struct radeon_device *rdev) */ /* workaround for RV530 */ if (rdev->family == CHIP_RV530) { - WREG32(0x4124, 1); WREG32(0x4128, 0xFF); } r420_pipes_init(rdev); diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h index 79ad98264e33..b519fb2fecbb 100644 --- a/drivers/gpu/drm/radeon/radeon.h +++ b/drivers/gpu/drm/radeon/radeon.h @@ -655,6 +655,7 @@ struct radeon_device { int usec_timeout; enum radeon_pll_errata pll_errata; int num_gb_pipes; + int num_z_pipes; int disp_priority; /* BIOS */ uint8_t *bios; diff --git a/drivers/gpu/drm/radeon/radeon_cp.c b/drivers/gpu/drm/radeon/radeon_cp.c index d8356827ef17..7a52c461145c 100644 --- a/drivers/gpu/drm/radeon/radeon_cp.c +++ b/drivers/gpu/drm/radeon/radeon_cp.c @@ -406,6 +406,15 @@ static void radeon_init_pipes(drm_radeon_private_t *dev_priv) { uint32_t gb_tile_config, gb_pipe_sel = 0; + if ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RV530) { + uint32_t z_pipe_sel = RADEON_READ(RV530_GB_PIPE_SELECT2); + if ((z_pipe_sel & 3) == 3) + dev_priv->num_z_pipes = 2; + else + dev_priv->num_z_pipes = 1; + } else + dev_priv->num_z_pipes = 1; + /* RS4xx/RS6xx/R4xx/R5xx */ if ((dev_priv->flags & RADEON_FAMILY_MASK) >= CHIP_R420) { gb_pipe_sel = RADEON_READ(R400_GB_PIPE_SELECT); diff --git a/drivers/gpu/drm/radeon/radeon_drv.h b/drivers/gpu/drm/radeon/radeon_drv.h index 3933f8216a34..6fa32dac4e97 100644 --- a/drivers/gpu/drm/radeon/radeon_drv.h +++ b/drivers/gpu/drm/radeon/radeon_drv.h @@ -100,9 +100,10 @@ * 1.28- Add support for VBL on CRTC2 * 1.29- R500 3D cmd buffer support * 1.30- Add support for occlusion queries + * 1.31- Add support for num Z pipes from GET_PARAM */ #define DRIVER_MAJOR 1 -#define DRIVER_MINOR 30 +#define DRIVER_MINOR 31 #define DRIVER_PATCHLEVEL 0 /* @@ -329,6 +330,7 @@ typedef struct drm_radeon_private { resource_size_t fb_aper_offset; int num_gb_pipes; + int num_z_pipes; int track_flush; drm_local_map_t *mmio; @@ -689,6 +691,7 @@ extern void r600_page_table_cleanup(struct drm_device *dev, struct drm_ati_pciga /* pipe config regs */ #define R400_GB_PIPE_SELECT 0x402c +#define RV530_GB_PIPE_SELECT2 0x4124 #define R500_DYN_SCLK_PWMEM_PIPE 0x000d /* PLL */ #define R300_GB_TILE_CONFIG 0x4018 # define R300_ENABLE_TILING (1 << 0) diff --git a/drivers/gpu/drm/radeon/radeon_kms.c b/drivers/gpu/drm/radeon/radeon_kms.c index 11ed672543b1..dce09ada32bc 100644 --- a/drivers/gpu/drm/radeon/radeon_kms.c +++ b/drivers/gpu/drm/radeon/radeon_kms.c @@ -95,6 +95,9 @@ int radeon_info_ioctl(struct drm_device *dev, void *data, struct drm_file *filp) case RADEON_INFO_NUM_GB_PIPES: value = rdev->num_gb_pipes; break; + case RADEON_INFO_NUM_Z_PIPES: + value = rdev->num_z_pipes; + break; default: DRM_DEBUG("Invalid request %d\n", info->request); return -EINVAL; diff --git a/drivers/gpu/drm/radeon/radeon_reg.h b/drivers/gpu/drm/radeon/radeon_reg.h index 5834497b366d..4df43f62c678 100644 --- a/drivers/gpu/drm/radeon/radeon_reg.h +++ b/drivers/gpu/drm/radeon/radeon_reg.h @@ -3574,4 +3574,6 @@ #define RADEON_SCRATCH_REG4 0x15f0 #define RADEON_SCRATCH_REG5 0x15f4 +#define RV530_GB_PIPE_SELECT2 0x4124 + #endif diff --git a/drivers/gpu/drm/radeon/radeon_state.c b/drivers/gpu/drm/radeon/radeon_state.c index 46645f3e0328..2882f40d5ec5 100644 --- a/drivers/gpu/drm/radeon/radeon_state.c +++ b/drivers/gpu/drm/radeon/radeon_state.c @@ -3081,6 +3081,9 @@ static int radeon_cp_getparam(struct drm_device *dev, void *data, struct drm_fil case RADEON_PARAM_NUM_GB_PIPES: value = dev_priv->num_gb_pipes; break; + case RADEON_PARAM_NUM_Z_PIPES: + value = dev_priv->num_z_pipes; + break; default: DRM_DEBUG("Invalid parameter %d\n", param->param); return -EINVAL; diff --git a/include/drm/radeon_drm.h b/include/drm/radeon_drm.h index b43925586b29..2ba61e18fc8b 100644 --- a/include/drm/radeon_drm.h +++ b/include/drm/radeon_drm.h @@ -709,6 +709,7 @@ typedef struct drm_radeon_indirect { #define RADEON_PARAM_FB_LOCATION 14 /* FB location */ #define RADEON_PARAM_NUM_GB_PIPES 15 /* num GB pipes */ #define RADEON_PARAM_DEVICE_ID 16 +#define RADEON_PARAM_NUM_Z_PIPES 17 /* num Z pipes */ typedef struct drm_radeon_getparam { int param; @@ -897,6 +898,7 @@ struct drm_radeon_cs { #define RADEON_INFO_DEVICE_ID 0x00 #define RADEON_INFO_NUM_GB_PIPES 0x01 +#define RADEON_INFO_NUM_Z_PIPES 0x02 struct drm_radeon_info { uint32_t request; -- cgit v1.2.3 From 4516fc0454e7ffe2f369e80045b23c2b32155004 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Fri, 14 Aug 2009 12:57:54 -0400 Subject: sunrpc: add routine for comparing addresses lockd needs these sort of routines, as does the NFSv4 callback code. Move lockd's routines into common code and rename them so that they can be used by others. Signed-off-by: Jeff Layton Acked-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/clntlock.c | 2 +- fs/lockd/host.c | 4 ++-- fs/lockd/mon.c | 2 +- fs/lockd/svcsubs.c | 2 +- include/linux/lockd/lockd.h | 43 ---------------------------------------- include/linux/sunrpc/clnt.h | 48 +++++++++++++++++++++++++++++++++++++++++++++ 6 files changed, 53 insertions(+), 48 deletions(-) (limited to 'include') diff --git a/fs/lockd/clntlock.c b/fs/lockd/clntlock.c index 1f3b0fc0d351..fc9032dc8862 100644 --- a/fs/lockd/clntlock.c +++ b/fs/lockd/clntlock.c @@ -166,7 +166,7 @@ __be32 nlmclnt_grant(const struct sockaddr *addr, const struct nlm_lock *lock) */ if (fl_blocked->fl_u.nfs_fl.owner->pid != lock->svid) continue; - if (!nlm_cmp_addr(nlm_addr(block->b_host), addr)) + if (!rpc_cmp_addr(nlm_addr(block->b_host), addr)) continue; if (nfs_compare_fh(NFS_FH(fl_blocked->fl_file->f_path.dentry->d_inode) ,fh) != 0) continue; diff --git a/fs/lockd/host.c b/fs/lockd/host.c index 7cb076ac6b45..4600c2037b8b 100644 --- a/fs/lockd/host.c +++ b/fs/lockd/host.c @@ -111,7 +111,7 @@ static struct nlm_host *nlm_lookup_host(struct nlm_lookup_host_info *ni) */ chain = &nlm_hosts[nlm_hash_address(ni->sap)]; hlist_for_each_entry(host, pos, chain, h_hash) { - if (!nlm_cmp_addr(nlm_addr(host), ni->sap)) + if (!rpc_cmp_addr(nlm_addr(host), ni->sap)) continue; /* See if we have an NSM handle for this client */ @@ -125,7 +125,7 @@ static struct nlm_host *nlm_lookup_host(struct nlm_lookup_host_info *ni) if (host->h_server != ni->server) continue; if (ni->server && - !nlm_cmp_addr(nlm_srcaddr(host), ni->src_sap)) + !rpc_cmp_addr(nlm_srcaddr(host), ni->src_sap)) continue; /* Move to head of hash chain. */ diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index 30c933188dd7..f956651d0f65 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -209,7 +209,7 @@ static struct nsm_handle *nsm_lookup_addr(const struct sockaddr *sap) struct nsm_handle *nsm; list_for_each_entry(nsm, &nsm_handles, sm_link) - if (nlm_cmp_addr(nsm_addr(nsm), sap)) + if (rpc_cmp_addr(nsm_addr(nsm), sap)) return nsm; return NULL; } diff --git a/fs/lockd/svcsubs.c b/fs/lockd/svcsubs.c index 9e4d6aab611b..ad478da7ca63 100644 --- a/fs/lockd/svcsubs.c +++ b/fs/lockd/svcsubs.c @@ -417,7 +417,7 @@ EXPORT_SYMBOL_GPL(nlmsvc_unlock_all_by_sb); static int nlmsvc_match_ip(void *datap, struct nlm_host *host) { - return nlm_cmp_addr(nlm_srcaddr(host), datap); + return rpc_cmp_addr(nlm_srcaddr(host), datap); } /** diff --git a/include/linux/lockd/lockd.h b/include/linux/lockd/lockd.h index c325b187966b..e7a251a988c0 100644 --- a/include/linux/lockd/lockd.h +++ b/include/linux/lockd/lockd.h @@ -338,49 +338,6 @@ static inline int nlm_privileged_requester(const struct svc_rqst *rqstp) } } -static inline int __nlm_cmp_addr4(const struct sockaddr *sap1, - const struct sockaddr *sap2) -{ - const struct sockaddr_in *sin1 = (const struct sockaddr_in *)sap1; - const struct sockaddr_in *sin2 = (const struct sockaddr_in *)sap2; - return sin1->sin_addr.s_addr == sin2->sin_addr.s_addr; -} - -#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE) -static inline int __nlm_cmp_addr6(const struct sockaddr *sap1, - const struct sockaddr *sap2) -{ - const struct sockaddr_in6 *sin1 = (const struct sockaddr_in6 *)sap1; - const struct sockaddr_in6 *sin2 = (const struct sockaddr_in6 *)sap2; - return ipv6_addr_equal(&sin1->sin6_addr, &sin2->sin6_addr); -} -#else /* !(CONFIG_IPV6 || CONFIG_IPV6_MODULE) */ -static inline int __nlm_cmp_addr6(const struct sockaddr *sap1, - const struct sockaddr *sap2) -{ - return 0; -} -#endif /* !(CONFIG_IPV6 || CONFIG_IPV6_MODULE) */ - -/* - * Compare two host addresses - * - * Return TRUE if the addresses are the same; otherwise FALSE. - */ -static inline int nlm_cmp_addr(const struct sockaddr *sap1, - const struct sockaddr *sap2) -{ - if (sap1->sa_family == sap2->sa_family) { - switch (sap1->sa_family) { - case AF_INET: - return __nlm_cmp_addr4(sap1, sap2); - case AF_INET6: - return __nlm_cmp_addr6(sap1, sap2); - } - } - return 0; -} - /* * Compare two NLM locks. * When the second lock is of type F_UNLCK, this acts like a wildcard. diff --git a/include/linux/sunrpc/clnt.h b/include/linux/sunrpc/clnt.h index ab3f6e90caa5..b17df361be82 100644 --- a/include/linux/sunrpc/clnt.h +++ b/include/linux/sunrpc/clnt.h @@ -22,6 +22,7 @@ #include #include #include +#include struct rpc_inode; @@ -188,5 +189,52 @@ static inline void rpc_set_port(struct sockaddr *sap, #define IPV6_SCOPE_DELIMITER '%' #define IPV6_SCOPE_ID_LEN sizeof("%nnnnnnnnnn") +static inline bool __rpc_cmp_addr4(const struct sockaddr *sap1, + const struct sockaddr *sap2) +{ + const struct sockaddr_in *sin1 = (const struct sockaddr_in *)sap1; + const struct sockaddr_in *sin2 = (const struct sockaddr_in *)sap2; + + return sin1->sin_addr.s_addr == sin2->sin_addr.s_addr; +} + +#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE) +static inline bool __rpc_cmp_addr6(const struct sockaddr *sap1, + const struct sockaddr *sap2) +{ + const struct sockaddr_in6 *sin1 = (const struct sockaddr_in6 *)sap1; + const struct sockaddr_in6 *sin2 = (const struct sockaddr_in6 *)sap2; + return ipv6_addr_equal(&sin1->sin6_addr, &sin2->sin6_addr); +} +#else /* !(CONFIG_IPV6 || CONFIG_IPV6_MODULE) */ +static inline bool __rpc_cmp_addr6(const struct sockaddr *sap1, + const struct sockaddr *sap2) +{ + return false; +} +#endif /* !(CONFIG_IPV6 || CONFIG_IPV6_MODULE) */ + +/** + * rpc_cmp_addr - compare the address portion of two sockaddrs. + * @sap1: first sockaddr + * @sap2: second sockaddr + * + * Just compares the family and address portion. Ignores port, scope, etc. + * Returns true if the addrs are equal, false if they aren't. + */ +static inline bool rpc_cmp_addr(const struct sockaddr *sap1, + const struct sockaddr *sap2) +{ + if (sap1->sa_family == sap2->sa_family) { + switch (sap1->sa_family) { + case AF_INET: + return __rpc_cmp_addr4(sap1, sap2); + case AF_INET6: + return __rpc_cmp_addr6(sap1, sap2); + } + } + return false; +} + #endif /* __KERNEL__ */ #endif /* _LINUX_SUNRPC_CLNT_H */ -- cgit v1.2.3 From be3ad6b0b675fd1d6b48362ca30bdee75fbef6b4 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Fri, 14 Aug 2009 12:57:55 -0400 Subject: sunrpc: add common routine for copying address portion of a sockaddr Signed-off-by: Jeff Layton Acked-by: Chuck Lever Signed-off-by: J. Bruce Fields --- include/linux/sunrpc/clnt.h | 50 +++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 50 insertions(+) (limited to 'include') diff --git a/include/linux/sunrpc/clnt.h b/include/linux/sunrpc/clnt.h index b17df361be82..044f531aee70 100644 --- a/include/linux/sunrpc/clnt.h +++ b/include/linux/sunrpc/clnt.h @@ -198,6 +198,17 @@ static inline bool __rpc_cmp_addr4(const struct sockaddr *sap1, return sin1->sin_addr.s_addr == sin2->sin_addr.s_addr; } +static inline bool __rpc_copy_addr4(struct sockaddr *dst, + const struct sockaddr *src) +{ + const struct sockaddr_in *ssin = (struct sockaddr_in *) src; + struct sockaddr_in *dsin = (struct sockaddr_in *) dst; + + dsin->sin_family = ssin->sin_family; + dsin->sin_addr.s_addr = ssin->sin_addr.s_addr; + return true; +} + #if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE) static inline bool __rpc_cmp_addr6(const struct sockaddr *sap1, const struct sockaddr *sap2) @@ -206,12 +217,29 @@ static inline bool __rpc_cmp_addr6(const struct sockaddr *sap1, const struct sockaddr_in6 *sin2 = (const struct sockaddr_in6 *)sap2; return ipv6_addr_equal(&sin1->sin6_addr, &sin2->sin6_addr); } + +static inline bool __rpc_copy_addr6(struct sockaddr *dst, + const struct sockaddr *src) +{ + const struct sockaddr_in6 *ssin6 = (const struct sockaddr_in6 *) src; + struct sockaddr_in6 *dsin6 = (struct sockaddr_in6 *) dst; + + dsin6->sin6_family = ssin6->sin6_family; + ipv6_addr_copy(&dsin6->sin6_addr, &ssin6->sin6_addr); + return true; +} #else /* !(CONFIG_IPV6 || CONFIG_IPV6_MODULE) */ static inline bool __rpc_cmp_addr6(const struct sockaddr *sap1, const struct sockaddr *sap2) { return false; } + +static inline bool __rpc_copy_addr6(struct sockaddr *dst, + const struct sockaddr *src) +{ + return false; +} #endif /* !(CONFIG_IPV6 || CONFIG_IPV6_MODULE) */ /** @@ -236,5 +264,27 @@ static inline bool rpc_cmp_addr(const struct sockaddr *sap1, return false; } +/** + * rpc_copy_addr - copy the address portion of one sockaddr to another + * @dst: destination sockaddr + * @src: source sockaddr + * + * Just copies the address portion and family. Ignores port, scope, etc. + * Caller is responsible for making certain that dst is large enough to hold + * the address in src. Returns true if address family is supported. Returns + * false otherwise. + */ +static inline bool rpc_copy_addr(struct sockaddr *dst, + const struct sockaddr *src) +{ + switch (src->sa_family) { + case AF_INET: + return __rpc_copy_addr4(dst, src); + case AF_INET6: + return __rpc_copy_addr6(dst, src); + } + return false; +} + #endif /* __KERNEL__ */ #endif /* _LINUX_SUNRPC_CLNT_H */ -- cgit v1.2.3 From 363168b4ea8ec26aeb982ac6024a09f907ecd27e Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Fri, 14 Aug 2009 12:57:56 -0400 Subject: nfsd: make nfs4_client->cl_addr a struct sockaddr_storage It's currently a __be32, which isn't big enough to hold an IPv6 address. Signed-off-by: Jeff Layton Acked-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 32 +++++++++++++++++++------------- include/linux/nfsd/state.h | 2 +- 2 files changed, 20 insertions(+), 14 deletions(-) (limited to 'include') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 9295c4b56bce..bfc14d879ea1 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -55,6 +55,7 @@ #include #include #include +#include #define NFSDDBG_FACILITY NFSDDBG_PROC @@ -1220,13 +1221,15 @@ nfsd4_exchange_id(struct svc_rqst *rqstp, int status; unsigned int strhashval; char dname[HEXDIR_LEN]; + char addr_str[INET6_ADDRSTRLEN]; nfs4_verifier verf = exid->verifier; - u32 ip_addr = svc_addr_in(rqstp)->sin_addr.s_addr; + struct sockaddr *sa = svc_addr(rqstp); + rpc_ntop(sa, addr_str, sizeof(addr_str)); dprintk("%s rqstp=%p exid=%p clname.len=%u clname.data=%p " - " ip_addr=%u flags %x, spa_how %d\n", + "ip_addr=%s flags %x, spa_how %d\n", __func__, rqstp, exid, exid->clname.len, exid->clname.data, - ip_addr, exid->flags, exid->spa_how); + addr_str, exid->flags, exid->spa_how); if (!check_name(exid->clname) || (exid->flags & ~EXCHGID4_FLAG_MASK_A)) return nfserr_inval; @@ -1315,7 +1318,7 @@ out_new: copy_verf(new, &verf); copy_cred(&new->cl_cred, &rqstp->rq_cred); - new->cl_addr = ip_addr; + rpc_copy_addr((struct sockaddr *) &new->cl_addr, sa); gen_clid(new); gen_confirm(new); add_to_unconfirmed(new, strhashval); @@ -1389,7 +1392,7 @@ nfsd4_create_session(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, struct nfsd4_create_session *cr_ses) { - u32 ip_addr = svc_addr_in(rqstp)->sin_addr.s_addr; + struct sockaddr *sa = svc_addr(rqstp); struct nfs4_client *conf, *unconf; struct nfsd4_clid_slot *cs_slot = NULL; int status = 0; @@ -1417,7 +1420,7 @@ nfsd4_create_session(struct svc_rqst *rqstp, cs_slot->sl_seqid++; } else if (unconf) { if (!same_creds(&unconf->cl_cred, &rqstp->rq_cred) || - (ip_addr != unconf->cl_addr)) { + !rpc_cmp_addr(sa, (struct sockaddr *) &unconf->cl_addr)) { status = nfserr_clid_inuse; goto out; } @@ -1564,7 +1567,7 @@ __be32 nfsd4_setclientid(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, struct nfsd4_setclientid *setclid) { - struct sockaddr_in *sin = svc_addr_in(rqstp); + struct sockaddr *sa = svc_addr(rqstp); struct xdr_netobj clname = { .len = setclid->se_namelen, .data = setclid->se_name, @@ -1596,8 +1599,11 @@ nfsd4_setclientid(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, /* RFC 3530 14.2.33 CASE 0: */ status = nfserr_clid_inuse; if (!same_creds(&conf->cl_cred, &rqstp->rq_cred)) { - dprintk("NFSD: setclientid: string in use by client" - " at %pI4\n", &conf->cl_addr); + char addr_str[INET6_ADDRSTRLEN]; + rpc_ntop((struct sockaddr *) &conf->cl_addr, addr_str, + sizeof(addr_str)); + dprintk("NFSD: setclientid: string in use by client " + "at %s\n", addr_str); goto out; } } @@ -1659,7 +1665,7 @@ nfsd4_setclientid(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, gen_clid(new); } copy_verf(new, &clverifier); - new->cl_addr = sin->sin_addr.s_addr; + rpc_copy_addr((struct sockaddr *) &new->cl_addr, sa); new->cl_flavor = rqstp->rq_flavor; princ = svc_gss_principal(rqstp); if (princ) { @@ -1693,7 +1699,7 @@ nfsd4_setclientid_confirm(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, struct nfsd4_setclientid_confirm *setclientid_confirm) { - struct sockaddr_in *sin = svc_addr_in(rqstp); + struct sockaddr *sa = svc_addr(rqstp); struct nfs4_client *conf, *unconf; nfs4_verifier confirm = setclientid_confirm->sc_confirm; clientid_t * clid = &setclientid_confirm->sc_clientid; @@ -1712,9 +1718,9 @@ nfsd4_setclientid_confirm(struct svc_rqst *rqstp, unconf = find_unconfirmed_client(clid); status = nfserr_clid_inuse; - if (conf && conf->cl_addr != sin->sin_addr.s_addr) + if (conf && !rpc_cmp_addr((struct sockaddr *) &conf->cl_addr, sa)) goto out; - if (unconf && unconf->cl_addr != sin->sin_addr.s_addr) + if (unconf && !rpc_cmp_addr((struct sockaddr *) &unconf->cl_addr, sa)) goto out; /* diff --git a/include/linux/nfsd/state.h b/include/linux/nfsd/state.h index 58bb19784e12..3510ddd4be49 100644 --- a/include/linux/nfsd/state.h +++ b/include/linux/nfsd/state.h @@ -200,7 +200,7 @@ struct nfs4_client { char cl_recdir[HEXDIR_LEN]; /* recovery dir */ nfs4_verifier cl_verifier; /* generated by client */ time_t cl_time; /* time of last lease renewal */ - __be32 cl_addr; /* client ipaddress */ + struct sockaddr_storage cl_addr; /* client ipaddress */ u32 cl_flavor; /* setclientid pseudoflavor */ char *cl_principal; /* setclientid principal name */ struct svc_cred cl_cred; /* setclientid principal */ -- cgit v1.2.3 From aa9a4ec7707a5391cde556f3fa1b0eb4bca3bcf6 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Fri, 14 Aug 2009 12:57:57 -0400 Subject: nfsd: convert nfs4_cb_conn struct to hold address in sockaddr_storage ...rather than as a separate address and port fields. This will be necessary for implementing callbacks over IPv6. Also, convert gen_callback to use the standard rpcuaddr2sockaddr routine rather than its own private one. Signed-off-by: Jeff Layton Acked-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4callback.c | 11 ++----- fs/nfsd/nfs4state.c | 81 ++++++---------------------------------------- include/linux/nfsd/state.h | 4 +-- 3 files changed, 13 insertions(+), 83 deletions(-) (limited to 'include') diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c index 3fd23f7aceca..81d1c5285dcc 100644 --- a/fs/nfsd/nfs4callback.c +++ b/fs/nfsd/nfs4callback.c @@ -377,7 +377,6 @@ static int max_cb_time(void) int setup_callback_client(struct nfs4_client *clp) { - struct sockaddr_in addr; struct nfs4_cb_conn *cb = &clp->cl_cb_conn; struct rpc_timeout timeparms = { .to_initval = max_cb_time(), @@ -385,8 +384,8 @@ int setup_callback_client(struct nfs4_client *clp) }; struct rpc_create_args args = { .protocol = IPPROTO_TCP, - .address = (struct sockaddr *)&addr, - .addrsize = sizeof(addr), + .address = (struct sockaddr *) &cb->cb_addr, + .addrsize = cb->cb_addrlen, .timeout = &timeparms, .program = &cb_program, .prognumber = cb->cb_prog, @@ -400,12 +399,6 @@ int setup_callback_client(struct nfs4_client *clp) if (!clp->cl_principal && (clp->cl_flavor >= RPC_AUTH_GSS_KRB5)) return -EINVAL; - /* Initialize address */ - memset(&addr, 0, sizeof(addr)); - addr.sin_family = AF_INET; - addr.sin_port = htons(cb->cb_port); - addr.sin_addr.s_addr = htonl(cb->cb_addr); - /* Create RPC client */ client = rpc_create(&args); if (IS_ERR(client)) { diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index bfc14d879ea1..96a742308cee 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -897,76 +897,6 @@ find_unconfirmed_client_by_str(const char *dname, unsigned int hashval, return NULL; } -/* a helper function for parse_callback */ -static int -parse_octet(unsigned int *lenp, char **addrp) -{ - unsigned int len = *lenp; - char *p = *addrp; - int n = -1; - char c; - - for (;;) { - if (!len) - break; - len--; - c = *p++; - if (c == '.') - break; - if ((c < '0') || (c > '9')) { - n = -1; - break; - } - if (n < 0) - n = 0; - n = (n * 10) + (c - '0'); - if (n > 255) { - n = -1; - break; - } - } - *lenp = len; - *addrp = p; - return n; -} - -/* parse and set the setclientid ipv4 callback address */ -static int -parse_ipv4(unsigned int addr_len, char *addr_val, unsigned int *cbaddrp, unsigned short *cbportp) -{ - int temp = 0; - u32 cbaddr = 0; - u16 cbport = 0; - u32 addrlen = addr_len; - char *addr = addr_val; - int i, shift; - - /* ipaddress */ - shift = 24; - for(i = 4; i > 0 ; i--) { - if ((temp = parse_octet(&addrlen, &addr)) < 0) { - return 0; - } - cbaddr |= (temp << shift); - if (shift > 0) - shift -= 8; - } - *cbaddrp = cbaddr; - - /* port */ - shift = 8; - for(i = 2; i > 0 ; i--) { - if ((temp = parse_octet(&addrlen, &addr)) < 0) { - return 0; - } - cbport |= (temp << shift); - if (shift > 0) - shift -= 8; - } - *cbportp = cbport; - return 1; -} - static void gen_callback(struct nfs4_client *clp, struct nfsd4_setclientid *se) { @@ -976,14 +906,21 @@ gen_callback(struct nfs4_client *clp, struct nfsd4_setclientid *se) if ((se->se_callback_netid_len != 3) || memcmp((char *)se->se_callback_netid_val, "tcp", 3)) goto out_err; - if ( !(parse_ipv4(se->se_callback_addr_len, se->se_callback_addr_val, - &cb->cb_addr, &cb->cb_port))) + cb->cb_addrlen = rpc_uaddr2sockaddr(se->se_callback_addr_val, + se->se_callback_addr_len, + (struct sockaddr *) &cb->cb_addr, + sizeof(cb->cb_addr)); + + if (!cb->cb_addrlen || cb->cb_addr.ss_family != AF_INET) goto out_err; + cb->cb_minorversion = 0; cb->cb_prog = se->se_callback_prog; cb->cb_ident = se->se_callback_ident; return; out_err: + cb->cb_addr.ss_family = AF_UNSPEC; + cb->cb_addrlen = 0; dprintk(KERN_INFO "NFSD: this client (clientid %08x/%08x) " "will not receive delegations\n", clp->cl_clientid.cl_boot, clp->cl_clientid.cl_id); diff --git a/include/linux/nfsd/state.h b/include/linux/nfsd/state.h index 3510ddd4be49..fb0c404c7c5c 100644 --- a/include/linux/nfsd/state.h +++ b/include/linux/nfsd/state.h @@ -81,8 +81,8 @@ struct nfs4_delegation { /* client delegation callback info */ struct nfs4_cb_conn { /* SETCLIENTID info */ - u32 cb_addr; - unsigned short cb_port; + struct sockaddr_storage cb_addr; + size_t cb_addrlen; u32 cb_prog; u32 cb_minorversion; u32 cb_ident; /* minorversion 0 only */ -- cgit v1.2.3 From fbf4665f41b02e757ab9d9198df65e319388e728 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Fri, 14 Aug 2009 12:57:59 -0400 Subject: nfsd: populate sin6_scope_id on callback address with scopeid from rq_addr on SETCLIENTID call When a SETCLIENTID call comes in, one of the args given is the svc_rqst. This struct contains an rq_addr field which holds the address that sent the call. If this is an IPv6 address, then we can use the sin6_scope_id field in this address to populate the sin6_scope_id field in the callback address. AFAICT, the rq_addr.sin6_scope_id is non-zero if and only if the client mounted the server's link-local address. Signed-off-by: Jeff Layton Acked-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 7 +++++-- include/linux/sunrpc/clnt.h | 15 +++++++++++++++ 2 files changed, 20 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 9ec0ca1ef4ea..d2a052480908 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -898,7 +898,7 @@ find_unconfirmed_client_by_str(const char *dname, unsigned int hashval, } static void -gen_callback(struct nfs4_client *clp, struct nfsd4_setclientid *se) +gen_callback(struct nfs4_client *clp, struct nfsd4_setclientid *se, u32 scopeid) { struct nfs4_cb_conn *cb = &clp->cl_cb_conn; unsigned short expected_family; @@ -921,6 +921,9 @@ gen_callback(struct nfs4_client *clp, struct nfsd4_setclientid *se) if (!cb->cb_addrlen || cb->cb_addr.ss_family != expected_family) goto out_err; + if (cb->cb_addr.ss_family == AF_INET6) + ((struct sockaddr_in6 *) &cb->cb_addr)->sin6_scope_id = scopeid; + cb->cb_minorversion = 0; cb->cb_prog = se->se_callback_prog; cb->cb_ident = se->se_callback_ident; @@ -1621,7 +1624,7 @@ nfsd4_setclientid(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, } copy_cred(&new->cl_cred, &rqstp->rq_cred); gen_confirm(new); - gen_callback(new, setclid); + gen_callback(new, setclid, rpc_get_scope_id(sa)); add_to_unconfirmed(new, strhashval); setclid->se_clientid.cl_boot = new->cl_clientid.cl_boot; setclid->se_clientid.cl_id = new->cl_clientid.cl_id; diff --git a/include/linux/sunrpc/clnt.h b/include/linux/sunrpc/clnt.h index 044f531aee70..3d025588e56e 100644 --- a/include/linux/sunrpc/clnt.h +++ b/include/linux/sunrpc/clnt.h @@ -286,5 +286,20 @@ static inline bool rpc_copy_addr(struct sockaddr *dst, return false; } +/** + * rpc_get_scope_id - return scopeid for a given sockaddr + * @sa: sockaddr to get scopeid from + * + * Returns the value of the sin6_scope_id for AF_INET6 addrs, or 0 if + * not an AF_INET6 address. + */ +static inline u32 rpc_get_scope_id(const struct sockaddr *sa) +{ + if (sa->sa_family != AF_INET6) + return 0; + + return ((struct sockaddr_in6 *) sa)->sin6_scope_id; +} + #endif /* __KERNEL__ */ #endif /* _LINUX_SUNRPC_CLNT_H */ -- cgit v1.2.3 From f4b0373b26567cafd421d91101852ed7a34e9e94 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Fri, 21 Aug 2009 09:26:15 -0700 Subject: Make bitmask 'and' operators return a result code When 'and'ing two bitmasks (where 'andnot' is a variation on it), some cases want to know whether the result is the empty set or not. In particular, the TLB IPI sending code wants to do cpumask operations and determine if there are any CPU's left in the final set. So this just makes the bitmask (and cpumask) functions return a boolean for whether the result has any bits set. Cc: stable@kernel.org (2.6.30, needed by TLB shootdown fix) Signed-off-by: Linus Torvalds --- include/linux/bitmap.h | 18 ++++++++---------- include/linux/cpumask.h | 20 ++++++++++---------- lib/bitmap.c | 12 ++++++++---- 3 files changed, 26 insertions(+), 24 deletions(-) (limited to 'include') diff --git a/include/linux/bitmap.h b/include/linux/bitmap.h index 2878811c6134..756d78b8c1c5 100644 --- a/include/linux/bitmap.h +++ b/include/linux/bitmap.h @@ -94,13 +94,13 @@ extern void __bitmap_shift_right(unsigned long *dst, const unsigned long *src, int shift, int bits); extern void __bitmap_shift_left(unsigned long *dst, const unsigned long *src, int shift, int bits); -extern void __bitmap_and(unsigned long *dst, const unsigned long *bitmap1, +extern int __bitmap_and(unsigned long *dst, const unsigned long *bitmap1, const unsigned long *bitmap2, int bits); extern void __bitmap_or(unsigned long *dst, const unsigned long *bitmap1, const unsigned long *bitmap2, int bits); extern void __bitmap_xor(unsigned long *dst, const unsigned long *bitmap1, const unsigned long *bitmap2, int bits); -extern void __bitmap_andnot(unsigned long *dst, const unsigned long *bitmap1, +extern int __bitmap_andnot(unsigned long *dst, const unsigned long *bitmap1, const unsigned long *bitmap2, int bits); extern int __bitmap_intersects(const unsigned long *bitmap1, const unsigned long *bitmap2, int bits); @@ -171,13 +171,12 @@ static inline void bitmap_copy(unsigned long *dst, const unsigned long *src, } } -static inline void bitmap_and(unsigned long *dst, const unsigned long *src1, +static inline int bitmap_and(unsigned long *dst, const unsigned long *src1, const unsigned long *src2, int nbits) { if (small_const_nbits(nbits)) - *dst = *src1 & *src2; - else - __bitmap_and(dst, src1, src2, nbits); + return (*dst = *src1 & *src2) != 0; + return __bitmap_and(dst, src1, src2, nbits); } static inline void bitmap_or(unsigned long *dst, const unsigned long *src1, @@ -198,13 +197,12 @@ static inline void bitmap_xor(unsigned long *dst, const unsigned long *src1, __bitmap_xor(dst, src1, src2, nbits); } -static inline void bitmap_andnot(unsigned long *dst, const unsigned long *src1, +static inline int bitmap_andnot(unsigned long *dst, const unsigned long *src1, const unsigned long *src2, int nbits) { if (small_const_nbits(nbits)) - *dst = *src1 & ~(*src2); - else - __bitmap_andnot(dst, src1, src2, nbits); + return (*dst = *src1 & ~(*src2)) != 0; + return __bitmap_andnot(dst, src1, src2, nbits); } static inline void bitmap_complement(unsigned long *dst, const unsigned long *src, diff --git a/include/linux/cpumask.h b/include/linux/cpumask.h index c5ac87ca7bc6..796df12091b7 100644 --- a/include/linux/cpumask.h +++ b/include/linux/cpumask.h @@ -43,10 +43,10 @@ * int cpu_isset(cpu, mask) true iff bit 'cpu' set in mask * int cpu_test_and_set(cpu, mask) test and set bit 'cpu' in mask * - * void cpus_and(dst, src1, src2) dst = src1 & src2 [intersection] + * int cpus_and(dst, src1, src2) dst = src1 & src2 [intersection] * void cpus_or(dst, src1, src2) dst = src1 | src2 [union] * void cpus_xor(dst, src1, src2) dst = src1 ^ src2 - * void cpus_andnot(dst, src1, src2) dst = src1 & ~src2 + * int cpus_andnot(dst, src1, src2) dst = src1 & ~src2 * void cpus_complement(dst, src) dst = ~src * * int cpus_equal(mask1, mask2) Does mask1 == mask2? @@ -179,10 +179,10 @@ static inline int __cpu_test_and_set(int cpu, cpumask_t *addr) } #define cpus_and(dst, src1, src2) __cpus_and(&(dst), &(src1), &(src2), NR_CPUS) -static inline void __cpus_and(cpumask_t *dstp, const cpumask_t *src1p, +static inline int __cpus_and(cpumask_t *dstp, const cpumask_t *src1p, const cpumask_t *src2p, int nbits) { - bitmap_and(dstp->bits, src1p->bits, src2p->bits, nbits); + return bitmap_and(dstp->bits, src1p->bits, src2p->bits, nbits); } #define cpus_or(dst, src1, src2) __cpus_or(&(dst), &(src1), &(src2), NR_CPUS) @@ -201,10 +201,10 @@ static inline void __cpus_xor(cpumask_t *dstp, const cpumask_t *src1p, #define cpus_andnot(dst, src1, src2) \ __cpus_andnot(&(dst), &(src1), &(src2), NR_CPUS) -static inline void __cpus_andnot(cpumask_t *dstp, const cpumask_t *src1p, +static inline int __cpus_andnot(cpumask_t *dstp, const cpumask_t *src1p, const cpumask_t *src2p, int nbits) { - bitmap_andnot(dstp->bits, src1p->bits, src2p->bits, nbits); + return bitmap_andnot(dstp->bits, src1p->bits, src2p->bits, nbits); } #define cpus_complement(dst, src) __cpus_complement(&(dst), &(src), NR_CPUS) @@ -738,11 +738,11 @@ static inline void cpumask_clear(struct cpumask *dstp) * @src1p: the first input * @src2p: the second input */ -static inline void cpumask_and(struct cpumask *dstp, +static inline int cpumask_and(struct cpumask *dstp, const struct cpumask *src1p, const struct cpumask *src2p) { - bitmap_and(cpumask_bits(dstp), cpumask_bits(src1p), + return bitmap_and(cpumask_bits(dstp), cpumask_bits(src1p), cpumask_bits(src2p), nr_cpumask_bits); } @@ -779,11 +779,11 @@ static inline void cpumask_xor(struct cpumask *dstp, * @src1p: the first input * @src2p: the second input */ -static inline void cpumask_andnot(struct cpumask *dstp, +static inline int cpumask_andnot(struct cpumask *dstp, const struct cpumask *src1p, const struct cpumask *src2p) { - bitmap_andnot(cpumask_bits(dstp), cpumask_bits(src1p), + return bitmap_andnot(cpumask_bits(dstp), cpumask_bits(src1p), cpumask_bits(src2p), nr_cpumask_bits); } diff --git a/lib/bitmap.c b/lib/bitmap.c index 35a1f7ff4149..702565821c99 100644 --- a/lib/bitmap.c +++ b/lib/bitmap.c @@ -179,14 +179,16 @@ void __bitmap_shift_left(unsigned long *dst, } EXPORT_SYMBOL(__bitmap_shift_left); -void __bitmap_and(unsigned long *dst, const unsigned long *bitmap1, +int __bitmap_and(unsigned long *dst, const unsigned long *bitmap1, const unsigned long *bitmap2, int bits) { int k; int nr = BITS_TO_LONGS(bits); + unsigned long result = 0; for (k = 0; k < nr; k++) - dst[k] = bitmap1[k] & bitmap2[k]; + result |= (dst[k] = bitmap1[k] & bitmap2[k]); + return result != 0; } EXPORT_SYMBOL(__bitmap_and); @@ -212,14 +214,16 @@ void __bitmap_xor(unsigned long *dst, const unsigned long *bitmap1, } EXPORT_SYMBOL(__bitmap_xor); -void __bitmap_andnot(unsigned long *dst, const unsigned long *bitmap1, +int __bitmap_andnot(unsigned long *dst, const unsigned long *bitmap1, const unsigned long *bitmap2, int bits) { int k; int nr = BITS_TO_LONGS(bits); + unsigned long result = 0; for (k = 0; k < nr; k++) - dst[k] = bitmap1[k] & ~bitmap2[k]; + result |= (dst[k] = bitmap1[k] & ~bitmap2[k]); + return result != 0; } EXPORT_SYMBOL(__bitmap_andnot); -- cgit v1.2.3 From da15cfdae03351c689736f8d142618592e3cebc3 Mon Sep 17 00:00:00 2001 From: john stultz Date: Wed, 19 Aug 2009 19:13:34 -0700 Subject: time: Introduce CLOCK_REALTIME_COARSE After talking with some application writers who want very fast, but not fine-grained timestamps, I decided to try to implement new clock_ids to clock_gettime(): CLOCK_REALTIME_COARSE and CLOCK_MONOTONIC_COARSE which returns the time at the last tick. This is very fast as we don't have to access any hardware (which can be very painful if you're using something like the acpi_pm clocksource), and we can even use the vdso clock_gettime() method to avoid the syscall. The only trade off is you only get low-res tick grained time resolution. This isn't a new idea, I know Ingo has a patch in the -rt tree that made the vsyscall gettimeofday() return coarse grained time when the vsyscall64 sysctrl was set to 2. However this affects all applications on a system. With this method, applications can choose the proper speed/granularity trade-off for themselves. Signed-off-by: John Stultz Cc: Andi Kleen Cc: nikolag@ca.ibm.com Cc: Darren Hart Cc: arjan@infradead.org Cc: jonathan@jonmasters.org LKML-Reference: <1250734414.6897.5.camel@localhost.localdomain> Signed-off-by: Thomas Gleixner --- arch/x86/include/asm/vgtod.h | 1 + arch/x86/kernel/vsyscall_64.c | 1 + arch/x86/vdso/vclock_gettime.c | 39 ++++++++++++++++++++++++++++++++++++--- include/linux/time.h | 4 ++++ kernel/posix-timers.c | 35 +++++++++++++++++++++++++++++++++++ kernel/time/timekeeping.c | 21 +++++++++++++++++++++ 6 files changed, 98 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/arch/x86/include/asm/vgtod.h b/arch/x86/include/asm/vgtod.h index dc27a69e5d2a..3d61e204826f 100644 --- a/arch/x86/include/asm/vgtod.h +++ b/arch/x86/include/asm/vgtod.h @@ -21,6 +21,7 @@ struct vsyscall_gtod_data { u32 shift; } clock; struct timespec wall_to_monotonic; + struct timespec wall_time_coarse; }; extern struct vsyscall_gtod_data __vsyscall_gtod_data __section_vsyscall_gtod_data; diff --git a/arch/x86/kernel/vsyscall_64.c b/arch/x86/kernel/vsyscall_64.c index 25ee06a80aad..cf53a78e2dcf 100644 --- a/arch/x86/kernel/vsyscall_64.c +++ b/arch/x86/kernel/vsyscall_64.c @@ -87,6 +87,7 @@ void update_vsyscall(struct timespec *wall_time, struct clocksource *clock) vsyscall_gtod_data.wall_time_sec = wall_time->tv_sec; vsyscall_gtod_data.wall_time_nsec = wall_time->tv_nsec; vsyscall_gtod_data.wall_to_monotonic = wall_to_monotonic; + vsyscall_gtod_data.wall_time_coarse = __current_kernel_time(); write_sequnlock_irqrestore(&vsyscall_gtod_data.lock, flags); } diff --git a/arch/x86/vdso/vclock_gettime.c b/arch/x86/vdso/vclock_gettime.c index 6a40b78b46aa..ee55754cc3c5 100644 --- a/arch/x86/vdso/vclock_gettime.c +++ b/arch/x86/vdso/vclock_gettime.c @@ -86,14 +86,47 @@ notrace static noinline int do_monotonic(struct timespec *ts) return 0; } +notrace static noinline int do_realtime_coarse(struct timespec *ts) +{ + unsigned long seq; + do { + seq = read_seqbegin(>od->lock); + ts->tv_sec = gtod->wall_time_coarse.tv_sec; + ts->tv_nsec = gtod->wall_time_coarse.tv_nsec; + } while (unlikely(read_seqretry(>od->lock, seq))); + return 0; +} + +notrace static noinline int do_monotonic_coarse(struct timespec *ts) +{ + unsigned long seq, ns, secs; + do { + seq = read_seqbegin(>od->lock); + secs = gtod->wall_time_coarse.tv_sec; + ns = gtod->wall_time_coarse.tv_nsec; + secs += gtod->wall_to_monotonic.tv_sec; + ns += gtod->wall_to_monotonic.tv_nsec; + } while (unlikely(read_seqretry(>od->lock, seq))); + vset_normalized_timespec(ts, secs, ns); + return 0; +} + notrace int __vdso_clock_gettime(clockid_t clock, struct timespec *ts) { - if (likely(gtod->sysctl_enabled && gtod->clock.vread)) + if (likely(gtod->sysctl_enabled)) switch (clock) { case CLOCK_REALTIME: - return do_realtime(ts); + if (likely(gtod->clock.vread)) + return do_realtime(ts); + break; case CLOCK_MONOTONIC: - return do_monotonic(ts); + if (likely(gtod->clock.vread)) + return do_monotonic(ts); + break; + case CLOCK_REALTIME_COARSE: + return do_realtime_coarse(ts); + case CLOCK_MONOTONIC_COARSE: + return do_monotonic_coarse(ts); } return vdso_fallback_gettime(clock, ts); } diff --git a/include/linux/time.h b/include/linux/time.h index f505988398e6..256232f7e5e6 100644 --- a/include/linux/time.h +++ b/include/linux/time.h @@ -110,6 +110,8 @@ extern int timekeeping_suspended; unsigned long get_seconds(void); struct timespec current_kernel_time(void); +struct timespec __current_kernel_time(void); /* does not hold xtime_lock */ +struct timespec get_monotonic_coarse(void); #define CURRENT_TIME (current_kernel_time()) #define CURRENT_TIME_SEC ((struct timespec) { get_seconds(), 0 }) @@ -243,6 +245,8 @@ struct itimerval { #define CLOCK_PROCESS_CPUTIME_ID 2 #define CLOCK_THREAD_CPUTIME_ID 3 #define CLOCK_MONOTONIC_RAW 4 +#define CLOCK_REALTIME_COARSE 5 +#define CLOCK_MONOTONIC_COARSE 6 /* * The IDs of various hardware clocks: diff --git a/kernel/posix-timers.c b/kernel/posix-timers.c index d089d052c4a9..495440779ce3 100644 --- a/kernel/posix-timers.c +++ b/kernel/posix-timers.c @@ -242,6 +242,25 @@ static int posix_get_monotonic_raw(clockid_t which_clock, struct timespec *tp) return 0; } + +static int posix_get_realtime_coarse(clockid_t which_clock, struct timespec *tp) +{ + *tp = current_kernel_time(); + return 0; +} + +static int posix_get_monotonic_coarse(clockid_t which_clock, + struct timespec *tp) +{ + *tp = get_monotonic_coarse(); + return 0; +} + +int posix_get_coarse_res(const clockid_t which_clock, struct timespec *tp) +{ + *tp = ktime_to_timespec(KTIME_LOW_RES); + return 0; +} /* * Initialize everything, well, just everything in Posix clocks/timers ;) */ @@ -262,10 +281,26 @@ static __init int init_posix_timers(void) .timer_create = no_timer_create, .nsleep = no_nsleep, }; + struct k_clock clock_realtime_coarse = { + .clock_getres = posix_get_coarse_res, + .clock_get = posix_get_realtime_coarse, + .clock_set = do_posix_clock_nosettime, + .timer_create = no_timer_create, + .nsleep = no_nsleep, + }; + struct k_clock clock_monotonic_coarse = { + .clock_getres = posix_get_coarse_res, + .clock_get = posix_get_monotonic_coarse, + .clock_set = do_posix_clock_nosettime, + .timer_create = no_timer_create, + .nsleep = no_nsleep, + }; register_posix_clock(CLOCK_REALTIME, &clock_realtime); register_posix_clock(CLOCK_MONOTONIC, &clock_monotonic); register_posix_clock(CLOCK_MONOTONIC_RAW, &clock_monotonic_raw); + register_posix_clock(CLOCK_REALTIME_COARSE, &clock_realtime_coarse); + register_posix_clock(CLOCK_MONOTONIC_COARSE, &clock_monotonic_coarse); posix_timers_cache = kmem_cache_create("posix_timers_cache", sizeof (struct k_itimer), 0, SLAB_PANIC, diff --git a/kernel/time/timekeeping.c b/kernel/time/timekeeping.c index 15e06defca55..03cbeb34d141 100644 --- a/kernel/time/timekeeping.c +++ b/kernel/time/timekeeping.c @@ -847,6 +847,10 @@ unsigned long get_seconds(void) } EXPORT_SYMBOL(get_seconds); +struct timespec __current_kernel_time(void) +{ + return xtime_cache; +} struct timespec current_kernel_time(void) { @@ -862,3 +866,20 @@ struct timespec current_kernel_time(void) return now; } EXPORT_SYMBOL(current_kernel_time); + +struct timespec get_monotonic_coarse(void) +{ + struct timespec now, mono; + unsigned long seq; + + do { + seq = read_seqbegin(&xtime_lock); + + now = xtime_cache; + mono = wall_to_monotonic; + } while (read_seqretry(&xtime_lock, seq)); + + set_normalized_timespec(&now, now.tv_sec + mono.tv_sec, + now.tv_nsec + mono.tv_nsec); + return now; +} -- cgit v1.2.3 From 8b5a10fc6fd02289ea03480f93382b1a99006142 Mon Sep 17 00:00:00 2001 From: Jan Beulich Date: Wed, 19 Aug 2009 08:40:48 +0100 Subject: x86: properly annotate alternatives.c Some of the NOPs tables aren't used on 64-bits, quite some code and data is needed post-init for module loading only, and a couple of functions aren't used outside that file (i.e. can be static, and don't need to be exported). The change to __INITDATA/__INITRODATA is needed to avoid an assembler warning. Signed-off-by: Jan Beulich LKML-Reference: <4A8BC8A00200007800010823@vpn.id2.novell.com> Acked-by: Sam Ravnborg Signed-off-by: H. Peter Anvin --- arch/x86/include/asm/alternative.h | 7 ----- arch/x86/kernel/alternative.c | 56 ++++++++++++++++++++++---------------- include/linux/init.h | 12 ++++++-- 3 files changed, 42 insertions(+), 33 deletions(-) (limited to 'include') diff --git a/arch/x86/include/asm/alternative.h b/arch/x86/include/asm/alternative.h index 1a37bcdc8606..c240efc74e00 100644 --- a/arch/x86/include/asm/alternative.h +++ b/arch/x86/include/asm/alternative.h @@ -73,8 +73,6 @@ static inline void alternatives_smp_module_del(struct module *mod) {} static inline void alternatives_smp_switch(int smp) {} #endif /* CONFIG_SMP */ -const unsigned char *const *find_nop_table(void); - /* alternative assembly primitive: */ #define ALTERNATIVE(oldinstr, newinstr, feature) \ \ @@ -144,8 +142,6 @@ static inline void apply_paravirt(struct paravirt_patch_site *start, #define __parainstructions_end NULL #endif -extern void add_nops(void *insns, unsigned int len); - /* * Clear and restore the kernel write-protection flag on the local CPU. * Allows the kernel to edit read-only pages. @@ -161,10 +157,7 @@ extern void add_nops(void *insns, unsigned int len); * Intel's errata. * On the local CPU you need to be protected again NMI or MCE handlers seeing an * inconsistent instruction while you patch. - * The _early version expects the memory to already be RW. */ - extern void *text_poke(void *addr, const void *opcode, size_t len); -extern void *text_poke_early(void *addr, const void *opcode, size_t len); #endif /* _ASM_X86_ALTERNATIVE_H */ diff --git a/arch/x86/kernel/alternative.c b/arch/x86/kernel/alternative.c index f57658702571..486935143e02 100644 --- a/arch/x86/kernel/alternative.c +++ b/arch/x86/kernel/alternative.c @@ -2,6 +2,7 @@ #include #include #include +#include #include #include #include @@ -32,7 +33,7 @@ __setup("smp-alt-boot", bootonly); #define smp_alt_once 1 #endif -static int debug_alternative; +static int __initdata_or_module debug_alternative; static int __init debug_alt(char *str) { @@ -51,7 +52,7 @@ static int __init setup_noreplace_smp(char *str) __setup("noreplace-smp", setup_noreplace_smp); #ifdef CONFIG_PARAVIRT -static int noreplace_paravirt = 0; +static int __initdata_or_module noreplace_paravirt = 0; static int __init setup_noreplace_paravirt(char *str) { @@ -64,16 +65,17 @@ __setup("noreplace-paravirt", setup_noreplace_paravirt); #define DPRINTK(fmt, args...) if (debug_alternative) \ printk(KERN_DEBUG fmt, args) -#ifdef GENERIC_NOP1 +#if defined(GENERIC_NOP1) && !defined(CONFIG_X86_64) /* Use inline assembly to define this because the nops are defined as inline assembly strings in the include files and we cannot get them easily into strings. */ -asm("\t.section .rodata, \"a\"\nintelnops: " +asm("\t" __stringify(__INITRODATA_OR_MODULE) "\nintelnops: " GENERIC_NOP1 GENERIC_NOP2 GENERIC_NOP3 GENERIC_NOP4 GENERIC_NOP5 GENERIC_NOP6 GENERIC_NOP7 GENERIC_NOP8 "\t.previous"); extern const unsigned char intelnops[]; -static const unsigned char *const intel_nops[ASM_NOP_MAX+1] = { +static const unsigned char *const __initconst_or_module +intel_nops[ASM_NOP_MAX+1] = { NULL, intelnops, intelnops + 1, @@ -87,12 +89,13 @@ static const unsigned char *const intel_nops[ASM_NOP_MAX+1] = { #endif #ifdef K8_NOP1 -asm("\t.section .rodata, \"a\"\nk8nops: " +asm("\t" __stringify(__INITRODATA_OR_MODULE) "\nk8nops: " K8_NOP1 K8_NOP2 K8_NOP3 K8_NOP4 K8_NOP5 K8_NOP6 K8_NOP7 K8_NOP8 "\t.previous"); extern const unsigned char k8nops[]; -static const unsigned char *const k8_nops[ASM_NOP_MAX+1] = { +static const unsigned char *const __initconst_or_module +k8_nops[ASM_NOP_MAX+1] = { NULL, k8nops, k8nops + 1, @@ -105,13 +108,14 @@ static const unsigned char *const k8_nops[ASM_NOP_MAX+1] = { }; #endif -#ifdef K7_NOP1 -asm("\t.section .rodata, \"a\"\nk7nops: " +#if defined(K7_NOP1) && !defined(CONFIG_X86_64) +asm("\t" __stringify(__INITRODATA_OR_MODULE) "\nk7nops: " K7_NOP1 K7_NOP2 K7_NOP3 K7_NOP4 K7_NOP5 K7_NOP6 K7_NOP7 K7_NOP8 "\t.previous"); extern const unsigned char k7nops[]; -static const unsigned char *const k7_nops[ASM_NOP_MAX+1] = { +static const unsigned char *const __initconst_or_module +k7_nops[ASM_NOP_MAX+1] = { NULL, k7nops, k7nops + 1, @@ -125,12 +129,13 @@ static const unsigned char *const k7_nops[ASM_NOP_MAX+1] = { #endif #ifdef P6_NOP1 -asm("\t.section .rodata, \"a\"\np6nops: " +asm("\t" __stringify(__INITRODATA_OR_MODULE) "\np6nops: " P6_NOP1 P6_NOP2 P6_NOP3 P6_NOP4 P6_NOP5 P6_NOP6 P6_NOP7 P6_NOP8 "\t.previous"); extern const unsigned char p6nops[]; -static const unsigned char *const p6_nops[ASM_NOP_MAX+1] = { +static const unsigned char *const __initconst_or_module +p6_nops[ASM_NOP_MAX+1] = { NULL, p6nops, p6nops + 1, @@ -146,7 +151,7 @@ static const unsigned char *const p6_nops[ASM_NOP_MAX+1] = { #ifdef CONFIG_X86_64 extern char __vsyscall_0; -const unsigned char *const *find_nop_table(void) +static const unsigned char *const *__init_or_module find_nop_table(void) { if (boot_cpu_data.x86_vendor == X86_VENDOR_INTEL && boot_cpu_has(X86_FEATURE_NOPL)) @@ -157,7 +162,7 @@ const unsigned char *const *find_nop_table(void) #else /* CONFIG_X86_64 */ -const unsigned char *const *find_nop_table(void) +static const unsigned char *const *__init_or_module find_nop_table(void) { if (boot_cpu_has(X86_FEATURE_K8)) return k8_nops; @@ -172,7 +177,7 @@ const unsigned char *const *find_nop_table(void) #endif /* CONFIG_X86_64 */ /* Use this to add nops to a buffer, then text_poke the whole buffer. */ -void add_nops(void *insns, unsigned int len) +static void __init_or_module add_nops(void *insns, unsigned int len) { const unsigned char *const *noptable = find_nop_table(); @@ -185,10 +190,10 @@ void add_nops(void *insns, unsigned int len) len -= noplen; } } -EXPORT_SYMBOL_GPL(add_nops); extern struct alt_instr __alt_instructions[], __alt_instructions_end[]; extern u8 *__smp_locks[], *__smp_locks_end[]; +static void *text_poke_early(void *addr, const void *opcode, size_t len); /* Replace instructions with better alternatives for this CPU type. This runs before SMP is initialized to avoid SMP problems with @@ -196,7 +201,8 @@ extern u8 *__smp_locks[], *__smp_locks_end[]; APs have less capabilities than the boot processor are not handled. Tough. Make sure you disable such features by hand. */ -void apply_alternatives(struct alt_instr *start, struct alt_instr *end) +void __init_or_module apply_alternatives(struct alt_instr *start, + struct alt_instr *end) { struct alt_instr *a; char insnbuf[MAX_PATCH_LEN]; @@ -279,9 +285,10 @@ static LIST_HEAD(smp_alt_modules); static DEFINE_MUTEX(smp_alt); static int smp_mode = 1; /* protected by smp_alt */ -void alternatives_smp_module_add(struct module *mod, char *name, - void *locks, void *locks_end, - void *text, void *text_end) +void __init_or_module alternatives_smp_module_add(struct module *mod, + char *name, + void *locks, void *locks_end, + void *text, void *text_end) { struct smp_alt_module *smp; @@ -317,7 +324,7 @@ void alternatives_smp_module_add(struct module *mod, char *name, mutex_unlock(&smp_alt); } -void alternatives_smp_module_del(struct module *mod) +void __init_or_module alternatives_smp_module_del(struct module *mod) { struct smp_alt_module *item; @@ -386,8 +393,8 @@ void alternatives_smp_switch(int smp) #endif #ifdef CONFIG_PARAVIRT -void apply_paravirt(struct paravirt_patch_site *start, - struct paravirt_patch_site *end) +void __init_or_module apply_paravirt(struct paravirt_patch_site *start, + struct paravirt_patch_site *end) { struct paravirt_patch_site *p; char insnbuf[MAX_PATCH_LEN]; @@ -485,7 +492,8 @@ void __init alternative_instructions(void) * instructions. And on the local CPU you need to be protected again NMI or MCE * handlers seeing an inconsistent instruction while you patch. */ -void *text_poke_early(void *addr, const void *opcode, size_t len) +static void *__init_or_module text_poke_early(void *addr, const void *opcode, + size_t len) { unsigned long flags; local_irq_save(flags); diff --git a/include/linux/init.h b/include/linux/init.h index 13b633ed695e..400adbb45414 100644 --- a/include/linux/init.h +++ b/include/linux/init.h @@ -103,8 +103,8 @@ #define __INIT .section ".init.text","ax" #define __FINIT .previous -#define __INITDATA .section ".init.data","aw" -#define __INITRODATA .section ".init.rodata","a" +#define __INITDATA .section ".init.data","aw",%progbits +#define __INITRODATA .section ".init.rodata","a",%progbits #define __FINITDATA .previous #define __DEVINIT .section ".devinit.text", "ax" @@ -305,9 +305,17 @@ void __init parse_early_options(char *cmdline); #ifdef CONFIG_MODULES #define __init_or_module #define __initdata_or_module +#define __initconst_or_module +#define __INIT_OR_MODULE .text +#define __INITDATA_OR_MODULE .data +#define __INITRODATA_OR_MODULE .section ".rodata","a",%progbits #else #define __init_or_module __init #define __initdata_or_module __initdata +#define __initconst_or_module __initconst +#define __INIT_OR_MODULE __INIT +#define __INITDATA_OR_MODULE __INITDATA +#define __INITRODATA_OR_MODULE __INITRODATA #endif /*CONFIG_MODULES*/ /* Functions marked as __devexit may be discarded at kernel link time, depending -- cgit v1.2.3 From b560d8ad8583803978aaaeba50ef29dc8e97a610 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Fri, 21 Aug 2009 22:08:51 -0700 Subject: rcu: Expunge lingering references to CONFIG_CLASSIC_RCU, optimize on !SMP A couple of references to CONFIG_CLASSIC_RCU have survived. Although these are harmless, it is past time for them to go. The one in hardirq.h is strictly a readability problem. The two in pagemap.h appear to disable a !SMP performance optimization (which this patch re-enables). This does raise the issue as to whether pagemap.h should really be referring to the CPU implementation. Long term, I intend to make the RCU implementation driven by CONFIG_PREEMPT, at which point these should change from defined(CONFIG_TREE_RCU) to !defined(CONFIG_PREEMPT). In the meantime, is there something else that could be done in pagemap.h? Signed-off-by: Paul E. McKenney Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <20090822050851.GA8414@linux.vnet.ibm.com> Signed-off-by: Ingo Molnar --- include/linux/hardirq.h | 4 ++-- include/linux/pagemap.h | 4 ++-- 2 files changed, 4 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/include/linux/hardirq.h b/include/linux/hardirq.h index 8246c697863d..330cb31bb496 100644 --- a/include/linux/hardirq.h +++ b/include/linux/hardirq.h @@ -132,7 +132,7 @@ static inline void account_system_vtime(struct task_struct *tsk) } #endif -#if defined(CONFIG_NO_HZ) && !defined(CONFIG_CLASSIC_RCU) +#if defined(CONFIG_NO_HZ) extern void rcu_irq_enter(void); extern void rcu_irq_exit(void); extern void rcu_nmi_enter(void); @@ -142,7 +142,7 @@ extern void rcu_nmi_exit(void); # define rcu_irq_exit() do { } while (0) # define rcu_nmi_enter() do { } while (0) # define rcu_nmi_exit() do { } while (0) -#endif /* #if defined(CONFIG_NO_HZ) && !defined(CONFIG_CLASSIC_RCU) */ +#endif /* #if defined(CONFIG_NO_HZ) */ /* * It is safe to do non-atomic ops on ->hardirq_context, diff --git a/include/linux/pagemap.h b/include/linux/pagemap.h index aec3252afcf5..ed5d7501e181 100644 --- a/include/linux/pagemap.h +++ b/include/linux/pagemap.h @@ -132,7 +132,7 @@ static inline int page_cache_get_speculative(struct page *page) { VM_BUG_ON(in_interrupt()); -#if !defined(CONFIG_SMP) && defined(CONFIG_CLASSIC_RCU) +#if !defined(CONFIG_SMP) && defined(CONFIG_TREE_RCU) # ifdef CONFIG_PREEMPT VM_BUG_ON(!in_atomic()); # endif @@ -170,7 +170,7 @@ static inline int page_cache_add_speculative(struct page *page, int count) { VM_BUG_ON(in_interrupt()); -#if !defined(CONFIG_SMP) && defined(CONFIG_CLASSIC_RCU) +#if !defined(CONFIG_SMP) && defined(CONFIG_TREE_RCU) # ifdef CONFIG_PREEMPT VM_BUG_ON(!in_atomic()); # endif -- cgit v1.2.3 From 9eba32b86d17ef87131fa0bce43c614904ab5781 Mon Sep 17 00:00:00 2001 From: Marcel Holtmann Date: Sat, 22 Aug 2009 14:19:26 -0700 Subject: Bluetooth: Add extra device reference counting for connections The device model itself has no real usable reference counting at the moment and this causes problems if parents are deleted before their children. The device model itself handles the memory details of this correctly, but the uevent order is not consistent. This causes various problems for systems like HAL or even X. So until device_put() does a proper cleanup, the device for Bluetooth connection will be protected with an extra reference counting to ensure the correct order of uevents when connections are terminated. This is not an automatic feature. Higher Bluetooth layers like HIDP or BNEP should grab this new reference to ensure that their uevents are send before the ones from the parent device. Based on a report by Brian Rogers Signed-off-by: Marcel Holtmann --- include/net/bluetooth/hci_core.h | 4 ++++ net/bluetooth/hci_conn.c | 17 ++++++++++++++++- net/bluetooth/hci_event.c | 2 ++ 3 files changed, 22 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/net/bluetooth/hci_core.h b/include/net/bluetooth/hci_core.h index c4ca4228b083..25b8a0345a6a 100644 --- a/include/net/bluetooth/hci_core.h +++ b/include/net/bluetooth/hci_core.h @@ -187,6 +187,7 @@ struct hci_conn { struct work_struct work_del; struct device dev; + atomic_t devref; struct hci_dev *hdev; void *l2cap_data; @@ -339,6 +340,9 @@ int hci_conn_switch_role(struct hci_conn *conn, __u8 role); void hci_conn_enter_active_mode(struct hci_conn *conn); void hci_conn_enter_sniff_mode(struct hci_conn *conn); +void hci_conn_hold_device(struct hci_conn *conn); +void hci_conn_put_device(struct hci_conn *conn); + static inline void hci_conn_hold(struct hci_conn *conn) { atomic_inc(&conn->refcnt); diff --git a/net/bluetooth/hci_conn.c b/net/bluetooth/hci_conn.c index fa47d5d84f5c..a9750984f772 100644 --- a/net/bluetooth/hci_conn.c +++ b/net/bluetooth/hci_conn.c @@ -246,6 +246,8 @@ struct hci_conn *hci_conn_add(struct hci_dev *hdev, int type, bdaddr_t *dst) if (hdev->notify) hdev->notify(hdev, HCI_NOTIFY_CONN_ADD); + atomic_set(&conn->devref, 0); + hci_conn_init_sysfs(conn); tasklet_enable(&hdev->tx_task); @@ -288,7 +290,7 @@ int hci_conn_del(struct hci_conn *conn) skb_queue_purge(&conn->data_q); - hci_conn_del_sysfs(conn); + hci_conn_put_device(conn); hci_dev_put(hdev); @@ -583,6 +585,19 @@ void hci_conn_check_pending(struct hci_dev *hdev) hci_dev_unlock(hdev); } +void hci_conn_hold_device(struct hci_conn *conn) +{ + atomic_inc(&conn->devref); +} +EXPORT_SYMBOL(hci_conn_hold_device); + +void hci_conn_put_device(struct hci_conn *conn) +{ + if (atomic_dec_and_test(&conn->devref)) + hci_conn_del_sysfs(conn); +} +EXPORT_SYMBOL(hci_conn_put_device); + int hci_get_conn_list(void __user *arg) { struct hci_conn_list_req req, *cl; diff --git a/net/bluetooth/hci_event.c b/net/bluetooth/hci_event.c index 184ba0a88ec0..e99fe385fba2 100644 --- a/net/bluetooth/hci_event.c +++ b/net/bluetooth/hci_event.c @@ -887,6 +887,7 @@ static inline void hci_conn_complete_evt(struct hci_dev *hdev, struct sk_buff *s } else conn->state = BT_CONNECTED; + hci_conn_hold_device(conn); hci_conn_add_sysfs(conn); if (test_bit(HCI_AUTH, &hdev->flags)) @@ -1693,6 +1694,7 @@ static inline void hci_sync_conn_complete_evt(struct hci_dev *hdev, struct sk_bu conn->handle = __le16_to_cpu(ev->handle); conn->state = BT_CONNECTED; + hci_conn_hold_device(conn); hci_conn_add_sysfs(conn); break; -- cgit v1.2.3 From a6a67efd7088702fdbbb780c5a3f8e1a74e77b63 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Sun, 26 Jul 2009 08:18:19 +0000 Subject: Bluetooth: Convert hdev->req_lock to a mutex hdev->req_lock is used as mutex so make it a mutex. Signed-off-by: Thomas Gleixner Signed-off-by: Marcel Holtmann --- include/net/bluetooth/hci_core.h | 6 +++--- net/bluetooth/hci_core.c | 2 +- 2 files changed, 4 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/include/net/bluetooth/hci_core.h b/include/net/bluetooth/hci_core.h index 25b8a0345a6a..7b640aeddb64 100644 --- a/include/net/bluetooth/hci_core.h +++ b/include/net/bluetooth/hci_core.h @@ -117,7 +117,7 @@ struct hci_dev { struct sk_buff *sent_cmd; struct sk_buff *reassembly[3]; - struct semaphore req_lock; + struct mutex req_lock; wait_queue_head_t req_wait_q; __u32 req_status; __u32 req_result; @@ -704,8 +704,8 @@ struct hci_sec_filter { #define HCI_REQ_PEND 1 #define HCI_REQ_CANCELED 2 -#define hci_req_lock(d) down(&d->req_lock) -#define hci_req_unlock(d) up(&d->req_lock) +#define hci_req_lock(d) mutex_lock(&d->req_lock) +#define hci_req_unlock(d) mutex_unlock(&d->req_lock) void hci_req_complete(struct hci_dev *hdev, int result); diff --git a/net/bluetooth/hci_core.c b/net/bluetooth/hci_core.c index 406ad07cdea1..e1da8f68759c 100644 --- a/net/bluetooth/hci_core.c +++ b/net/bluetooth/hci_core.c @@ -911,7 +911,7 @@ int hci_register_dev(struct hci_dev *hdev) hdev->reassembly[i] = NULL; init_waitqueue_head(&hdev->req_wait_q); - init_MUTEX(&hdev->req_lock); + mutex_init(&hdev->req_lock); inquiry_cache_init(hdev); -- cgit v1.2.3 From c6b03cf986eab00e20d0dbc852b233bb83472138 Mon Sep 17 00:00:00 2001 From: Marcel Holtmann Date: Sat, 2 May 2009 22:31:10 -0700 Subject: Bluetooth: Allow setting of L2CAP ERTM via socket option To enable Enhanced Retransmission mode it needs to be set via a socket option. A different mode can be set on a socket, but on listen() and connect() the mode is checked and ERTM is only allowed if it is enabled via the module parameter. Signed-off-by: Marcel Holtmann --- include/net/bluetooth/l2cap.h | 8 +++++--- net/bluetooth/l2cap.c | 37 +++++++++++++++++++++++++++++++++---- 2 files changed, 38 insertions(+), 7 deletions(-) (limited to 'include') diff --git a/include/net/bluetooth/l2cap.h b/include/net/bluetooth/l2cap.h index e919fca1072a..06b072fd6d54 100644 --- a/include/net/bluetooth/l2cap.h +++ b/include/net/bluetooth/l2cap.h @@ -190,7 +190,7 @@ struct l2cap_conf_rfc { #define L2CAP_MODE_RETRANS 0x01 #define L2CAP_MODE_FLOWCTL 0x02 #define L2CAP_MODE_ERTM 0x03 -#define L2CAP_MODE_STREAM 0x04 +#define L2CAP_MODE_STREAMING 0x04 struct l2cap_disconn_req { __le16 dcid; @@ -271,9 +271,11 @@ struct l2cap_pinfo { __u16 imtu; __u16 omtu; __u16 flush_to; - __u8 sec_level; + __u8 mode; + __u8 fcs; + __u8 sec_level; __u8 role_switch; - __u8 force_reliable; + __u8 force_reliable; __u8 conf_req[64]; __u8 conf_len; diff --git a/net/bluetooth/l2cap.c b/net/bluetooth/l2cap.c index 810a3c1a4188..8a59e57d9df1 100644 --- a/net/bluetooth/l2cap.c +++ b/net/bluetooth/l2cap.c @@ -717,12 +717,16 @@ static void l2cap_sock_init(struct sock *sk, struct sock *parent) pi->imtu = l2cap_pi(parent)->imtu; pi->omtu = l2cap_pi(parent)->omtu; + pi->mode = l2cap_pi(parent)->mode; + pi->fcs = l2cap_pi(parent)->fcs; pi->sec_level = l2cap_pi(parent)->sec_level; pi->role_switch = l2cap_pi(parent)->role_switch; pi->force_reliable = l2cap_pi(parent)->force_reliable; } else { pi->imtu = L2CAP_DEFAULT_MTU; pi->omtu = 0; + pi->mode = L2CAP_MODE_BASIC; + pi->fcs = L2CAP_FCS_CRC16; pi->sec_level = BT_SECURITY_LOW; pi->role_switch = 0; pi->force_reliable = 0; @@ -958,6 +962,18 @@ static int l2cap_sock_connect(struct socket *sock, struct sockaddr *addr, int al goto done; } + switch (l2cap_pi(sk)->mode) { + case L2CAP_MODE_BASIC: + break; + case L2CAP_MODE_ERTM: + if (enable_ertm) + break; + /* fall through */ + default: + err = -ENOTSUPP; + goto done; + } + switch (sk->sk_state) { case BT_CONNECT: case BT_CONNECT2: @@ -1009,6 +1025,18 @@ static int l2cap_sock_listen(struct socket *sock, int backlog) goto done; } + switch (l2cap_pi(sk)->mode) { + case L2CAP_MODE_BASIC: + break; + case L2CAP_MODE_ERTM: + if (enable_ertm) + break; + /* fall through */ + default: + err = -ENOTSUPP; + goto done; + } + if (!l2cap_pi(sk)->psm) { bdaddr_t *src = &bt_sk(sk)->src; u16 psm; @@ -1259,7 +1287,7 @@ static int l2cap_sock_setsockopt_old(struct socket *sock, int optname, char __us opts.imtu = l2cap_pi(sk)->imtu; opts.omtu = l2cap_pi(sk)->omtu; opts.flush_to = l2cap_pi(sk)->flush_to; - opts.mode = L2CAP_MODE_BASIC; + opts.mode = l2cap_pi(sk)->mode; len = min_t(unsigned int, sizeof(opts), optlen); if (copy_from_user((char *) &opts, optval, len)) { @@ -1267,8 +1295,9 @@ static int l2cap_sock_setsockopt_old(struct socket *sock, int optname, char __us break; } - l2cap_pi(sk)->imtu = opts.imtu; - l2cap_pi(sk)->omtu = opts.omtu; + l2cap_pi(sk)->imtu = opts.imtu; + l2cap_pi(sk)->omtu = opts.omtu; + l2cap_pi(sk)->mode = opts.mode; break; case L2CAP_LM: @@ -1381,7 +1410,7 @@ static int l2cap_sock_getsockopt_old(struct socket *sock, int optname, char __us opts.imtu = l2cap_pi(sk)->imtu; opts.omtu = l2cap_pi(sk)->omtu; opts.flush_to = l2cap_pi(sk)->flush_to; - opts.mode = L2CAP_MODE_BASIC; + opts.mode = l2cap_pi(sk)->mode; len = min_t(unsigned int, len, sizeof(opts)); if (copy_to_user(optval, (char *) &opts, len)) -- cgit v1.2.3 From f2fcfcd670257236ebf2088bbdf26f6a8ef459fe Mon Sep 17 00:00:00 2001 From: "Gustavo F. Padovan" Date: Sat, 4 Jul 2009 15:06:24 -0300 Subject: Bluetooth: Add configuration support for ERTM and Streaming mode Add support to config_req and config_rsp to configure ERTM and Streaming mode. If the remote device specifies ERTM or Streaming mode, then the same mode is proposed. Otherwise ERTM or Basic mode is used. And in case of a state 2 device, the remote device should propose the same mode. If not, then the channel gets disconnected. Signed-off-by: Gustavo F. Padovan Signed-off-by: Marcel Holtmann --- include/net/bluetooth/l2cap.h | 28 +++-- net/bluetooth/l2cap.c | 262 +++++++++++++++++++++++++++++++++++++----- 2 files changed, 255 insertions(+), 35 deletions(-) (limited to 'include') diff --git a/include/net/bluetooth/l2cap.h b/include/net/bluetooth/l2cap.h index 06b072fd6d54..6fc76986d70a 100644 --- a/include/net/bluetooth/l2cap.h +++ b/include/net/bluetooth/l2cap.h @@ -27,8 +27,9 @@ /* L2CAP defaults */ #define L2CAP_DEFAULT_MTU 672 +#define L2CAP_DEFAULT_MIN_MTU 48 #define L2CAP_DEFAULT_FLUSH_TO 0xffff -#define L2CAP_DEFAULT_RX_WINDOW 1 +#define L2CAP_DEFAULT_TX_WINDOW 1 #define L2CAP_DEFAULT_MAX_RECEIVE 1 #define L2CAP_DEFAULT_RETRANS_TO 300 /* 300 milliseconds */ #define L2CAP_DEFAULT_MONITOR_TO 1000 /* 1 second */ @@ -272,6 +273,9 @@ struct l2cap_pinfo { __u16 omtu; __u16 flush_to; __u8 mode; + __u8 num_conf_req; + __u8 num_conf_rsp; + __u8 fcs; __u8 sec_level; __u8 role_switch; @@ -280,10 +284,15 @@ struct l2cap_pinfo { __u8 conf_req[64]; __u8 conf_len; __u8 conf_state; - __u8 conf_retry; __u8 ident; + __u8 remote_tx_win; + __u8 remote_max_tx; + __u16 retrans_timeout; + __u16 monitor_timeout; + __u16 max_pdu_size; + __le16 sport; struct l2cap_conn *conn; @@ -291,12 +300,17 @@ struct l2cap_pinfo { struct sock *prev_c; }; -#define L2CAP_CONF_REQ_SENT 0x01 -#define L2CAP_CONF_INPUT_DONE 0x02 -#define L2CAP_CONF_OUTPUT_DONE 0x04 -#define L2CAP_CONF_CONNECT_PEND 0x80 +#define L2CAP_CONF_REQ_SENT 0x01 +#define L2CAP_CONF_INPUT_DONE 0x02 +#define L2CAP_CONF_OUTPUT_DONE 0x04 +#define L2CAP_CONF_MTU_DONE 0x08 +#define L2CAP_CONF_MODE_DONE 0x10 +#define L2CAP_CONF_CONNECT_PEND 0x20 +#define L2CAP_CONF_STATE2_DEVICE 0x80 + +#define L2CAP_CONF_MAX_CONF_REQ 2 +#define L2CAP_CONF_MAX_CONF_RSP 2 -#define L2CAP_CONF_MAX_RETRIES 2 void l2cap_load(void); diff --git a/net/bluetooth/l2cap.c b/net/bluetooth/l2cap.c index 7ce1a24735c8..af0fbf9ebfeb 100644 --- a/net/bluetooth/l2cap.c +++ b/net/bluetooth/l2cap.c @@ -966,6 +966,7 @@ static int l2cap_sock_connect(struct socket *sock, struct sockaddr *addr, int al case L2CAP_MODE_BASIC: break; case L2CAP_MODE_ERTM: + case L2CAP_MODE_STREAMING: if (enable_ertm) break; /* fall through */ @@ -1029,6 +1030,7 @@ static int l2cap_sock_listen(struct socket *sock, int backlog) case L2CAP_MODE_BASIC: break; case L2CAP_MODE_ERTM: + case L2CAP_MODE_STREAMING: if (enable_ertm) break; /* fall through */ @@ -1739,15 +1741,65 @@ static void l2cap_add_conf_opt(void **ptr, u8 type, u8 len, unsigned long val) *ptr += L2CAP_CONF_OPT_SIZE + len; } +static int l2cap_mode_supported(__u8 mode, __u32 feat_mask) +{ + u32 local_feat_mask = l2cap_feat_mask; + if (enable_ertm) + local_feat_mask |= L2CAP_FEAT_ERTM; + + switch (mode) { + case L2CAP_MODE_ERTM: + return L2CAP_FEAT_ERTM & feat_mask & local_feat_mask; + case L2CAP_MODE_STREAMING: + return L2CAP_FEAT_STREAMING & feat_mask & local_feat_mask; + default: + return 0x00; + } +} + +static inline __u8 l2cap_select_mode(__u8 mode, __u16 remote_feat_mask) +{ + switch (mode) { + case L2CAP_MODE_STREAMING: + case L2CAP_MODE_ERTM: + if (l2cap_mode_supported(mode, remote_feat_mask)) + return mode; + /* fall through */ + default: + return L2CAP_MODE_BASIC; + } +} + static int l2cap_build_conf_req(struct sock *sk, void *data) { struct l2cap_pinfo *pi = l2cap_pi(sk); struct l2cap_conf_req *req = data; - struct l2cap_conf_rfc rfc = { .mode = L2CAP_MODE_BASIC }; + struct l2cap_conf_rfc rfc = { .mode = L2CAP_MODE_ERTM }; void *ptr = req->data; BT_DBG("sk %p", sk); + if (pi->num_conf_req || pi->num_conf_rsp) + goto done; + + switch (pi->mode) { + case L2CAP_MODE_STREAMING: + case L2CAP_MODE_ERTM: + pi->conf_state |= L2CAP_CONF_STATE2_DEVICE; + if (!l2cap_mode_supported(pi->mode, pi->conn->feat_mask)) { + struct l2cap_disconn_req req; + req.dcid = cpu_to_le16(pi->dcid); + req.scid = cpu_to_le16(pi->scid); + l2cap_send_cmd(pi->conn, l2cap_get_ident(pi->conn), + L2CAP_DISCONN_REQ, sizeof(req), &req); + } + break; + default: + pi->mode = l2cap_select_mode(rfc.mode, pi->conn->feat_mask); + break; + } + +done: switch (pi->mode) { case L2CAP_MODE_BASIC: if (pi->imtu != L2CAP_DEFAULT_MTU) @@ -1756,10 +1808,22 @@ static int l2cap_build_conf_req(struct sock *sk, void *data) case L2CAP_MODE_ERTM: rfc.mode = L2CAP_MODE_ERTM; - rfc.txwin_size = L2CAP_DEFAULT_RX_WINDOW; + rfc.txwin_size = L2CAP_DEFAULT_TX_WINDOW; rfc.max_transmit = L2CAP_DEFAULT_MAX_RECEIVE; - rfc.retrans_timeout = cpu_to_le16(L2CAP_DEFAULT_RETRANS_TO); - rfc.monitor_timeout = cpu_to_le16(L2CAP_DEFAULT_MONITOR_TO); + rfc.retrans_timeout = 0; + rfc.monitor_timeout = 0; + rfc.max_pdu_size = cpu_to_le16(L2CAP_DEFAULT_MAX_RX_APDU); + + l2cap_add_conf_opt(&ptr, L2CAP_CONF_RFC, + sizeof(rfc), (unsigned long) &rfc); + break; + + case L2CAP_MODE_STREAMING: + rfc.mode = L2CAP_MODE_STREAMING; + rfc.txwin_size = 0; + rfc.max_transmit = 0; + rfc.retrans_timeout = 0; + rfc.monitor_timeout = 0; rfc.max_pdu_size = cpu_to_le16(L2CAP_DEFAULT_MAX_RX_APDU); l2cap_add_conf_opt(&ptr, L2CAP_CONF_RFC, @@ -1825,30 +1889,83 @@ static int l2cap_parse_conf_req(struct sock *sk, void *data) } } + if (pi->num_conf_rsp || pi->num_conf_req) + goto done; + + switch (pi->mode) { + case L2CAP_MODE_STREAMING: + case L2CAP_MODE_ERTM: + pi->conf_state |= L2CAP_CONF_STATE2_DEVICE; + if (!l2cap_mode_supported(pi->mode, pi->conn->feat_mask)) + return -ECONNREFUSED; + break; + default: + pi->mode = l2cap_select_mode(rfc.mode, pi->conn->feat_mask); + break; + } + +done: + if (pi->mode != rfc.mode) { + result = L2CAP_CONF_UNACCEPT; + rfc.mode = pi->mode; + + if (pi->num_conf_rsp == 1) + return -ECONNREFUSED; + + l2cap_add_conf_opt(&ptr, L2CAP_CONF_RFC, + sizeof(rfc), (unsigned long) &rfc); + } + + if (result == L2CAP_CONF_SUCCESS) { /* Configure output options and let the other side know * which ones we don't like. */ - if (rfc.mode == L2CAP_MODE_BASIC) { - if (mtu < pi->omtu) - result = L2CAP_CONF_UNACCEPT; - else { - pi->omtu = mtu; - pi->conf_state |= L2CAP_CONF_OUTPUT_DONE; - } + if (mtu < L2CAP_DEFAULT_MIN_MTU) + result = L2CAP_CONF_UNACCEPT; + else { + pi->omtu = mtu; + pi->conf_state |= L2CAP_CONF_MTU_DONE; + } + l2cap_add_conf_opt(&ptr, L2CAP_CONF_MTU, 2, pi->omtu); - l2cap_add_conf_opt(&ptr, L2CAP_CONF_MTU, 2, pi->omtu); - } else { + switch (rfc.mode) { + case L2CAP_MODE_BASIC: + pi->fcs = L2CAP_FCS_NONE; + pi->conf_state |= L2CAP_CONF_MODE_DONE; + break; + + case L2CAP_MODE_ERTM: + pi->remote_tx_win = rfc.txwin_size; + pi->remote_max_tx = rfc.max_transmit; + pi->max_pdu_size = rfc.max_pdu_size; + + rfc.retrans_timeout = L2CAP_DEFAULT_RETRANS_TO; + rfc.monitor_timeout = L2CAP_DEFAULT_MONITOR_TO; + + pi->conf_state |= L2CAP_CONF_MODE_DONE; + break; + + case L2CAP_MODE_STREAMING: + pi->remote_tx_win = rfc.txwin_size; + pi->max_pdu_size = rfc.max_pdu_size; + + pi->conf_state |= L2CAP_CONF_MODE_DONE; + break; + + default: result = L2CAP_CONF_UNACCEPT; memset(&rfc, 0, sizeof(rfc)); - rfc.mode = L2CAP_MODE_BASIC; + rfc.mode = pi->mode; + } - l2cap_add_conf_opt(&ptr, L2CAP_CONF_RFC, + l2cap_add_conf_opt(&ptr, L2CAP_CONF_RFC, sizeof(rfc), (unsigned long) &rfc); - } - } + if (result == L2CAP_CONF_SUCCESS) + pi->conf_state |= L2CAP_CONF_OUTPUT_DONE; + } rsp->scid = cpu_to_le16(pi->dcid); rsp->result = cpu_to_le16(result); rsp->flags = cpu_to_le16(0x0000); @@ -1856,6 +1973,73 @@ static int l2cap_parse_conf_req(struct sock *sk, void *data) return ptr - data; } +static int l2cap_parse_conf_rsp(struct sock *sk, void *rsp, int len, void *data, u16 *result) +{ + struct l2cap_pinfo *pi = l2cap_pi(sk); + struct l2cap_conf_req *req = data; + void *ptr = req->data; + int type, olen; + unsigned long val; + struct l2cap_conf_rfc rfc; + + BT_DBG("sk %p, rsp %p, len %d, req %p", sk, rsp, len, data); + + while (len >= L2CAP_CONF_OPT_SIZE) { + len -= l2cap_get_conf_opt(&rsp, &type, &olen, &val); + + switch (type) { + case L2CAP_CONF_MTU: + if (val < L2CAP_DEFAULT_MIN_MTU) { + *result = L2CAP_CONF_UNACCEPT; + pi->omtu = L2CAP_DEFAULT_MIN_MTU; + } else + pi->omtu = val; + l2cap_add_conf_opt(&ptr, L2CAP_CONF_MTU, 2, pi->omtu); + break; + + case L2CAP_CONF_FLUSH_TO: + pi->flush_to = val; + l2cap_add_conf_opt(&ptr, L2CAP_CONF_FLUSH_TO, + 2, pi->flush_to); + break; + + case L2CAP_CONF_RFC: + if (olen == sizeof(rfc)) + memcpy(&rfc, (void *)val, olen); + + if ((pi->conf_state & L2CAP_CONF_STATE2_DEVICE) && + rfc.mode != pi->mode) + return -ECONNREFUSED; + + pi->mode = rfc.mode; + pi->fcs = 0; + + l2cap_add_conf_opt(&ptr, L2CAP_CONF_RFC, + sizeof(rfc), (unsigned long) &rfc); + break; + } + } + + if (*result == L2CAP_CONF_SUCCESS) { + switch (rfc.mode) { + case L2CAP_MODE_ERTM: + pi->remote_tx_win = rfc.txwin_size; + pi->retrans_timeout = rfc.retrans_timeout; + pi->monitor_timeout = rfc.monitor_timeout; + pi->max_pdu_size = le16_to_cpu(rfc.max_pdu_size); + break; + case L2CAP_MODE_STREAMING: + pi->max_pdu_size = le16_to_cpu(rfc.max_pdu_size); + break; + } + } + + req->dcid = cpu_to_le16(pi->dcid); + req->flags = cpu_to_le16(0x0000); + + return ptr - data; +} + static int l2cap_build_conf_rsp(struct sock *sk, void *data, u16 result, u16 flags) { struct l2cap_conf_rsp *rsp = data; @@ -2042,6 +2226,7 @@ static inline int l2cap_connect_rsp(struct l2cap_conn *conn, struct l2cap_cmd_hd l2cap_send_cmd(conn, l2cap_get_ident(conn), L2CAP_CONF_REQ, l2cap_build_conf_req(sk, req), req); + l2cap_pi(sk)->num_conf_req++; break; case L2CAP_CR_PEND: @@ -2100,10 +2285,17 @@ static inline int l2cap_config_req(struct l2cap_conn *conn, struct l2cap_cmd_hdr /* Complete config. */ len = l2cap_parse_conf_req(sk, rsp); - if (len < 0) + if (len < 0) { + struct l2cap_disconn_req req; + req.dcid = cpu_to_le16(l2cap_pi(sk)->dcid); + req.scid = cpu_to_le16(l2cap_pi(sk)->scid); + l2cap_send_cmd(conn, l2cap_get_ident(conn), + L2CAP_DISCONN_REQ, sizeof(req), &req); goto unlock; + } l2cap_send_cmd(conn, cmd->ident, L2CAP_CONF_RSP, len, rsp); + l2cap_pi(sk)->num_conf_rsp++; /* Reset config buffer. */ l2cap_pi(sk)->conf_len = 0; @@ -2121,6 +2313,7 @@ static inline int l2cap_config_req(struct l2cap_conn *conn, struct l2cap_cmd_hdr u8 buf[64]; l2cap_send_cmd(conn, l2cap_get_ident(conn), L2CAP_CONF_REQ, l2cap_build_conf_req(sk, buf), buf); + l2cap_pi(sk)->num_conf_req++; } unlock: @@ -2150,16 +2343,29 @@ static inline int l2cap_config_rsp(struct l2cap_conn *conn, struct l2cap_cmd_hdr break; case L2CAP_CONF_UNACCEPT: - if (++l2cap_pi(sk)->conf_retry < L2CAP_CONF_MAX_RETRIES) { - char req[128]; - /* It does not make sense to adjust L2CAP parameters - * that are currently defined in the spec. We simply - * resend config request that we sent earlier. It is - * stupid, but it helps qualification testing which - * expects at least some response from us. */ - l2cap_send_cmd(conn, l2cap_get_ident(conn), L2CAP_CONF_REQ, - l2cap_build_conf_req(sk, req), req); - goto done; + if (l2cap_pi(sk)->num_conf_rsp <= L2CAP_CONF_MAX_CONF_RSP) { + int len = cmd->len - sizeof(*rsp); + char req[64]; + + /* throw out any old stored conf requests */ + result = L2CAP_CONF_SUCCESS; + len = l2cap_parse_conf_rsp(sk, rsp->data, + len, req, &result); + if (len < 0) { + struct l2cap_disconn_req req; + req.dcid = cpu_to_le16(l2cap_pi(sk)->dcid); + req.scid = cpu_to_le16(l2cap_pi(sk)->scid); + l2cap_send_cmd(conn, l2cap_get_ident(conn), + L2CAP_DISCONN_REQ, sizeof(req), &req); + goto done; + } + + l2cap_send_cmd(conn, l2cap_get_ident(conn), + L2CAP_CONF_REQ, len, req); + l2cap_pi(sk)->num_conf_req++; + if (result != L2CAP_CONF_SUCCESS) + goto done; + break; } default: -- cgit v1.2.3 From 1c2acffb76d4bc5fd27c4ea55cc27ad8ead10f9a Mon Sep 17 00:00:00 2001 From: "Gustavo F. Padovan" Date: Thu, 20 Aug 2009 22:25:57 -0300 Subject: Bluetooth: Add initial support for ERTM packets transfers This patch adds support for ERTM transfers, without retransmission, with txWindow up to 63 and with acknowledgement of packets received. Now the packets are queued before call l2cap_do_send(), so packets couldn't be sent at the time we call l2cap_sock_sendmsg(). They will be sent in an asynchronous way on later calls of l2cap_ertm_send(). Besides if an error occurs on calling l2cap_do_send() we disconnect the channel. Initially based on a patch from Nathan Holstein Signed-off-by: Gustavo F. Padovan Signed-off-by: Marcel Holtmann --- include/net/bluetooth/bluetooth.h | 3 +- include/net/bluetooth/l2cap.h | 54 +++++- net/bluetooth/l2cap.c | 384 +++++++++++++++++++++++++++++++++----- 3 files changed, 390 insertions(+), 51 deletions(-) (limited to 'include') diff --git a/include/net/bluetooth/bluetooth.h b/include/net/bluetooth/bluetooth.h index 968166a45f86..65a5cf868fd6 100644 --- a/include/net/bluetooth/bluetooth.h +++ b/include/net/bluetooth/bluetooth.h @@ -138,8 +138,9 @@ struct sock *bt_accept_dequeue(struct sock *parent, struct socket *newsock); struct bt_skb_cb { __u8 pkt_type; __u8 incoming; + __u8 tx_seq; }; -#define bt_cb(skb) ((struct bt_skb_cb *)(skb->cb)) +#define bt_cb(skb) ((struct bt_skb_cb *)((skb)->cb)) static inline struct sk_buff *bt_skb_alloc(unsigned int len, gfp_t how) { diff --git a/include/net/bluetooth/l2cap.h b/include/net/bluetooth/l2cap.h index 6fc76986d70a..9bbfbe74d400 100644 --- a/include/net/bluetooth/l2cap.h +++ b/include/net/bluetooth/l2cap.h @@ -29,7 +29,8 @@ #define L2CAP_DEFAULT_MTU 672 #define L2CAP_DEFAULT_MIN_MTU 48 #define L2CAP_DEFAULT_FLUSH_TO 0xffff -#define L2CAP_DEFAULT_TX_WINDOW 1 +#define L2CAP_DEFAULT_TX_WINDOW 63 +#define L2CAP_DEFAULT_NUM_TO_ACK (L2CAP_DEFAULT_TX_WINDOW/5) #define L2CAP_DEFAULT_MAX_RECEIVE 1 #define L2CAP_DEFAULT_RETRANS_TO 300 /* 300 milliseconds */ #define L2CAP_DEFAULT_MONITOR_TO 1000 /* 1 second */ @@ -94,6 +95,31 @@ struct l2cap_conninfo { #define L2CAP_FCS_NONE 0x00 #define L2CAP_FCS_CRC16 0x01 +/* L2CAP Control Field bit masks */ +#define L2CAP_CTRL_SAR 0xC000 +#define L2CAP_CTRL_REQSEQ 0x3F00 +#define L2CAP_CTRL_TXSEQ 0x007E +#define L2CAP_CTRL_RETRANS 0x0080 +#define L2CAP_CTRL_FINAL 0x0080 +#define L2CAP_CTRL_POLL 0x0010 +#define L2CAP_CTRL_SUPERVISE 0x000C +#define L2CAP_CTRL_FRAME_TYPE 0x0001 /* I- or S-Frame */ + +#define L2CAP_CTRL_TXSEQ_SHIFT 1 +#define L2CAP_CTRL_REQSEQ_SHIFT 8 + +/* L2CAP Supervisory Function */ +#define L2CAP_SUPER_RCV_READY 0x0000 +#define L2CAP_SUPER_REJECT 0x0004 +#define L2CAP_SUPER_RCV_NOT_READY 0x0008 +#define L2CAP_SUPER_SELECT_REJECT 0x000C + +/* L2CAP Segmentation and Reassembly */ +#define L2CAP_SDU_UNSEGMENTED 0x0000 +#define L2CAP_SDU_START 0x4000 +#define L2CAP_SDU_END 0x8000 +#define L2CAP_SDU_CONTINUE 0xC000 + /* L2CAP structures */ struct l2cap_hdr { __le16 len; @@ -262,6 +288,7 @@ struct l2cap_conn { /* ----- L2CAP channel and socket info ----- */ #define l2cap_pi(sk) ((struct l2cap_pinfo *) sk) +#define TX_QUEUE(sk) (&l2cap_pi(sk)->tx_queue) struct l2cap_pinfo { struct bt_sock bt; @@ -285,6 +312,13 @@ struct l2cap_pinfo { __u8 conf_len; __u8 conf_state; + __u8 next_tx_seq; + __u8 expected_ack_seq; + __u8 req_seq; + __u8 expected_tx_seq; + __u8 unacked_frames; + __u8 num_to_ack; + __u8 ident; __u8 remote_tx_win; @@ -295,6 +329,7 @@ struct l2cap_pinfo { __le16 sport; + struct sk_buff_head tx_queue; struct l2cap_conn *conn; struct sock *next_c; struct sock *prev_c; @@ -311,6 +346,23 @@ struct l2cap_pinfo { #define L2CAP_CONF_MAX_CONF_REQ 2 #define L2CAP_CONF_MAX_CONF_RSP 2 +static inline int l2cap_tx_window_full(struct sock *sk) +{ + struct l2cap_pinfo *pi = l2cap_pi(sk); + int sub; + + sub = (pi->next_tx_seq - pi->expected_ack_seq) % 64; + + if (sub < 0) + sub += 64; + + return (sub == pi->remote_tx_win); +} + +#define __get_txseq(ctrl) ((ctrl) & L2CAP_CTRL_TXSEQ) >> 1 +#define __get_reqseq(ctrl) ((ctrl) & L2CAP_CTRL_REQSEQ) >> 8 +#define __is_iframe(ctrl) !((ctrl) & L2CAP_CTRL_FRAME_TYPE) +#define __is_sframe(ctrl) (ctrl) & L2CAP_CTRL_FRAME_TYPE void l2cap_load(void); diff --git a/net/bluetooth/l2cap.c b/net/bluetooth/l2cap.c index c1b562085cb4..45b8697a7398 100644 --- a/net/bluetooth/l2cap.c +++ b/net/bluetooth/l2cap.c @@ -333,6 +333,30 @@ static inline int l2cap_send_cmd(struct l2cap_conn *conn, u8 ident, u8 code, u16 return hci_send_acl(conn->hcon, skb, 0); } +static inline int l2cap_send_sframe(struct l2cap_pinfo *pi, u16 control) +{ + struct sk_buff *skb; + struct l2cap_hdr *lh; + struct l2cap_conn *conn = pi->conn; + int count; + + BT_DBG("pi %p, control 0x%2.2x", pi, control); + + count = min_t(unsigned int, conn->mtu, L2CAP_HDR_SIZE + 2); + control |= L2CAP_CTRL_FRAME_TYPE; + + skb = bt_skb_alloc(count, GFP_ATOMIC); + if (!skb) + return -ENOMEM; + + lh = (struct l2cap_hdr *) skb_put(skb, L2CAP_HDR_SIZE); + lh->len = cpu_to_le16(2); + lh->cid = cpu_to_le16(pi->dcid); + put_unaligned_le16(control, skb_put(skb, 2)); + + return hci_send_acl(pi->conn->hcon, skb, 0); +} + static void l2cap_do_start(struct sock *sk) { struct l2cap_conn *conn = l2cap_pi(sk)->conn; @@ -1154,39 +1178,80 @@ static int l2cap_sock_getname(struct socket *sock, struct sockaddr *addr, int *l return 0; } -static inline int l2cap_do_send(struct sock *sk, struct msghdr *msg, int len) +static void l2cap_drop_acked_frames(struct sock *sk) { - struct l2cap_conn *conn = l2cap_pi(sk)->conn; - struct sk_buff *skb, **frag; - int err, hlen, count, sent = 0; - struct l2cap_hdr *lh; + struct sk_buff *skb; - BT_DBG("sk %p len %d", sk, len); + while ((skb = skb_peek(TX_QUEUE(sk)))) { + if (bt_cb(skb)->tx_seq == l2cap_pi(sk)->expected_ack_seq) + break; - /* First fragment (with L2CAP header) */ - if (sk->sk_type == SOCK_DGRAM) - hlen = L2CAP_HDR_SIZE + 2; - else - hlen = L2CAP_HDR_SIZE; + skb = skb_dequeue(TX_QUEUE(sk)); + kfree_skb(skb); - count = min_t(unsigned int, (conn->mtu - hlen), len); + l2cap_pi(sk)->unacked_frames--; + } - skb = bt_skb_send_alloc(sk, hlen + count, - msg->msg_flags & MSG_DONTWAIT, &err); - if (!skb) - return err; + return; +} - /* Create L2CAP header */ - lh = (struct l2cap_hdr *) skb_put(skb, L2CAP_HDR_SIZE); - lh->cid = cpu_to_le16(l2cap_pi(sk)->dcid); - lh->len = cpu_to_le16(len + (hlen - L2CAP_HDR_SIZE)); +static inline int l2cap_do_send(struct sock *sk, struct sk_buff *skb) +{ + struct l2cap_pinfo *pi = l2cap_pi(sk); + int err; + + BT_DBG("sk %p, skb %p len %d", sk, skb, skb->len); + + err = hci_send_acl(pi->conn->hcon, skb, 0); + if (err < 0) + kfree_skb(skb); + + return err; +} + +static int l2cap_ertm_send(struct sock *sk) +{ + struct sk_buff *skb, *tx_skb; + struct l2cap_pinfo *pi = l2cap_pi(sk); + u16 control; + int err; + + while ((skb = sk->sk_send_head) && (!l2cap_tx_window_full(sk))) { + tx_skb = skb_clone(skb, GFP_ATOMIC); - if (sk->sk_type == SOCK_DGRAM) - put_unaligned(l2cap_pi(sk)->psm, (__le16 *) skb_put(skb, 2)); + control = get_unaligned_le16(tx_skb->data + L2CAP_HDR_SIZE); + control |= (pi->req_seq << L2CAP_CTRL_REQSEQ_SHIFT) + | (pi->next_tx_seq << L2CAP_CTRL_TXSEQ_SHIFT); + put_unaligned_le16(control, tx_skb->data + L2CAP_HDR_SIZE); + + err = l2cap_do_send(sk, tx_skb); + if (err < 0) { + l2cap_send_disconn_req(pi->conn, sk); + return err; + } + + bt_cb(skb)->tx_seq = pi->next_tx_seq; + pi->next_tx_seq = (pi->next_tx_seq + 1) % 64; + + pi->unacked_frames++; + + if (skb_queue_is_last(TX_QUEUE(sk), skb)) + sk->sk_send_head = NULL; + else + sk->sk_send_head = skb_queue_next(TX_QUEUE(sk), skb); + } + + return 0; +} + +static inline int l2cap_skbuff_fromiovec(struct sock *sk, struct msghdr *msg, int len, int count, struct sk_buff *skb) +{ + struct l2cap_conn *conn = l2cap_pi(sk)->conn; + struct sk_buff **frag; + int err, sent = 0; if (memcpy_fromiovec(skb_put(skb, count), msg->msg_iov, count)) { - err = -EFAULT; - goto fail; + return -EFAULT; } sent += count; @@ -1199,33 +1264,112 @@ static inline int l2cap_do_send(struct sock *sk, struct msghdr *msg, int len) *frag = bt_skb_send_alloc(sk, count, msg->msg_flags & MSG_DONTWAIT, &err); if (!*frag) - goto fail; - - if (memcpy_fromiovec(skb_put(*frag, count), msg->msg_iov, count)) { - err = -EFAULT; - goto fail; - } + return -EFAULT; + if (memcpy_fromiovec(skb_put(*frag, count), msg->msg_iov, count)) + return -EFAULT; sent += count; len -= count; frag = &(*frag)->next; } - err = hci_send_acl(conn->hcon, skb, 0); - if (err < 0) - goto fail; return sent; +} -fail: - kfree_skb(skb); - return err; +static struct sk_buff *l2cap_create_connless_pdu(struct sock *sk, struct msghdr *msg, size_t len) +{ + struct l2cap_conn *conn = l2cap_pi(sk)->conn; + struct sk_buff *skb; + int err, count, hlen = L2CAP_HDR_SIZE + 2; + struct l2cap_hdr *lh; + + BT_DBG("sk %p len %d", sk, (int)len); + + count = min_t(unsigned int, (conn->mtu - hlen), len); + skb = bt_skb_send_alloc(sk, count + hlen, + msg->msg_flags & MSG_DONTWAIT, &err); + if (!skb) + return ERR_PTR(-ENOMEM); + + /* Create L2CAP header */ + lh = (struct l2cap_hdr *) skb_put(skb, L2CAP_HDR_SIZE); + lh->cid = cpu_to_le16(l2cap_pi(sk)->dcid); + lh->len = cpu_to_le16(len + (hlen - L2CAP_HDR_SIZE)); + put_unaligned_le16(l2cap_pi(sk)->psm, skb_put(skb, 2)); + + err = l2cap_skbuff_fromiovec(sk, msg, len, count, skb); + if (unlikely(err < 0)) { + kfree_skb(skb); + return ERR_PTR(err); + } + return skb; +} + +static struct sk_buff *l2cap_create_basic_pdu(struct sock *sk, struct msghdr *msg, size_t len) +{ + struct l2cap_conn *conn = l2cap_pi(sk)->conn; + struct sk_buff *skb; + int err, count, hlen = L2CAP_HDR_SIZE; + struct l2cap_hdr *lh; + + BT_DBG("sk %p len %d", sk, (int)len); + + count = min_t(unsigned int, (conn->mtu - hlen), len); + skb = bt_skb_send_alloc(sk, count + hlen, + msg->msg_flags & MSG_DONTWAIT, &err); + if (!skb) + return ERR_PTR(-ENOMEM); + + /* Create L2CAP header */ + lh = (struct l2cap_hdr *) skb_put(skb, L2CAP_HDR_SIZE); + lh->cid = cpu_to_le16(l2cap_pi(sk)->dcid); + lh->len = cpu_to_le16(len + (hlen - L2CAP_HDR_SIZE)); + + err = l2cap_skbuff_fromiovec(sk, msg, len, count, skb); + if (unlikely(err < 0)) { + kfree_skb(skb); + return ERR_PTR(err); + } + return skb; +} + +static struct sk_buff *l2cap_create_ertm_pdu(struct sock *sk, struct msghdr *msg, size_t len, u16 control) +{ + struct l2cap_conn *conn = l2cap_pi(sk)->conn; + struct sk_buff *skb; + int err, count, hlen = L2CAP_HDR_SIZE + 2; + struct l2cap_hdr *lh; + + BT_DBG("sk %p len %d", sk, (int)len); + + count = min_t(unsigned int, (conn->mtu - hlen), len); + skb = bt_skb_send_alloc(sk, count + hlen, + msg->msg_flags & MSG_DONTWAIT, &err); + if (!skb) + return ERR_PTR(-ENOMEM); + + /* Create L2CAP header */ + lh = (struct l2cap_hdr *) skb_put(skb, L2CAP_HDR_SIZE); + lh->cid = cpu_to_le16(l2cap_pi(sk)->dcid); + lh->len = cpu_to_le16(len + (hlen - L2CAP_HDR_SIZE)); + put_unaligned_le16(control, skb_put(skb, 2)); + + err = l2cap_skbuff_fromiovec(sk, msg, len, count, skb); + if (unlikely(err < 0)) { + kfree_skb(skb); + return ERR_PTR(err); + } + return skb; } static int l2cap_sock_sendmsg(struct kiocb *iocb, struct socket *sock, struct msghdr *msg, size_t len) { struct sock *sk = sock->sk; - int err = 0; + struct l2cap_pinfo *pi = l2cap_pi(sk); + struct sk_buff *skb; + u16 control; + int err; BT_DBG("sock %p, sk %p", sock, sk); @@ -1237,16 +1381,67 @@ static int l2cap_sock_sendmsg(struct kiocb *iocb, struct socket *sock, struct ms return -EOPNOTSUPP; /* Check outgoing MTU */ - if (sk->sk_type != SOCK_RAW && len > l2cap_pi(sk)->omtu) + if (sk->sk_type == SOCK_SEQPACKET && pi->mode == L2CAP_MODE_BASIC + && len > pi->omtu) return -EINVAL; lock_sock(sk); - if (sk->sk_state == BT_CONNECTED) - err = l2cap_do_send(sk, msg, len); - else + if (sk->sk_state != BT_CONNECTED) { err = -ENOTCONN; + goto done; + } + + /* Connectionless channel */ + if (sk->sk_type == SOCK_DGRAM) { + skb = l2cap_create_connless_pdu(sk, msg, len); + err = l2cap_do_send(sk, skb); + goto done; + } + switch (pi->mode) { + case L2CAP_MODE_BASIC: + /* Create a basic PDU */ + skb = l2cap_create_basic_pdu(sk, msg, len); + if (IS_ERR(skb)) { + err = PTR_ERR(skb); + goto done; + } + + err = l2cap_do_send(sk, skb); + if (!err) + err = len; + break; + + case L2CAP_MODE_ERTM: + /* Entire SDU fits into one PDU */ + if (len <= pi->omtu) { + control = L2CAP_SDU_UNSEGMENTED; + skb = l2cap_create_ertm_pdu(sk, msg, len, control); + if (IS_ERR(skb)) { + err = PTR_ERR(skb); + goto done; + } + } else { + /* FIXME: Segmentation will be added later */ + err = -EINVAL; + goto done; + } + __skb_queue_tail(TX_QUEUE(sk), skb); + if (sk->sk_send_head == NULL) + sk->sk_send_head = skb; + + err = l2cap_ertm_send(sk); + if (!err) + err = len; + break; + + default: + BT_DBG("bad state %1.1x", pi->mode); + err = -EINVAL; + } + +done: release_sock(sk); return err; } @@ -2301,6 +2496,10 @@ static inline int l2cap_config_req(struct l2cap_conn *conn, struct l2cap_cmd_hdr if (l2cap_pi(sk)->conf_state & L2CAP_CONF_INPUT_DONE) { sk->sk_state = BT_CONNECTED; + l2cap_pi(sk)->next_tx_seq = 0; + l2cap_pi(sk)->expected_ack_seq = 0; + l2cap_pi(sk)->unacked_frames = 0; + __skb_queue_head_init(TX_QUEUE(sk)); l2cap_chan_ready(sk); goto unlock; } @@ -2375,6 +2574,9 @@ static inline int l2cap_config_rsp(struct l2cap_conn *conn, struct l2cap_cmd_hdr if (l2cap_pi(sk)->conf_state & L2CAP_CONF_OUTPUT_DONE) { sk->sk_state = BT_CONNECTED; + l2cap_pi(sk)->expected_tx_seq = 0; + l2cap_pi(sk)->num_to_ack = 0; + __skb_queue_head_init(TX_QUEUE(sk)); l2cap_chan_ready(sk); } @@ -2405,6 +2607,8 @@ static inline int l2cap_disconnect_req(struct l2cap_conn *conn, struct l2cap_cmd sk->sk_shutdown = SHUTDOWN_MASK; + skb_queue_purge(TX_QUEUE(sk)); + l2cap_chan_del(sk, ECONNRESET); bh_unlock_sock(sk); @@ -2427,6 +2631,8 @@ static inline int l2cap_disconnect_rsp(struct l2cap_conn *conn, struct l2cap_cmd if (!sk) return 0; + skb_queue_purge(TX_QUEUE(sk)); + l2cap_chan_del(sk, 0); bh_unlock_sock(sk); @@ -2602,9 +2808,60 @@ static inline void l2cap_sig_channel(struct l2cap_conn *conn, struct sk_buff *sk kfree_skb(skb); } +static inline int l2cap_data_channel_iframe(struct sock *sk, u16 rx_control, struct sk_buff *skb) +{ + struct l2cap_pinfo *pi = l2cap_pi(sk); + u8 tx_seq = __get_txseq(rx_control); + u16 tx_control = 0; + int err = 0; + + BT_DBG("sk %p rx_control 0x%4.4x len %d", sk, rx_control, skb->len); + + if (tx_seq != pi->expected_tx_seq) + return -EINVAL; + + pi->expected_tx_seq = (pi->expected_tx_seq + 1) % 64; + err = sock_queue_rcv_skb(sk, skb); + if (err) + return err; + + pi->num_to_ack = (pi->num_to_ack + 1) % L2CAP_DEFAULT_NUM_TO_ACK; + if (pi->num_to_ack == L2CAP_DEFAULT_NUM_TO_ACK - 1) { + tx_control |= L2CAP_CTRL_FRAME_TYPE; + tx_control |= L2CAP_SUPER_RCV_READY; + tx_control |= pi->expected_tx_seq << L2CAP_CTRL_REQSEQ_SHIFT; + err = l2cap_send_sframe(pi, tx_control); + } + return err; +} + +static inline int l2cap_data_channel_sframe(struct sock *sk, u16 rx_control, struct sk_buff *skb) +{ + struct l2cap_pinfo *pi = l2cap_pi(sk); + + BT_DBG("sk %p rx_control 0x%4.4x len %d", sk, rx_control, skb->len); + + switch (rx_control & L2CAP_CTRL_SUPERVISE) { + case L2CAP_SUPER_RCV_READY: + pi->expected_ack_seq = __get_reqseq(rx_control); + l2cap_drop_acked_frames(sk); + l2cap_ertm_send(sk); + break; + + case L2CAP_SUPER_RCV_NOT_READY: + case L2CAP_SUPER_REJECT: + case L2CAP_SUPER_SELECT_REJECT: + break; + } + + return 0; +} + static inline int l2cap_data_channel(struct l2cap_conn *conn, u16 cid, struct sk_buff *skb) { struct sock *sk; + u16 control; + int err; sk = l2cap_get_chan_by_scid(&conn->chan_list, cid); if (!sk) { @@ -2617,16 +2874,40 @@ static inline int l2cap_data_channel(struct l2cap_conn *conn, u16 cid, struct sk if (sk->sk_state != BT_CONNECTED) goto drop; - if (l2cap_pi(sk)->imtu < skb->len) - goto drop; + switch (l2cap_pi(sk)->mode) { + case L2CAP_MODE_BASIC: + /* If socket recv buffers overflows we drop data here + * which is *bad* because L2CAP has to be reliable. + * But we don't have any other choice. L2CAP doesn't + * provide flow control mechanism. */ - /* If socket recv buffers overflows we drop data here - * which is *bad* because L2CAP has to be reliable. - * But we don't have any other choice. L2CAP doesn't - * provide flow control mechanism. */ + if (l2cap_pi(sk)->imtu < skb->len) + goto drop; - if (!sock_queue_rcv_skb(sk, skb)) - goto done; + if (!sock_queue_rcv_skb(sk, skb)) + goto done; + break; + + case L2CAP_MODE_ERTM: + control = get_unaligned_le16(skb->data); + skb_pull(skb, 2); + + if (l2cap_pi(sk)->imtu < skb->len) + goto drop; + + if (__is_iframe(control)) + err = l2cap_data_channel_iframe(sk, control, skb); + else + err = l2cap_data_channel_sframe(sk, control, skb); + + if (!err) + goto done; + break; + + default: + BT_DBG("sk %p: bad mode 0x%2.2x", sk, l2cap_pi(sk)->mode); + break; + } drop: kfree_skb(skb); @@ -2676,6 +2957,11 @@ static void l2cap_recv_frame(struct l2cap_conn *conn, struct sk_buff *skb) cid = __le16_to_cpu(lh->cid); len = __le16_to_cpu(lh->len); + if (len != skb->len) { + kfree_skb(skb); + return; + } + BT_DBG("len %d, cid 0x%4.4x", len, cid); switch (cid) { -- cgit v1.2.3 From c74e560cd0101455f1889515e1527e4c2e266113 Mon Sep 17 00:00:00 2001 From: "Gustavo F. Padovan" Date: Thu, 20 Aug 2009 22:25:58 -0300 Subject: Bluetooth: Add support for Segmentation and Reassembly of SDUs ERTM should use Segmentation and Reassembly to break down a SDU in many PDUs on sending data to the other side. On sending packets we queue all 'segments' until end of segmentation and just the add them to the queue for sending. On receiving we create a new SKB with the SDU reassembled. Initially based on a patch from Nathan Holstein Signed-off-by: Gustavo F. Padovan Signed-off-by: Marcel Holtmann --- include/net/bluetooth/l2cap.h | 9 ++- net/bluetooth/l2cap.c | 170 ++++++++++++++++++++++++++++++++++++++---- 2 files changed, 162 insertions(+), 17 deletions(-) (limited to 'include') diff --git a/include/net/bluetooth/l2cap.h b/include/net/bluetooth/l2cap.h index 9bbfbe74d400..0afde8d22b56 100644 --- a/include/net/bluetooth/l2cap.h +++ b/include/net/bluetooth/l2cap.h @@ -34,7 +34,7 @@ #define L2CAP_DEFAULT_MAX_RECEIVE 1 #define L2CAP_DEFAULT_RETRANS_TO 300 /* 300 milliseconds */ #define L2CAP_DEFAULT_MONITOR_TO 1000 /* 1 second */ -#define L2CAP_DEFAULT_MAX_RX_APDU 0xfff7 +#define L2CAP_DEFAULT_MAX_PDU_SIZE 672 #define L2CAP_CONN_TIMEOUT (40000) /* 40 seconds */ #define L2CAP_INFO_TIMEOUT (4000) /* 4 seconds */ @@ -311,6 +311,7 @@ struct l2cap_pinfo { __u8 conf_req[64]; __u8 conf_len; __u8 conf_state; + __u8 conn_state; __u8 next_tx_seq; __u8 expected_ack_seq; @@ -318,6 +319,9 @@ struct l2cap_pinfo { __u8 expected_tx_seq; __u8 unacked_frames; __u8 num_to_ack; + __u16 sdu_len; + __u16 partial_sdu_len; + struct sk_buff *sdu; __u8 ident; @@ -346,6 +350,8 @@ struct l2cap_pinfo { #define L2CAP_CONF_MAX_CONF_REQ 2 #define L2CAP_CONF_MAX_CONF_RSP 2 +#define L2CAP_CONN_SAR_SDU 0x01 + static inline int l2cap_tx_window_full(struct sock *sk) { struct l2cap_pinfo *pi = l2cap_pi(sk); @@ -363,6 +369,7 @@ static inline int l2cap_tx_window_full(struct sock *sk) #define __get_reqseq(ctrl) ((ctrl) & L2CAP_CTRL_REQSEQ) >> 8 #define __is_iframe(ctrl) !((ctrl) & L2CAP_CTRL_FRAME_TYPE) #define __is_sframe(ctrl) (ctrl) & L2CAP_CTRL_FRAME_TYPE +#define __is_sar_start(ctrl) ((ctrl) & L2CAP_CTRL_SAR) == L2CAP_SDU_START void l2cap_load(void); diff --git a/net/bluetooth/l2cap.c b/net/bluetooth/l2cap.c index 45b8697a7398..167e02532697 100644 --- a/net/bluetooth/l2cap.c +++ b/net/bluetooth/l2cap.c @@ -1334,7 +1334,7 @@ static struct sk_buff *l2cap_create_basic_pdu(struct sock *sk, struct msghdr *ms return skb; } -static struct sk_buff *l2cap_create_ertm_pdu(struct sock *sk, struct msghdr *msg, size_t len, u16 control) +static struct sk_buff *l2cap_create_ertm_pdu(struct sock *sk, struct msghdr *msg, size_t len, u16 control, u16 sdulen) { struct l2cap_conn *conn = l2cap_pi(sk)->conn; struct sk_buff *skb; @@ -1343,6 +1343,9 @@ static struct sk_buff *l2cap_create_ertm_pdu(struct sock *sk, struct msghdr *msg BT_DBG("sk %p len %d", sk, (int)len); + if (sdulen) + hlen += 2; + count = min_t(unsigned int, (conn->mtu - hlen), len); skb = bt_skb_send_alloc(sk, count + hlen, msg->msg_flags & MSG_DONTWAIT, &err); @@ -1354,6 +1357,8 @@ static struct sk_buff *l2cap_create_ertm_pdu(struct sock *sk, struct msghdr *msg lh->cid = cpu_to_le16(l2cap_pi(sk)->dcid); lh->len = cpu_to_le16(len + (hlen - L2CAP_HDR_SIZE)); put_unaligned_le16(control, skb_put(skb, 2)); + if (sdulen) + put_unaligned_le16(sdulen, skb_put(skb, 2)); err = l2cap_skbuff_fromiovec(sk, msg, len, count, skb); if (unlikely(err < 0)) { @@ -1363,6 +1368,54 @@ static struct sk_buff *l2cap_create_ertm_pdu(struct sock *sk, struct msghdr *msg return skb; } +static inline int l2cap_sar_segment_sdu(struct sock *sk, struct msghdr *msg, size_t len) +{ + struct l2cap_pinfo *pi = l2cap_pi(sk); + struct sk_buff *skb; + struct sk_buff_head sar_queue; + u16 control; + size_t size = 0; + + __skb_queue_head_init(&sar_queue); + control = L2CAP_SDU_START; + skb = l2cap_create_ertm_pdu(sk, msg, pi->max_pdu_size, control, len); + if (IS_ERR(skb)) + return PTR_ERR(skb); + + __skb_queue_tail(&sar_queue, skb); + len -= pi->max_pdu_size; + size +=pi->max_pdu_size; + control = 0; + + while (len > 0) { + size_t buflen; + + if (len > pi->max_pdu_size) { + control |= L2CAP_SDU_CONTINUE; + buflen = pi->max_pdu_size; + } else { + control |= L2CAP_SDU_END; + buflen = len; + } + + skb = l2cap_create_ertm_pdu(sk, msg, buflen, control, 0); + if (IS_ERR(skb)) { + skb_queue_purge(&sar_queue); + return PTR_ERR(skb); + } + + __skb_queue_tail(&sar_queue, skb); + len -= buflen; + size += buflen; + control = 0; + } + skb_queue_splice_tail(&sar_queue, TX_QUEUE(sk)); + if (sk->sk_send_head == NULL) + sk->sk_send_head = sar_queue.next; + + return size; +} + static int l2cap_sock_sendmsg(struct kiocb *iocb, struct socket *sock, struct msghdr *msg, size_t len) { struct sock *sk = sock->sk; @@ -1415,21 +1468,22 @@ static int l2cap_sock_sendmsg(struct kiocb *iocb, struct socket *sock, struct ms case L2CAP_MODE_ERTM: /* Entire SDU fits into one PDU */ - if (len <= pi->omtu) { + if (len <= pi->max_pdu_size) { control = L2CAP_SDU_UNSEGMENTED; - skb = l2cap_create_ertm_pdu(sk, msg, len, control); + skb = l2cap_create_ertm_pdu(sk, msg, len, control, 0); if (IS_ERR(skb)) { err = PTR_ERR(skb); goto done; } + __skb_queue_tail(TX_QUEUE(sk), skb); + if (sk->sk_send_head == NULL) + sk->sk_send_head = skb; } else { - /* FIXME: Segmentation will be added later */ - err = -EINVAL; - goto done; + /* Segment SDU into multiples PDUs */ + err = l2cap_sar_segment_sdu(sk, msg, len); + if (err < 0) + goto done; } - __skb_queue_tail(TX_QUEUE(sk), skb); - if (sk->sk_send_head == NULL) - sk->sk_send_head = skb; err = l2cap_ertm_send(sk); if (!err) @@ -2007,7 +2061,7 @@ done: rfc.max_transmit = L2CAP_DEFAULT_MAX_RECEIVE; rfc.retrans_timeout = 0; rfc.monitor_timeout = 0; - rfc.max_pdu_size = cpu_to_le16(L2CAP_DEFAULT_MAX_RX_APDU); + rfc.max_pdu_size = cpu_to_le16(L2CAP_DEFAULT_MAX_PDU_SIZE); l2cap_add_conf_opt(&ptr, L2CAP_CONF_RFC, sizeof(rfc), (unsigned long) &rfc); @@ -2019,7 +2073,7 @@ done: rfc.max_transmit = 0; rfc.retrans_timeout = 0; rfc.monitor_timeout = 0; - rfc.max_pdu_size = cpu_to_le16(L2CAP_DEFAULT_MAX_RX_APDU); + rfc.max_pdu_size = cpu_to_le16(L2CAP_DEFAULT_MAX_PDU_SIZE); l2cap_add_conf_opt(&ptr, L2CAP_CONF_RFC, sizeof(rfc), (unsigned long) &rfc); @@ -2808,6 +2862,86 @@ static inline void l2cap_sig_channel(struct l2cap_conn *conn, struct sk_buff *sk kfree_skb(skb); } +static int l2cap_sar_reassembly_sdu(struct sock *sk, struct sk_buff *skb, u16 control) +{ + struct l2cap_pinfo *pi = l2cap_pi(sk); + struct sk_buff *_skb; + int err = -EINVAL; + + switch (control & L2CAP_CTRL_SAR) { + case L2CAP_SDU_UNSEGMENTED: + if (pi->conn_state & L2CAP_CONN_SAR_SDU) { + kfree_skb(pi->sdu); + break; + } + + err = sock_queue_rcv_skb(sk, skb); + if (!err) + return 0; + + break; + + case L2CAP_SDU_START: + if (pi->conn_state & L2CAP_CONN_SAR_SDU) { + kfree_skb(pi->sdu); + break; + } + + pi->sdu_len = get_unaligned_le16(skb->data); + skb_pull(skb, 2); + + pi->sdu = bt_skb_alloc(pi->sdu_len, GFP_ATOMIC); + if (!pi->sdu) { + err = -ENOMEM; + break; + } + + memcpy(skb_put(pi->sdu, skb->len), skb->data, skb->len); + + pi->conn_state |= L2CAP_CONN_SAR_SDU; + pi->partial_sdu_len = skb->len; + err = 0; + break; + + case L2CAP_SDU_CONTINUE: + if (!(pi->conn_state & L2CAP_CONN_SAR_SDU)) + break; + + memcpy(skb_put(pi->sdu, skb->len), skb->data, skb->len); + + pi->partial_sdu_len += skb->len; + if (pi->partial_sdu_len > pi->sdu_len) + kfree_skb(pi->sdu); + else + err = 0; + + break; + + case L2CAP_SDU_END: + if (!(pi->conn_state & L2CAP_CONN_SAR_SDU)) + break; + + memcpy(skb_put(pi->sdu, skb->len), skb->data, skb->len); + + pi->conn_state &= ~L2CAP_CONN_SAR_SDU; + pi->partial_sdu_len += skb->len; + + if (pi->partial_sdu_len == pi->sdu_len) { + _skb = skb_clone(pi->sdu, GFP_ATOMIC); + err = sock_queue_rcv_skb(sk, _skb); + if (err < 0) + kfree_skb(_skb); + } + kfree_skb(pi->sdu); + err = 0; + + break; + } + + kfree_skb(skb); + return err; +} + static inline int l2cap_data_channel_iframe(struct sock *sk, u16 rx_control, struct sk_buff *skb) { struct l2cap_pinfo *pi = l2cap_pi(sk); @@ -2820,11 +2954,11 @@ static inline int l2cap_data_channel_iframe(struct sock *sk, u16 rx_control, str if (tx_seq != pi->expected_tx_seq) return -EINVAL; - pi->expected_tx_seq = (pi->expected_tx_seq + 1) % 64; - err = sock_queue_rcv_skb(sk, skb); - if (err) + err = l2cap_sar_reassembly_sdu(sk, skb, rx_control); + if (err < 0) return err; + pi->expected_tx_seq = (pi->expected_tx_seq + 1) % 64; pi->num_to_ack = (pi->num_to_ack + 1) % L2CAP_DEFAULT_NUM_TO_ACK; if (pi->num_to_ack == L2CAP_DEFAULT_NUM_TO_ACK - 1) { tx_control |= L2CAP_CTRL_FRAME_TYPE; @@ -2860,7 +2994,7 @@ static inline int l2cap_data_channel_sframe(struct sock *sk, u16 rx_control, str static inline int l2cap_data_channel(struct l2cap_conn *conn, u16 cid, struct sk_buff *skb) { struct sock *sk; - u16 control; + u16 control, len; int err; sk = l2cap_get_chan_by_scid(&conn->chan_list, cid); @@ -2891,8 +3025,12 @@ static inline int l2cap_data_channel(struct l2cap_conn *conn, u16 cid, struct sk case L2CAP_MODE_ERTM: control = get_unaligned_le16(skb->data); skb_pull(skb, 2); + len = skb->len; - if (l2cap_pi(sk)->imtu < skb->len) + if (__is_sar_start(control)) + len -= 2; + + if (len > L2CAP_DEFAULT_MAX_PDU_SIZE) goto drop; if (__is_iframe(control)) -- cgit v1.2.3 From 30afb5b2aa83adf4f69e5090d48e1bb04b64c58a Mon Sep 17 00:00:00 2001 From: "Gustavo F. Padovan" Date: Thu, 20 Aug 2009 22:25:59 -0300 Subject: Bluetooth: Initial support for retransmission of packets with REJ frames When receiving an I-frame with unexpected txSeq, receiver side start the recovery procedure by sending a REJ S-frame to the transmitter side. So the transmitter can re-send the lost I-frame. This patch just adds a basic support for retransmission, it doesn't mean that ERTM now has full support for packet retransmission. Signed-off-by: Gustavo F. Padovan Signed-off-by: Marcel Holtmann --- include/net/bluetooth/l2cap.h | 1 + net/bluetooth/l2cap.c | 57 ++++++++++++++++++++++++++++++++----------- 2 files changed, 44 insertions(+), 14 deletions(-) (limited to 'include') diff --git a/include/net/bluetooth/l2cap.h b/include/net/bluetooth/l2cap.h index 0afde8d22b56..a1d8ec468ef3 100644 --- a/include/net/bluetooth/l2cap.h +++ b/include/net/bluetooth/l2cap.h @@ -351,6 +351,7 @@ struct l2cap_pinfo { #define L2CAP_CONF_MAX_CONF_RSP 2 #define L2CAP_CONN_SAR_SDU 0x01 +#define L2CAP_CONN_UNDER_REJ 0x02 static inline int l2cap_tx_window_full(struct sock *sk) { diff --git a/net/bluetooth/l2cap.c b/net/bluetooth/l2cap.c index 167e02532697..35e9f5b80545 100644 --- a/net/bluetooth/l2cap.c +++ b/net/bluetooth/l2cap.c @@ -2951,22 +2951,36 @@ static inline int l2cap_data_channel_iframe(struct sock *sk, u16 rx_control, str BT_DBG("sk %p rx_control 0x%4.4x len %d", sk, rx_control, skb->len); - if (tx_seq != pi->expected_tx_seq) - return -EINVAL; + if (tx_seq == pi->expected_tx_seq) { + if (pi->conn_state & L2CAP_CONN_UNDER_REJ) + pi->conn_state &= ~L2CAP_CONN_UNDER_REJ; - err = l2cap_sar_reassembly_sdu(sk, skb, rx_control); - if (err < 0) - return err; + err = l2cap_sar_reassembly_sdu(sk, skb, rx_control); + if (err < 0) + return err; + + pi->expected_tx_seq = (pi->expected_tx_seq + 1) % 64; + pi->num_to_ack = (pi->num_to_ack + 1) % L2CAP_DEFAULT_NUM_TO_ACK; + if (pi->num_to_ack == L2CAP_DEFAULT_NUM_TO_ACK - 1) { + tx_control |= L2CAP_SUPER_RCV_READY; + tx_control |= pi->expected_tx_seq << L2CAP_CTRL_REQSEQ_SHIFT; + goto send; + } + } else { + /* Unexpected txSeq. Send a REJ S-frame */ + kfree_skb(skb); + if (!(pi->conn_state & L2CAP_CONN_UNDER_REJ)) { + tx_control |= L2CAP_SUPER_REJECT; + tx_control |= pi->expected_tx_seq << L2CAP_CTRL_REQSEQ_SHIFT; + pi->conn_state |= L2CAP_CONN_UNDER_REJ; - pi->expected_tx_seq = (pi->expected_tx_seq + 1) % 64; - pi->num_to_ack = (pi->num_to_ack + 1) % L2CAP_DEFAULT_NUM_TO_ACK; - if (pi->num_to_ack == L2CAP_DEFAULT_NUM_TO_ACK - 1) { - tx_control |= L2CAP_CTRL_FRAME_TYPE; - tx_control |= L2CAP_SUPER_RCV_READY; - tx_control |= pi->expected_tx_seq << L2CAP_CTRL_REQSEQ_SHIFT; - err = l2cap_send_sframe(pi, tx_control); + goto send; + } } - return err; + return 0; + +send: + return l2cap_send_sframe(pi, tx_control); } static inline int l2cap_data_channel_sframe(struct sock *sk, u16 rx_control, struct sk_buff *skb) @@ -2982,8 +2996,18 @@ static inline int l2cap_data_channel_sframe(struct sock *sk, u16 rx_control, str l2cap_ertm_send(sk); break; - case L2CAP_SUPER_RCV_NOT_READY: case L2CAP_SUPER_REJECT: + pi->expected_ack_seq = __get_reqseq(rx_control); + l2cap_drop_acked_frames(sk); + + sk->sk_send_head = TX_QUEUE(sk)->next; + pi->next_tx_seq = pi->expected_ack_seq; + + l2cap_ertm_send(sk); + + break; + + case L2CAP_SUPER_RCV_NOT_READY: case L2CAP_SUPER_SELECT_REJECT: break; } @@ -3030,6 +3054,11 @@ static inline int l2cap_data_channel(struct l2cap_conn *conn, u16 cid, struct sk if (__is_sar_start(control)) len -= 2; + /* + * We can just drop the corrupted I-frame here. + * Receiver will miss it and start proper recovery + * procedures and ask retransmission. + */ if (len > L2CAP_DEFAULT_MAX_PDU_SIZE) goto drop; -- cgit v1.2.3 From e90bac061b17cd81bd0df30606c64f4543bf5ca0 Mon Sep 17 00:00:00 2001 From: "Gustavo F. Padovan" Date: Thu, 20 Aug 2009 22:26:00 -0300 Subject: Bluetooth: Add support for Retransmission and Monitor Timers L2CAP uses retransmission and monitor timers to inquiry the other side about unacked I-frames. After sending each I-frame we (re)start the retransmission timer. If it expires, we start a monitor timer that send a S-frame with P bit set and wait for S-frame with F bit set. If monitor timer expires, try again, at a maximum of L2CAP_DEFAULT_MAX_TX. Signed-off-by: Gustavo F. Padovan Signed-off-by: Marcel Holtmann --- include/net/bluetooth/bluetooth.h | 1 + include/net/bluetooth/l2cap.h | 15 +++++-- net/bluetooth/l2cap.c | 86 +++++++++++++++++++++++++++++++++++++-- 3 files changed, 95 insertions(+), 7 deletions(-) (limited to 'include') diff --git a/include/net/bluetooth/bluetooth.h b/include/net/bluetooth/bluetooth.h index 65a5cf868fd6..b8b9a8479525 100644 --- a/include/net/bluetooth/bluetooth.h +++ b/include/net/bluetooth/bluetooth.h @@ -139,6 +139,7 @@ struct bt_skb_cb { __u8 pkt_type; __u8 incoming; __u8 tx_seq; + __u8 retries; }; #define bt_cb(skb) ((struct bt_skb_cb *)((skb)->cb)) diff --git a/include/net/bluetooth/l2cap.h b/include/net/bluetooth/l2cap.h index a1d8ec468ef3..2cf7003cb1b6 100644 --- a/include/net/bluetooth/l2cap.h +++ b/include/net/bluetooth/l2cap.h @@ -31,9 +31,9 @@ #define L2CAP_DEFAULT_FLUSH_TO 0xffff #define L2CAP_DEFAULT_TX_WINDOW 63 #define L2CAP_DEFAULT_NUM_TO_ACK (L2CAP_DEFAULT_TX_WINDOW/5) -#define L2CAP_DEFAULT_MAX_RECEIVE 1 -#define L2CAP_DEFAULT_RETRANS_TO 300 /* 300 milliseconds */ -#define L2CAP_DEFAULT_MONITOR_TO 1000 /* 1 second */ +#define L2CAP_DEFAULT_MAX_TX 3 +#define L2CAP_DEFAULT_RETRANS_TO 1000 /* 1 second */ +#define L2CAP_DEFAULT_MONITOR_TO 12000 /* 12 seconds */ #define L2CAP_DEFAULT_MAX_PDU_SIZE 672 #define L2CAP_CONN_TIMEOUT (40000) /* 40 seconds */ @@ -318,6 +318,7 @@ struct l2cap_pinfo { __u8 req_seq; __u8 expected_tx_seq; __u8 unacked_frames; + __u8 retry_count; __u8 num_to_ack; __u16 sdu_len; __u16 partial_sdu_len; @@ -333,6 +334,8 @@ struct l2cap_pinfo { __le16 sport; + struct timer_list retrans_timer; + struct timer_list monitor_timer; struct sk_buff_head tx_queue; struct l2cap_conn *conn; struct sock *next_c; @@ -352,6 +355,12 @@ struct l2cap_pinfo { #define L2CAP_CONN_SAR_SDU 0x01 #define L2CAP_CONN_UNDER_REJ 0x02 +#define L2CAP_CONN_WAIT_F 0x04 + +#define __mod_retrans_timer() mod_timer(&l2cap_pi(sk)->retrans_timer, \ + jiffies + msecs_to_jiffies(L2CAP_DEFAULT_RETRANS_TO)); +#define __mod_monitor_timer() mod_timer(&l2cap_pi(sk)->monitor_timer, \ + jiffies + msecs_to_jiffies(L2CAP_DEFAULT_MONITOR_TO)); static inline int l2cap_tx_window_full(struct sock *sk) { diff --git a/net/bluetooth/l2cap.c b/net/bluetooth/l2cap.c index 35e9f5b80545..97172f7c0a6a 100644 --- a/net/bluetooth/l2cap.c +++ b/net/bluetooth/l2cap.c @@ -1178,6 +1178,39 @@ static int l2cap_sock_getname(struct socket *sock, struct sockaddr *addr, int *l return 0; } +static void l2cap_monitor_timeout(unsigned long arg) +{ + struct sock *sk = (void *) arg; + u16 control; + + if (l2cap_pi(sk)->retry_count >= l2cap_pi(sk)->remote_max_tx) { + l2cap_send_disconn_req(l2cap_pi(sk)->conn, sk); + return; + } + + l2cap_pi(sk)->retry_count++; + __mod_monitor_timer(); + + control = L2CAP_CTRL_POLL; + control |= L2CAP_SUPER_RCV_READY; + l2cap_send_sframe(l2cap_pi(sk), control); +} + +static void l2cap_retrans_timeout(unsigned long arg) +{ + struct sock *sk = (void *) arg; + u16 control; + + l2cap_pi(sk)->retry_count = 1; + __mod_monitor_timer(); + + l2cap_pi(sk)->conn_state |= L2CAP_CONN_WAIT_F; + + control = L2CAP_CTRL_POLL; + control |= L2CAP_SUPER_RCV_READY; + l2cap_send_sframe(l2cap_pi(sk), control); +} + static void l2cap_drop_acked_frames(struct sock *sk) { struct sk_buff *skb; @@ -1192,6 +1225,9 @@ static void l2cap_drop_acked_frames(struct sock *sk) l2cap_pi(sk)->unacked_frames--; } + if (!l2cap_pi(sk)->unacked_frames) + del_timer(&l2cap_pi(sk)->retrans_timer); + return; } @@ -1216,19 +1252,32 @@ static int l2cap_ertm_send(struct sock *sk) u16 control; int err; + if (pi->conn_state & L2CAP_CONN_WAIT_F) + return 0; + while ((skb = sk->sk_send_head) && (!l2cap_tx_window_full(sk))) { tx_skb = skb_clone(skb, GFP_ATOMIC); + if (pi->remote_max_tx && + bt_cb(skb)->retries == pi->remote_max_tx) { + l2cap_send_disconn_req(pi->conn, sk); + break; + } + + bt_cb(skb)->retries++; + control = get_unaligned_le16(tx_skb->data + L2CAP_HDR_SIZE); control |= (pi->req_seq << L2CAP_CTRL_REQSEQ_SHIFT) | (pi->next_tx_seq << L2CAP_CTRL_TXSEQ_SHIFT); put_unaligned_le16(control, tx_skb->data + L2CAP_HDR_SIZE); + err = l2cap_do_send(sk, tx_skb); if (err < 0) { l2cap_send_disconn_req(pi->conn, sk); return err; } + __mod_retrans_timer(); bt_cb(skb)->tx_seq = pi->next_tx_seq; pi->next_tx_seq = (pi->next_tx_seq + 1) % 64; @@ -1365,6 +1414,8 @@ static struct sk_buff *l2cap_create_ertm_pdu(struct sock *sk, struct msghdr *msg kfree_skb(skb); return ERR_PTR(err); } + + bt_cb(skb)->retries = 0; return skb; } @@ -2058,7 +2109,7 @@ done: case L2CAP_MODE_ERTM: rfc.mode = L2CAP_MODE_ERTM; rfc.txwin_size = L2CAP_DEFAULT_TX_WINDOW; - rfc.max_transmit = L2CAP_DEFAULT_MAX_RECEIVE; + rfc.max_transmit = L2CAP_DEFAULT_MAX_TX; rfc.retrans_timeout = 0; rfc.monitor_timeout = 0; rfc.max_pdu_size = cpu_to_le16(L2CAP_DEFAULT_MAX_PDU_SIZE); @@ -2553,6 +2604,12 @@ static inline int l2cap_config_req(struct l2cap_conn *conn, struct l2cap_cmd_hdr l2cap_pi(sk)->next_tx_seq = 0; l2cap_pi(sk)->expected_ack_seq = 0; l2cap_pi(sk)->unacked_frames = 0; + + setup_timer(&l2cap_pi(sk)->retrans_timer, + l2cap_retrans_timeout, (unsigned long) sk); + setup_timer(&l2cap_pi(sk)->monitor_timer, + l2cap_monitor_timeout, (unsigned long) sk); + __skb_queue_head_init(TX_QUEUE(sk)); l2cap_chan_ready(sk); goto unlock; @@ -2662,6 +2719,8 @@ static inline int l2cap_disconnect_req(struct l2cap_conn *conn, struct l2cap_cmd sk->sk_shutdown = SHUTDOWN_MASK; skb_queue_purge(TX_QUEUE(sk)); + del_timer(&l2cap_pi(sk)->retrans_timer); + del_timer(&l2cap_pi(sk)->monitor_timer); l2cap_chan_del(sk, ECONNRESET); bh_unlock_sock(sk); @@ -2686,6 +2745,8 @@ static inline int l2cap_disconnect_rsp(struct l2cap_conn *conn, struct l2cap_cmd return 0; skb_queue_purge(TX_QUEUE(sk)); + del_timer(&l2cap_pi(sk)->retrans_timer); + del_timer(&l2cap_pi(sk)->monitor_timer); l2cap_chan_del(sk, 0); bh_unlock_sock(sk); @@ -2991,9 +3052,26 @@ static inline int l2cap_data_channel_sframe(struct sock *sk, u16 rx_control, str switch (rx_control & L2CAP_CTRL_SUPERVISE) { case L2CAP_SUPER_RCV_READY: - pi->expected_ack_seq = __get_reqseq(rx_control); - l2cap_drop_acked_frames(sk); - l2cap_ertm_send(sk); + if (rx_control & L2CAP_CTRL_POLL) { + u16 control = L2CAP_CTRL_FINAL; + control |= L2CAP_SUPER_RCV_READY; + l2cap_send_sframe(l2cap_pi(sk), control); + } else if (rx_control & L2CAP_CTRL_FINAL) { + if (!(pi->conn_state & L2CAP_CONN_WAIT_F)) + break; + + pi->conn_state &= ~L2CAP_CONN_WAIT_F; + del_timer(&pi->monitor_timer); + + if (pi->unacked_frames > 0) + __mod_retrans_timer(); + } else { + pi->expected_ack_seq = __get_reqseq(rx_control); + l2cap_drop_acked_frames(sk); + if (pi->unacked_frames > 0) + __mod_retrans_timer(); + l2cap_ertm_send(sk); + } break; case L2CAP_SUPER_REJECT: -- cgit v1.2.3 From fcc203c30d72dde82692f6b761a80e5ca5fdd8fa Mon Sep 17 00:00:00 2001 From: "Gustavo F. Padovan" Date: Thu, 20 Aug 2009 22:26:02 -0300 Subject: Bluetooth: Add support for FCS option to L2CAP Implement CRC16 check for L2CAP packets. FCS is used by Streaming Mode and Enhanced Retransmission Mode and is a extra check for the packet content. Using CRC16 is the default, L2CAP won't use FCS only when both side send a "No FCS" request. Initially based on a patch from Nathan Holstein Signed-off-by: Gustavo F. Padovan Signed-off-by: Marcel Holtmann --- include/net/bluetooth/l2cap.h | 2 + net/bluetooth/l2cap.c | 101 +++++++++++++++++++++++++++++++++++++++--- 2 files changed, 97 insertions(+), 6 deletions(-) (limited to 'include') diff --git a/include/net/bluetooth/l2cap.h b/include/net/bluetooth/l2cap.h index 2cf7003cb1b6..59b26bf10f30 100644 --- a/include/net/bluetooth/l2cap.h +++ b/include/net/bluetooth/l2cap.h @@ -54,6 +54,7 @@ struct l2cap_options { __u16 imtu; __u16 flush_to; __u8 mode; + __u8 fcs; }; #define L2CAP_CONNINFO 0x02 @@ -348,6 +349,7 @@ struct l2cap_pinfo { #define L2CAP_CONF_MTU_DONE 0x08 #define L2CAP_CONF_MODE_DONE 0x10 #define L2CAP_CONF_CONNECT_PEND 0x20 +#define L2CAP_CONF_NO_FCS_RECV 0x40 #define L2CAP_CONF_STATE2_DEVICE 0x80 #define L2CAP_CONF_MAX_CONF_REQ 2 diff --git a/net/bluetooth/l2cap.c b/net/bluetooth/l2cap.c index 7f835e761822..4c319003c290 100644 --- a/net/bluetooth/l2cap.c +++ b/net/bluetooth/l2cap.c @@ -41,6 +41,7 @@ #include #include #include +#include #include #include @@ -338,11 +339,14 @@ static inline int l2cap_send_sframe(struct l2cap_pinfo *pi, u16 control) struct sk_buff *skb; struct l2cap_hdr *lh; struct l2cap_conn *conn = pi->conn; - int count; + int count, hlen = L2CAP_HDR_SIZE + 2; + + if (pi->fcs == L2CAP_FCS_CRC16) + hlen += 2; BT_DBG("pi %p, control 0x%2.2x", pi, control); - count = min_t(unsigned int, conn->mtu, L2CAP_HDR_SIZE + 2); + count = min_t(unsigned int, conn->mtu, hlen); control |= L2CAP_CTRL_FRAME_TYPE; skb = bt_skb_alloc(count, GFP_ATOMIC); @@ -350,10 +354,15 @@ static inline int l2cap_send_sframe(struct l2cap_pinfo *pi, u16 control) return -ENOMEM; lh = (struct l2cap_hdr *) skb_put(skb, L2CAP_HDR_SIZE); - lh->len = cpu_to_le16(2); + lh->len = cpu_to_le16(hlen - L2CAP_HDR_SIZE); lh->cid = cpu_to_le16(pi->dcid); put_unaligned_le16(control, skb_put(skb, 2)); + if (pi->fcs == L2CAP_FCS_CRC16) { + u16 fcs = crc16(0, (u8 *)lh, count - 2); + put_unaligned_le16(fcs, skb_put(skb, 2)); + } + return hci_send_acl(pi->conn->hcon, skb, 0); } @@ -1249,7 +1258,7 @@ static int l2cap_streaming_send(struct sock *sk) { struct sk_buff *skb, *tx_skb; struct l2cap_pinfo *pi = l2cap_pi(sk); - u16 control; + u16 control, fcs; int err; while ((skb = sk->sk_send_head)) { @@ -1259,6 +1268,11 @@ static int l2cap_streaming_send(struct sock *sk) control |= pi->next_tx_seq << L2CAP_CTRL_TXSEQ_SHIFT; put_unaligned_le16(control, tx_skb->data + L2CAP_HDR_SIZE); + if (l2cap_pi(sk)->fcs == L2CAP_FCS_CRC16) { + fcs = crc16(0, (u8 *)tx_skb->data, tx_skb->len - 2); + put_unaligned_le16(fcs, tx_skb->data + tx_skb->len - 2); + } + err = l2cap_do_send(sk, tx_skb); if (err < 0) { l2cap_send_disconn_req(pi->conn, sk); @@ -1282,7 +1296,7 @@ static int l2cap_ertm_send(struct sock *sk) { struct sk_buff *skb, *tx_skb; struct l2cap_pinfo *pi = l2cap_pi(sk); - u16 control; + u16 control, fcs; int err; if (pi->conn_state & L2CAP_CONN_WAIT_F) @@ -1305,6 +1319,11 @@ static int l2cap_ertm_send(struct sock *sk) put_unaligned_le16(control, tx_skb->data + L2CAP_HDR_SIZE); + if (l2cap_pi(sk)->fcs == L2CAP_FCS_CRC16) { + fcs = crc16(0, (u8 *)skb->data, tx_skb->len - 2); + put_unaligned_le16(fcs, skb->data + tx_skb->len - 2); + } + err = l2cap_do_send(sk, tx_skb); if (err < 0) { l2cap_send_disconn_req(pi->conn, sk); @@ -1428,6 +1447,9 @@ static struct sk_buff *l2cap_create_iframe_pdu(struct sock *sk, struct msghdr *m if (sdulen) hlen += 2; + if (l2cap_pi(sk)->fcs == L2CAP_FCS_CRC16) + hlen += 2; + count = min_t(unsigned int, (conn->mtu - hlen), len); skb = bt_skb_send_alloc(sk, count + hlen, msg->msg_flags & MSG_DONTWAIT, &err); @@ -1448,6 +1470,9 @@ static struct sk_buff *l2cap_create_iframe_pdu(struct sock *sk, struct msghdr *m return ERR_PTR(err); } + if (l2cap_pi(sk)->fcs == L2CAP_FCS_CRC16) + put_unaligned_le16(0, skb_put(skb, 2)); + bt_cb(skb)->retries = 0; return skb; } @@ -1633,6 +1658,7 @@ static int l2cap_sock_setsockopt_old(struct socket *sock, int optname, char __us opts.omtu = l2cap_pi(sk)->omtu; opts.flush_to = l2cap_pi(sk)->flush_to; opts.mode = l2cap_pi(sk)->mode; + opts.fcs = l2cap_pi(sk)->fcs; len = min_t(unsigned int, sizeof(opts), optlen); if (copy_from_user((char *) &opts, optval, len)) { @@ -1643,6 +1669,7 @@ static int l2cap_sock_setsockopt_old(struct socket *sock, int optname, char __us l2cap_pi(sk)->imtu = opts.imtu; l2cap_pi(sk)->omtu = opts.omtu; l2cap_pi(sk)->mode = opts.mode; + l2cap_pi(sk)->fcs = opts.fcs; break; case L2CAP_LM: @@ -1756,6 +1783,7 @@ static int l2cap_sock_getsockopt_old(struct socket *sock, int optname, char __us opts.omtu = l2cap_pi(sk)->omtu; opts.flush_to = l2cap_pi(sk)->flush_to; opts.mode = l2cap_pi(sk)->mode; + opts.fcs = l2cap_pi(sk)->fcs; len = min_t(unsigned int, len, sizeof(opts)); if (copy_to_user(optval, (char *) &opts, len)) @@ -2154,6 +2182,15 @@ done: l2cap_add_conf_opt(&ptr, L2CAP_CONF_RFC, sizeof(rfc), (unsigned long) &rfc); + + if (!(pi->conn->feat_mask & L2CAP_FEAT_FCS)) + break; + + if (pi->fcs == L2CAP_FCS_NONE || + pi->conf_state & L2CAP_CONF_NO_FCS_RECV) { + pi->fcs = L2CAP_FCS_NONE; + l2cap_add_conf_opt(&ptr, L2CAP_CONF_FCS, 1, pi->fcs); + } break; case L2CAP_MODE_STREAMING: @@ -2166,6 +2203,15 @@ done: l2cap_add_conf_opt(&ptr, L2CAP_CONF_RFC, sizeof(rfc), (unsigned long) &rfc); + + if (!(pi->conn->feat_mask & L2CAP_FEAT_FCS)) + break; + + if (pi->fcs == L2CAP_FCS_NONE || + pi->conf_state & L2CAP_CONF_NO_FCS_RECV) { + pi->fcs = L2CAP_FCS_NONE; + l2cap_add_conf_opt(&ptr, L2CAP_CONF_FCS, 1, pi->fcs); + } break; } @@ -2217,6 +2263,12 @@ static int l2cap_parse_conf_req(struct sock *sk, void *data) memcpy(&rfc, (void *) val, olen); break; + case L2CAP_CONF_FCS: + if (val == L2CAP_FCS_NONE) + pi->conf_state |= L2CAP_CONF_NO_FCS_RECV; + + break; + default: if (hint) break; @@ -2638,6 +2690,10 @@ static inline int l2cap_config_req(struct l2cap_conn *conn, struct l2cap_cmd_hdr goto unlock; if (l2cap_pi(sk)->conf_state & L2CAP_CONF_INPUT_DONE) { + if (!(l2cap_pi(sk)->conf_state & L2CAP_CONF_NO_FCS_RECV) + || l2cap_pi(sk)->fcs != L2CAP_FCS_NONE) + l2cap_pi(sk)->fcs = L2CAP_FCS_CRC16; + sk->sk_state = BT_CONNECTED; l2cap_pi(sk)->next_tx_seq = 0; l2cap_pi(sk)->expected_ack_seq = 0; @@ -2722,6 +2778,10 @@ static inline int l2cap_config_rsp(struct l2cap_conn *conn, struct l2cap_cmd_hdr l2cap_pi(sk)->conf_state |= L2CAP_CONF_INPUT_DONE; if (l2cap_pi(sk)->conf_state & L2CAP_CONF_OUTPUT_DONE) { + if (!(l2cap_pi(sk)->conf_state & L2CAP_CONF_NO_FCS_RECV) + || l2cap_pi(sk)->fcs != L2CAP_FCS_NONE) + l2cap_pi(sk)->fcs = L2CAP_FCS_CRC16; + sk->sk_state = BT_CONNECTED; l2cap_pi(sk)->expected_tx_seq = 0; l2cap_pi(sk)->num_to_ack = 0; @@ -2809,7 +2869,8 @@ static inline int l2cap_information_req(struct l2cap_conn *conn, struct l2cap_cm rsp->type = cpu_to_le16(L2CAP_IT_FEAT_MASK); rsp->result = cpu_to_le16(L2CAP_IR_SUCCESS); if (enable_ertm) - feat_mask |= L2CAP_FEAT_ERTM | L2CAP_FEAT_STREAMING; + feat_mask |= L2CAP_FEAT_ERTM | L2CAP_FEAT_STREAMING + | L2CAP_FEAT_FCS; put_unaligned(cpu_to_le32(feat_mask), (__le32 *) rsp->data); l2cap_send_cmd(conn, cmd->ident, L2CAP_INFO_RSP, sizeof(buf), buf); @@ -2961,6 +3022,22 @@ static inline void l2cap_sig_channel(struct l2cap_conn *conn, struct sk_buff *sk kfree_skb(skb); } +static int l2cap_check_fcs(struct l2cap_pinfo *pi, struct sk_buff *skb) +{ + u16 our_fcs, rcv_fcs; + int hdr_size = L2CAP_HDR_SIZE + 2; + + if (pi->fcs == L2CAP_FCS_CRC16) { + skb_trim(skb, skb->len - 2); + rcv_fcs = get_unaligned_le16(skb->data + skb->len); + our_fcs = crc16(0, skb->data - hdr_size, skb->len + hdr_size); + + if (our_fcs != rcv_fcs) + return -EINVAL; + } + return 0; +} + static int l2cap_sar_reassembly_sdu(struct sock *sk, struct sk_buff *skb, u16 control) { struct l2cap_pinfo *pi = l2cap_pi(sk); @@ -3174,6 +3251,9 @@ static inline int l2cap_data_channel(struct l2cap_conn *conn, u16 cid, struct sk if (__is_sar_start(control)) len -= 2; + if (pi->fcs == L2CAP_FCS_CRC16) + len -= 2; + /* * We can just drop the corrupted I-frame here. * Receiver will miss it and start proper recovery @@ -3182,6 +3262,9 @@ static inline int l2cap_data_channel(struct l2cap_conn *conn, u16 cid, struct sk if (len > L2CAP_DEFAULT_MAX_PDU_SIZE) goto drop; + if (l2cap_check_fcs(pi, skb)) + goto drop; + if (__is_iframe(control)) err = l2cap_data_channel_iframe(sk, control, skb); else @@ -3199,9 +3282,15 @@ static inline int l2cap_data_channel(struct l2cap_conn *conn, u16 cid, struct sk if (__is_sar_start(control)) len -= 2; + if (pi->fcs == L2CAP_FCS_CRC16) + len -= 2; + if (len > L2CAP_DEFAULT_MAX_PDU_SIZE || __is_sframe(control)) goto drop; + if (l2cap_check_fcs(pi, skb)) + goto drop; + tx_seq = __get_txseq(control); if (pi->expected_tx_seq == tx_seq) -- cgit v1.2.3 From 8f17154f1f70fcc6faa31ac82164fcf7f0599f38 Mon Sep 17 00:00:00 2001 From: "Gustavo F. Padovan" Date: Thu, 20 Aug 2009 22:26:03 -0300 Subject: Bluetooth: Add support for L2CAP SREJ exception When L2CAP loses an I-frame we send a SREJ frame to the transmitter side requesting the lost packet. This patch implement all Recv I-frame events on SREJ_SENT state table except the ones that deal with SendRej (the REJ exception at receiver side is yet not implemented). Signed-off-by: Gustavo F. Padovan Signed-off-by: Marcel Holtmann --- include/net/bluetooth/bluetooth.h | 1 + include/net/bluetooth/l2cap.h | 14 ++- net/bluetooth/l2cap.c | 220 +++++++++++++++++++++++++++++++++----- 3 files changed, 210 insertions(+), 25 deletions(-) (limited to 'include') diff --git a/include/net/bluetooth/bluetooth.h b/include/net/bluetooth/bluetooth.h index b8b9a8479525..718394e2c01e 100644 --- a/include/net/bluetooth/bluetooth.h +++ b/include/net/bluetooth/bluetooth.h @@ -140,6 +140,7 @@ struct bt_skb_cb { __u8 incoming; __u8 tx_seq; __u8 retries; + __u8 sar; }; #define bt_cb(skb) ((struct bt_skb_cb *)((skb)->cb)) diff --git a/include/net/bluetooth/l2cap.h b/include/net/bluetooth/l2cap.h index 59b26bf10f30..9f2126a4f6f5 100644 --- a/include/net/bluetooth/l2cap.h +++ b/include/net/bluetooth/l2cap.h @@ -108,6 +108,7 @@ struct l2cap_conninfo { #define L2CAP_CTRL_TXSEQ_SHIFT 1 #define L2CAP_CTRL_REQSEQ_SHIFT 8 +#define L2CAP_CTRL_SAR_SHIFT 14 /* L2CAP Supervisory Function */ #define L2CAP_SUPER_RCV_READY 0x0000 @@ -290,6 +291,13 @@ struct l2cap_conn { /* ----- L2CAP channel and socket info ----- */ #define l2cap_pi(sk) ((struct l2cap_pinfo *) sk) #define TX_QUEUE(sk) (&l2cap_pi(sk)->tx_queue) +#define SREJ_QUEUE(sk) (&l2cap_pi(sk)->srej_queue) +#define SREJ_LIST(sk) (&l2cap_pi(sk)->srej_l.list) + +struct srej_list { + __u8 tx_seq; + struct list_head list; +}; struct l2cap_pinfo { struct bt_sock bt; @@ -318,6 +326,8 @@ struct l2cap_pinfo { __u8 expected_ack_seq; __u8 req_seq; __u8 expected_tx_seq; + __u8 buffer_seq; + __u8 buffer_seq_srej; __u8 unacked_frames; __u8 retry_count; __u8 num_to_ack; @@ -338,6 +348,8 @@ struct l2cap_pinfo { struct timer_list retrans_timer; struct timer_list monitor_timer; struct sk_buff_head tx_queue; + struct sk_buff_head srej_queue; + struct srej_list srej_l; struct l2cap_conn *conn; struct sock *next_c; struct sock *prev_c; @@ -356,7 +368,7 @@ struct l2cap_pinfo { #define L2CAP_CONF_MAX_CONF_RSP 2 #define L2CAP_CONN_SAR_SDU 0x01 -#define L2CAP_CONN_UNDER_REJ 0x02 +#define L2CAP_CONN_SREJ_SENT 0x02 #define L2CAP_CONN_WAIT_F 0x04 #define __mod_retrans_timer() mod_timer(&l2cap_pi(sk)->retrans_timer, \ diff --git a/net/bluetooth/l2cap.c b/net/bluetooth/l2cap.c index 4c319003c290..70aff921db8c 100644 --- a/net/bluetooth/l2cap.c +++ b/net/bluetooth/l2cap.c @@ -1292,6 +1292,50 @@ static int l2cap_streaming_send(struct sock *sk) return 0; } +static int l2cap_retransmit_frame(struct sock *sk, u8 tx_seq) +{ + struct l2cap_pinfo *pi = l2cap_pi(sk); + struct sk_buff *skb, *tx_skb; + u16 control, fcs; + int err; + + skb = skb_peek(TX_QUEUE(sk)); + do { + if (bt_cb(skb)->tx_seq != tx_seq) { + if (skb_queue_is_last(TX_QUEUE(sk), skb)) + break; + skb = skb_queue_next(TX_QUEUE(sk), skb); + continue; + } + + if (pi->remote_max_tx && + bt_cb(skb)->retries == pi->remote_max_tx) { + l2cap_send_disconn_req(pi->conn, sk); + break; + } + + tx_skb = skb_clone(skb, GFP_ATOMIC); + bt_cb(skb)->retries++; + control = get_unaligned_le16(tx_skb->data + L2CAP_HDR_SIZE); + control |= (pi->req_seq << L2CAP_CTRL_REQSEQ_SHIFT) + | (tx_seq << L2CAP_CTRL_TXSEQ_SHIFT); + put_unaligned_le16(control, tx_skb->data + L2CAP_HDR_SIZE); + + if (l2cap_pi(sk)->fcs == L2CAP_FCS_CRC16) { + fcs = crc16(0, (u8 *)tx_skb->data, tx_skb->len - 2); + put_unaligned_le16(fcs, tx_skb->data + tx_skb->len - 2); + } + + err = l2cap_do_send(sk, tx_skb); + if (err < 0) { + l2cap_send_disconn_req(pi->conn, sk); + return err; + } + break; + } while(1); + return 0; +} + static int l2cap_ertm_send(struct sock *sk) { struct sk_buff *skb, *tx_skb; @@ -2705,6 +2749,7 @@ static inline int l2cap_config_req(struct l2cap_conn *conn, struct l2cap_cmd_hdr l2cap_monitor_timeout, (unsigned long) sk); __skb_queue_head_init(TX_QUEUE(sk)); + __skb_queue_head_init(SREJ_QUEUE(sk)); l2cap_chan_ready(sk); goto unlock; } @@ -2784,8 +2829,10 @@ static inline int l2cap_config_rsp(struct l2cap_conn *conn, struct l2cap_cmd_hdr sk->sk_state = BT_CONNECTED; l2cap_pi(sk)->expected_tx_seq = 0; + l2cap_pi(sk)->buffer_seq = 0; l2cap_pi(sk)->num_to_ack = 0; __skb_queue_head_init(TX_QUEUE(sk)); + __skb_queue_head_init(SREJ_QUEUE(sk)); l2cap_chan_ready(sk); } @@ -2817,6 +2864,7 @@ static inline int l2cap_disconnect_req(struct l2cap_conn *conn, struct l2cap_cmd sk->sk_shutdown = SHUTDOWN_MASK; skb_queue_purge(TX_QUEUE(sk)); + skb_queue_purge(SREJ_QUEUE(sk)); del_timer(&l2cap_pi(sk)->retrans_timer); del_timer(&l2cap_pi(sk)->monitor_timer); @@ -2843,6 +2891,7 @@ static inline int l2cap_disconnect_rsp(struct l2cap_conn *conn, struct l2cap_cmd return 0; skb_queue_purge(TX_QUEUE(sk)); + skb_queue_purge(SREJ_QUEUE(sk)); del_timer(&l2cap_pi(sk)->retrans_timer); del_timer(&l2cap_pi(sk)->monitor_timer); @@ -3038,6 +3087,33 @@ static int l2cap_check_fcs(struct l2cap_pinfo *pi, struct sk_buff *skb) return 0; } +static void l2cap_add_to_srej_queue(struct sock *sk, struct sk_buff *skb, u8 tx_seq, u8 sar) +{ + struct sk_buff *next_skb; + + bt_cb(skb)->tx_seq = tx_seq; + bt_cb(skb)->sar = sar; + + next_skb = skb_peek(SREJ_QUEUE(sk)); + if (!next_skb) { + __skb_queue_tail(SREJ_QUEUE(sk), skb); + return; + } + + do { + if (bt_cb(next_skb)->tx_seq > tx_seq) { + __skb_queue_before(SREJ_QUEUE(sk), next_skb, skb); + return; + } + + if (skb_queue_is_last(SREJ_QUEUE(sk), next_skb)) + break; + + } while((next_skb = skb_queue_next(SREJ_QUEUE(sk), next_skb))); + + __skb_queue_tail(SREJ_QUEUE(sk), skb); +} + static int l2cap_sar_reassembly_sdu(struct sock *sk, struct sk_buff *skb, u16 control) { struct l2cap_pinfo *pi = l2cap_pi(sk); @@ -3118,50 +3194,143 @@ static int l2cap_sar_reassembly_sdu(struct sock *sk, struct sk_buff *skb, u16 co return err; } +static void l2cap_check_srej_gap(struct sock *sk, u8 tx_seq) +{ + struct sk_buff *skb; + u16 control = 0; + + while((skb = skb_peek(SREJ_QUEUE(sk)))) { + if (bt_cb(skb)->tx_seq != tx_seq) + break; + + skb = skb_dequeue(SREJ_QUEUE(sk)); + control |= bt_cb(skb)->sar << L2CAP_CTRL_SAR_SHIFT; + l2cap_sar_reassembly_sdu(sk, skb, control); + l2cap_pi(sk)->buffer_seq_srej = + (l2cap_pi(sk)->buffer_seq_srej + 1) % 64; + tx_seq++; + } +} + +static void l2cap_resend_srejframe(struct sock *sk, u8 tx_seq) +{ + struct l2cap_pinfo *pi = l2cap_pi(sk); + struct srej_list *l, *tmp; + u16 control; + + list_for_each_entry_safe(l,tmp, SREJ_LIST(sk), list) { + if (l->tx_seq == tx_seq) { + list_del(&l->list); + kfree(l); + return; + } + control = L2CAP_SUPER_SELECT_REJECT; + control |= l->tx_seq << L2CAP_CTRL_REQSEQ_SHIFT; + l2cap_send_sframe(pi, control); + list_del(&l->list); + list_add_tail(&l->list, SREJ_LIST(sk)); + } +} + +static void l2cap_send_srejframe(struct sock *sk, u8 tx_seq) +{ + struct l2cap_pinfo *pi = l2cap_pi(sk); + struct srej_list *new; + u16 control; + + while (tx_seq != pi->expected_tx_seq) { + control = L2CAP_SUPER_SELECT_REJECT; + control |= pi->expected_tx_seq << L2CAP_CTRL_REQSEQ_SHIFT; + l2cap_send_sframe(pi, control); + + new = kzalloc(sizeof(struct srej_list), GFP_ATOMIC); + new->tx_seq = pi->expected_tx_seq++; + list_add_tail(&new->list, SREJ_LIST(sk)); + } + pi->expected_tx_seq++; +} + static inline int l2cap_data_channel_iframe(struct sock *sk, u16 rx_control, struct sk_buff *skb) { struct l2cap_pinfo *pi = l2cap_pi(sk); u8 tx_seq = __get_txseq(rx_control); u16 tx_control = 0; + u8 sar = rx_control >> L2CAP_CTRL_SAR_SHIFT; int err = 0; BT_DBG("sk %p rx_control 0x%4.4x len %d", sk, rx_control, skb->len); - if (tx_seq == pi->expected_tx_seq) { - if (pi->conn_state & L2CAP_CONN_UNDER_REJ) - pi->conn_state &= ~L2CAP_CONN_UNDER_REJ; + if (tx_seq == pi->expected_tx_seq) + goto expected; - err = l2cap_sar_reassembly_sdu(sk, skb, rx_control); - if (err < 0) - return err; + if (pi->conn_state & L2CAP_CONN_SREJ_SENT) { + struct srej_list *first; - pi->expected_tx_seq = (pi->expected_tx_seq + 1) % 64; - pi->num_to_ack = (pi->num_to_ack + 1) % L2CAP_DEFAULT_NUM_TO_ACK; - if (pi->num_to_ack == L2CAP_DEFAULT_NUM_TO_ACK - 1) { - tx_control |= L2CAP_SUPER_RCV_READY; - tx_control |= pi->expected_tx_seq << L2CAP_CTRL_REQSEQ_SHIFT; - goto send; + first = list_first_entry(SREJ_LIST(sk), + struct srej_list, list); + if (tx_seq == first->tx_seq) { + l2cap_add_to_srej_queue(sk, skb, tx_seq, sar); + l2cap_check_srej_gap(sk, tx_seq); + + list_del(&first->list); + kfree(first); + + if (list_empty(SREJ_LIST(sk))) { + pi->buffer_seq = pi->buffer_seq_srej; + pi->conn_state &= ~L2CAP_CONN_SREJ_SENT; + } + } else { + struct srej_list *l; + l2cap_add_to_srej_queue(sk, skb, tx_seq, sar); + + list_for_each_entry(l, SREJ_LIST(sk), list) { + if (l->tx_seq == tx_seq) { + l2cap_resend_srejframe(sk, tx_seq); + return 0; + } + } + l2cap_send_srejframe(sk, tx_seq); } } else { - /* Unexpected txSeq. Send a REJ S-frame */ - kfree_skb(skb); - if (!(pi->conn_state & L2CAP_CONN_UNDER_REJ)) { - tx_control |= L2CAP_SUPER_REJECT; - tx_control |= pi->expected_tx_seq << L2CAP_CTRL_REQSEQ_SHIFT; - pi->conn_state |= L2CAP_CONN_UNDER_REJ; + pi->conn_state |= L2CAP_CONN_SREJ_SENT; - goto send; - } + INIT_LIST_HEAD(SREJ_LIST(sk)); + pi->buffer_seq_srej = pi->buffer_seq; + + __skb_queue_head_init(SREJ_QUEUE(sk)); + l2cap_add_to_srej_queue(sk, skb, tx_seq, sar); + + l2cap_send_srejframe(sk, tx_seq); } return 0; -send: - return l2cap_send_sframe(pi, tx_control); +expected: + pi->expected_tx_seq = (pi->expected_tx_seq + 1) % 64; + + if (pi->conn_state & L2CAP_CONN_SREJ_SENT) { + l2cap_add_to_srej_queue(sk, skb, tx_seq, sar); + return 0; + } + + pi->buffer_seq = (pi->buffer_seq + 1) % 64; + + err = l2cap_sar_reassembly_sdu(sk, skb, rx_control); + if (err < 0) + return err; + + pi->num_to_ack = (pi->num_to_ack + 1) % L2CAP_DEFAULT_NUM_TO_ACK; + if (pi->num_to_ack == L2CAP_DEFAULT_NUM_TO_ACK - 1) { + tx_control |= L2CAP_SUPER_RCV_READY; + tx_control |= pi->buffer_seq << L2CAP_CTRL_REQSEQ_SHIFT; + l2cap_send_sframe(pi, tx_control); + } + return 0; } static inline int l2cap_data_channel_sframe(struct sock *sk, u16 rx_control, struct sk_buff *skb) { struct l2cap_pinfo *pi = l2cap_pi(sk); + u8 tx_seq = __get_reqseq(rx_control); BT_DBG("sk %p rx_control 0x%4.4x len %d", sk, rx_control, skb->len); @@ -3181,7 +3350,7 @@ static inline int l2cap_data_channel_sframe(struct sock *sk, u16 rx_control, str if (pi->unacked_frames > 0) __mod_retrans_timer(); } else { - pi->expected_ack_seq = __get_reqseq(rx_control); + pi->expected_ack_seq = tx_seq; l2cap_drop_acked_frames(sk); if (pi->unacked_frames > 0) __mod_retrans_timer(); @@ -3200,8 +3369,11 @@ static inline int l2cap_data_channel_sframe(struct sock *sk, u16 rx_control, str break; - case L2CAP_SUPER_RCV_NOT_READY: case L2CAP_SUPER_SELECT_REJECT: + l2cap_retransmit_frame(sk, tx_seq); + break; + + case L2CAP_SUPER_RCV_NOT_READY: break; } -- cgit v1.2.3 From ef54fd937fbd5ebaeb023818524565bd526a5f36 Mon Sep 17 00:00:00 2001 From: "Gustavo F. Padovan" Date: Thu, 20 Aug 2009 22:26:04 -0300 Subject: Bluetooth: Full support for receiving L2CAP SREJ frames Support for receiving of SREJ frames as specified by the state table. Signed-off-by: Gustavo F. Padovan Signed-off-by: Marcel Holtmann --- include/net/bluetooth/l2cap.h | 3 +++ net/bluetooth/l2cap.c | 30 +++++++++++++++++++++++++++++- 2 files changed, 32 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/net/bluetooth/l2cap.h b/include/net/bluetooth/l2cap.h index 9f2126a4f6f5..7ca614ac5d43 100644 --- a/include/net/bluetooth/l2cap.h +++ b/include/net/bluetooth/l2cap.h @@ -328,6 +328,7 @@ struct l2cap_pinfo { __u8 expected_tx_seq; __u8 buffer_seq; __u8 buffer_seq_srej; + __u8 srej_save_reqseq; __u8 unacked_frames; __u8 retry_count; __u8 num_to_ack; @@ -370,6 +371,8 @@ struct l2cap_pinfo { #define L2CAP_CONN_SAR_SDU 0x01 #define L2CAP_CONN_SREJ_SENT 0x02 #define L2CAP_CONN_WAIT_F 0x04 +#define L2CAP_CONN_SREJ_ACT 0x08 +#define L2CAP_CONN_SEND_PBIT 0x10 #define __mod_retrans_timer() mod_timer(&l2cap_pi(sk)->retrans_timer, \ jiffies + msecs_to_jiffies(L2CAP_DEFAULT_RETRANS_TO)); diff --git a/net/bluetooth/l2cap.c b/net/bluetooth/l2cap.c index 70aff921db8c..c04526f3df2e 100644 --- a/net/bluetooth/l2cap.c +++ b/net/bluetooth/l2cap.c @@ -3241,6 +3241,10 @@ static void l2cap_send_srejframe(struct sock *sk, u8 tx_seq) while (tx_seq != pi->expected_tx_seq) { control = L2CAP_SUPER_SELECT_REJECT; control |= pi->expected_tx_seq << L2CAP_CTRL_REQSEQ_SHIFT; + if (pi->conn_state & L2CAP_CONN_SEND_PBIT) { + control |= L2CAP_CTRL_POLL; + pi->conn_state &= ~L2CAP_CONN_SEND_PBIT; + } l2cap_send_sframe(pi, control); new = kzalloc(sizeof(struct srej_list), GFP_ATOMIC); @@ -3300,6 +3304,8 @@ static inline int l2cap_data_channel_iframe(struct sock *sk, u16 rx_control, str __skb_queue_head_init(SREJ_QUEUE(sk)); l2cap_add_to_srej_queue(sk, skb, tx_seq, sar); + pi->conn_state |= L2CAP_CONN_SEND_PBIT; + l2cap_send_srejframe(sk, tx_seq); } return 0; @@ -3370,7 +3376,29 @@ static inline int l2cap_data_channel_sframe(struct sock *sk, u16 rx_control, str break; case L2CAP_SUPER_SELECT_REJECT: - l2cap_retransmit_frame(sk, tx_seq); + if (rx_control & L2CAP_CTRL_POLL) { + l2cap_retransmit_frame(sk, tx_seq); + pi->expected_ack_seq = tx_seq; + l2cap_drop_acked_frames(sk); + l2cap_ertm_send(sk); + if (pi->conn_state & L2CAP_CONN_WAIT_F) { + pi->srej_save_reqseq = tx_seq; + pi->conn_state |= L2CAP_CONN_SREJ_ACT; + } + } else if (rx_control & L2CAP_CTRL_FINAL) { + if ((pi->conn_state & L2CAP_CONN_SREJ_ACT) && + pi->srej_save_reqseq == tx_seq) + pi->srej_save_reqseq &= ~L2CAP_CONN_SREJ_ACT; + else + l2cap_retransmit_frame(sk, tx_seq); + } + else { + l2cap_retransmit_frame(sk, tx_seq); + if (pi->conn_state & L2CAP_CONN_WAIT_F) { + pi->srej_save_reqseq = tx_seq; + pi->conn_state |= L2CAP_CONN_SREJ_ACT; + } + } break; case L2CAP_SUPER_RCV_NOT_READY: -- cgit v1.2.3 From 5e928f77a09a07f9dd595bb8a489965d69a83458 Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Tue, 18 Aug 2009 23:38:32 +0200 Subject: PM: Introduce core framework for run-time PM of I/O devices (rev. 17) Introduce a core framework for run-time power management of I/O devices. Add device run-time PM fields to 'struct dev_pm_info' and device run-time PM callbacks to 'struct dev_pm_ops'. Introduce a run-time PM workqueue and define some device run-time PM helper functions at the core level. Document all these things. Special thanks to Alan Stern for his help with the design and multiple detailed reviews of the pereceding versions of this patch and to Magnus Damm for testing feedback. Signed-off-by: Rafael J. Wysocki Acked-by: Magnus Damm --- Documentation/power/runtime_pm.txt | 378 ++++++++++++++ drivers/base/dd.c | 11 + drivers/base/power/Makefile | 1 + drivers/base/power/main.c | 22 +- drivers/base/power/power.h | 31 +- drivers/base/power/runtime.c | 1011 ++++++++++++++++++++++++++++++++++++ include/linux/pm.h | 101 +++- include/linux/pm_runtime.h | 114 ++++ kernel/power/Kconfig | 14 + kernel/power/main.c | 17 + 10 files changed, 1689 insertions(+), 11 deletions(-) create mode 100644 Documentation/power/runtime_pm.txt create mode 100644 drivers/base/power/runtime.c create mode 100644 include/linux/pm_runtime.h (limited to 'include') diff --git a/Documentation/power/runtime_pm.txt b/Documentation/power/runtime_pm.txt new file mode 100644 index 000000000000..f49a33b704d2 --- /dev/null +++ b/Documentation/power/runtime_pm.txt @@ -0,0 +1,378 @@ +Run-time Power Management Framework for I/O Devices + +(C) 2009 Rafael J. Wysocki , Novell Inc. + +1. Introduction + +Support for run-time power management (run-time PM) of I/O devices is provided +at the power management core (PM core) level by means of: + +* The power management workqueue pm_wq in which bus types and device drivers can + put their PM-related work items. It is strongly recommended that pm_wq be + used for queuing all work items related to run-time PM, because this allows + them to be synchronized with system-wide power transitions (suspend to RAM, + hibernation and resume from system sleep states). pm_wq is declared in + include/linux/pm_runtime.h and defined in kernel/power/main.c. + +* A number of run-time PM fields in the 'power' member of 'struct device' (which + is of the type 'struct dev_pm_info', defined in include/linux/pm.h) that can + be used for synchronizing run-time PM operations with one another. + +* Three device run-time PM callbacks in 'struct dev_pm_ops' (defined in + include/linux/pm.h). + +* A set of helper functions defined in drivers/base/power/runtime.c that can be + used for carrying out run-time PM operations in such a way that the + synchronization between them is taken care of by the PM core. Bus types and + device drivers are encouraged to use these functions. + +The run-time PM callbacks present in 'struct dev_pm_ops', the device run-time PM +fields of 'struct dev_pm_info' and the core helper functions provided for +run-time PM are described below. + +2. Device Run-time PM Callbacks + +There are three device run-time PM callbacks defined in 'struct dev_pm_ops': + +struct dev_pm_ops { + ... + int (*runtime_suspend)(struct device *dev); + int (*runtime_resume)(struct device *dev); + void (*runtime_idle)(struct device *dev); + ... +}; + +The ->runtime_suspend() callback is executed by the PM core for the bus type of +the device being suspended. The bus type's callback is then _entirely_ +_responsible_ for handling the device as appropriate, which may, but need not +include executing the device driver's own ->runtime_suspend() callback (from the +PM core's point of view it is not necessary to implement a ->runtime_suspend() +callback in a device driver as long as the bus type's ->runtime_suspend() knows +what to do to handle the device). + + * Once the bus type's ->runtime_suspend() callback has completed successfully + for given device, the PM core regards the device as suspended, which need + not mean that the device has been put into a low power state. It is + supposed to mean, however, that the device will not process data and will + not communicate with the CPU(s) and RAM until its bus type's + ->runtime_resume() callback is executed for it. The run-time PM status of + a device after successful execution of its bus type's ->runtime_suspend() + callback is 'suspended'. + + * If the bus type's ->runtime_suspend() callback returns -EBUSY or -EAGAIN, + the device's run-time PM status is supposed to be 'active', which means that + the device _must_ be fully operational afterwards. + + * If the bus type's ->runtime_suspend() callback returns an error code + different from -EBUSY or -EAGAIN, the PM core regards this as a fatal + error and will refuse to run the helper functions described in Section 4 + for the device, until the status of it is directly set either to 'active' + or to 'suspended' (the PM core provides special helper functions for this + purpose). + +In particular, if the driver requires remote wakeup capability for proper +functioning and device_may_wakeup() returns 'false' for the device, then +->runtime_suspend() should return -EBUSY. On the other hand, if +device_may_wakeup() returns 'true' for the device and the device is put +into a low power state during the execution of its bus type's +->runtime_suspend(), it is expected that remote wake-up (i.e. hardware mechanism +allowing the device to request a change of its power state, such as PCI PME) +will be enabled for the device. Generally, remote wake-up should be enabled +for all input devices put into a low power state at run time. + +The ->runtime_resume() callback is executed by the PM core for the bus type of +the device being woken up. The bus type's callback is then _entirely_ +_responsible_ for handling the device as appropriate, which may, but need not +include executing the device driver's own ->runtime_resume() callback (from the +PM core's point of view it is not necessary to implement a ->runtime_resume() +callback in a device driver as long as the bus type's ->runtime_resume() knows +what to do to handle the device). + + * Once the bus type's ->runtime_resume() callback has completed successfully, + the PM core regards the device as fully operational, which means that the + device _must_ be able to complete I/O operations as needed. The run-time + PM status of the device is then 'active'. + + * If the bus type's ->runtime_resume() callback returns an error code, the PM + core regards this as a fatal error and will refuse to run the helper + functions described in Section 4 for the device, until its status is + directly set either to 'active' or to 'suspended' (the PM core provides + special helper functions for this purpose). + +The ->runtime_idle() callback is executed by the PM core for the bus type of +given device whenever the device appears to be idle, which is indicated to the +PM core by two counters, the device's usage counter and the counter of 'active' +children of the device. + + * If any of these counters is decreased using a helper function provided by + the PM core and it turns out to be equal to zero, the other counter is + checked. If that counter also is equal to zero, the PM core executes the + device bus type's ->runtime_idle() callback (with the device as an + argument). + +The action performed by a bus type's ->runtime_idle() callback is totally +dependent on the bus type in question, but the expected and recommended action +is to check if the device can be suspended (i.e. if all of the conditions +necessary for suspending the device are satisfied) and to queue up a suspend +request for the device in that case. + +The helper functions provided by the PM core, described in Section 4, guarantee +that the following constraints are met with respect to the bus type's run-time +PM callbacks: + +(1) The callbacks are mutually exclusive (e.g. it is forbidden to execute + ->runtime_suspend() in parallel with ->runtime_resume() or with another + instance of ->runtime_suspend() for the same device) with the exception that + ->runtime_suspend() or ->runtime_resume() can be executed in parallel with + ->runtime_idle() (although ->runtime_idle() will not be started while any + of the other callbacks is being executed for the same device). + +(2) ->runtime_idle() and ->runtime_suspend() can only be executed for 'active' + devices (i.e. the PM core will only execute ->runtime_idle() or + ->runtime_suspend() for the devices the run-time PM status of which is + 'active'). + +(3) ->runtime_idle() and ->runtime_suspend() can only be executed for a device + the usage counter of which is equal to zero _and_ either the counter of + 'active' children of which is equal to zero, or the 'power.ignore_children' + flag of which is set. + +(4) ->runtime_resume() can only be executed for 'suspended' devices (i.e. the + PM core will only execute ->runtime_resume() for the devices the run-time + PM status of which is 'suspended'). + +Additionally, the helper functions provided by the PM core obey the following +rules: + + * If ->runtime_suspend() is about to be executed or there's a pending request + to execute it, ->runtime_idle() will not be executed for the same device. + + * A request to execute or to schedule the execution of ->runtime_suspend() + will cancel any pending requests to execute ->runtime_idle() for the same + device. + + * If ->runtime_resume() is about to be executed or there's a pending request + to execute it, the other callbacks will not be executed for the same device. + + * A request to execute ->runtime_resume() will cancel any pending or + scheduled requests to execute the other callbacks for the same device. + +3. Run-time PM Device Fields + +The following device run-time PM fields are present in 'struct dev_pm_info', as +defined in include/linux/pm.h: + + struct timer_list suspend_timer; + - timer used for scheduling (delayed) suspend request + + unsigned long timer_expires; + - timer expiration time, in jiffies (if this is different from zero, the + timer is running and will expire at that time, otherwise the timer is not + running) + + struct work_struct work; + - work structure used for queuing up requests (i.e. work items in pm_wq) + + wait_queue_head_t wait_queue; + - wait queue used if any of the helper functions needs to wait for another + one to complete + + spinlock_t lock; + - lock used for synchronisation + + atomic_t usage_count; + - the usage counter of the device + + atomic_t child_count; + - the count of 'active' children of the device + + unsigned int ignore_children; + - if set, the value of child_count is ignored (but still updated) + + unsigned int disable_depth; + - used for disabling the helper funcions (they work normally if this is + equal to zero); the initial value of it is 1 (i.e. run-time PM is + initially disabled for all devices) + + unsigned int runtime_error; + - if set, there was a fatal error (one of the callbacks returned error code + as described in Section 2), so the helper funtions will not work until + this flag is cleared; this is the error code returned by the failing + callback + + unsigned int idle_notification; + - if set, ->runtime_idle() is being executed + + unsigned int request_pending; + - if set, there's a pending request (i.e. a work item queued up into pm_wq) + + enum rpm_request request; + - type of request that's pending (valid if request_pending is set) + + unsigned int deferred_resume; + - set if ->runtime_resume() is about to be run while ->runtime_suspend() is + being executed for that device and it is not practical to wait for the + suspend to complete; means "start a resume as soon as you've suspended" + + enum rpm_status runtime_status; + - the run-time PM status of the device; this field's initial value is + RPM_SUSPENDED, which means that each device is initially regarded by the + PM core as 'suspended', regardless of its real hardware status + +All of the above fields are members of the 'power' member of 'struct device'. + +4. Run-time PM Device Helper Functions + +The following run-time PM helper functions are defined in +drivers/base/power/runtime.c and include/linux/pm_runtime.h: + + void pm_runtime_init(struct device *dev); + - initialize the device run-time PM fields in 'struct dev_pm_info' + + void pm_runtime_remove(struct device *dev); + - make sure that the run-time PM of the device will be disabled after + removing the device from device hierarchy + + int pm_runtime_idle(struct device *dev); + - execute ->runtime_idle() for the device's bus type; returns 0 on success + or error code on failure, where -EINPROGRESS means that ->runtime_idle() + is already being executed + + int pm_runtime_suspend(struct device *dev); + - execute ->runtime_suspend() for the device's bus type; returns 0 on + success, 1 if the device's run-time PM status was already 'suspended', or + error code on failure, where -EAGAIN or -EBUSY means it is safe to attempt + to suspend the device again in future + + int pm_runtime_resume(struct device *dev); + - execute ->runtime_resume() for the device's bus type; returns 0 on + success, 1 if the device's run-time PM status was already 'active' or + error code on failure, where -EAGAIN means it may be safe to attempt to + resume the device again in future, but 'power.runtime_error' should be + checked additionally + + int pm_request_idle(struct device *dev); + - submit a request to execute ->runtime_idle() for the device's bus type + (the request is represented by a work item in pm_wq); returns 0 on success + or error code if the request has not been queued up + + int pm_schedule_suspend(struct device *dev, unsigned int delay); + - schedule the execution of ->runtime_suspend() for the device's bus type + in future, where 'delay' is the time to wait before queuing up a suspend + work item in pm_wq, in milliseconds (if 'delay' is zero, the work item is + queued up immediately); returns 0 on success, 1 if the device's PM + run-time status was already 'suspended', or error code if the request + hasn't been scheduled (or queued up if 'delay' is 0); if the execution of + ->runtime_suspend() is already scheduled and not yet expired, the new + value of 'delay' will be used as the time to wait + + int pm_request_resume(struct device *dev); + - submit a request to execute ->runtime_resume() for the device's bus type + (the request is represented by a work item in pm_wq); returns 0 on + success, 1 if the device's run-time PM status was already 'active', or + error code if the request hasn't been queued up + + void pm_runtime_get_noresume(struct device *dev); + - increment the device's usage counter + + int pm_runtime_get(struct device *dev); + - increment the device's usage counter, run pm_request_resume(dev) and + return its result + + int pm_runtime_get_sync(struct device *dev); + - increment the device's usage counter, run pm_runtime_resume(dev) and + return its result + + void pm_runtime_put_noidle(struct device *dev); + - decrement the device's usage counter + + int pm_runtime_put(struct device *dev); + - decrement the device's usage counter, run pm_request_idle(dev) and return + its result + + int pm_runtime_put_sync(struct device *dev); + - decrement the device's usage counter, run pm_runtime_idle(dev) and return + its result + + void pm_runtime_enable(struct device *dev); + - enable the run-time PM helper functions to run the device bus type's + run-time PM callbacks described in Section 2 + + int pm_runtime_disable(struct device *dev); + - prevent the run-time PM helper functions from running the device bus + type's run-time PM callbacks, make sure that all of the pending run-time + PM operations on the device are either completed or canceled; returns + 1 if there was a resume request pending and it was necessary to execute + ->runtime_resume() for the device's bus type to satisfy that request, + otherwise 0 is returned + + void pm_suspend_ignore_children(struct device *dev, bool enable); + - set/unset the power.ignore_children flag of the device + + int pm_runtime_set_active(struct device *dev); + - clear the device's 'power.runtime_error' flag, set the device's run-time + PM status to 'active' and update its parent's counter of 'active' + children as appropriate (it is only valid to use this function if + 'power.runtime_error' is set or 'power.disable_depth' is greater than + zero); it will fail and return error code if the device has a parent + which is not active and the 'power.ignore_children' flag of which is unset + + void pm_runtime_set_suspended(struct device *dev); + - clear the device's 'power.runtime_error' flag, set the device's run-time + PM status to 'suspended' and update its parent's counter of 'active' + children as appropriate (it is only valid to use this function if + 'power.runtime_error' is set or 'power.disable_depth' is greater than + zero) + +It is safe to execute the following helper functions from interrupt context: + +pm_request_idle() +pm_schedule_suspend() +pm_request_resume() +pm_runtime_get_noresume() +pm_runtime_get() +pm_runtime_put_noidle() +pm_runtime_put() +pm_suspend_ignore_children() +pm_runtime_set_active() +pm_runtime_set_suspended() +pm_runtime_enable() + +5. Run-time PM Initialization, Device Probing and Removal + +Initially, the run-time PM is disabled for all devices, which means that the +majority of the run-time PM helper funtions described in Section 4 will return +-EAGAIN until pm_runtime_enable() is called for the device. + +In addition to that, the initial run-time PM status of all devices is +'suspended', but it need not reflect the actual physical state of the device. +Thus, if the device is initially active (i.e. it is able to process I/O), its +run-time PM status must be changed to 'active', with the help of +pm_runtime_set_active(), before pm_runtime_enable() is called for the device. + +However, if the device has a parent and the parent's run-time PM is enabled, +calling pm_runtime_set_active() for the device will affect the parent, unless +the parent's 'power.ignore_children' flag is set. Namely, in that case the +parent won't be able to suspend at run time, using the PM core's helper +functions, as long as the child's status is 'active', even if the child's +run-time PM is still disabled (i.e. pm_runtime_enable() hasn't been called for +the child yet or pm_runtime_disable() has been called for it). For this reason, +once pm_runtime_set_active() has been called for the device, pm_runtime_enable() +should be called for it too as soon as reasonably possible or its run-time PM +status should be changed back to 'suspended' with the help of +pm_runtime_set_suspended(). + +If the default initial run-time PM status of the device (i.e. 'suspended') +reflects the actual state of the device, its bus type's or its driver's +->probe() callback will likely need to wake it up using one of the PM core's +helper functions described in Section 4. In that case, pm_runtime_resume() +should be used. Of course, for this purpose the device's run-time PM has to be +enabled earlier by calling pm_runtime_enable(). + +If the device bus type's or driver's ->probe() or ->remove() callback runs +pm_runtime_suspend() or pm_runtime_idle() or their asynchronous counterparts, +they will fail returning -EAGAIN, because the device's usage counter is +incremented by the core before executing ->probe() and ->remove(). Still, it +may be desirable to suspend the device as soon as ->probe() or ->remove() has +finished, so the PM core uses pm_runtime_idle_sync() to invoke the device bus +type's ->runtime_idle() callback at that time. diff --git a/drivers/base/dd.c b/drivers/base/dd.c index f0106875f01d..7b34b3a48f67 100644 --- a/drivers/base/dd.c +++ b/drivers/base/dd.c @@ -23,6 +23,7 @@ #include #include #include +#include #include "base.h" #include "power/power.h" @@ -202,7 +203,10 @@ int driver_probe_device(struct device_driver *drv, struct device *dev) pr_debug("bus: '%s': %s: matched device %s with driver %s\n", drv->bus->name, __func__, dev_name(dev), drv->name); + pm_runtime_get_noresume(dev); + pm_runtime_barrier(dev); ret = really_probe(dev, drv); + pm_runtime_put_sync(dev); return ret; } @@ -245,7 +249,9 @@ int device_attach(struct device *dev) ret = 0; } } else { + pm_runtime_get_noresume(dev); ret = bus_for_each_drv(dev->bus, NULL, dev, __device_attach); + pm_runtime_put_sync(dev); } up(&dev->sem); return ret; @@ -306,6 +312,9 @@ static void __device_release_driver(struct device *dev) drv = dev->driver; if (drv) { + pm_runtime_get_noresume(dev); + pm_runtime_barrier(dev); + driver_sysfs_remove(dev); if (dev->bus) @@ -324,6 +333,8 @@ static void __device_release_driver(struct device *dev) blocking_notifier_call_chain(&dev->bus->p->bus_notifier, BUS_NOTIFY_UNBOUND_DRIVER, dev); + + pm_runtime_put_sync(dev); } } diff --git a/drivers/base/power/Makefile b/drivers/base/power/Makefile index 911208b89259..3ce3519e8f30 100644 --- a/drivers/base/power/Makefile +++ b/drivers/base/power/Makefile @@ -1,5 +1,6 @@ obj-$(CONFIG_PM) += sysfs.o obj-$(CONFIG_PM_SLEEP) += main.o +obj-$(CONFIG_PM_RUNTIME) += runtime.o obj-$(CONFIG_PM_TRACE_RTC) += trace.o ccflags-$(CONFIG_DEBUG_DRIVER) := -DDEBUG diff --git a/drivers/base/power/main.c b/drivers/base/power/main.c index 1b1a786b7dec..86990011277b 100644 --- a/drivers/base/power/main.c +++ b/drivers/base/power/main.c @@ -21,6 +21,7 @@ #include #include #include +#include #include #include #include @@ -48,6 +49,16 @@ static DEFINE_MUTEX(dpm_list_mtx); */ static bool transition_started; +/** + * device_pm_init - Initialize the PM-related part of a device object + * @dev: Device object being initialized. + */ +void device_pm_init(struct device *dev) +{ + dev->power.status = DPM_ON; + pm_runtime_init(dev); +} + /** * device_pm_lock - lock the list of active devices used by the PM core */ @@ -105,6 +116,7 @@ void device_pm_remove(struct device *dev) mutex_lock(&dpm_list_mtx); list_del_init(&dev->power.entry); mutex_unlock(&dpm_list_mtx); + pm_runtime_remove(dev); } /** @@ -512,6 +524,7 @@ static void dpm_complete(pm_message_t state) mutex_unlock(&dpm_list_mtx); device_complete(dev, state); + pm_runtime_put_noidle(dev); mutex_lock(&dpm_list_mtx); } @@ -757,7 +770,14 @@ static int dpm_prepare(pm_message_t state) dev->power.status = DPM_PREPARING; mutex_unlock(&dpm_list_mtx); - error = device_prepare(dev, state); + pm_runtime_get_noresume(dev); + if (pm_runtime_barrier(dev) && device_may_wakeup(dev)) { + /* Wake-up requested during system sleep transition. */ + pm_runtime_put_noidle(dev); + error = -EBUSY; + } else { + error = device_prepare(dev, state); + } mutex_lock(&dpm_list_mtx); if (error) { diff --git a/drivers/base/power/power.h b/drivers/base/power/power.h index c7cb4fc3735c..b8fa1aa5225a 100644 --- a/drivers/base/power/power.h +++ b/drivers/base/power/power.h @@ -1,7 +1,14 @@ -static inline void device_pm_init(struct device *dev) -{ - dev->power.status = DPM_ON; -} +#ifdef CONFIG_PM_RUNTIME + +extern void pm_runtime_init(struct device *dev); +extern void pm_runtime_remove(struct device *dev); + +#else /* !CONFIG_PM_RUNTIME */ + +static inline void pm_runtime_init(struct device *dev) {} +static inline void pm_runtime_remove(struct device *dev) {} + +#endif /* !CONFIG_PM_RUNTIME */ #ifdef CONFIG_PM_SLEEP @@ -16,23 +23,33 @@ static inline struct device *to_device(struct list_head *entry) return container_of(entry, struct device, power.entry); } +extern void device_pm_init(struct device *dev); extern void device_pm_add(struct device *); extern void device_pm_remove(struct device *); extern void device_pm_move_before(struct device *, struct device *); extern void device_pm_move_after(struct device *, struct device *); extern void device_pm_move_last(struct device *); -#else /* CONFIG_PM_SLEEP */ +#else /* !CONFIG_PM_SLEEP */ + +static inline void device_pm_init(struct device *dev) +{ + pm_runtime_init(dev); +} + +static inline void device_pm_remove(struct device *dev) +{ + pm_runtime_remove(dev); +} static inline void device_pm_add(struct device *dev) {} -static inline void device_pm_remove(struct device *dev) {} static inline void device_pm_move_before(struct device *deva, struct device *devb) {} static inline void device_pm_move_after(struct device *deva, struct device *devb) {} static inline void device_pm_move_last(struct device *dev) {} -#endif +#endif /* !CONFIG_PM_SLEEP */ #ifdef CONFIG_PM diff --git a/drivers/base/power/runtime.c b/drivers/base/power/runtime.c new file mode 100644 index 000000000000..38556f6cc22d --- /dev/null +++ b/drivers/base/power/runtime.c @@ -0,0 +1,1011 @@ +/* + * drivers/base/power/runtime.c - Helper functions for device run-time PM + * + * Copyright (c) 2009 Rafael J. Wysocki , Novell Inc. + * + * This file is released under the GPLv2. + */ + +#include +#include +#include + +static int __pm_runtime_resume(struct device *dev, bool from_wq); +static int __pm_request_idle(struct device *dev); +static int __pm_request_resume(struct device *dev); + +/** + * pm_runtime_deactivate_timer - Deactivate given device's suspend timer. + * @dev: Device to handle. + */ +static void pm_runtime_deactivate_timer(struct device *dev) +{ + if (dev->power.timer_expires > 0) { + del_timer(&dev->power.suspend_timer); + dev->power.timer_expires = 0; + } +} + +/** + * pm_runtime_cancel_pending - Deactivate suspend timer and cancel requests. + * @dev: Device to handle. + */ +static void pm_runtime_cancel_pending(struct device *dev) +{ + pm_runtime_deactivate_timer(dev); + /* + * In case there's a request pending, make sure its work function will + * return without doing anything. + */ + dev->power.request = RPM_REQ_NONE; +} + +/** + * __pm_runtime_idle - Notify device bus type if the device can be suspended. + * @dev: Device to notify the bus type about. + * + * This function must be called under dev->power.lock with interrupts disabled. + */ +static int __pm_runtime_idle(struct device *dev) + __releases(&dev->power.lock) __acquires(&dev->power.lock) +{ + int retval = 0; + + dev_dbg(dev, "__pm_runtime_idle()!\n"); + + if (dev->power.runtime_error) + retval = -EINVAL; + else if (dev->power.idle_notification) + retval = -EINPROGRESS; + else if (atomic_read(&dev->power.usage_count) > 0 + || dev->power.disable_depth > 0 + || dev->power.runtime_status != RPM_ACTIVE) + retval = -EAGAIN; + else if (!pm_children_suspended(dev)) + retval = -EBUSY; + if (retval) + goto out; + + if (dev->power.request_pending) { + /* + * If an idle notification request is pending, cancel it. Any + * other pending request takes precedence over us. + */ + if (dev->power.request == RPM_REQ_IDLE) { + dev->power.request = RPM_REQ_NONE; + } else if (dev->power.request != RPM_REQ_NONE) { + retval = -EAGAIN; + goto out; + } + } + + dev->power.idle_notification = true; + + if (dev->bus && dev->bus->pm && dev->bus->pm->runtime_idle) { + spin_unlock_irq(&dev->power.lock); + + dev->bus->pm->runtime_idle(dev); + + spin_lock_irq(&dev->power.lock); + } + + dev->power.idle_notification = false; + wake_up_all(&dev->power.wait_queue); + + out: + dev_dbg(dev, "__pm_runtime_idle() returns %d!\n", retval); + + return retval; +} + +/** + * pm_runtime_idle - Notify device bus type if the device can be suspended. + * @dev: Device to notify the bus type about. + */ +int pm_runtime_idle(struct device *dev) +{ + int retval; + + spin_lock_irq(&dev->power.lock); + retval = __pm_runtime_idle(dev); + spin_unlock_irq(&dev->power.lock); + + return retval; +} +EXPORT_SYMBOL_GPL(pm_runtime_idle); + +/** + * __pm_runtime_suspend - Carry out run-time suspend of given device. + * @dev: Device to suspend. + * @from_wq: If set, the function has been called via pm_wq. + * + * Check if the device can be suspended and run the ->runtime_suspend() callback + * provided by its bus type. If another suspend has been started earlier, wait + * for it to finish. If an idle notification or suspend request is pending or + * scheduled, cancel it. + * + * This function must be called under dev->power.lock with interrupts disabled. + */ +int __pm_runtime_suspend(struct device *dev, bool from_wq) + __releases(&dev->power.lock) __acquires(&dev->power.lock) +{ + struct device *parent = NULL; + bool notify = false; + int retval = 0; + + dev_dbg(dev, "__pm_runtime_suspend()%s!\n", + from_wq ? " from workqueue" : ""); + + repeat: + if (dev->power.runtime_error) { + retval = -EINVAL; + goto out; + } + + /* Pending resume requests take precedence over us. */ + if (dev->power.request_pending + && dev->power.request == RPM_REQ_RESUME) { + retval = -EAGAIN; + goto out; + } + + /* Other scheduled or pending requests need to be canceled. */ + pm_runtime_cancel_pending(dev); + + if (dev->power.runtime_status == RPM_SUSPENDED) + retval = 1; + else if (dev->power.runtime_status == RPM_RESUMING + || dev->power.disable_depth > 0 + || atomic_read(&dev->power.usage_count) > 0) + retval = -EAGAIN; + else if (!pm_children_suspended(dev)) + retval = -EBUSY; + if (retval) + goto out; + + if (dev->power.runtime_status == RPM_SUSPENDING) { + DEFINE_WAIT(wait); + + if (from_wq) { + retval = -EINPROGRESS; + goto out; + } + + /* Wait for the other suspend running in parallel with us. */ + for (;;) { + prepare_to_wait(&dev->power.wait_queue, &wait, + TASK_UNINTERRUPTIBLE); + if (dev->power.runtime_status != RPM_SUSPENDING) + break; + + spin_unlock_irq(&dev->power.lock); + + schedule(); + + spin_lock_irq(&dev->power.lock); + } + finish_wait(&dev->power.wait_queue, &wait); + goto repeat; + } + + dev->power.runtime_status = RPM_SUSPENDING; + + if (dev->bus && dev->bus->pm && dev->bus->pm->runtime_suspend) { + spin_unlock_irq(&dev->power.lock); + + retval = dev->bus->pm->runtime_suspend(dev); + + spin_lock_irq(&dev->power.lock); + dev->power.runtime_error = retval; + } else { + retval = -ENOSYS; + } + + if (retval) { + dev->power.runtime_status = RPM_ACTIVE; + pm_runtime_cancel_pending(dev); + dev->power.deferred_resume = false; + + if (retval == -EAGAIN || retval == -EBUSY) { + notify = true; + dev->power.runtime_error = 0; + } + } else { + dev->power.runtime_status = RPM_SUSPENDED; + + if (dev->parent) { + parent = dev->parent; + atomic_add_unless(&parent->power.child_count, -1, 0); + } + } + wake_up_all(&dev->power.wait_queue); + + if (dev->power.deferred_resume) { + dev->power.deferred_resume = false; + __pm_runtime_resume(dev, false); + retval = -EAGAIN; + goto out; + } + + if (notify) + __pm_runtime_idle(dev); + + if (parent && !parent->power.ignore_children) { + spin_unlock_irq(&dev->power.lock); + + pm_request_idle(parent); + + spin_lock_irq(&dev->power.lock); + } + + out: + dev_dbg(dev, "__pm_runtime_suspend() returns %d!\n", retval); + + return retval; +} + +/** + * pm_runtime_suspend - Carry out run-time suspend of given device. + * @dev: Device to suspend. + */ +int pm_runtime_suspend(struct device *dev) +{ + int retval; + + spin_lock_irq(&dev->power.lock); + retval = __pm_runtime_suspend(dev, false); + spin_unlock_irq(&dev->power.lock); + + return retval; +} +EXPORT_SYMBOL_GPL(pm_runtime_suspend); + +/** + * __pm_runtime_resume - Carry out run-time resume of given device. + * @dev: Device to resume. + * @from_wq: If set, the function has been called via pm_wq. + * + * Check if the device can be woken up and run the ->runtime_resume() callback + * provided by its bus type. If another resume has been started earlier, wait + * for it to finish. If there's a suspend running in parallel with this + * function, wait for it to finish and resume the device. Cancel any scheduled + * or pending requests. + * + * This function must be called under dev->power.lock with interrupts disabled. + */ +int __pm_runtime_resume(struct device *dev, bool from_wq) + __releases(&dev->power.lock) __acquires(&dev->power.lock) +{ + struct device *parent = NULL; + int retval = 0; + + dev_dbg(dev, "__pm_runtime_resume()%s!\n", + from_wq ? " from workqueue" : ""); + + repeat: + if (dev->power.runtime_error) { + retval = -EINVAL; + goto out; + } + + pm_runtime_cancel_pending(dev); + + if (dev->power.runtime_status == RPM_ACTIVE) + retval = 1; + else if (dev->power.disable_depth > 0) + retval = -EAGAIN; + if (retval) + goto out; + + if (dev->power.runtime_status == RPM_RESUMING + || dev->power.runtime_status == RPM_SUSPENDING) { + DEFINE_WAIT(wait); + + if (from_wq) { + if (dev->power.runtime_status == RPM_SUSPENDING) + dev->power.deferred_resume = true; + retval = -EINPROGRESS; + goto out; + } + + /* Wait for the operation carried out in parallel with us. */ + for (;;) { + prepare_to_wait(&dev->power.wait_queue, &wait, + TASK_UNINTERRUPTIBLE); + if (dev->power.runtime_status != RPM_RESUMING + && dev->power.runtime_status != RPM_SUSPENDING) + break; + + spin_unlock_irq(&dev->power.lock); + + schedule(); + + spin_lock_irq(&dev->power.lock); + } + finish_wait(&dev->power.wait_queue, &wait); + goto repeat; + } + + if (!parent && dev->parent) { + /* + * Increment the parent's resume counter and resume it if + * necessary. + */ + parent = dev->parent; + spin_unlock_irq(&dev->power.lock); + + pm_runtime_get_noresume(parent); + + spin_lock_irq(&parent->power.lock); + /* + * We can resume if the parent's run-time PM is disabled or it + * is set to ignore children. + */ + if (!parent->power.disable_depth + && !parent->power.ignore_children) { + __pm_runtime_resume(parent, false); + if (parent->power.runtime_status != RPM_ACTIVE) + retval = -EBUSY; + } + spin_unlock_irq(&parent->power.lock); + + spin_lock_irq(&dev->power.lock); + if (retval) + goto out; + goto repeat; + } + + dev->power.runtime_status = RPM_RESUMING; + + if (dev->bus && dev->bus->pm && dev->bus->pm->runtime_resume) { + spin_unlock_irq(&dev->power.lock); + + retval = dev->bus->pm->runtime_resume(dev); + + spin_lock_irq(&dev->power.lock); + dev->power.runtime_error = retval; + } else { + retval = -ENOSYS; + } + + if (retval) { + dev->power.runtime_status = RPM_SUSPENDED; + pm_runtime_cancel_pending(dev); + } else { + dev->power.runtime_status = RPM_ACTIVE; + if (parent) + atomic_inc(&parent->power.child_count); + } + wake_up_all(&dev->power.wait_queue); + + if (!retval) + __pm_request_idle(dev); + + out: + if (parent) { + spin_unlock_irq(&dev->power.lock); + + pm_runtime_put(parent); + + spin_lock_irq(&dev->power.lock); + } + + dev_dbg(dev, "__pm_runtime_resume() returns %d!\n", retval); + + return retval; +} + +/** + * pm_runtime_resume - Carry out run-time resume of given device. + * @dev: Device to suspend. + */ +int pm_runtime_resume(struct device *dev) +{ + int retval; + + spin_lock_irq(&dev->power.lock); + retval = __pm_runtime_resume(dev, false); + spin_unlock_irq(&dev->power.lock); + + return retval; +} +EXPORT_SYMBOL_GPL(pm_runtime_resume); + +/** + * pm_runtime_work - Universal run-time PM work function. + * @work: Work structure used for scheduling the execution of this function. + * + * Use @work to get the device object the work is to be done for, determine what + * is to be done and execute the appropriate run-time PM function. + */ +static void pm_runtime_work(struct work_struct *work) +{ + struct device *dev = container_of(work, struct device, power.work); + enum rpm_request req; + + spin_lock_irq(&dev->power.lock); + + if (!dev->power.request_pending) + goto out; + + req = dev->power.request; + dev->power.request = RPM_REQ_NONE; + dev->power.request_pending = false; + + switch (req) { + case RPM_REQ_NONE: + break; + case RPM_REQ_IDLE: + __pm_runtime_idle(dev); + break; + case RPM_REQ_SUSPEND: + __pm_runtime_suspend(dev, true); + break; + case RPM_REQ_RESUME: + __pm_runtime_resume(dev, true); + break; + } + + out: + spin_unlock_irq(&dev->power.lock); +} + +/** + * __pm_request_idle - Submit an idle notification request for given device. + * @dev: Device to handle. + * + * Check if the device's run-time PM status is correct for suspending the device + * and queue up a request to run __pm_runtime_idle() for it. + * + * This function must be called under dev->power.lock with interrupts disabled. + */ +static int __pm_request_idle(struct device *dev) +{ + int retval = 0; + + if (dev->power.runtime_error) + retval = -EINVAL; + else if (atomic_read(&dev->power.usage_count) > 0 + || dev->power.disable_depth > 0 + || dev->power.runtime_status == RPM_SUSPENDED + || dev->power.runtime_status == RPM_SUSPENDING) + retval = -EAGAIN; + else if (!pm_children_suspended(dev)) + retval = -EBUSY; + if (retval) + return retval; + + if (dev->power.request_pending) { + /* Any requests other then RPM_REQ_IDLE take precedence. */ + if (dev->power.request == RPM_REQ_NONE) + dev->power.request = RPM_REQ_IDLE; + else if (dev->power.request != RPM_REQ_IDLE) + retval = -EAGAIN; + return retval; + } + + dev->power.request = RPM_REQ_IDLE; + dev->power.request_pending = true; + queue_work(pm_wq, &dev->power.work); + + return retval; +} + +/** + * pm_request_idle - Submit an idle notification request for given device. + * @dev: Device to handle. + */ +int pm_request_idle(struct device *dev) +{ + unsigned long flags; + int retval; + + spin_lock_irqsave(&dev->power.lock, flags); + retval = __pm_request_idle(dev); + spin_unlock_irqrestore(&dev->power.lock, flags); + + return retval; +} +EXPORT_SYMBOL_GPL(pm_request_idle); + +/** + * __pm_request_suspend - Submit a suspend request for given device. + * @dev: Device to suspend. + * + * This function must be called under dev->power.lock with interrupts disabled. + */ +static int __pm_request_suspend(struct device *dev) +{ + int retval = 0; + + if (dev->power.runtime_error) + return -EINVAL; + + if (dev->power.runtime_status == RPM_SUSPENDED) + retval = 1; + else if (atomic_read(&dev->power.usage_count) > 0 + || dev->power.disable_depth > 0) + retval = -EAGAIN; + else if (dev->power.runtime_status == RPM_SUSPENDING) + retval = -EINPROGRESS; + else if (!pm_children_suspended(dev)) + retval = -EBUSY; + if (retval < 0) + return retval; + + pm_runtime_deactivate_timer(dev); + + if (dev->power.request_pending) { + /* + * Pending resume requests take precedence over us, but we can + * overtake any other pending request. + */ + if (dev->power.request == RPM_REQ_RESUME) + retval = -EAGAIN; + else if (dev->power.request != RPM_REQ_SUSPEND) + dev->power.request = retval ? + RPM_REQ_NONE : RPM_REQ_SUSPEND; + return retval; + } else if (retval) { + return retval; + } + + dev->power.request = RPM_REQ_SUSPEND; + dev->power.request_pending = true; + queue_work(pm_wq, &dev->power.work); + + return 0; +} + +/** + * pm_suspend_timer_fn - Timer function for pm_schedule_suspend(). + * @data: Device pointer passed by pm_schedule_suspend(). + * + * Check if the time is right and execute __pm_request_suspend() in that case. + */ +static void pm_suspend_timer_fn(unsigned long data) +{ + struct device *dev = (struct device *)data; + unsigned long flags; + unsigned long expires; + + spin_lock_irqsave(&dev->power.lock, flags); + + expires = dev->power.timer_expires; + /* If 'expire' is after 'jiffies' we've been called too early. */ + if (expires > 0 && !time_after(expires, jiffies)) { + dev->power.timer_expires = 0; + __pm_request_suspend(dev); + } + + spin_unlock_irqrestore(&dev->power.lock, flags); +} + +/** + * pm_schedule_suspend - Set up a timer to submit a suspend request in future. + * @dev: Device to suspend. + * @delay: Time to wait before submitting a suspend request, in milliseconds. + */ +int pm_schedule_suspend(struct device *dev, unsigned int delay) +{ + unsigned long flags; + int retval = 0; + + spin_lock_irqsave(&dev->power.lock, flags); + + if (dev->power.runtime_error) { + retval = -EINVAL; + goto out; + } + + if (!delay) { + retval = __pm_request_suspend(dev); + goto out; + } + + pm_runtime_deactivate_timer(dev); + + if (dev->power.request_pending) { + /* + * Pending resume requests take precedence over us, but any + * other pending requests have to be canceled. + */ + if (dev->power.request == RPM_REQ_RESUME) { + retval = -EAGAIN; + goto out; + } + dev->power.request = RPM_REQ_NONE; + } + + if (dev->power.runtime_status == RPM_SUSPENDED) + retval = 1; + else if (dev->power.runtime_status == RPM_SUSPENDING) + retval = -EINPROGRESS; + else if (atomic_read(&dev->power.usage_count) > 0 + || dev->power.disable_depth > 0) + retval = -EAGAIN; + else if (!pm_children_suspended(dev)) + retval = -EBUSY; + if (retval) + goto out; + + dev->power.timer_expires = jiffies + msecs_to_jiffies(delay); + mod_timer(&dev->power.suspend_timer, dev->power.timer_expires); + + out: + spin_unlock_irqrestore(&dev->power.lock, flags); + + return retval; +} +EXPORT_SYMBOL_GPL(pm_schedule_suspend); + +/** + * pm_request_resume - Submit a resume request for given device. + * @dev: Device to resume. + * + * This function must be called under dev->power.lock with interrupts disabled. + */ +static int __pm_request_resume(struct device *dev) +{ + int retval = 0; + + if (dev->power.runtime_error) + return -EINVAL; + + if (dev->power.runtime_status == RPM_ACTIVE) + retval = 1; + else if (dev->power.runtime_status == RPM_RESUMING) + retval = -EINPROGRESS; + else if (dev->power.disable_depth > 0) + retval = -EAGAIN; + if (retval < 0) + return retval; + + pm_runtime_deactivate_timer(dev); + + if (dev->power.request_pending) { + /* If non-resume request is pending, we can overtake it. */ + dev->power.request = retval ? RPM_REQ_NONE : RPM_REQ_RESUME; + return retval; + } else if (retval) { + return retval; + } + + dev->power.request = RPM_REQ_RESUME; + dev->power.request_pending = true; + queue_work(pm_wq, &dev->power.work); + + return retval; +} + +/** + * pm_request_resume - Submit a resume request for given device. + * @dev: Device to resume. + */ +int pm_request_resume(struct device *dev) +{ + unsigned long flags; + int retval; + + spin_lock_irqsave(&dev->power.lock, flags); + retval = __pm_request_resume(dev); + spin_unlock_irqrestore(&dev->power.lock, flags); + + return retval; +} +EXPORT_SYMBOL_GPL(pm_request_resume); + +/** + * __pm_runtime_get - Reference count a device and wake it up, if necessary. + * @dev: Device to handle. + * @sync: If set and the device is suspended, resume it synchronously. + * + * Increment the usage count of the device and if it was zero previously, + * resume it or submit a resume request for it, depending on the value of @sync. + */ +int __pm_runtime_get(struct device *dev, bool sync) +{ + int retval = 1; + + if (atomic_add_return(1, &dev->power.usage_count) == 1) + retval = sync ? pm_runtime_resume(dev) : pm_request_resume(dev); + + return retval; +} +EXPORT_SYMBOL_GPL(__pm_runtime_get); + +/** + * __pm_runtime_put - Decrement the device's usage counter and notify its bus. + * @dev: Device to handle. + * @sync: If the device's bus type is to be notified, do that synchronously. + * + * Decrement the usage count of the device and if it reaches zero, carry out a + * synchronous idle notification or submit an idle notification request for it, + * depending on the value of @sync. + */ +int __pm_runtime_put(struct device *dev, bool sync) +{ + int retval = 0; + + if (atomic_dec_and_test(&dev->power.usage_count)) + retval = sync ? pm_runtime_idle(dev) : pm_request_idle(dev); + + return retval; +} +EXPORT_SYMBOL_GPL(__pm_runtime_put); + +/** + * __pm_runtime_set_status - Set run-time PM status of a device. + * @dev: Device to handle. + * @status: New run-time PM status of the device. + * + * If run-time PM of the device is disabled or its power.runtime_error field is + * different from zero, the status may be changed either to RPM_ACTIVE, or to + * RPM_SUSPENDED, as long as that reflects the actual state of the device. + * However, if the device has a parent and the parent is not active, and the + * parent's power.ignore_children flag is unset, the device's status cannot be + * set to RPM_ACTIVE, so -EBUSY is returned in that case. + * + * If successful, __pm_runtime_set_status() clears the power.runtime_error field + * and the device parent's counter of unsuspended children is modified to + * reflect the new status. If the new status is RPM_SUSPENDED, an idle + * notification request for the parent is submitted. + */ +int __pm_runtime_set_status(struct device *dev, unsigned int status) +{ + struct device *parent = dev->parent; + unsigned long flags; + bool notify_parent = false; + int error = 0; + + if (status != RPM_ACTIVE && status != RPM_SUSPENDED) + return -EINVAL; + + spin_lock_irqsave(&dev->power.lock, flags); + + if (!dev->power.runtime_error && !dev->power.disable_depth) { + error = -EAGAIN; + goto out; + } + + if (dev->power.runtime_status == status) + goto out_set; + + if (status == RPM_SUSPENDED) { + /* It always is possible to set the status to 'suspended'. */ + if (parent) { + atomic_add_unless(&parent->power.child_count, -1, 0); + notify_parent = !parent->power.ignore_children; + } + goto out_set; + } + + if (parent) { + spin_lock_irq(&parent->power.lock); + + /* + * It is invalid to put an active child under a parent that is + * not active, has run-time PM enabled and the + * 'power.ignore_children' flag unset. + */ + if (!parent->power.disable_depth + && !parent->power.ignore_children + && parent->power.runtime_status != RPM_ACTIVE) { + error = -EBUSY; + } else { + if (dev->power.runtime_status == RPM_SUSPENDED) + atomic_inc(&parent->power.child_count); + } + + spin_unlock_irq(&parent->power.lock); + + if (error) + goto out; + } + + out_set: + dev->power.runtime_status = status; + dev->power.runtime_error = 0; + out: + spin_unlock_irqrestore(&dev->power.lock, flags); + + if (notify_parent) + pm_request_idle(parent); + + return error; +} +EXPORT_SYMBOL_GPL(__pm_runtime_set_status); + +/** + * __pm_runtime_barrier - Cancel pending requests and wait for completions. + * @dev: Device to handle. + * + * Flush all pending requests for the device from pm_wq and wait for all + * run-time PM operations involving the device in progress to complete. + * + * Should be called under dev->power.lock with interrupts disabled. + */ +static void __pm_runtime_barrier(struct device *dev) +{ + pm_runtime_deactivate_timer(dev); + + if (dev->power.request_pending) { + dev->power.request = RPM_REQ_NONE; + spin_unlock_irq(&dev->power.lock); + + cancel_work_sync(&dev->power.work); + + spin_lock_irq(&dev->power.lock); + dev->power.request_pending = false; + } + + if (dev->power.runtime_status == RPM_SUSPENDING + || dev->power.runtime_status == RPM_RESUMING + || dev->power.idle_notification) { + DEFINE_WAIT(wait); + + /* Suspend, wake-up or idle notification in progress. */ + for (;;) { + prepare_to_wait(&dev->power.wait_queue, &wait, + TASK_UNINTERRUPTIBLE); + if (dev->power.runtime_status != RPM_SUSPENDING + && dev->power.runtime_status != RPM_RESUMING + && !dev->power.idle_notification) + break; + spin_unlock_irq(&dev->power.lock); + + schedule(); + + spin_lock_irq(&dev->power.lock); + } + finish_wait(&dev->power.wait_queue, &wait); + } +} + +/** + * pm_runtime_barrier - Flush pending requests and wait for completions. + * @dev: Device to handle. + * + * Prevent the device from being suspended by incrementing its usage counter and + * if there's a pending resume request for the device, wake the device up. + * Next, make sure that all pending requests for the device have been flushed + * from pm_wq and wait for all run-time PM operations involving the device in + * progress to complete. + * + * Return value: + * 1, if there was a resume request pending and the device had to be woken up, + * 0, otherwise + */ +int pm_runtime_barrier(struct device *dev) +{ + int retval = 0; + + pm_runtime_get_noresume(dev); + spin_lock_irq(&dev->power.lock); + + if (dev->power.request_pending + && dev->power.request == RPM_REQ_RESUME) { + __pm_runtime_resume(dev, false); + retval = 1; + } + + __pm_runtime_barrier(dev); + + spin_unlock_irq(&dev->power.lock); + pm_runtime_put_noidle(dev); + + return retval; +} +EXPORT_SYMBOL_GPL(pm_runtime_barrier); + +/** + * __pm_runtime_disable - Disable run-time PM of a device. + * @dev: Device to handle. + * @check_resume: If set, check if there's a resume request for the device. + * + * Increment power.disable_depth for the device and if was zero previously, + * cancel all pending run-time PM requests for the device and wait for all + * operations in progress to complete. The device can be either active or + * suspended after its run-time PM has been disabled. + * + * If @check_resume is set and there's a resume request pending when + * __pm_runtime_disable() is called and power.disable_depth is zero, the + * function will wake up the device before disabling its run-time PM. + */ +void __pm_runtime_disable(struct device *dev, bool check_resume) +{ + spin_lock_irq(&dev->power.lock); + + if (dev->power.disable_depth > 0) { + dev->power.disable_depth++; + goto out; + } + + /* + * Wake up the device if there's a resume request pending, because that + * means there probably is some I/O to process and disabling run-time PM + * shouldn't prevent the device from processing the I/O. + */ + if (check_resume && dev->power.request_pending + && dev->power.request == RPM_REQ_RESUME) { + /* + * Prevent suspends and idle notifications from being carried + * out after we have woken up the device. + */ + pm_runtime_get_noresume(dev); + + __pm_runtime_resume(dev, false); + + pm_runtime_put_noidle(dev); + } + + if (!dev->power.disable_depth++) + __pm_runtime_barrier(dev); + + out: + spin_unlock_irq(&dev->power.lock); +} +EXPORT_SYMBOL_GPL(__pm_runtime_disable); + +/** + * pm_runtime_enable - Enable run-time PM of a device. + * @dev: Device to handle. + */ +void pm_runtime_enable(struct device *dev) +{ + unsigned long flags; + + spin_lock_irqsave(&dev->power.lock, flags); + + if (dev->power.disable_depth > 0) + dev->power.disable_depth--; + else + dev_warn(dev, "Unbalanced %s!\n", __func__); + + spin_unlock_irqrestore(&dev->power.lock, flags); +} +EXPORT_SYMBOL_GPL(pm_runtime_enable); + +/** + * pm_runtime_init - Initialize run-time PM fields in given device object. + * @dev: Device object to initialize. + */ +void pm_runtime_init(struct device *dev) +{ + spin_lock_init(&dev->power.lock); + + dev->power.runtime_status = RPM_SUSPENDED; + dev->power.idle_notification = false; + + dev->power.disable_depth = 1; + atomic_set(&dev->power.usage_count, 0); + + dev->power.runtime_error = 0; + + atomic_set(&dev->power.child_count, 0); + pm_suspend_ignore_children(dev, false); + + dev->power.request_pending = false; + dev->power.request = RPM_REQ_NONE; + dev->power.deferred_resume = false; + INIT_WORK(&dev->power.work, pm_runtime_work); + + dev->power.timer_expires = 0; + setup_timer(&dev->power.suspend_timer, pm_suspend_timer_fn, + (unsigned long)dev); + + init_waitqueue_head(&dev->power.wait_queue); +} + +/** + * pm_runtime_remove - Prepare for removing a device from device hierarchy. + * @dev: Device object being removed from device hierarchy. + */ +void pm_runtime_remove(struct device *dev) +{ + __pm_runtime_disable(dev, false); + + /* Change the status back to 'suspended' to match the initial status. */ + if (dev->power.runtime_status == RPM_ACTIVE) + pm_runtime_set_suspended(dev); +} diff --git a/include/linux/pm.h b/include/linux/pm.h index b3f74764a586..2b6e20df0e52 100644 --- a/include/linux/pm.h +++ b/include/linux/pm.h @@ -22,6 +22,10 @@ #define _LINUX_PM_H #include +#include +#include +#include +#include /* * Callbacks for platform drivers to implement. @@ -165,6 +169,28 @@ typedef struct pm_message { * It is allowed to unregister devices while the above callbacks are being * executed. However, it is not allowed to unregister a device from within any * of its own callbacks. + * + * There also are the following callbacks related to run-time power management + * of devices: + * + * @runtime_suspend: Prepare the device for a condition in which it won't be + * able to communicate with the CPU(s) and RAM due to power management. + * This need not mean that the device should be put into a low power state. + * For example, if the device is behind a link which is about to be turned + * off, the device may remain at full power. If the device does go to low + * power and if device_may_wakeup(dev) is true, remote wake-up (i.e., a + * hardware mechanism allowing the device to request a change of its power + * state, such as PCI PME) should be enabled for it. + * + * @runtime_resume: Put the device into the fully active state in response to a + * wake-up event generated by hardware or at the request of software. If + * necessary, put the device into the full power state and restore its + * registers, so that it is fully operational. + * + * @runtime_idle: Device appears to be inactive and it might be put into a low + * power state if all of the necessary conditions are satisfied. Check + * these conditions and handle the device as appropriate, possibly queueing + * a suspend request for it. The return value is ignored by the PM core. */ struct dev_pm_ops { @@ -182,6 +208,9 @@ struct dev_pm_ops { int (*thaw_noirq)(struct device *dev); int (*poweroff_noirq)(struct device *dev); int (*restore_noirq)(struct device *dev); + int (*runtime_suspend)(struct device *dev); + int (*runtime_resume)(struct device *dev); + int (*runtime_idle)(struct device *dev); }; /** @@ -315,14 +344,80 @@ enum dpm_state { DPM_OFF_IRQ, }; +/** + * Device run-time power management status. + * + * These status labels are used internally by the PM core to indicate the + * current status of a device with respect to the PM core operations. They do + * not reflect the actual power state of the device or its status as seen by the + * driver. + * + * RPM_ACTIVE Device is fully operational. Indicates that the device + * bus type's ->runtime_resume() callback has completed + * successfully. + * + * RPM_SUSPENDED Device bus type's ->runtime_suspend() callback has + * completed successfully. The device is regarded as + * suspended. + * + * RPM_RESUMING Device bus type's ->runtime_resume() callback is being + * executed. + * + * RPM_SUSPENDING Device bus type's ->runtime_suspend() callback is being + * executed. + */ + +enum rpm_status { + RPM_ACTIVE = 0, + RPM_RESUMING, + RPM_SUSPENDED, + RPM_SUSPENDING, +}; + +/** + * Device run-time power management request types. + * + * RPM_REQ_NONE Do nothing. + * + * RPM_REQ_IDLE Run the device bus type's ->runtime_idle() callback + * + * RPM_REQ_SUSPEND Run the device bus type's ->runtime_suspend() callback + * + * RPM_REQ_RESUME Run the device bus type's ->runtime_resume() callback + */ + +enum rpm_request { + RPM_REQ_NONE = 0, + RPM_REQ_IDLE, + RPM_REQ_SUSPEND, + RPM_REQ_RESUME, +}; + struct dev_pm_info { pm_message_t power_state; - unsigned can_wakeup:1; - unsigned should_wakeup:1; + unsigned int can_wakeup:1; + unsigned int should_wakeup:1; enum dpm_state status; /* Owned by the PM core */ -#ifdef CONFIG_PM_SLEEP +#ifdef CONFIG_PM_SLEEP struct list_head entry; #endif +#ifdef CONFIG_PM_RUNTIME + struct timer_list suspend_timer; + unsigned long timer_expires; + struct work_struct work; + wait_queue_head_t wait_queue; + spinlock_t lock; + atomic_t usage_count; + atomic_t child_count; + unsigned int disable_depth:3; + unsigned int ignore_children:1; + unsigned int idle_notification:1; + unsigned int request_pending:1; + unsigned int deferred_resume:1; + enum rpm_request request; + enum rpm_status runtime_status; + int runtime_error; +#endif }; /* diff --git a/include/linux/pm_runtime.h b/include/linux/pm_runtime.h new file mode 100644 index 000000000000..44087044910f --- /dev/null +++ b/include/linux/pm_runtime.h @@ -0,0 +1,114 @@ +/* + * pm_runtime.h - Device run-time power management helper functions. + * + * Copyright (C) 2009 Rafael J. Wysocki + * + * This file is released under the GPLv2. + */ + +#ifndef _LINUX_PM_RUNTIME_H +#define _LINUX_PM_RUNTIME_H + +#include +#include + +#ifdef CONFIG_PM_RUNTIME + +extern struct workqueue_struct *pm_wq; + +extern int pm_runtime_idle(struct device *dev); +extern int pm_runtime_suspend(struct device *dev); +extern int pm_runtime_resume(struct device *dev); +extern int pm_request_idle(struct device *dev); +extern int pm_schedule_suspend(struct device *dev, unsigned int delay); +extern int pm_request_resume(struct device *dev); +extern int __pm_runtime_get(struct device *dev, bool sync); +extern int __pm_runtime_put(struct device *dev, bool sync); +extern int __pm_runtime_set_status(struct device *dev, unsigned int status); +extern int pm_runtime_barrier(struct device *dev); +extern void pm_runtime_enable(struct device *dev); +extern void __pm_runtime_disable(struct device *dev, bool check_resume); + +static inline bool pm_children_suspended(struct device *dev) +{ + return dev->power.ignore_children + || !atomic_read(&dev->power.child_count); +} + +static inline void pm_suspend_ignore_children(struct device *dev, bool enable) +{ + dev->power.ignore_children = enable; +} + +static inline void pm_runtime_get_noresume(struct device *dev) +{ + atomic_inc(&dev->power.usage_count); +} + +static inline void pm_runtime_put_noidle(struct device *dev) +{ + atomic_add_unless(&dev->power.usage_count, -1, 0); +} + +#else /* !CONFIG_PM_RUNTIME */ + +static inline int pm_runtime_idle(struct device *dev) { return -ENOSYS; } +static inline int pm_runtime_suspend(struct device *dev) { return -ENOSYS; } +static inline int pm_runtime_resume(struct device *dev) { return 0; } +static inline int pm_request_idle(struct device *dev) { return -ENOSYS; } +static inline int pm_schedule_suspend(struct device *dev, unsigned int delay) +{ + return -ENOSYS; +} +static inline int pm_request_resume(struct device *dev) { return 0; } +static inline int __pm_runtime_get(struct device *dev, bool sync) { return 1; } +static inline int __pm_runtime_put(struct device *dev, bool sync) { return 0; } +static inline int __pm_runtime_set_status(struct device *dev, + unsigned int status) { return 0; } +static inline int pm_runtime_barrier(struct device *dev) { return 0; } +static inline void pm_runtime_enable(struct device *dev) {} +static inline void __pm_runtime_disable(struct device *dev, bool c) {} + +static inline bool pm_children_suspended(struct device *dev) { return false; } +static inline void pm_suspend_ignore_children(struct device *dev, bool en) {} +static inline void pm_runtime_get_noresume(struct device *dev) {} +static inline void pm_runtime_put_noidle(struct device *dev) {} + +#endif /* !CONFIG_PM_RUNTIME */ + +static inline int pm_runtime_get(struct device *dev) +{ + return __pm_runtime_get(dev, false); +} + +static inline int pm_runtime_get_sync(struct device *dev) +{ + return __pm_runtime_get(dev, true); +} + +static inline int pm_runtime_put(struct device *dev) +{ + return __pm_runtime_put(dev, false); +} + +static inline int pm_runtime_put_sync(struct device *dev) +{ + return __pm_runtime_put(dev, true); +} + +static inline int pm_runtime_set_active(struct device *dev) +{ + return __pm_runtime_set_status(dev, RPM_ACTIVE); +} + +static inline void pm_runtime_set_suspended(struct device *dev) +{ + __pm_runtime_set_status(dev, RPM_SUSPENDED); +} + +static inline void pm_runtime_disable(struct device *dev) +{ + __pm_runtime_disable(dev, true); +} + +#endif diff --git a/kernel/power/Kconfig b/kernel/power/Kconfig index 72067cbdb37f..91e09d3b2eb2 100644 --- a/kernel/power/Kconfig +++ b/kernel/power/Kconfig @@ -208,3 +208,17 @@ config APM_EMULATION random kernel OOPSes or reboots that don't seem to be related to anything, try disabling/enabling this option (or disabling/enabling APM in your BIOS). + +config PM_RUNTIME + bool "Run-time PM core functionality" + depends on PM + ---help--- + Enable functionality allowing I/O devices to be put into energy-saving + (low power) states at run time (or autosuspended) after a specified + period of inactivity and woken up in response to a hardware-generated + wake-up event or a driver's request. + + Hardware support is generally required for this functionality to work + and the bus type drivers of the buses the devices are on are + responsible for the actual handling of the autosuspend requests and + wake-up events. diff --git a/kernel/power/main.c b/kernel/power/main.c index f710e36930cc..347d2cc88cd0 100644 --- a/kernel/power/main.c +++ b/kernel/power/main.c @@ -11,6 +11,7 @@ #include #include #include +#include #include "power.h" @@ -217,8 +218,24 @@ static struct attribute_group attr_group = { .attrs = g, }; +#ifdef CONFIG_PM_RUNTIME +struct workqueue_struct *pm_wq; + +static int __init pm_start_workqueue(void) +{ + pm_wq = create_freezeable_workqueue("pm"); + + return pm_wq ? 0 : -ENOMEM; +} +#else +static inline int pm_start_workqueue(void) { return 0; } +#endif + static int __init pm_init(void) { + int error = pm_start_workqueue(); + if (error) + return error; power_kobj = kobject_create_and_add("power", NULL); if (!power_kobj) return -ENOMEM; -- cgit v1.2.3 From 9e726b17422bade75fba94e625cd35fd1353e682 Mon Sep 17 00:00:00 2001 From: Luiz Augusto von Dentz Date: Wed, 15 Jul 2009 13:50:58 -0300 Subject: Bluetooth: Fix rejected connection not disconnecting ACL link When using DEFER_SETUP on a RFCOMM socket, a SABM frame triggers authorization which when rejected send a DM response. This is fine according to the RFCOMM spec: the responding implementation may replace the "proper" response on the Multiplexer Control channel with a DM frame, sent on the referenced DLCI to indicate that the DLCI is not open, and that the responder would not grant a request to open it later either. But some stacks doesn't seems to cope with this leaving DLCI 0 open after receiving DM frame. To fix it properly a timer was introduced to rfcomm_session which is used to set a timeout when the last active DLC of a session is unlinked, this will give the remote stack some time to reply with a proper DISC frame on DLCI 0 avoiding both sides sending DISC to each other on stacks that follow the specification and taking care of those who don't by taking down DLCI 0. Signed-off-by: Luiz Augusto von Dentz Signed-off-by: Marcel Holtmann --- include/net/bluetooth/rfcomm.h | 2 ++ net/bluetooth/rfcomm/core.c | 41 +++++++++++++++++++++++++++++++++++++++++ 2 files changed, 43 insertions(+) (limited to 'include') diff --git a/include/net/bluetooth/rfcomm.h b/include/net/bluetooth/rfcomm.h index c274993234e3..921d7b3c7f8d 100644 --- a/include/net/bluetooth/rfcomm.h +++ b/include/net/bluetooth/rfcomm.h @@ -29,6 +29,7 @@ #define RFCOMM_CONN_TIMEOUT (HZ * 30) #define RFCOMM_DISC_TIMEOUT (HZ * 20) #define RFCOMM_AUTH_TIMEOUT (HZ * 25) +#define RFCOMM_IDLE_TIMEOUT (HZ * 2) #define RFCOMM_DEFAULT_MTU 127 #define RFCOMM_DEFAULT_CREDITS 7 @@ -154,6 +155,7 @@ struct rfcomm_msc { struct rfcomm_session { struct list_head list; struct socket *sock; + struct timer_list timer; unsigned long state; unsigned long flags; atomic_t refcnt; diff --git a/net/bluetooth/rfcomm/core.c b/net/bluetooth/rfcomm/core.c index 26af4854af64..25692bc0a342 100644 --- a/net/bluetooth/rfcomm/core.c +++ b/net/bluetooth/rfcomm/core.c @@ -244,6 +244,33 @@ static inline int rfcomm_check_security(struct rfcomm_dlc *d) auth_type); } +static void rfcomm_session_timeout(unsigned long arg) +{ + struct rfcomm_session *s = (void *) arg; + + BT_DBG("session %p state %ld", s, s->state); + + set_bit(RFCOMM_TIMED_OUT, &s->flags); + rfcomm_session_put(s); + rfcomm_schedule(RFCOMM_SCHED_TIMEO); +} + +static void rfcomm_session_set_timer(struct rfcomm_session *s, long timeout) +{ + BT_DBG("session %p state %ld timeout %ld", s, s->state, timeout); + + if (!mod_timer(&s->timer, jiffies + timeout)) + rfcomm_session_hold(s); +} + +static void rfcomm_session_clear_timer(struct rfcomm_session *s) +{ + BT_DBG("session %p state %ld", s, s->state); + + if (timer_pending(&s->timer) && del_timer(&s->timer)) + rfcomm_session_put(s); +} + /* ---- RFCOMM DLCs ---- */ static void rfcomm_dlc_timeout(unsigned long arg) { @@ -320,6 +347,7 @@ static void rfcomm_dlc_link(struct rfcomm_session *s, struct rfcomm_dlc *d) rfcomm_session_hold(s); + rfcomm_session_clear_timer(s); rfcomm_dlc_hold(d); list_add(&d->list, &s->dlcs); d->session = s; @@ -335,6 +363,9 @@ static void rfcomm_dlc_unlink(struct rfcomm_dlc *d) d->session = NULL; rfcomm_dlc_put(d); + if (list_empty(&s->dlcs)) + rfcomm_session_set_timer(s, RFCOMM_IDLE_TIMEOUT); + rfcomm_session_put(s); } @@ -567,6 +598,8 @@ static struct rfcomm_session *rfcomm_session_add(struct socket *sock, int state) BT_DBG("session %p sock %p", s, sock); + setup_timer(&s->timer, rfcomm_session_timeout, (unsigned long) s); + INIT_LIST_HEAD(&s->dlcs); s->state = state; s->sock = sock; @@ -598,6 +631,7 @@ static void rfcomm_session_del(struct rfcomm_session *s) if (state == BT_CONNECTED) rfcomm_send_disc(s, 0); + rfcomm_session_clear_timer(s); sock_release(s->sock); kfree(s); @@ -639,6 +673,7 @@ static void rfcomm_session_close(struct rfcomm_session *s, int err) __rfcomm_dlc_close(d, err); } + rfcomm_session_clear_timer(s); rfcomm_session_put(s); } @@ -1879,6 +1914,12 @@ static inline void rfcomm_process_sessions(void) struct rfcomm_session *s; s = list_entry(p, struct rfcomm_session, list); + if (test_and_clear_bit(RFCOMM_TIMED_OUT, &s->flags)) { + s->state = BT_DISCONN; + rfcomm_send_disc(s, 0); + continue; + } + if (s->state == BT_LISTEN) { rfcomm_accept_connection(s); continue; -- cgit v1.2.3 From 6c10db72c94818573552fd71c89540da325efdfb Mon Sep 17 00:00:00 2001 From: Chandra Seetharaman Date: Fri, 26 Jun 2009 19:30:06 -0700 Subject: [SCSI] scsi_dh: Reference count scsi_dh_attach Problem reported: http://marc.info/?l=dm-devel&m=124585978305866&w=2 scsi_dh does not do a refernce count for attach/detach, and this affects the way it is supposed to work with multipath when a device is not in the dev_list of the hardware handler. This patch adds a reference count that counts each attach. Signed-off-by: Chandra Seetharaman Signed-off-by: James Bottomley Signed-off-by: James Bottomley --- drivers/scsi/device_handler/scsi_dh.c | 23 ++++++++++++++++------- include/scsi/scsi_device.h | 2 ++ 2 files changed, 18 insertions(+), 7 deletions(-) (limited to 'include') diff --git a/drivers/scsi/device_handler/scsi_dh.c b/drivers/scsi/device_handler/scsi_dh.c index a518f2eff19a..53a7385e1b4d 100644 --- a/drivers/scsi/device_handler/scsi_dh.c +++ b/drivers/scsi/device_handler/scsi_dh.c @@ -153,12 +153,24 @@ static int scsi_dh_handler_attach(struct scsi_device *sdev, if (sdev->scsi_dh_data) { if (sdev->scsi_dh_data->scsi_dh != scsi_dh) err = -EBUSY; - } else if (scsi_dh->attach) + else + kref_get(&sdev->scsi_dh_data->kref); + } else if (scsi_dh->attach) { err = scsi_dh->attach(sdev); - + if (!err) { + kref_init(&sdev->scsi_dh_data->kref); + sdev->scsi_dh_data->sdev = sdev; + } + } return err; } +static void __detach_handler (struct kref *kref) +{ + struct scsi_dh_data *scsi_dh_data = container_of(kref, struct scsi_dh_data, kref); + scsi_dh_data->scsi_dh->detach(scsi_dh_data->sdev); +} + /* * scsi_dh_handler_detach - Detach a device handler from a device * @sdev - SCSI device the device handler should be detached from @@ -180,7 +192,7 @@ static void scsi_dh_handler_detach(struct scsi_device *sdev, scsi_dh = sdev->scsi_dh_data->scsi_dh; if (scsi_dh && scsi_dh->detach) - scsi_dh->detach(sdev); + kref_put(&sdev->scsi_dh_data->kref, __detach_handler); } /* @@ -474,7 +486,6 @@ int scsi_dh_attach(struct request_queue *q, const char *name) if (!err) { err = scsi_dh_handler_attach(sdev, scsi_dh); - put_device(&sdev->sdev_gendev); } return err; @@ -505,10 +516,8 @@ void scsi_dh_detach(struct request_queue *q) return; if (sdev->scsi_dh_data) { - /* if sdev is not on internal list, detach */ scsi_dh = sdev->scsi_dh_data->scsi_dh; - if (!device_handler_match(scsi_dh, sdev)) - scsi_dh_handler_detach(sdev, scsi_dh); + scsi_dh_handler_detach(sdev, scsi_dh); } put_device(&sdev->sdev_gendev); } diff --git a/include/scsi/scsi_device.h b/include/scsi/scsi_device.h index 3f566af3f101..1f3a4c8044c0 100644 --- a/include/scsi/scsi_device.h +++ b/include/scsi/scsi_device.h @@ -191,6 +191,8 @@ struct scsi_device_handler { struct scsi_dh_data { struct scsi_device_handler *scsi_dh; + struct scsi_device *sdev; + struct kref kref; char buf[0]; }; -- cgit v1.2.3 From beb29a6d421f6dbd41d68d0621c1b28ad1d4a9f4 Mon Sep 17 00:00:00 2001 From: Joe Eykholt Date: Wed, 29 Jul 2009 17:04:06 -0700 Subject: [SCSI] libfc: remove extra semicolons from debug macros This is unlikely to cause any problems, but the libfc debug macros introduce extra undesirable semicolons. Signed-off-by: Joe Eykholt Signed-off-by: Robert Love Signed-off-by: James Bottomley Signed-off-by: James Bottomley --- include/scsi/libfc.h | 20 ++++++++++---------- 1 file changed, 10 insertions(+), 10 deletions(-) (limited to 'include') diff --git a/include/scsi/libfc.h b/include/scsi/libfc.h index b92584a8843a..ef04a2c52b8c 100644 --- a/include/scsi/libfc.h +++ b/include/scsi/libfc.h @@ -51,22 +51,22 @@ do { \ do { \ CMD; \ } while (0); \ -} while (0); +} while (0) #define FC_LIBFC_DBG(fmt, args...) \ FC_CHECK_LOGGING(FC_LIBFC_LOGGING, \ - printk(KERN_INFO "libfc: " fmt, ##args);) + printk(KERN_INFO "libfc: " fmt, ##args)) #define FC_LPORT_DBG(lport, fmt, args...) \ FC_CHECK_LOGGING(FC_LPORT_LOGGING, \ printk(KERN_INFO "lport: %6x: " fmt, \ - fc_host_port_id(lport->host), ##args);) + fc_host_port_id(lport->host), ##args)) #define FC_DISC_DBG(disc, fmt, args...) \ FC_CHECK_LOGGING(FC_DISC_LOGGING, \ printk(KERN_INFO "disc: %6x: " fmt, \ fc_host_port_id(disc->lport->host), \ - ##args);) + ##args)) #define FC_RPORT_DBG(rport, fmt, args...) \ do { \ @@ -75,31 +75,31 @@ do { \ FC_CHECK_LOGGING(FC_RPORT_LOGGING, \ printk(KERN_INFO "rport: %6x: %6x: " fmt, \ fc_host_port_id(lport->host), \ - rport->port_id, ##args);) \ -} while (0); + rport->port_id, ##args)); \ +} while (0) #define FC_FCP_DBG(pkt, fmt, args...) \ FC_CHECK_LOGGING(FC_FCP_LOGGING, \ printk(KERN_INFO "fcp: %6x: %6x: " fmt, \ fc_host_port_id(pkt->lp->host), \ - pkt->rport->port_id, ##args);) + pkt->rport->port_id, ##args)) #define FC_EM_DBG(em, fmt, args...) \ FC_CHECK_LOGGING(FC_EM_LOGGING, \ printk(KERN_INFO "em: %6x: " fmt, \ fc_host_port_id(em->lp->host), \ - ##args);) + ##args)) #define FC_EXCH_DBG(exch, fmt, args...) \ FC_CHECK_LOGGING(FC_EXCH_LOGGING, \ printk(KERN_INFO "exch: %6x: %4x: " fmt, \ fc_host_port_id(exch->lp->host), \ - exch->xid, ##args);) + exch->xid, ##args)) #define FC_SCSI_DBG(lport, fmt, args...) \ FC_CHECK_LOGGING(FC_SCSI_LOGGING, \ printk(KERN_INFO "scsi: %6x: " fmt, \ - fc_host_port_id(lport->host), ##args);) + fc_host_port_id(lport->host), ##args)) /* * libfc error codes -- cgit v1.2.3 From 7f74549ff630ad444b0b6bbcabf426f781910906 Mon Sep 17 00:00:00 2001 From: Joe Eykholt Date: Wed, 29 Jul 2009 17:04:12 -0700 Subject: [SCSI] libfc: change debug messages to give host number. libfc debug messages currently show 'lport: :' wher is the hex assigned port-id. When the lport is logged off, that will be zero, so its hard to distinguish which instance is involved. The FC-ID can change if the port is re-patched or changes VSANs. Two lports may even have the same FC-ID if connected to isolated SANs. Change the debug messages to print the SCSI host number "hostN:", which will not change for the life of the lport. Still show the FC_ID on lport messages. Also, add a macro to FC_RPORT_ID_DBG for rport debugging where there's no rdata structure involved. It takes the lport and port_id as parameters. Use this in fc_rport_recv_plogi_req() and fc_rport_recv_logo_req(). Signed-off-by: Joe Eykholt Signed-off-by: Robert Love Signed-off-by: James Bottomley Signed-off-by: James Bottomley --- include/scsi/libfc.h | 36 ++++++++++++++++++++---------------- 1 file changed, 20 insertions(+), 16 deletions(-) (limited to 'include') diff --git a/include/scsi/libfc.h b/include/scsi/libfc.h index ef04a2c52b8c..efdb6ba310e5 100644 --- a/include/scsi/libfc.h +++ b/include/scsi/libfc.h @@ -59,47 +59,51 @@ do { \ #define FC_LPORT_DBG(lport, fmt, args...) \ FC_CHECK_LOGGING(FC_LPORT_LOGGING, \ - printk(KERN_INFO "lport: %6x: " fmt, \ - fc_host_port_id(lport->host), ##args)) + printk(KERN_INFO "host%u: lport %6x: " fmt, \ + (lport)->host->host_no, \ + fc_host_port_id((lport)->host), ##args)) #define FC_DISC_DBG(disc, fmt, args...) \ FC_CHECK_LOGGING(FC_DISC_LOGGING, \ - printk(KERN_INFO "disc: %6x: " fmt, \ - fc_host_port_id(disc->lport->host), \ + printk(KERN_INFO "host%u: disc: " fmt, \ + (disc)->lport->host->host_no, \ ##args)) +#define FC_RPORT_ID_DBG(lport, port_id, fmt, args...) \ + FC_CHECK_LOGGING(FC_RPORT_LOGGING, \ + printk(KERN_INFO "host%u: rport %6x: " fmt, \ + (lport)->host->host_no, \ + (port_id), ##args)) + #define FC_RPORT_DBG(rport, fmt, args...) \ do { \ struct fc_rport_libfc_priv *rdata = rport->dd_data; \ struct fc_lport *lport = rdata->local_port; \ - FC_CHECK_LOGGING(FC_RPORT_LOGGING, \ - printk(KERN_INFO "rport: %6x: %6x: " fmt, \ - fc_host_port_id(lport->host), \ - rport->port_id, ##args)); \ + FC_RPORT_ID_DBG(lport, rport->port_id, fmt, ##args); \ } while (0) #define FC_FCP_DBG(pkt, fmt, args...) \ FC_CHECK_LOGGING(FC_FCP_LOGGING, \ - printk(KERN_INFO "fcp: %6x: %6x: " fmt, \ - fc_host_port_id(pkt->lp->host), \ + printk(KERN_INFO "host%u: fcp: %6x: " fmt, \ + (pkt)->lp->host->host_no, \ pkt->rport->port_id, ##args)) #define FC_EM_DBG(em, fmt, args...) \ FC_CHECK_LOGGING(FC_EM_LOGGING, \ - printk(KERN_INFO "em: %6x: " fmt, \ - fc_host_port_id(em->lp->host), \ + printk(KERN_INFO "host%u: em: " fmt, \ + (em)->lp->host->host_no, \ ##args)) #define FC_EXCH_DBG(exch, fmt, args...) \ FC_CHECK_LOGGING(FC_EXCH_LOGGING, \ - printk(KERN_INFO "exch: %6x: %4x: " fmt, \ - fc_host_port_id(exch->lp->host), \ + printk(KERN_INFO "host%u: xid %4x: " fmt, \ + (exch)->lp->host->host_no, \ exch->xid, ##args)) #define FC_SCSI_DBG(lport, fmt, args...) \ FC_CHECK_LOGGING(FC_SCSI_LOGGING, \ - printk(KERN_INFO "scsi: %6x: " fmt, \ - fc_host_port_id(lport->host), ##args)) + printk(KERN_INFO "host%u: scsi: " fmt, \ + (lport)->host->host_no, ##args)) /* * libfc error codes -- cgit v1.2.3 From b1d9fd5574763abe5c763e32e3547a4adee9bd88 Mon Sep 17 00:00:00 2001 From: Joe Eykholt Date: Wed, 29 Jul 2009 17:04:22 -0700 Subject: [SCSI] libfc: rename lport NONE state to DISABLED The state NONE was meant to be invalid, but has been used as the initial state. Rename it to be DISABLED, as more descriptive. Further patches will make it the like the RESET state, except it won't transition to FLOGI until fc_lport_fabric_login() is called. Signed-off-by: Joe Eykholt Signed-off-by: Robert Love Signed-off-by: James Bottomley Signed-off-by: James Bottomley --- drivers/scsi/libfc/fc_exch.c | 2 +- drivers/scsi/libfc/fc_lport.c | 12 ++++++------ include/scsi/libfc.h | 2 +- 3 files changed, 8 insertions(+), 8 deletions(-) (limited to 'include') diff --git a/drivers/scsi/libfc/fc_exch.c b/drivers/scsi/libfc/fc_exch.c index 145ab9ba55ea..e6d82d780acd 100644 --- a/drivers/scsi/libfc/fc_exch.c +++ b/drivers/scsi/libfc/fc_exch.c @@ -1875,7 +1875,7 @@ void fc_exch_recv(struct fc_lport *lp, struct fc_exch_mgr *mp, u32 f_ctl; /* lport lock ? */ - if (!lp || !mp || (lp->state == LPORT_ST_NONE)) { + if (!lp || !mp || lp->state == LPORT_ST_DISABLED) { FC_LPORT_DBG(lp, "Receiving frames for an lport that " "has not been initialized correctly\n"); fc_frame_free(fp); diff --git a/drivers/scsi/libfc/fc_lport.c b/drivers/scsi/libfc/fc_lport.c index 745fa5555d6a..3b28190ca2eb 100644 --- a/drivers/scsi/libfc/fc_lport.c +++ b/drivers/scsi/libfc/fc_lport.c @@ -113,7 +113,7 @@ static void fc_lport_enter_ready(struct fc_lport *); static void fc_lport_enter_logo(struct fc_lport *); static const char *fc_lport_state_names[] = { - [LPORT_ST_NONE] = "none", + [LPORT_ST_DISABLED] = "disabled", [LPORT_ST_FLOGI] = "FLOGI", [LPORT_ST_DNS] = "dNS", [LPORT_ST_RPN_ID] = "RPN_ID", @@ -550,7 +550,7 @@ int fc_fabric_login(struct fc_lport *lport) int rc = -1; mutex_lock(&lport->lp_mutex); - if (lport->state == LPORT_ST_NONE) { + if (lport->state == LPORT_ST_DISABLED) { fc_lport_enter_reset(lport); rc = 0; } @@ -637,7 +637,7 @@ EXPORT_SYMBOL(fc_fabric_logoff); int fc_lport_destroy(struct fc_lport *lport) { mutex_lock(&lport->lp_mutex); - lport->state = LPORT_ST_NONE; + lport->state = LPORT_ST_DISABLED; lport->link_up = 0; lport->tt.frame_send = fc_frame_drop; mutex_unlock(&lport->lp_mutex); @@ -992,7 +992,7 @@ static void fc_lport_error(struct fc_lport *lport, struct fc_frame *fp) schedule_delayed_work(&lport->retry_work, delay); } else { switch (lport->state) { - case LPORT_ST_NONE: + case LPORT_ST_DISABLED: case LPORT_ST_READY: case LPORT_ST_RESET: case LPORT_ST_RPN_ID: @@ -1316,7 +1316,7 @@ static void fc_lport_timeout(struct work_struct *work) mutex_lock(&lport->lp_mutex); switch (lport->state) { - case LPORT_ST_NONE: + case LPORT_ST_DISABLED: case LPORT_ST_READY: case LPORT_ST_RESET: WARN_ON(1); @@ -1550,7 +1550,7 @@ int fc_lport_config(struct fc_lport *lport) INIT_DELAYED_WORK(&lport->retry_work, fc_lport_timeout); mutex_init(&lport->lp_mutex); - fc_lport_state_enter(lport, LPORT_ST_NONE); + fc_lport_state_enter(lport, LPORT_ST_DISABLED); fc_lport_add_fc4_type(lport, FC_TYPE_FCP); fc_lport_add_fc4_type(lport, FC_TYPE_CT); diff --git a/include/scsi/libfc.h b/include/scsi/libfc.h index efdb6ba310e5..b5c9b285b462 100644 --- a/include/scsi/libfc.h +++ b/include/scsi/libfc.h @@ -129,7 +129,7 @@ do { \ * FC HBA status */ enum fc_lport_state { - LPORT_ST_NONE = 0, + LPORT_ST_DISABLED = 0, LPORT_ST_FLOGI, LPORT_ST_DNS, LPORT_ST_RPN_ID, -- cgit v1.2.3 From 141940548c6919c22bf0573c68fd59d961e22475 Mon Sep 17 00:00:00 2001 From: Joe Eykholt Date: Wed, 29 Jul 2009 17:04:43 -0700 Subject: [SCSI] libfc: rename rport state "NONE" to "DELETE". State RPORT_ST_NONE was intented to be an invalid state (0), never used. This was a misguided attempt to be sure it was always initialized. Having an extra state meaning nothing requires switch statements to have a case covering that state. State NONE has been used instead to mean the remote port is being deleted. Changing the name to RPORT_ST_DELETE. Signed-off-by: Joe Eykholt Signed-off-by: Robert Love Signed-off-by: James Bottomley Signed-off-by: James Bottomley --- drivers/scsi/libfc/fc_rport.c | 28 ++++++++++++++-------------- include/scsi/libfc.h | 2 +- 2 files changed, 15 insertions(+), 15 deletions(-) (limited to 'include') diff --git a/drivers/scsi/libfc/fc_rport.c b/drivers/scsi/libfc/fc_rport.c index 7162385f52eb..bf7364fc16cb 100644 --- a/drivers/scsi/libfc/fc_rport.c +++ b/drivers/scsi/libfc/fc_rport.c @@ -77,13 +77,13 @@ static void fc_rport_error_retry(struct fc_rport *, struct fc_frame *); static void fc_rport_work(struct work_struct *); static const char *fc_rport_state_names[] = { - [RPORT_ST_NONE] = "None", [RPORT_ST_INIT] = "Init", [RPORT_ST_PLOGI] = "PLOGI", [RPORT_ST_PRLI] = "PRLI", [RPORT_ST_RTV] = "RTV", [RPORT_ST_READY] = "Ready", [RPORT_ST_LOGO] = "LOGO", + [RPORT_ST_DELETE] = "Delete", }; static void fc_rport_rogue_destroy(struct device *dev) @@ -326,8 +326,8 @@ int fc_rport_logoff(struct fc_rport *rport) FC_RPORT_DBG(rport, "Remove port\n"); - if (rdata->rp_state == RPORT_ST_NONE) { - FC_RPORT_DBG(rport, "Port in NONE state, not removing\n"); + if (rdata->rp_state == RPORT_ST_DELETE) { + FC_RPORT_DBG(rport, "Port in Delete state, not removing\n"); mutex_unlock(&rdata->rp_mutex); goto out; } @@ -335,10 +335,10 @@ int fc_rport_logoff(struct fc_rport *rport) fc_rport_enter_logo(rport); /* - * Change the state to NONE so that we discard + * Change the state to Delete so that we discard * the response. */ - fc_rport_state_enter(rport, RPORT_ST_NONE); + fc_rport_state_enter(rport, RPORT_ST_DELETE); mutex_unlock(&rdata->rp_mutex); @@ -405,7 +405,7 @@ static void fc_rport_timeout(struct work_struct *work) break; case RPORT_ST_READY: case RPORT_ST_INIT: - case RPORT_ST_NONE: + case RPORT_ST_DELETE: break; } @@ -433,14 +433,14 @@ static void fc_rport_error(struct fc_rport *rport, struct fc_frame *fp) case RPORT_ST_PRLI: case RPORT_ST_LOGO: rdata->event = RPORT_EV_FAILED; - fc_rport_state_enter(rport, RPORT_ST_NONE); + fc_rport_state_enter(rport, RPORT_ST_DELETE); queue_work(rport_event_queue, &rdata->event_work); break; case RPORT_ST_RTV: fc_rport_enter_ready(rport); break; - case RPORT_ST_NONE: + case RPORT_ST_DELETE: case RPORT_ST_READY: case RPORT_ST_INIT: break; @@ -652,7 +652,7 @@ static void fc_rport_prli_resp(struct fc_seq *sp, struct fc_frame *fp, } else { FC_RPORT_DBG(rport, "Bad ELS response for PRLI command\n"); rdata->event = RPORT_EV_FAILED; - fc_rport_state_enter(rport, RPORT_ST_NONE); + fc_rport_state_enter(rport, RPORT_ST_DELETE); queue_work(rport_event_queue, &rdata->event_work); } @@ -703,7 +703,7 @@ static void fc_rport_logo_resp(struct fc_seq *sp, struct fc_frame *fp, } else { FC_RPORT_DBG(rport, "Bad ELS response for LOGO command\n"); rdata->event = RPORT_EV_LOGO; - fc_rport_state_enter(rport, RPORT_ST_NONE); + fc_rport_state_enter(rport, RPORT_ST_DELETE); queue_work(rport_event_queue, &rdata->event_work); } @@ -1012,7 +1012,7 @@ static void fc_rport_recv_plogi_req(struct fc_rport *rport, "- ignored for now\n", rdata->rp_state); /* XXX TBD - should reset */ break; - case RPORT_ST_NONE: + case RPORT_ST_DELETE: default: FC_RPORT_DBG(rport, "Received PLOGI in unexpected " "state %d\n", rdata->rp_state); @@ -1238,7 +1238,7 @@ static void fc_rport_recv_prlo_req(struct fc_rport *rport, struct fc_seq *sp, FC_RPORT_DBG(rport, "Received PRLO request while in state %s\n", fc_rport_state(rport)); - if (rdata->rp_state == RPORT_ST_NONE) { + if (rdata->rp_state == RPORT_ST_DELETE) { fc_frame_free(fp); return; } @@ -1271,13 +1271,13 @@ static void fc_rport_recv_logo_req(struct fc_rport *rport, struct fc_seq *sp, FC_RPORT_DBG(rport, "Received LOGO request while in state %s\n", fc_rport_state(rport)); - if (rdata->rp_state == RPORT_ST_NONE) { + if (rdata->rp_state == RPORT_ST_DELETE) { fc_frame_free(fp); return; } rdata->event = RPORT_EV_LOGO; - fc_rport_state_enter(rport, RPORT_ST_NONE); + fc_rport_state_enter(rport, RPORT_ST_DELETE); queue_work(rport_event_queue, &rdata->event_work); lport->tt.seq_els_rsp_send(sp, ELS_LS_ACC, NULL); diff --git a/include/scsi/libfc.h b/include/scsi/libfc.h index b5c9b285b462..04db7a9e631b 100644 --- a/include/scsi/libfc.h +++ b/include/scsi/libfc.h @@ -147,13 +147,13 @@ enum fc_disc_event { }; enum fc_rport_state { - RPORT_ST_NONE = 0, RPORT_ST_INIT, /* initialized */ RPORT_ST_PLOGI, /* waiting for PLOGI completion */ RPORT_ST_PRLI, /* waiting for PRLI completion */ RPORT_ST_RTV, /* waiting for RTV completion */ RPORT_ST_READY, /* ready for use */ RPORT_ST_LOGO, /* port logout sent */ + RPORT_ST_DELETE, /* port being deleted */ }; enum fc_rport_trans_state { -- cgit v1.2.3 From 96316099ac3cb259eac2d6891f3c75b38b29d26e Mon Sep 17 00:00:00 2001 From: Vasu Dev Date: Wed, 29 Jul 2009 17:05:00 -0700 Subject: [SCSI] fcoe, libfc: adds exchange manager(EM) anchor list per lport and related APIs Adds EM list using a anchor struct fc_exch_mgr_anchor, anchor is used to allow same EM instance sharing across more than one lport on a eth device, this implementation is per discussed design posted at http://www.open-fcoe.org/pipermail/devel/2009-June/002566.html. The shared EM is required for multiple lports on eth device when using multiple VLANs or NPIV. Adds fc_exch_mgr_add API to add a EM to the lport and fc_exch_mgr_del API to delete previously added EM. Also adds function fc_exch_mgr_destroy() to destroy allocated EM. The kref is added to the EM to keep track of EM usage count, the EM is destroyed when no longer in use upon kref reaching to zero. The caller can specify match function to fc_exch_mgr_add, this will be used in determining exchange allocation from its EM or not. Moved calling of fcoe_em_config below fcoe_libfc_config calling, so that list head lp->ema_list is initialized before configuring EM. Signed-off-by: Vasu Dev Signed-off-by: Robert Love Signed-off-by: James Bottomley Signed-off-by: James Bottomley --- drivers/scsi/fcoe/fcoe.c | 14 ++++++------- drivers/scsi/libfc/fc_exch.c | 48 +++++++++++++++++++++++++++++++++++++++++++ drivers/scsi/libfc/fc_lport.c | 1 + include/scsi/libfc.h | 24 ++++++++++++++++++++++ 4 files changed, 80 insertions(+), 7 deletions(-) (limited to 'include') diff --git a/drivers/scsi/fcoe/fcoe.c b/drivers/scsi/fcoe/fcoe.c index 14a4017a1535..719a99d4a438 100644 --- a/drivers/scsi/fcoe/fcoe.c +++ b/drivers/scsi/fcoe/fcoe.c @@ -603,18 +603,18 @@ static int fcoe_if_create(struct net_device *netdev) goto out_netdev_cleanup; } - /* lport exch manager allocation */ - rc = fcoe_em_config(lp); + /* Initialize the library */ + rc = fcoe_libfc_config(lp, &fcoe_libfc_fcn_templ); if (rc) { - FCOE_NETDEV_DBG(netdev, "Could not configure the EM for the " + FCOE_NETDEV_DBG(netdev, "Could not configure libfc for the " "interface\n"); - goto out_netdev_cleanup; + goto out_lp_destroy; } - /* Initialize the library */ - rc = fcoe_libfc_config(lp, &fcoe_libfc_fcn_templ); + /* lport exch manager allocation */ + rc = fcoe_em_config(lp); if (rc) { - FCOE_NETDEV_DBG(netdev, "Could not configure libfc for the " + FCOE_NETDEV_DBG(netdev, "Could not configure the EM for the " "interface\n"); goto out_lp_destroy; } diff --git a/drivers/scsi/libfc/fc_exch.c b/drivers/scsi/libfc/fc_exch.c index cab54996375c..f1fa2b196e98 100644 --- a/drivers/scsi/libfc/fc_exch.c +++ b/drivers/scsi/libfc/fc_exch.c @@ -55,6 +55,7 @@ static struct kmem_cache *fc_em_cachep; /* cache for exchanges */ */ struct fc_exch_mgr { enum fc_class class; /* default class for sequences */ + struct kref kref; /* exchange mgr reference count */ spinlock_t em_lock; /* exchange manager lock, must be taken before ex_lock */ u16 last_xid; /* last allocated exchange ID */ @@ -84,6 +85,12 @@ struct fc_exch_mgr { }; #define fc_seq_exch(sp) container_of(sp, struct fc_exch, seq) +struct fc_exch_mgr_anchor { + struct list_head ema_list; + struct fc_exch_mgr *mp; + bool (*match)(struct fc_frame *); +}; + static void fc_exch_rrq(struct fc_exch *); static void fc_seq_ls_acc(struct fc_seq *); static void fc_seq_ls_rjt(struct fc_seq *, enum fc_els_rjt_reason, @@ -1729,6 +1736,47 @@ reject: fc_frame_free(fp); } +struct fc_exch_mgr_anchor *fc_exch_mgr_add(struct fc_lport *lport, + struct fc_exch_mgr *mp, + bool (*match)(struct fc_frame *)) +{ + struct fc_exch_mgr_anchor *ema; + + ema = kmalloc(sizeof(*ema), GFP_ATOMIC); + if (!ema) + return ema; + + ema->mp = mp; + ema->match = match; + /* add EM anchor to EM anchors list */ + list_add_tail(&ema->ema_list, &lport->ema_list); + kref_get(&mp->kref); + return ema; +} +EXPORT_SYMBOL(fc_exch_mgr_add); + +static void fc_exch_mgr_destroy(struct kref *kref) +{ + struct fc_exch_mgr *mp = container_of(kref, struct fc_exch_mgr, kref); + + /* + * The total exch count must be zero + * before freeing exchange manager. + */ + WARN_ON(mp->total_exches != 0); + mempool_destroy(mp->ep_pool); + kfree(mp); +} + +void fc_exch_mgr_del(struct fc_exch_mgr_anchor *ema) +{ + /* remove EM anchor from EM anchors list */ + list_del(&ema->ema_list); + kref_put(&ema->mp->kref, fc_exch_mgr_destroy); + kfree(ema); +} +EXPORT_SYMBOL(fc_exch_mgr_del); + struct fc_exch_mgr *fc_exch_mgr_alloc(struct fc_lport *lp, enum fc_class class, u16 min_xid, u16 max_xid) diff --git a/drivers/scsi/libfc/fc_lport.c b/drivers/scsi/libfc/fc_lport.c index a430335ebf59..ca8ea264b684 100644 --- a/drivers/scsi/libfc/fc_lport.c +++ b/drivers/scsi/libfc/fc_lport.c @@ -1618,6 +1618,7 @@ int fc_lport_init(struct fc_lport *lport) if (lport->link_supported_speeds & FC_PORTSPEED_10GBIT) fc_host_supported_speeds(lport->host) |= FC_PORTSPEED_10GBIT; + INIT_LIST_HEAD(&lport->ema_list); return 0; } EXPORT_SYMBOL(fc_lport_init); diff --git a/include/scsi/libfc.h b/include/scsi/libfc.h index 04db7a9e631b..b381b1ca9aec 100644 --- a/include/scsi/libfc.h +++ b/include/scsi/libfc.h @@ -348,6 +348,7 @@ static inline bool fc_fcp_is_read(const struct fc_fcp_pkt *fsp) */ struct fc_exch_mgr; +struct fc_exch_mgr_anchor; /* * Sequence. @@ -709,6 +710,7 @@ struct fc_lport { /* Associations */ struct Scsi_Host *host; struct fc_exch_mgr *emp; + struct list_head ema_list; struct fc_rport *dns_rp; struct fc_rport *ptp_rp; void *scsi_priv; @@ -963,6 +965,28 @@ int fc_elsct_init(struct fc_lport *lp); */ int fc_exch_init(struct fc_lport *lp); +/* + * Adds Exchange Manager (EM) mp to lport. + * + * Adds specified mp to lport using struct fc_exch_mgr_anchor, + * the struct fc_exch_mgr_anchor allows same EM sharing by + * more than one lport with their specified match function, + * the match function is used in allocating exchange from + * added mp. + */ +struct fc_exch_mgr_anchor *fc_exch_mgr_add(struct fc_lport *lport, + struct fc_exch_mgr *mp, + bool (*match)(struct fc_frame *)); + +/* + * Deletes Exchange Manager (EM) from lport by removing + * its anchor ema from lport. + * + * If removed anchor ema was the last user of its associated EM + * then also destroys associated EM. + */ +void fc_exch_mgr_del(struct fc_exch_mgr_anchor *ema); + /* * Allocates an Exchange Manager (EM). * -- cgit v1.2.3 From d459b7ea1b4c7aa3dacfeee174d02b2f7a95850d Mon Sep 17 00:00:00 2001 From: Robert Love Date: Wed, 29 Jul 2009 17:05:05 -0700 Subject: [SCSI] libfc: Remove the FC_EM_DBG macro Currently there is a 1:1 relationship between the lport and exchange manager. This macro takes an EM as an argument and determines the lport from it. However, later patches will use an EM list per lport, so we will no longer have this 1:1 relationship- this macro must change. The FC_EM_DBG macro is rarely used. There are four callers, two can use FC_LPORT_DBG instead and two can be removed since they're not necessary. This patch makes those changes and removes the macro. Signed-off-by: Robert Love Signed-off-by: James Bottomley Signed-off-by: James Bottomley --- drivers/scsi/libfc/fc_exch.c | 13 ++++++------- include/scsi/libfc.h | 6 ------ 2 files changed, 6 insertions(+), 13 deletions(-) (limited to 'include') diff --git a/drivers/scsi/libfc/fc_exch.c b/drivers/scsi/libfc/fc_exch.c index f1fa2b196e98..3ad7f88e7ae3 100644 --- a/drivers/scsi/libfc/fc_exch.c +++ b/drivers/scsi/libfc/fc_exch.c @@ -1129,7 +1129,7 @@ static void fc_exch_recv_req(struct fc_lport *lp, struct fc_exch_mgr *mp, lp->tt.lport_recv(lp, sp, fp); fc_exch_release(ep); /* release from lookup */ } else { - FC_EM_DBG(mp, "exch/seq lookup failed: reject %x\n", reject); + FC_LPORT_DBG(lp, "exch/seq lookup failed: reject %x\n", reject); fc_frame_free(fp); } } @@ -1235,13 +1235,12 @@ static void fc_exch_recv_resp(struct fc_exch_mgr *mp, struct fc_frame *fp) struct fc_seq *sp; sp = fc_seq_lookup_orig(mp, fp); /* doesn't hold sequence */ - if (!sp) { + + if (!sp) atomic_inc(&mp->stats.xid_not_found); - FC_EM_DBG(mp, "seq lookup failed\n"); - } else { + else atomic_inc(&mp->stats.non_bls_resp); - FC_EM_DBG(mp, "non-BLS response to sequence"); - } + fc_frame_free(fp); } @@ -1950,7 +1949,7 @@ void fc_exch_recv(struct fc_lport *lp, struct fc_exch_mgr *mp, fc_exch_recv_req(lp, mp, fp); break; default: - FC_EM_DBG(mp, "dropping invalid frame (eof %x)", fr_eof(fp)); + FC_LPORT_DBG(lp, "dropping invalid frame (eof %x)", fr_eof(fp)); fc_frame_free(fp); break; } diff --git a/include/scsi/libfc.h b/include/scsi/libfc.h index b381b1ca9aec..f1bde91f98a2 100644 --- a/include/scsi/libfc.h +++ b/include/scsi/libfc.h @@ -88,12 +88,6 @@ do { \ (pkt)->lp->host->host_no, \ pkt->rport->port_id, ##args)) -#define FC_EM_DBG(em, fmt, args...) \ - FC_CHECK_LOGGING(FC_EM_LOGGING, \ - printk(KERN_INFO "host%u: em: " fmt, \ - (em)->lp->host->host_no, \ - ##args)) - #define FC_EXCH_DBG(exch, fmt, args...) \ FC_CHECK_LOGGING(FC_EXCH_LOGGING, \ printk(KERN_INFO "host%u: xid %4x: " fmt, \ -- cgit v1.2.3 From 52ff878c912215210f53c0a080552dd6ba3055a2 Mon Sep 17 00:00:00 2001 From: Vasu Dev Date: Wed, 29 Jul 2009 17:05:10 -0700 Subject: [SCSI] fcoe, fnic, libfc: modifies current code paths to use EM anchor list Modifies current code to use EM anchor list in EM allocation, EM free, EM reset, exch allocation and exch lookup code paths. 1. Modifies fc_exch_mgr_alloc to accept EM match function and then have allocated EM added to the lport using fc_exch_mgr_add API while also updating EM kref for newly added EM. 2. Updates fc_exch_mgr_free API to accept only lport pointer instead EM and then have this API free all EMs of the lport from EM anchor list. 3. Removes single lport pointer link from the EM, which was used in associating lport pointer in newly allocated exchange. Instead have lport pointer passed along new exchange allocation call path and then store passed lport pointer in newly allocated exchange, this will allow a single EM instance to be used across more than one lport and used in EM reset to reset only lport specific exchanges. 4. Modifies fc_exch_mgr_reset to reset all EMs from the EM anchor list of the lport, adds additional exch lport pointer (ep->lp) check for shared EM case to reset exchange specific to a lport requested reset. 5. Updates exch allocation API fc_exch_alloc to use EM anchor list and its anchor match func pointer. The fc_exch_alloc will walk the list of EMs until it finds a match, a match will be either null match func pointer or call to match function returning true value. 6. Updates fc_exch_recv to accept incoming frame on local port using only lport pointer and frame pointer without specifying EM instance of incoming frame. Instead modified fc_exch_recv to locate EM for the incoming frame by matching xid of incoming frame against a EM xid range. This change was required to use EM list in libfc Rx path and after this change the lport fc_exch_mgr pointer emp is not needed anymore, so removed emp pointer. 7. Updates fnic for removed lport emp pointer and above modified libfc APIs fc_exch_recv, fc_exch_mgr_alloc and fc_exch_mgr_free. 8. Removes exch_get and exch_put from libfc_function_template as these are no longer needed with EM anchor list and its match function use. Also removes its default function fc_exch_get. A defect this patch introduced regarding the libfc initialization order in the fnic driver was fixed by Joe Eykholt . Signed-off-by: Vasu Dev Signed-off-by: Robert Love Signed-off-by: James Bottomley Signed-off-by: James Bottomley --- drivers/scsi/fcoe/fcoe.c | 16 ++-- drivers/scsi/fcoe/libfcoe.c | 2 +- drivers/scsi/fnic/fnic_fcs.c | 2 +- drivers/scsi/fnic/fnic_main.c | 20 ++--- drivers/scsi/libfc/fc_exch.c | 191 ++++++++++++++++++++++++++---------------- include/scsi/libfc.h | 48 ++--------- 6 files changed, 142 insertions(+), 137 deletions(-) (limited to 'include') diff --git a/drivers/scsi/fcoe/fcoe.c b/drivers/scsi/fcoe/fcoe.c index 719a99d4a438..ebf2e20370d7 100644 --- a/drivers/scsi/fcoe/fcoe.c +++ b/drivers/scsi/fcoe/fcoe.c @@ -423,11 +423,8 @@ static int fcoe_shost_config(struct fc_lport *lp, struct Scsi_Host *shost, */ static inline int fcoe_em_config(struct fc_lport *lp) { - BUG_ON(lp->emp); - - lp->emp = fc_exch_mgr_alloc(lp, FC_CLASS_3, - FCOE_MIN_XID, FCOE_MAX_XID); - if (!lp->emp) + if (!fc_exch_mgr_alloc(lp, FC_CLASS_3, FCOE_MIN_XID, + FCOE_MAX_XID, NULL)) return -ENOMEM; return 0; @@ -478,8 +475,7 @@ static int fcoe_if_destroy(struct net_device *netdev) scsi_remove_host(lp->host); /* There are no more rports or I/O, free the EM */ - if (lp->emp) - fc_exch_mgr_free(lp->emp); + fc_exch_mgr_free(lp); /* Free existing skbs */ fcoe_clean_pending_queue(lp); @@ -634,7 +630,7 @@ static int fcoe_if_create(struct net_device *netdev) return rc; out_lp_destroy: - fc_exch_mgr_free(lp->emp); /* Free the EM */ + fc_exch_mgr_free(lp); out_netdev_cleanup: fcoe_netdev_cleanup(fc); out_host_put: @@ -1277,7 +1273,7 @@ int fcoe_percpu_receive_thread(void *arg) fh = fc_frame_header_get(fp); if (fh->fh_r_ctl == FC_RCTL_DD_SOL_DATA && fh->fh_type == FC_TYPE_FCP) { - fc_exch_recv(lp, lp->emp, fp); + fc_exch_recv(lp, fp); continue; } if (fr_flags(fp) & FCPHF_CRC_UNCHECKED) { @@ -1298,7 +1294,7 @@ int fcoe_percpu_receive_thread(void *arg) fc_frame_free(fp); continue; } - fc_exch_recv(lp, lp->emp, fp); + fc_exch_recv(lp, fp); } return 0; } diff --git a/drivers/scsi/fcoe/libfcoe.c b/drivers/scsi/fcoe/libfcoe.c index 78caa6be1130..4db719d6ada1 100644 --- a/drivers/scsi/fcoe/libfcoe.c +++ b/drivers/scsi/fcoe/libfcoe.c @@ -885,7 +885,7 @@ static void fcoe_ctlr_recv_els(struct fcoe_ctlr *fip, struct sk_buff *skb) stats->RxFrames++; stats->RxWords += skb->len / FIP_BPW; - fc_exch_recv(lp, lp->emp, fp); + fc_exch_recv(lp, fp); return; len_err: diff --git a/drivers/scsi/fnic/fnic_fcs.c b/drivers/scsi/fnic/fnic_fcs.c index 07e6eedb83ce..50db3e36a619 100644 --- a/drivers/scsi/fnic/fnic_fcs.c +++ b/drivers/scsi/fnic/fnic_fcs.c @@ -115,7 +115,7 @@ void fnic_handle_frame(struct work_struct *work) } spin_unlock_irqrestore(&fnic->fnic_lock, flags); - fc_exch_recv(lp, lp->emp, fp); + fc_exch_recv(lp, fp); } } diff --git a/drivers/scsi/fnic/fnic_main.c b/drivers/scsi/fnic/fnic_main.c index 2c266c01dc5a..71c7bbe26d05 100644 --- a/drivers/scsi/fnic/fnic_main.c +++ b/drivers/scsi/fnic/fnic_main.c @@ -671,14 +671,6 @@ static int __devinit fnic_probe(struct pci_dev *pdev, lp->link_up = 0; lp->tt = fnic_transport_template; - lp->emp = fc_exch_mgr_alloc(lp, FC_CLASS_3, - FCPIO_HOST_EXCH_RANGE_START, - FCPIO_HOST_EXCH_RANGE_END); - if (!lp->emp) { - err = -ENOMEM; - goto err_out_remove_scsi_host; - } - lp->max_retry_count = fnic->config.flogi_retries; lp->max_rport_retry_count = fnic->config.plogi_retries; lp->service_params = (FCP_SPPF_INIT_FCN | FCP_SPPF_RD_XRDY_DIS | @@ -693,12 +685,18 @@ static int __devinit fnic_probe(struct pci_dev *pdev, fc_set_wwnn(lp, fnic->config.node_wwn); fc_set_wwpn(lp, fnic->config.port_wwn); - fc_exch_init(lp); fc_lport_init(lp); + fc_exch_init(lp); fc_elsct_init(lp); fc_rport_init(lp); fc_disc_init(lp); + if (!fc_exch_mgr_alloc(lp, FC_CLASS_3, FCPIO_HOST_EXCH_RANGE_START, + FCPIO_HOST_EXCH_RANGE_END, NULL)) { + err = -ENOMEM; + goto err_out_remove_scsi_host; + } + fc_lport_config(lp); if (fc_set_mfs(lp, fnic->config.maxdatafieldsize + @@ -738,7 +736,7 @@ static int __devinit fnic_probe(struct pci_dev *pdev, return 0; err_out_free_exch_mgr: - fc_exch_mgr_free(lp->emp); + fc_exch_mgr_free(lp); err_out_remove_scsi_host: fc_remove_host(fnic->lport->host); scsi_remove_host(fnic->lport->host); @@ -827,7 +825,7 @@ static void __devexit fnic_remove(struct pci_dev *pdev) fc_remove_host(fnic->lport->host); scsi_remove_host(fnic->lport->host); - fc_exch_mgr_free(fnic->lport->emp); + fc_exch_mgr_free(fnic->lport); vnic_dev_notify_unset(fnic->vdev); fnic_free_vnic_resources(fnic); fnic_free_intr(fnic); diff --git a/drivers/scsi/libfc/fc_exch.c b/drivers/scsi/libfc/fc_exch.c index 3ad7f88e7ae3..324589a5cc03 100644 --- a/drivers/scsi/libfc/fc_exch.c +++ b/drivers/scsi/libfc/fc_exch.c @@ -65,7 +65,6 @@ struct fc_exch_mgr { u16 last_read; /* last xid allocated for read */ u32 total_exches; /* total allocated exchanges */ struct list_head ex_list; /* allocated exchanges list */ - struct fc_lport *lp; /* fc device instance */ mempool_t *ep_pool; /* reserve ep's */ /* @@ -275,8 +274,6 @@ static void fc_exch_release(struct fc_exch *ep) mp = ep->em; if (ep->destructor) ep->destructor(&ep->seq, ep->arg); - if (ep->lp->tt.exch_put) - ep->lp->tt.exch_put(ep->lp, mp, ep->xid); WARN_ON(!(ep->esb_stat & ESB_ST_COMPLETE)); mempool_free(ep, mp->ep_pool); } @@ -513,17 +510,20 @@ static u16 fc_em_alloc_xid(struct fc_exch_mgr *mp, const struct fc_frame *fp) return xid; } -/* - * fc_exch_alloc - allocate an exchange. - * @mp : ptr to the exchange manager - * @xid: input xid +/** + * fc_exch_em_alloc() - allocate an exchange from a specified EM. + * @lport: ptr to the local port + * @mp: ptr to the exchange manager + * @fp: ptr to the FC frame + * @xid: input xid * * if xid is supplied zero then assign next free exchange ID * from exchange manager, otherwise use supplied xid. * Returns with exch lock held. */ -struct fc_exch *fc_exch_alloc(struct fc_exch_mgr *mp, - struct fc_frame *fp, u16 xid) +static struct fc_exch *fc_exch_em_alloc(struct fc_lport *lport, + struct fc_exch_mgr *mp, + struct fc_frame *fp, u16 xid) { struct fc_exch *ep; @@ -566,7 +566,7 @@ struct fc_exch *fc_exch_alloc(struct fc_exch_mgr *mp, */ ep->oxid = ep->xid = xid; ep->em = mp; - ep->lp = mp->lp; + ep->lp = lport; ep->f_ctl = FC_FC_FIRST_SEQ; /* next seq is first seq */ ep->rxid = FC_XID_UNKNOWN; ep->class = mp->class; @@ -579,6 +579,31 @@ err: mempool_free(ep, mp->ep_pool); return NULL; } + +/** + * fc_exch_alloc() - allocate an exchange. + * @lport: ptr to the local port + * @fp: ptr to the FC frame + * + * This function walks the list of the exchange manager(EM) + * anchors to select a EM for new exchange allocation. The + * EM is selected having either a NULL match function pointer + * or call to match function returning true. + */ +struct fc_exch *fc_exch_alloc(struct fc_lport *lport, struct fc_frame *fp) +{ + struct fc_exch_mgr_anchor *ema; + struct fc_exch *ep; + + list_for_each_entry(ema, &lport->ema_list, ema_list) { + if (!ema->match || ema->match(fp)) { + ep = fc_exch_em_alloc(lport, ema->mp, fp, 0); + if (ep) + return ep; + } + } + return NULL; +} EXPORT_SYMBOL(fc_exch_alloc); /* @@ -617,12 +642,14 @@ EXPORT_SYMBOL(fc_exch_done); * Allocate a new exchange as responder. * Sets the responder ID in the frame header. */ -static struct fc_exch *fc_exch_resp(struct fc_exch_mgr *mp, struct fc_frame *fp) +static struct fc_exch *fc_exch_resp(struct fc_lport *lport, + struct fc_exch_mgr *mp, + struct fc_frame *fp) { struct fc_exch *ep; struct fc_frame_header *fh; - ep = mp->lp->tt.exch_get(mp->lp, fp); + ep = fc_exch_alloc(lport, fp); if (ep) { ep->class = fc_frame_class(fp); @@ -648,7 +675,7 @@ static struct fc_exch *fc_exch_resp(struct fc_exch_mgr *mp, struct fc_frame *fp) ep->esb_stat &= ~ESB_ST_SEQ_INIT; fc_exch_hold(ep); /* hold for caller */ - spin_unlock_bh(&ep->ex_lock); /* lock from exch_get */ + spin_unlock_bh(&ep->ex_lock); /* lock from fc_exch_alloc */ } return ep; } @@ -658,7 +685,8 @@ static struct fc_exch *fc_exch_resp(struct fc_exch_mgr *mp, struct fc_frame *fp) * If fc_pf_rjt_reason is FC_RJT_NONE then this function will have a hold * on the ep that should be released by the caller. */ -static enum fc_pf_rjt_reason fc_seq_lookup_recip(struct fc_exch_mgr *mp, +static enum fc_pf_rjt_reason fc_seq_lookup_recip(struct fc_lport *lport, + struct fc_exch_mgr *mp, struct fc_frame *fp) { struct fc_frame_header *fh = fc_frame_header_get(fp); @@ -712,7 +740,7 @@ static enum fc_pf_rjt_reason fc_seq_lookup_recip(struct fc_exch_mgr *mp, reject = FC_RJT_RX_ID; goto rel; } - ep = fc_exch_resp(mp, fp); + ep = fc_exch_resp(lport, mp, fp); if (!ep) { reject = FC_RJT_EXCH_EST; /* XXX */ goto out; @@ -1103,7 +1131,7 @@ static void fc_exch_recv_req(struct fc_lport *lp, struct fc_exch_mgr *mp, enum fc_pf_rjt_reason reject; fr_seq(fp) = NULL; - reject = fc_seq_lookup_recip(mp, fp); + reject = fc_seq_lookup_recip(lp, mp, fp); if (reject == FC_RJT_NONE) { sp = fr_seq(fp); /* sequence will be held */ ep = fc_seq_exch(sp); @@ -1467,29 +1495,34 @@ void fc_exch_mgr_reset(struct fc_lport *lp, u32 sid, u32 did) { struct fc_exch *ep; struct fc_exch *next; - struct fc_exch_mgr *mp = lp->emp; + struct fc_exch_mgr *mp; + struct fc_exch_mgr_anchor *ema; - spin_lock_bh(&mp->em_lock); + list_for_each_entry(ema, &lp->ema_list, ema_list) { + mp = ema->mp; + spin_lock_bh(&mp->em_lock); restart: - list_for_each_entry_safe(ep, next, &mp->ex_list, ex_list) { - if ((sid == 0 || sid == ep->sid) && - (did == 0 || did == ep->did)) { - fc_exch_hold(ep); - spin_unlock_bh(&mp->em_lock); - - fc_exch_reset(ep); - - fc_exch_release(ep); - spin_lock_bh(&mp->em_lock); - - /* - * must restart loop incase while lock was down - * multiple eps were released. - */ - goto restart; + list_for_each_entry_safe(ep, next, &mp->ex_list, ex_list) { + if ((lp == ep->lp) && + (sid == 0 || sid == ep->sid) && + (did == 0 || did == ep->did)) { + fc_exch_hold(ep); + spin_unlock_bh(&mp->em_lock); + + fc_exch_reset(ep); + + fc_exch_release(ep); + spin_lock_bh(&mp->em_lock); + + /* + * must restart loop incase while lock + * was down multiple eps were released. + */ + goto restart; + } } + spin_unlock_bh(&mp->em_lock); } - spin_unlock_bh(&mp->em_lock); } EXPORT_SYMBOL(fc_exch_mgr_reset); @@ -1778,7 +1811,8 @@ EXPORT_SYMBOL(fc_exch_mgr_del); struct fc_exch_mgr *fc_exch_mgr_alloc(struct fc_lport *lp, enum fc_class class, - u16 min_xid, u16 max_xid) + u16 min_xid, u16 max_xid, + bool (*match)(struct fc_frame *)) { struct fc_exch_mgr *mp; size_t len; @@ -1803,7 +1837,6 @@ struct fc_exch_mgr *fc_exch_mgr_alloc(struct fc_lport *lp, mp->class = class; mp->total_exches = 0; mp->exches = (struct fc_exch **)(mp + 1); - mp->lp = lp; /* adjust em exch xid range for offload */ mp->min_xid = min_xid; mp->max_xid = max_xid; @@ -1826,6 +1859,18 @@ struct fc_exch_mgr *fc_exch_mgr_alloc(struct fc_lport *lp, if (!mp->ep_pool) goto free_mp; + kref_init(&mp->kref); + if (!fc_exch_mgr_add(lp, mp, match)) { + mempool_destroy(mp->ep_pool); + goto free_mp; + } + + /* + * Above kref_init() sets mp->kref to 1 and then + * call to fc_exch_mgr_add incremented mp->kref again, + * so adjust that extra increment. + */ + kref_put(&mp->kref, fc_exch_mgr_destroy); return mp; free_mp: @@ -1834,27 +1879,15 @@ free_mp: } EXPORT_SYMBOL(fc_exch_mgr_alloc); -void fc_exch_mgr_free(struct fc_exch_mgr *mp) +void fc_exch_mgr_free(struct fc_lport *lport) { - WARN_ON(!mp); - /* - * The total exch count must be zero - * before freeing exchange manager. - */ - WARN_ON(mp->total_exches != 0); - mempool_destroy(mp->ep_pool); - kfree(mp); + struct fc_exch_mgr_anchor *ema, *next; + + list_for_each_entry_safe(ema, next, &lport->ema_list, ema_list) + fc_exch_mgr_del(ema); } EXPORT_SYMBOL(fc_exch_mgr_free); -struct fc_exch *fc_exch_get(struct fc_lport *lp, struct fc_frame *fp) -{ - if (!lp || !lp->emp) - return NULL; - - return fc_exch_alloc(lp->emp, fp, 0); -} -EXPORT_SYMBOL(fc_exch_get); struct fc_seq *fc_exch_seq_send(struct fc_lport *lp, struct fc_frame *fp, @@ -1869,7 +1902,7 @@ struct fc_seq *fc_exch_seq_send(struct fc_lport *lp, struct fc_frame_header *fh; int rc = 1; - ep = lp->tt.exch_get(lp, fp); + ep = fc_exch_alloc(lp, fp); if (!ep) { fc_frame_free(fp); return NULL; @@ -1914,24 +1947,44 @@ EXPORT_SYMBOL(fc_exch_seq_send); /* * Receive a frame */ -void fc_exch_recv(struct fc_lport *lp, struct fc_exch_mgr *mp, - struct fc_frame *fp) +void fc_exch_recv(struct fc_lport *lp, struct fc_frame *fp) { struct fc_frame_header *fh = fc_frame_header_get(fp); - u32 f_ctl; + struct fc_exch_mgr_anchor *ema; + u32 f_ctl, found = 0; + u16 oxid; /* lport lock ? */ - if (!lp || !mp || lp->state == LPORT_ST_DISABLED) { + if (!lp || lp->state == LPORT_ST_DISABLED) { FC_LPORT_DBG(lp, "Receiving frames for an lport that " "has not been initialized correctly\n"); fc_frame_free(fp); return; } + f_ctl = ntoh24(fh->fh_f_ctl); + oxid = ntohs(fh->fh_ox_id); + if (f_ctl & FC_FC_EX_CTX) { + list_for_each_entry(ema, &lp->ema_list, ema_list) { + if ((oxid >= ema->mp->min_xid) && + (oxid <= ema->mp->max_xid)) { + found = 1; + break; + } + } + + if (!found) { + FC_LPORT_DBG(lp, "Received response for out " + "of range oxid:%hx\n", oxid); + fc_frame_free(fp); + return; + } + } else + ema = list_entry(lp->ema_list.prev, typeof(*ema), ema_list); + /* * If frame is marked invalid, just drop it. */ - f_ctl = ntoh24(fh->fh_f_ctl); switch (fr_eof(fp)) { case FC_EOF_T: if (f_ctl & FC_FC_END_SEQ) @@ -1939,34 +1992,24 @@ void fc_exch_recv(struct fc_lport *lp, struct fc_exch_mgr *mp, /* fall through */ case FC_EOF_N: if (fh->fh_type == FC_TYPE_BLS) - fc_exch_recv_bls(mp, fp); + fc_exch_recv_bls(ema->mp, fp); else if ((f_ctl & (FC_FC_EX_CTX | FC_FC_SEQ_CTX)) == FC_FC_EX_CTX) - fc_exch_recv_seq_resp(mp, fp); + fc_exch_recv_seq_resp(ema->mp, fp); else if (f_ctl & FC_FC_SEQ_CTX) - fc_exch_recv_resp(mp, fp); + fc_exch_recv_resp(ema->mp, fp); else - fc_exch_recv_req(lp, mp, fp); + fc_exch_recv_req(lp, ema->mp, fp); break; default: FC_LPORT_DBG(lp, "dropping invalid frame (eof %x)", fr_eof(fp)); fc_frame_free(fp); - break; } } EXPORT_SYMBOL(fc_exch_recv); int fc_exch_init(struct fc_lport *lp) { - if (!lp->tt.exch_get) { - /* - * exch_put() should be NULL if - * exch_get() is NULL - */ - WARN_ON(lp->tt.exch_put); - lp->tt.exch_get = fc_exch_get; - } - if (!lp->tt.seq_start_next) lp->tt.seq_start_next = fc_seq_start_next; diff --git a/include/scsi/libfc.h b/include/scsi/libfc.h index f1bde91f98a2..c2b928cfafb9 100644 --- a/include/scsi/libfc.h +++ b/include/scsi/libfc.h @@ -517,25 +517,6 @@ struct libfc_function_template { */ void (*exch_done)(struct fc_seq *sp); - /* - * Assigns a EM and a free XID for an new exchange and then - * allocates a new exchange and sequence pair. - * The fp can be used to determine free XID. - * - * STATUS: OPTIONAL - */ - struct fc_exch *(*exch_get)(struct fc_lport *lp, struct fc_frame *fp); - - /* - * Release previously assigned XID by exch_get API. - * The LLD may implement this if XID is assigned by LLD - * in exch_get(). - * - * STATUS: OPTIONAL - */ - void (*exch_put)(struct fc_lport *lp, struct fc_exch_mgr *mp, - u16 ex_id); - /* * Start a new sequence on the same exchange/sequence tuple. * @@ -703,7 +684,6 @@ struct fc_lport { /* Associations */ struct Scsi_Host *host; - struct fc_exch_mgr *emp; struct list_head ema_list; struct fc_rport *dns_rp; struct fc_rport *ptp_rp; @@ -996,27 +976,25 @@ void fc_exch_mgr_del(struct fc_exch_mgr_anchor *ema); * a new exchange. * The LLD may choose to have multiple EMs, * e.g. one EM instance per CPU receive thread in LLD. - * The LLD can use exch_get() of struct libfc_function_template - * to specify XID for a new exchange within - * a specified EM instance. * - * The em_idx to uniquely identify an EM instance. + * Specified match function is used in allocating exchanges + * from newly allocated EM. */ struct fc_exch_mgr *fc_exch_mgr_alloc(struct fc_lport *lp, enum fc_class class, u16 min_xid, - u16 max_xid); + u16 max_xid, + bool (*match)(struct fc_frame *)); /* - * Free an exchange manager. + * Free all exchange managers of a lport. */ -void fc_exch_mgr_free(struct fc_exch_mgr *mp); +void fc_exch_mgr_free(struct fc_lport *lport); /* * Receive a frame on specified local port and exchange manager. */ -void fc_exch_recv(struct fc_lport *lp, struct fc_exch_mgr *mp, - struct fc_frame *fp); +void fc_exch_recv(struct fc_lport *lp, struct fc_frame *fp); /* * This function is for exch_seq_send function pointer in @@ -1057,20 +1035,10 @@ int fc_seq_exch_abort(const struct fc_seq *req_sp, unsigned int timer_msec); */ void fc_exch_done(struct fc_seq *sp); -/* - * Assigns a EM and XID for a frame and then allocates - * a new exchange and sequence pair. - * The fp can be used to determine free XID. - */ -struct fc_exch *fc_exch_get(struct fc_lport *lp, struct fc_frame *fp); - /* * Allocate a new exchange and sequence pair. - * if ex_id is zero then next free exchange id - * from specified exchange manger mp will be assigned. */ -struct fc_exch *fc_exch_alloc(struct fc_exch_mgr *mp, - struct fc_frame *fp, u16 ex_id); +struct fc_exch *fc_exch_alloc(struct fc_lport *lport, struct fc_frame *fp); /* * Start a new sequence on the same exchange as the supplied sequence. */ -- cgit v1.2.3 From 537029f8e950776951ca2a3fe30121d5c05643d1 Mon Sep 17 00:00:00 2001 From: Yi Zou Date: Wed, 29 Jul 2009 17:05:32 -0700 Subject: [SCSI] libfc: Remove FC_FRAME_SG_LEN in fc_fcp_send_data FC_FRAME_SG_LEN is 4 which is too small when offload is enabled. Actually, the WARN_ON() in fc_fcp_send_data() should be: WARN_ON(skb_shinfo(fp_skb(fp))->nr_frags > MAX_SKB_FRAGS); But since we will not get anything more than 64K anyway, so there is no need to do this anyway here. Therefore, I am getting rid of FC_FRAME_SG_LEN here and the WARN_ON here. Signed-off-by: Yi Zou Signed-off-by: Robert Love Signed-off-by: James Bottomley Signed-off-by: James Bottomley --- drivers/scsi/libfc/fc_fcp.c | 2 -- include/scsi/fc_frame.h | 7 ------- 2 files changed, 9 deletions(-) (limited to 'include') diff --git a/drivers/scsi/libfc/fc_fcp.c b/drivers/scsi/libfc/fc_fcp.c index e303e0d12c4b..2069edf80268 100644 --- a/drivers/scsi/libfc/fc_fcp.c +++ b/drivers/scsi/libfc/fc_fcp.c @@ -569,8 +569,6 @@ static int fc_fcp_send_data(struct fc_fcp_pkt *fsp, struct fc_seq *seq, } sg_bytes = min(tlen, sg->length - offset); if (using_sg) { - WARN_ON(skb_shinfo(fp_skb(fp))->nr_frags > - FC_FRAME_SG_LEN); get_page(sg_page(sg)); skb_fill_page_desc(fp_skb(fp), skb_shinfo(fp_skb(fp))->nr_frags, diff --git a/include/scsi/fc_frame.h b/include/scsi/fc_frame.h index 59511057cee0..c35d2383cc26 100644 --- a/include/scsi/fc_frame.h +++ b/include/scsi/fc_frame.h @@ -37,13 +37,6 @@ #define FC_FRAME_HEADROOM 32 /* headroom for VLAN + FCoE headers */ #define FC_FRAME_TAILROOM 8 /* trailer space for FCoE */ -/* - * Information about an individual fibre channel frame received or to be sent. - * The buffer may be in up to 4 additional non-contiguous sections, - * but the linear section must hold the frame header. - */ -#define FC_FRAME_SG_LEN 4 /* scatter/gather list maximum length */ - #define fp_skb(fp) (&((fp)->skb)) #define fr_hdr(fp) ((fp)->skb.data) #define fr_len(fp) ((fp)->skb.len) -- cgit v1.2.3 From 163f52b6cf3a639df6a72c7937e0eb88b20f1ef3 Mon Sep 17 00:00:00 2001 From: James Bottomley Date: Sat, 1 Aug 2009 00:39:36 +0000 Subject: [SCSI] ses: fix hotplug with multiple devices and expanders In a situation either with expanders or with multiple enclosure devices, hot add doesn't always work. This is because we try to find a single enclosure device attached to the host. Fix this by looping over all enclosure devices attached to the host and also by making the find loop recognise that the enclosure devices may be expander remote (i.e. not parented by the host). Signed-off-by: James Bottomley Signed-off-by: James Bottomley --- drivers/misc/enclosure.c | 44 ++++++++++++++++++++++++++++++++------------ drivers/scsi/ses.c | 10 ++++++---- include/linux/enclosure.h | 3 ++- 3 files changed, 40 insertions(+), 17 deletions(-) (limited to 'include') diff --git a/drivers/misc/enclosure.c b/drivers/misc/enclosure.c index 348443bdb23b..789d12128c24 100644 --- a/drivers/misc/enclosure.c +++ b/drivers/misc/enclosure.c @@ -33,24 +33,44 @@ static DEFINE_MUTEX(container_list_lock); static struct class enclosure_class; /** - * enclosure_find - find an enclosure given a device - * @dev: the device to find for + * enclosure_find - find an enclosure given a parent device + * @dev: the parent to match against + * @start: Optional enclosure device to start from (NULL if none) * - * Looks through the list of registered enclosures to see - * if it can find a match for a device. Returns NULL if no - * enclosure is found. Obtains a reference to the enclosure class - * device which must be released with device_put(). + * Looks through the list of registered enclosures to find all those + * with @dev as a parent. Returns NULL if no enclosure is + * found. @start can be used as a starting point to obtain multiple + * enclosures per parent (should begin with NULL and then be set to + * each returned enclosure device). Obtains a reference to the + * enclosure class device which must be released with device_put(). + * If @start is not NULL, a reference must be taken on it which is + * released before returning (this allows a loop through all + * enclosures to exit with only the reference on the enclosure of + * interest held). Note that the @dev may correspond to the actual + * device housing the enclosure, in which case no iteration via @start + * is required. */ -struct enclosure_device *enclosure_find(struct device *dev) +struct enclosure_device *enclosure_find(struct device *dev, + struct enclosure_device *start) { struct enclosure_device *edev; mutex_lock(&container_list_lock); - list_for_each_entry(edev, &container_list, node) { - if (edev->edev.parent == dev) { - get_device(&edev->edev); - mutex_unlock(&container_list_lock); - return edev; + edev = list_prepare_entry(start, &container_list, node); + if (start) + put_device(&start->edev); + + list_for_each_entry_continue(edev, &container_list, node) { + struct device *parent = edev->edev.parent; + /* parent might not be immediate, so iterate up to + * the root of the tree if necessary */ + while (parent) { + if (parent == dev) { + get_device(&edev->edev); + mutex_unlock(&container_list_lock); + return edev; + } + parent = parent->parent; } } mutex_unlock(&container_list_lock); diff --git a/drivers/scsi/ses.c b/drivers/scsi/ses.c index 4f618f487356..e1b8c828f03a 100644 --- a/drivers/scsi/ses.c +++ b/drivers/scsi/ses.c @@ -413,10 +413,11 @@ static int ses_intf_add(struct device *cdev, if (!scsi_device_enclosure(sdev)) { /* not an enclosure, but might be in one */ - edev = enclosure_find(&sdev->host->shost_gendev); - if (edev) { + struct enclosure_device *prev = NULL; + + while ((edev = enclosure_find(&sdev->host->shost_gendev, prev)) != NULL) { ses_match_to_enclosure(edev, sdev); - put_device(&edev->edev); + prev = edev; } return -ENODEV; } @@ -625,7 +626,8 @@ static void ses_intf_remove(struct device *cdev, if (!scsi_device_enclosure(sdev)) return; - edev = enclosure_find(cdev->parent); + /* exact match to this enclosure */ + edev = enclosure_find(cdev->parent, NULL); if (!edev) return; diff --git a/include/linux/enclosure.h b/include/linux/enclosure.h index 4332442b1b57..d77811e9ed84 100644 --- a/include/linux/enclosure.h +++ b/include/linux/enclosure.h @@ -123,7 +123,8 @@ enclosure_component_register(struct enclosure_device *, unsigned int, int enclosure_add_device(struct enclosure_device *enclosure, int component, struct device *dev); int enclosure_remove_device(struct enclosure_device *enclosure, int component); -struct enclosure_device *enclosure_find(struct device *dev); +struct enclosure_device *enclosure_find(struct device *dev, + struct enclosure_device *start); int enclosure_for_each_device(int (*fn)(struct enclosure_device *, void *), void *data); -- cgit v1.2.3 From 43d8eb9cfd0aea93be32181c64e18191b69c211c Mon Sep 17 00:00:00 2001 From: James Bottomley Date: Sat, 1 Aug 2009 00:41:22 +0000 Subject: [SCSI] ses: add support for enclosure component hot removal Right at the moment, hot removal of a device within an enclosure does nothing (because the intf_remove only copes with enclosure removal not with component removal). Fix this by adding a function to remove the component. Also needed to fix the prototype of enclosure_remove_device, since we know the device we've removed but not the internal component number Signed-off-by: James Bottomley Signed-off-by: James Bottomley --- drivers/misc/enclosure.c | 22 ++++++++++++++-------- drivers/scsi/ses.c | 33 ++++++++++++++++++++++++++------- include/linux/enclosure.h | 2 +- 3 files changed, 41 insertions(+), 16 deletions(-) (limited to 'include') diff --git a/drivers/misc/enclosure.c b/drivers/misc/enclosure.c index 789d12128c24..850706a5e553 100644 --- a/drivers/misc/enclosure.c +++ b/drivers/misc/enclosure.c @@ -332,19 +332,25 @@ EXPORT_SYMBOL_GPL(enclosure_add_device); * Returns zero on success or an error. * */ -int enclosure_remove_device(struct enclosure_device *edev, int component) +int enclosure_remove_device(struct enclosure_device *edev, struct device *dev) { struct enclosure_component *cdev; + int i; - if (!edev || component >= edev->components) + if (!edev || !dev) return -EINVAL; - cdev = &edev->component[component]; - - device_del(&cdev->cdev); - put_device(cdev->dev); - cdev->dev = NULL; - return device_add(&cdev->cdev); + for (i = 0; i < edev->components; i++) { + cdev = &edev->component[i]; + if (cdev->dev == dev) { + enclosure_remove_links(cdev); + device_del(&cdev->cdev); + put_device(dev); + cdev->dev = NULL; + return device_add(&cdev->cdev); + } + } + return -ENODEV; } EXPORT_SYMBOL_GPL(enclosure_remove_device); diff --git a/drivers/scsi/ses.c b/drivers/scsi/ses.c index e1b8c828f03a..be593c8525b5 100644 --- a/drivers/scsi/ses.c +++ b/drivers/scsi/ses.c @@ -616,18 +616,26 @@ static int ses_remove(struct device *dev) return 0; } -static void ses_intf_remove(struct device *cdev, - struct class_interface *intf) +static void ses_intf_remove_component(struct scsi_device *sdev) +{ + struct enclosure_device *edev, *prev = NULL; + + while ((edev = enclosure_find(&sdev->host->shost_gendev, prev)) != NULL) { + prev = edev; + if (!enclosure_remove_device(edev, &sdev->sdev_gendev)) + break; + } + if (edev) + put_device(&edev->edev); +} + +static void ses_intf_remove_enclosure(struct scsi_device *sdev) { - struct scsi_device *sdev = to_scsi_device(cdev->parent); struct enclosure_device *edev; struct ses_device *ses_dev; - if (!scsi_device_enclosure(sdev)) - return; - /* exact match to this enclosure */ - edev = enclosure_find(cdev->parent, NULL); + edev = enclosure_find(&sdev->sdev_gendev, NULL); if (!edev) return; @@ -645,6 +653,17 @@ static void ses_intf_remove(struct device *cdev, enclosure_unregister(edev); } +static void ses_intf_remove(struct device *cdev, + struct class_interface *intf) +{ + struct scsi_device *sdev = to_scsi_device(cdev->parent); + + if (!scsi_device_enclosure(sdev)) + ses_intf_remove_component(sdev); + else + ses_intf_remove_enclosure(sdev); +} + static struct class_interface ses_interface = { .add_dev = ses_intf_add, .remove_dev = ses_intf_remove, diff --git a/include/linux/enclosure.h b/include/linux/enclosure.h index d77811e9ed84..90d1c2184112 100644 --- a/include/linux/enclosure.h +++ b/include/linux/enclosure.h @@ -122,7 +122,7 @@ enclosure_component_register(struct enclosure_device *, unsigned int, enum enclosure_component_type, const char *); int enclosure_add_device(struct enclosure_device *enclosure, int component, struct device *dev); -int enclosure_remove_device(struct enclosure_device *enclosure, int component); +int enclosure_remove_device(struct enclosure_device *, struct device *); struct enclosure_device *enclosure_find(struct device *dev, struct enclosure_device *start); int enclosure_for_each_device(int (*fn)(struct enclosure_device *, void *), -- cgit v1.2.3 From 18ee70c9d7b2dcd312a1f8c6536841e7c0fea5ca Mon Sep 17 00:00:00 2001 From: Chandra Seetharaman Date: Mon, 3 Aug 2009 12:42:33 -0700 Subject: [SCSI] scsi_dh: add the interface scsi_dh_set_params() When we moved the device handler functionality from dm layer to SCSI layer we dropped the parameter functionality. This path adds an interface to scsi dh layer to set device handler parameters. Basically, multipath layer need to create a string with all the parameters and call scsi_dh_set_params() after it called scsi_dh_attach() on a device. If a device handler provides such an interface it will handle the parameters as it expects them. Reported-by: Eddie Williams Signed-off-by: Chandra Seetharaman Tested-by: Eddie Williams Signed-off-by: James Bottomley Signed-off-by: James Bottomley --- drivers/scsi/device_handler/scsi_dh.c | 33 +++++++++++++++++++++++++++++++++ include/scsi/scsi_device.h | 1 + include/scsi/scsi_dh.h | 5 +++++ 3 files changed, 39 insertions(+) (limited to 'include') diff --git a/drivers/scsi/device_handler/scsi_dh.c b/drivers/scsi/device_handler/scsi_dh.c index 53a7385e1b4d..3ee1cbc89479 100644 --- a/drivers/scsi/device_handler/scsi_dh.c +++ b/drivers/scsi/device_handler/scsi_dh.c @@ -451,6 +451,39 @@ int scsi_dh_activate(struct request_queue *q) } EXPORT_SYMBOL_GPL(scsi_dh_activate); +/* + * scsi_dh_set_params - set the parameters for the device as per the + * string specified in params. + * @q - Request queue that is associated with the scsi_device for + * which the parameters to be set. + * @params - parameters in the following format + * "no_of_params\0param1\0param2\0param3\0...\0" + * for example, string for 2 parameters with value 10 and 21 + * is specified as "2\010\021\0". + */ +int scsi_dh_set_params(struct request_queue *q, const char *params) +{ + int err = -SCSI_DH_NOSYS; + unsigned long flags; + struct scsi_device *sdev; + struct scsi_device_handler *scsi_dh = NULL; + + spin_lock_irqsave(q->queue_lock, flags); + sdev = q->queuedata; + if (sdev && sdev->scsi_dh_data) + scsi_dh = sdev->scsi_dh_data->scsi_dh; + if (scsi_dh && scsi_dh->set_params && get_device(&sdev->sdev_gendev)) + err = 0; + spin_unlock_irqrestore(q->queue_lock, flags); + + if (err) + return err; + err = scsi_dh->set_params(sdev, params); + put_device(&sdev->sdev_gendev); + return err; +} +EXPORT_SYMBOL_GPL(scsi_dh_set_params); + /* * scsi_dh_handler_exist - Return TRUE(1) if a device handler exists for * the given name. FALSE(0) otherwise. diff --git a/include/scsi/scsi_device.h b/include/scsi/scsi_device.h index 1f3a4c8044c0..9af48cbf0036 100644 --- a/include/scsi/scsi_device.h +++ b/include/scsi/scsi_device.h @@ -187,6 +187,7 @@ struct scsi_device_handler { void (*detach)(struct scsi_device *); int (*activate)(struct scsi_device *); int (*prep_fn)(struct scsi_device *, struct request *); + int (*set_params)(struct scsi_device *, const char *); }; struct scsi_dh_data { diff --git a/include/scsi/scsi_dh.h b/include/scsi/scsi_dh.h index 33efce20c26c..ff2407405b42 100644 --- a/include/scsi/scsi_dh.h +++ b/include/scsi/scsi_dh.h @@ -60,6 +60,7 @@ extern int scsi_dh_activate(struct request_queue *); extern int scsi_dh_handler_exist(const char *); extern int scsi_dh_attach(struct request_queue *, const char *); extern void scsi_dh_detach(struct request_queue *); +extern int scsi_dh_set_params(struct request_queue *, const char *); #else static inline int scsi_dh_activate(struct request_queue *req) { @@ -77,4 +78,8 @@ static inline void scsi_dh_detach(struct request_queue *q) { return; } +static inline int scsi_dh_set_params(struct request_queue *req, const char *params) +{ + return -SCSI_DH_NOSYS; +} #endif -- cgit v1.2.3 From ee5f9757ea17759e1ce5503bdae2b07e48e32af9 Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Fri, 21 Aug 2009 16:33:34 -0700 Subject: pkt_sched: Convert qdisc_watchdog to tasklet_hrtimer None of this stuff should execute in hw IRQ context, therefore use a tasklet_hrtimer so that it runs in softirq context. Signed-off-by: David S. Miller Acked-by: Thomas Gleixner --- include/net/pkt_sched.h | 4 ++-- net/sched/sch_api.c | 8 ++++---- 2 files changed, 6 insertions(+), 6 deletions(-) (limited to 'include') diff --git a/include/net/pkt_sched.h b/include/net/pkt_sched.h index 82a3191375f5..7eafb8d54470 100644 --- a/include/net/pkt_sched.h +++ b/include/net/pkt_sched.h @@ -61,8 +61,8 @@ psched_tdiff_bounded(psched_time_t tv1, psched_time_t tv2, psched_time_t bound) } struct qdisc_watchdog { - struct hrtimer timer; - struct Qdisc *qdisc; + struct tasklet_hrtimer timer; + struct Qdisc *qdisc; }; extern void qdisc_watchdog_init(struct qdisc_watchdog *wd, struct Qdisc *qdisc); diff --git a/net/sched/sch_api.c b/net/sched/sch_api.c index 24d17ce9c294..e1c2bf7e9ba4 100644 --- a/net/sched/sch_api.c +++ b/net/sched/sch_api.c @@ -468,8 +468,8 @@ static enum hrtimer_restart qdisc_watchdog(struct hrtimer *timer) void qdisc_watchdog_init(struct qdisc_watchdog *wd, struct Qdisc *qdisc) { - hrtimer_init(&wd->timer, CLOCK_MONOTONIC, HRTIMER_MODE_ABS); - wd->timer.function = qdisc_watchdog; + tasklet_hrtimer_init(&wd->timer, qdisc_watchdog, + CLOCK_MONOTONIC, HRTIMER_MODE_ABS); wd->qdisc = qdisc; } EXPORT_SYMBOL(qdisc_watchdog_init); @@ -485,13 +485,13 @@ void qdisc_watchdog_schedule(struct qdisc_watchdog *wd, psched_time_t expires) wd->qdisc->flags |= TCQ_F_THROTTLED; time = ktime_set(0, 0); time = ktime_add_ns(time, PSCHED_TICKS2NS(expires)); - hrtimer_start(&wd->timer, time, HRTIMER_MODE_ABS); + tasklet_hrtimer_start(&wd->timer, time, HRTIMER_MODE_ABS); } EXPORT_SYMBOL(qdisc_watchdog_schedule); void qdisc_watchdog_cancel(struct qdisc_watchdog *wd) { - hrtimer_cancel(&wd->timer); + tasklet_hrtimer_cancel(&wd->timer); wd->qdisc->flags &= ~TCQ_F_THROTTLED; } EXPORT_SYMBOL(qdisc_watchdog_cancel); -- cgit v1.2.3 From 9f77da9f40045253e91f55c12d4481254b513d2d Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Sat, 22 Aug 2009 13:56:45 -0700 Subject: rcu: Move private definitions from include/linux/rcutree.h to kernel/rcutree.h Some information hiding that makes it easier to merge preemptability into rcutree without descending into #include hell. Signed-off-by: Paul E. McKenney Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <1250974613373-git-send-email-> Signed-off-by: Ingo Molnar --- include/linux/rcutree.h | 211 ------------------------------------------ kernel/rcutree.c | 2 + kernel/rcutree.h | 238 ++++++++++++++++++++++++++++++++++++++++++++++++ kernel/rcutree_trace.c | 1 + 4 files changed, 241 insertions(+), 211 deletions(-) (limited to 'include') diff --git a/include/linux/rcutree.h b/include/linux/rcutree.h index d4dfd2489633..e37d5e2a8353 100644 --- a/include/linux/rcutree.h +++ b/include/linux/rcutree.h @@ -30,217 +30,6 @@ #ifndef __LINUX_RCUTREE_H #define __LINUX_RCUTREE_H -#include -#include -#include -#include -#include - -/* - * Define shape of hierarchy based on NR_CPUS and CONFIG_RCU_FANOUT. - * In theory, it should be possible to add more levels straightforwardly. - * In practice, this has not been tested, so there is probably some - * bug somewhere. - */ -#define MAX_RCU_LVLS 3 -#define RCU_FANOUT (CONFIG_RCU_FANOUT) -#define RCU_FANOUT_SQ (RCU_FANOUT * RCU_FANOUT) -#define RCU_FANOUT_CUBE (RCU_FANOUT_SQ * RCU_FANOUT) - -#if NR_CPUS <= RCU_FANOUT -# define NUM_RCU_LVLS 1 -# define NUM_RCU_LVL_0 1 -# define NUM_RCU_LVL_1 (NR_CPUS) -# define NUM_RCU_LVL_2 0 -# define NUM_RCU_LVL_3 0 -#elif NR_CPUS <= RCU_FANOUT_SQ -# define NUM_RCU_LVLS 2 -# define NUM_RCU_LVL_0 1 -# define NUM_RCU_LVL_1 (((NR_CPUS) + RCU_FANOUT - 1) / RCU_FANOUT) -# define NUM_RCU_LVL_2 (NR_CPUS) -# define NUM_RCU_LVL_3 0 -#elif NR_CPUS <= RCU_FANOUT_CUBE -# define NUM_RCU_LVLS 3 -# define NUM_RCU_LVL_0 1 -# define NUM_RCU_LVL_1 (((NR_CPUS) + RCU_FANOUT_SQ - 1) / RCU_FANOUT_SQ) -# define NUM_RCU_LVL_2 (((NR_CPUS) + (RCU_FANOUT) - 1) / (RCU_FANOUT)) -# define NUM_RCU_LVL_3 NR_CPUS -#else -# error "CONFIG_RCU_FANOUT insufficient for NR_CPUS" -#endif /* #if (NR_CPUS) <= RCU_FANOUT */ - -#define RCU_SUM (NUM_RCU_LVL_0 + NUM_RCU_LVL_1 + NUM_RCU_LVL_2 + NUM_RCU_LVL_3) -#define NUM_RCU_NODES (RCU_SUM - NR_CPUS) - -/* - * Dynticks per-CPU state. - */ -struct rcu_dynticks { - int dynticks_nesting; /* Track nesting level, sort of. */ - int dynticks; /* Even value for dynticks-idle, else odd. */ - int dynticks_nmi; /* Even value for either dynticks-idle or */ - /* not in nmi handler, else odd. So this */ - /* remains even for nmi from irq handler. */ -}; - -/* - * Definition for node within the RCU grace-period-detection hierarchy. - */ -struct rcu_node { - spinlock_t lock; - unsigned long qsmask; /* CPUs or groups that need to switch in */ - /* order for current grace period to proceed.*/ - unsigned long qsmaskinit; - /* Per-GP initialization for qsmask. */ - unsigned long grpmask; /* Mask to apply to parent qsmask. */ - int grplo; /* lowest-numbered CPU or group here. */ - int grphi; /* highest-numbered CPU or group here. */ - u8 grpnum; /* CPU/group number for next level up. */ - u8 level; /* root is at level 0. */ - struct rcu_node *parent; -} ____cacheline_internodealigned_in_smp; - -/* Index values for nxttail array in struct rcu_data. */ -#define RCU_DONE_TAIL 0 /* Also RCU_WAIT head. */ -#define RCU_WAIT_TAIL 1 /* Also RCU_NEXT_READY head. */ -#define RCU_NEXT_READY_TAIL 2 /* Also RCU_NEXT head. */ -#define RCU_NEXT_TAIL 3 -#define RCU_NEXT_SIZE 4 - -/* Per-CPU data for read-copy update. */ -struct rcu_data { - /* 1) quiescent-state and grace-period handling : */ - long completed; /* Track rsp->completed gp number */ - /* in order to detect GP end. */ - long gpnum; /* Highest gp number that this CPU */ - /* is aware of having started. */ - long passed_quiesc_completed; - /* Value of completed at time of qs. */ - bool passed_quiesc; /* User-mode/idle loop etc. */ - bool qs_pending; /* Core waits for quiesc state. */ - bool beenonline; /* CPU online at least once. */ - struct rcu_node *mynode; /* This CPU's leaf of hierarchy */ - unsigned long grpmask; /* Mask to apply to leaf qsmask. */ - - /* 2) batch handling */ - /* - * If nxtlist is not NULL, it is partitioned as follows. - * Any of the partitions might be empty, in which case the - * pointer to that partition will be equal to the pointer for - * the following partition. When the list is empty, all of - * the nxttail elements point to nxtlist, which is NULL. - * - * [*nxttail[RCU_NEXT_READY_TAIL], NULL = *nxttail[RCU_NEXT_TAIL]): - * Entries that might have arrived after current GP ended - * [*nxttail[RCU_WAIT_TAIL], *nxttail[RCU_NEXT_READY_TAIL]): - * Entries known to have arrived before current GP ended - * [*nxttail[RCU_DONE_TAIL], *nxttail[RCU_WAIT_TAIL]): - * Entries that batch # <= ->completed - 1: waiting for current GP - * [nxtlist, *nxttail[RCU_DONE_TAIL]): - * Entries that batch # <= ->completed - * The grace period for these entries has completed, and - * the other grace-period-completed entries may be moved - * here temporarily in rcu_process_callbacks(). - */ - struct rcu_head *nxtlist; - struct rcu_head **nxttail[RCU_NEXT_SIZE]; - long qlen; /* # of queued callbacks */ - long blimit; /* Upper limit on a processed batch */ - -#ifdef CONFIG_NO_HZ - /* 3) dynticks interface. */ - struct rcu_dynticks *dynticks; /* Shared per-CPU dynticks state. */ - int dynticks_snap; /* Per-GP tracking for dynticks. */ - int dynticks_nmi_snap; /* Per-GP tracking for dynticks_nmi. */ -#endif /* #ifdef CONFIG_NO_HZ */ - - /* 4) reasons this CPU needed to be kicked by force_quiescent_state */ -#ifdef CONFIG_NO_HZ - unsigned long dynticks_fqs; /* Kicked due to dynticks idle. */ -#endif /* #ifdef CONFIG_NO_HZ */ - unsigned long offline_fqs; /* Kicked due to being offline. */ - unsigned long resched_ipi; /* Sent a resched IPI. */ - - /* 5) __rcu_pending() statistics. */ - long n_rcu_pending; /* rcu_pending() calls since boot. */ - long n_rp_qs_pending; - long n_rp_cb_ready; - long n_rp_cpu_needs_gp; - long n_rp_gp_completed; - long n_rp_gp_started; - long n_rp_need_fqs; - long n_rp_need_nothing; - - int cpu; -}; - -/* Values for signaled field in struct rcu_state. */ -#define RCU_GP_INIT 0 /* Grace period being initialized. */ -#define RCU_SAVE_DYNTICK 1 /* Need to scan dyntick state. */ -#define RCU_FORCE_QS 2 /* Need to force quiescent state. */ -#ifdef CONFIG_NO_HZ -#define RCU_SIGNAL_INIT RCU_SAVE_DYNTICK -#else /* #ifdef CONFIG_NO_HZ */ -#define RCU_SIGNAL_INIT RCU_FORCE_QS -#endif /* #else #ifdef CONFIG_NO_HZ */ - -#define RCU_JIFFIES_TILL_FORCE_QS 3 /* for rsp->jiffies_force_qs */ -#ifdef CONFIG_RCU_CPU_STALL_DETECTOR -#define RCU_SECONDS_TILL_STALL_CHECK (10 * HZ) /* for rsp->jiffies_stall */ -#define RCU_SECONDS_TILL_STALL_RECHECK (30 * HZ) /* for rsp->jiffies_stall */ -#define RCU_STALL_RAT_DELAY 2 /* Allow other CPUs time */ - /* to take at least one */ - /* scheduling clock irq */ - /* before ratting on them. */ - -#endif /* #ifdef CONFIG_RCU_CPU_STALL_DETECTOR */ - -/* - * RCU global state, including node hierarchy. This hierarchy is - * represented in "heap" form in a dense array. The root (first level) - * of the hierarchy is in ->node[0] (referenced by ->level[0]), the second - * level in ->node[1] through ->node[m] (->node[1] referenced by ->level[1]), - * and the third level in ->node[m+1] and following (->node[m+1] referenced - * by ->level[2]). The number of levels is determined by the number of - * CPUs and by CONFIG_RCU_FANOUT. Small systems will have a "hierarchy" - * consisting of a single rcu_node. - */ -struct rcu_state { - struct rcu_node node[NUM_RCU_NODES]; /* Hierarchy. */ - struct rcu_node *level[NUM_RCU_LVLS]; /* Hierarchy levels. */ - u32 levelcnt[MAX_RCU_LVLS + 1]; /* # nodes in each level. */ - u8 levelspread[NUM_RCU_LVLS]; /* kids/node in each level. */ - struct rcu_data *rda[NR_CPUS]; /* array of rdp pointers. */ - - /* The following fields are guarded by the root rcu_node's lock. */ - - u8 signaled ____cacheline_internodealigned_in_smp; - /* Force QS state. */ - long gpnum; /* Current gp number. */ - long completed; /* # of last completed gp. */ - spinlock_t onofflock; /* exclude on/offline and */ - /* starting new GP. */ - spinlock_t fqslock; /* Only one task forcing */ - /* quiescent states. */ - unsigned long jiffies_force_qs; /* Time at which to invoke */ - /* force_quiescent_state(). */ - unsigned long n_force_qs; /* Number of calls to */ - /* force_quiescent_state(). */ - unsigned long n_force_qs_lh; /* ~Number of calls leaving */ - /* due to lock unavailable. */ - unsigned long n_force_qs_ngp; /* Number of calls leaving */ - /* due to no GP active. */ -#ifdef CONFIG_RCU_CPU_STALL_DETECTOR - unsigned long gp_start; /* Time at which GP started, */ - /* but in jiffies. */ - unsigned long jiffies_stall; /* Time at which to check */ - /* for CPU stalls. */ -#endif /* #ifdef CONFIG_RCU_CPU_STALL_DETECTOR */ -#ifdef CONFIG_NO_HZ - long dynticks_completed; /* Value of completed @ snap. */ -#endif /* #ifdef CONFIG_NO_HZ */ -}; - extern void rcu_qsctr_inc(int cpu); extern void rcu_bh_qsctr_inc(int cpu); diff --git a/kernel/rcutree.c b/kernel/rcutree.c index 75762cddbe03..a162f859dd32 100644 --- a/kernel/rcutree.c +++ b/kernel/rcutree.c @@ -46,6 +46,8 @@ #include #include +#include "rcutree.h" + #ifdef CONFIG_DEBUG_LOCK_ALLOC static struct lock_class_key rcu_lock_key; struct lockdep_map rcu_lock_map = diff --git a/kernel/rcutree.h b/kernel/rcutree.h index 5e872bbf07f5..7cc830a1c44a 100644 --- a/kernel/rcutree.h +++ b/kernel/rcutree.h @@ -1,3 +1,239 @@ +/* + * Read-Copy Update mechanism for mutual exclusion (tree-based version) + * Internal non-public definitions. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. + * + * Copyright IBM Corporation, 2008 + * + * Author: Ingo Molnar + * Paul E. McKenney + */ + +#include +#include +#include +#include +#include + +/* + * Define shape of hierarchy based on NR_CPUS and CONFIG_RCU_FANOUT. + * In theory, it should be possible to add more levels straightforwardly. + * In practice, this has not been tested, so there is probably some + * bug somewhere. + */ +#define MAX_RCU_LVLS 3 +#define RCU_FANOUT (CONFIG_RCU_FANOUT) +#define RCU_FANOUT_SQ (RCU_FANOUT * RCU_FANOUT) +#define RCU_FANOUT_CUBE (RCU_FANOUT_SQ * RCU_FANOUT) + +#if NR_CPUS <= RCU_FANOUT +# define NUM_RCU_LVLS 1 +# define NUM_RCU_LVL_0 1 +# define NUM_RCU_LVL_1 (NR_CPUS) +# define NUM_RCU_LVL_2 0 +# define NUM_RCU_LVL_3 0 +#elif NR_CPUS <= RCU_FANOUT_SQ +# define NUM_RCU_LVLS 2 +# define NUM_RCU_LVL_0 1 +# define NUM_RCU_LVL_1 (((NR_CPUS) + RCU_FANOUT - 1) / RCU_FANOUT) +# define NUM_RCU_LVL_2 (NR_CPUS) +# define NUM_RCU_LVL_3 0 +#elif NR_CPUS <= RCU_FANOUT_CUBE +# define NUM_RCU_LVLS 3 +# define NUM_RCU_LVL_0 1 +# define NUM_RCU_LVL_1 (((NR_CPUS) + RCU_FANOUT_SQ - 1) / RCU_FANOUT_SQ) +# define NUM_RCU_LVL_2 (((NR_CPUS) + (RCU_FANOUT) - 1) / (RCU_FANOUT)) +# define NUM_RCU_LVL_3 NR_CPUS +#else +# error "CONFIG_RCU_FANOUT insufficient for NR_CPUS" +#endif /* #if (NR_CPUS) <= RCU_FANOUT */ + +#define RCU_SUM (NUM_RCU_LVL_0 + NUM_RCU_LVL_1 + NUM_RCU_LVL_2 + NUM_RCU_LVL_3) +#define NUM_RCU_NODES (RCU_SUM - NR_CPUS) + +/* + * Dynticks per-CPU state. + */ +struct rcu_dynticks { + int dynticks_nesting; /* Track nesting level, sort of. */ + int dynticks; /* Even value for dynticks-idle, else odd. */ + int dynticks_nmi; /* Even value for either dynticks-idle or */ + /* not in nmi handler, else odd. So this */ + /* remains even for nmi from irq handler. */ +}; + +/* + * Definition for node within the RCU grace-period-detection hierarchy. + */ +struct rcu_node { + spinlock_t lock; + unsigned long qsmask; /* CPUs or groups that need to switch in */ + /* order for current grace period to proceed.*/ + unsigned long qsmaskinit; + /* Per-GP initialization for qsmask. */ + unsigned long grpmask; /* Mask to apply to parent qsmask. */ + int grplo; /* lowest-numbered CPU or group here. */ + int grphi; /* highest-numbered CPU or group here. */ + u8 grpnum; /* CPU/group number for next level up. */ + u8 level; /* root is at level 0. */ + struct rcu_node *parent; +} ____cacheline_internodealigned_in_smp; + +/* Index values for nxttail array in struct rcu_data. */ +#define RCU_DONE_TAIL 0 /* Also RCU_WAIT head. */ +#define RCU_WAIT_TAIL 1 /* Also RCU_NEXT_READY head. */ +#define RCU_NEXT_READY_TAIL 2 /* Also RCU_NEXT head. */ +#define RCU_NEXT_TAIL 3 +#define RCU_NEXT_SIZE 4 + +/* Per-CPU data for read-copy update. */ +struct rcu_data { + /* 1) quiescent-state and grace-period handling : */ + long completed; /* Track rsp->completed gp number */ + /* in order to detect GP end. */ + long gpnum; /* Highest gp number that this CPU */ + /* is aware of having started. */ + long passed_quiesc_completed; + /* Value of completed at time of qs. */ + bool passed_quiesc; /* User-mode/idle loop etc. */ + bool qs_pending; /* Core waits for quiesc state. */ + bool beenonline; /* CPU online at least once. */ + struct rcu_node *mynode; /* This CPU's leaf of hierarchy */ + unsigned long grpmask; /* Mask to apply to leaf qsmask. */ + + /* 2) batch handling */ + /* + * If nxtlist is not NULL, it is partitioned as follows. + * Any of the partitions might be empty, in which case the + * pointer to that partition will be equal to the pointer for + * the following partition. When the list is empty, all of + * the nxttail elements point to nxtlist, which is NULL. + * + * [*nxttail[RCU_NEXT_READY_TAIL], NULL = *nxttail[RCU_NEXT_TAIL]): + * Entries that might have arrived after current GP ended + * [*nxttail[RCU_WAIT_TAIL], *nxttail[RCU_NEXT_READY_TAIL]): + * Entries known to have arrived before current GP ended + * [*nxttail[RCU_DONE_TAIL], *nxttail[RCU_WAIT_TAIL]): + * Entries that batch # <= ->completed - 1: waiting for current GP + * [nxtlist, *nxttail[RCU_DONE_TAIL]): + * Entries that batch # <= ->completed + * The grace period for these entries has completed, and + * the other grace-period-completed entries may be moved + * here temporarily in rcu_process_callbacks(). + */ + struct rcu_head *nxtlist; + struct rcu_head **nxttail[RCU_NEXT_SIZE]; + long qlen; /* # of queued callbacks */ + long blimit; /* Upper limit on a processed batch */ + +#ifdef CONFIG_NO_HZ + /* 3) dynticks interface. */ + struct rcu_dynticks *dynticks; /* Shared per-CPU dynticks state. */ + int dynticks_snap; /* Per-GP tracking for dynticks. */ + int dynticks_nmi_snap; /* Per-GP tracking for dynticks_nmi. */ +#endif /* #ifdef CONFIG_NO_HZ */ + + /* 4) reasons this CPU needed to be kicked by force_quiescent_state */ +#ifdef CONFIG_NO_HZ + unsigned long dynticks_fqs; /* Kicked due to dynticks idle. */ +#endif /* #ifdef CONFIG_NO_HZ */ + unsigned long offline_fqs; /* Kicked due to being offline. */ + unsigned long resched_ipi; /* Sent a resched IPI. */ + + /* 5) __rcu_pending() statistics. */ + long n_rcu_pending; /* rcu_pending() calls since boot. */ + long n_rp_qs_pending; + long n_rp_cb_ready; + long n_rp_cpu_needs_gp; + long n_rp_gp_completed; + long n_rp_gp_started; + long n_rp_need_fqs; + long n_rp_need_nothing; + + int cpu; +}; + +/* Values for signaled field in struct rcu_state. */ +#define RCU_GP_INIT 0 /* Grace period being initialized. */ +#define RCU_SAVE_DYNTICK 1 /* Need to scan dyntick state. */ +#define RCU_FORCE_QS 2 /* Need to force quiescent state. */ +#ifdef CONFIG_NO_HZ +#define RCU_SIGNAL_INIT RCU_SAVE_DYNTICK +#else /* #ifdef CONFIG_NO_HZ */ +#define RCU_SIGNAL_INIT RCU_FORCE_QS +#endif /* #else #ifdef CONFIG_NO_HZ */ + +#define RCU_JIFFIES_TILL_FORCE_QS 3 /* for rsp->jiffies_force_qs */ +#ifdef CONFIG_RCU_CPU_STALL_DETECTOR +#define RCU_SECONDS_TILL_STALL_CHECK (10 * HZ) /* for rsp->jiffies_stall */ +#define RCU_SECONDS_TILL_STALL_RECHECK (30 * HZ) /* for rsp->jiffies_stall */ +#define RCU_STALL_RAT_DELAY 2 /* Allow other CPUs time */ + /* to take at least one */ + /* scheduling clock irq */ + /* before ratting on them. */ + +#endif /* #ifdef CONFIG_RCU_CPU_STALL_DETECTOR */ + +/* + * RCU global state, including node hierarchy. This hierarchy is + * represented in "heap" form in a dense array. The root (first level) + * of the hierarchy is in ->node[0] (referenced by ->level[0]), the second + * level in ->node[1] through ->node[m] (->node[1] referenced by ->level[1]), + * and the third level in ->node[m+1] and following (->node[m+1] referenced + * by ->level[2]). The number of levels is determined by the number of + * CPUs and by CONFIG_RCU_FANOUT. Small systems will have a "hierarchy" + * consisting of a single rcu_node. + */ +struct rcu_state { + struct rcu_node node[NUM_RCU_NODES]; /* Hierarchy. */ + struct rcu_node *level[NUM_RCU_LVLS]; /* Hierarchy levels. */ + u32 levelcnt[MAX_RCU_LVLS + 1]; /* # nodes in each level. */ + u8 levelspread[NUM_RCU_LVLS]; /* kids/node in each level. */ + struct rcu_data *rda[NR_CPUS]; /* array of rdp pointers. */ + + /* The following fields are guarded by the root rcu_node's lock. */ + + u8 signaled ____cacheline_internodealigned_in_smp; + /* Force QS state. */ + long gpnum; /* Current gp number. */ + long completed; /* # of last completed gp. */ + spinlock_t onofflock; /* exclude on/offline and */ + /* starting new GP. */ + spinlock_t fqslock; /* Only one task forcing */ + /* quiescent states. */ + unsigned long jiffies_force_qs; /* Time at which to invoke */ + /* force_quiescent_state(). */ + unsigned long n_force_qs; /* Number of calls to */ + /* force_quiescent_state(). */ + unsigned long n_force_qs_lh; /* ~Number of calls leaving */ + /* due to lock unavailable. */ + unsigned long n_force_qs_ngp; /* Number of calls leaving */ + /* due to no GP active. */ +#ifdef CONFIG_RCU_CPU_STALL_DETECTOR + unsigned long gp_start; /* Time at which GP started, */ + /* but in jiffies. */ + unsigned long jiffies_stall; /* Time at which to check */ + /* for CPU stalls. */ +#endif /* #ifdef CONFIG_RCU_CPU_STALL_DETECTOR */ +#ifdef CONFIG_NO_HZ + long dynticks_completed; /* Value of completed @ snap. */ +#endif /* #ifdef CONFIG_NO_HZ */ +}; + +#ifdef RCU_TREE_NONCORE /* * RCU implementation internal declarations: @@ -8,3 +244,5 @@ DECLARE_PER_CPU(struct rcu_data, rcu_data); extern struct rcu_state rcu_bh_state; DECLARE_PER_CPU(struct rcu_data, rcu_bh_data); +#endif /* #ifdef RCU_TREE_NONCORE */ + diff --git a/kernel/rcutree_trace.c b/kernel/rcutree_trace.c index fe1dcdbf1ca3..0cb52b887758 100644 --- a/kernel/rcutree_trace.c +++ b/kernel/rcutree_trace.c @@ -43,6 +43,7 @@ #include #include +#define RCU_TREE_NONCORE #include "rcutree.h" static void print_one_rcu_data(struct seq_file *m, struct rcu_data *rdp) -- cgit v1.2.3 From d6714c22b43fbcbead7e7b706ff270e15f04a791 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Sat, 22 Aug 2009 13:56:46 -0700 Subject: rcu: Renamings to increase RCU clarity Make RCU-sched, RCU-bh, and RCU-preempt be underlying implementations, with "RCU" defined in terms of one of the three. Update the outdated rcu_qsctr_inc() names, as these functions no longer increment anything. Signed-off-by: Paul E. McKenney Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746132696-git-send-email-> Signed-off-by: Ingo Molnar --- Documentation/RCU/trace.txt | 7 ++-- include/linux/rcupdate.h | 21 +++++++++--- include/linux/rcupreempt.h | 4 +-- include/linux/rcutree.h | 8 +++-- kernel/rcupreempt.c | 8 ++--- kernel/rcutree.c | 80 ++++++++++++++++++++++++++++----------------- kernel/rcutree.h | 4 +-- kernel/rcutree_trace.c | 20 ++++++------ kernel/sched.c | 2 +- kernel/softirq.c | 4 +-- 10 files changed, 95 insertions(+), 63 deletions(-) (limited to 'include') diff --git a/Documentation/RCU/trace.txt b/Documentation/RCU/trace.txt index 02cced183b2d..187bbf10c923 100644 --- a/Documentation/RCU/trace.txt +++ b/Documentation/RCU/trace.txt @@ -191,8 +191,7 @@ rcu/rcuhier (which displays the struct rcu_node hierarchy). The output of "cat rcu/rcudata" looks as follows: -rcu: -rcu: +rcu_sched: 0 c=17829 g=17829 pq=1 pqc=17829 qp=0 dt=10951/1 dn=0 df=1101 of=0 ri=36 ql=0 b=10 1 c=17829 g=17829 pq=1 pqc=17829 qp=0 dt=16117/1 dn=0 df=1015 of=0 ri=0 ql=0 b=10 2 c=17829 g=17829 pq=1 pqc=17829 qp=0 dt=1445/1 dn=0 df=1839 of=0 ri=0 ql=0 b=10 @@ -306,7 +305,7 @@ comma-separated-variable spreadsheet format. The output of "cat rcu/rcugp" looks as follows: -rcu: completed=33062 gpnum=33063 +rcu_sched: completed=33062 gpnum=33063 rcu_bh: completed=464 gpnum=464 Again, this output is for both "rcu" and "rcu_bh". The fields are @@ -413,7 +412,7 @@ o Each element of the form "1/1 0:127 ^0" represents one struct The output of "cat rcu/rcu_pending" looks as follows: -rcu: +rcu_sched: 0 np=255892 qsp=53936 cbr=0 cng=14417 gpc=10033 gps=24320 nf=6445 nn=146741 1 np=261224 qsp=54638 cbr=0 cng=25723 gpc=16310 gps=2849 nf=5912 nn=155792 2 np=237496 qsp=49664 cbr=0 cng=2762 gpc=45478 gps=1762 nf=1201 nn=136629 diff --git a/include/linux/rcupdate.h b/include/linux/rcupdate.h index 3c89d6a2591f..e920f0fd59d8 100644 --- a/include/linux/rcupdate.h +++ b/include/linux/rcupdate.h @@ -157,17 +157,28 @@ extern int rcu_scheduler_active; * - call_rcu_sched() and rcu_barrier_sched() * on the write-side to insure proper synchronization. */ -#define rcu_read_lock_sched() preempt_disable() -#define rcu_read_lock_sched_notrace() preempt_disable_notrace() +static inline void rcu_read_lock_sched(void) +{ + preempt_disable(); +} +static inline void rcu_read_lock_sched_notrace(void) +{ + preempt_disable_notrace(); +} /* * rcu_read_unlock_sched - marks the end of a RCU-classic critical section * * See rcu_read_lock_sched for more information. */ -#define rcu_read_unlock_sched() preempt_enable() -#define rcu_read_unlock_sched_notrace() preempt_enable_notrace() - +static inline void rcu_read_unlock_sched(void) +{ + preempt_enable(); +} +static inline void rcu_read_unlock_sched_notrace(void) +{ + preempt_enable_notrace(); +} /** diff --git a/include/linux/rcupreempt.h b/include/linux/rcupreempt.h index f164ac9b7807..2963f080e48d 100644 --- a/include/linux/rcupreempt.h +++ b/include/linux/rcupreempt.h @@ -40,8 +40,8 @@ #include #include -extern void rcu_qsctr_inc(int cpu); -static inline void rcu_bh_qsctr_inc(int cpu) { } +extern void rcu_sched_qs(int cpu); +static inline void rcu_bh_qs(int cpu) { } /* * Someone might want to pass call_rcu_bh as a function pointer. diff --git a/include/linux/rcutree.h b/include/linux/rcutree.h index e37d5e2a8353..a0852d0d915b 100644 --- a/include/linux/rcutree.h +++ b/include/linux/rcutree.h @@ -30,8 +30,8 @@ #ifndef __LINUX_RCUTREE_H #define __LINUX_RCUTREE_H -extern void rcu_qsctr_inc(int cpu); -extern void rcu_bh_qsctr_inc(int cpu); +extern void rcu_sched_qs(int cpu); +extern void rcu_bh_qs(int cpu); extern int rcu_pending(int cpu); extern int rcu_needs_cpu(int cpu); @@ -73,7 +73,8 @@ static inline void __rcu_read_unlock_bh(void) #define __synchronize_sched() synchronize_rcu() -#define call_rcu_sched(head, func) call_rcu(head, func) +extern void call_rcu_sched(struct rcu_head *head, + void (*func)(struct rcu_head *rcu)); static inline void synchronize_rcu_expedited(void) { @@ -91,6 +92,7 @@ extern void rcu_restart_cpu(int cpu); extern long rcu_batches_completed(void); extern long rcu_batches_completed_bh(void); +extern long rcu_batches_completed_sched(void); static inline void rcu_init_sched(void) { diff --git a/kernel/rcupreempt.c b/kernel/rcupreempt.c index 510898a7bd69..7d777c9f394c 100644 --- a/kernel/rcupreempt.c +++ b/kernel/rcupreempt.c @@ -159,7 +159,7 @@ static DEFINE_PER_CPU_SHARED_ALIGNED(struct rcu_dyntick_sched, rcu_dyntick_sched .dynticks = 1, }; -void rcu_qsctr_inc(int cpu) +void rcu_sched_qs(int cpu) { struct rcu_dyntick_sched *rdssp = &per_cpu(rcu_dyntick_sched, cpu); @@ -967,12 +967,12 @@ void rcu_check_callbacks(int cpu, int user) * If this CPU took its interrupt from user mode or from the * idle loop, and this is not a nested interrupt, then * this CPU has to have exited all prior preept-disable - * sections of code. So increment the counter to note this. + * sections of code. So invoke rcu_sched_qs() to note this. * * The memory barrier is needed to handle the case where * writes from a preempt-disable section of code get reordered * into schedule() by this CPU's write buffer. So the memory - * barrier makes sure that the rcu_qsctr_inc() is seen by other + * barrier makes sure that the rcu_sched_qs() is seen by other * CPUs to happen after any such write. */ @@ -980,7 +980,7 @@ void rcu_check_callbacks(int cpu, int user) (idle_cpu(cpu) && !in_softirq() && hardirq_count() <= (1 << HARDIRQ_SHIFT))) { smp_mb(); /* Guard against aggressive schedule(). */ - rcu_qsctr_inc(cpu); + rcu_sched_qs(cpu); } rcu_check_mb(cpu); diff --git a/kernel/rcutree.c b/kernel/rcutree.c index a162f859dd32..4d71d4e8b5a8 100644 --- a/kernel/rcutree.c +++ b/kernel/rcutree.c @@ -74,26 +74,25 @@ EXPORT_SYMBOL_GPL(rcu_lock_map); .n_force_qs_ngp = 0, \ } -struct rcu_state rcu_state = RCU_STATE_INITIALIZER(rcu_state); -DEFINE_PER_CPU(struct rcu_data, rcu_data); +struct rcu_state rcu_sched_state = RCU_STATE_INITIALIZER(rcu_sched_state); +DEFINE_PER_CPU(struct rcu_data, rcu_sched_data); struct rcu_state rcu_bh_state = RCU_STATE_INITIALIZER(rcu_bh_state); DEFINE_PER_CPU(struct rcu_data, rcu_bh_data); /* - * Increment the quiescent state counter. - * The counter is a bit degenerated: We do not need to know + * Note a quiescent state. Because we do not need to know * how many quiescent states passed, just if there was at least - * one since the start of the grace period. Thus just a flag. + * one since the start of the grace period, this just sets a flag. */ -void rcu_qsctr_inc(int cpu) +void rcu_sched_qs(int cpu) { - struct rcu_data *rdp = &per_cpu(rcu_data, cpu); + struct rcu_data *rdp = &per_cpu(rcu_sched_data, cpu); rdp->passed_quiesc = 1; rdp->passed_quiesc_completed = rdp->completed; } -void rcu_bh_qsctr_inc(int cpu) +void rcu_bh_qs(int cpu) { struct rcu_data *rdp = &per_cpu(rcu_bh_data, cpu); rdp->passed_quiesc = 1; @@ -113,12 +112,22 @@ static int qlowmark = 100; /* Once only this many pending, use blimit. */ static void force_quiescent_state(struct rcu_state *rsp, int relaxed); +/* + * Return the number of RCU-sched batches processed thus far for debug & stats. + */ +long rcu_batches_completed_sched(void) +{ + return rcu_sched_state.completed; +} +EXPORT_SYMBOL_GPL(rcu_batches_completed_sched); + /* * Return the number of RCU batches processed thus far for debug & stats. + * @@@ placeholder, maps to rcu_batches_completed_sched(). */ long rcu_batches_completed(void) { - return rcu_state.completed; + return rcu_batches_completed_sched(); } EXPORT_SYMBOL_GPL(rcu_batches_completed); @@ -310,7 +319,7 @@ void rcu_irq_exit(void) WARN_ON_RATELIMIT(rdtp->dynticks & 0x1, &rcu_rs); /* If the interrupt queued a callback, get out of dyntick mode. */ - if (__get_cpu_var(rcu_data).nxtlist || + if (__get_cpu_var(rcu_sched_data).nxtlist || __get_cpu_var(rcu_bh_data).nxtlist) set_need_resched(); } @@ -847,7 +856,7 @@ static void __rcu_offline_cpu(int cpu, struct rcu_state *rsp) /* * Move callbacks from the outgoing CPU to the running CPU. * Note that the outgoing CPU is now quiscent, so it is now - * (uncharacteristically) safe to access it rcu_data structure. + * (uncharacteristically) safe to access its rcu_data structure. * Note also that we must carefully retain the order of the * outgoing CPU's callbacks in order for rcu_barrier() to work * correctly. Finally, note that we start all the callbacks @@ -878,7 +887,7 @@ static void __rcu_offline_cpu(int cpu, struct rcu_state *rsp) */ static void rcu_offline_cpu(int cpu) { - __rcu_offline_cpu(cpu, &rcu_state); + __rcu_offline_cpu(cpu, &rcu_sched_state); __rcu_offline_cpu(cpu, &rcu_bh_state); } @@ -973,17 +982,16 @@ void rcu_check_callbacks(int cpu, int user) * Get here if this CPU took its interrupt from user * mode or from the idle loop, and if this is not a * nested interrupt. In this case, the CPU is in - * a quiescent state, so count it. + * a quiescent state, so note it. * * No memory barrier is required here because both - * rcu_qsctr_inc() and rcu_bh_qsctr_inc() reference - * only CPU-local variables that other CPUs neither - * access nor modify, at least not while the corresponding - * CPU is online. + * rcu_sched_qs() and rcu_bh_qs() reference only CPU-local + * variables that other CPUs neither access nor modify, + * at least not while the corresponding CPU is online. */ - rcu_qsctr_inc(cpu); - rcu_bh_qsctr_inc(cpu); + rcu_sched_qs(cpu); + rcu_bh_qs(cpu); } else if (!in_softirq()) { @@ -991,10 +999,10 @@ void rcu_check_callbacks(int cpu, int user) * Get here if this CPU did not take its interrupt from * softirq, in other words, if it is not interrupting * a rcu_bh read-side critical section. This is an _bh - * critical section, so count it. + * critical section, so note it. */ - rcu_bh_qsctr_inc(cpu); + rcu_bh_qs(cpu); } raise_softirq(RCU_SOFTIRQ); } @@ -1174,7 +1182,8 @@ static void rcu_process_callbacks(struct softirq_action *unused) */ smp_mb(); /* See above block comment. */ - __rcu_process_callbacks(&rcu_state, &__get_cpu_var(rcu_data)); + __rcu_process_callbacks(&rcu_sched_state, + &__get_cpu_var(rcu_sched_data)); __rcu_process_callbacks(&rcu_bh_state, &__get_cpu_var(rcu_bh_data)); /* @@ -1231,14 +1240,25 @@ __call_rcu(struct rcu_head *head, void (*func)(struct rcu_head *rcu), } /* - * Queue an RCU callback for invocation after a grace period. + * Queue an RCU-sched callback for invocation after a grace period. + */ +void call_rcu_sched(struct rcu_head *head, void (*func)(struct rcu_head *rcu)) +{ + __call_rcu(head, func, &rcu_sched_state); +} +EXPORT_SYMBOL_GPL(call_rcu_sched); + +/* + * @@@ Queue an RCU callback for invocation after a grace period. + * @@@ Placeholder pending rcutree_plugin.h. */ void call_rcu(struct rcu_head *head, void (*func)(struct rcu_head *rcu)) { - __call_rcu(head, func, &rcu_state); + call_rcu_sched(head, func); } EXPORT_SYMBOL_GPL(call_rcu); + /* * Queue an RCU for invocation after a quicker grace period. */ @@ -1311,7 +1331,7 @@ static int __rcu_pending(struct rcu_state *rsp, struct rcu_data *rdp) */ int rcu_pending(int cpu) { - return __rcu_pending(&rcu_state, &per_cpu(rcu_data, cpu)) || + return __rcu_pending(&rcu_sched_state, &per_cpu(rcu_sched_data, cpu)) || __rcu_pending(&rcu_bh_state, &per_cpu(rcu_bh_data, cpu)); } @@ -1324,7 +1344,7 @@ int rcu_pending(int cpu) int rcu_needs_cpu(int cpu) { /* RCU callbacks either ready or pending? */ - return per_cpu(rcu_data, cpu).nxtlist || + return per_cpu(rcu_sched_data, cpu).nxtlist || per_cpu(rcu_bh_data, cpu).nxtlist; } @@ -1418,7 +1438,7 @@ rcu_init_percpu_data(int cpu, struct rcu_state *rsp) static void __cpuinit rcu_online_cpu(int cpu) { - rcu_init_percpu_data(cpu, &rcu_state); + rcu_init_percpu_data(cpu, &rcu_sched_state); rcu_init_percpu_data(cpu, &rcu_bh_state); } @@ -1545,10 +1565,10 @@ void __init __rcu_init(void) #ifdef CONFIG_RCU_CPU_STALL_DETECTOR printk(KERN_INFO "RCU-based detection of stalled CPUs is enabled.\n"); #endif /* #ifdef CONFIG_RCU_CPU_STALL_DETECTOR */ - rcu_init_one(&rcu_state); - RCU_DATA_PTR_INIT(&rcu_state, rcu_data); + rcu_init_one(&rcu_sched_state); + RCU_DATA_PTR_INIT(&rcu_sched_state, rcu_sched_data); for_each_possible_cpu(i) - rcu_boot_init_percpu_data(i, &rcu_state); + rcu_boot_init_percpu_data(i, &rcu_sched_state); rcu_init_one(&rcu_bh_state); RCU_DATA_PTR_INIT(&rcu_bh_state, rcu_bh_data); for_each_possible_cpu(i) diff --git a/kernel/rcutree.h b/kernel/rcutree.h index 7cc830a1c44a..0024e5ddcc68 100644 --- a/kernel/rcutree.h +++ b/kernel/rcutree.h @@ -238,8 +238,8 @@ struct rcu_state { /* * RCU implementation internal declarations: */ -extern struct rcu_state rcu_state; -DECLARE_PER_CPU(struct rcu_data, rcu_data); +extern struct rcu_state rcu_sched_state; +DECLARE_PER_CPU(struct rcu_data, rcu_sched_data); extern struct rcu_state rcu_bh_state; DECLARE_PER_CPU(struct rcu_data, rcu_bh_data); diff --git a/kernel/rcutree_trace.c b/kernel/rcutree_trace.c index 0cb52b887758..236c0504fee2 100644 --- a/kernel/rcutree_trace.c +++ b/kernel/rcutree_trace.c @@ -77,8 +77,8 @@ static void print_one_rcu_data(struct seq_file *m, struct rcu_data *rdp) static int show_rcudata(struct seq_file *m, void *unused) { - seq_puts(m, "rcu:\n"); - PRINT_RCU_DATA(rcu_data, print_one_rcu_data, m); + seq_puts(m, "rcu_sched:\n"); + PRINT_RCU_DATA(rcu_sched_data, print_one_rcu_data, m); seq_puts(m, "rcu_bh:\n"); PRINT_RCU_DATA(rcu_bh_data, print_one_rcu_data, m); return 0; @@ -125,8 +125,8 @@ static int show_rcudata_csv(struct seq_file *m, void *unused) seq_puts(m, "\"dt\",\"dt nesting\",\"dn\",\"df\","); #endif /* #ifdef CONFIG_NO_HZ */ seq_puts(m, "\"of\",\"ri\",\"ql\",\"b\"\n"); - seq_puts(m, "\"rcu:\"\n"); - PRINT_RCU_DATA(rcu_data, print_one_rcu_data_csv, m); + seq_puts(m, "\"rcu_sched:\"\n"); + PRINT_RCU_DATA(rcu_sched_data, print_one_rcu_data_csv, m); seq_puts(m, "\"rcu_bh:\"\n"); PRINT_RCU_DATA(rcu_bh_data, print_one_rcu_data_csv, m); return 0; @@ -172,8 +172,8 @@ static void print_one_rcu_state(struct seq_file *m, struct rcu_state *rsp) static int show_rcuhier(struct seq_file *m, void *unused) { - seq_puts(m, "rcu:\n"); - print_one_rcu_state(m, &rcu_state); + seq_puts(m, "rcu_sched:\n"); + print_one_rcu_state(m, &rcu_sched_state); seq_puts(m, "rcu_bh:\n"); print_one_rcu_state(m, &rcu_bh_state); return 0; @@ -194,8 +194,8 @@ static struct file_operations rcuhier_fops = { static int show_rcugp(struct seq_file *m, void *unused) { - seq_printf(m, "rcu: completed=%ld gpnum=%ld\n", - rcu_state.completed, rcu_state.gpnum); + seq_printf(m, "rcu_sched: completed=%ld gpnum=%ld\n", + rcu_sched_state.completed, rcu_sched_state.gpnum); seq_printf(m, "rcu_bh: completed=%ld gpnum=%ld\n", rcu_bh_state.completed, rcu_bh_state.gpnum); return 0; @@ -244,8 +244,8 @@ static void print_rcu_pendings(struct seq_file *m, struct rcu_state *rsp) static int show_rcu_pending(struct seq_file *m, void *unused) { - seq_puts(m, "rcu:\n"); - print_rcu_pendings(m, &rcu_state); + seq_puts(m, "rcu_sched:\n"); + print_rcu_pendings(m, &rcu_sched_state); seq_puts(m, "rcu_bh:\n"); print_rcu_pendings(m, &rcu_bh_state); return 0; diff --git a/kernel/sched.c b/kernel/sched.c index cda8b81f8801..c9beca67a53e 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -5325,7 +5325,7 @@ need_resched: preempt_disable(); cpu = smp_processor_id(); rq = cpu_rq(cpu); - rcu_qsctr_inc(cpu); + rcu_sched_qs(cpu); prev = rq->curr; switch_count = &prev->nivcsw; diff --git a/kernel/softirq.c b/kernel/softirq.c index eb5e131a0485..7db25067cd2d 100644 --- a/kernel/softirq.c +++ b/kernel/softirq.c @@ -227,7 +227,7 @@ restart: preempt_count() = prev_count; } - rcu_bh_qsctr_inc(cpu); + rcu_bh_qs(cpu); } h++; pending >>= 1; @@ -721,7 +721,7 @@ static int ksoftirqd(void * __bind_cpu) preempt_enable_no_resched(); cond_resched(); preempt_disable(); - rcu_qsctr_inc((long)__bind_cpu); + rcu_sched_qs((long)__bind_cpu); } preempt_enable(); set_current_state(TASK_INTERRUPTIBLE); -- cgit v1.2.3 From bc33f24bdca8b6e97376e3a182ab69e6cdefa989 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Sat, 22 Aug 2009 13:56:47 -0700 Subject: rcu: Consolidate sparse and lockdep declarations in include/linux/rcupdate.h Signed-off-by: Paul E. McKenney Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746132349-git-send-email-> Signed-off-by: Ingo Molnar --- include/linux/rcupdate.h | 46 ++++++++++++++++++++++++++++++++++++++++++---- include/linux/rcupreempt.h | 4 ++-- include/linux/rcutree.h | 18 ------------------ 3 files changed, 44 insertions(+), 24 deletions(-) (limited to 'include') diff --git a/include/linux/rcupdate.h b/include/linux/rcupdate.h index e920f0fd59d8..9d85ee19492a 100644 --- a/include/linux/rcupdate.h +++ b/include/linux/rcupdate.h @@ -80,6 +80,16 @@ extern int rcu_scheduler_active; (ptr)->next = NULL; (ptr)->func = NULL; \ } while (0) +#ifdef CONFIG_DEBUG_LOCK_ALLOC +extern struct lockdep_map rcu_lock_map; +# define rcu_read_acquire() \ + lock_acquire(&rcu_lock_map, 0, 0, 2, 1, NULL, _THIS_IP_) +# define rcu_read_release() lock_release(&rcu_lock_map, 1, _THIS_IP_) +#else +# define rcu_read_acquire() do { } while (0) +# define rcu_read_release() do { } while (0) +#endif + /** * rcu_read_lock - mark the beginning of an RCU read-side critical section. * @@ -109,7 +119,12 @@ extern int rcu_scheduler_active; * * It is illegal to block while in an RCU read-side critical section. */ -#define rcu_read_lock() __rcu_read_lock() +static inline void rcu_read_lock(void) +{ + __rcu_read_lock(); + __acquire(RCU); + rcu_read_acquire(); +} /** * rcu_read_unlock - marks the end of an RCU read-side critical section. @@ -126,7 +141,12 @@ extern int rcu_scheduler_active; * used as well. RCU does not care how the writers keep out of each * others' way, as long as they do so. */ -#define rcu_read_unlock() __rcu_read_unlock() +static inline void rcu_read_unlock(void) +{ + rcu_read_release(); + __release(RCU); + __rcu_read_unlock(); +} /** * rcu_read_lock_bh - mark the beginning of a softirq-only RCU critical section @@ -139,14 +159,24 @@ extern int rcu_scheduler_active; * can use just rcu_read_lock(). * */ -#define rcu_read_lock_bh() __rcu_read_lock_bh() +static inline void rcu_read_lock_bh(void) +{ + __rcu_read_lock_bh(); + __acquire(RCU_BH); + rcu_read_acquire(); +} /* * rcu_read_unlock_bh - marks the end of a softirq-only RCU critical section * * See rcu_read_lock_bh() for more information. */ -#define rcu_read_unlock_bh() __rcu_read_unlock_bh() +static inline void rcu_read_unlock_bh(void) +{ + rcu_read_release(); + __release(RCU_BH); + __rcu_read_unlock_bh(); +} /** * rcu_read_lock_sched - mark the beginning of a RCU-classic critical section @@ -160,10 +190,14 @@ extern int rcu_scheduler_active; static inline void rcu_read_lock_sched(void) { preempt_disable(); + __acquire(RCU_SCHED); + rcu_read_acquire(); } static inline void rcu_read_lock_sched_notrace(void) { preempt_disable_notrace(); + __acquire(RCU_SCHED); + rcu_read_acquire(); } /* @@ -173,10 +207,14 @@ static inline void rcu_read_lock_sched_notrace(void) */ static inline void rcu_read_unlock_sched(void) { + rcu_read_release(); + __release(RCU_SCHED); preempt_enable(); } static inline void rcu_read_unlock_sched_notrace(void) { + rcu_read_release(); + __release(RCU_SCHED); preempt_enable_notrace(); } diff --git a/include/linux/rcupreempt.h b/include/linux/rcupreempt.h index 2963f080e48d..6c9dd9cf8b8e 100644 --- a/include/linux/rcupreempt.h +++ b/include/linux/rcupreempt.h @@ -64,8 +64,8 @@ static inline void rcu_bh_qs(int cpu) { } extern void call_rcu_sched(struct rcu_head *head, void (*func)(struct rcu_head *head)); -extern void __rcu_read_lock(void) __acquires(RCU); -extern void __rcu_read_unlock(void) __releases(RCU); +extern void __rcu_read_lock(void); +extern void __rcu_read_unlock(void); extern int rcu_pending(int cpu); extern int rcu_needs_cpu(int cpu); diff --git a/include/linux/rcutree.h b/include/linux/rcutree.h index a0852d0d915b..8a0222ce3b13 100644 --- a/include/linux/rcutree.h +++ b/include/linux/rcutree.h @@ -36,38 +36,20 @@ extern void rcu_bh_qs(int cpu); extern int rcu_pending(int cpu); extern int rcu_needs_cpu(int cpu); -#ifdef CONFIG_DEBUG_LOCK_ALLOC -extern struct lockdep_map rcu_lock_map; -# define rcu_read_acquire() \ - lock_acquire(&rcu_lock_map, 0, 0, 2, 1, NULL, _THIS_IP_) -# define rcu_read_release() lock_release(&rcu_lock_map, 1, _THIS_IP_) -#else -# define rcu_read_acquire() do { } while (0) -# define rcu_read_release() do { } while (0) -#endif - static inline void __rcu_read_lock(void) { preempt_disable(); - __acquire(RCU); - rcu_read_acquire(); } static inline void __rcu_read_unlock(void) { - rcu_read_release(); - __release(RCU); preempt_enable(); } static inline void __rcu_read_lock_bh(void) { local_bh_disable(); - __acquire(RCU_BH); - rcu_read_acquire(); } static inline void __rcu_read_unlock_bh(void) { - rcu_read_release(); - __release(RCU_BH); local_bh_enable(); } -- cgit v1.2.3 From a157229cabd6dd8cfa82525fc9bf730c94cc9ac2 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Sat, 22 Aug 2009 13:56:51 -0700 Subject: rcu: Simplify rcu_pending()/rcu_check_callbacks() API All calls from outside RCU are of the form: if (rcu_pending(cpu)) rcu_check_callbacks(cpu, user); This is silly, instead we put a call to rcu_pending() in rcu_check_callbacks(), and then make the outside calls be to rcu_check_callbacks(). This cuts down on the code a bit and also gives the compiler a better chance of optimizing. Signed-off-by: Paul E. McKenney Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <125097461311-git-send-email-> Signed-off-by: Ingo Molnar --- arch/ia64/xen/time.c | 3 +-- include/linux/rcupreempt.h | 1 - include/linux/rcutree.h | 1 - kernel/rcupreempt.c | 10 ++++++++-- kernel/rcutree.c | 5 ++++- kernel/timer.c | 3 +-- 6 files changed, 14 insertions(+), 9 deletions(-) (limited to 'include') diff --git a/arch/ia64/xen/time.c b/arch/ia64/xen/time.c index fb8332690179..dbeadb9c8e20 100644 --- a/arch/ia64/xen/time.c +++ b/arch/ia64/xen/time.c @@ -133,8 +133,7 @@ consider_steal_time(unsigned long new_itm) account_idle_ticks(blocked); run_local_timers(); - if (rcu_pending(cpu)) - rcu_check_callbacks(cpu, user_mode(get_irq_regs())); + rcu_check_callbacks(cpu, user_mode(get_irq_regs())); scheduler_tick(); run_posix_cpu_timers(p); diff --git a/include/linux/rcupreempt.h b/include/linux/rcupreempt.h index 6c9dd9cf8b8e..aff4772fb49e 100644 --- a/include/linux/rcupreempt.h +++ b/include/linux/rcupreempt.h @@ -66,7 +66,6 @@ extern void call_rcu_sched(struct rcu_head *head, extern void __rcu_read_lock(void); extern void __rcu_read_unlock(void); -extern int rcu_pending(int cpu); extern int rcu_needs_cpu(int cpu); #define __rcu_read_lock_bh() { rcu_read_lock(); local_bh_disable(); } diff --git a/include/linux/rcutree.h b/include/linux/rcutree.h index 8a0222ce3b13..c739d90f5e68 100644 --- a/include/linux/rcutree.h +++ b/include/linux/rcutree.h @@ -33,7 +33,6 @@ extern void rcu_sched_qs(int cpu); extern void rcu_bh_qs(int cpu); -extern int rcu_pending(int cpu); extern int rcu_needs_cpu(int cpu); static inline void __rcu_read_lock(void) diff --git a/kernel/rcupreempt.c b/kernel/rcupreempt.c index 7d777c9f394c..0053ce56e326 100644 --- a/kernel/rcupreempt.c +++ b/kernel/rcupreempt.c @@ -159,6 +159,8 @@ static DEFINE_PER_CPU_SHARED_ALIGNED(struct rcu_dyntick_sched, rcu_dyntick_sched .dynticks = 1, }; +static int rcu_pending(int cpu); + void rcu_sched_qs(int cpu) { struct rcu_dyntick_sched *rdssp = &per_cpu(rcu_dyntick_sched, cpu); @@ -961,7 +963,10 @@ static void rcu_check_mb(int cpu) void rcu_check_callbacks(int cpu, int user) { unsigned long flags; - struct rcu_data *rdp = RCU_DATA_CPU(cpu); + struct rcu_data *rdp; + + if (!rcu_pending(cpu)) + return; /* if nothing for RCU to do. */ /* * If this CPU took its interrupt from user mode or from the @@ -976,6 +981,7 @@ void rcu_check_callbacks(int cpu, int user) * CPUs to happen after any such write. */ + rdp = RCU_DATA_CPU(cpu); if (user || (idle_cpu(cpu) && !in_softirq() && hardirq_count() <= (1 << HARDIRQ_SHIFT))) { @@ -1382,7 +1388,7 @@ int rcu_needs_cpu(int cpu) rdp->waitschedlist != NULL); } -int rcu_pending(int cpu) +static int rcu_pending(int cpu) { struct rcu_data *rdp = RCU_DATA_CPU(cpu); diff --git a/kernel/rcutree.c b/kernel/rcutree.c index 7c515082ae84..4ce3adcfa94d 100644 --- a/kernel/rcutree.c +++ b/kernel/rcutree.c @@ -111,6 +111,7 @@ static int qhimark = 10000; /* If this many pending, ignore blimit. */ static int qlowmark = 100; /* Once only this many pending, use blimit. */ static void force_quiescent_state(struct rcu_state *rsp, int relaxed); +static int rcu_pending(int cpu); /* * Return the number of RCU-sched batches processed thus far for debug & stats. @@ -974,6 +975,8 @@ static void rcu_do_batch(struct rcu_data *rdp) */ void rcu_check_callbacks(int cpu, int user) { + if (!rcu_pending(cpu)) + return; /* if nothing for RCU to do. */ if (user || (idle_cpu(cpu) && rcu_scheduler_active && !in_softirq() && hardirq_count() <= (1 << HARDIRQ_SHIFT))) { @@ -1329,7 +1332,7 @@ static int __rcu_pending(struct rcu_state *rsp, struct rcu_data *rdp) * by the current CPU, returning 1 if so. This function is part of the * RCU implementation; it is -not- an exported member of the RCU API. */ -int rcu_pending(int cpu) +static int rcu_pending(int cpu) { return __rcu_pending(&rcu_sched_state, &per_cpu(rcu_sched_data, cpu)) || __rcu_pending(&rcu_bh_state, &per_cpu(rcu_bh_data, cpu)); diff --git a/kernel/timer.c b/kernel/timer.c index a7f07d5a6241..a3d25f415019 100644 --- a/kernel/timer.c +++ b/kernel/timer.c @@ -1156,8 +1156,7 @@ void update_process_times(int user_tick) /* Note: this timer irq context must be accounted for as well. */ account_process_tick(p, user_tick); run_local_timers(); - if (rcu_pending(cpu)) - rcu_check_callbacks(cpu, user_tick); + rcu_check_callbacks(cpu, user_tick); printk_tick(); scheduler_tick(); run_posix_cpu_timers(p); -- cgit v1.2.3 From f41d911f8c49a5d65c86504c19e8204bb605c4fd Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Sat, 22 Aug 2009 13:56:52 -0700 Subject: rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar --- include/linux/init_task.h | 15 ++ include/linux/rcupdate.h | 2 +- include/linux/rcupreempt.h | 4 + include/linux/rcutree.h | 16 ++ include/linux/sched.h | 37 ++++ init/Kconfig | 22 ++- kernel/Makefile | 1 + kernel/exit.c | 1 + kernel/fork.c | 5 +- kernel/rcutree.c | 135 +++++++++----- kernel/rcutree.h | 9 + kernel/rcutree_plugin.h | 447 +++++++++++++++++++++++++++++++++++++++++++++ kernel/rcutree_trace.c | 20 ++ lib/Kconfig.debug | 2 +- 14 files changed, 661 insertions(+), 55 deletions(-) create mode 100644 kernel/rcutree_plugin.h (limited to 'include') diff --git a/include/linux/init_task.h b/include/linux/init_task.h index 7fc01b13be43..971a968831bf 100644 --- a/include/linux/init_task.h +++ b/include/linux/init_task.h @@ -94,6 +94,20 @@ extern struct group_info init_groups; # define CAP_INIT_BSET CAP_INIT_EFF_SET #endif +#ifdef CONFIG_PREEMPT_RCU +#define INIT_TASK_RCU_PREEMPT(tsk) \ + .rcu_read_lock_nesting = 0, \ + .rcu_flipctr_idx = 0, +#elif defined(CONFIG_TREE_PREEMPT_RCU) +#define INIT_TASK_RCU_PREEMPT(tsk) \ + .rcu_read_lock_nesting = 0, \ + .rcu_read_unlock_special = 0, \ + .rcu_blocked_cpu = -1, \ + .rcu_node_entry = LIST_HEAD_INIT(tsk.rcu_node_entry), +#else +#define INIT_TASK_RCU_PREEMPT(tsk) +#endif + extern struct cred init_cred; #ifdef CONFIG_PERF_COUNTERS @@ -173,6 +187,7 @@ extern struct cred init_cred; INIT_LOCKDEP \ INIT_FTRACE_GRAPH \ INIT_TRACE_RECURSION \ + INIT_TASK_RCU_PREEMPT(tsk) \ } diff --git a/include/linux/rcupdate.h b/include/linux/rcupdate.h index 9d85ee19492a..26892f5e7bd8 100644 --- a/include/linux/rcupdate.h +++ b/include/linux/rcupdate.h @@ -66,7 +66,7 @@ extern void rcu_scheduler_starting(void); extern int rcu_needs_cpu(int cpu); extern int rcu_scheduler_active; -#if defined(CONFIG_TREE_RCU) +#if defined(CONFIG_TREE_RCU) || defined(CONFIG_TREE_PREEMPT_RCU) #include #elif defined(CONFIG_PREEMPT_RCU) #include diff --git a/include/linux/rcupreempt.h b/include/linux/rcupreempt.h index aff4772fb49e..a42ab88e9210 100644 --- a/include/linux/rcupreempt.h +++ b/include/linux/rcupreempt.h @@ -98,6 +98,10 @@ static inline long rcu_batches_completed_bh(void) return rcu_batches_completed(); } +static inline void exit_rcu(void) +{ +} + #ifdef CONFIG_RCU_TRACE struct rcupreempt_trace; extern long *rcupreempt_flipctr(int cpu); diff --git a/include/linux/rcutree.h b/include/linux/rcutree.h index c739d90f5e68..a89307717825 100644 --- a/include/linux/rcutree.h +++ b/include/linux/rcutree.h @@ -35,14 +35,30 @@ extern void rcu_bh_qs(int cpu); extern int rcu_needs_cpu(int cpu); +#ifdef CONFIG_TREE_PREEMPT_RCU + +extern void __rcu_read_lock(void); +extern void __rcu_read_unlock(void); +extern void exit_rcu(void); + +#else /* #ifdef CONFIG_TREE_PREEMPT_RCU */ + static inline void __rcu_read_lock(void) { preempt_disable(); } + static inline void __rcu_read_unlock(void) { preempt_enable(); } + +static inline void exit_rcu(void) +{ +} + +#endif /* #else #ifdef CONFIG_TREE_PREEMPT_RCU */ + static inline void __rcu_read_lock_bh(void) { local_bh_disable(); diff --git a/include/linux/sched.h b/include/linux/sched.h index 3ab08e4bb6b8..d7f98f637a2a 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -1210,6 +1210,13 @@ struct task_struct { int rcu_flipctr_idx; #endif /* #ifdef CONFIG_PREEMPT_RCU */ +#ifdef CONFIG_TREE_PREEMPT_RCU + int rcu_read_lock_nesting; + char rcu_read_unlock_special; + int rcu_blocked_cpu; + struct list_head rcu_node_entry; +#endif /* #ifdef CONFIG_TREE_PREEMPT_RCU */ + #if defined(CONFIG_SCHEDSTATS) || defined(CONFIG_TASK_DELAY_ACCT) struct sched_info sched_info; #endif @@ -1723,6 +1730,36 @@ extern cputime_t task_gtime(struct task_struct *p); #define tsk_used_math(p) ((p)->flags & PF_USED_MATH) #define used_math() tsk_used_math(current) +#ifdef CONFIG_TREE_PREEMPT_RCU + +#define RCU_READ_UNLOCK_BLOCKED (1 << 0) /* blocked while in RCU read-side. */ +#define RCU_READ_UNLOCK_NEED_QS (1 << 1) /* RCU core needs CPU response. */ +#define RCU_READ_UNLOCK_GOT_QS (1 << 2) /* CPU has responded to RCU core. */ + +static inline void rcu_copy_process(struct task_struct *p) +{ + p->rcu_read_lock_nesting = 0; + p->rcu_read_unlock_special = 0; + p->rcu_blocked_cpu = -1; + INIT_LIST_HEAD(&p->rcu_node_entry); +} + +#elif defined(CONFIG_PREEMPT_RCU) + +static inline void rcu_copy_process(struct task_struct *p) +{ + p->rcu_read_lock_nesting = 0; + p->rcu_flipctr_idx = 0; +} + +#else + +static inline void rcu_copy_process(struct task_struct *p) +{ +} + +#endif + #ifdef CONFIG_SMP extern int set_cpus_allowed_ptr(struct task_struct *p, const struct cpumask *new_mask); diff --git a/init/Kconfig b/init/Kconfig index 25373cf32672..f88da2d1c1fb 100644 --- a/init/Kconfig +++ b/init/Kconfig @@ -335,11 +335,20 @@ config PREEMPT_RCU now-naive assumptions about each RCU read-side critical section remaining on a given CPU through its execution. +config TREE_PREEMPT_RCU + bool "Preemptable tree-based hierarchical RCU" + depends on PREEMPT + help + This option selects the RCU implementation that is + designed for very large SMP systems with hundreds or + thousands of CPUs, but for which real-time response + is also required. + endchoice config RCU_TRACE bool "Enable tracing for RCU" - depends on TREE_RCU || PREEMPT_RCU + depends on TREE_RCU || PREEMPT_RCU || TREE_PREEMPT_RCU help This option provides tracing in RCU which presents stats in debugfs for debugging RCU implementation. @@ -351,7 +360,7 @@ config RCU_FANOUT int "Tree-based hierarchical RCU fanout value" range 2 64 if 64BIT range 2 32 if !64BIT - depends on TREE_RCU + depends on TREE_RCU || TREE_PREEMPT_RCU default 64 if 64BIT default 32 if !64BIT help @@ -366,7 +375,7 @@ config RCU_FANOUT config RCU_FANOUT_EXACT bool "Disable tree-based hierarchical RCU auto-balancing" - depends on TREE_RCU + depends on TREE_RCU || TREE_PREEMPT_RCU default n help This option forces use of the exact RCU_FANOUT value specified, @@ -379,11 +388,12 @@ config RCU_FANOUT_EXACT Say N if unsure. config TREE_RCU_TRACE - def_bool RCU_TRACE && TREE_RCU + def_bool RCU_TRACE && ( TREE_RCU || TREE_PREEMPT_RCU ) select DEBUG_FS help - This option provides tracing for the TREE_RCU implementation, - permitting Makefile to trivially select kernel/rcutree_trace.c. + This option provides tracing for the TREE_RCU and + TREE_PREEMPT_RCU implementations, permitting Makefile to + trivially select kernel/rcutree_trace.c. config PREEMPT_RCU_TRACE def_bool RCU_TRACE && PREEMPT_RCU diff --git a/kernel/Makefile b/kernel/Makefile index 2419c9d43918..1a38b4789dda 100644 --- a/kernel/Makefile +++ b/kernel/Makefile @@ -81,6 +81,7 @@ obj-$(CONFIG_GENERIC_HARDIRQS) += irq/ obj-$(CONFIG_SECCOMP) += seccomp.o obj-$(CONFIG_RCU_TORTURE_TEST) += rcutorture.o obj-$(CONFIG_TREE_RCU) += rcutree.o +obj-$(CONFIG_TREE_PREEMPT_RCU) += rcutree.o obj-$(CONFIG_PREEMPT_RCU) += rcupreempt.o obj-$(CONFIG_TREE_RCU_TRACE) += rcutree_trace.o obj-$(CONFIG_PREEMPT_RCU_TRACE) += rcupreempt_trace.o diff --git a/kernel/exit.c b/kernel/exit.c index 869dc221733e..263f95ed7201 100644 --- a/kernel/exit.c +++ b/kernel/exit.c @@ -1010,6 +1010,7 @@ NORET_TYPE void do_exit(long code) __free_pipe_info(tsk->splice_pipe); preempt_disable(); + exit_rcu(); /* causes final put_task_struct in finish_task_switch(). */ tsk->state = TASK_DEAD; schedule(); diff --git a/kernel/fork.c b/kernel/fork.c index 021e1138556e..642e8b5edf00 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -1022,10 +1022,7 @@ static struct task_struct *copy_process(unsigned long clone_flags, copy_flags(clone_flags, p); INIT_LIST_HEAD(&p->children); INIT_LIST_HEAD(&p->sibling); -#ifdef CONFIG_PREEMPT_RCU - p->rcu_read_lock_nesting = 0; - p->rcu_flipctr_idx = 0; -#endif /* #ifdef CONFIG_PREEMPT_RCU */ + rcu_copy_process(p); p->vfork_done = NULL; spin_lock_init(&p->alloc_lock); diff --git a/kernel/rcutree.c b/kernel/rcutree.c index 4ce3adcfa94d..cc0255714075 100644 --- a/kernel/rcutree.c +++ b/kernel/rcutree.c @@ -80,6 +80,21 @@ DEFINE_PER_CPU(struct rcu_data, rcu_sched_data); struct rcu_state rcu_bh_state = RCU_STATE_INITIALIZER(rcu_bh_state); DEFINE_PER_CPU(struct rcu_data, rcu_bh_data); +extern long rcu_batches_completed_sched(void); +static void cpu_quiet_msk(unsigned long mask, struct rcu_state *rsp, + struct rcu_node *rnp, unsigned long flags); +static void cpu_quiet_msk_finish(struct rcu_state *rsp, unsigned long flags); +static void __rcu_process_callbacks(struct rcu_state *rsp, + struct rcu_data *rdp); +static void __call_rcu(struct rcu_head *head, + void (*func)(struct rcu_head *rcu), + struct rcu_state *rsp); +static int __rcu_pending(struct rcu_state *rsp, struct rcu_data *rdp); +static void __cpuinit rcu_init_percpu_data(int cpu, struct rcu_state *rsp, + int preemptable); + +#include "rcutree_plugin.h" + /* * Note a quiescent state. Because we do not need to know * how many quiescent states passed, just if there was at least @@ -87,16 +102,27 @@ DEFINE_PER_CPU(struct rcu_data, rcu_bh_data); */ void rcu_sched_qs(int cpu) { - struct rcu_data *rdp = &per_cpu(rcu_sched_data, cpu); + unsigned long flags; + struct rcu_data *rdp; + + local_irq_save(flags); + rdp = &per_cpu(rcu_sched_data, cpu); rdp->passed_quiesc = 1; rdp->passed_quiesc_completed = rdp->completed; + rcu_preempt_qs(cpu); + local_irq_restore(flags); } void rcu_bh_qs(int cpu) { - struct rcu_data *rdp = &per_cpu(rcu_bh_data, cpu); + unsigned long flags; + struct rcu_data *rdp; + + local_irq_save(flags); + rdp = &per_cpu(rcu_bh_data, cpu); rdp->passed_quiesc = 1; rdp->passed_quiesc_completed = rdp->completed; + local_irq_restore(flags); } #ifdef CONFIG_NO_HZ @@ -122,16 +148,6 @@ long rcu_batches_completed_sched(void) } EXPORT_SYMBOL_GPL(rcu_batches_completed_sched); -/* - * Return the number of RCU batches processed thus far for debug & stats. - * @@@ placeholder, maps to rcu_batches_completed_sched(). - */ -long rcu_batches_completed(void) -{ - return rcu_batches_completed_sched(); -} -EXPORT_SYMBOL_GPL(rcu_batches_completed); - /* * Return the number of RCU BH batches processed thus far for debug & stats. */ @@ -193,6 +209,10 @@ static int rcu_implicit_offline_qs(struct rcu_data *rdp) return 1; } + /* If preemptable RCU, no point in sending reschedule IPI. */ + if (rdp->preemptable) + return 0; + /* The CPU is online, so send it a reschedule IPI. */ if (rdp->cpu != smp_processor_id()) smp_send_reschedule(rdp->cpu); @@ -473,6 +493,7 @@ static void print_other_cpu_stall(struct rcu_state *rsp) printk(KERN_ERR "INFO: RCU detected CPU stalls:"); for (; rnp_cur < rnp_end; rnp_cur++) { + rcu_print_task_stall(rnp); if (rnp_cur->qsmask == 0) continue; for (cpu = 0; cpu <= rnp_cur->grphi - rnp_cur->grplo; cpu++) @@ -685,6 +706,19 @@ rcu_process_gp_end(struct rcu_state *rsp, struct rcu_data *rdp) local_irq_restore(flags); } +/* + * Clean up after the prior grace period and let rcu_start_gp() start up + * the next grace period if one is needed. Note that the caller must + * hold rnp->lock, as required by rcu_start_gp(), which will release it. + */ +static void cpu_quiet_msk_finish(struct rcu_state *rsp, unsigned long flags) + __releases(rnp->lock) +{ + rsp->completed = rsp->gpnum; + rcu_process_gp_end(rsp, rsp->rda[smp_processor_id()]); + rcu_start_gp(rsp, flags); /* releases root node's rnp->lock. */ +} + /* * Similar to cpu_quiet(), for which it is a helper function. Allows * a group of CPUs to be quieted at one go, though all the CPUs in the @@ -706,7 +740,7 @@ cpu_quiet_msk(unsigned long mask, struct rcu_state *rsp, struct rcu_node *rnp, return; } rnp->qsmask &= ~mask; - if (rnp->qsmask != 0) { + if (rnp->qsmask != 0 || rcu_preempted_readers(rnp)) { /* Other bits still set at this level, so done. */ spin_unlock_irqrestore(&rnp->lock, flags); @@ -726,14 +760,10 @@ cpu_quiet_msk(unsigned long mask, struct rcu_state *rsp, struct rcu_node *rnp, /* * Get here if we are the last CPU to pass through a quiescent - * state for this grace period. Clean up and let rcu_start_gp() - * start up the next grace period if one is needed. Note that - * we still hold rnp->lock, as required by rcu_start_gp(), which - * will release it. + * state for this grace period. Invoke cpu_quiet_msk_finish() + * to clean up and start the next grace period if one is needed. */ - rsp->completed = rsp->gpnum; - rcu_process_gp_end(rsp, rsp->rda[smp_processor_id()]); - rcu_start_gp(rsp, flags); /* releases rnp->lock. */ + cpu_quiet_msk_finish(rsp, flags); /* releases rnp->lock. */ } /* @@ -840,11 +870,11 @@ static void __rcu_offline_cpu(int cpu, struct rcu_state *rsp) spin_lock(&rnp->lock); /* irqs already disabled. */ rnp->qsmaskinit &= ~mask; if (rnp->qsmaskinit != 0) { - spin_unlock(&rnp->lock); /* irqs already disabled. */ + spin_unlock(&rnp->lock); /* irqs remain disabled. */ break; } mask = rnp->grpmask; - spin_unlock(&rnp->lock); /* irqs already disabled. */ + spin_unlock(&rnp->lock); /* irqs remain disabled. */ rnp = rnp->parent; } while (rnp != NULL); lastcomp = rsp->completed; @@ -1007,6 +1037,7 @@ void rcu_check_callbacks(int cpu, int user) rcu_bh_qs(cpu); } + rcu_preempt_check_callbacks(cpu); raise_softirq(RCU_SOFTIRQ); } @@ -1188,6 +1219,7 @@ static void rcu_process_callbacks(struct softirq_action *unused) __rcu_process_callbacks(&rcu_sched_state, &__get_cpu_var(rcu_sched_data)); __rcu_process_callbacks(&rcu_bh_state, &__get_cpu_var(rcu_bh_data)); + rcu_preempt_process_callbacks(); /* * Memory references from any later RCU read-side critical sections @@ -1251,17 +1283,6 @@ void call_rcu_sched(struct rcu_head *head, void (*func)(struct rcu_head *rcu)) } EXPORT_SYMBOL_GPL(call_rcu_sched); -/* - * @@@ Queue an RCU callback for invocation after a grace period. - * @@@ Placeholder pending rcutree_plugin.h. - */ -void call_rcu(struct rcu_head *head, void (*func)(struct rcu_head *rcu)) -{ - call_rcu_sched(head, func); -} -EXPORT_SYMBOL_GPL(call_rcu); - - /* * Queue an RCU for invocation after a quicker grace period. */ @@ -1335,7 +1356,8 @@ static int __rcu_pending(struct rcu_state *rsp, struct rcu_data *rdp) static int rcu_pending(int cpu) { return __rcu_pending(&rcu_sched_state, &per_cpu(rcu_sched_data, cpu)) || - __rcu_pending(&rcu_bh_state, &per_cpu(rcu_bh_data, cpu)); + __rcu_pending(&rcu_bh_state, &per_cpu(rcu_bh_data, cpu)) || + rcu_preempt_pending(cpu); } /* @@ -1348,7 +1370,8 @@ int rcu_needs_cpu(int cpu) { /* RCU callbacks either ready or pending? */ return per_cpu(rcu_sched_data, cpu).nxtlist || - per_cpu(rcu_bh_data, cpu).nxtlist; + per_cpu(rcu_bh_data, cpu).nxtlist || + rcu_preempt_needs_cpu(cpu); } /* @@ -1383,7 +1406,7 @@ rcu_boot_init_percpu_data(int cpu, struct rcu_state *rsp) * that this CPU cannot possibly have any RCU callbacks in flight yet. */ static void __cpuinit -rcu_init_percpu_data(int cpu, struct rcu_state *rsp) +rcu_init_percpu_data(int cpu, struct rcu_state *rsp, int preemptable) { unsigned long flags; long lastcomp; @@ -1399,6 +1422,7 @@ rcu_init_percpu_data(int cpu, struct rcu_state *rsp) rdp->passed_quiesc = 0; /* We could be racing with new GP, */ rdp->qs_pending = 1; /* so set up to respond to current GP. */ rdp->beenonline = 1; /* We have now been online. */ + rdp->preemptable = preemptable; rdp->passed_quiesc_completed = lastcomp - 1; rdp->blimit = blimit; spin_unlock(&rnp->lock); /* irqs remain disabled. */ @@ -1441,12 +1465,13 @@ rcu_init_percpu_data(int cpu, struct rcu_state *rsp) static void __cpuinit rcu_online_cpu(int cpu) { - rcu_init_percpu_data(cpu, &rcu_sched_state); - rcu_init_percpu_data(cpu, &rcu_bh_state); + rcu_init_percpu_data(cpu, &rcu_sched_state, 0); + rcu_init_percpu_data(cpu, &rcu_bh_state, 0); + rcu_preempt_init_percpu_data(cpu); } /* - * Handle CPU online/offline notifcation events. + * Handle CPU online/offline notification events. */ int __cpuinit rcu_cpu_notify(struct notifier_block *self, unsigned long action, void *hcpu) @@ -1521,6 +1546,7 @@ static void __init rcu_init_one(struct rcu_state *rsp) rnp = rsp->level[i]; for (j = 0; j < rsp->levelcnt[i]; j++, rnp++) { spin_lock_init(&rnp->lock); + rnp->gpnum = 0; rnp->qsmask = 0; rnp->qsmaskinit = 0; rnp->grplo = j * cpustride; @@ -1538,13 +1564,16 @@ static void __init rcu_init_one(struct rcu_state *rsp) j / rsp->levelspread[i - 1]; } rnp->level = i; + INIT_LIST_HEAD(&rnp->blocked_tasks[0]); + INIT_LIST_HEAD(&rnp->blocked_tasks[1]); } } } /* - * Helper macro for __rcu_init(). To be used nowhere else! - * Assigns leaf node pointers into each CPU's rcu_data structure. + * Helper macro for __rcu_init() and __rcu_init_preempt(). To be used + * nowhere else! Assigns leaf node pointers into each CPU's rcu_data + * structure. */ #define RCU_INIT_FLAVOR(rsp, rcu_data) \ do { \ @@ -1560,18 +1589,38 @@ do { \ } \ } while (0) +#ifdef CONFIG_TREE_PREEMPT_RCU + +void __init __rcu_init_preempt(void) +{ + int i; /* All used by RCU_INIT_FLAVOR(). */ + int j; + struct rcu_node *rnp; + + RCU_INIT_FLAVOR(&rcu_preempt_state, rcu_preempt_data); +} + +#else /* #ifdef CONFIG_TREE_PREEMPT_RCU */ + +void __init __rcu_init_preempt(void) +{ +} + +#endif /* #else #ifdef CONFIG_TREE_PREEMPT_RCU */ + void __init __rcu_init(void) { - int i; /* All used by RCU_DATA_PTR_INIT(). */ + int i; /* All used by RCU_INIT_FLAVOR(). */ int j; struct rcu_node *rnp; - printk(KERN_INFO "Hierarchical RCU implementation.\n"); + rcu_bootup_announce(); #ifdef CONFIG_RCU_CPU_STALL_DETECTOR printk(KERN_INFO "RCU-based detection of stalled CPUs is enabled.\n"); #endif /* #ifdef CONFIG_RCU_CPU_STALL_DETECTOR */ RCU_INIT_FLAVOR(&rcu_sched_state, rcu_sched_data); RCU_INIT_FLAVOR(&rcu_bh_state, rcu_bh_data); + __rcu_init_preempt(); open_softirq(RCU_SOFTIRQ, rcu_process_callbacks); } diff --git a/kernel/rcutree.h b/kernel/rcutree.h index 0024e5ddcc68..ca560364d8cd 100644 --- a/kernel/rcutree.h +++ b/kernel/rcutree.h @@ -80,6 +80,7 @@ struct rcu_dynticks { */ struct rcu_node { spinlock_t lock; + long gpnum; /* Current grace period for this node. */ unsigned long qsmask; /* CPUs or groups that need to switch in */ /* order for current grace period to proceed.*/ unsigned long qsmaskinit; @@ -90,6 +91,8 @@ struct rcu_node { u8 grpnum; /* CPU/group number for next level up. */ u8 level; /* root is at level 0. */ struct rcu_node *parent; + struct list_head blocked_tasks[2]; + /* Tasks blocked in RCU read-side critsect. */ } ____cacheline_internodealigned_in_smp; /* Index values for nxttail array in struct rcu_data. */ @@ -111,6 +114,7 @@ struct rcu_data { bool passed_quiesc; /* User-mode/idle loop etc. */ bool qs_pending; /* Core waits for quiesc state. */ bool beenonline; /* CPU online at least once. */ + bool preemptable; /* Preemptable RCU? */ struct rcu_node *mynode; /* This CPU's leaf of hierarchy */ unsigned long grpmask; /* Mask to apply to leaf qsmask. */ @@ -244,5 +248,10 @@ DECLARE_PER_CPU(struct rcu_data, rcu_sched_data); extern struct rcu_state rcu_bh_state; DECLARE_PER_CPU(struct rcu_data, rcu_bh_data); +#ifdef CONFIG_TREE_PREEMPT_RCU +extern struct rcu_state rcu_preempt_state; +DECLARE_PER_CPU(struct rcu_data, rcu_preempt_data); +#endif /* #ifdef CONFIG_TREE_PREEMPT_RCU */ + #endif /* #ifdef RCU_TREE_NONCORE */ diff --git a/kernel/rcutree_plugin.h b/kernel/rcutree_plugin.h new file mode 100644 index 000000000000..cd2ab67400c6 --- /dev/null +++ b/kernel/rcutree_plugin.h @@ -0,0 +1,447 @@ +/* + * Read-Copy Update mechanism for mutual exclusion (tree-based version) + * Internal non-public definitions that provide either classic + * or preemptable semantics. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. + * + * Copyright Red Hat, 2009 + * Copyright IBM Corporation, 2009 + * + * Author: Ingo Molnar + * Paul E. McKenney + */ + + +#ifdef CONFIG_TREE_PREEMPT_RCU + +struct rcu_state rcu_preempt_state = RCU_STATE_INITIALIZER(rcu_preempt_state); +DEFINE_PER_CPU(struct rcu_data, rcu_preempt_data); + +/* + * Tell them what RCU they are running. + */ +static inline void rcu_bootup_announce(void) +{ + printk(KERN_INFO + "Experimental preemptable hierarchical RCU implementation.\n"); +} + +/* + * Return the number of RCU-preempt batches processed thus far + * for debug and statistics. + */ +long rcu_batches_completed_preempt(void) +{ + return rcu_preempt_state.completed; +} +EXPORT_SYMBOL_GPL(rcu_batches_completed_preempt); + +/* + * Return the number of RCU batches processed thus far for debug & stats. + */ +long rcu_batches_completed(void) +{ + return rcu_batches_completed_preempt(); +} +EXPORT_SYMBOL_GPL(rcu_batches_completed); + +/* + * Record a preemptable-RCU quiescent state for the specified CPU. Note + * that this just means that the task currently running on the CPU is + * not in a quiescent state. There might be any number of tasks blocked + * while in an RCU read-side critical section. + */ +static void rcu_preempt_qs_record(int cpu) +{ + struct rcu_data *rdp = &per_cpu(rcu_preempt_data, cpu); + rdp->passed_quiesc = 1; + rdp->passed_quiesc_completed = rdp->completed; +} + +/* + * We have entered the scheduler or are between softirqs in ksoftirqd. + * If we are in an RCU read-side critical section, we need to reflect + * that in the state of the rcu_node structure corresponding to this CPU. + * Caller must disable hardirqs. + */ +static void rcu_preempt_qs(int cpu) +{ + struct task_struct *t = current; + int phase; + struct rcu_data *rdp; + struct rcu_node *rnp; + + if (t->rcu_read_lock_nesting && + (t->rcu_read_unlock_special & RCU_READ_UNLOCK_BLOCKED) == 0) { + + /* Possibly blocking in an RCU read-side critical section. */ + rdp = rcu_preempt_state.rda[cpu]; + rnp = rdp->mynode; + spin_lock(&rnp->lock); + t->rcu_read_unlock_special |= RCU_READ_UNLOCK_BLOCKED; + t->rcu_blocked_cpu = cpu; + + /* + * If this CPU has already checked in, then this task + * will hold up the next grace period rather than the + * current grace period. Queue the task accordingly. + * If the task is queued for the current grace period + * (i.e., this CPU has not yet passed through a quiescent + * state for the current grace period), then as long + * as that task remains queued, the current grace period + * cannot end. + */ + phase = !(rnp->qsmask & rdp->grpmask) ^ (rnp->gpnum & 0x1); + list_add(&t->rcu_node_entry, &rnp->blocked_tasks[phase]); + smp_mb(); /* Ensure later ctxt swtch seen after above. */ + spin_unlock(&rnp->lock); + } + + /* + * Either we were not in an RCU read-side critical section to + * begin with, or we have now recorded that critical section + * globally. Either way, we can now note a quiescent state + * for this CPU. Again, if we were in an RCU read-side critical + * section, and if that critical section was blocking the current + * grace period, then the fact that the task has been enqueued + * means that we continue to block the current grace period. + */ + rcu_preempt_qs_record(cpu); + t->rcu_read_unlock_special &= ~(RCU_READ_UNLOCK_NEED_QS | + RCU_READ_UNLOCK_GOT_QS); +} + +/* + * Tree-preemptable RCU implementation for rcu_read_lock(). + * Just increment ->rcu_read_lock_nesting, shared state will be updated + * if we block. + */ +void __rcu_read_lock(void) +{ + ACCESS_ONCE(current->rcu_read_lock_nesting)++; + barrier(); /* needed if we ever invoke rcu_read_lock in rcutree.c */ +} +EXPORT_SYMBOL_GPL(__rcu_read_lock); + +static void rcu_read_unlock_special(struct task_struct *t) +{ + int empty; + unsigned long flags; + unsigned long mask; + struct rcu_node *rnp; + int special; + + /* NMI handlers cannot block and cannot safely manipulate state. */ + if (in_nmi()) + return; + + local_irq_save(flags); + + /* + * If RCU core is waiting for this CPU to exit critical section, + * let it know that we have done so. + */ + special = t->rcu_read_unlock_special; + if (special & RCU_READ_UNLOCK_NEED_QS) { + t->rcu_read_unlock_special &= ~RCU_READ_UNLOCK_NEED_QS; + t->rcu_read_unlock_special |= RCU_READ_UNLOCK_GOT_QS; + } + + /* Hardware IRQ handlers cannot block. */ + if (in_irq()) { + local_irq_restore(flags); + return; + } + + /* Clean up if blocked during RCU read-side critical section. */ + if (special & RCU_READ_UNLOCK_BLOCKED) { + t->rcu_read_unlock_special &= ~RCU_READ_UNLOCK_BLOCKED; + + /* Remove this task from the list it blocked on. */ + rnp = rcu_preempt_state.rda[t->rcu_blocked_cpu]->mynode; + spin_lock(&rnp->lock); + empty = list_empty(&rnp->blocked_tasks[rnp->gpnum & 0x1]); + list_del_init(&t->rcu_node_entry); + t->rcu_blocked_cpu = -1; + + /* + * If this was the last task on the current list, and if + * we aren't waiting on any CPUs, report the quiescent state. + * Note that both cpu_quiet_msk_finish() and cpu_quiet_msk() + * drop rnp->lock and restore irq. + */ + if (!empty && rnp->qsmask == 0 && + list_empty(&rnp->blocked_tasks[rnp->gpnum & 0x1])) { + t->rcu_read_unlock_special &= + ~(RCU_READ_UNLOCK_NEED_QS | + RCU_READ_UNLOCK_GOT_QS); + if (rnp->parent == NULL) { + /* Only one rcu_node in the tree. */ + cpu_quiet_msk_finish(&rcu_preempt_state, flags); + return; + } + /* Report up the rest of the hierarchy. */ + mask = rnp->grpmask; + spin_unlock_irqrestore(&rnp->lock, flags); + rnp = rnp->parent; + spin_lock_irqsave(&rnp->lock, flags); + cpu_quiet_msk(mask, &rcu_preempt_state, rnp, flags); + return; + } + spin_unlock(&rnp->lock); + } + local_irq_restore(flags); +} + +/* + * Tree-preemptable RCU implementation for rcu_read_unlock(). + * Decrement ->rcu_read_lock_nesting. If the result is zero (outermost + * rcu_read_unlock()) and ->rcu_read_unlock_special is non-zero, then + * invoke rcu_read_unlock_special() to clean up after a context switch + * in an RCU read-side critical section and other special cases. + */ +void __rcu_read_unlock(void) +{ + struct task_struct *t = current; + + barrier(); /* needed if we ever invoke rcu_read_unlock in rcutree.c */ + if (--ACCESS_ONCE(t->rcu_read_lock_nesting) == 0 && + unlikely(ACCESS_ONCE(t->rcu_read_unlock_special))) + rcu_read_unlock_special(t); +} +EXPORT_SYMBOL_GPL(__rcu_read_unlock); + +#ifdef CONFIG_RCU_CPU_STALL_DETECTOR + +/* + * Scan the current list of tasks blocked within RCU read-side critical + * sections, printing out the tid of each. + */ +static void rcu_print_task_stall(struct rcu_node *rnp) +{ + unsigned long flags; + struct list_head *lp; + int phase = rnp->gpnum & 0x1; + struct task_struct *t; + + if (!list_empty(&rnp->blocked_tasks[phase])) { + spin_lock_irqsave(&rnp->lock, flags); + phase = rnp->gpnum & 0x1; /* re-read under lock. */ + lp = &rnp->blocked_tasks[phase]; + list_for_each_entry(t, lp, rcu_node_entry) + printk(" P%d", t->pid); + spin_unlock_irqrestore(&rnp->lock, flags); + } +} + +#endif /* #ifdef CONFIG_RCU_CPU_STALL_DETECTOR */ + +/* + * Check for preempted RCU readers for the specified rcu_node structure. + * If the caller needs a reliable answer, it must hold the rcu_node's + * >lock. + */ +static int rcu_preempted_readers(struct rcu_node *rnp) +{ + return !list_empty(&rnp->blocked_tasks[rnp->gpnum & 0x1]); +} + +/* + * Check for a quiescent state from the current CPU. When a task blocks, + * the task is recorded in the corresponding CPU's rcu_node structure, + * which is checked elsewhere. + * + * Caller must disable hard irqs. + */ +static void rcu_preempt_check_callbacks(int cpu) +{ + struct task_struct *t = current; + + if (t->rcu_read_lock_nesting == 0) { + t->rcu_read_unlock_special &= + ~(RCU_READ_UNLOCK_NEED_QS | RCU_READ_UNLOCK_GOT_QS); + rcu_preempt_qs_record(cpu); + return; + } + if (per_cpu(rcu_preempt_data, cpu).qs_pending) { + if (t->rcu_read_unlock_special & RCU_READ_UNLOCK_GOT_QS) { + rcu_preempt_qs_record(cpu); + t->rcu_read_unlock_special &= ~RCU_READ_UNLOCK_GOT_QS; + } else if (!(t->rcu_read_unlock_special & + RCU_READ_UNLOCK_NEED_QS)) { + t->rcu_read_unlock_special |= RCU_READ_UNLOCK_NEED_QS; + } + } +} + +/* + * Process callbacks for preemptable RCU. + */ +static void rcu_preempt_process_callbacks(void) +{ + __rcu_process_callbacks(&rcu_preempt_state, + &__get_cpu_var(rcu_preempt_data)); +} + +/* + * Queue a preemptable-RCU callback for invocation after a grace period. + */ +void call_rcu(struct rcu_head *head, void (*func)(struct rcu_head *rcu)) +{ + __call_rcu(head, func, &rcu_preempt_state); +} +EXPORT_SYMBOL_GPL(call_rcu); + +/* + * Check to see if there is any immediate preemptable-RCU-related work + * to be done. + */ +static int rcu_preempt_pending(int cpu) +{ + return __rcu_pending(&rcu_preempt_state, + &per_cpu(rcu_preempt_data, cpu)); +} + +/* + * Does preemptable RCU need the CPU to stay out of dynticks mode? + */ +static int rcu_preempt_needs_cpu(int cpu) +{ + return !!per_cpu(rcu_preempt_data, cpu).nxtlist; +} + +/* + * Initialize preemptable RCU's per-CPU data. + */ +static void __cpuinit rcu_preempt_init_percpu_data(int cpu) +{ + rcu_init_percpu_data(cpu, &rcu_preempt_state, 1); +} + +/* + * Check for a task exiting while in a preemptable-RCU read-side + * critical section, clean up if so. No need to issue warnings, + * as debug_check_no_locks_held() already does this if lockdep + * is enabled. + */ +void exit_rcu(void) +{ + struct task_struct *t = current; + + if (t->rcu_read_lock_nesting == 0) + return; + t->rcu_read_lock_nesting = 1; + rcu_read_unlock(); +} + +#else /* #ifdef CONFIG_TREE_PREEMPT_RCU */ + +/* + * Tell them what RCU they are running. + */ +static inline void rcu_bootup_announce(void) +{ + printk(KERN_INFO "Hierarchical RCU implementation.\n"); +} + +/* + * Return the number of RCU batches processed thus far for debug & stats. + */ +long rcu_batches_completed(void) +{ + return rcu_batches_completed_sched(); +} +EXPORT_SYMBOL_GPL(rcu_batches_completed); + +/* + * Because preemptable RCU does not exist, we never have to check for + * CPUs being in quiescent states. + */ +static void rcu_preempt_qs(int cpu) +{ +} + +#ifdef CONFIG_RCU_CPU_STALL_DETECTOR + +/* + * Because preemptable RCU does not exist, we never have to check for + * tasks blocked within RCU read-side critical sections. + */ +static void rcu_print_task_stall(struct rcu_node *rnp) +{ +} + +#endif /* #ifdef CONFIG_RCU_CPU_STALL_DETECTOR */ + +/* + * Because preemptable RCU does not exist, there are never any preempted + * RCU readers. + */ +static int rcu_preempted_readers(struct rcu_node *rnp) +{ + return 0; +} + +/* + * Because preemptable RCU does not exist, it never has any callbacks + * to check. + */ +void rcu_preempt_check_callbacks(int cpu) +{ +} + +/* + * Because preemptable RCU does not exist, it never has any callbacks + * to process. + */ +void rcu_preempt_process_callbacks(void) +{ +} + +/* + * In classic RCU, call_rcu() is just call_rcu_sched(). + */ +void call_rcu(struct rcu_head *head, void (*func)(struct rcu_head *rcu)) +{ + call_rcu_sched(head, func); +} +EXPORT_SYMBOL_GPL(call_rcu); + +/* + * Because preemptable RCU does not exist, it never has any work to do. + */ +static int rcu_preempt_pending(int cpu) +{ + return 0; +} + +/* + * Because preemptable RCU does not exist, it never needs any CPU. + */ +static int rcu_preempt_needs_cpu(int cpu) +{ + return 0; +} + +/* + * Because preemptable RCU does not exist, there is no per-CPU + * data to initialize. + */ +static void __cpuinit rcu_preempt_init_percpu_data(int cpu) +{ +} + +#endif /* #else #ifdef CONFIG_TREE_PREEMPT_RCU */ diff --git a/kernel/rcutree_trace.c b/kernel/rcutree_trace.c index 31af3a0fb6d5..0ea1bff69727 100644 --- a/kernel/rcutree_trace.c +++ b/kernel/rcutree_trace.c @@ -77,6 +77,10 @@ static void print_one_rcu_data(struct seq_file *m, struct rcu_data *rdp) static int show_rcudata(struct seq_file *m, void *unused) { +#ifdef CONFIG_TREE_PREEMPT_RCU + seq_puts(m, "rcu_preempt:\n"); + PRINT_RCU_DATA(rcu_preempt_data, print_one_rcu_data, m); +#endif /* #ifdef CONFIG_TREE_PREEMPT_RCU */ seq_puts(m, "rcu_sched:\n"); PRINT_RCU_DATA(rcu_sched_data, print_one_rcu_data, m); seq_puts(m, "rcu_bh:\n"); @@ -125,6 +129,10 @@ static int show_rcudata_csv(struct seq_file *m, void *unused) seq_puts(m, "\"dt\",\"dt nesting\",\"dn\",\"df\","); #endif /* #ifdef CONFIG_NO_HZ */ seq_puts(m, "\"of\",\"ri\",\"ql\",\"b\"\n"); +#ifdef CONFIG_TREE_PREEMPT_RCU + seq_puts(m, "\"rcu_preempt:\"\n"); + PRINT_RCU_DATA(rcu_preempt_data, print_one_rcu_data_csv, m); +#endif /* #ifdef CONFIG_TREE_PREEMPT_RCU */ seq_puts(m, "\"rcu_sched:\"\n"); PRINT_RCU_DATA(rcu_sched_data, print_one_rcu_data_csv, m); seq_puts(m, "\"rcu_bh:\"\n"); @@ -172,6 +180,10 @@ static void print_one_rcu_state(struct seq_file *m, struct rcu_state *rsp) static int show_rcuhier(struct seq_file *m, void *unused) { +#ifdef CONFIG_TREE_PREEMPT_RCU + seq_puts(m, "rcu_preempt:\n"); + print_one_rcu_state(m, &rcu_preempt_state); +#endif /* #ifdef CONFIG_TREE_PREEMPT_RCU */ seq_puts(m, "rcu_sched:\n"); print_one_rcu_state(m, &rcu_sched_state); seq_puts(m, "rcu_bh:\n"); @@ -194,6 +206,10 @@ static struct file_operations rcuhier_fops = { static int show_rcugp(struct seq_file *m, void *unused) { +#ifdef CONFIG_TREE_PREEMPT_RCU + seq_printf(m, "rcu_preempt: completed=%ld gpnum=%ld\n", + rcu_preempt_state.completed, rcu_preempt_state.gpnum); +#endif /* #ifdef CONFIG_TREE_PREEMPT_RCU */ seq_printf(m, "rcu_sched: completed=%ld gpnum=%ld\n", rcu_sched_state.completed, rcu_sched_state.gpnum); seq_printf(m, "rcu_bh: completed=%ld gpnum=%ld\n", @@ -244,6 +260,10 @@ static void print_rcu_pendings(struct seq_file *m, struct rcu_state *rsp) static int show_rcu_pending(struct seq_file *m, void *unused) { +#ifdef CONFIG_TREE_PREEMPT_RCU + seq_puts(m, "rcu_preempt:\n"); + print_rcu_pendings(m, &rcu_preempt_state); +#endif /* #ifdef CONFIG_TREE_PREEMPT_RCU */ seq_puts(m, "rcu_sched:\n"); print_rcu_pendings(m, &rcu_sched_state); seq_puts(m, "rcu_bh:\n"); diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index 12327b2bb785..f87fb0c8f924 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -725,7 +725,7 @@ config RCU_TORTURE_TEST_RUNNABLE config RCU_CPU_STALL_DETECTOR bool "Check for stalled CPUs delaying RCU grace periods" - depends on CLASSIC_RCU || TREE_RCU + depends on CLASSIC_RCU || TREE_RCU || TREE_PREEMPT_RCU default n help This option causes RCU to printk information on which -- cgit v1.2.3 From 6b3ef48adf847f7adf11c870e3ffacac150f1564 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Sat, 22 Aug 2009 13:56:53 -0700 Subject: rcu: Remove CONFIG_PREEMPT_RCU Now that CONFIG_TREE_PREEMPT_RCU is in place, there is no further need for CONFIG_PREEMPT_RCU. Remove it, along with whatever subtle bugs it may (or may not) contain. Signed-off-by: Paul E. McKenney Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <125097461396-git-send-email-> Signed-off-by: Ingo Molnar --- Documentation/RCU/rcu.txt | 10 +- Documentation/RCU/whatisRCU.txt | 8 +- include/linux/init_task.h | 6 +- include/linux/rcupdate.h | 4 +- include/linux/rcupreempt.h | 140 ---- include/linux/rcupreempt_trace.h | 97 --- include/linux/sched.h | 13 - init/Kconfig | 20 +- kernel/Makefile | 2 - kernel/rcupreempt.c | 1518 -------------------------------------- kernel/rcupreempt_trace.c | 335 --------- lib/Kconfig.debug | 2 +- 12 files changed, 13 insertions(+), 2142 deletions(-) delete mode 100644 include/linux/rcupreempt.h delete mode 100644 include/linux/rcupreempt_trace.h delete mode 100644 kernel/rcupreempt.c delete mode 100644 kernel/rcupreempt_trace.c (limited to 'include') diff --git a/Documentation/RCU/rcu.txt b/Documentation/RCU/rcu.txt index 7aa2002ade77..2a23523ce471 100644 --- a/Documentation/RCU/rcu.txt +++ b/Documentation/RCU/rcu.txt @@ -36,7 +36,7 @@ o How can the updater tell when a grace period has completed executed in user mode, or executed in the idle loop, we can safely free up that item. - Preemptible variants of RCU (CONFIG_PREEMPT_RCU) get the + Preemptible variants of RCU (CONFIG_TREE_PREEMPT_RCU) get the same effect, but require that the readers manipulate CPU-local counters. These counters allow limited types of blocking within RCU read-side critical sections. SRCU also uses @@ -79,10 +79,10 @@ o I hear that RCU is patented? What is with that? o I hear that RCU needs work in order to support realtime kernels? This work is largely completed. Realtime-friendly RCU can be - enabled via the CONFIG_PREEMPT_RCU kernel configuration parameter. - However, work is in progress for enabling priority boosting of - preempted RCU read-side critical sections. This is needed if you - have CPU-bound realtime threads. + enabled via the CONFIG_TREE_PREEMPT_RCU kernel configuration + parameter. However, work is in progress for enabling priority + boosting of preempted RCU read-side critical sections. This is + needed if you have CPU-bound realtime threads. o Where can I find more information on RCU? diff --git a/Documentation/RCU/whatisRCU.txt b/Documentation/RCU/whatisRCU.txt index 97ded2432c59..e41a7fecf0d3 100644 --- a/Documentation/RCU/whatisRCU.txt +++ b/Documentation/RCU/whatisRCU.txt @@ -136,10 +136,10 @@ rcu_read_lock() Used by a reader to inform the reclaimer that the reader is entering an RCU read-side critical section. It is illegal to block while in an RCU read-side critical section, though - kernels built with CONFIG_PREEMPT_RCU can preempt RCU read-side - critical sections. Any RCU-protected data structure accessed - during an RCU read-side critical section is guaranteed to remain - unreclaimed for the full duration of that critical section. + kernels built with CONFIG_TREE_PREEMPT_RCU can preempt RCU + read-side critical sections. Any RCU-protected data structure + accessed during an RCU read-side critical section is guaranteed to + remain unreclaimed for the full duration of that critical section. Reference counts may be used in conjunction with RCU to maintain longer-term references to data structures. diff --git a/include/linux/init_task.h b/include/linux/init_task.h index 971a968831bf..79d4baee31b6 100644 --- a/include/linux/init_task.h +++ b/include/linux/init_task.h @@ -94,11 +94,7 @@ extern struct group_info init_groups; # define CAP_INIT_BSET CAP_INIT_EFF_SET #endif -#ifdef CONFIG_PREEMPT_RCU -#define INIT_TASK_RCU_PREEMPT(tsk) \ - .rcu_read_lock_nesting = 0, \ - .rcu_flipctr_idx = 0, -#elif defined(CONFIG_TREE_PREEMPT_RCU) +#ifdef CONFIG_TREE_PREEMPT_RCU #define INIT_TASK_RCU_PREEMPT(tsk) \ .rcu_read_lock_nesting = 0, \ .rcu_read_unlock_special = 0, \ diff --git a/include/linux/rcupdate.h b/include/linux/rcupdate.h index 26892f5e7bd8..ec90fc34fea9 100644 --- a/include/linux/rcupdate.h +++ b/include/linux/rcupdate.h @@ -68,11 +68,9 @@ extern int rcu_scheduler_active; #if defined(CONFIG_TREE_RCU) || defined(CONFIG_TREE_PREEMPT_RCU) #include -#elif defined(CONFIG_PREEMPT_RCU) -#include #else #error "Unknown RCU implementation specified to kernel configuration" -#endif /* #else #if defined(CONFIG_CLASSIC_RCU) */ +#endif #define RCU_HEAD_INIT { .next = NULL, .func = NULL } #define RCU_HEAD(head) struct rcu_head head = RCU_HEAD_INIT diff --git a/include/linux/rcupreempt.h b/include/linux/rcupreempt.h deleted file mode 100644 index a42ab88e9210..000000000000 --- a/include/linux/rcupreempt.h +++ /dev/null @@ -1,140 +0,0 @@ -/* - * Read-Copy Update mechanism for mutual exclusion (RT implementation) - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. - * - * Copyright (C) IBM Corporation, 2006 - * - * Author: Paul McKenney - * - * Based on the original work by Paul McKenney - * and inputs from Rusty Russell, Andrea Arcangeli and Andi Kleen. - * Papers: - * http://www.rdrop.com/users/paulmck/paper/rclockpdcsproof.pdf - * http://lse.sourceforge.net/locking/rclock_OLS.2001.05.01c.sc.pdf (OLS2001) - * - * For detailed explanation of Read-Copy Update mechanism see - - * Documentation/RCU - * - */ - -#ifndef __LINUX_RCUPREEMPT_H -#define __LINUX_RCUPREEMPT_H - -#include -#include -#include -#include -#include -#include - -extern void rcu_sched_qs(int cpu); -static inline void rcu_bh_qs(int cpu) { } - -/* - * Someone might want to pass call_rcu_bh as a function pointer. - * So this needs to just be a rename and not a macro function. - * (no parentheses) - */ -#define call_rcu_bh call_rcu - -/** - * call_rcu_sched - Queue RCU callback for invocation after sched grace period. - * @head: structure to be used for queueing the RCU updates. - * @func: actual update function to be invoked after the grace period - * - * The update function will be invoked some time after a full - * synchronize_sched()-style grace period elapses, in other words after - * all currently executing preempt-disabled sections of code (including - * hardirq handlers, NMI handlers, and local_irq_save() blocks) have - * completed. - */ -extern void call_rcu_sched(struct rcu_head *head, - void (*func)(struct rcu_head *head)); - -extern void __rcu_read_lock(void); -extern void __rcu_read_unlock(void); -extern int rcu_needs_cpu(int cpu); - -#define __rcu_read_lock_bh() { rcu_read_lock(); local_bh_disable(); } -#define __rcu_read_unlock_bh() { local_bh_enable(); rcu_read_unlock(); } - -extern void __synchronize_sched(void); - -static inline void synchronize_rcu_expedited(void) -{ - synchronize_rcu(); /* Placeholder for new rcupreempt implementation. */ -} - -static inline void synchronize_rcu_bh_expedited(void) -{ - synchronize_rcu_bh(); /* Placeholder for new rcupreempt impl. */ -} - -extern void __rcu_init(void); -extern void rcu_init_sched(void); -extern void rcu_check_callbacks(int cpu, int user); -extern void rcu_restart_cpu(int cpu); -extern long rcu_batches_completed(void); - -/* - * Return the number of RCU batches processed thus far. Useful for debug - * and statistic. The _bh variant is identifcal to straight RCU - */ -static inline long rcu_batches_completed_bh(void) -{ - return rcu_batches_completed(); -} - -static inline void exit_rcu(void) -{ -} - -#ifdef CONFIG_RCU_TRACE -struct rcupreempt_trace; -extern long *rcupreempt_flipctr(int cpu); -extern long rcupreempt_data_completed(void); -extern int rcupreempt_flip_flag(int cpu); -extern int rcupreempt_mb_flag(int cpu); -extern char *rcupreempt_try_flip_state_name(void); -extern struct rcupreempt_trace *rcupreempt_trace_cpu(int cpu); -#endif - -struct softirq_action; - -#ifdef CONFIG_NO_HZ -extern void rcu_enter_nohz(void); -extern void rcu_exit_nohz(void); -#else -# define rcu_enter_nohz() do { } while (0) -# define rcu_exit_nohz() do { } while (0) -#endif - -/* - * A context switch is a grace period for rcupreempt synchronize_rcu() - * only during early boot, before the scheduler has been initialized. - * So, how the heck do we get a context switch? Well, if the caller - * invokes synchronize_rcu(), they are willing to accept a context - * switch, so we simply pretend that one happened. - * - * After boot, there might be a blocked or preempted task in an RCU - * read-side critical section, so we cannot then take the fastpath. - */ -static inline int rcu_blocking_is_gp(void) -{ - return num_online_cpus() == 1 && !rcu_scheduler_active; -} - -#endif /* __LINUX_RCUPREEMPT_H */ diff --git a/include/linux/rcupreempt_trace.h b/include/linux/rcupreempt_trace.h deleted file mode 100644 index b99ae073192a..000000000000 --- a/include/linux/rcupreempt_trace.h +++ /dev/null @@ -1,97 +0,0 @@ -/* - * Read-Copy Update mechanism for mutual exclusion (RT implementation) - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. - * - * Copyright (C) IBM Corporation, 2006 - * - * Author: Paul McKenney - * - * Based on the original work by Paul McKenney - * and inputs from Rusty Russell, Andrea Arcangeli and Andi Kleen. - * Papers: - * http://www.rdrop.com/users/paulmck/paper/rclockpdcsproof.pdf - * http://lse.sourceforge.net/locking/rclock_OLS.2001.05.01c.sc.pdf (OLS2001) - * - * For detailed explanation of the Preemptible Read-Copy Update mechanism see - - * http://lwn.net/Articles/253651/ - */ - -#ifndef __LINUX_RCUPREEMPT_TRACE_H -#define __LINUX_RCUPREEMPT_TRACE_H - -#include -#include - -#include - -/* - * PREEMPT_RCU data structures. - */ - -struct rcupreempt_trace { - long next_length; - long next_add; - long wait_length; - long wait_add; - long done_length; - long done_add; - long done_remove; - atomic_t done_invoked; - long rcu_check_callbacks; - atomic_t rcu_try_flip_1; - atomic_t rcu_try_flip_e1; - long rcu_try_flip_i1; - long rcu_try_flip_ie1; - long rcu_try_flip_g1; - long rcu_try_flip_a1; - long rcu_try_flip_ae1; - long rcu_try_flip_a2; - long rcu_try_flip_z1; - long rcu_try_flip_ze1; - long rcu_try_flip_z2; - long rcu_try_flip_m1; - long rcu_try_flip_me1; - long rcu_try_flip_m2; -}; - -#ifdef CONFIG_RCU_TRACE -#define RCU_TRACE(fn, arg) fn(arg); -#else -#define RCU_TRACE(fn, arg) -#endif - -extern void rcupreempt_trace_move2done(struct rcupreempt_trace *trace); -extern void rcupreempt_trace_move2wait(struct rcupreempt_trace *trace); -extern void rcupreempt_trace_try_flip_1(struct rcupreempt_trace *trace); -extern void rcupreempt_trace_try_flip_e1(struct rcupreempt_trace *trace); -extern void rcupreempt_trace_try_flip_i1(struct rcupreempt_trace *trace); -extern void rcupreempt_trace_try_flip_ie1(struct rcupreempt_trace *trace); -extern void rcupreempt_trace_try_flip_g1(struct rcupreempt_trace *trace); -extern void rcupreempt_trace_try_flip_a1(struct rcupreempt_trace *trace); -extern void rcupreempt_trace_try_flip_ae1(struct rcupreempt_trace *trace); -extern void rcupreempt_trace_try_flip_a2(struct rcupreempt_trace *trace); -extern void rcupreempt_trace_try_flip_z1(struct rcupreempt_trace *trace); -extern void rcupreempt_trace_try_flip_ze1(struct rcupreempt_trace *trace); -extern void rcupreempt_trace_try_flip_z2(struct rcupreempt_trace *trace); -extern void rcupreempt_trace_try_flip_m1(struct rcupreempt_trace *trace); -extern void rcupreempt_trace_try_flip_me1(struct rcupreempt_trace *trace); -extern void rcupreempt_trace_try_flip_m2(struct rcupreempt_trace *trace); -extern void rcupreempt_trace_check_callbacks(struct rcupreempt_trace *trace); -extern void rcupreempt_trace_done_remove(struct rcupreempt_trace *trace); -extern void rcupreempt_trace_invoke(struct rcupreempt_trace *trace); -extern void rcupreempt_trace_next_add(struct rcupreempt_trace *trace); - -#endif /* __LINUX_RCUPREEMPT_TRACE_H */ diff --git a/include/linux/sched.h b/include/linux/sched.h index d7f98f637a2a..bfca26d63b13 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -1205,11 +1205,6 @@ struct task_struct { unsigned int policy; cpumask_t cpus_allowed; -#ifdef CONFIG_PREEMPT_RCU - int rcu_read_lock_nesting; - int rcu_flipctr_idx; -#endif /* #ifdef CONFIG_PREEMPT_RCU */ - #ifdef CONFIG_TREE_PREEMPT_RCU int rcu_read_lock_nesting; char rcu_read_unlock_special; @@ -1744,14 +1739,6 @@ static inline void rcu_copy_process(struct task_struct *p) INIT_LIST_HEAD(&p->rcu_node_entry); } -#elif defined(CONFIG_PREEMPT_RCU) - -static inline void rcu_copy_process(struct task_struct *p) -{ - p->rcu_read_lock_nesting = 0; - p->rcu_flipctr_idx = 0; -} - #else static inline void rcu_copy_process(struct task_struct *p) diff --git a/init/Kconfig b/init/Kconfig index f88da2d1c1fb..8e8b76d8a272 100644 --- a/init/Kconfig +++ b/init/Kconfig @@ -324,17 +324,6 @@ config TREE_RCU thousands of CPUs. It also scales down nicely to smaller systems. -config PREEMPT_RCU - bool "Preemptible RCU" - depends on PREEMPT - help - This option reduces the latency of the kernel by making certain - RCU sections preemptible. Normally RCU code is non-preemptible, if - this option is selected then read-only RCU sections become - preemptible. This helps latency, but may expose bugs due to - now-naive assumptions about each RCU read-side critical section - remaining on a given CPU through its execution. - config TREE_PREEMPT_RCU bool "Preemptable tree-based hierarchical RCU" depends on PREEMPT @@ -348,7 +337,7 @@ endchoice config RCU_TRACE bool "Enable tracing for RCU" - depends on TREE_RCU || PREEMPT_RCU || TREE_PREEMPT_RCU + depends on TREE_RCU || TREE_PREEMPT_RCU help This option provides tracing in RCU which presents stats in debugfs for debugging RCU implementation. @@ -395,13 +384,6 @@ config TREE_RCU_TRACE TREE_PREEMPT_RCU implementations, permitting Makefile to trivially select kernel/rcutree_trace.c. -config PREEMPT_RCU_TRACE - def_bool RCU_TRACE && PREEMPT_RCU - select DEBUG_FS - help - This option provides tracing for the PREEMPT_RCU implementation, - permitting Makefile to trivially select kernel/rcupreempt_trace.c. - endmenu # "RCU Subsystem" config IKCONFIG diff --git a/kernel/Makefile b/kernel/Makefile index 1a38b4789dda..b833bd5cc127 100644 --- a/kernel/Makefile +++ b/kernel/Makefile @@ -82,9 +82,7 @@ obj-$(CONFIG_SECCOMP) += seccomp.o obj-$(CONFIG_RCU_TORTURE_TEST) += rcutorture.o obj-$(CONFIG_TREE_RCU) += rcutree.o obj-$(CONFIG_TREE_PREEMPT_RCU) += rcutree.o -obj-$(CONFIG_PREEMPT_RCU) += rcupreempt.o obj-$(CONFIG_TREE_RCU_TRACE) += rcutree_trace.o -obj-$(CONFIG_PREEMPT_RCU_TRACE) += rcupreempt_trace.o obj-$(CONFIG_RELAY) += relay.o obj-$(CONFIG_SYSCTL) += utsname_sysctl.o obj-$(CONFIG_TASK_DELAY_ACCT) += delayacct.o diff --git a/kernel/rcupreempt.c b/kernel/rcupreempt.c deleted file mode 100644 index 0053ce56e326..000000000000 --- a/kernel/rcupreempt.c +++ /dev/null @@ -1,1518 +0,0 @@ -/* - * Read-Copy Update mechanism for mutual exclusion, realtime implementation - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. - * - * Copyright IBM Corporation, 2006 - * - * Authors: Paul E. McKenney - * With thanks to Esben Nielsen, Bill Huey, and Ingo Molnar - * for pushing me away from locks and towards counters, and - * to Suparna Bhattacharya for pushing me completely away - * from atomic instructions on the read side. - * - * - Added handling of Dynamic Ticks - * Copyright 2007 - Paul E. Mckenney - * - Steven Rostedt - * - * Papers: http://www.rdrop.com/users/paulmck/RCU - * - * Design Document: http://lwn.net/Articles/253651/ - * - * For detailed explanation of Read-Copy Update mechanism see - - * Documentation/RCU/ *.txt - * - */ -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include - -/* - * PREEMPT_RCU data structures. - */ - -/* - * GP_STAGES specifies the number of times the state machine has - * to go through the all the rcu_try_flip_states (see below) - * in a single Grace Period. - * - * GP in GP_STAGES stands for Grace Period ;) - */ -#define GP_STAGES 2 -struct rcu_data { - spinlock_t lock; /* Protect rcu_data fields. */ - long completed; /* Number of last completed batch. */ - int waitlistcount; - struct rcu_head *nextlist; - struct rcu_head **nexttail; - struct rcu_head *waitlist[GP_STAGES]; - struct rcu_head **waittail[GP_STAGES]; - struct rcu_head *donelist; /* from waitlist & waitschedlist */ - struct rcu_head **donetail; - long rcu_flipctr[2]; - struct rcu_head *nextschedlist; - struct rcu_head **nextschedtail; - struct rcu_head *waitschedlist; - struct rcu_head **waitschedtail; - int rcu_sched_sleeping; -#ifdef CONFIG_RCU_TRACE - struct rcupreempt_trace trace; -#endif /* #ifdef CONFIG_RCU_TRACE */ -}; - -/* - * States for rcu_try_flip() and friends. - */ - -enum rcu_try_flip_states { - - /* - * Stay here if nothing is happening. Flip the counter if somthing - * starts happening. Denoted by "I" - */ - rcu_try_flip_idle_state, - - /* - * Wait here for all CPUs to notice that the counter has flipped. This - * prevents the old set of counters from ever being incremented once - * we leave this state, which in turn is necessary because we cannot - * test any individual counter for zero -- we can only check the sum. - * Denoted by "A". - */ - rcu_try_flip_waitack_state, - - /* - * Wait here for the sum of the old per-CPU counters to reach zero. - * Denoted by "Z". - */ - rcu_try_flip_waitzero_state, - - /* - * Wait here for each of the other CPUs to execute a memory barrier. - * This is necessary to ensure that these other CPUs really have - * completed executing their RCU read-side critical sections, despite - * their CPUs wildly reordering memory. Denoted by "M". - */ - rcu_try_flip_waitmb_state, -}; - -/* - * States for rcu_ctrlblk.rcu_sched_sleep. - */ - -enum rcu_sched_sleep_states { - rcu_sched_not_sleeping, /* Not sleeping, callbacks need GP. */ - rcu_sched_sleep_prep, /* Thinking of sleeping, rechecking. */ - rcu_sched_sleeping, /* Sleeping, awaken if GP needed. */ -}; - -struct rcu_ctrlblk { - spinlock_t fliplock; /* Protect state-machine transitions. */ - long completed; /* Number of last completed batch. */ - enum rcu_try_flip_states rcu_try_flip_state; /* The current state of - the rcu state machine */ - spinlock_t schedlock; /* Protect rcu_sched sleep state. */ - enum rcu_sched_sleep_states sched_sleep; /* rcu_sched state. */ - wait_queue_head_t sched_wq; /* Place for rcu_sched to sleep. */ -}; - -struct rcu_dyntick_sched { - int dynticks; - int dynticks_snap; - int sched_qs; - int sched_qs_snap; - int sched_dynticks_snap; -}; - -static DEFINE_PER_CPU_SHARED_ALIGNED(struct rcu_dyntick_sched, rcu_dyntick_sched) = { - .dynticks = 1, -}; - -static int rcu_pending(int cpu); - -void rcu_sched_qs(int cpu) -{ - struct rcu_dyntick_sched *rdssp = &per_cpu(rcu_dyntick_sched, cpu); - - rdssp->sched_qs++; -} - -#ifdef CONFIG_NO_HZ - -void rcu_enter_nohz(void) -{ - static DEFINE_RATELIMIT_STATE(rs, 10 * HZ, 1); - - smp_mb(); /* CPUs seeing ++ must see prior RCU read-side crit sects */ - __get_cpu_var(rcu_dyntick_sched).dynticks++; - WARN_ON_RATELIMIT(__get_cpu_var(rcu_dyntick_sched).dynticks & 0x1, &rs); -} - -void rcu_exit_nohz(void) -{ - static DEFINE_RATELIMIT_STATE(rs, 10 * HZ, 1); - - __get_cpu_var(rcu_dyntick_sched).dynticks++; - smp_mb(); /* CPUs seeing ++ must see later RCU read-side crit sects */ - WARN_ON_RATELIMIT(!(__get_cpu_var(rcu_dyntick_sched).dynticks & 0x1), - &rs); -} - -#endif /* CONFIG_NO_HZ */ - - -static DEFINE_PER_CPU(struct rcu_data, rcu_data); - -static struct rcu_ctrlblk rcu_ctrlblk = { - .fliplock = __SPIN_LOCK_UNLOCKED(rcu_ctrlblk.fliplock), - .completed = 0, - .rcu_try_flip_state = rcu_try_flip_idle_state, - .schedlock = __SPIN_LOCK_UNLOCKED(rcu_ctrlblk.schedlock), - .sched_sleep = rcu_sched_not_sleeping, - .sched_wq = __WAIT_QUEUE_HEAD_INITIALIZER(rcu_ctrlblk.sched_wq), -}; - -static struct task_struct *rcu_sched_grace_period_task; - -#ifdef CONFIG_RCU_TRACE -static char *rcu_try_flip_state_names[] = - { "idle", "waitack", "waitzero", "waitmb" }; -#endif /* #ifdef CONFIG_RCU_TRACE */ - -static DECLARE_BITMAP(rcu_cpu_online_map, NR_CPUS) __read_mostly - = CPU_BITS_NONE; - -/* - * Enum and per-CPU flag to determine when each CPU has seen - * the most recent counter flip. - */ - -enum rcu_flip_flag_values { - rcu_flip_seen, /* Steady/initial state, last flip seen. */ - /* Only GP detector can update. */ - rcu_flipped /* Flip just completed, need confirmation. */ - /* Only corresponding CPU can update. */ -}; -static DEFINE_PER_CPU_SHARED_ALIGNED(enum rcu_flip_flag_values, rcu_flip_flag) - = rcu_flip_seen; - -/* - * Enum and per-CPU flag to determine when each CPU has executed the - * needed memory barrier to fence in memory references from its last RCU - * read-side critical section in the just-completed grace period. - */ - -enum rcu_mb_flag_values { - rcu_mb_done, /* Steady/initial state, no mb()s required. */ - /* Only GP detector can update. */ - rcu_mb_needed /* Flip just completed, need an mb(). */ - /* Only corresponding CPU can update. */ -}; -static DEFINE_PER_CPU_SHARED_ALIGNED(enum rcu_mb_flag_values, rcu_mb_flag) - = rcu_mb_done; - -/* - * RCU_DATA_ME: find the current CPU's rcu_data structure. - * RCU_DATA_CPU: find the specified CPU's rcu_data structure. - */ -#define RCU_DATA_ME() (&__get_cpu_var(rcu_data)) -#define RCU_DATA_CPU(cpu) (&per_cpu(rcu_data, cpu)) - -/* - * Helper macro for tracing when the appropriate rcu_data is not - * cached in a local variable, but where the CPU number is so cached. - */ -#define RCU_TRACE_CPU(f, cpu) RCU_TRACE(f, &(RCU_DATA_CPU(cpu)->trace)); - -/* - * Helper macro for tracing when the appropriate rcu_data is not - * cached in a local variable. - */ -#define RCU_TRACE_ME(f) RCU_TRACE(f, &(RCU_DATA_ME()->trace)); - -/* - * Helper macro for tracing when the appropriate rcu_data is pointed - * to by a local variable. - */ -#define RCU_TRACE_RDP(f, rdp) RCU_TRACE(f, &((rdp)->trace)); - -#define RCU_SCHED_BATCH_TIME (HZ / 50) - -/* - * Return the number of RCU batches processed thus far. Useful - * for debug and statistics. - */ -long rcu_batches_completed(void) -{ - return rcu_ctrlblk.completed; -} -EXPORT_SYMBOL_GPL(rcu_batches_completed); - -void __rcu_read_lock(void) -{ - int idx; - struct task_struct *t = current; - int nesting; - - nesting = ACCESS_ONCE(t->rcu_read_lock_nesting); - if (nesting != 0) { - - /* An earlier rcu_read_lock() covers us, just count it. */ - - t->rcu_read_lock_nesting = nesting + 1; - - } else { - unsigned long flags; - - /* - * We disable interrupts for the following reasons: - * - If we get scheduling clock interrupt here, and we - * end up acking the counter flip, it's like a promise - * that we will never increment the old counter again. - * Thus we will break that promise if that - * scheduling clock interrupt happens between the time - * we pick the .completed field and the time that we - * increment our counter. - * - * - We don't want to be preempted out here. - * - * NMIs can still occur, of course, and might themselves - * contain rcu_read_lock(). - */ - - local_irq_save(flags); - - /* - * Outermost nesting of rcu_read_lock(), so increment - * the current counter for the current CPU. Use volatile - * casts to prevent the compiler from reordering. - */ - - idx = ACCESS_ONCE(rcu_ctrlblk.completed) & 0x1; - ACCESS_ONCE(RCU_DATA_ME()->rcu_flipctr[idx])++; - - /* - * Now that the per-CPU counter has been incremented, we - * are protected from races with rcu_read_lock() invoked - * from NMI handlers on this CPU. We can therefore safely - * increment the nesting counter, relieving further NMIs - * of the need to increment the per-CPU counter. - */ - - ACCESS_ONCE(t->rcu_read_lock_nesting) = nesting + 1; - - /* - * Now that we have preventing any NMIs from storing - * to the ->rcu_flipctr_idx, we can safely use it to - * remember which counter to decrement in the matching - * rcu_read_unlock(). - */ - - ACCESS_ONCE(t->rcu_flipctr_idx) = idx; - local_irq_restore(flags); - } -} -EXPORT_SYMBOL_GPL(__rcu_read_lock); - -void __rcu_read_unlock(void) -{ - int idx; - struct task_struct *t = current; - int nesting; - - nesting = ACCESS_ONCE(t->rcu_read_lock_nesting); - if (nesting > 1) { - - /* - * We are still protected by the enclosing rcu_read_lock(), - * so simply decrement the counter. - */ - - t->rcu_read_lock_nesting = nesting - 1; - - } else { - unsigned long flags; - - /* - * Disable local interrupts to prevent the grace-period - * detection state machine from seeing us half-done. - * NMIs can still occur, of course, and might themselves - * contain rcu_read_lock() and rcu_read_unlock(). - */ - - local_irq_save(flags); - - /* - * Outermost nesting of rcu_read_unlock(), so we must - * decrement the current counter for the current CPU. - * This must be done carefully, because NMIs can - * occur at any point in this code, and any rcu_read_lock() - * and rcu_read_unlock() pairs in the NMI handlers - * must interact non-destructively with this code. - * Lots of volatile casts, and -very- careful ordering. - * - * Changes to this code, including this one, must be - * inspected, validated, and tested extremely carefully!!! - */ - - /* - * First, pick up the index. - */ - - idx = ACCESS_ONCE(t->rcu_flipctr_idx); - - /* - * Now that we have fetched the counter index, it is - * safe to decrement the per-task RCU nesting counter. - * After this, any interrupts or NMIs will increment and - * decrement the per-CPU counters. - */ - ACCESS_ONCE(t->rcu_read_lock_nesting) = nesting - 1; - - /* - * It is now safe to decrement this task's nesting count. - * NMIs that occur after this statement will route their - * rcu_read_lock() calls through this "else" clause, and - * will thus start incrementing the per-CPU counter on - * their own. They will also clobber ->rcu_flipctr_idx, - * but that is OK, since we have already fetched it. - */ - - ACCESS_ONCE(RCU_DATA_ME()->rcu_flipctr[idx])--; - local_irq_restore(flags); - } -} -EXPORT_SYMBOL_GPL(__rcu_read_unlock); - -/* - * If a global counter flip has occurred since the last time that we - * advanced callbacks, advance them. Hardware interrupts must be - * disabled when calling this function. - */ -static void __rcu_advance_callbacks(struct rcu_data *rdp) -{ - int cpu; - int i; - int wlc = 0; - - if (rdp->completed != rcu_ctrlblk.completed) { - if (rdp->waitlist[GP_STAGES - 1] != NULL) { - *rdp->donetail = rdp->waitlist[GP_STAGES - 1]; - rdp->donetail = rdp->waittail[GP_STAGES - 1]; - RCU_TRACE_RDP(rcupreempt_trace_move2done, rdp); - } - for (i = GP_STAGES - 2; i >= 0; i--) { - if (rdp->waitlist[i] != NULL) { - rdp->waitlist[i + 1] = rdp->waitlist[i]; - rdp->waittail[i + 1] = rdp->waittail[i]; - wlc++; - } else { - rdp->waitlist[i + 1] = NULL; - rdp->waittail[i + 1] = - &rdp->waitlist[i + 1]; - } - } - if (rdp->nextlist != NULL) { - rdp->waitlist[0] = rdp->nextlist; - rdp->waittail[0] = rdp->nexttail; - wlc++; - rdp->nextlist = NULL; - rdp->nexttail = &rdp->nextlist; - RCU_TRACE_RDP(rcupreempt_trace_move2wait, rdp); - } else { - rdp->waitlist[0] = NULL; - rdp->waittail[0] = &rdp->waitlist[0]; - } - rdp->waitlistcount = wlc; - rdp->completed = rcu_ctrlblk.completed; - } - - /* - * Check to see if this CPU needs to report that it has seen - * the most recent counter flip, thereby declaring that all - * subsequent rcu_read_lock() invocations will respect this flip. - */ - - cpu = raw_smp_processor_id(); - if (per_cpu(rcu_flip_flag, cpu) == rcu_flipped) { - smp_mb(); /* Subsequent counter accesses must see new value */ - per_cpu(rcu_flip_flag, cpu) = rcu_flip_seen; - smp_mb(); /* Subsequent RCU read-side critical sections */ - /* seen -after- acknowledgement. */ - } -} - -#ifdef CONFIG_NO_HZ -static DEFINE_PER_CPU(int, rcu_update_flag); - -/** - * rcu_irq_enter - Called from Hard irq handlers and NMI/SMI. - * - * If the CPU was idle with dynamic ticks active, this updates the - * rcu_dyntick_sched.dynticks to let the RCU handling know that the - * CPU is active. - */ -void rcu_irq_enter(void) -{ - int cpu = smp_processor_id(); - struct rcu_dyntick_sched *rdssp = &per_cpu(rcu_dyntick_sched, cpu); - - if (per_cpu(rcu_update_flag, cpu)) - per_cpu(rcu_update_flag, cpu)++; - - /* - * Only update if we are coming from a stopped ticks mode - * (rcu_dyntick_sched.dynticks is even). - */ - if (!in_interrupt() && - (rdssp->dynticks & 0x1) == 0) { - /* - * The following might seem like we could have a race - * with NMI/SMIs. But this really isn't a problem. - * Here we do a read/modify/write, and the race happens - * when an NMI/SMI comes in after the read and before - * the write. But NMI/SMIs will increment this counter - * twice before returning, so the zero bit will not - * be corrupted by the NMI/SMI which is the most important - * part. - * - * The only thing is that we would bring back the counter - * to a postion that it was in during the NMI/SMI. - * But the zero bit would be set, so the rest of the - * counter would again be ignored. - * - * On return from the IRQ, the counter may have the zero - * bit be 0 and the counter the same as the return from - * the NMI/SMI. If the state machine was so unlucky to - * see that, it still doesn't matter, since all - * RCU read-side critical sections on this CPU would - * have already completed. - */ - rdssp->dynticks++; - /* - * The following memory barrier ensures that any - * rcu_read_lock() primitives in the irq handler - * are seen by other CPUs to follow the above - * increment to rcu_dyntick_sched.dynticks. This is - * required in order for other CPUs to correctly - * determine when it is safe to advance the RCU - * grace-period state machine. - */ - smp_mb(); /* see above block comment. */ - /* - * Since we can't determine the dynamic tick mode from - * the rcu_dyntick_sched.dynticks after this routine, - * we use a second flag to acknowledge that we came - * from an idle state with ticks stopped. - */ - per_cpu(rcu_update_flag, cpu)++; - /* - * If we take an NMI/SMI now, they will also increment - * the rcu_update_flag, and will not update the - * rcu_dyntick_sched.dynticks on exit. That is for - * this IRQ to do. - */ - } -} - -/** - * rcu_irq_exit - Called from exiting Hard irq context. - * - * If the CPU was idle with dynamic ticks active, update the - * rcu_dyntick_sched.dynticks to let the RCU handling be - * aware that the CPU is going back to idle with no ticks. - */ -void rcu_irq_exit(void) -{ - int cpu = smp_processor_id(); - struct rcu_dyntick_sched *rdssp = &per_cpu(rcu_dyntick_sched, cpu); - - /* - * rcu_update_flag is set if we interrupted the CPU - * when it was idle with ticks stopped. - * Once this occurs, we keep track of interrupt nesting - * because a NMI/SMI could also come in, and we still - * only want the IRQ that started the increment of the - * rcu_dyntick_sched.dynticks to be the one that modifies - * it on exit. - */ - if (per_cpu(rcu_update_flag, cpu)) { - if (--per_cpu(rcu_update_flag, cpu)) - return; - - /* This must match the interrupt nesting */ - WARN_ON(in_interrupt()); - - /* - * If an NMI/SMI happens now we are still - * protected by the rcu_dyntick_sched.dynticks being odd. - */ - - /* - * The following memory barrier ensures that any - * rcu_read_unlock() primitives in the irq handler - * are seen by other CPUs to preceed the following - * increment to rcu_dyntick_sched.dynticks. This - * is required in order for other CPUs to determine - * when it is safe to advance the RCU grace-period - * state machine. - */ - smp_mb(); /* see above block comment. */ - rdssp->dynticks++; - WARN_ON(rdssp->dynticks & 0x1); - } -} - -void rcu_nmi_enter(void) -{ - rcu_irq_enter(); -} - -void rcu_nmi_exit(void) -{ - rcu_irq_exit(); -} - -static void dyntick_save_progress_counter(int cpu) -{ - struct rcu_dyntick_sched *rdssp = &per_cpu(rcu_dyntick_sched, cpu); - - rdssp->dynticks_snap = rdssp->dynticks; -} - -static inline int -rcu_try_flip_waitack_needed(int cpu) -{ - long curr; - long snap; - struct rcu_dyntick_sched *rdssp = &per_cpu(rcu_dyntick_sched, cpu); - - curr = rdssp->dynticks; - snap = rdssp->dynticks_snap; - smp_mb(); /* force ordering with cpu entering/leaving dynticks. */ - - /* - * If the CPU remained in dynticks mode for the entire time - * and didn't take any interrupts, NMIs, SMIs, or whatever, - * then it cannot be in the middle of an rcu_read_lock(), so - * the next rcu_read_lock() it executes must use the new value - * of the counter. So we can safely pretend that this CPU - * already acknowledged the counter. - */ - - if ((curr == snap) && ((curr & 0x1) == 0)) - return 0; - - /* - * If the CPU passed through or entered a dynticks idle phase with - * no active irq handlers, then, as above, we can safely pretend - * that this CPU already acknowledged the counter. - */ - - if ((curr - snap) > 2 || (curr & 0x1) == 0) - return 0; - - /* We need this CPU to explicitly acknowledge the counter flip. */ - - return 1; -} - -static inline int -rcu_try_flip_waitmb_needed(int cpu) -{ - long curr; - long snap; - struct rcu_dyntick_sched *rdssp = &per_cpu(rcu_dyntick_sched, cpu); - - curr = rdssp->dynticks; - snap = rdssp->dynticks_snap; - smp_mb(); /* force ordering with cpu entering/leaving dynticks. */ - - /* - * If the CPU remained in dynticks mode for the entire time - * and didn't take any interrupts, NMIs, SMIs, or whatever, - * then it cannot have executed an RCU read-side critical section - * during that time, so there is no need for it to execute a - * memory barrier. - */ - - if ((curr == snap) && ((curr & 0x1) == 0)) - return 0; - - /* - * If the CPU either entered or exited an outermost interrupt, - * SMI, NMI, or whatever handler, then we know that it executed - * a memory barrier when doing so. So we don't need another one. - */ - if (curr != snap) - return 0; - - /* We need the CPU to execute a memory barrier. */ - - return 1; -} - -static void dyntick_save_progress_counter_sched(int cpu) -{ - struct rcu_dyntick_sched *rdssp = &per_cpu(rcu_dyntick_sched, cpu); - - rdssp->sched_dynticks_snap = rdssp->dynticks; -} - -static int rcu_qsctr_inc_needed_dyntick(int cpu) -{ - long curr; - long snap; - struct rcu_dyntick_sched *rdssp = &per_cpu(rcu_dyntick_sched, cpu); - - curr = rdssp->dynticks; - snap = rdssp->sched_dynticks_snap; - smp_mb(); /* force ordering with cpu entering/leaving dynticks. */ - - /* - * If the CPU remained in dynticks mode for the entire time - * and didn't take any interrupts, NMIs, SMIs, or whatever, - * then it cannot be in the middle of an rcu_read_lock(), so - * the next rcu_read_lock() it executes must use the new value - * of the counter. Therefore, this CPU has been in a quiescent - * state the entire time, and we don't need to wait for it. - */ - - if ((curr == snap) && ((curr & 0x1) == 0)) - return 0; - - /* - * If the CPU passed through or entered a dynticks idle phase with - * no active irq handlers, then, as above, this CPU has already - * passed through a quiescent state. - */ - - if ((curr - snap) > 2 || (snap & 0x1) == 0) - return 0; - - /* We need this CPU to go through a quiescent state. */ - - return 1; -} - -#else /* !CONFIG_NO_HZ */ - -# define dyntick_save_progress_counter(cpu) do { } while (0) -# define rcu_try_flip_waitack_needed(cpu) (1) -# define rcu_try_flip_waitmb_needed(cpu) (1) - -# define dyntick_save_progress_counter_sched(cpu) do { } while (0) -# define rcu_qsctr_inc_needed_dyntick(cpu) (1) - -#endif /* CONFIG_NO_HZ */ - -static void save_qsctr_sched(int cpu) -{ - struct rcu_dyntick_sched *rdssp = &per_cpu(rcu_dyntick_sched, cpu); - - rdssp->sched_qs_snap = rdssp->sched_qs; -} - -static inline int rcu_qsctr_inc_needed(int cpu) -{ - struct rcu_dyntick_sched *rdssp = &per_cpu(rcu_dyntick_sched, cpu); - - /* - * If there has been a quiescent state, no more need to wait - * on this CPU. - */ - - if (rdssp->sched_qs != rdssp->sched_qs_snap) { - smp_mb(); /* force ordering with cpu entering schedule(). */ - return 0; - } - - /* We need this CPU to go through a quiescent state. */ - - return 1; -} - -/* - * Get here when RCU is idle. Decide whether we need to - * move out of idle state, and return non-zero if so. - * "Straightforward" approach for the moment, might later - * use callback-list lengths, grace-period duration, or - * some such to determine when to exit idle state. - * Might also need a pre-idle test that does not acquire - * the lock, but let's get the simple case working first... - */ - -static int -rcu_try_flip_idle(void) -{ - int cpu; - - RCU_TRACE_ME(rcupreempt_trace_try_flip_i1); - if (!rcu_pending(smp_processor_id())) { - RCU_TRACE_ME(rcupreempt_trace_try_flip_ie1); - return 0; - } - - /* - * Do the flip. - */ - - RCU_TRACE_ME(rcupreempt_trace_try_flip_g1); - rcu_ctrlblk.completed++; /* stands in for rcu_try_flip_g2 */ - - /* - * Need a memory barrier so that other CPUs see the new - * counter value before they see the subsequent change of all - * the rcu_flip_flag instances to rcu_flipped. - */ - - smp_mb(); /* see above block comment. */ - - /* Now ask each CPU for acknowledgement of the flip. */ - - for_each_cpu(cpu, to_cpumask(rcu_cpu_online_map)) { - per_cpu(rcu_flip_flag, cpu) = rcu_flipped; - dyntick_save_progress_counter(cpu); - } - - return 1; -} - -/* - * Wait for CPUs to acknowledge the flip. - */ - -static int -rcu_try_flip_waitack(void) -{ - int cpu; - - RCU_TRACE_ME(rcupreempt_trace_try_flip_a1); - for_each_cpu(cpu, to_cpumask(rcu_cpu_online_map)) - if (rcu_try_flip_waitack_needed(cpu) && - per_cpu(rcu_flip_flag, cpu) != rcu_flip_seen) { - RCU_TRACE_ME(rcupreempt_trace_try_flip_ae1); - return 0; - } - - /* - * Make sure our checks above don't bleed into subsequent - * waiting for the sum of the counters to reach zero. - */ - - smp_mb(); /* see above block comment. */ - RCU_TRACE_ME(rcupreempt_trace_try_flip_a2); - return 1; -} - -/* - * Wait for collective ``last'' counter to reach zero, - * then tell all CPUs to do an end-of-grace-period memory barrier. - */ - -static int -rcu_try_flip_waitzero(void) -{ - int cpu; - int lastidx = !(rcu_ctrlblk.completed & 0x1); - int sum = 0; - - /* Check to see if the sum of the "last" counters is zero. */ - - RCU_TRACE_ME(rcupreempt_trace_try_flip_z1); - for_each_possible_cpu(cpu) - sum += RCU_DATA_CPU(cpu)->rcu_flipctr[lastidx]; - if (sum != 0) { - RCU_TRACE_ME(rcupreempt_trace_try_flip_ze1); - return 0; - } - - /* - * This ensures that the other CPUs see the call for - * memory barriers -after- the sum to zero has been - * detected here - */ - smp_mb(); /* ^^^^^^^^^^^^ */ - - /* Call for a memory barrier from each CPU. */ - for_each_cpu(cpu, to_cpumask(rcu_cpu_online_map)) { - per_cpu(rcu_mb_flag, cpu) = rcu_mb_needed; - dyntick_save_progress_counter(cpu); - } - - RCU_TRACE_ME(rcupreempt_trace_try_flip_z2); - return 1; -} - -/* - * Wait for all CPUs to do their end-of-grace-period memory barrier. - * Return 0 once all CPUs have done so. - */ - -static int -rcu_try_flip_waitmb(void) -{ - int cpu; - - RCU_TRACE_ME(rcupreempt_trace_try_flip_m1); - for_each_cpu(cpu, to_cpumask(rcu_cpu_online_map)) - if (rcu_try_flip_waitmb_needed(cpu) && - per_cpu(rcu_mb_flag, cpu) != rcu_mb_done) { - RCU_TRACE_ME(rcupreempt_trace_try_flip_me1); - return 0; - } - - smp_mb(); /* Ensure that the above checks precede any following flip. */ - RCU_TRACE_ME(rcupreempt_trace_try_flip_m2); - return 1; -} - -/* - * Attempt a single flip of the counters. Remember, a single flip does - * -not- constitute a grace period. Instead, the interval between - * at least GP_STAGES consecutive flips is a grace period. - * - * If anyone is nuts enough to run this CONFIG_PREEMPT_RCU implementation - * on a large SMP, they might want to use a hierarchical organization of - * the per-CPU-counter pairs. - */ -static void rcu_try_flip(void) -{ - unsigned long flags; - - RCU_TRACE_ME(rcupreempt_trace_try_flip_1); - if (unlikely(!spin_trylock_irqsave(&rcu_ctrlblk.fliplock, flags))) { - RCU_TRACE_ME(rcupreempt_trace_try_flip_e1); - return; - } - - /* - * Take the next transition(s) through the RCU grace-period - * flip-counter state machine. - */ - - switch (rcu_ctrlblk.rcu_try_flip_state) { - case rcu_try_flip_idle_state: - if (rcu_try_flip_idle()) - rcu_ctrlblk.rcu_try_flip_state = - rcu_try_flip_waitack_state; - break; - case rcu_try_flip_waitack_state: - if (rcu_try_flip_waitack()) - rcu_ctrlblk.rcu_try_flip_state = - rcu_try_flip_waitzero_state; - break; - case rcu_try_flip_waitzero_state: - if (rcu_try_flip_waitzero()) - rcu_ctrlblk.rcu_try_flip_state = - rcu_try_flip_waitmb_state; - break; - case rcu_try_flip_waitmb_state: - if (rcu_try_flip_waitmb()) - rcu_ctrlblk.rcu_try_flip_state = - rcu_try_flip_idle_state; - } - spin_unlock_irqrestore(&rcu_ctrlblk.fliplock, flags); -} - -/* - * Check to see if this CPU needs to do a memory barrier in order to - * ensure that any prior RCU read-side critical sections have committed - * their counter manipulations and critical-section memory references - * before declaring the grace period to be completed. - */ -static void rcu_check_mb(int cpu) -{ - if (per_cpu(rcu_mb_flag, cpu) == rcu_mb_needed) { - smp_mb(); /* Ensure RCU read-side accesses are visible. */ - per_cpu(rcu_mb_flag, cpu) = rcu_mb_done; - } -} - -void rcu_check_callbacks(int cpu, int user) -{ - unsigned long flags; - struct rcu_data *rdp; - - if (!rcu_pending(cpu)) - return; /* if nothing for RCU to do. */ - - /* - * If this CPU took its interrupt from user mode or from the - * idle loop, and this is not a nested interrupt, then - * this CPU has to have exited all prior preept-disable - * sections of code. So invoke rcu_sched_qs() to note this. - * - * The memory barrier is needed to handle the case where - * writes from a preempt-disable section of code get reordered - * into schedule() by this CPU's write buffer. So the memory - * barrier makes sure that the rcu_sched_qs() is seen by other - * CPUs to happen after any such write. - */ - - rdp = RCU_DATA_CPU(cpu); - if (user || - (idle_cpu(cpu) && !in_softirq() && - hardirq_count() <= (1 << HARDIRQ_SHIFT))) { - smp_mb(); /* Guard against aggressive schedule(). */ - rcu_sched_qs(cpu); - } - - rcu_check_mb(cpu); - if (rcu_ctrlblk.completed == rdp->completed) - rcu_try_flip(); - spin_lock_irqsave(&rdp->lock, flags); - RCU_TRACE_RDP(rcupreempt_trace_check_callbacks, rdp); - __rcu_advance_callbacks(rdp); - if (rdp->donelist == NULL) { - spin_unlock_irqrestore(&rdp->lock, flags); - } else { - spin_unlock_irqrestore(&rdp->lock, flags); - raise_softirq(RCU_SOFTIRQ); - } -} - -/* - * Needed by dynticks, to make sure all RCU processing has finished - * when we go idle: - */ -void rcu_advance_callbacks(int cpu, int user) -{ - unsigned long flags; - struct rcu_data *rdp = RCU_DATA_CPU(cpu); - - if (rcu_ctrlblk.completed == rdp->completed) { - rcu_try_flip(); - if (rcu_ctrlblk.completed == rdp->completed) - return; - } - spin_lock_irqsave(&rdp->lock, flags); - RCU_TRACE_RDP(rcupreempt_trace_check_callbacks, rdp); - __rcu_advance_callbacks(rdp); - spin_unlock_irqrestore(&rdp->lock, flags); -} - -#ifdef CONFIG_HOTPLUG_CPU -#define rcu_offline_cpu_enqueue(srclist, srctail, dstlist, dsttail) do { \ - *dsttail = srclist; \ - if (srclist != NULL) { \ - dsttail = srctail; \ - srclist = NULL; \ - srctail = &srclist;\ - } \ - } while (0) - -void rcu_offline_cpu(int cpu) -{ - int i; - struct rcu_head *list = NULL; - unsigned long flags; - struct rcu_data *rdp = RCU_DATA_CPU(cpu); - struct rcu_head *schedlist = NULL; - struct rcu_head **schedtail = &schedlist; - struct rcu_head **tail = &list; - - /* - * Remove all callbacks from the newly dead CPU, retaining order. - * Otherwise rcu_barrier() will fail - */ - - spin_lock_irqsave(&rdp->lock, flags); - rcu_offline_cpu_enqueue(rdp->donelist, rdp->donetail, list, tail); - for (i = GP_STAGES - 1; i >= 0; i--) - rcu_offline_cpu_enqueue(rdp->waitlist[i], rdp->waittail[i], - list, tail); - rcu_offline_cpu_enqueue(rdp->nextlist, rdp->nexttail, list, tail); - rcu_offline_cpu_enqueue(rdp->waitschedlist, rdp->waitschedtail, - schedlist, schedtail); - rcu_offline_cpu_enqueue(rdp->nextschedlist, rdp->nextschedtail, - schedlist, schedtail); - rdp->rcu_sched_sleeping = 0; - spin_unlock_irqrestore(&rdp->lock, flags); - rdp->waitlistcount = 0; - - /* Disengage the newly dead CPU from the grace-period computation. */ - - spin_lock_irqsave(&rcu_ctrlblk.fliplock, flags); - rcu_check_mb(cpu); - if (per_cpu(rcu_flip_flag, cpu) == rcu_flipped) { - smp_mb(); /* Subsequent counter accesses must see new value */ - per_cpu(rcu_flip_flag, cpu) = rcu_flip_seen; - smp_mb(); /* Subsequent RCU read-side critical sections */ - /* seen -after- acknowledgement. */ - } - - cpumask_clear_cpu(cpu, to_cpumask(rcu_cpu_online_map)); - - spin_unlock_irqrestore(&rcu_ctrlblk.fliplock, flags); - - /* - * Place the removed callbacks on the current CPU's queue. - * Make them all start a new grace period: simple approach, - * in theory could starve a given set of callbacks, but - * you would need to be doing some serious CPU hotplugging - * to make this happen. If this becomes a problem, adding - * a synchronize_rcu() to the hotplug path would be a simple - * fix. - */ - - local_irq_save(flags); /* disable preempt till we know what lock. */ - rdp = RCU_DATA_ME(); - spin_lock(&rdp->lock); - *rdp->nexttail = list; - if (list) - rdp->nexttail = tail; - *rdp->nextschedtail = schedlist; - if (schedlist) - rdp->nextschedtail = schedtail; - spin_unlock_irqrestore(&rdp->lock, flags); -} - -#else /* #ifdef CONFIG_HOTPLUG_CPU */ - -void rcu_offline_cpu(int cpu) -{ -} - -#endif /* #else #ifdef CONFIG_HOTPLUG_CPU */ - -void __cpuinit rcu_online_cpu(int cpu) -{ - unsigned long flags; - struct rcu_data *rdp; - - spin_lock_irqsave(&rcu_ctrlblk.fliplock, flags); - cpumask_set_cpu(cpu, to_cpumask(rcu_cpu_online_map)); - spin_unlock_irqrestore(&rcu_ctrlblk.fliplock, flags); - - /* - * The rcu_sched grace-period processing might have bypassed - * this CPU, given that it was not in the rcu_cpu_online_map - * when the grace-period scan started. This means that the - * grace-period task might sleep. So make sure that if this - * should happen, the first callback posted to this CPU will - * wake up the grace-period task if need be. - */ - - rdp = RCU_DATA_CPU(cpu); - spin_lock_irqsave(&rdp->lock, flags); - rdp->rcu_sched_sleeping = 1; - spin_unlock_irqrestore(&rdp->lock, flags); -} - -static void rcu_process_callbacks(struct softirq_action *unused) -{ - unsigned long flags; - struct rcu_head *next, *list; - struct rcu_data *rdp; - - local_irq_save(flags); - rdp = RCU_DATA_ME(); - spin_lock(&rdp->lock); - list = rdp->donelist; - if (list == NULL) { - spin_unlock_irqrestore(&rdp->lock, flags); - return; - } - rdp->donelist = NULL; - rdp->donetail = &rdp->donelist; - RCU_TRACE_RDP(rcupreempt_trace_done_remove, rdp); - spin_unlock_irqrestore(&rdp->lock, flags); - while (list) { - next = list->next; - list->func(list); - list = next; - RCU_TRACE_ME(rcupreempt_trace_invoke); - } -} - -void call_rcu(struct rcu_head *head, void (*func)(struct rcu_head *rcu)) -{ - unsigned long flags; - struct rcu_data *rdp; - - head->func = func; - head->next = NULL; - local_irq_save(flags); - rdp = RCU_DATA_ME(); - spin_lock(&rdp->lock); - __rcu_advance_callbacks(rdp); - *rdp->nexttail = head; - rdp->nexttail = &head->next; - RCU_TRACE_RDP(rcupreempt_trace_next_add, rdp); - spin_unlock_irqrestore(&rdp->lock, flags); -} -EXPORT_SYMBOL_GPL(call_rcu); - -void call_rcu_sched(struct rcu_head *head, void (*func)(struct rcu_head *rcu)) -{ - unsigned long flags; - struct rcu_data *rdp; - int wake_gp = 0; - - head->func = func; - head->next = NULL; - local_irq_save(flags); - rdp = RCU_DATA_ME(); - spin_lock(&rdp->lock); - *rdp->nextschedtail = head; - rdp->nextschedtail = &head->next; - if (rdp->rcu_sched_sleeping) { - - /* Grace-period processing might be sleeping... */ - - rdp->rcu_sched_sleeping = 0; - wake_gp = 1; - } - spin_unlock_irqrestore(&rdp->lock, flags); - if (wake_gp) { - - /* Wake up grace-period processing, unless someone beat us. */ - - spin_lock_irqsave(&rcu_ctrlblk.schedlock, flags); - if (rcu_ctrlblk.sched_sleep != rcu_sched_sleeping) - wake_gp = 0; - rcu_ctrlblk.sched_sleep = rcu_sched_not_sleeping; - spin_unlock_irqrestore(&rcu_ctrlblk.schedlock, flags); - if (wake_gp) - wake_up_interruptible(&rcu_ctrlblk.sched_wq); - } -} -EXPORT_SYMBOL_GPL(call_rcu_sched); - -/* - * Wait until all currently running preempt_disable() code segments - * (including hardware-irq-disable segments) complete. Note that - * in -rt this does -not- necessarily result in all currently executing - * interrupt -handlers- having completed. - */ -void __synchronize_sched(void) -{ - struct rcu_synchronize rcu; - - if (num_online_cpus() == 1) - return; /* blocking is gp if only one CPU! */ - - init_completion(&rcu.completion); - /* Will wake me after RCU finished. */ - call_rcu_sched(&rcu.head, wakeme_after_rcu); - /* Wait for it. */ - wait_for_completion(&rcu.completion); -} -EXPORT_SYMBOL_GPL(__synchronize_sched); - -/* - * kthread function that manages call_rcu_sched grace periods. - */ -static int rcu_sched_grace_period(void *arg) -{ - int couldsleep; /* might sleep after current pass. */ - int couldsleepnext = 0; /* might sleep after next pass. */ - int cpu; - unsigned long flags; - struct rcu_data *rdp; - int ret; - - /* - * Each pass through the following loop handles one - * rcu_sched grace period cycle. - */ - do { - /* Save each CPU's current state. */ - - for_each_online_cpu(cpu) { - dyntick_save_progress_counter_sched(cpu); - save_qsctr_sched(cpu); - } - - /* - * Sleep for about an RCU grace-period's worth to - * allow better batching and to consume less CPU. - */ - schedule_timeout_interruptible(RCU_SCHED_BATCH_TIME); - - /* - * If there was nothing to do last time, prepare to - * sleep at the end of the current grace period cycle. - */ - couldsleep = couldsleepnext; - couldsleepnext = 1; - if (couldsleep) { - spin_lock_irqsave(&rcu_ctrlblk.schedlock, flags); - rcu_ctrlblk.sched_sleep = rcu_sched_sleep_prep; - spin_unlock_irqrestore(&rcu_ctrlblk.schedlock, flags); - } - - /* - * Wait on each CPU in turn to have either visited - * a quiescent state or been in dynticks-idle mode. - */ - for_each_online_cpu(cpu) { - while (rcu_qsctr_inc_needed(cpu) && - rcu_qsctr_inc_needed_dyntick(cpu)) { - /* resched_cpu(cpu); @@@ */ - schedule_timeout_interruptible(1); - } - } - - /* Advance callbacks for each CPU. */ - - for_each_online_cpu(cpu) { - - rdp = RCU_DATA_CPU(cpu); - spin_lock_irqsave(&rdp->lock, flags); - - /* - * We are running on this CPU irq-disabled, so no - * CPU can go offline until we re-enable irqs. - * The current CPU might have already gone - * offline (between the for_each_offline_cpu and - * the spin_lock_irqsave), but in that case all its - * callback lists will be empty, so no harm done. - * - * Advance the callbacks! We share normal RCU's - * donelist, since callbacks are invoked the - * same way in either case. - */ - if (rdp->waitschedlist != NULL) { - *rdp->donetail = rdp->waitschedlist; - rdp->donetail = rdp->waitschedtail; - - /* - * Next rcu_check_callbacks() will - * do the required raise_softirq(). - */ - } - if (rdp->nextschedlist != NULL) { - rdp->waitschedlist = rdp->nextschedlist; - rdp->waitschedtail = rdp->nextschedtail; - couldsleep = 0; - couldsleepnext = 0; - } else { - rdp->waitschedlist = NULL; - rdp->waitschedtail = &rdp->waitschedlist; - } - rdp->nextschedlist = NULL; - rdp->nextschedtail = &rdp->nextschedlist; - - /* Mark sleep intention. */ - - rdp->rcu_sched_sleeping = couldsleep; - - spin_unlock_irqrestore(&rdp->lock, flags); - } - - /* If we saw callbacks on the last scan, go deal with them. */ - - if (!couldsleep) - continue; - - /* Attempt to block... */ - - spin_lock_irqsave(&rcu_ctrlblk.schedlock, flags); - if (rcu_ctrlblk.sched_sleep != rcu_sched_sleep_prep) { - - /* - * Someone posted a callback after we scanned. - * Go take care of it. - */ - spin_unlock_irqrestore(&rcu_ctrlblk.schedlock, flags); - couldsleepnext = 0; - continue; - } - - /* Block until the next person posts a callback. */ - - rcu_ctrlblk.sched_sleep = rcu_sched_sleeping; - spin_unlock_irqrestore(&rcu_ctrlblk.schedlock, flags); - ret = 0; /* unused */ - __wait_event_interruptible(rcu_ctrlblk.sched_wq, - rcu_ctrlblk.sched_sleep != rcu_sched_sleeping, - ret); - - couldsleepnext = 0; - - } while (!kthread_should_stop()); - - return (0); -} - -/* - * Check to see if any future RCU-related work will need to be done - * by the current CPU, even if none need be done immediately, returning - * 1 if so. Assumes that notifiers would take care of handling any - * outstanding requests from the RCU core. - * - * This function is part of the RCU implementation; it is -not- - * an exported member of the RCU API. - */ -int rcu_needs_cpu(int cpu) -{ - struct rcu_data *rdp = RCU_DATA_CPU(cpu); - - return (rdp->donelist != NULL || - !!rdp->waitlistcount || - rdp->nextlist != NULL || - rdp->nextschedlist != NULL || - rdp->waitschedlist != NULL); -} - -static int rcu_pending(int cpu) -{ - struct rcu_data *rdp = RCU_DATA_CPU(cpu); - - /* The CPU has at least one callback queued somewhere. */ - - if (rdp->donelist != NULL || - !!rdp->waitlistcount || - rdp->nextlist != NULL || - rdp->nextschedlist != NULL || - rdp->waitschedlist != NULL) - return 1; - - /* The RCU core needs an acknowledgement from this CPU. */ - - if ((per_cpu(rcu_flip_flag, cpu) == rcu_flipped) || - (per_cpu(rcu_mb_flag, cpu) == rcu_mb_needed)) - return 1; - - /* This CPU has fallen behind the global grace-period number. */ - - if (rdp->completed != rcu_ctrlblk.completed) - return 1; - - /* Nothing needed from this CPU. */ - - return 0; -} - -int __cpuinit rcu_cpu_notify(struct notifier_block *self, - unsigned long action, void *hcpu) -{ - long cpu = (long)hcpu; - - switch (action) { - case CPU_UP_PREPARE: - case CPU_UP_PREPARE_FROZEN: - rcu_online_cpu(cpu); - break; - case CPU_UP_CANCELED: - case CPU_UP_CANCELED_FROZEN: - case CPU_DEAD: - case CPU_DEAD_FROZEN: - rcu_offline_cpu(cpu); - break; - default: - break; - } - return NOTIFY_OK; -} - -void __init __rcu_init(void) -{ - int cpu; - int i; - struct rcu_data *rdp; - - printk(KERN_NOTICE "Preemptible RCU implementation.\n"); - for_each_possible_cpu(cpu) { - rdp = RCU_DATA_CPU(cpu); - spin_lock_init(&rdp->lock); - rdp->completed = 0; - rdp->waitlistcount = 0; - rdp->nextlist = NULL; - rdp->nexttail = &rdp->nextlist; - for (i = 0; i < GP_STAGES; i++) { - rdp->waitlist[i] = NULL; - rdp->waittail[i] = &rdp->waitlist[i]; - } - rdp->donelist = NULL; - rdp->donetail = &rdp->donelist; - rdp->rcu_flipctr[0] = 0; - rdp->rcu_flipctr[1] = 0; - rdp->nextschedlist = NULL; - rdp->nextschedtail = &rdp->nextschedlist; - rdp->waitschedlist = NULL; - rdp->waitschedtail = &rdp->waitschedlist; - rdp->rcu_sched_sleeping = 0; - } - open_softirq(RCU_SOFTIRQ, rcu_process_callbacks); -} - -/* - * Late-boot-time RCU initialization that must wait until after scheduler - * has been initialized. - */ -void __init rcu_init_sched(void) -{ - rcu_sched_grace_period_task = kthread_run(rcu_sched_grace_period, - NULL, - "rcu_sched_grace_period"); - WARN_ON(IS_ERR(rcu_sched_grace_period_task)); -} - -#ifdef CONFIG_RCU_TRACE -long *rcupreempt_flipctr(int cpu) -{ - return &RCU_DATA_CPU(cpu)->rcu_flipctr[0]; -} -EXPORT_SYMBOL_GPL(rcupreempt_flipctr); - -int rcupreempt_flip_flag(int cpu) -{ - return per_cpu(rcu_flip_flag, cpu); -} -EXPORT_SYMBOL_GPL(rcupreempt_flip_flag); - -int rcupreempt_mb_flag(int cpu) -{ - return per_cpu(rcu_mb_flag, cpu); -} -EXPORT_SYMBOL_GPL(rcupreempt_mb_flag); - -char *rcupreempt_try_flip_state_name(void) -{ - return rcu_try_flip_state_names[rcu_ctrlblk.rcu_try_flip_state]; -} -EXPORT_SYMBOL_GPL(rcupreempt_try_flip_state_name); - -struct rcupreempt_trace *rcupreempt_trace_cpu(int cpu) -{ - struct rcu_data *rdp = RCU_DATA_CPU(cpu); - - return &rdp->trace; -} -EXPORT_SYMBOL_GPL(rcupreempt_trace_cpu); - -#endif /* #ifdef RCU_TRACE */ diff --git a/kernel/rcupreempt_trace.c b/kernel/rcupreempt_trace.c deleted file mode 100644 index 11640346a507..000000000000 --- a/kernel/rcupreempt_trace.c +++ /dev/null @@ -1,335 +0,0 @@ -/* - * Read-Copy Update tracing for realtime implementation - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. - * - * Copyright IBM Corporation, 2006 - * - * Papers: http://www.rdrop.com/users/paulmck/RCU - * - * For detailed explanation of Read-Copy Update mechanism see - - * Documentation/RCU/ *.txt - * - */ -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include - -static struct mutex rcupreempt_trace_mutex; -static char *rcupreempt_trace_buf; -#define RCUPREEMPT_TRACE_BUF_SIZE 4096 - -void rcupreempt_trace_move2done(struct rcupreempt_trace *trace) -{ - trace->done_length += trace->wait_length; - trace->done_add += trace->wait_length; - trace->wait_length = 0; -} -void rcupreempt_trace_move2wait(struct rcupreempt_trace *trace) -{ - trace->wait_length += trace->next_length; - trace->wait_add += trace->next_length; - trace->next_length = 0; -} -void rcupreempt_trace_try_flip_1(struct rcupreempt_trace *trace) -{ - atomic_inc(&trace->rcu_try_flip_1); -} -void rcupreempt_trace_try_flip_e1(struct rcupreempt_trace *trace) -{ - atomic_inc(&trace->rcu_try_flip_e1); -} -void rcupreempt_trace_try_flip_i1(struct rcupreempt_trace *trace) -{ - trace->rcu_try_flip_i1++; -} -void rcupreempt_trace_try_flip_ie1(struct rcupreempt_trace *trace) -{ - trace->rcu_try_flip_ie1++; -} -void rcupreempt_trace_try_flip_g1(struct rcupreempt_trace *trace) -{ - trace->rcu_try_flip_g1++; -} -void rcupreempt_trace_try_flip_a1(struct rcupreempt_trace *trace) -{ - trace->rcu_try_flip_a1++; -} -void rcupreempt_trace_try_flip_ae1(struct rcupreempt_trace *trace) -{ - trace->rcu_try_flip_ae1++; -} -void rcupreempt_trace_try_flip_a2(struct rcupreempt_trace *trace) -{ - trace->rcu_try_flip_a2++; -} -void rcupreempt_trace_try_flip_z1(struct rcupreempt_trace *trace) -{ - trace->rcu_try_flip_z1++; -} -void rcupreempt_trace_try_flip_ze1(struct rcupreempt_trace *trace) -{ - trace->rcu_try_flip_ze1++; -} -void rcupreempt_trace_try_flip_z2(struct rcupreempt_trace *trace) -{ - trace->rcu_try_flip_z2++; -} -void rcupreempt_trace_try_flip_m1(struct rcupreempt_trace *trace) -{ - trace->rcu_try_flip_m1++; -} -void rcupreempt_trace_try_flip_me1(struct rcupreempt_trace *trace) -{ - trace->rcu_try_flip_me1++; -} -void rcupreempt_trace_try_flip_m2(struct rcupreempt_trace *trace) -{ - trace->rcu_try_flip_m2++; -} -void rcupreempt_trace_check_callbacks(struct rcupreempt_trace *trace) -{ - trace->rcu_check_callbacks++; -} -void rcupreempt_trace_done_remove(struct rcupreempt_trace *trace) -{ - trace->done_remove += trace->done_length; - trace->done_length = 0; -} -void rcupreempt_trace_invoke(struct rcupreempt_trace *trace) -{ - atomic_inc(&trace->done_invoked); -} -void rcupreempt_trace_next_add(struct rcupreempt_trace *trace) -{ - trace->next_add++; - trace->next_length++; -} - -static void rcupreempt_trace_sum(struct rcupreempt_trace *sp) -{ - struct rcupreempt_trace *cp; - int cpu; - - memset(sp, 0, sizeof(*sp)); - for_each_possible_cpu(cpu) { - cp = rcupreempt_trace_cpu(cpu); - sp->next_length += cp->next_length; - sp->next_add += cp->next_add; - sp->wait_length += cp->wait_length; - sp->wait_add += cp->wait_add; - sp->done_length += cp->done_length; - sp->done_add += cp->done_add; - sp->done_remove += cp->done_remove; - atomic_add(atomic_read(&cp->done_invoked), &sp->done_invoked); - sp->rcu_check_callbacks += cp->rcu_check_callbacks; - atomic_add(atomic_read(&cp->rcu_try_flip_1), - &sp->rcu_try_flip_1); - atomic_add(atomic_read(&cp->rcu_try_flip_e1), - &sp->rcu_try_flip_e1); - sp->rcu_try_flip_i1 += cp->rcu_try_flip_i1; - sp->rcu_try_flip_ie1 += cp->rcu_try_flip_ie1; - sp->rcu_try_flip_g1 += cp->rcu_try_flip_g1; - sp->rcu_try_flip_a1 += cp->rcu_try_flip_a1; - sp->rcu_try_flip_ae1 += cp->rcu_try_flip_ae1; - sp->rcu_try_flip_a2 += cp->rcu_try_flip_a2; - sp->rcu_try_flip_z1 += cp->rcu_try_flip_z1; - sp->rcu_try_flip_ze1 += cp->rcu_try_flip_ze1; - sp->rcu_try_flip_z2 += cp->rcu_try_flip_z2; - sp->rcu_try_flip_m1 += cp->rcu_try_flip_m1; - sp->rcu_try_flip_me1 += cp->rcu_try_flip_me1; - sp->rcu_try_flip_m2 += cp->rcu_try_flip_m2; - } -} - -static ssize_t rcustats_read(struct file *filp, char __user *buffer, - size_t count, loff_t *ppos) -{ - struct rcupreempt_trace trace; - ssize_t bcount; - int cnt = 0; - - rcupreempt_trace_sum(&trace); - mutex_lock(&rcupreempt_trace_mutex); - snprintf(&rcupreempt_trace_buf[cnt], RCUPREEMPT_TRACE_BUF_SIZE - cnt, - "ggp=%ld rcc=%ld\n", - rcu_batches_completed(), - trace.rcu_check_callbacks); - snprintf(&rcupreempt_trace_buf[cnt], RCUPREEMPT_TRACE_BUF_SIZE - cnt, - "na=%ld nl=%ld wa=%ld wl=%ld da=%ld dl=%ld dr=%ld di=%d\n" - "1=%d e1=%d i1=%ld ie1=%ld g1=%ld a1=%ld ae1=%ld a2=%ld\n" - "z1=%ld ze1=%ld z2=%ld m1=%ld me1=%ld m2=%ld\n", - - trace.next_add, trace.next_length, - trace.wait_add, trace.wait_length, - trace.done_add, trace.done_length, - trace.done_remove, atomic_read(&trace.done_invoked), - atomic_read(&trace.rcu_try_flip_1), - atomic_read(&trace.rcu_try_flip_e1), - trace.rcu_try_flip_i1, trace.rcu_try_flip_ie1, - trace.rcu_try_flip_g1, - trace.rcu_try_flip_a1, trace.rcu_try_flip_ae1, - trace.rcu_try_flip_a2, - trace.rcu_try_flip_z1, trace.rcu_try_flip_ze1, - trace.rcu_try_flip_z2, - trace.rcu_try_flip_m1, trace.rcu_try_flip_me1, - trace.rcu_try_flip_m2); - bcount = simple_read_from_buffer(buffer, count, ppos, - rcupreempt_trace_buf, strlen(rcupreempt_trace_buf)); - mutex_unlock(&rcupreempt_trace_mutex); - return bcount; -} - -static ssize_t rcugp_read(struct file *filp, char __user *buffer, - size_t count, loff_t *ppos) -{ - long oldgp = rcu_batches_completed(); - ssize_t bcount; - - mutex_lock(&rcupreempt_trace_mutex); - synchronize_rcu(); - snprintf(rcupreempt_trace_buf, RCUPREEMPT_TRACE_BUF_SIZE, - "oldggp=%ld newggp=%ld\n", oldgp, rcu_batches_completed()); - bcount = simple_read_from_buffer(buffer, count, ppos, - rcupreempt_trace_buf, strlen(rcupreempt_trace_buf)); - mutex_unlock(&rcupreempt_trace_mutex); - return bcount; -} - -static ssize_t rcuctrs_read(struct file *filp, char __user *buffer, - size_t count, loff_t *ppos) -{ - int cnt = 0; - int cpu; - int f = rcu_batches_completed() & 0x1; - ssize_t bcount; - - mutex_lock(&rcupreempt_trace_mutex); - - cnt += snprintf(&rcupreempt_trace_buf[cnt], RCUPREEMPT_TRACE_BUF_SIZE, - "CPU last cur F M\n"); - for_each_possible_cpu(cpu) { - long *flipctr = rcupreempt_flipctr(cpu); - cnt += snprintf(&rcupreempt_trace_buf[cnt], - RCUPREEMPT_TRACE_BUF_SIZE - cnt, - "%3d%c %4ld %3ld %d %d\n", - cpu, - cpu_is_offline(cpu) ? '!' : ' ', - flipctr[!f], - flipctr[f], - rcupreempt_flip_flag(cpu), - rcupreempt_mb_flag(cpu)); - } - cnt += snprintf(&rcupreempt_trace_buf[cnt], - RCUPREEMPT_TRACE_BUF_SIZE - cnt, - "ggp = %ld, state = %s\n", - rcu_batches_completed(), - rcupreempt_try_flip_state_name()); - cnt += snprintf(&rcupreempt_trace_buf[cnt], - RCUPREEMPT_TRACE_BUF_SIZE - cnt, - "\n"); - bcount = simple_read_from_buffer(buffer, count, ppos, - rcupreempt_trace_buf, strlen(rcupreempt_trace_buf)); - mutex_unlock(&rcupreempt_trace_mutex); - return bcount; -} - -static struct file_operations rcustats_fops = { - .owner = THIS_MODULE, - .read = rcustats_read, -}; - -static struct file_operations rcugp_fops = { - .owner = THIS_MODULE, - .read = rcugp_read, -}; - -static struct file_operations rcuctrs_fops = { - .owner = THIS_MODULE, - .read = rcuctrs_read, -}; - -static struct dentry *rcudir, *statdir, *ctrsdir, *gpdir; -static int rcupreempt_debugfs_init(void) -{ - rcudir = debugfs_create_dir("rcu", NULL); - if (!rcudir) - goto out; - statdir = debugfs_create_file("rcustats", 0444, rcudir, - NULL, &rcustats_fops); - if (!statdir) - goto free_out; - - gpdir = debugfs_create_file("rcugp", 0444, rcudir, NULL, &rcugp_fops); - if (!gpdir) - goto free_out; - - ctrsdir = debugfs_create_file("rcuctrs", 0444, rcudir, - NULL, &rcuctrs_fops); - if (!ctrsdir) - goto free_out; - return 0; -free_out: - if (statdir) - debugfs_remove(statdir); - if (gpdir) - debugfs_remove(gpdir); - debugfs_remove(rcudir); -out: - return 1; -} - -static int __init rcupreempt_trace_init(void) -{ - int ret; - - mutex_init(&rcupreempt_trace_mutex); - rcupreempt_trace_buf = kmalloc(RCUPREEMPT_TRACE_BUF_SIZE, GFP_KERNEL); - if (!rcupreempt_trace_buf) - return 1; - ret = rcupreempt_debugfs_init(); - if (ret) - kfree(rcupreempt_trace_buf); - return ret; -} - -static void __exit rcupreempt_trace_cleanup(void) -{ - debugfs_remove(statdir); - debugfs_remove(gpdir); - debugfs_remove(ctrsdir); - debugfs_remove(rcudir); - kfree(rcupreempt_trace_buf); -} - - -module_init(rcupreempt_trace_init); -module_exit(rcupreempt_trace_cleanup); diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index f87fb0c8f924..82fbc49728df 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -725,7 +725,7 @@ config RCU_TORTURE_TEST_RUNNABLE config RCU_CPU_STALL_DETECTOR bool "Check for stalled CPUs delaying RCU grace periods" - depends on CLASSIC_RCU || TREE_RCU || TREE_PREEMPT_RCU + depends on TREE_RCU || TREE_PREEMPT_RCU default n help This option causes RCU to printk information on which -- cgit v1.2.3 From 70041088e3b976627ba9a183b812f39ef8a9ba0e Mon Sep 17 00:00:00 2001 From: Andy Grover Date: Fri, 21 Aug 2009 12:28:31 +0000 Subject: RDS: Add TCP transport to RDS This code allows RDS to be tunneled over a TCP connection. RDMA operations are disabled when using TCP transport, but this frees RDS from the IB/RDMA stack dependency, and allows it to be used with standard Ethernet adapters, or in a VM. Signed-off-by: Andy Grover Signed-off-by: David S. Miller --- include/linux/rds.h | 12 ++ net/rds/tcp.c | 319 ++++++++++++++++++++++++++++++++++++++++++++ net/rds/tcp.h | 93 +++++++++++++ net/rds/tcp_connect.c | 153 ++++++++++++++++++++++ net/rds/tcp_listen.c | 199 ++++++++++++++++++++++++++++ net/rds/tcp_recv.c | 356 ++++++++++++++++++++++++++++++++++++++++++++++++++ net/rds/tcp_send.c | 263 +++++++++++++++++++++++++++++++++++++ net/rds/tcp_stats.c | 74 +++++++++++ 8 files changed, 1469 insertions(+) create mode 100644 net/rds/tcp.c create mode 100644 net/rds/tcp.h create mode 100644 net/rds/tcp_connect.c create mode 100644 net/rds/tcp_listen.c create mode 100644 net/rds/tcp_recv.c create mode 100644 net/rds/tcp_send.c create mode 100644 net/rds/tcp_stats.c (limited to 'include') diff --git a/include/linux/rds.h b/include/linux/rds.h index d91dc91f5443..89d46e1afbb1 100644 --- a/include/linux/rds.h +++ b/include/linux/rds.h @@ -147,6 +147,18 @@ struct rds_info_socket { u_int64_t inum; } __attribute__((packed)); +struct rds_info_tcp_socket { + __be32 local_addr; + __be16 local_port; + __be32 peer_addr; + __be16 peer_port; + u_int64_t hdr_rem; + u_int64_t data_rem; + u_int32_t last_sent_nxt; + u_int32_t last_expected_una; + u_int32_t last_seen_una; +} __attribute__((packed)); + #define RDS_IB_GID_LEN 16 struct rds_info_rdma_connection { __be32 src_addr; diff --git a/net/rds/tcp.c b/net/rds/tcp.c new file mode 100644 index 000000000000..e0ac9009db1a --- /dev/null +++ b/net/rds/tcp.c @@ -0,0 +1,319 @@ +/* + * Copyright (c) 2006 Oracle. All rights reserved. + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * OpenIB.org BSD license below: + * + * Redistribution and use in source and binary forms, with or + * without modification, are permitted provided that the following + * conditions are met: + * + * - Redistributions of source code must retain the above + * copyright notice, this list of conditions and the following + * disclaimer. + * + * - Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials + * provided with the distribution. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ +#include +#include +#include + +#include "rds.h" +#include "tcp.h" + +/* only for info exporting */ +static DEFINE_SPINLOCK(rds_tcp_tc_list_lock); +static LIST_HEAD(rds_tcp_tc_list); +unsigned int rds_tcp_tc_count; + +/* Track rds_tcp_connection structs so they can be cleaned up */ +static DEFINE_SPINLOCK(rds_tcp_conn_lock); +static LIST_HEAD(rds_tcp_conn_list); + +static struct kmem_cache *rds_tcp_conn_slab; + +#define RDS_TCP_DEFAULT_BUFSIZE (128 * 1024) + +/* doing it this way avoids calling tcp_sk() */ +void rds_tcp_nonagle(struct socket *sock) +{ + mm_segment_t oldfs = get_fs(); + int val = 1; + + set_fs(KERNEL_DS); + sock->ops->setsockopt(sock, SOL_TCP, TCP_NODELAY, (char __user *)&val, + sizeof(val)); + set_fs(oldfs); +} + +void rds_tcp_tune(struct socket *sock) +{ + struct sock *sk = sock->sk; + + rds_tcp_nonagle(sock); + + /* + * We're trying to saturate gigabit with the default, + * see svc_sock_setbufsize(). + */ + lock_sock(sk); + sk->sk_sndbuf = RDS_TCP_DEFAULT_BUFSIZE; + sk->sk_rcvbuf = RDS_TCP_DEFAULT_BUFSIZE; + sk->sk_userlocks |= SOCK_SNDBUF_LOCK|SOCK_RCVBUF_LOCK; + release_sock(sk); +} + +u32 rds_tcp_snd_nxt(struct rds_tcp_connection *tc) +{ + return tcp_sk(tc->t_sock->sk)->snd_nxt; +} + +u32 rds_tcp_snd_una(struct rds_tcp_connection *tc) +{ + return tcp_sk(tc->t_sock->sk)->snd_una; +} + +void rds_tcp_restore_callbacks(struct socket *sock, + struct rds_tcp_connection *tc) +{ + rdsdebug("restoring sock %p callbacks from tc %p\n", sock, tc); + write_lock_bh(&sock->sk->sk_callback_lock); + + /* done under the callback_lock to serialize with write_space */ + spin_lock(&rds_tcp_tc_list_lock); + list_del_init(&tc->t_list_item); + rds_tcp_tc_count--; + spin_unlock(&rds_tcp_tc_list_lock); + + tc->t_sock = NULL; + + sock->sk->sk_write_space = tc->t_orig_write_space; + sock->sk->sk_data_ready = tc->t_orig_data_ready; + sock->sk->sk_state_change = tc->t_orig_state_change; + sock->sk->sk_user_data = NULL; + + write_unlock_bh(&sock->sk->sk_callback_lock); +} + +/* + * This is the only path that sets tc->t_sock. Send and receive trust that + * it is set. The RDS_CONN_CONNECTED bit protects those paths from being + * called while it isn't set. + */ +void rds_tcp_set_callbacks(struct socket *sock, struct rds_connection *conn) +{ + struct rds_tcp_connection *tc = conn->c_transport_data; + + rdsdebug("setting sock %p callbacks to tc %p\n", sock, tc); + write_lock_bh(&sock->sk->sk_callback_lock); + + /* done under the callback_lock to serialize with write_space */ + spin_lock(&rds_tcp_tc_list_lock); + list_add_tail(&tc->t_list_item, &rds_tcp_tc_list); + rds_tcp_tc_count++; + spin_unlock(&rds_tcp_tc_list_lock); + + /* accepted sockets need our listen data ready undone */ + if (sock->sk->sk_data_ready == rds_tcp_listen_data_ready) + sock->sk->sk_data_ready = sock->sk->sk_user_data; + + tc->t_sock = sock; + tc->conn = conn; + tc->t_orig_data_ready = sock->sk->sk_data_ready; + tc->t_orig_write_space = sock->sk->sk_write_space; + tc->t_orig_state_change = sock->sk->sk_state_change; + + sock->sk->sk_user_data = conn; + sock->sk->sk_data_ready = rds_tcp_data_ready; + sock->sk->sk_write_space = rds_tcp_write_space; + sock->sk->sk_state_change = rds_tcp_state_change; + + write_unlock_bh(&sock->sk->sk_callback_lock); +} + +static void rds_tcp_tc_info(struct socket *sock, unsigned int len, + struct rds_info_iterator *iter, + struct rds_info_lengths *lens) +{ + struct rds_info_tcp_socket tsinfo; + struct rds_tcp_connection *tc; + unsigned long flags; + struct sockaddr_in sin; + int sinlen; + + spin_lock_irqsave(&rds_tcp_tc_list_lock, flags); + + if (len / sizeof(tsinfo) < rds_tcp_tc_count) + goto out; + + list_for_each_entry(tc, &rds_tcp_tc_list, t_list_item) { + + sock->ops->getname(sock, (struct sockaddr *)&sin, &sinlen, 0); + tsinfo.local_addr = sin.sin_addr.s_addr; + tsinfo.local_port = sin.sin_port; + sock->ops->getname(sock, (struct sockaddr *)&sin, &sinlen, 1); + tsinfo.peer_addr = sin.sin_addr.s_addr; + tsinfo.peer_port = sin.sin_port; + + tsinfo.hdr_rem = tc->t_tinc_hdr_rem; + tsinfo.data_rem = tc->t_tinc_data_rem; + tsinfo.last_sent_nxt = tc->t_last_sent_nxt; + tsinfo.last_expected_una = tc->t_last_expected_una; + tsinfo.last_seen_una = tc->t_last_seen_una; + + rds_info_copy(iter, &tsinfo, sizeof(tsinfo)); + } + +out: + lens->nr = rds_tcp_tc_count; + lens->each = sizeof(tsinfo); + + spin_unlock_irqrestore(&rds_tcp_tc_list_lock, flags); +} + +static int rds_tcp_laddr_check(__be32 addr) +{ + if (inet_addr_type(&init_net, addr) == RTN_LOCAL) + return 0; + return -EADDRNOTAVAIL; +} + +static int rds_tcp_conn_alloc(struct rds_connection *conn, gfp_t gfp) +{ + struct rds_tcp_connection *tc; + + tc = kmem_cache_alloc(rds_tcp_conn_slab, gfp); + if (tc == NULL) + return -ENOMEM; + + tc->t_sock = NULL; + tc->t_tinc = NULL; + tc->t_tinc_hdr_rem = sizeof(struct rds_header); + tc->t_tinc_data_rem = 0; + + conn->c_transport_data = tc; + + spin_lock_irq(&rds_tcp_conn_lock); + list_add_tail(&tc->t_tcp_node, &rds_tcp_conn_list); + spin_unlock_irq(&rds_tcp_conn_lock); + + rdsdebug("alloced tc %p\n", conn->c_transport_data); + return 0; +} + +static void rds_tcp_conn_free(void *arg) +{ + struct rds_tcp_connection *tc = arg; + rdsdebug("freeing tc %p\n", tc); + kmem_cache_free(rds_tcp_conn_slab, tc); +} + +static void rds_tcp_destroy_conns(void) +{ + struct rds_tcp_connection *tc, *_tc; + LIST_HEAD(tmp_list); + + /* avoid calling conn_destroy with irqs off */ + spin_lock_irq(&rds_tcp_conn_lock); + list_splice(&rds_tcp_conn_list, &tmp_list); + INIT_LIST_HEAD(&rds_tcp_conn_list); + spin_unlock_irq(&rds_tcp_conn_lock); + + list_for_each_entry_safe(tc, _tc, &tmp_list, t_tcp_node) { + if (tc->conn->c_passive) + rds_conn_destroy(tc->conn->c_passive); + rds_conn_destroy(tc->conn); + } +} + +void rds_tcp_exit(void) +{ + rds_info_deregister_func(RDS_INFO_TCP_SOCKETS, rds_tcp_tc_info); + rds_tcp_listen_stop(); + rds_tcp_destroy_conns(); + rds_trans_unregister(&rds_tcp_transport); + rds_tcp_recv_exit(); + kmem_cache_destroy(rds_tcp_conn_slab); +} +module_exit(rds_tcp_exit); + +struct rds_transport rds_tcp_transport = { + .laddr_check = rds_tcp_laddr_check, + .xmit_prepare = rds_tcp_xmit_prepare, + .xmit_complete = rds_tcp_xmit_complete, + .xmit_cong_map = rds_tcp_xmit_cong_map, + .xmit = rds_tcp_xmit, + .recv = rds_tcp_recv, + .conn_alloc = rds_tcp_conn_alloc, + .conn_free = rds_tcp_conn_free, + .conn_connect = rds_tcp_conn_connect, + .conn_shutdown = rds_tcp_conn_shutdown, + .inc_copy_to_user = rds_tcp_inc_copy_to_user, + .inc_purge = rds_tcp_inc_purge, + .inc_free = rds_tcp_inc_free, + .stats_info_copy = rds_tcp_stats_info_copy, + .exit = rds_tcp_exit, + .t_owner = THIS_MODULE, + .t_name = "tcp", + .t_prefer_loopback = 1, +}; + +int __init rds_tcp_init(void) +{ + int ret; + + rds_tcp_conn_slab = kmem_cache_create("rds_tcp_connection", + sizeof(struct rds_tcp_connection), + 0, 0, NULL); + if (rds_tcp_conn_slab == NULL) { + ret = -ENOMEM; + goto out; + } + + ret = rds_tcp_recv_init(); + if (ret) + goto out_slab; + + ret = rds_trans_register(&rds_tcp_transport); + if (ret) + goto out_recv; + + ret = rds_tcp_listen_init(); + if (ret) + goto out_register; + + rds_info_register_func(RDS_INFO_TCP_SOCKETS, rds_tcp_tc_info); + + goto out; + +out_register: + rds_trans_unregister(&rds_tcp_transport); +out_recv: + rds_tcp_recv_exit(); +out_slab: + kmem_cache_destroy(rds_tcp_conn_slab); +out: + return ret; +} +module_init(rds_tcp_init); + +MODULE_AUTHOR("Oracle Corporation "); +MODULE_DESCRIPTION("RDS: TCP transport"); +MODULE_LICENSE("Dual BSD/GPL"); + diff --git a/net/rds/tcp.h b/net/rds/tcp.h new file mode 100644 index 000000000000..844fa6b9cf5a --- /dev/null +++ b/net/rds/tcp.h @@ -0,0 +1,93 @@ +#ifndef _RDS_TCP_H +#define _RDS_TCP_H + +#define RDS_TCP_PORT 16385 + +struct rds_tcp_incoming { + struct rds_incoming ti_inc; + struct sk_buff_head ti_skb_list; +}; + +struct rds_tcp_connection { + + struct list_head t_tcp_node; + struct rds_connection *conn; + struct socket *t_sock; + void *t_orig_write_space; + void *t_orig_data_ready; + void *t_orig_state_change; + + struct rds_tcp_incoming *t_tinc; + size_t t_tinc_hdr_rem; + size_t t_tinc_data_rem; + + /* XXX error report? */ + struct work_struct t_conn_w; + struct work_struct t_send_w; + struct work_struct t_down_w; + struct work_struct t_recv_w; + + /* for info exporting only */ + struct list_head t_list_item; + u32 t_last_sent_nxt; + u32 t_last_expected_una; + u32 t_last_seen_una; +}; + +struct rds_tcp_statistics { + uint64_t s_tcp_data_ready_calls; + uint64_t s_tcp_write_space_calls; + uint64_t s_tcp_sndbuf_full; + uint64_t s_tcp_connect_raced; + uint64_t s_tcp_listen_closed_stale; +}; + +/* tcp.c */ +int __init rds_tcp_init(void); +void rds_tcp_exit(void); +void rds_tcp_tune(struct socket *sock); +void rds_tcp_nonagle(struct socket *sock); +void rds_tcp_set_callbacks(struct socket *sock, struct rds_connection *conn); +void rds_tcp_restore_callbacks(struct socket *sock, + struct rds_tcp_connection *tc); +u32 rds_tcp_snd_nxt(struct rds_tcp_connection *tc); +u32 rds_tcp_snd_una(struct rds_tcp_connection *tc); +u64 rds_tcp_map_seq(struct rds_tcp_connection *tc, u32 seq); +extern struct rds_transport rds_tcp_transport; + +/* tcp_connect.c */ +int rds_tcp_conn_connect(struct rds_connection *conn); +void rds_tcp_conn_shutdown(struct rds_connection *conn); +void rds_tcp_state_change(struct sock *sk); + +/* tcp_listen.c */ +int __init rds_tcp_listen_init(void); +void rds_tcp_listen_stop(void); +void rds_tcp_listen_data_ready(struct sock *sk, int bytes); + +/* tcp_recv.c */ +int __init rds_tcp_recv_init(void); +void rds_tcp_recv_exit(void); +void rds_tcp_data_ready(struct sock *sk, int bytes); +int rds_tcp_recv(struct rds_connection *conn); +void rds_tcp_inc_purge(struct rds_incoming *inc); +void rds_tcp_inc_free(struct rds_incoming *inc); +int rds_tcp_inc_copy_to_user(struct rds_incoming *inc, struct iovec *iov, + size_t size); + +/* tcp_send.c */ +void rds_tcp_xmit_prepare(struct rds_connection *conn); +void rds_tcp_xmit_complete(struct rds_connection *conn); +int rds_tcp_xmit(struct rds_connection *conn, struct rds_message *rm, + unsigned int hdr_off, unsigned int sg, unsigned int off); +void rds_tcp_write_space(struct sock *sk); +int rds_tcp_xmit_cong_map(struct rds_connection *conn, + struct rds_cong_map *map, unsigned long offset); + +/* tcp_stats.c */ +DECLARE_PER_CPU(struct rds_tcp_statistics, rds_tcp_stats); +#define rds_tcp_stats_inc(member) rds_stats_inc_which(rds_tcp_stats, member) +unsigned int rds_tcp_stats_info_copy(struct rds_info_iterator *iter, + unsigned int avail); + +#endif diff --git a/net/rds/tcp_connect.c b/net/rds/tcp_connect.c new file mode 100644 index 000000000000..211522f9a9a2 --- /dev/null +++ b/net/rds/tcp_connect.c @@ -0,0 +1,153 @@ +/* + * Copyright (c) 2006 Oracle. All rights reserved. + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * OpenIB.org BSD license below: + * + * Redistribution and use in source and binary forms, with or + * without modification, are permitted provided that the following + * conditions are met: + * + * - Redistributions of source code must retain the above + * copyright notice, this list of conditions and the following + * disclaimer. + * + * - Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials + * provided with the distribution. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ +#include +#include +#include + +#include "rds.h" +#include "tcp.h" + +void rds_tcp_state_change(struct sock *sk) +{ + void (*state_change)(struct sock *sk); + struct rds_connection *conn; + struct rds_tcp_connection *tc; + + read_lock(&sk->sk_callback_lock); + conn = sk->sk_user_data; + if (conn == NULL) { + state_change = sk->sk_state_change; + goto out; + } + tc = conn->c_transport_data; + state_change = tc->t_orig_state_change; + + rdsdebug("sock %p state_change to %d\n", tc->t_sock, sk->sk_state); + + switch(sk->sk_state) { + /* ignore connecting sockets as they make progress */ + case TCP_SYN_SENT: + case TCP_SYN_RECV: + break; + case TCP_ESTABLISHED: + rds_connect_complete(conn); + break; + case TCP_CLOSE: + rds_conn_drop(conn); + default: + break; + } +out: + read_unlock(&sk->sk_callback_lock); + state_change(sk); +} + +int rds_tcp_conn_connect(struct rds_connection *conn) +{ + struct socket *sock = NULL; + struct sockaddr_in src, dest; + int ret; + + ret = sock_create(PF_INET, SOCK_STREAM, IPPROTO_TCP, &sock); + if (ret < 0) + goto out; + + rds_tcp_tune(sock); + + src.sin_family = AF_INET; + src.sin_addr.s_addr = (__force u32)conn->c_laddr; + src.sin_port = (__force u16)htons(0); + + ret = sock->ops->bind(sock, (struct sockaddr *)&src, sizeof(src)); + if (ret) { + rdsdebug("bind failed with %d at address %u.%u.%u.%u\n", + ret, NIPQUAD(conn->c_laddr)); + goto out; + } + + dest.sin_family = AF_INET; + dest.sin_addr.s_addr = (__force u32)conn->c_faddr; + dest.sin_port = (__force u16)htons(RDS_TCP_PORT); + + /* + * once we call connect() we can start getting callbacks and they + * own the socket + */ + rds_tcp_set_callbacks(sock, conn); + ret = sock->ops->connect(sock, (struct sockaddr *)&dest, sizeof(dest), + O_NONBLOCK); + sock = NULL; + + rdsdebug("connect to address %u.%u.%u.%u returned %d\n", + NIPQUAD(conn->c_faddr), ret); + if (ret == -EINPROGRESS) + ret = 0; + +out: + if (sock) + sock_release(sock); + return ret; +} + +/* + * Before killing the tcp socket this needs to serialize with callbacks. The + * caller has already grabbed the sending sem so we're serialized with other + * senders. + * + * TCP calls the callbacks with the sock lock so we hold it while we reset the + * callbacks to those set by TCP. Our callbacks won't execute again once we + * hold the sock lock. + */ +void rds_tcp_conn_shutdown(struct rds_connection *conn) +{ + struct rds_tcp_connection *tc = conn->c_transport_data; + struct socket *sock = tc->t_sock; + + rdsdebug("shutting down conn %p tc %p sock %p\n", conn, tc, sock); + + if (sock) { + sock->ops->shutdown(sock, RCV_SHUTDOWN | SEND_SHUTDOWN); + lock_sock(sock->sk); + rds_tcp_restore_callbacks(sock, tc); /* tc->tc_sock = NULL */ + + release_sock(sock->sk); + sock_release(sock); + }; + + if (tc->t_tinc) { + rds_inc_put(&tc->t_tinc->ti_inc); + tc->t_tinc = NULL; + } + tc->t_tinc_hdr_rem = sizeof(struct rds_header); + tc->t_tinc_data_rem = 0; +} diff --git a/net/rds/tcp_listen.c b/net/rds/tcp_listen.c new file mode 100644 index 000000000000..24b743eb0b1b --- /dev/null +++ b/net/rds/tcp_listen.c @@ -0,0 +1,199 @@ +/* + * Copyright (c) 2006 Oracle. All rights reserved. + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * OpenIB.org BSD license below: + * + * Redistribution and use in source and binary forms, with or + * without modification, are permitted provided that the following + * conditions are met: + * + * - Redistributions of source code must retain the above + * copyright notice, this list of conditions and the following + * disclaimer. + * + * - Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials + * provided with the distribution. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ +#include +#include +#include + +#include "rds.h" +#include "tcp.h" + +/* + * cheesy, but simple.. + */ +static void rds_tcp_accept_worker(struct work_struct *work); +static DECLARE_WORK(rds_tcp_listen_work, rds_tcp_accept_worker); +static struct socket *rds_tcp_listen_sock; + +static int rds_tcp_accept_one(struct socket *sock) +{ + struct socket *new_sock = NULL; + struct rds_connection *conn; + int ret; + struct inet_sock *inet; + + ret = sock_create_lite(sock->sk->sk_family, sock->sk->sk_type, + sock->sk->sk_protocol, &new_sock); + if (ret) + goto out; + + new_sock->type = sock->type; + new_sock->ops = sock->ops; + ret = sock->ops->accept(sock, new_sock, O_NONBLOCK); + if (ret < 0) + goto out; + + rds_tcp_tune(new_sock); + + inet = inet_sk(new_sock->sk); + + rdsdebug("accepted tcp %u.%u.%u.%u:%u -> %u.%u.%u.%u:%u\n", + NIPQUAD(inet->saddr), ntohs(inet->sport), + NIPQUAD(inet->daddr), ntohs(inet->dport)); + + conn = rds_conn_create(inet->saddr, inet->daddr, &rds_tcp_transport, + GFP_KERNEL); + if (IS_ERR(conn)) { + ret = PTR_ERR(conn); + goto out; + } + + /* + * see the comment above rds_queue_delayed_reconnect() + */ + if (!rds_conn_transition(conn, RDS_CONN_DOWN, RDS_CONN_CONNECTING)) { + if (rds_conn_state(conn) == RDS_CONN_UP) + rds_tcp_stats_inc(s_tcp_listen_closed_stale); + else + rds_tcp_stats_inc(s_tcp_connect_raced); + rds_conn_drop(conn); + ret = 0; + goto out; + } + + rds_tcp_set_callbacks(new_sock, conn); + rds_connect_complete(conn); + new_sock = NULL; + ret = 0; + +out: + if (new_sock) + sock_release(new_sock); + return ret; +} + +static void rds_tcp_accept_worker(struct work_struct *work) +{ + while (rds_tcp_accept_one(rds_tcp_listen_sock) == 0) + cond_resched(); +} + +void rds_tcp_listen_data_ready(struct sock *sk, int bytes) +{ + void (*ready)(struct sock *sk, int bytes); + + rdsdebug("listen data ready sk %p\n", sk); + + read_lock(&sk->sk_callback_lock); + ready = sk->sk_user_data; + if (ready == NULL) { /* check for teardown race */ + ready = sk->sk_data_ready; + goto out; + } + + /* + * ->sk_data_ready is also called for a newly established child socket + * before it has been accepted and the accepter has set up their + * data_ready.. we only want to queue listen work for our listening + * socket + */ + if (sk->sk_state == TCP_LISTEN) + queue_work(rds_wq, &rds_tcp_listen_work); + +out: + read_unlock(&sk->sk_callback_lock); + ready(sk, bytes); +} + +int __init rds_tcp_listen_init(void) +{ + struct sockaddr_in sin; + struct socket *sock = NULL; + int ret; + + ret = sock_create(PF_INET, SOCK_STREAM, IPPROTO_TCP, &sock); + if (ret < 0) + goto out; + + sock->sk->sk_reuse = 1; + rds_tcp_nonagle(sock); + + write_lock_bh(&sock->sk->sk_callback_lock); + sock->sk->sk_user_data = sock->sk->sk_data_ready; + sock->sk->sk_data_ready = rds_tcp_listen_data_ready; + write_unlock_bh(&sock->sk->sk_callback_lock); + + sin.sin_family = PF_INET, + sin.sin_addr.s_addr = (__force u32)htonl(INADDR_ANY); + sin.sin_port = (__force u16)htons(RDS_TCP_PORT); + + ret = sock->ops->bind(sock, (struct sockaddr *)&sin, sizeof(sin)); + if (ret < 0) + goto out; + + ret = sock->ops->listen(sock, 64); + if (ret < 0) + goto out; + + rds_tcp_listen_sock = sock; + sock = NULL; +out: + if (sock) + sock_release(sock); + return ret; +} + +void rds_tcp_listen_stop(void) +{ + struct socket *sock = rds_tcp_listen_sock; + struct sock *sk; + + if (sock == NULL) + return; + + sk = sock->sk; + + /* serialize with and prevent further callbacks */ + lock_sock(sk); + write_lock_bh(&sk->sk_callback_lock); + if (sk->sk_user_data) { + sk->sk_data_ready = sk->sk_user_data; + sk->sk_user_data = NULL; + } + write_unlock_bh(&sk->sk_callback_lock); + release_sock(sk); + + /* wait for accepts to stop and close the socket */ + flush_workqueue(rds_wq); + sock_release(sock); + rds_tcp_listen_sock = NULL; +} diff --git a/net/rds/tcp_recv.c b/net/rds/tcp_recv.c new file mode 100644 index 000000000000..c00dafffbb5a --- /dev/null +++ b/net/rds/tcp_recv.c @@ -0,0 +1,356 @@ +/* + * Copyright (c) 2006 Oracle. All rights reserved. + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * OpenIB.org BSD license below: + * + * Redistribution and use in source and binary forms, with or + * without modification, are permitted provided that the following + * conditions are met: + * + * - Redistributions of source code must retain the above + * copyright notice, this list of conditions and the following + * disclaimer. + * + * - Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials + * provided with the distribution. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ +#include +#include + +#include "rds.h" +#include "tcp.h" + +static struct kmem_cache *rds_tcp_incoming_slab; + +void rds_tcp_inc_purge(struct rds_incoming *inc) +{ + struct rds_tcp_incoming *tinc; + tinc = container_of(inc, struct rds_tcp_incoming, ti_inc); + rdsdebug("purging tinc %p inc %p\n", tinc, inc); + skb_queue_purge(&tinc->ti_skb_list); +} + +void rds_tcp_inc_free(struct rds_incoming *inc) +{ + struct rds_tcp_incoming *tinc; + tinc = container_of(inc, struct rds_tcp_incoming, ti_inc); + rds_tcp_inc_purge(inc); + rdsdebug("freeing tinc %p inc %p\n", tinc, inc); + kmem_cache_free(rds_tcp_incoming_slab, tinc); +} + +/* + * this is pretty lame, but, whatever. + */ +int rds_tcp_inc_copy_to_user(struct rds_incoming *inc, struct iovec *first_iov, + size_t size) +{ + struct rds_tcp_incoming *tinc; + struct iovec *iov, tmp; + struct sk_buff *skb; + unsigned long to_copy, skb_off; + int ret = 0; + + if (size == 0) + goto out; + + tinc = container_of(inc, struct rds_tcp_incoming, ti_inc); + iov = first_iov; + tmp = *iov; + + skb_queue_walk(&tinc->ti_skb_list, skb) { + skb_off = 0; + while (skb_off < skb->len) { + while (tmp.iov_len == 0) { + iov++; + tmp = *iov; + } + + to_copy = min(tmp.iov_len, size); + to_copy = min(to_copy, skb->len - skb_off); + + rdsdebug("ret %d size %zu skb %p skb_off %lu " + "skblen %d iov_base %p iov_len %zu cpy %lu\n", + ret, size, skb, skb_off, skb->len, + tmp.iov_base, tmp.iov_len, to_copy); + + /* modifies tmp as it copies */ + if (skb_copy_datagram_iovec(skb, skb_off, &tmp, + to_copy)) { + ret = -EFAULT; + goto out; + } + + size -= to_copy; + ret += to_copy; + skb_off += to_copy; + if (size == 0) + goto out; + } + } +out: + return ret; +} + +/* + * We have a series of skbs that have fragmented pieces of the congestion + * bitmap. They must add up to the exact size of the congestion bitmap. We + * use the skb helpers to copy those into the pages that make up the in-memory + * congestion bitmap for the remote address of this connection. We then tell + * the congestion core that the bitmap has been changed so that it can wake up + * sleepers. + * + * This is racing with sending paths which are using test_bit to see if the + * bitmap indicates that their recipient is congested. + */ + +static void rds_tcp_cong_recv(struct rds_connection *conn, + struct rds_tcp_incoming *tinc) +{ + struct sk_buff *skb; + unsigned int to_copy, skb_off; + unsigned int map_off; + unsigned int map_page; + struct rds_cong_map *map; + int ret; + + /* catch completely corrupt packets */ + if (be32_to_cpu(tinc->ti_inc.i_hdr.h_len) != RDS_CONG_MAP_BYTES) + return; + + map_page = 0; + map_off = 0; + map = conn->c_fcong; + + skb_queue_walk(&tinc->ti_skb_list, skb) { + skb_off = 0; + while (skb_off < skb->len) { + to_copy = min_t(unsigned int, PAGE_SIZE - map_off, + skb->len - skb_off); + + BUG_ON(map_page >= RDS_CONG_MAP_PAGES); + + /* only returns 0 or -error */ + ret = skb_copy_bits(skb, skb_off, + (void *)map->m_page_addrs[map_page] + map_off, + to_copy); + BUG_ON(ret != 0); + + skb_off += to_copy; + map_off += to_copy; + if (map_off == PAGE_SIZE) { + map_off = 0; + map_page++; + } + } + } + + rds_cong_map_updated(map, ~(u64) 0); +} + +struct rds_tcp_desc_arg { + struct rds_connection *conn; + gfp_t gfp; + enum km_type km; +}; + +static int rds_tcp_data_recv(read_descriptor_t *desc, struct sk_buff *skb, + unsigned int offset, size_t len) +{ + struct rds_tcp_desc_arg *arg = desc->arg.data; + struct rds_connection *conn = arg->conn; + struct rds_tcp_connection *tc = conn->c_transport_data; + struct rds_tcp_incoming *tinc = tc->t_tinc; + struct sk_buff *clone; + size_t left = len, to_copy; + + rdsdebug("tcp data tc %p skb %p offset %u len %zu\n", tc, skb, offset, + len); + + /* + * tcp_read_sock() interprets partial progress as an indication to stop + * processing. + */ + while (left) { + if (tinc == NULL) { + tinc = kmem_cache_alloc(rds_tcp_incoming_slab, + arg->gfp); + if (tinc == NULL) { + desc->error = -ENOMEM; + goto out; + } + tc->t_tinc = tinc; + rdsdebug("alloced tinc %p\n", tinc); + rds_inc_init(&tinc->ti_inc, conn, conn->c_faddr); + /* + * XXX * we might be able to use the __ variants when + * we've already serialized at a higher level. + */ + skb_queue_head_init(&tinc->ti_skb_list); + } + + if (left && tc->t_tinc_hdr_rem) { + to_copy = min(tc->t_tinc_hdr_rem, left); + rdsdebug("copying %zu header from skb %p\n", to_copy, + skb); + skb_copy_bits(skb, offset, + (char *)&tinc->ti_inc.i_hdr + + sizeof(struct rds_header) - + tc->t_tinc_hdr_rem, + to_copy); + tc->t_tinc_hdr_rem -= to_copy; + left -= to_copy; + offset += to_copy; + + if (tc->t_tinc_hdr_rem == 0) { + /* could be 0 for a 0 len message */ + tc->t_tinc_data_rem = + be32_to_cpu(tinc->ti_inc.i_hdr.h_len); + } + } + + if (left && tc->t_tinc_data_rem) { + clone = skb_clone(skb, arg->gfp); + if (clone == NULL) { + desc->error = -ENOMEM; + goto out; + } + + to_copy = min(tc->t_tinc_data_rem, left); + pskb_pull(clone, offset); + pskb_trim(clone, to_copy); + skb_queue_tail(&tinc->ti_skb_list, clone); + + rdsdebug("skb %p data %p len %d off %u to_copy %zu -> " + "clone %p data %p len %d\n", + skb, skb->data, skb->len, offset, to_copy, + clone, clone->data, clone->len); + + tc->t_tinc_data_rem -= to_copy; + left -= to_copy; + offset += to_copy; + } + + if (tc->t_tinc_hdr_rem == 0 && tc->t_tinc_data_rem == 0) { + if (tinc->ti_inc.i_hdr.h_flags == RDS_FLAG_CONG_BITMAP) + rds_tcp_cong_recv(conn, tinc); + else + rds_recv_incoming(conn, conn->c_faddr, + conn->c_laddr, &tinc->ti_inc, + arg->gfp, arg->km); + + tc->t_tinc_hdr_rem = sizeof(struct rds_header); + tc->t_tinc_data_rem = 0; + tc->t_tinc = NULL; + rds_inc_put(&tinc->ti_inc); + tinc = NULL; + } + } +out: + rdsdebug("returning len %zu left %zu skb len %d rx queue depth %d\n", + len, left, skb->len, + skb_queue_len(&tc->t_sock->sk->sk_receive_queue)); + return len - left; +} + +/* the caller has to hold the sock lock */ +int rds_tcp_read_sock(struct rds_connection *conn, gfp_t gfp, enum km_type km) +{ + struct rds_tcp_connection *tc = conn->c_transport_data; + struct socket *sock = tc->t_sock; + read_descriptor_t desc; + struct rds_tcp_desc_arg arg; + + /* It's like glib in the kernel! */ + arg.conn = conn; + arg.gfp = gfp; + arg.km = km; + desc.arg.data = &arg; + desc.error = 0; + desc.count = 1; /* give more than one skb per call */ + + tcp_read_sock(sock->sk, &desc, rds_tcp_data_recv); + rdsdebug("tcp_read_sock for tc %p gfp 0x%x returned %d\n", tc, gfp, + desc.error); + + return desc.error; +} + +/* + * We hold the sock lock to serialize our rds_tcp_recv->tcp_read_sock from + * data_ready. + * + * if we fail to allocate we're in trouble.. blindly wait some time before + * trying again to see if the VM can free up something for us. + */ +int rds_tcp_recv(struct rds_connection *conn) +{ + struct rds_tcp_connection *tc = conn->c_transport_data; + struct socket *sock = tc->t_sock; + int ret = 0; + + rdsdebug("recv worker conn %p tc %p sock %p\n", conn, tc, sock); + + lock_sock(sock->sk); + ret = rds_tcp_read_sock(conn, GFP_KERNEL, KM_USER0); + release_sock(sock->sk); + + return ret; +} + +void rds_tcp_data_ready(struct sock *sk, int bytes) +{ + void (*ready)(struct sock *sk, int bytes); + struct rds_connection *conn; + struct rds_tcp_connection *tc; + + rdsdebug("data ready sk %p bytes %d\n", sk, bytes); + + read_lock(&sk->sk_callback_lock); + conn = sk->sk_user_data; + if (conn == NULL) { /* check for teardown race */ + ready = sk->sk_data_ready; + goto out; + } + + tc = conn->c_transport_data; + ready = tc->t_orig_data_ready; + rds_tcp_stats_inc(s_tcp_data_ready_calls); + + if (rds_tcp_read_sock(conn, GFP_ATOMIC, KM_SOFTIRQ0) == -ENOMEM) + queue_delayed_work(rds_wq, &conn->c_recv_w, 0); +out: + read_unlock(&sk->sk_callback_lock); + ready(sk, bytes); +} + +int __init rds_tcp_recv_init(void) +{ + rds_tcp_incoming_slab = kmem_cache_create("rds_tcp_incoming", + sizeof(struct rds_tcp_incoming), + 0, 0, NULL); + if (rds_tcp_incoming_slab == NULL) + return -ENOMEM; + return 0; +} + +void rds_tcp_recv_exit(void) +{ + kmem_cache_destroy(rds_tcp_incoming_slab); +} diff --git a/net/rds/tcp_send.c b/net/rds/tcp_send.c new file mode 100644 index 000000000000..ab545e0cd5d6 --- /dev/null +++ b/net/rds/tcp_send.c @@ -0,0 +1,263 @@ +/* + * Copyright (c) 2006 Oracle. All rights reserved. + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * OpenIB.org BSD license below: + * + * Redistribution and use in source and binary forms, with or + * without modification, are permitted provided that the following + * conditions are met: + * + * - Redistributions of source code must retain the above + * copyright notice, this list of conditions and the following + * disclaimer. + * + * - Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials + * provided with the distribution. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ +#include +#include +#include + +#include "rds.h" +#include "tcp.h" + +static void rds_tcp_cork(struct socket *sock, int val) +{ + mm_segment_t oldfs; + + oldfs = get_fs(); + set_fs(KERNEL_DS); + sock->ops->setsockopt(sock, SOL_TCP, TCP_CORK, (char __user *)&val, + sizeof(val)); + set_fs(oldfs); +} + +void rds_tcp_xmit_prepare(struct rds_connection *conn) +{ + struct rds_tcp_connection *tc = conn->c_transport_data; + + rds_tcp_cork(tc->t_sock, 1); +} + +void rds_tcp_xmit_complete(struct rds_connection *conn) +{ + struct rds_tcp_connection *tc = conn->c_transport_data; + + rds_tcp_cork(tc->t_sock, 0); +} + +/* the core send_sem serializes this with other xmit and shutdown */ +int rds_tcp_sendmsg(struct socket *sock, void *data, unsigned int len) +{ + struct kvec vec = { + .iov_base = data, + .iov_len = len, + }; + struct msghdr msg = { + .msg_flags = MSG_DONTWAIT | MSG_NOSIGNAL, + }; + + return kernel_sendmsg(sock, &msg, &vec, 1, vec.iov_len); +} + +/* the core send_sem serializes this with other xmit and shutdown */ +int rds_tcp_xmit_cong_map(struct rds_connection *conn, + struct rds_cong_map *map, unsigned long offset) +{ + static struct rds_header rds_tcp_map_header = { + .h_flags = RDS_FLAG_CONG_BITMAP, + }; + struct rds_tcp_connection *tc = conn->c_transport_data; + unsigned long i; + int ret; + int copied = 0; + + /* Some problem claims cpu_to_be32(constant) isn't a constant. */ + rds_tcp_map_header.h_len = cpu_to_be32(RDS_CONG_MAP_BYTES); + + if (offset < sizeof(struct rds_header)) { + ret = rds_tcp_sendmsg(tc->t_sock, + (void *)&rds_tcp_map_header + offset, + sizeof(struct rds_header) - offset); + if (ret <= 0) + return ret; + offset += ret; + copied = ret; + if (offset < sizeof(struct rds_header)) + return ret; + } + + offset -= sizeof(struct rds_header); + i = offset / PAGE_SIZE; + offset = offset % PAGE_SIZE; + BUG_ON(i >= RDS_CONG_MAP_PAGES); + + do { + ret = tc->t_sock->ops->sendpage(tc->t_sock, + virt_to_page(map->m_page_addrs[i]), + offset, PAGE_SIZE - offset, + MSG_DONTWAIT); + if (ret <= 0) + break; + copied += ret; + offset += ret; + if (offset == PAGE_SIZE) { + offset = 0; + i++; + } + } while (i < RDS_CONG_MAP_PAGES); + + return copied ? copied : ret; +} + +/* the core send_sem serializes this with other xmit and shutdown */ +int rds_tcp_xmit(struct rds_connection *conn, struct rds_message *rm, + unsigned int hdr_off, unsigned int sg, unsigned int off) +{ + struct rds_tcp_connection *tc = conn->c_transport_data; + int done = 0; + int ret = 0; + + if (hdr_off == 0) { + /* + * m_ack_seq is set to the sequence number of the last byte of + * header and data. see rds_tcp_is_acked(). + */ + tc->t_last_sent_nxt = rds_tcp_snd_nxt(tc); + rm->m_ack_seq = tc->t_last_sent_nxt + + sizeof(struct rds_header) + + be32_to_cpu(rm->m_inc.i_hdr.h_len) - 1; + smp_mb__before_clear_bit(); + set_bit(RDS_MSG_HAS_ACK_SEQ, &rm->m_flags); + tc->t_last_expected_una = rm->m_ack_seq + 1; + + rdsdebug("rm %p tcp nxt %u ack_seq %llu\n", + rm, rds_tcp_snd_nxt(tc), + (unsigned long long)rm->m_ack_seq); + } + + if (hdr_off < sizeof(struct rds_header)) { + /* see rds_tcp_write_space() */ + set_bit(SOCK_NOSPACE, &tc->t_sock->sk->sk_socket->flags); + + ret = rds_tcp_sendmsg(tc->t_sock, + (void *)&rm->m_inc.i_hdr + hdr_off, + sizeof(rm->m_inc.i_hdr) - hdr_off); + if (ret < 0) + goto out; + done += ret; + if (hdr_off + done != sizeof(struct rds_header)) + goto out; + } + + while (sg < rm->m_nents) { + ret = tc->t_sock->ops->sendpage(tc->t_sock, + sg_page(&rm->m_sg[sg]), + rm->m_sg[sg].offset + off, + rm->m_sg[sg].length - off, + MSG_DONTWAIT|MSG_NOSIGNAL); + rdsdebug("tcp sendpage %p:%u:%u ret %d\n", (void *)sg_page(&rm->m_sg[sg]), + rm->m_sg[sg].offset + off, rm->m_sg[sg].length - off, + ret); + if (ret <= 0) + break; + + off += ret; + done += ret; + if (off == rm->m_sg[sg].length) { + off = 0; + sg++; + } + } + +out: + if (ret <= 0) { + /* write_space will hit after EAGAIN, all else fatal */ + if (ret == -EAGAIN) { + rds_tcp_stats_inc(s_tcp_sndbuf_full); + ret = 0; + } else { + printk(KERN_WARNING "RDS/tcp: send to %u.%u.%u.%u " + "returned %d, disconnecting and reconnecting\n", + NIPQUAD(conn->c_faddr), ret); + rds_conn_drop(conn); + } + } + if (done == 0) + done = ret; + return done; +} + +/* + * rm->m_ack_seq is set to the tcp sequence number that corresponds to the + * last byte of the message, including the header. This means that the + * entire message has been received if rm->m_ack_seq is "before" the next + * unacked byte of the TCP sequence space. We have to do very careful + * wrapping 32bit comparisons here. + */ +static int rds_tcp_is_acked(struct rds_message *rm, uint64_t ack) +{ + if (!test_bit(RDS_MSG_HAS_ACK_SEQ, &rm->m_flags)) + return 0; + return (__s32)((u32)rm->m_ack_seq - (u32)ack) < 0; +} + +void rds_tcp_write_space(struct sock *sk) +{ + void (*write_space)(struct sock *sk); + struct rds_connection *conn; + struct rds_tcp_connection *tc; + + read_lock(&sk->sk_callback_lock); + conn = sk->sk_user_data; + if (conn == NULL) { + write_space = sk->sk_write_space; + goto out; + } + + tc = conn->c_transport_data; + rdsdebug("write_space for tc %p\n", tc); + write_space = tc->t_orig_write_space; + rds_tcp_stats_inc(s_tcp_write_space_calls); + + rdsdebug("tcp una %u\n", rds_tcp_snd_una(tc)); + tc->t_last_seen_una = rds_tcp_snd_una(tc); + rds_send_drop_acked(conn, rds_tcp_snd_una(tc), rds_tcp_is_acked); + + queue_delayed_work(rds_wq, &conn->c_send_w, 0); +out: + read_unlock(&sk->sk_callback_lock); + + /* + * write_space is only called when data leaves tcp's send queue if + * SOCK_NOSPACE is set. We set SOCK_NOSPACE every time we put + * data in tcp's send queue because we use write_space to parse the + * sequence numbers and notice that rds messages have been fully + * received. + * + * tcp's write_space clears SOCK_NOSPACE if the send queue has more + * than a certain amount of space. So we need to set it again *after* + * we call tcp's write_space or else we might only get called on the + * first of a series of incoming tcp acks. + */ + write_space(sk); + + if (sk->sk_socket) + set_bit(SOCK_NOSPACE, &sk->sk_socket->flags); +} diff --git a/net/rds/tcp_stats.c b/net/rds/tcp_stats.c new file mode 100644 index 000000000000..d5898d03cd68 --- /dev/null +++ b/net/rds/tcp_stats.c @@ -0,0 +1,74 @@ +/* + * Copyright (c) 2006 Oracle. All rights reserved. + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * OpenIB.org BSD license below: + * + * Redistribution and use in source and binary forms, with or + * without modification, are permitted provided that the following + * conditions are met: + * + * - Redistributions of source code must retain the above + * copyright notice, this list of conditions and the following + * disclaimer. + * + * - Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials + * provided with the distribution. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ +#include +#include +#include + +#include "rds.h" +#include "tcp.h" + +DEFINE_PER_CPU(struct rds_tcp_statistics, rds_tcp_stats) + ____cacheline_aligned; + +static const char const *rds_tcp_stat_names[] = { + "tcp_data_ready_calls", + "tcp_write_space_calls", + "tcp_sndbuf_full", + "tcp_connect_raced", + "tcp_listen_closed_stale", +}; + +unsigned int rds_tcp_stats_info_copy(struct rds_info_iterator *iter, + unsigned int avail) +{ + struct rds_tcp_statistics stats = {0, }; + uint64_t *src; + uint64_t *sum; + size_t i; + int cpu; + + if (avail < ARRAY_SIZE(rds_tcp_stat_names)) + goto out; + + for_each_online_cpu(cpu) { + src = (uint64_t *)&(per_cpu(rds_tcp_stats, cpu)); + sum = (uint64_t *)&stats; + for (i = 0; i < sizeof(stats) / sizeof(uint64_t); i++) + *(sum++) += *(src++); + } + + rds_stats_info_copy(iter, (uint64_t *)&stats, rds_tcp_stat_names, + ARRAY_SIZE(rds_tcp_stat_names)); +out: + return ARRAY_SIZE(rds_tcp_stat_names); +} -- cgit v1.2.3 From 6777d773a463ac045d333b989d4e44660f8d92ad Mon Sep 17 00:00:00 2001 From: Mimi Zohar Date: Fri, 21 Aug 2009 14:32:48 -0400 Subject: kernel_read: redefine offset type vfs_read() offset is defined as loff_t, but kernel_read() offset is only defined as unsigned long. Redefine kernel_read() offset as loff_t. Cc: stable@kernel.org Signed-off-by: Mimi Zohar Signed-off-by: James Morris --- fs/exec.c | 4 ++-- include/linux/fs.h | 2 +- 2 files changed, 3 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/fs/exec.c b/fs/exec.c index 4a8849e45b21..fb4f3cdda78c 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -678,8 +678,8 @@ exit: } EXPORT_SYMBOL(open_exec); -int kernel_read(struct file *file, unsigned long offset, - char *addr, unsigned long count) +int kernel_read(struct file *file, loff_t offset, + char *addr, unsigned long count) { mm_segment_t old_fs; loff_t pos = offset; diff --git a/include/linux/fs.h b/include/linux/fs.h index 67888a9e0655..73e9b643e455 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -2123,7 +2123,7 @@ extern struct file *do_filp_open(int dfd, const char *pathname, int open_flag, int mode, int acc_mode); extern int may_open(struct path *, int, int); -extern int kernel_read(struct file *, unsigned long, char *, unsigned long); +extern int kernel_read(struct file *, loff_t, char *, unsigned long); extern struct file * open_exec(const char *); /* fs/dcache.c -- generic fs support functions */ -- cgit v1.2.3 From 05ecd5a1f76c183cca381705b3adb7d77c9a0439 Mon Sep 17 00:00:00 2001 From: Pawel Moll Date: Mon, 24 Aug 2009 19:52:38 +0900 Subject: sh: Simplify "multi-evt" interrupt handling. This patch changes the way in which "multi-evt" interrups are handled. The intc_evt2irq_table and related intc_evt2irq() have been removed and the "redirecting" handler is installed for the coupled interrupts. Thanks to that the do_IRQ() function don't have to use another level of indirection for all the interrupts... Signed-off-by: Pawel Moll Signed-off-by: Stuart Menefy Signed-off-by: Paul Mundt --- arch/sh/kernel/irq.c | 2 +- drivers/sh/intc.c | 54 ++++++++++++++++--------------------------------- include/linux/sh_intc.h | 1 - 3 files changed, 18 insertions(+), 39 deletions(-) (limited to 'include') diff --git a/arch/sh/kernel/irq.c b/arch/sh/kernel/irq.c index 278c68c60488..d1053392e287 100644 --- a/arch/sh/kernel/irq.c +++ b/arch/sh/kernel/irq.c @@ -114,7 +114,7 @@ asmlinkage int do_IRQ(unsigned int irq, struct pt_regs *regs) #endif irq_enter(); - irq = irq_demux(intc_evt2irq(irq)); + irq = irq_demux(evt2irq(irq)); #ifdef CONFIG_IRQSTACKS curctx = (union irq_ctx *)current_thread_info(); diff --git a/drivers/sh/intc.c b/drivers/sh/intc.c index 4b1ca9d28353..a9174ec72853 100644 --- a/drivers/sh/intc.c +++ b/drivers/sh/intc.c @@ -663,16 +663,9 @@ static unsigned int __init save_reg(struct intc_desc_int *d, return 0; } -static unsigned char *intc_evt2irq_table; - -unsigned int intc_evt2irq(unsigned int vector) +static void intc_redirect_irq(unsigned int irq, struct irq_desc *desc) { - unsigned int irq = evt2irq(vector); - - if (intc_evt2irq_table && intc_evt2irq_table[irq]) - irq = intc_evt2irq_table[irq]; - - return irq; + generic_handle_irq((unsigned int)get_irq_data(irq)); } void __init register_intc_controller(struct intc_desc *desc) @@ -745,34 +738,6 @@ void __init register_intc_controller(struct intc_desc *desc) BUG_ON(k > 256); /* _INTC_ADDR_E() and _INTC_ADDR_D() are 8 bits */ - /* keep the first vector only if same enum is used multiple times */ - for (i = 0; i < desc->nr_vectors; i++) { - struct intc_vect *vect = desc->vectors + i; - int first_irq = evt2irq(vect->vect); - - if (!vect->enum_id) - continue; - - for (k = i + 1; k < desc->nr_vectors; k++) { - struct intc_vect *vect2 = desc->vectors + k; - - if (vect->enum_id != vect2->enum_id) - continue; - - vect2->enum_id = 0; - - if (!intc_evt2irq_table) - intc_evt2irq_table = kzalloc(NR_IRQS, GFP_NOWAIT); - - if (!intc_evt2irq_table) { - pr_warning("intc: cannot allocate evt2irq!\n"); - continue; - } - - intc_evt2irq_table[evt2irq(vect2->vect)] = first_irq; - } - } - /* register the vectors one by one */ for (i = 0; i < desc->nr_vectors; i++) { struct intc_vect *vect = desc->vectors + i; @@ -789,6 +754,21 @@ void __init register_intc_controller(struct intc_desc *desc) } intc_register_irq(desc, d, vect->enum_id, irq); + + for (k = i + 1; k < desc->nr_vectors; k++) { + struct intc_vect *vect2 = desc->vectors + k; + unsigned int irq2 = evt2irq(vect2->vect); + + if (vect->enum_id != vect2->enum_id) + continue; + + vect2->enum_id = 0; + + /* redirect this interrupts to the first one */ + set_irq_chip_and_handler_name(irq2, &d->chip, + intc_redirect_irq, "redirect"); + set_irq_data(irq2, (void *)irq); + } } } diff --git a/include/linux/sh_intc.h b/include/linux/sh_intc.h index eb1423a0078d..68e212ff9dde 100644 --- a/include/linux/sh_intc.h +++ b/include/linux/sh_intc.h @@ -85,7 +85,6 @@ struct intc_desc symbol __initdata = { \ } #endif -unsigned int intc_evt2irq(unsigned int vector); void __init register_intc_controller(struct intc_desc *desc); int intc_set_priority(unsigned int irq, unsigned int prio); -- cgit v1.2.3 From 35aad0ffdf548617940ca1e78be1f2e0bafc4496 Mon Sep 17 00:00:00 2001 From: Jan Engelhardt Date: Mon, 24 Aug 2009 14:56:30 +0200 Subject: netfilter: xtables: mark initial tables constant The inputted table is never modified, so should be considered const. Signed-off-by: Jan Engelhardt Signed-off-by: Patrick McHardy --- include/linux/netfilter/x_tables.h | 2 +- include/linux/netfilter_arp/arp_tables.h | 2 +- include/linux/netfilter_bridge/ebtables.h | 2 +- include/linux/netfilter_ipv4/ip_tables.h | 2 +- include/linux/netfilter_ipv6/ip6_tables.h | 2 +- net/bridge/netfilter/ebtable_broute.c | 2 +- net/bridge/netfilter/ebtable_filter.c | 2 +- net/bridge/netfilter/ebtables.c | 13 +++++++------ net/ipv4/netfilter/arp_tables.c | 3 ++- net/ipv4/netfilter/arptable_filter.c | 4 ++-- net/ipv4/netfilter/ip_tables.c | 3 ++- net/ipv4/netfilter/iptable_filter.c | 2 +- net/ipv4/netfilter/iptable_mangle.c | 4 ++-- net/ipv4/netfilter/iptable_raw.c | 4 ++-- net/ipv4/netfilter/iptable_security.c | 4 ++-- net/ipv4/netfilter/nf_nat_rule.c | 4 ++-- net/ipv6/netfilter/ip6_tables.c | 3 ++- net/ipv6/netfilter/ip6table_filter.c | 2 +- net/ipv6/netfilter/ip6table_mangle.c | 4 ++-- net/ipv6/netfilter/ip6table_raw.c | 4 ++-- net/ipv6/netfilter/ip6table_security.c | 4 ++-- net/netfilter/x_tables.c | 7 ++++--- 22 files changed, 42 insertions(+), 37 deletions(-) (limited to 'include') diff --git a/include/linux/netfilter/x_tables.h b/include/linux/netfilter/x_tables.h index 4fa6e4c263e0..812cb153cabb 100644 --- a/include/linux/netfilter/x_tables.h +++ b/include/linux/netfilter/x_tables.h @@ -407,7 +407,7 @@ extern int xt_check_target(struct xt_tgchk_param *, unsigned int size, u_int8_t proto, bool inv_proto); extern struct xt_table *xt_register_table(struct net *net, - struct xt_table *table, + const struct xt_table *table, struct xt_table_info *bootstrap, struct xt_table_info *newinfo); extern void *xt_unregister_table(struct xt_table *table); diff --git a/include/linux/netfilter_arp/arp_tables.h b/include/linux/netfilter_arp/arp_tables.h index 590ac3d6d5d6..6fe3e6aa10db 100644 --- a/include/linux/netfilter_arp/arp_tables.h +++ b/include/linux/netfilter_arp/arp_tables.h @@ -265,7 +265,7 @@ struct arpt_error } extern struct xt_table *arpt_register_table(struct net *net, - struct xt_table *table, + const struct xt_table *table, const struct arpt_replace *repl); extern void arpt_unregister_table(struct xt_table *table); extern unsigned int arpt_do_table(struct sk_buff *skb, diff --git a/include/linux/netfilter_bridge/ebtables.h b/include/linux/netfilter_bridge/ebtables.h index e40ddb94b1af..ea281e6a2048 100644 --- a/include/linux/netfilter_bridge/ebtables.h +++ b/include/linux/netfilter_bridge/ebtables.h @@ -301,7 +301,7 @@ struct ebt_table #define EBT_ALIGN(s) (((s) + (__alignof__(struct ebt_replace)-1)) & \ ~(__alignof__(struct ebt_replace)-1)) extern struct ebt_table *ebt_register_table(struct net *net, - struct ebt_table *table); + const struct ebt_table *table); extern void ebt_unregister_table(struct ebt_table *table); extern unsigned int ebt_do_table(unsigned int hook, struct sk_buff *skb, const struct net_device *in, const struct net_device *out, diff --git a/include/linux/netfilter_ipv4/ip_tables.h b/include/linux/netfilter_ipv4/ip_tables.h index 092bd50581a9..61fafc868a7b 100644 --- a/include/linux/netfilter_ipv4/ip_tables.h +++ b/include/linux/netfilter_ipv4/ip_tables.h @@ -245,7 +245,7 @@ ipt_get_target(struct ipt_entry *e) extern void ipt_init(void) __init; extern struct xt_table *ipt_register_table(struct net *net, - struct xt_table *table, + const struct xt_table *table, const struct ipt_replace *repl); extern void ipt_unregister_table(struct xt_table *table); diff --git a/include/linux/netfilter_ipv6/ip6_tables.h b/include/linux/netfilter_ipv6/ip6_tables.h index 1089e33cf633..a64e1451ac38 100644 --- a/include/linux/netfilter_ipv6/ip6_tables.h +++ b/include/linux/netfilter_ipv6/ip6_tables.h @@ -306,7 +306,7 @@ ip6t_get_target(struct ip6t_entry *e) extern void ip6t_init(void) __init; extern struct xt_table *ip6t_register_table(struct net *net, - struct xt_table *table, + const struct xt_table *table, const struct ip6t_replace *repl); extern void ip6t_unregister_table(struct xt_table *table); extern unsigned int ip6t_do_table(struct sk_buff *skb, diff --git a/net/bridge/netfilter/ebtable_broute.c b/net/bridge/netfilter/ebtable_broute.c index c751111440f8..d32ab13e728c 100644 --- a/net/bridge/netfilter/ebtable_broute.c +++ b/net/bridge/netfilter/ebtable_broute.c @@ -41,7 +41,7 @@ static int check(const struct ebt_table_info *info, unsigned int valid_hooks) return 0; } -static struct ebt_table broute_table = +static const struct ebt_table broute_table = { .name = "broute", .table = &initial_table, diff --git a/net/bridge/netfilter/ebtable_filter.c b/net/bridge/netfilter/ebtable_filter.c index 4b988db3cd4d..60b1a6ca7185 100644 --- a/net/bridge/netfilter/ebtable_filter.c +++ b/net/bridge/netfilter/ebtable_filter.c @@ -50,7 +50,7 @@ static int check(const struct ebt_table_info *info, unsigned int valid_hooks) return 0; } -static struct ebt_table frame_filter = +static const struct ebt_table frame_filter = { .name = "filter", .table = &initial_table, diff --git a/net/bridge/netfilter/ebtables.c b/net/bridge/netfilter/ebtables.c index 37928d5f2840..bd1c65425d4f 100644 --- a/net/bridge/netfilter/ebtables.c +++ b/net/bridge/netfilter/ebtables.c @@ -1103,23 +1103,24 @@ free_newinfo: return ret; } -struct ebt_table *ebt_register_table(struct net *net, struct ebt_table *table) +struct ebt_table * +ebt_register_table(struct net *net, const struct ebt_table *input_table) { struct ebt_table_info *newinfo; - struct ebt_table *t; + struct ebt_table *t, *table; struct ebt_replace_kernel *repl; int ret, i, countersize; void *p; - if (!table || !(repl = table->table) || !repl->entries || - repl->entries_size == 0 || - repl->counters || table->private) { + if (input_table == NULL || (repl = input_table->table) == NULL || + repl->entries == 0 || repl->entries_size == 0 || + repl->counters != NULL || input_table->private != NULL) { BUGPRINT("Bad table data for ebt_register_table!!!\n"); return ERR_PTR(-EINVAL); } /* Don't add one table to multiple lists. */ - table = kmemdup(table, sizeof(struct ebt_table), GFP_KERNEL); + table = kmemdup(input_table, sizeof(struct ebt_table), GFP_KERNEL); if (!table) { ret = -ENOMEM; goto out; diff --git a/net/ipv4/netfilter/arp_tables.c b/net/ipv4/netfilter/arp_tables.c index 7bc11ffbb845..27774c99d888 100644 --- a/net/ipv4/netfilter/arp_tables.c +++ b/net/ipv4/netfilter/arp_tables.c @@ -1778,7 +1778,8 @@ static int do_arpt_get_ctl(struct sock *sk, int cmd, void __user *user, int *len return ret; } -struct xt_table *arpt_register_table(struct net *net, struct xt_table *table, +struct xt_table *arpt_register_table(struct net *net, + const struct xt_table *table, const struct arpt_replace *repl) { int ret; diff --git a/net/ipv4/netfilter/arptable_filter.c b/net/ipv4/netfilter/arptable_filter.c index 6ecfdae7c589..97337601827a 100644 --- a/net/ipv4/netfilter/arptable_filter.c +++ b/net/ipv4/netfilter/arptable_filter.c @@ -15,7 +15,7 @@ MODULE_DESCRIPTION("arptables filter table"); #define FILTER_VALID_HOOKS ((1 << NF_ARP_IN) | (1 << NF_ARP_OUT) | \ (1 << NF_ARP_FORWARD)) -static struct +static const struct { struct arpt_replace repl; struct arpt_standard entries[3]; @@ -45,7 +45,7 @@ static struct .term = ARPT_ERROR_INIT, }; -static struct xt_table packet_filter = { +static const struct xt_table packet_filter = { .name = "filter", .valid_hooks = FILTER_VALID_HOOKS, .me = THIS_MODULE, diff --git a/net/ipv4/netfilter/ip_tables.c b/net/ipv4/netfilter/ip_tables.c index 0b43fd7ca04a..cde755d5eeab 100644 --- a/net/ipv4/netfilter/ip_tables.c +++ b/net/ipv4/netfilter/ip_tables.c @@ -2065,7 +2065,8 @@ do_ipt_get_ctl(struct sock *sk, int cmd, void __user *user, int *len) return ret; } -struct xt_table *ipt_register_table(struct net *net, struct xt_table *table, +struct xt_table *ipt_register_table(struct net *net, + const struct xt_table *table, const struct ipt_replace *repl) { int ret; diff --git a/net/ipv4/netfilter/iptable_filter.c b/net/ipv4/netfilter/iptable_filter.c index 97dbd94a8e37..df566cbd68e5 100644 --- a/net/ipv4/netfilter/iptable_filter.c +++ b/net/ipv4/netfilter/iptable_filter.c @@ -53,7 +53,7 @@ static struct .term = IPT_ERROR_INIT, /* ERROR */ }; -static struct xt_table packet_filter = { +static const struct xt_table packet_filter = { .name = "filter", .valid_hooks = FILTER_VALID_HOOKS, .me = THIS_MODULE, diff --git a/net/ipv4/netfilter/iptable_mangle.c b/net/ipv4/netfilter/iptable_mangle.c index 28647f10aa7e..036047f9b0f2 100644 --- a/net/ipv4/netfilter/iptable_mangle.c +++ b/net/ipv4/netfilter/iptable_mangle.c @@ -28,7 +28,7 @@ MODULE_DESCRIPTION("iptables mangle table"); (1 << NF_INET_POST_ROUTING)) /* Ouch - five different hooks? Maybe this should be a config option..... -- BC */ -static struct +static const struct { struct ipt_replace repl; struct ipt_standard entries[5]; @@ -64,7 +64,7 @@ static struct .term = IPT_ERROR_INIT, /* ERROR */ }; -static struct xt_table packet_mangler = { +static const struct xt_table packet_mangler = { .name = "mangle", .valid_hooks = MANGLE_VALID_HOOKS, .me = THIS_MODULE, diff --git a/net/ipv4/netfilter/iptable_raw.c b/net/ipv4/netfilter/iptable_raw.c index 494784c999eb..993edc23be09 100644 --- a/net/ipv4/netfilter/iptable_raw.c +++ b/net/ipv4/netfilter/iptable_raw.c @@ -9,7 +9,7 @@ #define RAW_VALID_HOOKS ((1 << NF_INET_PRE_ROUTING) | (1 << NF_INET_LOCAL_OUT)) -static struct +static const struct { struct ipt_replace repl; struct ipt_standard entries[2]; @@ -36,7 +36,7 @@ static struct .term = IPT_ERROR_INIT, /* ERROR */ }; -static struct xt_table packet_raw = { +static const struct xt_table packet_raw = { .name = "raw", .valid_hooks = RAW_VALID_HOOKS, .me = THIS_MODULE, diff --git a/net/ipv4/netfilter/iptable_security.c b/net/ipv4/netfilter/iptable_security.c index 8804e1a0f915..99eb76c65d25 100644 --- a/net/ipv4/netfilter/iptable_security.c +++ b/net/ipv4/netfilter/iptable_security.c @@ -27,7 +27,7 @@ MODULE_DESCRIPTION("iptables security table, for MAC rules"); (1 << NF_INET_FORWARD) | \ (1 << NF_INET_LOCAL_OUT) -static struct +static const struct { struct ipt_replace repl; struct ipt_standard entries[3]; @@ -57,7 +57,7 @@ static struct .term = IPT_ERROR_INIT, /* ERROR */ }; -static struct xt_table security_table = { +static const struct xt_table security_table = { .name = "security", .valid_hooks = SECURITY_VALID_HOOKS, .me = THIS_MODULE, diff --git a/net/ipv4/netfilter/nf_nat_rule.c b/net/ipv4/netfilter/nf_nat_rule.c index 6448a9b7d6f0..9e81e0dfb4ec 100644 --- a/net/ipv4/netfilter/nf_nat_rule.c +++ b/net/ipv4/netfilter/nf_nat_rule.c @@ -28,7 +28,7 @@ (1 << NF_INET_POST_ROUTING) | \ (1 << NF_INET_LOCAL_OUT)) -static struct +static const struct { struct ipt_replace repl; struct ipt_standard entries[3]; @@ -58,7 +58,7 @@ static struct .term = IPT_ERROR_INIT, /* ERROR */ }; -static struct xt_table nat_table = { +static const struct xt_table nat_table = { .name = "nat", .valid_hooks = NAT_VALID_HOOKS, .me = THIS_MODULE, diff --git a/net/ipv6/netfilter/ip6_tables.c b/net/ipv6/netfilter/ip6_tables.c index a5d0c27cc26f..cc9f8ef303fd 100644 --- a/net/ipv6/netfilter/ip6_tables.c +++ b/net/ipv6/netfilter/ip6_tables.c @@ -2100,7 +2100,8 @@ do_ip6t_get_ctl(struct sock *sk, int cmd, void __user *user, int *len) return ret; } -struct xt_table *ip6t_register_table(struct net *net, struct xt_table *table, +struct xt_table *ip6t_register_table(struct net *net, + const struct xt_table *table, const struct ip6t_replace *repl) { int ret; diff --git a/net/ipv6/netfilter/ip6table_filter.c b/net/ipv6/netfilter/ip6table_filter.c index 0a3ae48ac4d5..6f4383ad86f9 100644 --- a/net/ipv6/netfilter/ip6table_filter.c +++ b/net/ipv6/netfilter/ip6table_filter.c @@ -51,7 +51,7 @@ static struct .term = IP6T_ERROR_INIT, /* ERROR */ }; -static struct xt_table packet_filter = { +static const struct xt_table packet_filter = { .name = "filter", .valid_hooks = FILTER_VALID_HOOKS, .me = THIS_MODULE, diff --git a/net/ipv6/netfilter/ip6table_mangle.c b/net/ipv6/netfilter/ip6table_mangle.c index 0f49e005a8c5..0ad91433ed61 100644 --- a/net/ipv6/netfilter/ip6table_mangle.c +++ b/net/ipv6/netfilter/ip6table_mangle.c @@ -21,7 +21,7 @@ MODULE_DESCRIPTION("ip6tables mangle table"); (1 << NF_INET_LOCAL_OUT) | \ (1 << NF_INET_POST_ROUTING)) -static struct +static const struct { struct ip6t_replace repl; struct ip6t_standard entries[5]; @@ -57,7 +57,7 @@ static struct .term = IP6T_ERROR_INIT, /* ERROR */ }; -static struct xt_table packet_mangler = { +static const struct xt_table packet_mangler = { .name = "mangle", .valid_hooks = MANGLE_VALID_HOOKS, .me = THIS_MODULE, diff --git a/net/ipv6/netfilter/ip6table_raw.c b/net/ipv6/netfilter/ip6table_raw.c index 679865e3d5ff..ed1a1180f3b3 100644 --- a/net/ipv6/netfilter/ip6table_raw.c +++ b/net/ipv6/netfilter/ip6table_raw.c @@ -8,7 +8,7 @@ #define RAW_VALID_HOOKS ((1 << NF_INET_PRE_ROUTING) | (1 << NF_INET_LOCAL_OUT)) -static struct +static const struct { struct ip6t_replace repl; struct ip6t_standard entries[2]; @@ -35,7 +35,7 @@ static struct .term = IP6T_ERROR_INIT, /* ERROR */ }; -static struct xt_table packet_raw = { +static const struct xt_table packet_raw = { .name = "raw", .valid_hooks = RAW_VALID_HOOKS, .me = THIS_MODULE, diff --git a/net/ipv6/netfilter/ip6table_security.c b/net/ipv6/netfilter/ip6table_security.c index 822afabbdc88..41b444c60934 100644 --- a/net/ipv6/netfilter/ip6table_security.c +++ b/net/ipv6/netfilter/ip6table_security.c @@ -26,7 +26,7 @@ MODULE_DESCRIPTION("ip6tables security table, for MAC rules"); (1 << NF_INET_FORWARD) | \ (1 << NF_INET_LOCAL_OUT) -static struct +static const struct { struct ip6t_replace repl; struct ip6t_standard entries[3]; @@ -56,7 +56,7 @@ static struct .term = IP6T_ERROR_INIT, /* ERROR */ }; -static struct xt_table security_table = { +static const struct xt_table security_table = { .name = "security", .valid_hooks = SECURITY_VALID_HOOKS, .me = THIS_MODULE, diff --git a/net/netfilter/x_tables.c b/net/netfilter/x_tables.c index 025d1a0af78b..a6ac83a93348 100644 --- a/net/netfilter/x_tables.c +++ b/net/netfilter/x_tables.c @@ -736,16 +736,17 @@ xt_replace_table(struct xt_table *table, } EXPORT_SYMBOL_GPL(xt_replace_table); -struct xt_table *xt_register_table(struct net *net, struct xt_table *table, +struct xt_table *xt_register_table(struct net *net, + const struct xt_table *input_table, struct xt_table_info *bootstrap, struct xt_table_info *newinfo) { int ret; struct xt_table_info *private; - struct xt_table *t; + struct xt_table *t, *table; /* Don't add one object to multiple lists. */ - table = kmemdup(table, sizeof(struct xt_table), GFP_KERNEL); + table = kmemdup(input_table, sizeof(struct xt_table), GFP_KERNEL); if (!table) { ret = -ENOMEM; goto out; -- cgit v1.2.3 From 7c614d6461399acca5c0ba444f5db49cb332fc08 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Mon, 24 Aug 2009 09:42:00 -0700 Subject: rcu: Add "notrace" to RCU function headers used by ftrace Both rcu_read_lock_sched_notrace() and rcu_read_unlock_sched_notrace() are used by ftrace, and thus need to be marked "notrace". Unfortunately, my naive assumption that gcc would see the inner "notrace" does not hold. Kudos to Lai Jiangshan for noting this. Reported-by: Ingo Molnar Bug-spotted-by: Lai Jiangshan Signed-off-by: Paul E. McKenney Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12511321213243-git-send-email-> Signed-off-by: Ingo Molnar --- include/linux/rcupdate.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/include/linux/rcupdate.h b/include/linux/rcupdate.h index ec90fc34fea9..8b4422c558da 100644 --- a/include/linux/rcupdate.h +++ b/include/linux/rcupdate.h @@ -191,7 +191,7 @@ static inline void rcu_read_lock_sched(void) __acquire(RCU_SCHED); rcu_read_acquire(); } -static inline void rcu_read_lock_sched_notrace(void) +static inline notrace void rcu_read_lock_sched_notrace(void) { preempt_disable_notrace(); __acquire(RCU_SCHED); @@ -209,7 +209,7 @@ static inline void rcu_read_unlock_sched(void) __release(RCU_SCHED); preempt_enable(); } -static inline void rcu_read_unlock_sched_notrace(void) +static inline notrace void rcu_read_unlock_sched_notrace(void) { rcu_read_release(); __release(RCU_SCHED); -- cgit v1.2.3 From 353d5c30c666580347515da609dd74a2b8e9b828 Mon Sep 17 00:00:00 2001 From: Hugh Dickins Date: Mon, 24 Aug 2009 16:30:28 +0100 Subject: mm: fix hugetlb bug due to user_shm_unlock call 2.6.30's commit 8a0bdec194c21c8fdef840989d0d7b742bb5d4bc removed user_shm_lock() calls in hugetlb_file_setup() but left the user_shm_unlock call in shm_destroy(). In detail: Assume that can_do_hugetlb_shm() returns true and hence user_shm_lock() is not called in hugetlb_file_setup(). However, user_shm_unlock() is called in any case in shm_destroy() and in the following atomic_dec_and_lock(&up->__count) in free_uid() is executed and if up->__count gets zero, also cleanup_user_struct() is scheduled. Note that sched_destroy_user() is empty if CONFIG_USER_SCHED is not set. However, the ref counter up->__count gets unexpectedly non-positive and the corresponding structs are freed even though there are live references to them, resulting in a kernel oops after a lots of shmget(SHM_HUGETLB)/shmctl(IPC_RMID) cycles and CONFIG_USER_SCHED set. Hugh changed Stefan's suggested patch: can_do_hugetlb_shm() at the time of shm_destroy() may give a different answer from at the time of hugetlb_file_setup(). And fixed newseg()'s no_id error path, which has missed user_shm_unlock() ever since it came in 2.6.9. Reported-by: Stefan Huber Signed-off-by: Hugh Dickins Tested-by: Stefan Huber Cc: stable@kernel.org Signed-off-by: Linus Torvalds --- fs/hugetlbfs/inode.c | 20 ++++++++++++-------- include/linux/hugetlb.h | 6 ++++-- ipc/shm.c | 8 +++++--- 3 files changed, 21 insertions(+), 13 deletions(-) (limited to 'include') diff --git a/fs/hugetlbfs/inode.c b/fs/hugetlbfs/inode.c index 941c8425c10b..cb88dac8ccaa 100644 --- a/fs/hugetlbfs/inode.c +++ b/fs/hugetlbfs/inode.c @@ -935,26 +935,28 @@ static int can_do_hugetlb_shm(void) return capable(CAP_IPC_LOCK) || in_group_p(sysctl_hugetlb_shm_group); } -struct file *hugetlb_file_setup(const char *name, size_t size, int acctflag) +struct file *hugetlb_file_setup(const char *name, size_t size, int acctflag, + struct user_struct **user) { int error = -ENOMEM; - int unlock_shm = 0; struct file *file; struct inode *inode; struct dentry *dentry, *root; struct qstr quick_string; - struct user_struct *user = current_user(); + *user = NULL; if (!hugetlbfs_vfsmount) return ERR_PTR(-ENOENT); if (!can_do_hugetlb_shm()) { - if (user_shm_lock(size, user)) { - unlock_shm = 1; + *user = current_user(); + if (user_shm_lock(size, *user)) { WARN_ONCE(1, "Using mlock ulimits for SHM_HUGETLB deprecated\n"); - } else + } else { + *user = NULL; return ERR_PTR(-EPERM); + } } root = hugetlbfs_vfsmount->mnt_root; @@ -996,8 +998,10 @@ out_inode: out_dentry: dput(dentry); out_shm_unlock: - if (unlock_shm) - user_shm_unlock(size, user); + if (*user) { + user_shm_unlock(size, *user); + *user = NULL; + } return ERR_PTR(error); } diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h index 2723513a5651..5cbc620bdfe0 100644 --- a/include/linux/hugetlb.h +++ b/include/linux/hugetlb.h @@ -10,6 +10,7 @@ #include struct ctl_table; +struct user_struct; int PageHuge(struct page *page); @@ -146,7 +147,8 @@ static inline struct hugetlbfs_sb_info *HUGETLBFS_SB(struct super_block *sb) extern const struct file_operations hugetlbfs_file_operations; extern struct vm_operations_struct hugetlb_vm_ops; -struct file *hugetlb_file_setup(const char *name, size_t, int); +struct file *hugetlb_file_setup(const char *name, size_t size, int acct, + struct user_struct **user); int hugetlb_get_quota(struct address_space *mapping, long delta); void hugetlb_put_quota(struct address_space *mapping, long delta); @@ -168,7 +170,7 @@ static inline void set_file_hugepages(struct file *file) #define is_file_hugepages(file) 0 #define set_file_hugepages(file) BUG() -#define hugetlb_file_setup(name,size,acctflag) ERR_PTR(-ENOSYS) +#define hugetlb_file_setup(name,size,acct,user) ERR_PTR(-ENOSYS) #endif /* !CONFIG_HUGETLBFS */ diff --git a/ipc/shm.c b/ipc/shm.c index 15dd238e5338..1bc4701ef4f0 100644 --- a/ipc/shm.c +++ b/ipc/shm.c @@ -174,7 +174,7 @@ static void shm_destroy(struct ipc_namespace *ns, struct shmid_kernel *shp) shm_unlock(shp); if (!is_file_hugepages(shp->shm_file)) shmem_lock(shp->shm_file, 0, shp->mlock_user); - else + else if (shp->mlock_user) user_shm_unlock(shp->shm_file->f_path.dentry->d_inode->i_size, shp->mlock_user); fput (shp->shm_file); @@ -369,8 +369,8 @@ static int newseg(struct ipc_namespace *ns, struct ipc_params *params) /* hugetlb_file_setup applies strict accounting */ if (shmflg & SHM_NORESERVE) acctflag = VM_NORESERVE; - file = hugetlb_file_setup(name, size, acctflag); - shp->mlock_user = current_user(); + file = hugetlb_file_setup(name, size, acctflag, + &shp->mlock_user); } else { /* * Do not allow no accounting for OVERCOMMIT_NEVER, even @@ -410,6 +410,8 @@ static int newseg(struct ipc_namespace *ns, struct ipc_params *params) return error; no_id: + if (shp->mlock_user) /* shmflg & SHM_HUGETLB case */ + user_shm_unlock(size, shp->mlock_user); fput(file); no_file: security_shm_free(shp); -- cgit v1.2.3 From a4be7c2778d1fd8e0a8a5607bec91fa221ab2c91 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Wed, 19 Aug 2009 11:18:27 +0200 Subject: perf_counter: Allow sharing of output channels Provide the ability to configure a counter to send its output to another (already existing) counter's output stream. Signed-off-by: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: Frederic Weisbecker Cc: Mike Galbraith Cc: stephane eranian Cc: Paul Mackerras LKML-Reference: <20090819092023.980284148@chello.nl> Signed-off-by: Ingo Molnar --- include/linux/perf_counter.h | 5 +++ kernel/perf_counter.c | 83 ++++++++++++++++++++++++++++++++++++++++++-- 2 files changed, 85 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/include/linux/perf_counter.h b/include/linux/perf_counter.h index b53f7006cc4e..e022b847c90d 100644 --- a/include/linux/perf_counter.h +++ b/include/linux/perf_counter.h @@ -216,6 +216,7 @@ struct perf_counter_attr { #define PERF_COUNTER_IOC_REFRESH _IO ('$', 2) #define PERF_COUNTER_IOC_RESET _IO ('$', 3) #define PERF_COUNTER_IOC_PERIOD _IOW('$', 4, u64) +#define PERF_COUNTER_IOC_SET_OUTPUT _IO ('$', 5) enum perf_counter_ioc_flags { PERF_IOC_FLAG_GROUP = 1U << 0, @@ -415,6 +416,9 @@ enum perf_callchain_context { PERF_CONTEXT_MAX = (__u64)-4095, }; +#define PERF_FLAG_FD_NO_GROUP (1U << 0) +#define PERF_FLAG_FD_OUTPUT (1U << 1) + #ifdef __KERNEL__ /* * Kernel-internal data types and definitions: @@ -536,6 +540,7 @@ struct perf_counter { struct list_head sibling_list; int nr_siblings; struct perf_counter *group_leader; + struct perf_counter *output; const struct pmu *pmu; enum perf_counter_active_state state; diff --git a/kernel/perf_counter.c b/kernel/perf_counter.c index 06bf6a4f2608..53abcbefa0bf 100644 --- a/kernel/perf_counter.c +++ b/kernel/perf_counter.c @@ -1692,6 +1692,11 @@ static void free_counter(struct perf_counter *counter) atomic_dec(&nr_task_counters); } + if (counter->output) { + fput(counter->output->filp); + counter->output = NULL; + } + if (counter->destroy) counter->destroy(counter); @@ -1977,6 +1982,8 @@ unlock: return ret; } +int perf_counter_set_output(struct perf_counter *counter, int output_fd); + static long perf_ioctl(struct file *file, unsigned int cmd, unsigned long arg) { struct perf_counter *counter = file->private_data; @@ -2000,6 +2007,9 @@ static long perf_ioctl(struct file *file, unsigned int cmd, unsigned long arg) case PERF_COUNTER_IOC_PERIOD: return perf_counter_period(counter, (u64 __user *)arg); + case PERF_COUNTER_IOC_SET_OUTPUT: + return perf_counter_set_output(counter, arg); + default: return -ENOTTY; } @@ -2270,6 +2280,11 @@ static int perf_mmap(struct file *file, struct vm_area_struct *vma) WARN_ON_ONCE(counter->ctx->parent_ctx); mutex_lock(&counter->mmap_mutex); + if (counter->output) { + ret = -EINVAL; + goto unlock; + } + if (atomic_inc_not_zero(&counter->mmap_count)) { if (nr_pages != counter->data->nr_pages) ret = -EINVAL; @@ -2655,6 +2670,7 @@ static int perf_output_begin(struct perf_output_handle *handle, struct perf_counter *counter, unsigned int size, int nmi, int sample) { + struct perf_counter *output_counter; struct perf_mmap_data *data; unsigned int offset, head; int have_lost; @@ -2664,13 +2680,17 @@ static int perf_output_begin(struct perf_output_handle *handle, u64 lost; } lost_event; + rcu_read_lock(); /* * For inherited counters we send all the output towards the parent. */ if (counter->parent) counter = counter->parent; - rcu_read_lock(); + output_counter = rcu_dereference(counter->output); + if (output_counter) + counter = output_counter; + data = rcu_dereference(counter->data); if (!data) goto out; @@ -4218,6 +4238,57 @@ err_size: goto out; } +int perf_counter_set_output(struct perf_counter *counter, int output_fd) +{ + struct perf_counter *output_counter = NULL; + struct file *output_file = NULL; + struct perf_counter *old_output; + int fput_needed = 0; + int ret = -EINVAL; + + if (!output_fd) + goto set; + + output_file = fget_light(output_fd, &fput_needed); + if (!output_file) + return -EBADF; + + if (output_file->f_op != &perf_fops) + goto out; + + output_counter = output_file->private_data; + + /* Don't chain output fds */ + if (output_counter->output) + goto out; + + /* Don't set an output fd when we already have an output channel */ + if (counter->data) + goto out; + + atomic_long_inc(&output_file->f_count); + +set: + mutex_lock(&counter->mmap_mutex); + old_output = counter->output; + rcu_assign_pointer(counter->output, output_counter); + mutex_unlock(&counter->mmap_mutex); + + if (old_output) { + /* + * we need to make sure no existing perf_output_*() + * is still referencing this counter. + */ + synchronize_rcu(); + fput(old_output->filp); + } + + ret = 0; +out: + fput_light(output_file, fput_needed); + return ret; +} + /** * sys_perf_counter_open - open a performance counter, associate it to a task/cpu * @@ -4240,7 +4311,7 @@ SYSCALL_DEFINE5(perf_counter_open, int ret; /* for future expandability... */ - if (flags) + if (flags & ~(PERF_FLAG_FD_NO_GROUP | PERF_FLAG_FD_OUTPUT)) return -EINVAL; ret = perf_copy_attr(attr_uptr, &attr); @@ -4268,7 +4339,7 @@ SYSCALL_DEFINE5(perf_counter_open, * Look up the group leader (we will attach this counter to it): */ group_leader = NULL; - if (group_fd != -1) { + if (group_fd != -1 && !(flags & PERF_FLAG_FD_NO_GROUP)) { ret = -EINVAL; group_file = fget_light(group_fd, &fput_needed); if (!group_file) @@ -4310,6 +4381,12 @@ SYSCALL_DEFINE5(perf_counter_open, if (!counter_file) goto err_free_put_context; + if (flags & PERF_FLAG_FD_OUTPUT) { + ret = perf_counter_set_output(counter, group_fd); + if (ret) + goto err_free_put_context; + } + counter->filp = counter_file; WARN_ON_ONCE(ctx->parent_ctx); mutex_lock(&ctx->mutex); -- cgit v1.2.3 From 31b47cf7609288893a10706c648faa932c7aef90 Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Mon, 24 Aug 2009 20:28:04 +0100 Subject: genirq: Add prototype for handle_nested_irq() The function is supposed to be called from the primary IRQ handler for a demultiplexing chip so make a protype visible for them. Signed-off-by: Mark Brown Cc: Mark Brown LKML-Reference: <1251142084-9852-1-git-send-email-broonie@opensource.wolfsonmicro.com> Signed-off-by: Ingo Molnar --- include/linux/irq.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include') diff --git a/include/linux/irq.h b/include/linux/irq.h index 6956df9961ab..9e9eb76faf81 100644 --- a/include/linux/irq.h +++ b/include/linux/irq.h @@ -282,6 +282,7 @@ extern void handle_edge_irq(unsigned int irq, struct irq_desc *desc); extern void handle_simple_irq(unsigned int irq, struct irq_desc *desc); extern void handle_percpu_irq(unsigned int irq, struct irq_desc *desc); extern void handle_bad_irq(unsigned int irq, struct irq_desc *desc); +extern void handle_nested_irq(unsigned int irq); /* * Monolithic do_IRQ implementation. -- cgit v1.2.3 From b6ed2e03df1e2c6ee41cf0e2e2699f2410671916 Mon Sep 17 00:00:00 2001 From: Steven Whitehouse Date: Tue, 25 Aug 2009 13:44:04 +0100 Subject: GFS2: Add explanation of extended attr on-disk format Some useful info regarding the on-disk representation of GFS2 extended attributes. Signed-off-by: Steven Whitehouse --- include/linux/gfs2_ondisk.h | 22 ++++++++++++++++++++++ 1 file changed, 22 insertions(+) (limited to 'include') diff --git a/include/linux/gfs2_ondisk.h b/include/linux/gfs2_ondisk.h index c56b4bce56d0..b80c88dedbbb 100644 --- a/include/linux/gfs2_ondisk.h +++ b/include/linux/gfs2_ondisk.h @@ -333,6 +333,28 @@ struct gfs2_leaf { /* * Extended attribute header format + * + * This works in a similar way to dirents. There is a fixed size header + * followed by a variable length section made up of the name and the + * associated data. In the case of a "stuffed" entry, the value is + * inline directly after the name, the ea_num_ptrs entry will be + * zero in that case. For non-"stuffed" entries, there will be + * a set of pointers (aligned to 8 byte boundary) to the block(s) + * containing the value. + * + * The blocks containing the values and the blocks containing the + * extended attribute headers themselves all start with the common + * metadata header. Each inode, if it has extended attributes, will + * have either a single block containing the extended attribute headers + * or a single indirect block pointing to blocks containing the + * extended attribure headers. + * + * The maximim size of the data part of an extended attribute is 64k + * so the number of blocks required depends upon block size. Since the + * block size also determines the number of pointers in an indirect + * block, its a fairly complicated calculation to work out the maximum + * number of blocks that an inode may have relating to extended attributes. + * */ #define GFS2_EA_MAX_NAME_LEN 255 -- cgit v1.2.3 From 3a6c2b419b7768703cfb2cabdb894517c5065e33 Mon Sep 17 00:00:00 2001 From: Patrick McHardy Date: Tue, 25 Aug 2009 16:07:40 +0200 Subject: netlink: constify nlmsghdr arguments Consitfy nlmsghdr arguments to a couple of functions as preparation for the next patch, which will constify the netlink message data in all nfnetlink users. Signed-off-by: Patrick McHardy --- include/linux/netlink.h | 15 ++++++++------- include/net/netlink.h | 4 ++-- include/net/rtnetlink.h | 2 +- net/netlink/af_netlink.c | 2 +- net/sched/act_api.c | 2 +- 5 files changed, 13 insertions(+), 12 deletions(-) (limited to 'include') diff --git a/include/linux/netlink.h b/include/linux/netlink.h index 5ba398e90304..0fbecbbe8e9e 100644 --- a/include/linux/netlink.h +++ b/include/linux/netlink.h @@ -217,12 +217,13 @@ int netlink_sendskb(struct sock *sk, struct sk_buff *skb); struct netlink_callback { - struct sk_buff *skb; - struct nlmsghdr *nlh; - int (*dump)(struct sk_buff * skb, struct netlink_callback *cb); - int (*done)(struct netlink_callback *cb); - int family; - long args[6]; + struct sk_buff *skb; + const struct nlmsghdr *nlh; + int (*dump)(struct sk_buff * skb, + struct netlink_callback *cb); + int (*done)(struct netlink_callback *cb); + int family; + long args[6]; }; struct netlink_notify @@ -258,7 +259,7 @@ __nlmsg_put(struct sk_buff *skb, u32 pid, u32 seq, int type, int len, int flags) NLMSG_NEW(skb, pid, seq, type, len, 0) extern int netlink_dump_start(struct sock *ssk, struct sk_buff *skb, - struct nlmsghdr *nlh, + const struct nlmsghdr *nlh, int (*dump)(struct sk_buff *skb, struct netlink_callback*), int (*done)(struct netlink_callback*)); diff --git a/include/net/netlink.h b/include/net/netlink.h index 007bdb07dabb..a63b2192ac1c 100644 --- a/include/net/netlink.h +++ b/include/net/netlink.h @@ -365,7 +365,7 @@ static inline struct nlmsghdr *nlmsg_next(struct nlmsghdr *nlh, int *remaining) * * See nla_parse() */ -static inline int nlmsg_parse(struct nlmsghdr *nlh, int hdrlen, +static inline int nlmsg_parse(const struct nlmsghdr *nlh, int hdrlen, struct nlattr *tb[], int maxtype, const struct nla_policy *policy) { @@ -414,7 +414,7 @@ static inline int nlmsg_validate(struct nlmsghdr *nlh, int hdrlen, int maxtype, * * Returns 1 if a report back to the application is requested. */ -static inline int nlmsg_report(struct nlmsghdr *nlh) +static inline int nlmsg_report(const struct nlmsghdr *nlh) { return !!(nlh->nlmsg_flags & NLM_F_ECHO); } diff --git a/include/net/rtnetlink.h b/include/net/rtnetlink.h index 3c1895e54b7f..85ba560332ed 100644 --- a/include/net/rtnetlink.h +++ b/include/net/rtnetlink.h @@ -14,7 +14,7 @@ extern void rtnl_register(int protocol, int msgtype, extern int rtnl_unregister(int protocol, int msgtype); extern void rtnl_unregister_all(int protocol); -static inline int rtnl_msg_family(struct nlmsghdr *nlh) +static inline int rtnl_msg_family(const struct nlmsghdr *nlh) { if (nlmsg_len(nlh) >= sizeof(struct rtgenmsg)) return ((struct rtgenmsg *) nlmsg_data(nlh))->rtgen_family; diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c index da3163d15ef0..d0ff382c40ca 100644 --- a/net/netlink/af_netlink.c +++ b/net/netlink/af_netlink.c @@ -1705,7 +1705,7 @@ errout: } int netlink_dump_start(struct sock *ssk, struct sk_buff *skb, - struct nlmsghdr *nlh, + const struct nlmsghdr *nlh, int (*dump)(struct sk_buff *skb, struct netlink_callback *), int (*done)(struct netlink_callback *)) diff --git a/net/sched/act_api.c b/net/sched/act_api.c index 9d03cc33b6cc..2dfb3e7a040d 100644 --- a/net/sched/act_api.c +++ b/net/sched/act_api.c @@ -1011,7 +1011,7 @@ replay: } static struct nlattr * -find_dump_kind(struct nlmsghdr *n) +find_dump_kind(const struct nlmsghdr *n) { struct nlattr *tb1, *tb2[TCA_ACT_MAX+1]; struct nlattr *tb[TCA_ACT_MAX_PRIO + 1]; -- cgit v1.2.3 From 3993832464dd4e14a4c926583a11f0fa92c1f0f0 Mon Sep 17 00:00:00 2001 From: Patrick McHardy Date: Tue, 25 Aug 2009 16:07:58 +0200 Subject: netfilter: nfnetlink: constify message attributes and headers Signed-off-by: Patrick McHardy --- include/linux/netfilter/nfnetlink.h | 3 +- include/net/netfilter/nf_nat_core.h | 2 +- net/ipv4/netfilter/nf_nat_core.c | 6 ++-- net/netfilter/nf_conntrack_core.c | 2 +- net/netfilter/nf_conntrack_netlink.c | 54 ++++++++++++++++++++++-------------- net/netfilter/nfnetlink.c | 2 +- net/netfilter/nfnetlink_log.c | 6 ++-- net/netfilter/nfnetlink_queue.c | 9 ++++-- net/netfilter/xt_osf.c | 6 ++-- 9 files changed, 55 insertions(+), 35 deletions(-) (limited to 'include') diff --git a/include/linux/netfilter/nfnetlink.h b/include/linux/netfilter/nfnetlink.h index bff4d5741d98..9f00da287f2c 100644 --- a/include/linux/netfilter/nfnetlink.h +++ b/include/linux/netfilter/nfnetlink.h @@ -58,7 +58,8 @@ struct nfgenmsg { struct nfnl_callback { int (*call)(struct sock *nl, struct sk_buff *skb, - struct nlmsghdr *nlh, struct nlattr *cda[]); + const struct nlmsghdr *nlh, + const struct nlattr * const cda[]); const struct nla_policy *policy; /* netlink attribute policy */ const u_int16_t attr_count; /* number of nlattr's */ }; diff --git a/include/net/netfilter/nf_nat_core.h b/include/net/netfilter/nf_nat_core.h index 58684066388c..33602ab66190 100644 --- a/include/net/netfilter/nf_nat_core.h +++ b/include/net/netfilter/nf_nat_core.h @@ -31,6 +31,6 @@ struct nlattr; extern int (*nfnetlink_parse_nat_setup_hook)(struct nf_conn *ct, enum nf_nat_manip_type manip, - struct nlattr *attr); + const struct nlattr *attr); #endif /* _NF_NAT_CORE_H */ diff --git a/net/ipv4/netfilter/nf_nat_core.c b/net/ipv4/netfilter/nf_nat_core.c index b6ddd5633045..68afc6ecd343 100644 --- a/net/ipv4/netfilter/nf_nat_core.c +++ b/net/ipv4/netfilter/nf_nat_core.c @@ -620,7 +620,7 @@ static const struct nla_policy nat_nla_policy[CTA_NAT_MAX+1] = { }; static int -nfnetlink_parse_nat(struct nlattr *nat, +nfnetlink_parse_nat(const struct nlattr *nat, const struct nf_conn *ct, struct nf_nat_range *range) { struct nlattr *tb[CTA_NAT_MAX+1]; @@ -656,7 +656,7 @@ nfnetlink_parse_nat(struct nlattr *nat, static int nfnetlink_parse_nat_setup(struct nf_conn *ct, enum nf_nat_manip_type manip, - struct nlattr *attr) + const struct nlattr *attr) { struct nf_nat_range range; @@ -671,7 +671,7 @@ nfnetlink_parse_nat_setup(struct nf_conn *ct, static int nfnetlink_parse_nat_setup(struct nf_conn *ct, enum nf_nat_manip_type manip, - struct nlattr *attr) + const struct nlattr *attr) { return -EOPNOTSUPP; } diff --git a/net/netfilter/nf_conntrack_core.c b/net/netfilter/nf_conntrack_core.c index b5869b9574b0..565c3a86423f 100644 --- a/net/netfilter/nf_conntrack_core.c +++ b/net/netfilter/nf_conntrack_core.c @@ -47,7 +47,7 @@ int (*nfnetlink_parse_nat_setup_hook)(struct nf_conn *ct, enum nf_nat_manip_type manip, - struct nlattr *attr) __read_mostly; + const struct nlattr *attr) __read_mostly; EXPORT_SYMBOL_GPL(nfnetlink_parse_nat_setup_hook); DEFINE_SPINLOCK(nf_conntrack_lock); diff --git a/net/netfilter/nf_conntrack_netlink.c b/net/netfilter/nf_conntrack_netlink.c index 49479d194570..59d8064eb522 100644 --- a/net/netfilter/nf_conntrack_netlink.c +++ b/net/netfilter/nf_conntrack_netlink.c @@ -704,7 +704,8 @@ ctnetlink_parse_tuple_proto(struct nlattr *attr, } static int -ctnetlink_parse_tuple(struct nlattr *cda[], struct nf_conntrack_tuple *tuple, +ctnetlink_parse_tuple(const struct nlattr * const cda[], + struct nf_conntrack_tuple *tuple, enum ctattr_tuple type, u_int8_t l3num) { struct nlattr *tb[CTA_TUPLE_MAX+1]; @@ -740,7 +741,7 @@ ctnetlink_parse_tuple(struct nlattr *cda[], struct nf_conntrack_tuple *tuple, } static inline int -ctnetlink_parse_help(struct nlattr *attr, char **helper_name) +ctnetlink_parse_help(const struct nlattr *attr, char **helper_name) { struct nlattr *tb[CTA_HELP_MAX+1]; @@ -764,7 +765,8 @@ static const struct nla_policy ct_nla_policy[CTA_MAX+1] = { static int ctnetlink_del_conntrack(struct sock *ctnl, struct sk_buff *skb, - struct nlmsghdr *nlh, struct nlattr *cda[]) + const struct nlmsghdr *nlh, + const struct nlattr * const cda[]) { struct nf_conntrack_tuple_hash *h; struct nf_conntrack_tuple tuple; @@ -823,7 +825,8 @@ ctnetlink_del_conntrack(struct sock *ctnl, struct sk_buff *skb, static int ctnetlink_get_conntrack(struct sock *ctnl, struct sk_buff *skb, - struct nlmsghdr *nlh, struct nlattr *cda[]) + const struct nlmsghdr *nlh, + const struct nlattr * const cda[]) { struct nf_conntrack_tuple_hash *h; struct nf_conntrack_tuple tuple; @@ -884,7 +887,7 @@ out: static int ctnetlink_parse_nat_setup(struct nf_conn *ct, enum nf_nat_manip_type manip, - struct nlattr *attr) + const struct nlattr *attr) { typeof(nfnetlink_parse_nat_setup_hook) parse_nat_setup; @@ -914,7 +917,7 @@ ctnetlink_parse_nat_setup(struct nf_conn *ct, #endif static int -ctnetlink_change_status(struct nf_conn *ct, struct nlattr *cda[]) +ctnetlink_change_status(struct nf_conn *ct, const struct nlattr * const cda[]) { unsigned long d; unsigned int status = ntohl(nla_get_be32(cda[CTA_STATUS])); @@ -940,7 +943,7 @@ ctnetlink_change_status(struct nf_conn *ct, struct nlattr *cda[]) } static int -ctnetlink_change_nat(struct nf_conn *ct, struct nlattr *cda[]) +ctnetlink_change_nat(struct nf_conn *ct, const struct nlattr * const cda[]) { #ifdef CONFIG_NF_NAT_NEEDED int ret; @@ -966,7 +969,7 @@ ctnetlink_change_nat(struct nf_conn *ct, struct nlattr *cda[]) } static inline int -ctnetlink_change_helper(struct nf_conn *ct, struct nlattr *cda[]) +ctnetlink_change_helper(struct nf_conn *ct, const struct nlattr * const cda[]) { struct nf_conntrack_helper *helper; struct nf_conn_help *help = nfct_help(ct); @@ -1028,7 +1031,7 @@ ctnetlink_change_helper(struct nf_conn *ct, struct nlattr *cda[]) } static inline int -ctnetlink_change_timeout(struct nf_conn *ct, struct nlattr *cda[]) +ctnetlink_change_timeout(struct nf_conn *ct, const struct nlattr * const cda[]) { u_int32_t timeout = ntohl(nla_get_be32(cda[CTA_TIMEOUT])); @@ -1042,9 +1045,10 @@ ctnetlink_change_timeout(struct nf_conn *ct, struct nlattr *cda[]) } static inline int -ctnetlink_change_protoinfo(struct nf_conn *ct, struct nlattr *cda[]) +ctnetlink_change_protoinfo(struct nf_conn *ct, const struct nlattr * const cda[]) { - struct nlattr *tb[CTA_PROTOINFO_MAX+1], *attr = cda[CTA_PROTOINFO]; + const struct nlattr *attr = cda[CTA_PROTOINFO]; + struct nlattr *tb[CTA_PROTOINFO_MAX+1]; struct nf_conntrack_l4proto *l4proto; int err = 0; @@ -1061,7 +1065,7 @@ ctnetlink_change_protoinfo(struct nf_conn *ct, struct nlattr *cda[]) #ifdef CONFIG_NF_NAT_NEEDED static inline int -change_nat_seq_adj(struct nf_nat_seq *natseq, struct nlattr *attr) +change_nat_seq_adj(struct nf_nat_seq *natseq, const struct nlattr * const attr) { struct nlattr *cda[CTA_NAT_SEQ_MAX+1]; @@ -1089,7 +1093,8 @@ change_nat_seq_adj(struct nf_nat_seq *natseq, struct nlattr *attr) } static int -ctnetlink_change_nat_seq_adj(struct nf_conn *ct, struct nlattr *cda[]) +ctnetlink_change_nat_seq_adj(struct nf_conn *ct, + const struct nlattr * const cda[]) { int ret = 0; struct nf_conn_nat *nat = nfct_nat(ct); @@ -1120,7 +1125,8 @@ ctnetlink_change_nat_seq_adj(struct nf_conn *ct, struct nlattr *cda[]) #endif static int -ctnetlink_change_conntrack(struct nf_conn *ct, struct nlattr *cda[]) +ctnetlink_change_conntrack(struct nf_conn *ct, + const struct nlattr * const cda[]) { int err; @@ -1169,7 +1175,7 @@ ctnetlink_change_conntrack(struct nf_conn *ct, struct nlattr *cda[]) } static struct nf_conn * -ctnetlink_create_conntrack(struct nlattr *cda[], +ctnetlink_create_conntrack(const struct nlattr * const cda[], struct nf_conntrack_tuple *otuple, struct nf_conntrack_tuple *rtuple, u8 u3) @@ -1304,7 +1310,8 @@ err1: static int ctnetlink_new_conntrack(struct sock *ctnl, struct sk_buff *skb, - struct nlmsghdr *nlh, struct nlattr *cda[]) + const struct nlmsghdr *nlh, + const struct nlattr * const cda[]) { struct nf_conntrack_tuple otuple, rtuple; struct nf_conntrack_tuple_hash *h = NULL; @@ -1629,7 +1636,8 @@ static const struct nla_policy exp_nla_policy[CTA_EXPECT_MAX+1] = { static int ctnetlink_get_expect(struct sock *ctnl, struct sk_buff *skb, - struct nlmsghdr *nlh, struct nlattr *cda[]) + const struct nlmsghdr *nlh, + const struct nlattr * const cda[]) { struct nf_conntrack_tuple tuple; struct nf_conntrack_expect *exp; @@ -1689,7 +1697,8 @@ out: static int ctnetlink_del_expect(struct sock *ctnl, struct sk_buff *skb, - struct nlmsghdr *nlh, struct nlattr *cda[]) + const struct nlmsghdr *nlh, + const struct nlattr * const cda[]) { struct nf_conntrack_expect *exp; struct nf_conntrack_tuple tuple; @@ -1767,13 +1776,15 @@ ctnetlink_del_expect(struct sock *ctnl, struct sk_buff *skb, return 0; } static int -ctnetlink_change_expect(struct nf_conntrack_expect *x, struct nlattr *cda[]) +ctnetlink_change_expect(struct nf_conntrack_expect *x, + const struct nlattr * const cda[]) { return -EOPNOTSUPP; } static int -ctnetlink_create_expect(struct nlattr *cda[], u_int8_t u3, u32 pid, int report) +ctnetlink_create_expect(const struct nlattr * const cda[], u_int8_t u3, + u32 pid, int report) { struct nf_conntrack_tuple tuple, mask, master_tuple; struct nf_conntrack_tuple_hash *h = NULL; @@ -1831,7 +1842,8 @@ out: static int ctnetlink_new_expect(struct sock *ctnl, struct sk_buff *skb, - struct nlmsghdr *nlh, struct nlattr *cda[]) + const struct nlmsghdr *nlh, + const struct nlattr * const cda[]) { struct nf_conntrack_tuple tuple; struct nf_conntrack_expect *exp; diff --git a/net/netfilter/nfnetlink.c b/net/netfilter/nfnetlink.c index 92761a988375..eedc0c1ac7a4 100644 --- a/net/netfilter/nfnetlink.c +++ b/net/netfilter/nfnetlink.c @@ -170,7 +170,7 @@ replay: if (err < 0) return err; - err = nc->call(nfnl, skb, nlh, cda); + err = nc->call(nfnl, skb, nlh, (const struct nlattr **)cda); if (err == -EAGAIN) goto replay; return err; diff --git a/net/netfilter/nfnetlink_log.c b/net/netfilter/nfnetlink_log.c index 66a6dd5c519a..f900dc3194af 100644 --- a/net/netfilter/nfnetlink_log.c +++ b/net/netfilter/nfnetlink_log.c @@ -694,7 +694,8 @@ static struct notifier_block nfulnl_rtnl_notifier = { static int nfulnl_recv_unsupp(struct sock *ctnl, struct sk_buff *skb, - struct nlmsghdr *nlh, struct nlattr *nfqa[]) + const struct nlmsghdr *nlh, + const struct nlattr * const nfqa[]) { return -ENOTSUPP; } @@ -716,7 +717,8 @@ static const struct nla_policy nfula_cfg_policy[NFULA_CFG_MAX+1] = { static int nfulnl_recv_config(struct sock *ctnl, struct sk_buff *skb, - struct nlmsghdr *nlh, struct nlattr *nfula[]) + const struct nlmsghdr *nlh, + const struct nlattr * const nfula[]) { struct nfgenmsg *nfmsg = NLMSG_DATA(nlh); u_int16_t group_num = ntohs(nfmsg->res_id); diff --git a/net/netfilter/nfnetlink_queue.c b/net/netfilter/nfnetlink_queue.c index 71daa0934b6c..7a9dec9fb822 100644 --- a/net/netfilter/nfnetlink_queue.c +++ b/net/netfilter/nfnetlink_queue.c @@ -608,7 +608,8 @@ static const struct nla_policy nfqa_verdict_policy[NFQA_MAX+1] = { static int nfqnl_recv_verdict(struct sock *ctnl, struct sk_buff *skb, - struct nlmsghdr *nlh, struct nlattr *nfqa[]) + const struct nlmsghdr *nlh, + const struct nlattr * const nfqa[]) { struct nfgenmsg *nfmsg = NLMSG_DATA(nlh); u_int16_t queue_num = ntohs(nfmsg->res_id); @@ -670,7 +671,8 @@ err_out_unlock: static int nfqnl_recv_unsupp(struct sock *ctnl, struct sk_buff *skb, - struct nlmsghdr *nlh, struct nlattr *nfqa[]) + const struct nlmsghdr *nlh, + const struct nlattr * const nfqa[]) { return -ENOTSUPP; } @@ -687,7 +689,8 @@ static const struct nf_queue_handler nfqh = { static int nfqnl_recv_config(struct sock *ctnl, struct sk_buff *skb, - struct nlmsghdr *nlh, struct nlattr *nfqa[]) + const struct nlmsghdr *nlh, + const struct nlattr * const nfqa[]) { struct nfgenmsg *nfmsg = NLMSG_DATA(nlh); u_int16_t queue_num = ntohs(nfmsg->res_id); diff --git a/net/netfilter/xt_osf.c b/net/netfilter/xt_osf.c index 0f482e2440b4..63e190504656 100644 --- a/net/netfilter/xt_osf.c +++ b/net/netfilter/xt_osf.c @@ -70,7 +70,8 @@ static void xt_osf_finger_free_rcu(struct rcu_head *rcu_head) } static int xt_osf_add_callback(struct sock *ctnl, struct sk_buff *skb, - struct nlmsghdr *nlh, struct nlattr *osf_attrs[]) + const struct nlmsghdr *nlh, + const struct nlattr * const osf_attrs[]) { struct xt_osf_user_finger *f; struct xt_osf_finger *kf = NULL, *sf; @@ -112,7 +113,8 @@ static int xt_osf_add_callback(struct sock *ctnl, struct sk_buff *skb, } static int xt_osf_remove_callback(struct sock *ctnl, struct sk_buff *skb, - struct nlmsghdr *nlh, struct nlattr *osf_attrs[]) + const struct nlmsghdr *nlh, + const struct nlattr * const osf_attrs[]) { struct xt_osf_user_finger *f; struct xt_osf_finger *sf; -- cgit v1.2.3 From 3d27d8cb34fc156beb86de2338ca4029873a5cc6 Mon Sep 17 00:00:00 2001 From: Josh Stone Date: Mon, 24 Aug 2009 14:43:12 -0700 Subject: tracing: Make syscall tracepoints conditional The syscall enter/exit tracepoints are only supported on archs that HAVE_SYSCALL_TRACEPOINTS, so the declarations should be #ifdef'ed. Also, the definition of syscall_regfunc and syscall_unregfunc should depend on this same config, rather than the ftrace-specific one. Signed-off-by: Josh Stone Cc: Jason Baron Cc: Frederic Weisbecker Cc: Ingo Molnar Cc: Li Zefan Cc: Steven Rostedt Cc: Peter Zijlstra Cc: Mathieu Desnoyers Cc: Jiaying Zhang Cc: Martin Bligh Cc: Lai Jiangshan LKML-Reference: <1251150194-1713-3-git-send-email-jistone@redhat.com> Signed-off-by: Frederic Weisbecker --- include/trace/syscall.h | 4 ++++ kernel/tracepoint.c | 2 +- 2 files changed, 5 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/trace/syscall.h b/include/trace/syscall.h index 9661dd406b93..5dcb7e3a544c 100644 --- a/include/trace/syscall.h +++ b/include/trace/syscall.h @@ -8,6 +8,8 @@ #include +#ifdef CONFIG_HAVE_SYSCALL_TRACEPOINTS + extern void syscall_regfunc(void); extern void syscall_unregfunc(void); @@ -25,6 +27,8 @@ DECLARE_TRACE_WITH_CALLBACK(syscall_exit, syscall_unregfunc ); +#endif + /* * A syscall entry in the ftrace syscalls array. * diff --git a/kernel/tracepoint.c b/kernel/tracepoint.c index be86b9a01a09..9e0a36f0e2a9 100644 --- a/kernel/tracepoint.c +++ b/kernel/tracepoint.c @@ -576,7 +576,7 @@ __initcall(init_tracepoints); #endif /* CONFIG_MODULES */ -#ifdef CONFIG_FTRACE_SYSCALLS +#ifdef CONFIG_HAVE_SYSCALL_TRACEPOINTS static DEFINE_MUTEX(regfunc_mutex); static int sys_tracepoint_refcount; -- cgit v1.2.3 From 97419875865859fd2403e66266c02ce028e2f5ab Mon Sep 17 00:00:00 2001 From: Josh Stone Date: Mon, 24 Aug 2009 14:43:13 -0700 Subject: tracing: Move tracepoint callbacks from declaration to definition It's not strictly correct for the tracepoint reg/unreg callbacks to occur when a client is hooking up, because the actual tracepoint may not be present yet. This happens to be fine for syscall, since that's in the core kernel, but it would cause problems for tracepoints defined in a module that hasn't been loaded yet. It also means the reg/unreg has to be EXPORTed for any modules to use the tracepoint (as in SystemTap). This patch removes DECLARE_TRACE_WITH_CALLBACK, and instead introduces DEFINE_TRACE_FN which stores the callbacks in struct tracepoint. The callbacks are used now when the active state of the tracepoint changes in set_tracepoint & disable_tracepoint. This also introduces TRACE_EVENT_FN, so ftrace events can also provide registration callbacks if needed. Signed-off-by: Josh Stone Cc: Jason Baron Cc: Frederic Weisbecker Cc: Ingo Molnar Cc: Li Zefan Cc: Steven Rostedt Cc: Peter Zijlstra Cc: Mathieu Desnoyers Cc: Jiaying Zhang Cc: Martin Bligh Cc: Lai Jiangshan Cc: Paul Mundt Cc: Martin Schwidefsky Cc: Heiko Carstens LKML-Reference: <1251150194-1713-4-git-send-email-jistone@redhat.com> Signed-off-by: Frederic Weisbecker --- arch/s390/kernel/ptrace.c | 4 ++-- arch/x86/kernel/ptrace.c | 4 ++-- include/linux/tracepoint.h | 46 +++++++++++++++++--------------------------- include/trace/define_trace.h | 5 +++++ include/trace/ftrace.h | 9 +++++++++ include/trace/syscall.h | 12 ++++-------- kernel/tracepoint.c | 14 +++++++++----- 7 files changed, 49 insertions(+), 45 deletions(-) (limited to 'include') diff --git a/arch/s390/kernel/ptrace.c b/arch/s390/kernel/ptrace.c index 9d3dcfa79ea2..c05b44b80c23 100644 --- a/arch/s390/kernel/ptrace.c +++ b/arch/s390/kernel/ptrace.c @@ -51,8 +51,8 @@ #include "compat_ptrace.h" #endif -DEFINE_TRACE(syscall_enter); -DEFINE_TRACE(syscall_exit); +DEFINE_TRACE_FN(syscall_enter, syscall_regfunc, syscall_unregfunc); +DEFINE_TRACE_FN(syscall_exit, syscall_regfunc, syscall_unregfunc); enum s390_regset { REGSET_GENERAL, diff --git a/arch/x86/kernel/ptrace.c b/arch/x86/kernel/ptrace.c index a909afef44f4..31e9b97ec4d6 100644 --- a/arch/x86/kernel/ptrace.c +++ b/arch/x86/kernel/ptrace.c @@ -37,8 +37,8 @@ #include -DEFINE_TRACE(syscall_enter); -DEFINE_TRACE(syscall_exit); +DEFINE_TRACE_FN(syscall_enter, syscall_regfunc, syscall_unregfunc); +DEFINE_TRACE_FN(syscall_exit, syscall_regfunc, syscall_unregfunc); #include "tls.h" diff --git a/include/linux/tracepoint.h b/include/linux/tracepoint.h index 5984ed04c03b..846a4ae501eb 100644 --- a/include/linux/tracepoint.h +++ b/include/linux/tracepoint.h @@ -23,6 +23,8 @@ struct tracepoint; struct tracepoint { const char *name; /* Tracepoint name */ int state; /* State. */ + void (*regfunc)(void); + void (*unregfunc)(void); void **funcs; } __attribute__((aligned(32))); /* * Aligned on 32 bytes because it is @@ -60,10 +62,8 @@ struct tracepoint { * Make sure the alignment of the structure in the __tracepoints section will * not add unwanted padding between the beginning of the section and the * structure. Force alignment to the same alignment as the section start. - * An optional set of (un)registration functions can be passed to perform any - * additional (un)registration work. */ -#define DECLARE_TRACE_WITH_CALLBACK(name, proto, args, reg, unreg) \ +#define DECLARE_TRACE(name, proto, args) \ extern struct tracepoint __tracepoint_##name; \ static inline void trace_##name(proto) \ { \ @@ -73,36 +73,23 @@ struct tracepoint { } \ static inline int register_trace_##name(void (*probe)(proto)) \ { \ - int ret; \ - void (*func)(void) = reg; \ - \ - ret = tracepoint_probe_register(#name, (void *)probe); \ - if (func && !ret) \ - func(); \ - return ret; \ + return tracepoint_probe_register(#name, (void *)probe); \ } \ static inline int unregister_trace_##name(void (*probe)(proto)) \ { \ - int ret; \ - void (*func)(void) = unreg; \ - \ - ret = tracepoint_probe_unregister(#name, (void *)probe);\ - if (func && !ret) \ - func(); \ - return ret; \ + return tracepoint_probe_unregister(#name, (void *)probe);\ } -#define DECLARE_TRACE(name, proto, args) \ - DECLARE_TRACE_WITH_CALLBACK(name, TP_PROTO(proto), TP_ARGS(args),\ - NULL, NULL); - -#define DEFINE_TRACE(name) \ +#define DEFINE_TRACE_FN(name, reg, unreg) \ static const char __tpstrtab_##name[] \ __attribute__((section("__tracepoints_strings"))) = #name; \ struct tracepoint __tracepoint_##name \ __attribute__((section("__tracepoints"), aligned(32))) = \ - { __tpstrtab_##name, 0, NULL } + { __tpstrtab_##name, 0, reg, unreg, NULL } + +#define DEFINE_TRACE(name) \ + DEFINE_TRACE_FN(name, NULL, NULL); #define EXPORT_TRACEPOINT_SYMBOL_GPL(name) \ EXPORT_SYMBOL_GPL(__tracepoint_##name) @@ -113,7 +100,7 @@ extern void tracepoint_update_probe_range(struct tracepoint *begin, struct tracepoint *end); #else /* !CONFIG_TRACEPOINTS */ -#define DECLARE_TRACE_WITH_CALLBACK(name, proto, args, reg, unreg) \ +#define DECLARE_TRACE(name, proto, args) \ static inline void _do_trace_##name(struct tracepoint *tp, proto) \ { } \ static inline void trace_##name(proto) \ @@ -127,10 +114,7 @@ extern void tracepoint_update_probe_range(struct tracepoint *begin, return -ENOSYS; \ } -#define DECLARE_TRACE(name, proto, args) \ - DECLARE_TRACE_WITH_CALLBACK(name, TP_PROTO(proto), TP_ARGS(args),\ - NULL, NULL); - +#define DEFINE_TRACE_FN(name, reg, unreg) #define DEFINE_TRACE(name) #define EXPORT_TRACEPOINT_SYMBOL_GPL(name) #define EXPORT_TRACEPOINT_SYMBOL(name) @@ -282,10 +266,16 @@ static inline void tracepoint_synchronize_unregister(void) * can also by used by generic instrumentation like SystemTap), and * it is also used to expose a structured trace record in * /sys/kernel/debug/tracing/events/. + * + * A set of (un)registration functions can be passed to the variant + * TRACE_EVENT_FN to perform any (un)registration work. */ #define TRACE_EVENT(name, proto, args, struct, assign, print) \ DECLARE_TRACE(name, PARAMS(proto), PARAMS(args)) +#define TRACE_EVENT_FN(name, proto, args, struct, \ + assign, print, reg, unreg) \ + DECLARE_TRACE(name, PARAMS(proto), PARAMS(args)) #endif #endif diff --git a/include/trace/define_trace.h b/include/trace/define_trace.h index f7a7ae1e8f90..2a969850736d 100644 --- a/include/trace/define_trace.h +++ b/include/trace/define_trace.h @@ -26,6 +26,11 @@ #define TRACE_EVENT(name, proto, args, tstruct, assign, print) \ DEFINE_TRACE(name) +#undef TRACE_EVENT_FN +#define TRACE_EVENT_FN(name, proto, args, tstruct, \ + assign, print, reg, unreg) \ + DEFINE_TRACE_FN(name, reg, unreg) + #undef DECLARE_TRACE #define DECLARE_TRACE(name, proto, args) \ DEFINE_TRACE(name) diff --git a/include/trace/ftrace.h b/include/trace/ftrace.h index 127400255e4c..3a0b44bdabf7 100644 --- a/include/trace/ftrace.h +++ b/include/trace/ftrace.h @@ -42,6 +42,15 @@ }; \ static struct ftrace_event_call event_##name +/* Callbacks are meaningless to ftrace. */ +#undef TRACE_EVENT_FN +#define TRACE_EVENT_FN(name, proto, args, tstruct, \ + assign, print, reg, unreg) \ + TRACE_EVENT(name, TP_PROTO(proto), TP_ARGS(args), \ + TP_STRUCT__entry(tstruct), \ + TP_fast_assign(assign), \ + TP_printk(print)) + #include TRACE_INCLUDE(TRACE_INCLUDE_FILE) diff --git a/include/trace/syscall.h b/include/trace/syscall.h index 5dcb7e3a544c..4e1943001854 100644 --- a/include/trace/syscall.h +++ b/include/trace/syscall.h @@ -13,18 +13,14 @@ extern void syscall_regfunc(void); extern void syscall_unregfunc(void); -DECLARE_TRACE_WITH_CALLBACK(syscall_enter, +DECLARE_TRACE(syscall_enter, TP_PROTO(struct pt_regs *regs, long id), - TP_ARGS(regs, id), - syscall_regfunc, - syscall_unregfunc + TP_ARGS(regs, id) ); -DECLARE_TRACE_WITH_CALLBACK(syscall_exit, +DECLARE_TRACE(syscall_exit, TP_PROTO(struct pt_regs *regs, long ret), - TP_ARGS(regs, ret), - syscall_regfunc, - syscall_unregfunc + TP_ARGS(regs, ret) ); #endif diff --git a/kernel/tracepoint.c b/kernel/tracepoint.c index 9e0a36f0e2a9..1a6a453b7efb 100644 --- a/kernel/tracepoint.c +++ b/kernel/tracepoint.c @@ -243,6 +243,11 @@ static void set_tracepoint(struct tracepoint_entry **entry, { WARN_ON(strcmp((*entry)->name, elem->name) != 0); + if (elem->regfunc && !elem->state && active) + elem->regfunc(); + else if (elem->unregfunc && elem->state && !active) + elem->unregfunc(); + /* * rcu_assign_pointer has a smp_wmb() which makes sure that the new * probe callbacks array is consistent before setting a pointer to it. @@ -262,6 +267,9 @@ static void set_tracepoint(struct tracepoint_entry **entry, */ static void disable_tracepoint(struct tracepoint *elem) { + if (elem->unregfunc && elem->state) + elem->unregfunc(); + elem->state = 0; rcu_assign_pointer(elem->funcs, NULL); } @@ -578,7 +586,7 @@ __initcall(init_tracepoints); #ifdef CONFIG_HAVE_SYSCALL_TRACEPOINTS -static DEFINE_MUTEX(regfunc_mutex); +/* NB: reg/unreg are called while guarded with the tracepoints_mutex */ static int sys_tracepoint_refcount; void syscall_regfunc(void) @@ -586,7 +594,6 @@ void syscall_regfunc(void) unsigned long flags; struct task_struct *g, *t; - mutex_lock(®func_mutex); if (!sys_tracepoint_refcount) { read_lock_irqsave(&tasklist_lock, flags); do_each_thread(g, t) { @@ -595,7 +602,6 @@ void syscall_regfunc(void) read_unlock_irqrestore(&tasklist_lock, flags); } sys_tracepoint_refcount++; - mutex_unlock(®func_mutex); } void syscall_unregfunc(void) @@ -603,7 +609,6 @@ void syscall_unregfunc(void) unsigned long flags; struct task_struct *g, *t; - mutex_lock(®func_mutex); sys_tracepoint_refcount--; if (!sys_tracepoint_refcount) { read_lock_irqsave(&tasklist_lock, flags); @@ -612,6 +617,5 @@ void syscall_unregfunc(void) } while_each_thread(g, t); read_unlock_irqrestore(&tasklist_lock, flags); } - mutex_unlock(®func_mutex); } #endif -- cgit v1.2.3 From 1c569f0264ea629c10bbab471dd0626ce4d3f19f Mon Sep 17 00:00:00 2001 From: Josh Stone Date: Mon, 24 Aug 2009 14:43:14 -0700 Subject: tracing: Create generic syscall TRACE_EVENTs This converts the syscall_enter/exit tracepoints into TRACE_EVENTs, so you can have generic ftrace events that capture all system calls with arguments and return values. These generic events are also renamed to sys_enter/exit, so they're more closely aligned to the specific sys_enter_foo events. Signed-off-by: Josh Stone Cc: Jason Baron Cc: Frederic Weisbecker Cc: Ingo Molnar Cc: Li Zefan Cc: Steven Rostedt Cc: Peter Zijlstra Cc: Mathieu Desnoyers Cc: Jiaying Zhang Cc: Martin Bligh Cc: Lai Jiangshan Cc: Paul Mundt Cc: Martin Schwidefsky Cc: Heiko Carstens LKML-Reference: <1251150194-1713-5-git-send-email-jistone@redhat.com> Signed-off-by: Frederic Weisbecker --- arch/s390/kernel/ptrace.c | 8 ++--- arch/x86/kernel/ptrace.c | 12 +++---- include/trace/events/syscalls.h | 70 +++++++++++++++++++++++++++++++++++++++++ include/trace/syscall.h | 17 ---------- kernel/trace/trace_syscalls.c | 17 +++++----- 5 files changed, 88 insertions(+), 36 deletions(-) create mode 100644 include/trace/events/syscalls.h (limited to 'include') diff --git a/arch/s390/kernel/ptrace.c b/arch/s390/kernel/ptrace.c index c05b44b80c23..f3ddd7ac06c5 100644 --- a/arch/s390/kernel/ptrace.c +++ b/arch/s390/kernel/ptrace.c @@ -51,8 +51,8 @@ #include "compat_ptrace.h" #endif -DEFINE_TRACE_FN(syscall_enter, syscall_regfunc, syscall_unregfunc); -DEFINE_TRACE_FN(syscall_exit, syscall_regfunc, syscall_unregfunc); +#define CREATE_TRACE_POINTS +#include enum s390_regset { REGSET_GENERAL, @@ -665,7 +665,7 @@ asmlinkage long do_syscall_trace_enter(struct pt_regs *regs) } if (unlikely(test_thread_flag(TIF_SYSCALL_TRACEPOINT))) - trace_syscall_enter(regs, regs->gprs[2]); + trace_sys_enter(regs, regs->gprs[2]); if (unlikely(current->audit_context)) audit_syscall_entry(is_compat_task() ? @@ -683,7 +683,7 @@ asmlinkage void do_syscall_trace_exit(struct pt_regs *regs) regs->gprs[2]); if (unlikely(test_thread_flag(TIF_SYSCALL_TRACEPOINT))) - trace_syscall_exit(regs, regs->gprs[2]); + trace_sys_exit(regs, regs->gprs[2]); if (test_thread_flag(TIF_SYSCALL_TRACE)) tracehook_report_syscall_exit(regs, 0); diff --git a/arch/x86/kernel/ptrace.c b/arch/x86/kernel/ptrace.c index 31e9b97ec4d6..8d7d5c9c1be3 100644 --- a/arch/x86/kernel/ptrace.c +++ b/arch/x86/kernel/ptrace.c @@ -35,13 +35,11 @@ #include #include -#include - -DEFINE_TRACE_FN(syscall_enter, syscall_regfunc, syscall_unregfunc); -DEFINE_TRACE_FN(syscall_exit, syscall_regfunc, syscall_unregfunc); - #include "tls.h" +#define CREATE_TRACE_POINTS +#include + enum x86_regset { REGSET_GENERAL, REGSET_FP, @@ -1501,7 +1499,7 @@ asmregparm long syscall_trace_enter(struct pt_regs *regs) ret = -1L; if (unlikely(test_thread_flag(TIF_SYSCALL_TRACEPOINT))) - trace_syscall_enter(regs, regs->orig_ax); + trace_sys_enter(regs, regs->orig_ax); if (unlikely(current->audit_context)) { if (IS_IA32) @@ -1527,7 +1525,7 @@ asmregparm void syscall_trace_leave(struct pt_regs *regs) audit_syscall_exit(AUDITSC_RESULT(regs->ax), regs->ax); if (unlikely(test_thread_flag(TIF_SYSCALL_TRACEPOINT))) - trace_syscall_exit(regs, regs->ax); + trace_sys_exit(regs, regs->ax); if (test_thread_flag(TIF_SYSCALL_TRACE)) tracehook_report_syscall_exit(regs, 0); diff --git a/include/trace/events/syscalls.h b/include/trace/events/syscalls.h new file mode 100644 index 000000000000..397dff2dbd5a --- /dev/null +++ b/include/trace/events/syscalls.h @@ -0,0 +1,70 @@ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM syscalls + +#if !defined(_TRACE_EVENTS_SYSCALLS_H) || defined(TRACE_HEADER_MULTI_READ) +#define _TRACE_EVENTS_SYSCALLS_H + +#include + +#include +#include + + +#ifdef CONFIG_HAVE_SYSCALL_TRACEPOINTS + +extern void syscall_regfunc(void); +extern void syscall_unregfunc(void); + +TRACE_EVENT_FN(sys_enter, + + TP_PROTO(struct pt_regs *regs, long id), + + TP_ARGS(regs, id), + + TP_STRUCT__entry( + __field( long, id ) + __array( unsigned long, args, 6 ) + ), + + TP_fast_assign( + __entry->id = id; + syscall_get_arguments(current, regs, 0, 6, __entry->args); + ), + + TP_printk("NR %ld (%lx, %lx, %lx, %lx, %lx, %lx)", + __entry->id, + __entry->args[0], __entry->args[1], __entry->args[2], + __entry->args[3], __entry->args[4], __entry->args[5]), + + syscall_regfunc, syscall_unregfunc +); + +TRACE_EVENT_FN(sys_exit, + + TP_PROTO(struct pt_regs *regs, long ret), + + TP_ARGS(regs, ret), + + TP_STRUCT__entry( + __field( long, id ) + __field( long, ret ) + ), + + TP_fast_assign( + __entry->id = syscall_get_nr(current, regs); + __entry->ret = ret; + ), + + TP_printk("NR %ld = %ld", + __entry->id, __entry->ret), + + syscall_regfunc, syscall_unregfunc +); + +#endif /* CONFIG_HAVE_SYSCALL_TRACEPOINTS */ + +#endif /* _TRACE_EVENTS_SYSCALLS_H */ + +/* This part must be outside protection */ +#include + diff --git a/include/trace/syscall.h b/include/trace/syscall.h index 4e1943001854..5dc283ba5ae0 100644 --- a/include/trace/syscall.h +++ b/include/trace/syscall.h @@ -8,23 +8,6 @@ #include -#ifdef CONFIG_HAVE_SYSCALL_TRACEPOINTS - -extern void syscall_regfunc(void); -extern void syscall_unregfunc(void); - -DECLARE_TRACE(syscall_enter, - TP_PROTO(struct pt_regs *regs, long id), - TP_ARGS(regs, id) -); - -DECLARE_TRACE(syscall_exit, - TP_PROTO(struct pt_regs *regs, long ret), - TP_ARGS(regs, ret) -); - -#endif - /* * A syscall entry in the ftrace syscalls array. * diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c index 46c1b977a2cb..2698fe401ebd 100644 --- a/kernel/trace/trace_syscalls.c +++ b/kernel/trace/trace_syscalls.c @@ -1,4 +1,5 @@ #include +#include #include #include #include @@ -286,7 +287,7 @@ int reg_event_syscall_enter(void *ptr) return -ENOSYS; mutex_lock(&syscall_trace_lock); if (!sys_refcount_enter) - ret = register_trace_syscall_enter(ftrace_syscall_enter); + ret = register_trace_sys_enter(ftrace_syscall_enter); if (ret) { pr_info("event trace: Could not activate" "syscall entry trace point"); @@ -311,7 +312,7 @@ void unreg_event_syscall_enter(void *ptr) sys_refcount_enter--; clear_bit(num, enabled_enter_syscalls); if (!sys_refcount_enter) - unregister_trace_syscall_enter(ftrace_syscall_enter); + unregister_trace_sys_enter(ftrace_syscall_enter); mutex_unlock(&syscall_trace_lock); } @@ -327,7 +328,7 @@ int reg_event_syscall_exit(void *ptr) return -ENOSYS; mutex_lock(&syscall_trace_lock); if (!sys_refcount_exit) - ret = register_trace_syscall_exit(ftrace_syscall_exit); + ret = register_trace_sys_exit(ftrace_syscall_exit); if (ret) { pr_info("event trace: Could not activate" "syscall exit trace point"); @@ -352,7 +353,7 @@ void unreg_event_syscall_exit(void *ptr) sys_refcount_exit--; clear_bit(num, enabled_exit_syscalls); if (!sys_refcount_exit) - unregister_trace_syscall_exit(ftrace_syscall_exit); + unregister_trace_sys_exit(ftrace_syscall_exit); mutex_unlock(&syscall_trace_lock); } @@ -418,7 +419,7 @@ int reg_prof_syscall_enter(char *name) mutex_lock(&syscall_trace_lock); if (!sys_prof_refcount_enter) - ret = register_trace_syscall_enter(prof_syscall_enter); + ret = register_trace_sys_enter(prof_syscall_enter); if (ret) { pr_info("event trace: Could not activate" "syscall entry trace point"); @@ -442,7 +443,7 @@ void unreg_prof_syscall_enter(char *name) sys_prof_refcount_enter--; clear_bit(num, enabled_prof_enter_syscalls); if (!sys_prof_refcount_enter) - unregister_trace_syscall_enter(prof_syscall_enter); + unregister_trace_sys_enter(prof_syscall_enter); mutex_unlock(&syscall_trace_lock); } @@ -479,7 +480,7 @@ int reg_prof_syscall_exit(char *name) mutex_lock(&syscall_trace_lock); if (!sys_prof_refcount_exit) - ret = register_trace_syscall_exit(prof_syscall_exit); + ret = register_trace_sys_exit(prof_syscall_exit); if (ret) { pr_info("event trace: Could not activate" "syscall entry trace point"); @@ -503,7 +504,7 @@ void unreg_prof_syscall_exit(char *name) sys_prof_refcount_exit--; clear_bit(num, enabled_prof_exit_syscalls); if (!sys_prof_refcount_exit) - unregister_trace_syscall_exit(prof_syscall_exit); + unregister_trace_sys_exit(prof_syscall_exit); mutex_unlock(&syscall_trace_lock); } -- cgit v1.2.3 From f0693c8bd5c50380b299e19d19e7640024640b42 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Thu, 6 Aug 2009 14:59:32 -0400 Subject: tracing/sched: show CPU task wakes up on in trace event While debugging the scheduler push / pull algorithm, I found it very annoying that the sched wake up events did not show the CPU that the task was waking on. In order to analyze the scheduler, I needed that information. This patch adds recording of the CPU that a task is waking up on. Signed-off-by: Steven Rostedt --- include/trace/events/sched.h | 12 ++++++++---- 1 file changed, 8 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/include/trace/events/sched.h b/include/trace/events/sched.h index 8949bb7eb082..a581ef211ff5 100644 --- a/include/trace/events/sched.h +++ b/include/trace/events/sched.h @@ -94,6 +94,7 @@ TRACE_EVENT(sched_wakeup, __field( pid_t, pid ) __field( int, prio ) __field( int, success ) + __field( int, cpu ) ), TP_fast_assign( @@ -101,11 +102,12 @@ TRACE_EVENT(sched_wakeup, __entry->pid = p->pid; __entry->prio = p->prio; __entry->success = success; + __entry->cpu = task_cpu(p); ), - TP_printk("task %s:%d [%d] success=%d", + TP_printk("task %s:%d [%d] success=%d [%03d]", __entry->comm, __entry->pid, __entry->prio, - __entry->success) + __entry->success, __entry->cpu) ); /* @@ -125,6 +127,7 @@ TRACE_EVENT(sched_wakeup_new, __field( pid_t, pid ) __field( int, prio ) __field( int, success ) + __field( int, cpu ) ), TP_fast_assign( @@ -132,11 +135,12 @@ TRACE_EVENT(sched_wakeup_new, __entry->pid = p->pid; __entry->prio = p->prio; __entry->success = success; + __entry->cpu = task_cpu(p); ), - TP_printk("task %s:%d [%d] success=%d", + TP_printk("task %s:%d [%d] success=%d [%03d]", __entry->comm, __entry->pid, __entry->prio, - __entry->success) + __entry->success, __entry->cpu) ); /* -- cgit v1.2.3 From 43b51ead3f752a3935116e5b1a94254b8573734f Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Fri, 7 Aug 2009 10:33:22 +0800 Subject: tracing/filters: Add __field_ext() to TRACE_EVENT Add __field_ext(), so a field can be assigned to a specific filter_type, which matches a corresponding filter function. For example, a later patch will allow this: __field_ext(const char *, str, FILTER_PTR_STR); Signed-off-by: Li Zefan LKML-Reference: <4A7B9272.6050709@cn.fujitsu.com> [ Fixed a -1 to FILTER_OTHER Forward ported to latest kernel. ] Signed-off-by: Steven Rostedt --- include/linux/ftrace_event.h | 9 ++++++++- include/trace/ftrace.h | 31 ++++++++++++++++++++++++------- kernel/trace/trace_events.c | 11 ++++++++--- kernel/trace/trace_events_filter.c | 6 ------ kernel/trace/trace_export.c | 8 +++++--- kernel/trace/trace_syscalls.c | 6 ++++-- 6 files changed, 49 insertions(+), 22 deletions(-) (limited to 'include') diff --git a/include/linux/ftrace_event.h b/include/linux/ftrace_event.h index df5b085c4150..0440bea8f6bb 100644 --- a/include/linux/ftrace_event.h +++ b/include/linux/ftrace_event.h @@ -140,9 +140,16 @@ extern int filter_current_check_discard(struct ftrace_event_call *call, void *rec, struct ring_buffer_event *event); +enum { + FILTER_OTHER = 0, + FILTER_STATIC_STRING, + FILTER_DYN_STRING, +}; + extern int trace_define_field(struct ftrace_event_call *call, const char *type, const char *name, - int offset, int size, int is_signed); + int offset, int size, int is_signed, + int filter_type); extern int trace_define_common_fields(struct ftrace_event_call *call); #define is_signed_type(type) (((type)(-1)) < 0) diff --git a/include/trace/ftrace.h b/include/trace/ftrace.h index 127400255e4c..1b1f742a6045 100644 --- a/include/trace/ftrace.h +++ b/include/trace/ftrace.h @@ -21,6 +21,9 @@ #undef __field #define __field(type, item) type item; +#undef __field_ext +#define __field_ext(type, item, filter_type) type item; + #undef __array #define __array(type, item, len) type item[len]; @@ -62,7 +65,10 @@ */ #undef __field -#define __field(type, item); +#define __field(type, item) + +#undef __field_ext +#define __field_ext(type, item, filter_type) #undef __array #define __array(type, item, len) @@ -110,6 +116,9 @@ if (!ret) \ return 0; +#undef __field_ext +#define __field_ext(type, item, filter_type) __field(type, item) + #undef __array #define __array(type, item, len) \ ret = trace_seq_printf(s, "\tfield:" #type " " #item "[" #len "];\t" \ @@ -265,28 +274,33 @@ ftrace_raw_output_##call(struct trace_iterator *iter, int flags) \ #include TRACE_INCLUDE(TRACE_INCLUDE_FILE) -#undef __field -#define __field(type, item) \ +#undef __field_ext +#define __field_ext(type, item, filter_type) \ ret = trace_define_field(event_call, #type, #item, \ offsetof(typeof(field), item), \ - sizeof(field.item), is_signed_type(type)); \ + sizeof(field.item), \ + is_signed_type(type), filter_type); \ if (ret) \ return ret; +#undef __field +#define __field(type, item) __field_ext(type, item, FILTER_OTHER) + #undef __array #define __array(type, item, len) \ BUILD_BUG_ON(len > MAX_FILTER_STR_VAL); \ ret = trace_define_field(event_call, #type "[" #len "]", #item, \ offsetof(typeof(field), item), \ - sizeof(field.item), 0); \ + sizeof(field.item), 0, FILTER_OTHER); \ if (ret) \ return ret; #undef __dynamic_array #define __dynamic_array(type, item, len) \ ret = trace_define_field(event_call, "__data_loc " #type "[]", #item, \ - offsetof(typeof(field), __data_loc_##item), \ - sizeof(field.__data_loc_##item), 0); + offsetof(typeof(field), __data_loc_##item), \ + sizeof(field.__data_loc_##item), 0, \ + FILTER_OTHER); #undef __string #define __string(item, src) __dynamic_array(char, item, -1) @@ -320,6 +334,9 @@ ftrace_define_fields_##call(struct ftrace_event_call *event_call) \ #undef __field #define __field(type, item) +#undef __field_ext +#define __field_ext(type, item, filter_type) + #undef __array #define __array(type, item, len) diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c index 5740e90f4ca1..d33bcdeffe69 100644 --- a/kernel/trace/trace_events.c +++ b/kernel/trace/trace_events.c @@ -28,7 +28,8 @@ DEFINE_MUTEX(event_mutex); LIST_HEAD(ftrace_events); int trace_define_field(struct ftrace_event_call *call, const char *type, - const char *name, int offset, int size, int is_signed) + const char *name, int offset, int size, int is_signed, + int filter_type) { struct ftrace_event_field *field; @@ -44,7 +45,11 @@ int trace_define_field(struct ftrace_event_call *call, const char *type, if (!field->type) goto err; - field->filter_type = filter_assign_type(type); + if (filter_type == FILTER_OTHER) + field->filter_type = filter_assign_type(type); + else + field->filter_type = filter_type; + field->offset = offset; field->size = size; field->is_signed = is_signed; @@ -68,7 +73,7 @@ EXPORT_SYMBOL_GPL(trace_define_field); ret = trace_define_field(call, #type, "common_" #item, \ offsetof(typeof(ent), item), \ sizeof(ent.item), \ - is_signed_type(type)); \ + is_signed_type(type), FILTER_OTHER); \ if (ret) \ return ret; diff --git a/kernel/trace/trace_events_filter.c b/kernel/trace/trace_events_filter.c index 22e6d822bbaa..8a8e576733fc 100644 --- a/kernel/trace/trace_events_filter.c +++ b/kernel/trace/trace_events_filter.c @@ -475,12 +475,6 @@ static int filter_add_pred_fn(struct filter_parse_state *ps, return 0; } -enum { - FILTER_OTHER = 0, - FILTER_STATIC_STRING, - FILTER_DYN_STRING, -}; - int filter_assign_type(const char *type) { if (strstr(type, "__data_loc") && strstr(type, "char")) diff --git a/kernel/trace/trace_export.c b/kernel/trace/trace_export.c index 70875303ae46..029a91f42287 100644 --- a/kernel/trace/trace_export.c +++ b/kernel/trace/trace_export.c @@ -158,7 +158,8 @@ __attribute__((section("_ftrace_events"))) event_##call = { \ #define TRACE_FIELD(type, item, assign) \ ret = trace_define_field(event_call, #type, #item, \ offsetof(typeof(field), item), \ - sizeof(field.item), is_signed_type(type)); \ + sizeof(field.item), \ + is_signed_type(type), FILTER_OTHER); \ if (ret) \ return ret; @@ -166,7 +167,7 @@ __attribute__((section("_ftrace_events"))) event_##call = { \ #define TRACE_FIELD_SPECIAL(type, item, len, cmd) \ ret = trace_define_field(event_call, #type "[" #len "]", #item, \ offsetof(typeof(field), item), \ - sizeof(field.item), 0); \ + sizeof(field.item), 0, FILTER_OTHER); \ if (ret) \ return ret; @@ -174,7 +175,8 @@ __attribute__((section("_ftrace_events"))) event_##call = { \ #define TRACE_FIELD_SIGN(type, item, assign, is_signed) \ ret = trace_define_field(event_call, #type, #item, \ offsetof(typeof(field), item), \ - sizeof(field.item), is_signed); \ + sizeof(field.item), is_signed, \ + FILTER_OTHER); \ if (ret) \ return ret; diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c index 46c1b977a2cb..97a2454760b0 100644 --- a/kernel/trace/trace_syscalls.c +++ b/kernel/trace/trace_syscalls.c @@ -194,7 +194,8 @@ int syscall_enter_define_fields(struct ftrace_event_call *call) for (i = 0; i < meta->nb_args; i++) { ret = trace_define_field(call, meta->types[i], meta->args[i], offset, - sizeof(unsigned long), 0); + sizeof(unsigned long), 0, + FILTER_OTHER); offset += sizeof(unsigned long); } @@ -210,7 +211,8 @@ int syscall_exit_define_fields(struct ftrace_event_call *call) if (ret) return ret; - ret = trace_define_field(call, SYSCALL_FIELD(unsigned long, ret), 0); + ret = trace_define_field(call, SYSCALL_FIELD(unsigned long, ret), 0, + FILTER_OTHER); return ret; } -- cgit v1.2.3 From 87a342f5db69d53ea70493bb1ec69c9047677038 Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Fri, 7 Aug 2009 10:33:43 +0800 Subject: tracing/filters: Support filtering for char * strings Usually, char * entries are dangerous in traces because the string can be released whereas a pointer to it can still wait to be read from the ring buffer. But sometimes we can assume it's safe, like in case of RO data (eg: __file__ or __line__, used in bkl trace event). If these RO data are in a module and so is the call to the trace event, then it's safe, because the ring buffer will be flushed once this module get unloaded. To allow char * to be treated as a string: TRACE_EVENT(..., TP_STRUCT__entry( __field_ext(const char *, name, FILTER_PTR_STRING) ... ) ... ); The filtering will not dereference "char *" unless the developer explicitly sets FILTER_PTR_STR in __field_ext. Signed-off-by: Li Zefan LKML-Reference: <4A7B9287.90205@cn.fujitsu.com> Signed-off-by: Steven Rostedt --- include/linux/ftrace_event.h | 1 + kernel/trace/trace_events_filter.c | 26 +++++++++++++++++++++++--- 2 files changed, 24 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/include/linux/ftrace_event.h b/include/linux/ftrace_event.h index 0440bea8f6bb..ace2da9e0a0d 100644 --- a/include/linux/ftrace_event.h +++ b/include/linux/ftrace_event.h @@ -144,6 +144,7 @@ enum { FILTER_OTHER = 0, FILTER_STATIC_STRING, FILTER_DYN_STRING, + FILTER_PTR_STRING, }; extern int trace_define_field(struct ftrace_event_call *call, diff --git a/kernel/trace/trace_events_filter.c b/kernel/trace/trace_events_filter.c index 8a8e576733fc..9f03082c81d8 100644 --- a/kernel/trace/trace_events_filter.c +++ b/kernel/trace/trace_events_filter.c @@ -163,6 +163,20 @@ static int filter_pred_string(struct filter_pred *pred, void *event, return match; } +/* Filter predicate for char * pointers */ +static int filter_pred_pchar(struct filter_pred *pred, void *event, + int val1, int val2) +{ + char **addr = (char **)(event + pred->offset); + int cmp, match; + + cmp = strncmp(*addr, pred->str_val, pred->str_len); + + match = (!cmp) ^ pred->not; + + return match; +} + /* * Filter predicate for dynamic sized arrays of characters. * These are implemented through a list of strings at the end @@ -489,7 +503,8 @@ int filter_assign_type(const char *type) static bool is_string_field(struct ftrace_event_field *field) { return field->filter_type == FILTER_DYN_STRING || - field->filter_type == FILTER_STATIC_STRING; + field->filter_type == FILTER_STATIC_STRING || + field->filter_type == FILTER_PTR_STRING; } static int is_legal_op(struct ftrace_event_field *field, int op) @@ -579,11 +594,16 @@ static int filter_add_pred(struct filter_parse_state *ps, } if (is_string_field(field)) { + pred->str_len = field->size; + if (field->filter_type == FILTER_STATIC_STRING) fn = filter_pred_string; - else + else if (field->filter_type == FILTER_DYN_STRING) fn = filter_pred_strloc; - pred->str_len = field->size; + else { + fn = filter_pred_pchar; + pred->str_len = strlen(pred->str_val); + } } else { if (field->is_signed) ret = strict_strtoll(pred->str_val, 0, &val); -- cgit v1.2.3 From 5ac35daa9343936038a3c9c4f4d6d3fe6a2a7bd8 Mon Sep 17 00:00:00 2001 From: Xiao Guangrong Date: Tue, 25 Aug 2009 14:06:22 +0800 Subject: tracing/events: fix the include file dependencies The TRACE_EVENT depends on the include/linux/tracepoint.h first and include/trace/ftrace.h later, if we include the ftrace.h early, a building error will occur. Both define TRACE_EVENT in trace_a.h and trace_b.h, if we include those in .c file, like this: #define CREATE_TRACE_POINTS include include The above will not work, because the TRACE_EVENT was re-defined by the previous .h file. Reported-by: Wei Yongjun Signed-off-by: Xiao Guangrong LKML-Reference: <4A937F5E.3020802@cn.fujitsu.com> Signed-off-by: Steven Rostedt --- include/linux/tracepoint.h | 3 +-- include/trace/define_trace.h | 1 + 2 files changed, 2 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/include/linux/tracepoint.h b/include/linux/tracepoint.h index 5984ed04c03b..81709854f7ab 100644 --- a/include/linux/tracepoint.h +++ b/include/linux/tracepoint.h @@ -180,6 +180,7 @@ static inline void tracepoint_synchronize_unregister(void) } #define PARAMS(args...) args +#endif #ifndef TRACE_EVENT /* @@ -287,5 +288,3 @@ static inline void tracepoint_synchronize_unregister(void) #define TRACE_EVENT(name, proto, args, struct, assign, print) \ DECLARE_TRACE(name, PARAMS(proto), PARAMS(args)) #endif - -#endif diff --git a/include/trace/define_trace.h b/include/trace/define_trace.h index f7a7ae1e8f90..cd150b9d8e32 100644 --- a/include/trace/define_trace.h +++ b/include/trace/define_trace.h @@ -56,6 +56,7 @@ #include #endif +#undef TRACE_EVENT #undef TRACE_HEADER_MULTI_READ /* Only undef what we defined in this file */ -- cgit v1.2.3 From 7cb2e3ee2aeec5b83ecadba929a2dc575dd4008f Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Wed, 26 Aug 2009 00:32:37 -0400 Subject: tracing: add comments to explain TRACE_EVENT out of protection The commit: commit 5ac35daa9343936038a3c9c4f4d6d3fe6a2a7bd8 Author: Xiao Guangrong tracing/events: fix the include file dependencies Moved the TRACE_EVENT out of the ifdef protection of tracepoints.h but uses the define of TRACE_EVENT itself as protection. This patch adds comments to explain why. Signed-off-by: Steven Rostedt --- include/linux/tracepoint.h | 13 +++++++++++-- 1 file changed, 11 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/include/linux/tracepoint.h b/include/linux/tracepoint.h index 81709854f7ab..0341f2e2698a 100644 --- a/include/linux/tracepoint.h +++ b/include/linux/tracepoint.h @@ -180,7 +180,15 @@ static inline void tracepoint_synchronize_unregister(void) } #define PARAMS(args...) args -#endif + +#endif /* _LINUX_TRACEPOINT_H */ + +/* + * Note: we keep the TRACE_EVENT outside the include file ifdef protection. + * This is due to the way trace events work. If a file includes two + * trace event headers under one "CREATE_TRACE_POINTS" the first include + * will override the TRACE_EVENT and break the second include. + */ #ifndef TRACE_EVENT /* @@ -287,4 +295,5 @@ static inline void tracepoint_synchronize_unregister(void) #define TRACE_EVENT(name, proto, args, struct, assign, print) \ DECLARE_TRACE(name, PARAMS(proto), PARAMS(args)) -#endif + +#endif /* ifdef TRACE_EVENT (see note above) */ -- cgit v1.2.3 From 2246b2f1b43f3fbd128e72b129dcbbd3202cc592 Mon Sep 17 00:00:00 2001 From: "Gustavo F. Padovan" Date: Wed, 26 Aug 2009 04:04:02 -0300 Subject: Bluetooth: Handle L2CAP case when the remote receiver is busy Implement all issues related to RemoteBusy in the RECV state table. Signed-off-by: Gustavo F. Padovan Signed-off-by: Marcel Holtmann --- include/net/bluetooth/l2cap.h | 2 ++ net/bluetooth/l2cap.c | 25 +++++++++++++++++++++++-- 2 files changed, 25 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/include/net/bluetooth/l2cap.h b/include/net/bluetooth/l2cap.h index 7ca614ac5d43..9516f4b4a3c2 100644 --- a/include/net/bluetooth/l2cap.h +++ b/include/net/bluetooth/l2cap.h @@ -373,6 +373,8 @@ struct l2cap_pinfo { #define L2CAP_CONN_WAIT_F 0x04 #define L2CAP_CONN_SREJ_ACT 0x08 #define L2CAP_CONN_SEND_PBIT 0x10 +#define L2CAP_CONN_REMOTE_BUSY 0x20 +#define L2CAP_CONN_LOCAL_BUSY 0x40 #define __mod_retrans_timer() mod_timer(&l2cap_pi(sk)->retrans_timer, \ jiffies + msecs_to_jiffies(L2CAP_DEFAULT_RETRANS_TO)); diff --git a/net/bluetooth/l2cap.c b/net/bluetooth/l2cap.c index 0a36c61c011f..40fbf5cb1f7e 100644 --- a/net/bluetooth/l2cap.c +++ b/net/bluetooth/l2cap.c @@ -1350,7 +1350,8 @@ static int l2cap_ertm_send(struct sock *sk) if (pi->conn_state & L2CAP_CONN_WAIT_F) return 0; - while ((skb = sk->sk_send_head) && (!l2cap_tx_window_full(sk))) { + while ((skb = sk->sk_send_head) && (!l2cap_tx_window_full(sk)) + && !(pi->conn_state & L2CAP_CONN_REMOTE_BUSY)) { tx_skb = skb_clone(skb, GFP_ATOMIC); if (pi->remote_max_tx && @@ -3351,7 +3352,10 @@ static inline int l2cap_data_channel_sframe(struct sock *sk, u16 rx_control, str control |= L2CAP_SUPER_RCV_READY | (pi->buffer_seq << L2CAP_CTRL_REQSEQ_SHIFT); l2cap_send_sframe(l2cap_pi(sk), control); + pi->conn_state &= ~L2CAP_CONN_REMOTE_BUSY; + } else if (rx_control & L2CAP_CTRL_FINAL) { + pi->conn_state &= ~L2CAP_CONN_REMOTE_BUSY; pi->expected_ack_seq = tx_seq; l2cap_drop_acked_frames(sk); @@ -3366,13 +3370,19 @@ static inline int l2cap_data_channel_sframe(struct sock *sk, u16 rx_control, str } else { pi->expected_ack_seq = tx_seq; l2cap_drop_acked_frames(sk); - if (pi->unacked_frames > 0) + + if ((pi->conn_state & L2CAP_CONN_REMOTE_BUSY) + && (pi->unacked_frames > 0)) __mod_retrans_timer(); + l2cap_ertm_send(sk); + pi->conn_state &= ~L2CAP_CONN_REMOTE_BUSY; } break; case L2CAP_SUPER_REJECT: + pi->conn_state &= ~L2CAP_CONN_REMOTE_BUSY; + pi->expected_ack_seq = __get_reqseq(rx_control); l2cap_drop_acked_frames(sk); @@ -3384,6 +3394,8 @@ static inline int l2cap_data_channel_sframe(struct sock *sk, u16 rx_control, str break; case L2CAP_SUPER_SELECT_REJECT: + pi->conn_state &= ~L2CAP_CONN_REMOTE_BUSY; + if (rx_control & L2CAP_CTRL_POLL) { l2cap_retransmit_frame(sk, tx_seq); pi->expected_ack_seq = tx_seq; @@ -3410,6 +3422,15 @@ static inline int l2cap_data_channel_sframe(struct sock *sk, u16 rx_control, str break; case L2CAP_SUPER_RCV_NOT_READY: + pi->conn_state |= L2CAP_CONN_REMOTE_BUSY; + pi->expected_ack_seq = tx_seq; + l2cap_drop_acked_frames(sk); + + del_timer(&l2cap_pi(sk)->retrans_timer); + if (rx_control & L2CAP_CTRL_POLL) { + u16 control = L2CAP_CTRL_FINAL | L2CAP_SUPER_RCV_READY; + l2cap_send_sframe(l2cap_pi(sk), control); + } break; } -- cgit v1.2.3 From 06e799764eb7c2e4640888d438c3524d756613e1 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Tue, 25 Aug 2009 18:53:37 -0700 Subject: rcu: Remove lockdep annotations from RCU's _notrace() API members The lockdep annotations rcu_read_acquire() and rcu_read_release() might lead to infinite looping if called from lockdep. So this patch removes them. Formal repost of http://lkml.org/lkml/2009/8/25/309 on the strength of Lai Jiangshan's review. Suggested-by: Lai Jiangshan Suggested-by: Mathieu Desnoyers Reviewed-by: Lai Jiangshan Signed-off-by: Paul E. McKenney Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <20090826015337.GA18904@linux.vnet.ibm.com> Signed-off-by: Ingo Molnar --- include/linux/rcupdate.h | 2 -- 1 file changed, 2 deletions(-) (limited to 'include') diff --git a/include/linux/rcupdate.h b/include/linux/rcupdate.h index 8b4422c558da..95e0615f4d75 100644 --- a/include/linux/rcupdate.h +++ b/include/linux/rcupdate.h @@ -195,7 +195,6 @@ static inline notrace void rcu_read_lock_sched_notrace(void) { preempt_disable_notrace(); __acquire(RCU_SCHED); - rcu_read_acquire(); } /* @@ -211,7 +210,6 @@ static inline void rcu_read_unlock_sched(void) } static inline notrace void rcu_read_unlock_sched_notrace(void) { - rcu_read_release(); __release(RCU_SCHED); preempt_enable_notrace(); } -- cgit v1.2.3 From 0396c215f301e92677d1e9a064b405e31501dc1d Mon Sep 17 00:00:00 2001 From: David Vrabel Date: Tue, 25 Aug 2009 16:41:06 +0100 Subject: uwb: avoid radio controller reset loops If a radio controller reset attempt occurs while a probe() or remove() is in progress it fails and is retried endlessly, potentially preventing the probe() or remove() from completing. If a reset fails, sleep for a bit before retrying the reset. This allows the probe()/remove() to complete. Signed-off-by: David Vrabel --- drivers/uwb/hwa-rc.c | 3 +-- drivers/uwb/reset.c | 21 +++++++++++---------- drivers/uwb/umc-bus.c | 2 +- drivers/uwb/whc-rc.c | 3 +-- include/linux/uwb.h | 2 +- 5 files changed, 15 insertions(+), 16 deletions(-) (limited to 'include') diff --git a/drivers/uwb/hwa-rc.c b/drivers/uwb/hwa-rc.c index 9052bcb4f528..e7eeb63fab23 100644 --- a/drivers/uwb/hwa-rc.c +++ b/drivers/uwb/hwa-rc.c @@ -887,8 +887,7 @@ static int hwarc_post_reset(struct usb_interface *iface) struct hwarc *hwarc = usb_get_intfdata(iface); struct uwb_rc *uwb_rc = hwarc->uwb_rc; - uwb_rc_post_reset(uwb_rc); - return 0; + return uwb_rc_post_reset(uwb_rc); } /** USB device ID's that we handle */ diff --git a/drivers/uwb/reset.c b/drivers/uwb/reset.c index 70f8050221ff..7f0512e43d9d 100644 --- a/drivers/uwb/reset.c +++ b/drivers/uwb/reset.c @@ -30,6 +30,7 @@ */ #include #include +#include #include "uwb-internal.h" @@ -323,13 +324,15 @@ int uwbd_msg_handle_reset(struct uwb_event *evt) dev_info(&rc->uwb_dev.dev, "resetting radio controller\n"); ret = rc->reset(rc); - if (ret) { + if (ret < 0) { dev_err(&rc->uwb_dev.dev, "failed to reset hardware: %d\n", ret); goto error; } return 0; error: - /* Nothing can be done except try the reset again. */ + /* Nothing can be done except try the reset again. Wait a bit + to avoid reset loops during probe() or remove(). */ + msleep(1000); uwb_rc_reset_all(rc); return ret; } @@ -368,22 +371,20 @@ void uwb_rc_pre_reset(struct uwb_rc *rc) } EXPORT_SYMBOL_GPL(uwb_rc_pre_reset); -void uwb_rc_post_reset(struct uwb_rc *rc) +int uwb_rc_post_reset(struct uwb_rc *rc) { int ret; ret = rc->start(rc); if (ret) - goto error; + goto out; ret = uwb_rc_mac_addr_set(rc, &rc->uwb_dev.mac_addr); if (ret) - goto error; + goto out; ret = uwb_rc_dev_addr_set(rc, &rc->uwb_dev.dev_addr); if (ret) - goto error; - return; -error: - /* Nothing can be done except try the reset again. */ - uwb_rc_reset_all(rc); + goto out; +out: + return ret; } EXPORT_SYMBOL_GPL(uwb_rc_post_reset); diff --git a/drivers/uwb/umc-bus.c b/drivers/uwb/umc-bus.c index 5ad36164c13b..cdd6c8efc9f8 100644 --- a/drivers/uwb/umc-bus.c +++ b/drivers/uwb/umc-bus.c @@ -66,7 +66,7 @@ int umc_controller_reset(struct umc_dev *umc) return -EAGAIN; ret = device_for_each_child(parent, parent, umc_bus_pre_reset_helper); if (ret >= 0) - device_for_each_child(parent, parent, umc_bus_post_reset_helper); + ret = device_for_each_child(parent, parent, umc_bus_post_reset_helper); up(&parent->sem); return ret; diff --git a/drivers/uwb/whc-rc.c b/drivers/uwb/whc-rc.c index 19a1dd129212..1d9a6f54658e 100644 --- a/drivers/uwb/whc-rc.c +++ b/drivers/uwb/whc-rc.c @@ -443,8 +443,7 @@ static int whcrc_post_reset(struct umc_dev *umc) struct whcrc *whcrc = umc_get_drvdata(umc); struct uwb_rc *uwb_rc = whcrc->uwb_rc; - uwb_rc_post_reset(uwb_rc); - return 0; + return uwb_rc_post_reset(uwb_rc); } /* PCI device ID's that we handle [so it gets loaded] */ diff --git a/include/linux/uwb.h b/include/linux/uwb.h index c02128991ff7..7fc9746f22cd 100644 --- a/include/linux/uwb.h +++ b/include/linux/uwb.h @@ -597,7 +597,7 @@ void uwb_rc_neh_grok(struct uwb_rc *, void *, size_t); void uwb_rc_neh_error(struct uwb_rc *, int); void uwb_rc_reset_all(struct uwb_rc *rc); void uwb_rc_pre_reset(struct uwb_rc *rc); -void uwb_rc_post_reset(struct uwb_rc *rc); +int uwb_rc_post_reset(struct uwb_rc *rc); /** * uwb_rsv_is_owner - is the owner of this reservation the RC? -- cgit v1.2.3 From 9e36fda0b359d2a6ae039c3d7e71a04502a77898 Mon Sep 17 00:00:00 2001 From: Venkatesh Pallipadi Date: Fri, 10 Jul 2009 09:57:35 -0700 Subject: x86, pat: Add PAT reserve free to io_mapping* APIs io_mapping_* interfaces were added, mainly for graphics drivers. Make this interface go through the PAT reserve/free, instead of hardcoding WC mapping. This makes sure that there are no aliases due to unconditional WC setting. Signed-off-by: Venkatesh Pallipadi Signed-off-by: Suresh Siddha Signed-off-by: H. Peter Anvin --- arch/x86/include/asm/iomap.h | 9 ++++++--- arch/x86/mm/iomap_32.c | 27 +++++++++++++++++++++++++-- include/linux/io-mapping.h | 17 ++++++++++++----- 3 files changed, 43 insertions(+), 10 deletions(-) (limited to 'include') diff --git a/arch/x86/include/asm/iomap.h b/arch/x86/include/asm/iomap.h index 0e9fe1d9d971..f35eb45d6576 100644 --- a/arch/x86/include/asm/iomap.h +++ b/arch/x86/include/asm/iomap.h @@ -26,13 +26,16 @@ #include #include -int -is_io_mapping_possible(resource_size_t base, unsigned long size); - void * iomap_atomic_prot_pfn(unsigned long pfn, enum km_type type, pgprot_t prot); void iounmap_atomic(void *kvaddr, enum km_type type); +int +iomap_create_wc(resource_size_t base, unsigned long size, pgprot_t *prot); + +void +iomap_free(resource_size_t base, unsigned long size); + #endif /* _ASM_X86_IOMAP_H */ diff --git a/arch/x86/mm/iomap_32.c b/arch/x86/mm/iomap_32.c index fe6f84ca121e..84e236ce76ba 100644 --- a/arch/x86/mm/iomap_32.c +++ b/arch/x86/mm/iomap_32.c @@ -21,7 +21,7 @@ #include #include -int is_io_mapping_possible(resource_size_t base, unsigned long size) +static int is_io_mapping_possible(resource_size_t base, unsigned long size) { #if !defined(CONFIG_X86_PAE) && defined(CONFIG_PHYS_ADDR_T_64BIT) /* There is no way to map greater than 1 << 32 address without PAE */ @@ -30,7 +30,30 @@ int is_io_mapping_possible(resource_size_t base, unsigned long size) #endif return 1; } -EXPORT_SYMBOL_GPL(is_io_mapping_possible); + +int iomap_create_wc(resource_size_t base, unsigned long size, pgprot_t *prot) +{ + unsigned long flag = _PAGE_CACHE_WC; + int ret; + + if (!is_io_mapping_possible(base, size)) + return -EINVAL; + + ret = io_reserve_memtype(base, base + size, &flag); + if (ret) + return ret; + + *prot = __pgprot(__PAGE_KERNEL | flag); + return 0; +} +EXPORT_SYMBOL_GPL(iomap_create_wc); + +void +iomap_free(resource_size_t base, unsigned long size) +{ + io_free_memtype(base, base + size); +} +EXPORT_SYMBOL_GPL(iomap_free); void *kmap_atomic_prot_pfn(unsigned long pfn, enum km_type type, pgprot_t prot) { diff --git a/include/linux/io-mapping.h b/include/linux/io-mapping.h index 0adb0f91568c..97eb928b4924 100644 --- a/include/linux/io-mapping.h +++ b/include/linux/io-mapping.h @@ -49,23 +49,30 @@ static inline struct io_mapping * io_mapping_create_wc(resource_size_t base, unsigned long size) { struct io_mapping *iomap; - - if (!is_io_mapping_possible(base, size)) - return NULL; + pgprot_t prot; iomap = kmalloc(sizeof(*iomap), GFP_KERNEL); if (!iomap) - return NULL; + goto out_err; + + if (iomap_create_wc(base, size, &prot)) + goto out_free; iomap->base = base; iomap->size = size; - iomap->prot = pgprot_writecombine(__pgprot(__PAGE_KERNEL)); + iomap->prot = prot; return iomap; + +out_free: + kfree(iomap); +out_err: + return NULL; } static inline void io_mapping_free(struct io_mapping *mapping) { + iomap_free(mapping->base, mapping->size); kfree(mapping); } -- cgit v1.2.3 From 46cf98cdaef5471926010b5bddf84c44ec177fdd Mon Sep 17 00:00:00 2001 From: Venkatesh Pallipadi Date: Fri, 10 Jul 2009 09:57:37 -0700 Subject: x86, pat: Generalize the use of page flag PG_uncached Only IA64 was using PG_uncached as of now. We now intend to use this bit in x86 as well, to keep track of memory type of those addresses that have page struct for them. So, generalize the use of that bit across ia64 and x86. Signed-off-by: Venkatesh Pallipadi Signed-off-by: Suresh Siddha Signed-off-by: H. Peter Anvin --- arch/ia64/Kconfig | 4 ++++ arch/x86/Kconfig | 4 ++++ include/linux/page-flags.h | 4 ++-- 3 files changed, 10 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/arch/ia64/Kconfig b/arch/ia64/Kconfig index 170042b420d4..e6246119932a 100644 --- a/arch/ia64/Kconfig +++ b/arch/ia64/Kconfig @@ -112,6 +112,10 @@ config IA64_UNCACHED_ALLOCATOR bool select GENERIC_ALLOCATOR +config ARCH_USES_PG_UNCACHED + def_bool y + depends on IA64_UNCACHED_ALLOCATOR + config AUDIT_ARCH bool default y diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index c07f72205909..8e1595382196 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -1414,6 +1414,10 @@ config X86_PAT If unsure, say Y. +config ARCH_USES_PG_UNCACHED + def_bool y + depends on X86_PAT + config EFI bool "EFI runtime service support" depends on ACPI diff --git a/include/linux/page-flags.h b/include/linux/page-flags.h index e2e5ce543595..2b87acfc5f87 100644 --- a/include/linux/page-flags.h +++ b/include/linux/page-flags.h @@ -99,7 +99,7 @@ enum pageflags { #ifdef CONFIG_HAVE_MLOCKED_PAGE_BIT PG_mlocked, /* Page is vma mlocked */ #endif -#ifdef CONFIG_IA64_UNCACHED_ALLOCATOR +#ifdef CONFIG_ARCH_USES_PG_UNCACHED PG_uncached, /* Page has been mapped as uncached */ #endif __NR_PAGEFLAGS, @@ -257,7 +257,7 @@ PAGEFLAG_FALSE(Mlocked) SETPAGEFLAG_NOOP(Mlocked) TESTCLEARFLAG_FALSE(Mlocked) #endif -#ifdef CONFIG_IA64_UNCACHED_ALLOCATOR +#ifdef CONFIG_ARCH_USES_PG_UNCACHED PAGEFLAG(Uncached, uncached) #else PAGEFLAG_FALSE(Uncached) -- cgit v1.2.3 From c9c97b8c75019814d8c007059bc827bb475be917 Mon Sep 17 00:00:00 2001 From: Dave Airlie Date: Thu, 27 Aug 2009 09:53:47 +1000 Subject: drm/ttm: consolidate cache flushing code in one place. This merges the TTM and drm cache flushing into one file in the drm core. Signed-off-by: Dave Airlie --- drivers/gpu/drm/drm_cache.c | 51 +++++++++++++++++++++++++++------ drivers/gpu/drm/ttm/ttm_tt.c | 67 ++------------------------------------------ include/drm/drm_cache.h | 38 +++++++++++++++++++++++++ 3 files changed, 82 insertions(+), 74 deletions(-) create mode 100644 include/drm/drm_cache.h (limited to 'include') diff --git a/drivers/gpu/drm/drm_cache.c b/drivers/gpu/drm/drm_cache.c index 0e994a0e46d4..3a5575e638db 100644 --- a/drivers/gpu/drm/drm_cache.c +++ b/drivers/gpu/drm/drm_cache.c @@ -45,25 +45,58 @@ drm_clflush_page(struct page *page) clflush(page_virtual + i); kunmap_atomic(page_virtual, KM_USER0); } -#endif +static void drm_cache_flush_clflush(struct page *pages[], + unsigned long num_pages) +{ + unsigned long i; + + mb(); + for (i = 0; i < num_pages; i++) + drm_clflush_page(*pages++); + mb(); +} + +static void +drm_clflush_ipi_handler(void *null) +{ + wbinvd(); +} +#elif !defined(__powerpc__) +static void drm_cache_ipi_handler(void *dummy) +{ +} +#endif void drm_clflush_pages(struct page *pages[], unsigned long num_pages) { #if defined(CONFIG_X86) if (cpu_has_clflush) { - unsigned long i; - - mb(); - for (i = 0; i < num_pages; ++i) - drm_clflush_page(*pages++); - mb(); - + drm_cache_flush_clflush(pages, num_pages); return; } - wbinvd(); + if (on_each_cpu(drm_clflush_ipi_handler, NULL, 1) != 0) + printk(KERN_ERR "Timed out waiting for cache flush.\n"); + +#elif defined(__powerpc__) + unsigned long i; + for (i = 0; i < num_pages; i++) { + struct page *page = pages[i]; + void *page_virtual; + + if (unlikely(page == NULL)) + continue; + + page_virtual = kmap_atomic(page, KM_USER0); + flush_dcache_range((unsigned long)page_virtual, + (unsigned long)page_virtual + PAGE_SIZE); + kunmap_atomic(page_virtual, KM_USER0); + } +#else + if (on_each_cpu(drm_clflush_ipi_handler, NULL, 1) != 0) + printk(KERN_ERR "Timed out waiting for drm cache flush\n"); #endif } EXPORT_SYMBOL(drm_clflush_pages); diff --git a/drivers/gpu/drm/ttm/ttm_tt.c b/drivers/gpu/drm/ttm/ttm_tt.c index 42cca5519761..a55ee1a56c16 100644 --- a/drivers/gpu/drm/ttm/ttm_tt.c +++ b/drivers/gpu/drm/ttm/ttm_tt.c @@ -34,76 +34,13 @@ #include #include #include +#include "drm_cache.h" #include "ttm/ttm_module.h" #include "ttm/ttm_bo_driver.h" #include "ttm/ttm_placement.h" static int ttm_tt_swapin(struct ttm_tt *ttm); -#if defined(CONFIG_X86) -static void ttm_tt_clflush_page(struct page *page) -{ - uint8_t *page_virtual; - unsigned int i; - - if (unlikely(page == NULL)) - return; - - page_virtual = kmap_atomic(page, KM_USER0); - - for (i = 0; i < PAGE_SIZE; i += boot_cpu_data.x86_clflush_size) - clflush(page_virtual + i); - - kunmap_atomic(page_virtual, KM_USER0); -} - -static void ttm_tt_cache_flush_clflush(struct page *pages[], - unsigned long num_pages) -{ - unsigned long i; - - mb(); - for (i = 0; i < num_pages; ++i) - ttm_tt_clflush_page(*pages++); - mb(); -} -#elif !defined(__powerpc__) -static void ttm_tt_ipi_handler(void *null) -{ - ; -} -#endif - -void ttm_tt_cache_flush(struct page *pages[], unsigned long num_pages) -{ - -#if defined(CONFIG_X86) - if (cpu_has_clflush) { - ttm_tt_cache_flush_clflush(pages, num_pages); - return; - } -#elif defined(__powerpc__) - unsigned long i; - - for (i = 0; i < num_pages; ++i) { - struct page *page = pages[i]; - void *page_virtual; - - if (unlikely(page == NULL)) - continue; - - page_virtual = kmap_atomic(page, KM_USER0); - flush_dcache_range((unsigned long) page_virtual, - (unsigned long) page_virtual + PAGE_SIZE); - kunmap_atomic(page_virtual, KM_USER0); - } -#else - if (on_each_cpu(ttm_tt_ipi_handler, NULL, 1) != 0) - printk(KERN_ERR TTM_PFX - "Timed out waiting for drm cache flush.\n"); -#endif -} - /** * Allocates storage for pointers to the pages that back the ttm. * @@ -302,7 +239,7 @@ static int ttm_tt_set_caching(struct ttm_tt *ttm, } if (ttm->caching_state == tt_cached) - ttm_tt_cache_flush(ttm->pages, ttm->num_pages); + drm_clflush_pages(ttm->pages, ttm->num_pages); for (i = 0; i < ttm->num_pages; ++i) { cur_page = ttm->pages[i]; diff --git a/include/drm/drm_cache.h b/include/drm/drm_cache.h new file mode 100644 index 000000000000..7bfb063029d8 --- /dev/null +++ b/include/drm/drm_cache.h @@ -0,0 +1,38 @@ +/************************************************************************** + * + * Copyright 2009 Red Hat Inc. + * All Rights Reserved. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the + * "Software"), to deal in the Software without restriction, including + * without limitation the rights to use, copy, modify, merge, publish, + * distribute, sub license, and/or sell copies of the Software, and to + * permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice (including the + * next paragraph) shall be included in all copies or substantial portions + * of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDERS, AUTHORS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM, + * DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR + * OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE + * USE OR OTHER DEALINGS IN THE SOFTWARE. + * + * + **************************************************************************/ +/* + * Authors: + * Dave Airlie + */ + +#ifndef _DRM_CACHE_H_ +#define _DRM_CACHE_H_ + +void drm_clflush_pages(struct page *pages[], unsigned long num_pages); + +#endif -- cgit v1.2.3 From a1a2d1d32250f6fcc317419e9dfb4a5a6946d2e6 Mon Sep 17 00:00:00 2001 From: Pekka Paalanen Date: Sun, 23 Aug 2009 12:40:55 +0300 Subject: drm: GEM handles are u32, not int Several functions in the GEM kernel API used int as handle type, but user API has it __u32 which is also the intended type. Replace int with u32. Signed-off-by: Pekka Paalanen Signed-off-by: Dave Airlie --- drivers/gpu/drm/drm_gem.c | 11 +++++------ drivers/gpu/drm/i915/i915_gem.c | 3 ++- include/drm/drmP.h | 4 ++-- 3 files changed, 9 insertions(+), 9 deletions(-) (limited to 'include') diff --git a/drivers/gpu/drm/drm_gem.c b/drivers/gpu/drm/drm_gem.c index ffe8f4394d50..230c9ffdd5e9 100644 --- a/drivers/gpu/drm/drm_gem.c +++ b/drivers/gpu/drm/drm_gem.c @@ -164,7 +164,7 @@ EXPORT_SYMBOL(drm_gem_object_alloc); * Removes the mapping from handle to filp for this object. */ static int -drm_gem_handle_delete(struct drm_file *filp, int handle) +drm_gem_handle_delete(struct drm_file *filp, u32 handle) { struct drm_device *dev; struct drm_gem_object *obj; @@ -207,7 +207,7 @@ drm_gem_handle_delete(struct drm_file *filp, int handle) int drm_gem_handle_create(struct drm_file *file_priv, struct drm_gem_object *obj, - int *handlep) + u32 *handlep) { int ret; @@ -221,7 +221,7 @@ again: /* do the allocation under our spinlock */ spin_lock(&file_priv->table_lock); - ret = idr_get_new_above(&file_priv->object_idr, obj, 1, handlep); + ret = idr_get_new_above(&file_priv->object_idr, obj, 1, (int *)handlep); spin_unlock(&file_priv->table_lock); if (ret == -EAGAIN) goto again; @@ -237,7 +237,7 @@ EXPORT_SYMBOL(drm_gem_handle_create); /** Returns a reference to the object named by the handle. */ struct drm_gem_object * drm_gem_object_lookup(struct drm_device *dev, struct drm_file *filp, - int handle) + u32 handle) { struct drm_gem_object *obj; @@ -344,7 +344,7 @@ drm_gem_open_ioctl(struct drm_device *dev, void *data, struct drm_gem_open *args = data; struct drm_gem_object *obj; int ret; - int handle; + u32 handle; if (!(dev->driver->driver_features & DRIVER_GEM)) return -ENODEV; @@ -539,7 +539,6 @@ int drm_gem_mmap(struct file *filp, struct vm_area_struct *vma) vma->vm_flags |= VM_RESERVED | VM_IO | VM_PFNMAP | VM_DONTEXPAND; vma->vm_ops = obj->dev->driver->gem_vm_ops; vma->vm_private_data = map->handle; - /* FIXME: use pgprot_writecombine when available */ vma->vm_page_prot = pgprot_writecombine(vma->vm_page_prot); /* Take a ref for this mapping of the object, so that the fault diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index 140bee142fc2..0e6c9cca897c 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -111,7 +111,8 @@ i915_gem_create_ioctl(struct drm_device *dev, void *data, { struct drm_i915_gem_create *args = data; struct drm_gem_object *obj; - int handle, ret; + int ret; + u32 handle; args->size = roundup(args->size, PAGE_SIZE); diff --git a/include/drm/drmP.h b/include/drm/drmP.h index e0f1c1fee58b..eeefb6369e19 100644 --- a/include/drm/drmP.h +++ b/include/drm/drmP.h @@ -1441,7 +1441,7 @@ drm_gem_object_unreference(struct drm_gem_object *obj) int drm_gem_handle_create(struct drm_file *file_priv, struct drm_gem_object *obj, - int *handlep); + u32 *handlep); static inline void drm_gem_object_handle_reference(struct drm_gem_object *obj) @@ -1467,7 +1467,7 @@ drm_gem_object_handle_unreference(struct drm_gem_object *obj) struct drm_gem_object *drm_gem_object_lookup(struct drm_device *dev, struct drm_file *filp, - int handle); + u32 handle); int drm_gem_close_ioctl(struct drm_device *dev, void *data, struct drm_file *file_priv); int drm_gem_flink_ioctl(struct drm_device *dev, void *data, -- cgit v1.2.3 From 8e7ee27095aee87b5db1b0061e2ceea5878a1bbd Mon Sep 17 00:00:00 2001 From: David Rientjes Date: Wed, 26 Aug 2009 14:29:21 -0700 Subject: flex_array: declare parts member to have incomplete type The `parts' member of struct flex_array should evaluate to an incomplete type so that sizeof() cannot be used and C99 does not require the zero-length specification. Signed-off-by: David Rientjes Acked-by: Dave Hansen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/flex_array.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/flex_array.h b/include/linux/flex_array.h index 23c1ec79a31b..603160db7c98 100644 --- a/include/linux/flex_array.h +++ b/include/linux/flex_array.h @@ -21,7 +21,7 @@ struct flex_array { struct { int element_size; int total_nr_elements; - struct flex_array_part *parts[0]; + struct flex_array_part *parts[]; }; /* * This little trick makes sure that -- cgit v1.2.3 From b62e408c05228f40e69bb38a48db8961cac6cd23 Mon Sep 17 00:00:00 2001 From: David Rientjes Date: Wed, 26 Aug 2009 14:29:22 -0700 Subject: flex_array: convert element_nr formals to unsigned It's problematic to allow signed element_nr's or total's to be passed as part of the flex array API. flex_array_alloc() allows total_nr_elements to be set to a negative quantity, which is obviously erroneous. flex_array_get() and flex_array_put() allows negative array indices in dereferencing an array part, which could address memory mapped before struct flex_array. The fix is to convert all existing element_nr formals to be qualified as unsigned. Existing checks to compare it to total_nr_elements or the max array size based on element_size need not be changed. Signed-off-by: David Rientjes Cc: Dave Hansen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/flex_array.h | 10 ++++++---- lib/flex_array.c | 24 +++++++++++++----------- 2 files changed, 19 insertions(+), 15 deletions(-) (limited to 'include') diff --git a/include/linux/flex_array.h b/include/linux/flex_array.h index 603160db7c98..45ff18491514 100644 --- a/include/linux/flex_array.h +++ b/include/linux/flex_array.h @@ -36,12 +36,14 @@ struct flex_array { .total_nr_elements = (total), \ } } } -struct flex_array *flex_array_alloc(int element_size, int total, gfp_t flags); -int flex_array_prealloc(struct flex_array *fa, int start, int end, gfp_t flags); +struct flex_array *flex_array_alloc(int element_size, unsigned int total, + gfp_t flags); +int flex_array_prealloc(struct flex_array *fa, unsigned int start, + unsigned int end, gfp_t flags); void flex_array_free(struct flex_array *fa); void flex_array_free_parts(struct flex_array *fa); -int flex_array_put(struct flex_array *fa, int element_nr, void *src, +int flex_array_put(struct flex_array *fa, unsigned int element_nr, void *src, gfp_t flags); -void *flex_array_get(struct flex_array *fa, int element_nr); +void *flex_array_get(struct flex_array *fa, unsigned int element_nr); #endif /* _FLEX_ARRAY_H */ diff --git a/lib/flex_array.c b/lib/flex_array.c index cf4e372cae90..7baed2fc3bc8 100644 --- a/lib/flex_array.c +++ b/lib/flex_array.c @@ -99,7 +99,8 @@ static inline int elements_fit_in_base(struct flex_array *fa) * capacity in the base structure. Also note that no effort is made * to efficiently pack objects across page boundaries. */ -struct flex_array *flex_array_alloc(int element_size, int total, gfp_t flags) +struct flex_array *flex_array_alloc(int element_size, unsigned int total, + gfp_t flags) { struct flex_array *ret; int max_size = nr_base_part_ptrs() * __elements_per_part(element_size); @@ -115,7 +116,8 @@ struct flex_array *flex_array_alloc(int element_size, int total, gfp_t flags) return ret; } -static int fa_element_to_part_nr(struct flex_array *fa, int element_nr) +static int fa_element_to_part_nr(struct flex_array *fa, + unsigned int element_nr) { return element_nr / __elements_per_part(fa->element_size); } @@ -143,14 +145,12 @@ void flex_array_free(struct flex_array *fa) kfree(fa); } -static int fa_index_inside_part(struct flex_array *fa, int element_nr) +static unsigned int index_inside_part(struct flex_array *fa, + unsigned int element_nr) { - return element_nr % __elements_per_part(fa->element_size); -} + unsigned int part_offset; -static int index_inside_part(struct flex_array *fa, int element_nr) -{ - int part_offset = fa_index_inside_part(fa, element_nr); + part_offset = element_nr % __elements_per_part(fa->element_size); return part_offset * fa->element_size; } @@ -185,7 +185,8 @@ __fa_get_part(struct flex_array *fa, int part_nr, gfp_t flags) * * Locking must be provided by the caller. */ -int flex_array_put(struct flex_array *fa, int element_nr, void *src, gfp_t flags) +int flex_array_put(struct flex_array *fa, unsigned int element_nr, void *src, + gfp_t flags) { int part_nr = fa_element_to_part_nr(fa, element_nr); struct flex_array_part *part; @@ -217,7 +218,8 @@ int flex_array_put(struct flex_array *fa, int element_nr, void *src, gfp_t flags * * Locking must be provided by the caller. */ -int flex_array_prealloc(struct flex_array *fa, int start, int end, gfp_t flags) +int flex_array_prealloc(struct flex_array *fa, unsigned int start, + unsigned int end, gfp_t flags) { int start_part; int end_part; @@ -248,7 +250,7 @@ int flex_array_prealloc(struct flex_array *fa, int start, int end, gfp_t flags) * * Locking must be provided by the caller. */ -void *flex_array_get(struct flex_array *fa, int element_nr) +void *flex_array_get(struct flex_array *fa, unsigned int element_nr) { int part_nr = fa_element_to_part_nr(fa, element_nr); struct flex_array_part *part; -- cgit v1.2.3 From 2a908002c7b1b666616103e9df2419b38d7c6f1f Mon Sep 17 00:00:00 2001 From: Frans Pop Date: Wed, 26 Aug 2009 14:29:29 -0700 Subject: ACPI processor: force throttling state when BIOS returns incorrect value If the BIOS reports an invalid throttling state (which seems to be fairly common after system boot), a reset is done to state T0. Because of a check in acpi_processor_get_throttling_ptc(), the reset never actually gets executed, which results in the error reoccurring on every access of for example /proc/acpi/processor/CPU0/throttling. Add a 'force' option to acpi_processor_set_throttling() to ensure the reset really takes effect. Addresses http://bugzilla.kernel.org/show_bug.cgi?id=13389 This patch, together with the next one, fixes a regression introduced in 2.6.30, listed on the regression list. They have been available for 2.5 months now in bugzilla, but have not been picked up, despite various reminders and without any reason given. Google shows that numerous people are hitting this issue. The issue is in itself relatively minor, but the bug in the code is clear. The patches have been in all my kernels and today testing has shown that throttling works correctly with the patches applied when the system overheats (http://bugzilla.kernel.org/show_bug.cgi?id=13918#c14). Signed-off-by: Frans Pop Acked-by: Zhang Rui Cc: Len Brown Cc: "Rafael J. Wysocki" Cc: Rusty Russell Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/acpi/processor_thermal.c | 6 +++--- drivers/acpi/processor_throttling.c | 26 ++++++++++++++------------ include/acpi/processor.h | 5 +++-- 3 files changed, 20 insertions(+), 17 deletions(-) (limited to 'include') diff --git a/drivers/acpi/processor_thermal.c b/drivers/acpi/processor_thermal.c index 39838c666032..31adda1099e0 100644 --- a/drivers/acpi/processor_thermal.c +++ b/drivers/acpi/processor_thermal.c @@ -66,7 +66,7 @@ static int acpi_processor_apply_limit(struct acpi_processor *pr) if (pr->limit.thermal.tx > tx) tx = pr->limit.thermal.tx; - result = acpi_processor_set_throttling(pr, tx); + result = acpi_processor_set_throttling(pr, tx, false); if (result) goto end; } @@ -421,12 +421,12 @@ processor_set_cur_state(struct thermal_cooling_device *cdev, if (state <= max_pstate) { if (pr->flags.throttling && pr->throttling.state) - result = acpi_processor_set_throttling(pr, 0); + result = acpi_processor_set_throttling(pr, 0, false); cpufreq_set_cur_state(pr->id, state); } else { cpufreq_set_cur_state(pr->id, max_pstate); result = acpi_processor_set_throttling(pr, - state - max_pstate); + state - max_pstate, false); } return result; } diff --git a/drivers/acpi/processor_throttling.c b/drivers/acpi/processor_throttling.c index 227543789ba9..841be4ee2109 100644 --- a/drivers/acpi/processor_throttling.c +++ b/drivers/acpi/processor_throttling.c @@ -62,7 +62,8 @@ struct throttling_tstate { #define THROTTLING_POSTCHANGE (2) static int acpi_processor_get_throttling(struct acpi_processor *pr); -int acpi_processor_set_throttling(struct acpi_processor *pr, int state); +int acpi_processor_set_throttling(struct acpi_processor *pr, + int state, bool force); static int acpi_processor_update_tsd_coord(void) { @@ -361,7 +362,7 @@ int acpi_processor_tstate_has_changed(struct acpi_processor *pr) */ target_state = throttling_limit; } - return acpi_processor_set_throttling(pr, target_state); + return acpi_processor_set_throttling(pr, target_state, false); } /* @@ -842,7 +843,7 @@ static int acpi_processor_get_throttling_ptc(struct acpi_processor *pr) ACPI_WARNING((AE_INFO, "Invalid throttling state, reset")); state = 0; - ret = acpi_processor_set_throttling(pr, state); + ret = acpi_processor_set_throttling(pr, state, true); if (ret) return ret; } @@ -915,7 +916,7 @@ static int acpi_processor_get_fadt_info(struct acpi_processor *pr) } static int acpi_processor_set_throttling_fadt(struct acpi_processor *pr, - int state) + int state, bool force) { u32 value = 0; u32 duty_mask = 0; @@ -930,7 +931,7 @@ static int acpi_processor_set_throttling_fadt(struct acpi_processor *pr, if (!pr->flags.throttling) return -ENODEV; - if (state == pr->throttling.state) + if (!force && (state == pr->throttling.state)) return 0; if (state < pr->throttling_platform_limit) @@ -988,7 +989,7 @@ static int acpi_processor_set_throttling_fadt(struct acpi_processor *pr, } static int acpi_processor_set_throttling_ptc(struct acpi_processor *pr, - int state) + int state, bool force) { int ret; acpi_integer value; @@ -1002,7 +1003,7 @@ static int acpi_processor_set_throttling_ptc(struct acpi_processor *pr, if (!pr->flags.throttling) return -ENODEV; - if (state == pr->throttling.state) + if (!force && (state == pr->throttling.state)) return 0; if (state < pr->throttling_platform_limit) @@ -1018,7 +1019,8 @@ static int acpi_processor_set_throttling_ptc(struct acpi_processor *pr, return 0; } -int acpi_processor_set_throttling(struct acpi_processor *pr, int state) +int acpi_processor_set_throttling(struct acpi_processor *pr, + int state, bool force) { cpumask_var_t saved_mask; int ret = 0; @@ -1070,7 +1072,7 @@ int acpi_processor_set_throttling(struct acpi_processor *pr, int state) /* FIXME: use work_on_cpu() */ set_cpus_allowed_ptr(current, cpumask_of(pr->id)); ret = p_throttling->acpi_processor_set_throttling(pr, - t_state.target_state); + t_state.target_state, force); } else { /* * When the T-state coordination is SW_ALL or HW_ALL, @@ -1103,7 +1105,7 @@ int acpi_processor_set_throttling(struct acpi_processor *pr, int state) set_cpus_allowed_ptr(current, cpumask_of(i)); ret = match_pr->throttling. acpi_processor_set_throttling( - match_pr, t_state.target_state); + match_pr, t_state.target_state, force); } } /* @@ -1201,7 +1203,7 @@ int acpi_processor_get_throttling_info(struct acpi_processor *pr) ACPI_DEBUG_PRINT((ACPI_DB_INFO, "Disabling throttling (was T%d)\n", pr->throttling.state)); - result = acpi_processor_set_throttling(pr, 0); + result = acpi_processor_set_throttling(pr, 0, false); if (result) goto end; } @@ -1307,7 +1309,7 @@ static ssize_t acpi_processor_write_throttling(struct file *file, if (strcmp(tmpbuf, charp) != 0) return -EINVAL; - result = acpi_processor_set_throttling(pr, state_val); + result = acpi_processor_set_throttling(pr, state_val, false); if (result) return result; diff --git a/include/acpi/processor.h b/include/acpi/processor.h index baf1e0a9a7ee..740ac3ad8fd0 100644 --- a/include/acpi/processor.h +++ b/include/acpi/processor.h @@ -174,7 +174,7 @@ struct acpi_processor_throttling { cpumask_var_t shared_cpu_map; int (*acpi_processor_get_throttling) (struct acpi_processor * pr); int (*acpi_processor_set_throttling) (struct acpi_processor * pr, - int state); + int state, bool force); u32 address; u8 duty_offset; @@ -321,7 +321,8 @@ static inline int acpi_processor_ppc_has_changed(struct acpi_processor *pr) /* in processor_throttling.c */ int acpi_processor_tstate_has_changed(struct acpi_processor *pr); int acpi_processor_get_throttling_info(struct acpi_processor *pr); -extern int acpi_processor_set_throttling(struct acpi_processor *pr, int state); +extern int acpi_processor_set_throttling(struct acpi_processor *pr, + int state, bool force); extern const struct file_operations acpi_processor_throttling_fops; extern void acpi_processor_throttling_init(void); /* in processor_idle.c */ -- cgit v1.2.3 From a6186d89c913b176e7339f37a4ec6ccb38b2c5c0 Mon Sep 17 00:00:00 2001 From: Catalin Marinas Date: Thu, 27 Aug 2009 14:29:16 +0100 Subject: kmemleak: Mark the early log buffer as __initdata This buffer isn't needed after kmemleak was initialised so it can be freed together with the .init.data section. This patch also marks functions conditionally accessing the early log variables with __ref. Signed-off-by: Catalin Marinas --- include/linux/kmemleak.h | 18 +++++++++--------- mm/kmemleak.c | 26 ++++++++++++++------------ 2 files changed, 23 insertions(+), 21 deletions(-) (limited to 'include') diff --git a/include/linux/kmemleak.h b/include/linux/kmemleak.h index 6a63807f714e..3c7497d46ee9 100644 --- a/include/linux/kmemleak.h +++ b/include/linux/kmemleak.h @@ -23,18 +23,18 @@ #ifdef CONFIG_DEBUG_KMEMLEAK -extern void kmemleak_init(void); +extern void kmemleak_init(void) __ref; extern void kmemleak_alloc(const void *ptr, size_t size, int min_count, - gfp_t gfp); -extern void kmemleak_free(const void *ptr); -extern void kmemleak_free_part(const void *ptr, size_t size); + gfp_t gfp) __ref; +extern void kmemleak_free(const void *ptr) __ref; +extern void kmemleak_free_part(const void *ptr, size_t size) __ref; extern void kmemleak_padding(const void *ptr, unsigned long offset, - size_t size); -extern void kmemleak_not_leak(const void *ptr); -extern void kmemleak_ignore(const void *ptr); + size_t size) __ref; +extern void kmemleak_not_leak(const void *ptr) __ref; +extern void kmemleak_ignore(const void *ptr) __ref; extern void kmemleak_scan_area(const void *ptr, unsigned long offset, - size_t length, gfp_t gfp); -extern void kmemleak_no_scan(const void *ptr); + size_t length, gfp_t gfp) __ref; +extern void kmemleak_no_scan(const void *ptr) __ref; static inline void kmemleak_alloc_recursive(const void *ptr, size_t size, int min_count, unsigned long flags, diff --git a/mm/kmemleak.c b/mm/kmemleak.c index c977f7a2f0e4..576c0a4cec52 100644 --- a/mm/kmemleak.c +++ b/mm/kmemleak.c @@ -232,8 +232,9 @@ struct early_log { }; /* early logging buffer and current position */ -static struct early_log early_log[CONFIG_DEBUG_KMEMLEAK_EARLY_LOG_SIZE]; -static int crt_early_log; +static struct early_log + early_log[CONFIG_DEBUG_KMEMLEAK_EARLY_LOG_SIZE] __initdata; +static int crt_early_log __initdata; static void kmemleak_disable(void); @@ -718,8 +719,8 @@ static void object_no_scan(unsigned long ptr) * Log an early kmemleak_* call to the early_log buffer. These calls will be * processed later once kmemleak is fully initialized. */ -static void log_early(int op_type, const void *ptr, size_t size, - int min_count, unsigned long offset, size_t length) +static void __init log_early(int op_type, const void *ptr, size_t size, + int min_count, unsigned long offset, size_t length) { unsigned long flags; struct early_log *log; @@ -751,7 +752,8 @@ static void log_early(int op_type, const void *ptr, size_t size, * kernel allocators when a new block is allocated (kmem_cache_alloc, kmalloc, * vmalloc etc.). */ -void kmemleak_alloc(const void *ptr, size_t size, int min_count, gfp_t gfp) +void __ref kmemleak_alloc(const void *ptr, size_t size, int min_count, + gfp_t gfp) { pr_debug("%s(0x%p, %zu, %d)\n", __func__, ptr, size, min_count); @@ -766,7 +768,7 @@ EXPORT_SYMBOL_GPL(kmemleak_alloc); * Memory freeing function callback. This function is called from the kernel * allocators when a block is freed (kmem_cache_free, kfree, vfree etc.). */ -void kmemleak_free(const void *ptr) +void __ref kmemleak_free(const void *ptr) { pr_debug("%s(0x%p)\n", __func__, ptr); @@ -781,7 +783,7 @@ EXPORT_SYMBOL_GPL(kmemleak_free); * Partial memory freeing function callback. This function is usually called * from bootmem allocator when (part of) a memory block is freed. */ -void kmemleak_free_part(const void *ptr, size_t size) +void __ref kmemleak_free_part(const void *ptr, size_t size) { pr_debug("%s(0x%p)\n", __func__, ptr); @@ -796,7 +798,7 @@ EXPORT_SYMBOL_GPL(kmemleak_free_part); * Mark an already allocated memory block as a false positive. This will cause * the block to no longer be reported as leak and always be scanned. */ -void kmemleak_not_leak(const void *ptr) +void __ref kmemleak_not_leak(const void *ptr) { pr_debug("%s(0x%p)\n", __func__, ptr); @@ -812,7 +814,7 @@ EXPORT_SYMBOL(kmemleak_not_leak); * corresponding block is not a leak and does not contain any references to * other allocated memory blocks. */ -void kmemleak_ignore(const void *ptr) +void __ref kmemleak_ignore(const void *ptr) { pr_debug("%s(0x%p)\n", __func__, ptr); @@ -826,8 +828,8 @@ EXPORT_SYMBOL(kmemleak_ignore); /* * Limit the range to be scanned in an allocated memory block. */ -void kmemleak_scan_area(const void *ptr, unsigned long offset, size_t length, - gfp_t gfp) +void __ref kmemleak_scan_area(const void *ptr, unsigned long offset, + size_t length, gfp_t gfp) { pr_debug("%s(0x%p)\n", __func__, ptr); @@ -841,7 +843,7 @@ EXPORT_SYMBOL(kmemleak_scan_area); /* * Inform kmemleak not to scan the given memory block. */ -void kmemleak_no_scan(const void *ptr) +void __ref kmemleak_no_scan(const void *ptr) { pr_debug("%s(0x%p)\n", __func__, ptr); -- cgit v1.2.3 From f8d80cdf40fe4d2393159012b38ce9f85a488686 Mon Sep 17 00:00:00 2001 From: Bob Moore Date: Tue, 2 Jun 2009 13:28:13 +0800 Subject: ACPICA: Remove duplicate extern declarations for public globals Some were defined twice, causes a warning with gcc -Wredundant-decls. Signed-off-by: Bob Moore Signed-off-by: Lin Ming Signed-off-by: Len Brown --- drivers/acpi/acpica/acglobal.h | 33 ++++++++++++++++++--------------- include/acpi/acpixf.h | 1 + 2 files changed, 19 insertions(+), 15 deletions(-) (limited to 'include') diff --git a/drivers/acpi/acpica/acglobal.h b/drivers/acpi/acpica/acglobal.h index 3d87362d17ed..0b73b31c1b53 100644 --- a/drivers/acpi/acpica/acglobal.h +++ b/drivers/acpi/acpica/acglobal.h @@ -58,6 +58,10 @@ #define ACPI_INIT_GLOBAL(a,b) a #endif +#ifdef DEFINE_ACPI_GLOBALS + +/* Public globals, available from outside ACPICA subsystem */ + /***************************************************************************** * * Runtime configuration (static defaults that can be overriden at runtime) @@ -78,7 +82,7 @@ * 5) Allow unresolved references (invalid target name) in package objects * 6) Enable warning messages for behavior that is not ACPI spec compliant */ -ACPI_EXTERN u8 ACPI_INIT_GLOBAL(acpi_gbl_enable_interpreter_slack, FALSE); +u8 ACPI_INIT_GLOBAL(acpi_gbl_enable_interpreter_slack, FALSE); /* * Automatically serialize ALL control methods? Default is FALSE, meaning @@ -86,27 +90,36 @@ ACPI_EXTERN u8 ACPI_INIT_GLOBAL(acpi_gbl_enable_interpreter_slack, FALSE); * Only change this if the ASL code is poorly written and cannot handle * reentrancy even though methods are marked "NotSerialized". */ -ACPI_EXTERN u8 ACPI_INIT_GLOBAL(acpi_gbl_all_methods_serialized, FALSE); +u8 ACPI_INIT_GLOBAL(acpi_gbl_all_methods_serialized, FALSE); /* * Create the predefined _OSI method in the namespace? Default is TRUE * because ACPI CA is fully compatible with other ACPI implementations. * Changing this will revert ACPI CA (and machine ASL) to pre-OSI behavior. */ -ACPI_EXTERN u8 ACPI_INIT_GLOBAL(acpi_gbl_create_osi_method, TRUE); +u8 ACPI_INIT_GLOBAL(acpi_gbl_create_osi_method, TRUE); /* * Disable wakeup GPEs during runtime? Default is TRUE because WAKE and * RUNTIME GPEs should never be shared, and WAKE GPEs should typically only * be enabled just before going to sleep. */ -ACPI_EXTERN u8 ACPI_INIT_GLOBAL(acpi_gbl_leave_wake_gpes_disabled, TRUE); +u8 ACPI_INIT_GLOBAL(acpi_gbl_leave_wake_gpes_disabled, TRUE); /* * Optionally use default values for the ACPI register widths. Set this to * TRUE to use the defaults, if an FADT contains incorrect widths/lengths. */ -ACPI_EXTERN u8 ACPI_INIT_GLOBAL(acpi_gbl_use_default_register_widths, TRUE); +u8 ACPI_INIT_GLOBAL(acpi_gbl_use_default_register_widths, TRUE); + +/* acpi_gbl_FADT is a local copy of the FADT, converted to a common format. */ + +struct acpi_table_fadt acpi_gbl_FADT; +u32 acpi_current_gpe_count; +u32 acpi_gbl_trace_flags; +acpi_name acpi_gbl_trace_method_name; + +#endif /***************************************************************************** * @@ -114,11 +127,6 @@ ACPI_EXTERN u8 ACPI_INIT_GLOBAL(acpi_gbl_use_default_register_widths, TRUE); * ****************************************************************************/ -/* Runtime configuration of debug print levels */ - -extern u32 acpi_dbg_level; -extern u32 acpi_dbg_layer; - /* Procedure nesting level for debug output */ extern u32 acpi_gbl_nesting_level; @@ -127,10 +135,8 @@ extern u32 acpi_gbl_nesting_level; ACPI_EXTERN u32 acpi_gbl_original_dbg_level; ACPI_EXTERN u32 acpi_gbl_original_dbg_layer; -ACPI_EXTERN acpi_name acpi_gbl_trace_method_name; ACPI_EXTERN u32 acpi_gbl_trace_dbg_level; ACPI_EXTERN u32 acpi_gbl_trace_dbg_layer; -ACPI_EXTERN u32 acpi_gbl_trace_flags; /***************************************************************************** * @@ -142,10 +148,8 @@ ACPI_EXTERN u32 acpi_gbl_trace_flags; * acpi_gbl_root_table_list is the master list of ACPI tables found in the * RSDT/XSDT. * - * acpi_gbl_FADT is a local copy of the FADT, converted to a common format. */ ACPI_EXTERN struct acpi_internal_rsdt acpi_gbl_root_table_list; -ACPI_EXTERN struct acpi_table_fadt acpi_gbl_FADT; ACPI_EXTERN struct acpi_table_facs *acpi_gbl_FACS; /* These addresses are calculated from the FADT Event Block addresses */ @@ -340,7 +344,6 @@ ACPI_EXTERN struct acpi_fixed_event_handler ACPI_EXTERN struct acpi_gpe_xrupt_info *acpi_gbl_gpe_xrupt_list_head; ACPI_EXTERN struct acpi_gpe_block_info *acpi_gbl_gpe_fadt_blocks[ACPI_MAX_GPE_BLOCKS]; -ACPI_EXTERN u32 acpi_current_gpe_count; /***************************************************************************** * diff --git a/include/acpi/acpixf.h b/include/acpi/acpixf.h index 82ec6a3c0500..2aecaa5cc06c 100644 --- a/include/acpi/acpixf.h +++ b/include/acpi/acpixf.h @@ -64,6 +64,7 @@ extern u8 acpi_gbl_enable_interpreter_slack; extern u8 acpi_gbl_all_methods_serialized; extern u8 acpi_gbl_create_osi_method; extern u8 acpi_gbl_leave_wake_gpes_disabled; +extern u8 acpi_gbl_use_default_register_widths; extern acpi_name acpi_gbl_trace_method_name; extern u32 acpi_gbl_trace_flags; -- cgit v1.2.3 From c6b5774caafa4c12b6019366e2fdaaff117e95a4 Mon Sep 17 00:00:00 2001 From: Bob Moore Date: Wed, 24 Jun 2009 09:44:06 +0800 Subject: ACPICA: Add 64-bit support to acpi_read and acpi_write Needed by drivers for new ACPi tables. Internal versions of these functions still use 32-bit max transfers, in order to minimize disruption and stack use for the standard ACPI registers (FADT-based). Signed-off-by: Bob Moore Signed-off-by: Lin Ming Signed-off-by: Len Brown --- drivers/acpi/acpica/achware.h | 8 ++ drivers/acpi/acpica/evgpe.c | 8 +- drivers/acpi/acpica/evgpeblk.c | 4 +- drivers/acpi/acpica/hwgpe.c | 34 +++---- drivers/acpi/acpica/hwregs.c | 206 ++++++++++++++++++++++++++++++++++++++--- drivers/acpi/acpica/hwtimer.c | 2 +- drivers/acpi/acpica/hwxface.c | 166 +++++++++++++++++++-------------- include/acpi/acpixf.h | 4 +- 8 files changed, 327 insertions(+), 105 deletions(-) (limited to 'include') diff --git a/drivers/acpi/acpica/achware.h b/drivers/acpi/acpica/achware.h index 4afa3d8e0efb..36192f142fbb 100644 --- a/drivers/acpi/acpica/achware.h +++ b/drivers/acpi/acpica/achware.h @@ -62,6 +62,14 @@ u32 acpi_hw_get_mode(void); /* * hwregs - ACPI Register I/O */ +acpi_status +acpi_hw_validate_register(struct acpi_generic_address *reg, + u8 max_bit_width, u64 *address); + +acpi_status acpi_hw_read(u32 *value, struct acpi_generic_address *reg); + +acpi_status acpi_hw_write(u32 value, struct acpi_generic_address *reg); + struct acpi_bit_register_info *acpi_hw_get_bit_register_info(u32 register_id); acpi_status acpi_hw_write_pm1_control(u32 pm1a_control, u32 pm1b_control); diff --git a/drivers/acpi/acpica/evgpe.c b/drivers/acpi/acpica/evgpe.c index b9d8ee69ca6c..afacf4416c73 100644 --- a/drivers/acpi/acpica/evgpe.c +++ b/drivers/acpi/acpica/evgpe.c @@ -424,8 +424,8 @@ u32 acpi_ev_gpe_detect(struct acpi_gpe_xrupt_info * gpe_xrupt_list) /* Read the Status Register */ status = - acpi_read(&status_reg, - &gpe_register_info->status_address); + acpi_hw_read(&status_reg, + &gpe_register_info->status_address); if (ACPI_FAILURE(status)) { goto unlock_and_exit; } @@ -433,8 +433,8 @@ u32 acpi_ev_gpe_detect(struct acpi_gpe_xrupt_info * gpe_xrupt_list) /* Read the Enable Register */ status = - acpi_read(&enable_reg, - &gpe_register_info->enable_address); + acpi_hw_read(&enable_reg, + &gpe_register_info->enable_address); if (ACPI_FAILURE(status)) { goto unlock_and_exit; } diff --git a/drivers/acpi/acpica/evgpeblk.c b/drivers/acpi/acpica/evgpeblk.c index 7b3463639422..a60aaa7635f3 100644 --- a/drivers/acpi/acpica/evgpeblk.c +++ b/drivers/acpi/acpica/evgpeblk.c @@ -843,14 +843,14 @@ acpi_ev_create_gpe_info_blocks(struct acpi_gpe_block_info *gpe_block) /* Disable all GPEs within this register */ - status = acpi_write(0x00, &this_register->enable_address); + status = acpi_hw_write(0x00, &this_register->enable_address); if (ACPI_FAILURE(status)) { goto error_exit; } /* Clear any pending GPE events within this register */ - status = acpi_write(0xFF, &this_register->status_address); + status = acpi_hw_write(0xFF, &this_register->status_address); if (ACPI_FAILURE(status)) { goto error_exit; } diff --git a/drivers/acpi/acpica/hwgpe.c b/drivers/acpi/acpica/hwgpe.c index d3b7e37c9eed..c28c41b3180b 100644 --- a/drivers/acpi/acpica/hwgpe.c +++ b/drivers/acpi/acpica/hwgpe.c @@ -82,7 +82,7 @@ acpi_status acpi_hw_low_disable_gpe(struct acpi_gpe_event_info *gpe_event_info) /* Get current value of the enable register that contains this GPE */ - status = acpi_read(&enable_mask, &gpe_register_info->enable_address); + status = acpi_hw_read(&enable_mask, &gpe_register_info->enable_address); if (ACPI_FAILURE(status)) { return (status); } @@ -95,7 +95,7 @@ acpi_status acpi_hw_low_disable_gpe(struct acpi_gpe_event_info *gpe_event_info) /* Write the updated enable mask */ - status = acpi_write(enable_mask, &gpe_register_info->enable_address); + status = acpi_hw_write(enable_mask, &gpe_register_info->enable_address); return (status); } @@ -130,8 +130,8 @@ acpi_hw_write_gpe_enable_reg(struct acpi_gpe_event_info * gpe_event_info) /* Write the entire GPE (runtime) enable register */ - status = acpi_write(gpe_register_info->enable_for_run, - &gpe_register_info->enable_address); + status = acpi_hw_write(gpe_register_info->enable_for_run, + &gpe_register_info->enable_address); return (status); } @@ -163,8 +163,8 @@ acpi_status acpi_hw_clear_gpe(struct acpi_gpe_event_info * gpe_event_info) * Write a one to the appropriate bit in the status register to * clear this GPE. */ - status = acpi_write(register_bit, - &gpe_event_info->register_info->status_address); + status = acpi_hw_write(register_bit, + &gpe_event_info->register_info->status_address); return (status); } @@ -222,7 +222,7 @@ acpi_hw_get_gpe_status(struct acpi_gpe_event_info * gpe_event_info, /* GPE currently active (status bit == 1)? */ - status = acpi_read(&in_byte, &gpe_register_info->status_address); + status = acpi_hw_read(&in_byte, &gpe_register_info->status_address); if (ACPI_FAILURE(status)) { goto unlock_and_exit; } @@ -266,8 +266,8 @@ acpi_hw_disable_gpe_block(struct acpi_gpe_xrupt_info *gpe_xrupt_info, /* Disable all GPEs in this register */ status = - acpi_write(0x00, - &gpe_block->register_info[i].enable_address); + acpi_hw_write(0x00, + &gpe_block->register_info[i].enable_address); if (ACPI_FAILURE(status)) { return (status); } @@ -303,8 +303,8 @@ acpi_hw_clear_gpe_block(struct acpi_gpe_xrupt_info *gpe_xrupt_info, /* Clear status on all GPEs in this register */ status = - acpi_write(0xFF, - &gpe_block->register_info[i].status_address); + acpi_hw_write(0xFF, + &gpe_block->register_info[i].status_address); if (ACPI_FAILURE(status)) { return (status); } @@ -345,9 +345,9 @@ acpi_hw_enable_runtime_gpe_block(struct acpi_gpe_xrupt_info *gpe_xrupt_info, /* Enable all "runtime" GPEs in this register */ - status = acpi_write(gpe_block->register_info[i].enable_for_run, - &gpe_block->register_info[i]. - enable_address); + status = + acpi_hw_write(gpe_block->register_info[i].enable_for_run, + &gpe_block->register_info[i].enable_address); if (ACPI_FAILURE(status)) { return (status); } @@ -387,9 +387,9 @@ acpi_hw_enable_wakeup_gpe_block(struct acpi_gpe_xrupt_info *gpe_xrupt_info, /* Enable all "wake" GPEs in this register */ - status = acpi_write(gpe_block->register_info[i].enable_for_wake, - &gpe_block->register_info[i]. - enable_address); + status = + acpi_hw_write(gpe_block->register_info[i].enable_for_wake, + &gpe_block->register_info[i].enable_address); if (ACPI_FAILURE(status)) { return (status); } diff --git a/drivers/acpi/acpica/hwregs.c b/drivers/acpi/acpica/hwregs.c index 23d5505cb1f7..15c9ed2be853 100644 --- a/drivers/acpi/acpica/hwregs.c +++ b/drivers/acpi/acpica/hwregs.c @@ -62,6 +62,184 @@ acpi_hw_write_multiple(u32 value, struct acpi_generic_address *register_a, struct acpi_generic_address *register_b); +/****************************************************************************** + * + * FUNCTION: acpi_hw_validate_register + * + * PARAMETERS: Reg - GAS register structure + * max_bit_width - Max bit_width supported (32 or 64) + * Address - Pointer to where the gas->address + * is returned + * + * RETURN: Status + * + * DESCRIPTION: Validate the contents of a GAS register. Checks the GAS + * pointer, Address, space_id, bit_width, and bit_offset. + * + ******************************************************************************/ + +acpi_status +acpi_hw_validate_register(struct acpi_generic_address *reg, + u8 max_bit_width, u64 *address) +{ + + /* Must have a valid pointer to a GAS structure */ + + if (!reg) { + return (AE_BAD_PARAMETER); + } + + /* + * Copy the target address. This handles possible alignment issues. + * Address must not be null. A null address also indicates an optional + * ACPI register that is not supported, so no error message. + */ + ACPI_MOVE_64_TO_64(address, ®->address); + if (!(*address)) { + return (AE_BAD_ADDRESS); + } + + /* Validate the space_iD */ + + if ((reg->space_id != ACPI_ADR_SPACE_SYSTEM_MEMORY) && + (reg->space_id != ACPI_ADR_SPACE_SYSTEM_IO)) { + ACPI_ERROR((AE_INFO, + "Unsupported address space: 0x%X", reg->space_id)); + return (AE_SUPPORT); + } + + /* Validate the bit_width */ + + if ((reg->bit_width != 8) && + (reg->bit_width != 16) && + (reg->bit_width != 32) && (reg->bit_width != max_bit_width)) { + ACPI_ERROR((AE_INFO, + "Unsupported register bit width: 0x%X", + reg->bit_width)); + return (AE_SUPPORT); + } + + /* Validate the bit_offset. Just a warning for now. */ + + if (reg->bit_offset != 0) { + ACPI_WARNING((AE_INFO, + "Unsupported register bit offset: 0x%X", + reg->bit_offset)); + } + + return (AE_OK); +} + +/****************************************************************************** + * + * FUNCTION: acpi_hw_read + * + * PARAMETERS: Value - Where the value is returned + * Reg - GAS register structure + * + * RETURN: Status + * + * DESCRIPTION: Read from either memory or IO space. This is a 32-bit max + * version of acpi_read, used internally since the overhead of + * 64-bit values is not needed. + * + * LIMITATIONS: + * bit_width must be exactly 8, 16, or 32. + * space_iD must be system_memory or system_iO. + * bit_offset and access_width are currently ignored, as there has + * not been a need to implement these. + * + ******************************************************************************/ + +acpi_status acpi_hw_read(u32 *value, struct acpi_generic_address *reg) +{ + u64 address; + acpi_status status; + + ACPI_FUNCTION_NAME(hw_read); + + /* Validate contents of the GAS register */ + + status = acpi_hw_validate_register(reg, 32, &address); + if (ACPI_FAILURE(status)) { + return (status); + } + + /* Initialize entire 32-bit return value to zero */ + + *value = 0; + + /* + * Two address spaces supported: Memory or IO. PCI_Config is + * not supported here because the GAS structure is insufficient + */ + if (reg->space_id == ACPI_ADR_SPACE_SYSTEM_MEMORY) { + status = acpi_os_read_memory((acpi_physical_address) + address, value, reg->bit_width); + } else { /* ACPI_ADR_SPACE_SYSTEM_IO, validated earlier */ + + status = acpi_hw_read_port((acpi_io_address) + address, value, reg->bit_width); + } + + ACPI_DEBUG_PRINT((ACPI_DB_IO, + "Read: %8.8X width %2d from %8.8X%8.8X (%s)\n", + *value, reg->bit_width, ACPI_FORMAT_UINT64(address), + acpi_ut_get_region_name(reg->space_id))); + + return (status); +} + +/****************************************************************************** + * + * FUNCTION: acpi_hw_write + * + * PARAMETERS: Value - Value to be written + * Reg - GAS register structure + * + * RETURN: Status + * + * DESCRIPTION: Write to either memory or IO space. This is a 32-bit max + * version of acpi_write, used internally since the overhead of + * 64-bit values is not needed. + * + ******************************************************************************/ + +acpi_status acpi_hw_write(u32 value, struct acpi_generic_address *reg) +{ + u64 address; + acpi_status status; + + ACPI_FUNCTION_NAME(hw_write); + + /* Validate contents of the GAS register */ + + status = acpi_hw_validate_register(reg, 32, &address); + if (ACPI_FAILURE(status)) { + return (status); + } + + /* + * Two address spaces supported: Memory or IO. PCI_Config is + * not supported here because the GAS structure is insufficient + */ + if (reg->space_id == ACPI_ADR_SPACE_SYSTEM_MEMORY) { + status = acpi_os_write_memory((acpi_physical_address) + address, value, reg->bit_width); + } else { /* ACPI_ADR_SPACE_SYSTEM_IO, validated earlier */ + + status = acpi_hw_write_port((acpi_io_address) + address, value, reg->bit_width); + } + + ACPI_DEBUG_PRINT((ACPI_DB_IO, + "Wrote: %8.8X width %2d to %8.8X%8.8X (%s)\n", + value, reg->bit_width, ACPI_FORMAT_UINT64(address), + acpi_ut_get_region_name(reg->space_id))); + + return (status); +} + /******************************************************************************* * * FUNCTION: acpi_hw_clear_acpi_status @@ -152,15 +330,16 @@ acpi_status acpi_hw_write_pm1_control(u32 pm1a_control, u32 pm1b_control) ACPI_FUNCTION_TRACE(hw_write_pm1_control); - status = acpi_write(pm1a_control, &acpi_gbl_FADT.xpm1a_control_block); + status = + acpi_hw_write(pm1a_control, &acpi_gbl_FADT.xpm1a_control_block); if (ACPI_FAILURE(status)) { return_ACPI_STATUS(status); } if (acpi_gbl_FADT.xpm1b_control_block.address) { status = - acpi_write(pm1b_control, - &acpi_gbl_FADT.xpm1b_control_block); + acpi_hw_write(pm1b_control, + &acpi_gbl_FADT.xpm1b_control_block); } return_ACPI_STATUS(status); } @@ -218,12 +397,13 @@ acpi_hw_register_read(u32 register_id, u32 * return_value) case ACPI_REGISTER_PM2_CONTROL: /* 8-bit access */ - status = acpi_read(&value, &acpi_gbl_FADT.xpm2_control_block); + status = + acpi_hw_read(&value, &acpi_gbl_FADT.xpm2_control_block); break; case ACPI_REGISTER_PM_TIMER: /* 32-bit access */ - status = acpi_read(&value, &acpi_gbl_FADT.xpm_timer_block); + status = acpi_hw_read(&value, &acpi_gbl_FADT.xpm_timer_block); break; case ACPI_REGISTER_SMI_COMMAND_BLOCK: /* 8-bit access */ @@ -340,7 +520,8 @@ acpi_status acpi_hw_register_write(u32 register_id, u32 value) * as per the ACPI spec. */ status = - acpi_read(&read_value, &acpi_gbl_FADT.xpm2_control_block); + acpi_hw_read(&read_value, + &acpi_gbl_FADT.xpm2_control_block); if (ACPI_FAILURE(status)) { goto exit; } @@ -350,12 +531,13 @@ acpi_status acpi_hw_register_write(u32 register_id, u32 value) ACPI_INSERT_BITS(value, ACPI_PM2_CONTROL_PRESERVED_BITS, read_value); - status = acpi_write(value, &acpi_gbl_FADT.xpm2_control_block); + status = + acpi_hw_write(value, &acpi_gbl_FADT.xpm2_control_block); break; case ACPI_REGISTER_PM_TIMER: /* 32-bit access */ - status = acpi_write(value, &acpi_gbl_FADT.xpm_timer_block); + status = acpi_hw_write(value, &acpi_gbl_FADT.xpm_timer_block); break; case ACPI_REGISTER_SMI_COMMAND_BLOCK: /* 8-bit access */ @@ -401,7 +583,7 @@ acpi_hw_read_multiple(u32 *value, /* The first register is always required */ - status = acpi_read(&value_a, register_a); + status = acpi_hw_read(&value_a, register_a); if (ACPI_FAILURE(status)) { return (status); } @@ -409,7 +591,7 @@ acpi_hw_read_multiple(u32 *value, /* Second register is optional */ if (register_b->address) { - status = acpi_read(&value_b, register_b); + status = acpi_hw_read(&value_b, register_b); if (ACPI_FAILURE(status)) { return (status); } @@ -452,7 +634,7 @@ acpi_hw_write_multiple(u32 value, /* The first register is always required */ - status = acpi_write(value, register_a); + status = acpi_hw_write(value, register_a); if (ACPI_FAILURE(status)) { return (status); } @@ -470,7 +652,7 @@ acpi_hw_write_multiple(u32 value, * and writes have no side effects" */ if (register_b->address) { - status = acpi_write(value, register_b); + status = acpi_hw_write(value, register_b); } return (status); diff --git a/drivers/acpi/acpica/hwtimer.c b/drivers/acpi/acpica/hwtimer.c index b7f522c8f023..6b282e85d039 100644 --- a/drivers/acpi/acpica/hwtimer.c +++ b/drivers/acpi/acpica/hwtimer.c @@ -100,7 +100,7 @@ acpi_status acpi_get_timer(u32 * ticks) } status = - acpi_hw_low_level_read(32, ticks, &acpi_gbl_FADT.xpm_timer_block); + acpi_hw_read(ticks, &acpi_gbl_FADT.xpm_timer_block); return_ACPI_STATUS(status); } diff --git a/drivers/acpi/acpica/hwxface.c b/drivers/acpi/acpica/hwxface.c index 9829979f2bdd..4ead85f29215 100644 --- a/drivers/acpi/acpica/hwxface.c +++ b/drivers/acpi/acpica/hwxface.c @@ -80,7 +80,7 @@ acpi_status acpi_reset(void) /* Write the reset value to the reset register */ - status = acpi_write(acpi_gbl_FADT.reset_value, reset_reg); + status = acpi_hw_write(acpi_gbl_FADT.reset_value, reset_reg); return_ACPI_STATUS(status); } @@ -97,67 +97,92 @@ ACPI_EXPORT_SYMBOL(acpi_reset) * * DESCRIPTION: Read from either memory or IO space. * + * LIMITATIONS: + * bit_width must be exactly 8, 16, 32, or 64. + * space_iD must be system_memory or system_iO. + * bit_offset and access_width are currently ignored, as there has + * not been a need to implement these. + * ******************************************************************************/ -acpi_status acpi_read(u32 *value, struct acpi_generic_address *reg) +acpi_status acpi_read(u64 *return_value, struct acpi_generic_address *reg) { + u32 value; u32 width; u64 address; acpi_status status; ACPI_FUNCTION_NAME(acpi_read); - /* - * Must have a valid pointer to a GAS structure, and a non-zero address - * within. - */ - if (!reg) { + if (!return_value) { return (AE_BAD_PARAMETER); } - /* Get a local copy of the address. Handles possible alignment issues */ + /* Validate contents of the GAS register. Allow 64-bit transfers */ - ACPI_MOVE_64_TO_64(&address, ®->address); - if (!address) { - return (AE_BAD_ADDRESS); + status = acpi_hw_validate_register(reg, 64, &address); + if (ACPI_FAILURE(status)) { + return (status); } - /* Supported widths are 8/16/32 */ - width = reg->bit_width; - if ((width != 8) && (width != 16) && (width != 32)) { - return (AE_SUPPORT); + if (width == 64) { + width = 32; /* Break into two 32-bit transfers */ } - /* Initialize entire 32-bit return value to zero */ + /* Initialize entire 64-bit return value to zero */ - *value = 0; + *return_value = 0; + value = 0; /* * Two address spaces supported: Memory or IO. PCI_Config is * not supported here because the GAS structure is insufficient */ - switch (reg->space_id) { - case ACPI_ADR_SPACE_SYSTEM_MEMORY: + if (reg->space_id == ACPI_ADR_SPACE_SYSTEM_MEMORY) { + status = acpi_os_read_memory((acpi_physical_address) + address, &value, width); + if (ACPI_FAILURE(status)) { + return (status); + } + *return_value = value; + + if (reg->bit_width == 64) { - status = acpi_os_read_memory((acpi_physical_address) address, - value, width); - break; + /* Read the top 32 bits */ - case ACPI_ADR_SPACE_SYSTEM_IO: + status = acpi_os_read_memory((acpi_physical_address) + (address + 4), &value, 32); + if (ACPI_FAILURE(status)) { + return (status); + } + *return_value |= ((u64)value << 32); + } + } else { /* ACPI_ADR_SPACE_SYSTEM_IO, validated earlier */ - status = - acpi_hw_read_port((acpi_io_address) address, value, width); - break; + status = acpi_hw_read_port((acpi_io_address) + address, &value, width); + if (ACPI_FAILURE(status)) { + return (status); + } + *return_value = value; - default: - ACPI_ERROR((AE_INFO, - "Unsupported address space: %X", reg->space_id)); - return (AE_BAD_PARAMETER); + if (reg->bit_width == 64) { + + /* Read the top 32 bits */ + + status = acpi_hw_read_port((acpi_io_address) + (address + 4), &value, 32); + if (ACPI_FAILURE(status)) { + return (status); + } + *return_value |= ((u64)value << 32); + } } ACPI_DEBUG_PRINT((ACPI_DB_IO, - "Read: %8.8X width %2d from %8.8X%8.8X (%s)\n", - *value, width, ACPI_FORMAT_UINT64(address), + "Read: %8.8X%8.8X width %2d from %8.8X%8.8X (%s)\n", + ACPI_FORMAT_UINT64(*return_value), reg->bit_width, + ACPI_FORMAT_UINT64(address), acpi_ut_get_region_name(reg->space_id))); return (status); @@ -169,7 +194,7 @@ ACPI_EXPORT_SYMBOL(acpi_read) * * FUNCTION: acpi_write * - * PARAMETERS: Value - To be written + * PARAMETERS: Value - Value to be written * Reg - GAS register structure * * RETURN: Status @@ -177,7 +202,7 @@ ACPI_EXPORT_SYMBOL(acpi_read) * DESCRIPTION: Write to either memory or IO space. * ******************************************************************************/ -acpi_status acpi_write(u32 value, struct acpi_generic_address *reg) +acpi_status acpi_write(u64 value, struct acpi_generic_address *reg) { u32 width; u64 address; @@ -185,54 +210,61 @@ acpi_status acpi_write(u32 value, struct acpi_generic_address *reg) ACPI_FUNCTION_NAME(acpi_write); - /* - * Must have a valid pointer to a GAS structure, and a non-zero address - * within. - */ - if (!reg) { - return (AE_BAD_PARAMETER); - } + /* Validate contents of the GAS register. Allow 64-bit transfers */ - /* Get a local copy of the address. Handles possible alignment issues */ - - ACPI_MOVE_64_TO_64(&address, ®->address); - if (!address) { - return (AE_BAD_ADDRESS); + status = acpi_hw_validate_register(reg, 64, &address); + if (ACPI_FAILURE(status)) { + return (status); } - /* Supported widths are 8/16/32 */ - width = reg->bit_width; - if ((width != 8) && (width != 16) && (width != 32)) { - return (AE_SUPPORT); + if (width == 64) { + width = 32; /* Break into two 32-bit transfers */ } /* - * Two address spaces supported: Memory or IO. - * PCI_Config is not supported here because the GAS struct is insufficient + * Two address spaces supported: Memory or IO. PCI_Config is + * not supported here because the GAS structure is insufficient */ - switch (reg->space_id) { - case ACPI_ADR_SPACE_SYSTEM_MEMORY: - - status = acpi_os_write_memory((acpi_physical_address) address, - value, width); - break; + if (reg->space_id == ACPI_ADR_SPACE_SYSTEM_MEMORY) { + status = acpi_os_write_memory((acpi_physical_address) + address, ACPI_LODWORD(value), + width); + if (ACPI_FAILURE(status)) { + return (status); + } - case ACPI_ADR_SPACE_SYSTEM_IO: + if (reg->bit_width == 64) { + status = acpi_os_write_memory((acpi_physical_address) + (address + 4), + ACPI_HIDWORD(value), 32); + if (ACPI_FAILURE(status)) { + return (status); + } + } + } else { /* ACPI_ADR_SPACE_SYSTEM_IO, validated earlier */ - status = acpi_hw_write_port((acpi_io_address) address, value, + status = acpi_hw_write_port((acpi_io_address) + address, ACPI_LODWORD(value), width); - break; + if (ACPI_FAILURE(status)) { + return (status); + } - default: - ACPI_ERROR((AE_INFO, - "Unsupported address space: %X", reg->space_id)); - return (AE_BAD_PARAMETER); + if (reg->bit_width == 64) { + status = acpi_hw_write_port((acpi_io_address) + (address + 4), + ACPI_HIDWORD(value), 32); + if (ACPI_FAILURE(status)) { + return (status); + } + } } ACPI_DEBUG_PRINT((ACPI_DB_IO, - "Wrote: %8.8X width %2d to %8.8X%8.8X (%s)\n", - value, width, ACPI_FORMAT_UINT64(address), + "Wrote: %8.8X%8.8X width %2d to %8.8X%8.8X (%s)\n", + ACPI_FORMAT_UINT64(value), reg->bit_width, + ACPI_FORMAT_UINT64(address), acpi_ut_get_region_name(reg->space_id))); return (status); diff --git a/include/acpi/acpixf.h b/include/acpi/acpixf.h index 2aecaa5cc06c..b450a195319a 100644 --- a/include/acpi/acpixf.h +++ b/include/acpi/acpixf.h @@ -360,9 +360,9 @@ acpi_status acpi_set_firmware_waking_vector(u32 physical_address); acpi_status acpi_set_firmware_waking_vector64(u64 physical_address); #endif -acpi_status acpi_read(u32 *value, struct acpi_generic_address *reg); +acpi_status acpi_read(u64 *value, struct acpi_generic_address *reg); -acpi_status acpi_write(u32 value, struct acpi_generic_address *reg); +acpi_status acpi_write(u64 value, struct acpi_generic_address *reg); acpi_status acpi_get_sleep_type_data(u8 sleep_state, u8 * slp_typ_a, u8 * slp_typ_b); -- cgit v1.2.3 From 15b8dd53f5ffaf8e2d9095c423f713423f576c0f Mon Sep 17 00:00:00 2001 From: Bob Moore Date: Mon, 29 Jun 2009 13:39:29 +0800 Subject: ACPICA: Major update for acpi_get_object_info external interface Completed a major update for the acpi_get_object_info external interface. Changes include: - Support for variable, unlimited length HID, UID, and CID strings - Support Processor objects the same as Devices (HID,UID,CID,ADR,STA, etc.) - Call the _SxW power methods on behalf of a device object - Determine if a device is a PCI root bridge - Change the ACPI_BUFFER parameter to ACPI_DEVICE_INFO. These changes will require an update to all callers of this interface. See the ACPICA Programmer Reference for details. Also, update all invocations of acpi_get_object_info interface Signed-off-by: Bob Moore Signed-off-by: Lin Ming Signed-off-by: Len Brown --- arch/ia64/hp/common/sba_iommu.c | 7 +- drivers/acpi/acpi_memhotplug.c | 11 +- drivers/acpi/acpica/Makefile | 2 +- drivers/acpi/acpica/acconfig.h | 5 + drivers/acpi/acpica/acglobal.h | 3 +- drivers/acpi/acpica/acinterp.h | 4 +- drivers/acpi/acpica/acutils.h | 24 ++- drivers/acpi/acpica/evrgnini.c | 45 +---- drivers/acpi/acpica/exutils.c | 53 +++-- drivers/acpi/acpica/nsdumpdv.c | 7 +- drivers/acpi/acpica/nsxfeval.c | 23 ++- drivers/acpi/acpica/nsxfname.c | 237 +++++++++++++++++------ drivers/acpi/acpica/uteval.c | 375 ++++-------------------------------- drivers/acpi/acpica/utglobal.c | 10 +- drivers/acpi/acpica/utids.c | 382 +++++++++++++++++++++++++++++++++++++ drivers/acpi/acpica/utmisc.c | 28 +++ drivers/acpi/container.c | 11 +- drivers/acpi/dock.c | 8 +- drivers/acpi/glue.c | 6 +- drivers/acpi/scan.c | 153 +++++++++------ drivers/char/agp/hp-agp.c | 9 +- drivers/ide/ide-acpi.c | 5 +- drivers/pci/hotplug/acpiphp_ibm.c | 12 +- drivers/platform/x86/sony-laptop.c | 7 +- drivers/pnp/pnpacpi/core.c | 6 +- include/acpi/acpi_bus.h | 8 +- include/acpi/acpixf.h | 3 +- include/acpi/actypes.h | 87 +++++---- 28 files changed, 901 insertions(+), 630 deletions(-) create mode 100644 drivers/acpi/acpica/utids.c (limited to 'include') diff --git a/arch/ia64/hp/common/sba_iommu.c b/arch/ia64/hp/common/sba_iommu.c index 8cfb001092ab..674a8374c6d9 100644 --- a/arch/ia64/hp/common/sba_iommu.c +++ b/arch/ia64/hp/common/sba_iommu.c @@ -2026,24 +2026,21 @@ acpi_sba_ioc_add(struct acpi_device *device) struct ioc *ioc; acpi_status status; u64 hpa, length; - struct acpi_buffer buffer; struct acpi_device_info *dev_info; status = hp_acpi_csr_space(device->handle, &hpa, &length); if (ACPI_FAILURE(status)) return 1; - buffer.length = ACPI_ALLOCATE_LOCAL_BUFFER; - status = acpi_get_object_info(device->handle, &buffer); + status = acpi_get_object_info(device->handle, &dev_info); if (ACPI_FAILURE(status)) return 1; - dev_info = buffer.pointer; /* * For HWP0001, only SBA appears in ACPI namespace. It encloses the PCI * root bridges, and its CSR space includes the IOC function. */ - if (strncmp("HWP0001", dev_info->hardware_id.value, 7) == 0) { + if (strncmp("HWP0001", dev_info->hardware_id.string, 7) == 0) { hpa += ZX1_IOC_OFFSET; /* zx1 based systems default to kernel page size iommu pages */ if (!iovp_shift) diff --git a/drivers/acpi/acpi_memhotplug.c b/drivers/acpi/acpi_memhotplug.c index 9a62224cc278..80eacbe157e2 100644 --- a/drivers/acpi/acpi_memhotplug.c +++ b/drivers/acpi/acpi_memhotplug.c @@ -481,26 +481,23 @@ static acpi_status is_memory_device(acpi_handle handle) { char *hardware_id; acpi_status status; - struct acpi_buffer buffer = { ACPI_ALLOCATE_BUFFER, NULL }; struct acpi_device_info *info; - - status = acpi_get_object_info(handle, &buffer); + status = acpi_get_object_info(handle, &info); if (ACPI_FAILURE(status)) return status; - info = buffer.pointer; if (!(info->valid & ACPI_VALID_HID)) { - kfree(buffer.pointer); + kfree(info); return AE_ERROR; } - hardware_id = info->hardware_id.value; + hardware_id = info->hardware_id.string; if ((hardware_id == NULL) || (strcmp(hardware_id, ACPI_MEMORY_DEVICE_HID))) status = AE_ERROR; - kfree(buffer.pointer); + kfree(info); return status; } diff --git a/drivers/acpi/acpica/Makefile b/drivers/acpi/acpica/Makefile index 72ac28da14e3..0e7d56185f6d 100644 --- a/drivers/acpi/acpica/Makefile +++ b/drivers/acpi/acpica/Makefile @@ -44,4 +44,4 @@ acpi-y += tbxface.o tbinstal.o tbutils.o tbfind.o tbfadt.o tbxfroot.o acpi-y += utalloc.o utdebug.o uteval.o utinit.o utmisc.o utxface.o \ utcopy.o utdelete.o utglobal.o utmath.o utobject.o \ - utstate.o utmutex.o utobject.o utresrc.o utlock.o + utstate.o utmutex.o utobject.o utresrc.o utlock.o utids.o diff --git a/drivers/acpi/acpica/acconfig.h b/drivers/acpi/acpica/acconfig.h index e6777fb883d2..6c1fb2d9f4d5 100644 --- a/drivers/acpi/acpica/acconfig.h +++ b/drivers/acpi/acpica/acconfig.h @@ -203,6 +203,11 @@ #define ACPI_SMBUS_BUFFER_SIZE 34 +/* _sx_d and _sx_w control methods */ + +#define ACPI_NUM_sx_d_METHODS 4 +#define ACPI_NUM_sx_w_METHODS 5 + /****************************************************************************** * * ACPI AML Debugger diff --git a/drivers/acpi/acpica/acglobal.h b/drivers/acpi/acpica/acglobal.h index 0b73b31c1b53..6389f7c1de59 100644 --- a/drivers/acpi/acpica/acglobal.h +++ b/drivers/acpi/acpica/acglobal.h @@ -265,7 +265,8 @@ ACPI_EXTERN u8 acpi_gbl_osi_data; extern u8 acpi_gbl_shutdown; extern u32 acpi_gbl_startup_flags; extern const char *acpi_gbl_sleep_state_names[ACPI_S_STATE_COUNT]; -extern const char *acpi_gbl_highest_dstate_names[4]; +extern const char *acpi_gbl_lowest_dstate_names[ACPI_NUM_sx_w_METHODS]; +extern const char *acpi_gbl_highest_dstate_names[ACPI_NUM_sx_d_METHODS]; extern const struct acpi_opcode_info acpi_gbl_aml_op_info[AML_NUM_OPCODES]; extern const char *acpi_gbl_region_types[ACPI_NUM_PREDEFINED_REGIONS]; diff --git a/drivers/acpi/acpica/acinterp.h b/drivers/acpi/acpica/acinterp.h index e8db7a3143a5..5db9f2916f7c 100644 --- a/drivers/acpi/acpica/acinterp.h +++ b/drivers/acpi/acpica/acinterp.h @@ -461,9 +461,9 @@ void acpi_ex_acquire_global_lock(u32 rule); void acpi_ex_release_global_lock(u32 rule); -void acpi_ex_eisa_id_to_string(u32 numeric_id, char *out_string); +void acpi_ex_eisa_id_to_string(char *dest, acpi_integer compressed_id); -void acpi_ex_unsigned_integer_to_string(acpi_integer value, char *out_string); +void acpi_ex_integer_to_string(char *dest, acpi_integer value); /* * exregion - default op_region handlers diff --git a/drivers/acpi/acpica/acutils.h b/drivers/acpi/acpica/acutils.h index 897810ba0ccc..b0add85de308 100644 --- a/drivers/acpi/acpica/acutils.h +++ b/drivers/acpi/acpica/acutils.h @@ -324,26 +324,30 @@ acpi_ut_evaluate_object(struct acpi_namespace_node *prefix_node, acpi_status acpi_ut_evaluate_numeric_object(char *object_name, struct acpi_namespace_node *device_node, - acpi_integer * address); + acpi_integer *value); acpi_status -acpi_ut_execute_HID(struct acpi_namespace_node *device_node, - struct acpica_device_id *hid); +acpi_ut_execute_STA(struct acpi_namespace_node *device_node, u32 *status_flags); acpi_status -acpi_ut_execute_CID(struct acpi_namespace_node *device_node, - struct acpi_compatible_id_list **return_cid_list); +acpi_ut_execute_power_methods(struct acpi_namespace_node *device_node, + const char **method_names, + u8 method_count, u8 *out_values); +/* + * utids - device ID support + */ acpi_status -acpi_ut_execute_STA(struct acpi_namespace_node *device_node, - u32 * status_flags); +acpi_ut_execute_HID(struct acpi_namespace_node *device_node, + struct acpica_device_id **return_id); acpi_status acpi_ut_execute_UID(struct acpi_namespace_node *device_node, - struct acpica_device_id *uid); + struct acpica_device_id **return_id); acpi_status -acpi_ut_execute_sxds(struct acpi_namespace_node *device_node, u8 * highest); +acpi_ut_execute_CID(struct acpi_namespace_node *device_node, + struct acpica_device_id_list **return_cid_list); /* * utlock - reader/writer locks @@ -445,6 +449,8 @@ acpi_ut_short_divide(acpi_integer in_dividend, */ const char *acpi_ut_validate_exception(acpi_status status); +u8 acpi_ut_is_pci_root_bridge(char *id); + u8 acpi_ut_is_aml_table(struct acpi_table_header *table); acpi_status acpi_ut_allocate_owner_id(acpi_owner_id * owner_id); diff --git a/drivers/acpi/acpica/evrgnini.c b/drivers/acpi/acpica/evrgnini.c index 284a7becbe96..cf29c4953028 100644 --- a/drivers/acpi/acpica/evrgnini.c +++ b/drivers/acpi/acpica/evrgnini.c @@ -50,8 +50,6 @@ ACPI_MODULE_NAME("evrgnini") /* Local prototypes */ -static u8 acpi_ev_match_pci_root_bridge(char *id); - static u8 acpi_ev_is_pci_root_bridge(struct acpi_namespace_node *node); /******************************************************************************* @@ -330,37 +328,6 @@ acpi_ev_pci_config_region_setup(acpi_handle handle, return_ACPI_STATUS(AE_OK); } -/******************************************************************************* - * - * FUNCTION: acpi_ev_match_pci_root_bridge - * - * PARAMETERS: Id - The HID/CID in string format - * - * RETURN: TRUE if the Id is a match for a PCI/PCI-Express Root Bridge - * - * DESCRIPTION: Determine if the input ID is a PCI Root Bridge ID. - * - ******************************************************************************/ - -static u8 acpi_ev_match_pci_root_bridge(char *id) -{ - - /* - * Check if this is a PCI root. - * ACPI 3.0+: check for a PCI Express root also. - */ - if (!(ACPI_STRNCMP(id, - PCI_ROOT_HID_STRING, - sizeof(PCI_ROOT_HID_STRING))) || - !(ACPI_STRNCMP(id, - PCI_EXPRESS_ROOT_HID_STRING, - sizeof(PCI_EXPRESS_ROOT_HID_STRING)))) { - return (TRUE); - } - - return (FALSE); -} - /******************************************************************************* * * FUNCTION: acpi_ev_is_pci_root_bridge @@ -377,9 +344,10 @@ static u8 acpi_ev_match_pci_root_bridge(char *id) static u8 acpi_ev_is_pci_root_bridge(struct acpi_namespace_node *node) { acpi_status status; - struct acpica_device_id hid; - struct acpi_compatible_id_list *cid; + struct acpica_device_id *hid; + struct acpica_device_id_list *cid; u32 i; + u8 match; /* Get the _HID and check for a PCI Root Bridge */ @@ -388,7 +356,10 @@ static u8 acpi_ev_is_pci_root_bridge(struct acpi_namespace_node *node) return (FALSE); } - if (acpi_ev_match_pci_root_bridge(hid.value)) { + match = acpi_ut_is_pci_root_bridge(hid->string); + ACPI_FREE(hid); + + if (match) { return (TRUE); } @@ -402,7 +373,7 @@ static u8 acpi_ev_is_pci_root_bridge(struct acpi_namespace_node *node) /* Check all _CIDs in the returned list */ for (i = 0; i < cid->count; i++) { - if (acpi_ev_match_pci_root_bridge(cid->id[i].value)) { + if (acpi_ut_is_pci_root_bridge(cid->ids[i].string)) { ACPI_FREE(cid); return (TRUE); } diff --git a/drivers/acpi/acpica/exutils.c b/drivers/acpi/acpica/exutils.c index 87730e944132..7d41f99f7052 100644 --- a/drivers/acpi/acpica/exutils.c +++ b/drivers/acpi/acpica/exutils.c @@ -358,50 +358,67 @@ static u32 acpi_ex_digits_needed(acpi_integer value, u32 base) * * FUNCTION: acpi_ex_eisa_id_to_string * - * PARAMETERS: numeric_id - EISA ID to be converted + * PARAMETERS: compressed_id - EISAID to be converted * out_string - Where to put the converted string (8 bytes) * * RETURN: None * - * DESCRIPTION: Convert a numeric EISA ID to string representation + * DESCRIPTION: Convert a numeric EISAID to string representation. Return + * buffer must be large enough to hold the string. The string + * returned is always exactly of length ACPI_EISAID_STRING_SIZE + * (includes null terminator). The EISAID is always 32 bits. * ******************************************************************************/ -void acpi_ex_eisa_id_to_string(u32 numeric_id, char *out_string) +void acpi_ex_eisa_id_to_string(char *out_string, acpi_integer compressed_id) { - u32 eisa_id; + u32 swapped_id; ACPI_FUNCTION_ENTRY(); + /* The EISAID should be a 32-bit integer */ + + if (compressed_id > ACPI_UINT32_MAX) { + ACPI_WARNING((AE_INFO, + "Expected EISAID is larger than 32 bits: 0x%8.8X%8.8X, truncating", + ACPI_FORMAT_UINT64(compressed_id))); + } + /* Swap ID to big-endian to get contiguous bits */ - eisa_id = acpi_ut_dword_byte_swap(numeric_id); + swapped_id = acpi_ut_dword_byte_swap((u32)compressed_id); - out_string[0] = (char)('@' + (((unsigned long)eisa_id >> 26) & 0x1f)); - out_string[1] = (char)('@' + ((eisa_id >> 21) & 0x1f)); - out_string[2] = (char)('@' + ((eisa_id >> 16) & 0x1f)); - out_string[3] = acpi_ut_hex_to_ascii_char((acpi_integer) eisa_id, 12); - out_string[4] = acpi_ut_hex_to_ascii_char((acpi_integer) eisa_id, 8); - out_string[5] = acpi_ut_hex_to_ascii_char((acpi_integer) eisa_id, 4); - out_string[6] = acpi_ut_hex_to_ascii_char((acpi_integer) eisa_id, 0); + /* First 3 bytes are uppercase letters. Next 4 bytes are hexadecimal */ + + out_string[0] = + (char)(0x40 + (((unsigned long)swapped_id >> 26) & 0x1F)); + out_string[1] = (char)(0x40 + ((swapped_id >> 21) & 0x1F)); + out_string[2] = (char)(0x40 + ((swapped_id >> 16) & 0x1F)); + out_string[3] = acpi_ut_hex_to_ascii_char((acpi_integer)swapped_id, 12); + out_string[4] = acpi_ut_hex_to_ascii_char((acpi_integer)swapped_id, 8); + out_string[5] = acpi_ut_hex_to_ascii_char((acpi_integer)swapped_id, 4); + out_string[6] = acpi_ut_hex_to_ascii_char((acpi_integer)swapped_id, 0); out_string[7] = 0; } /******************************************************************************* * - * FUNCTION: acpi_ex_unsigned_integer_to_string + * FUNCTION: acpi_ex_integer_to_string * - * PARAMETERS: Value - Value to be converted - * out_string - Where to put the converted string (8 bytes) + * PARAMETERS: out_string - Where to put the converted string. At least + * 21 bytes are needed to hold the largest + * possible 64-bit integer. + * Value - Value to be converted * * RETURN: None, string * - * DESCRIPTION: Convert a number to string representation. Assumes string - * buffer is large enough to hold the string. + * DESCRIPTION: Convert a 64-bit integer to decimal string representation. + * Assumes string buffer is large enough to hold the string. The + * largest string is (ACPI_MAX64_DECIMAL_DIGITS + 1). * ******************************************************************************/ -void acpi_ex_unsigned_integer_to_string(acpi_integer value, char *out_string) +void acpi_ex_integer_to_string(char *out_string, acpi_integer value) { u32 count; u32 digits_needed; diff --git a/drivers/acpi/acpica/nsdumpdv.c b/drivers/acpi/acpica/nsdumpdv.c index 41994fe7fbb8..0fe87f1aef16 100644 --- a/drivers/acpi/acpica/nsdumpdv.c +++ b/drivers/acpi/acpica/nsdumpdv.c @@ -70,7 +70,6 @@ static acpi_status acpi_ns_dump_one_device(acpi_handle obj_handle, u32 level, void *context, void **return_value) { - struct acpi_buffer buffer; struct acpi_device_info *info; acpi_status status; u32 i; @@ -80,17 +79,15 @@ acpi_ns_dump_one_device(acpi_handle obj_handle, status = acpi_ns_dump_one_object(obj_handle, level, context, return_value); - buffer.length = ACPI_ALLOCATE_LOCAL_BUFFER; - status = acpi_get_object_info(obj_handle, &buffer); + status = acpi_get_object_info(obj_handle, &info); if (ACPI_SUCCESS(status)) { - info = buffer.pointer; for (i = 0; i < level; i++) { ACPI_DEBUG_PRINT_RAW((ACPI_DB_TABLES, " ")); } ACPI_DEBUG_PRINT_RAW((ACPI_DB_TABLES, " HID: %s, ADR: %8.8X%8.8X, Status: %X\n", - info->hardware_id.value, + info->hardware_id.string, ACPI_FORMAT_UINT64(info->address), info->current_status)); ACPI_FREE(info); diff --git a/drivers/acpi/acpica/nsxfeval.c b/drivers/acpi/acpica/nsxfeval.c index daf4ad37896d..4929dbdbc8f0 100644 --- a/drivers/acpi/acpica/nsxfeval.c +++ b/drivers/acpi/acpica/nsxfeval.c @@ -535,10 +535,11 @@ acpi_ns_get_device_callback(acpi_handle obj_handle, acpi_status status; struct acpi_namespace_node *node; u32 flags; - struct acpica_device_id hid; - struct acpi_compatible_id_list *cid; + struct acpica_device_id *hid; + struct acpica_device_id_list *cid; u32 i; - int found; + u8 found; + int no_match; status = acpi_ut_acquire_mutex(ACPI_MTX_NAMESPACE); if (ACPI_FAILURE(status)) { @@ -582,10 +583,14 @@ acpi_ns_get_device_callback(acpi_handle obj_handle, return (AE_CTRL_DEPTH); } - if (ACPI_STRNCMP(hid.value, info->hid, sizeof(hid.value)) != 0) { - - /* Get the list of Compatible IDs */ + no_match = ACPI_STRCMP(hid->string, info->hid); + ACPI_FREE(hid); + if (no_match) { + /* + * HID does not match, attempt match within the + * list of Compatible IDs (CIDs) + */ status = acpi_ut_execute_CID(node, &cid); if (status == AE_NOT_FOUND) { return (AE_OK); @@ -597,10 +602,8 @@ acpi_ns_get_device_callback(acpi_handle obj_handle, found = 0; for (i = 0; i < cid->count; i++) { - if (ACPI_STRNCMP(cid->id[i].value, info->hid, - sizeof(struct - acpi_compatible_id)) == - 0) { + if (ACPI_STRCMP(cid->ids[i].string, info->hid) + == 0) { found = 1; break; } diff --git a/drivers/acpi/acpica/nsxfname.c b/drivers/acpi/acpica/nsxfname.c index f23593d6add4..ddc84af6336e 100644 --- a/drivers/acpi/acpica/nsxfname.c +++ b/drivers/acpi/acpica/nsxfname.c @@ -51,6 +51,11 @@ #define _COMPONENT ACPI_NAMESPACE ACPI_MODULE_NAME("nsxfname") +/* Local prototypes */ +static char *acpi_ns_copy_device_id(struct acpica_device_id *dest, + struct acpica_device_id *source, + char *string_area); + /****************************************************************************** * * FUNCTION: acpi_get_handle @@ -68,6 +73,7 @@ ACPI_MODULE_NAME("nsxfname") * namespace handle. * ******************************************************************************/ + acpi_status acpi_get_handle(acpi_handle parent, acpi_string pathname, acpi_handle * ret_handle) @@ -208,12 +214,40 @@ acpi_get_name(acpi_handle handle, u32 name_type, struct acpi_buffer * buffer) ACPI_EXPORT_SYMBOL(acpi_get_name) +/****************************************************************************** + * + * FUNCTION: acpi_ns_copy_device_id + * + * PARAMETERS: Dest - Pointer to the destination DEVICE_ID + * Source - Pointer to the source DEVICE_ID + * string_area - Pointer to where to copy the dest string + * + * RETURN: Pointer to the next string area + * + * DESCRIPTION: Copy a single DEVICE_ID, including the string data. + * + ******************************************************************************/ +static char *acpi_ns_copy_device_id(struct acpica_device_id *dest, + struct acpica_device_id *source, + char *string_area) +{ + /* Create the destination DEVICE_ID */ + + dest->string = string_area; + dest->length = source->length; + + /* Copy actual string and return a pointer to the next string area */ + + ACPI_MEMCPY(string_area, source->string, source->length); + return (string_area + source->length); +} + /****************************************************************************** * * FUNCTION: acpi_get_object_info * - * PARAMETERS: Handle - Object Handle - * Buffer - Where the info is returned + * PARAMETERS: Handle - Object Handle + * return_buffer - Where the info is returned * * RETURN: Status * @@ -221,33 +255,37 @@ ACPI_EXPORT_SYMBOL(acpi_get_name) * namespace node and possibly by running several standard * control methods (Such as in the case of a device.) * + * For Device and Processor objects, run the Device _HID, _UID, _CID, _STA, + * _ADR, _sx_w, and _sx_d methods. + * + * Note: Allocates the return buffer, must be freed by the caller. + * ******************************************************************************/ + acpi_status -acpi_get_object_info(acpi_handle handle, struct acpi_buffer * buffer) +acpi_get_object_info(acpi_handle handle, + struct acpi_device_info **return_buffer) { - acpi_status status; struct acpi_namespace_node *node; struct acpi_device_info *info; - struct acpi_device_info *return_info; - struct acpi_compatible_id_list *cid_list = NULL; - acpi_size size; + struct acpica_device_id_list *cid_list = NULL; + struct acpica_device_id *hid = NULL; + struct acpica_device_id *uid = NULL; + char *next_id_string; + acpi_object_type type; + acpi_name name; + u8 param_count = 0; + u8 valid = 0; + u32 info_size; + u32 i; + acpi_status status; /* Parameter validation */ - if (!handle || !buffer) { + if (!handle || !return_buffer) { return (AE_BAD_PARAMETER); } - status = acpi_ut_validate_buffer(buffer); - if (ACPI_FAILURE(status)) { - return (status); - } - - info = ACPI_ALLOCATE_ZEROED(sizeof(struct acpi_device_info)); - if (!info) { - return (AE_NO_MEMORY); - } - status = acpi_ut_acquire_mutex(ACPI_MTX_NAMESPACE); if (ACPI_FAILURE(status)) { goto cleanup; @@ -256,66 +294,91 @@ acpi_get_object_info(acpi_handle handle, struct acpi_buffer * buffer) node = acpi_ns_map_handle_to_node(handle); if (!node) { (void)acpi_ut_release_mutex(ACPI_MTX_NAMESPACE); - status = AE_BAD_PARAMETER; - goto cleanup; + return (AE_BAD_PARAMETER); } - /* Init return structure */ - - size = sizeof(struct acpi_device_info); + /* Get the namespace node data while the namespace is locked */ - info->type = node->type; - info->name = node->name.integer; - info->valid = 0; + info_size = sizeof(struct acpi_device_info); + type = node->type; + name = node->name.integer; if (node->type == ACPI_TYPE_METHOD) { - info->param_count = node->object->method.param_count; + param_count = node->object->method.param_count; } status = acpi_ut_release_mutex(ACPI_MTX_NAMESPACE); if (ACPI_FAILURE(status)) { - goto cleanup; + return (status); } - /* If not a device, we are all done */ - - if (info->type == ACPI_TYPE_DEVICE) { + if ((type == ACPI_TYPE_DEVICE) || (type == ACPI_TYPE_PROCESSOR)) { /* - * Get extra info for ACPI Devices objects only: - * Run the Device _HID, _UID, _CID, _STA, _ADR and _sx_d methods. + * Get extra info for ACPI Device/Processor objects only: + * Run the Device _HID, _UID, and _CID methods. * * Note: none of these methods are required, so they may or may - * not be present for this device. The Info->Valid bitfield is used - * to indicate which methods were found and ran successfully. + * not be present for this device. The Info->Valid bitfield is used + * to indicate which methods were found and run successfully. */ /* Execute the Device._HID method */ - status = acpi_ut_execute_HID(node, &info->hardware_id); + status = acpi_ut_execute_HID(node, &hid); if (ACPI_SUCCESS(status)) { - info->valid |= ACPI_VALID_HID; + info_size += hid->length; + valid |= ACPI_VALID_HID; } /* Execute the Device._UID method */ - status = acpi_ut_execute_UID(node, &info->unique_id); + status = acpi_ut_execute_UID(node, &uid); if (ACPI_SUCCESS(status)) { - info->valid |= ACPI_VALID_UID; + info_size += uid->length; + valid |= ACPI_VALID_UID; } /* Execute the Device._CID method */ status = acpi_ut_execute_CID(node, &cid_list); if (ACPI_SUCCESS(status)) { - size += cid_list->size; - info->valid |= ACPI_VALID_CID; + + /* Add size of CID strings and CID pointer array */ + + info_size += + (cid_list->list_size - + sizeof(struct acpica_device_id_list)); + valid |= ACPI_VALID_CID; } + } + + /* + * Now that we have the variable-length data, we can allocate the + * return buffer + */ + info = ACPI_ALLOCATE_ZEROED(info_size); + if (!info) { + status = AE_NO_MEMORY; + goto cleanup; + } + + /* Get the fixed-length data */ + + if ((type == ACPI_TYPE_DEVICE) || (type == ACPI_TYPE_PROCESSOR)) { + /* + * Get extra info for ACPI Device/Processor objects only: + * Run the _STA, _ADR and, sx_w, and _sx_d methods. + * + * Note: none of these methods are required, so they may or may + * not be present for this device. The Info->Valid bitfield is used + * to indicate which methods were found and run successfully. + */ /* Execute the Device._STA method */ status = acpi_ut_execute_STA(node, &info->current_status); if (ACPI_SUCCESS(status)) { - info->valid |= ACPI_VALID_STA; + valid |= ACPI_VALID_STA; } /* Execute the Device._ADR method */ @@ -323,36 +386,100 @@ acpi_get_object_info(acpi_handle handle, struct acpi_buffer * buffer) status = acpi_ut_evaluate_numeric_object(METHOD_NAME__ADR, node, &info->address); if (ACPI_SUCCESS(status)) { - info->valid |= ACPI_VALID_ADR; + valid |= ACPI_VALID_ADR; + } + + /* Execute the Device._sx_w methods */ + + status = acpi_ut_execute_power_methods(node, + acpi_gbl_lowest_dstate_names, + ACPI_NUM_sx_w_METHODS, + info->lowest_dstates); + if (ACPI_SUCCESS(status)) { + valid |= ACPI_VALID_SXWS; } /* Execute the Device._sx_d methods */ - status = acpi_ut_execute_sxds(node, info->highest_dstates); + status = acpi_ut_execute_power_methods(node, + acpi_gbl_highest_dstate_names, + ACPI_NUM_sx_d_METHODS, + info->highest_dstates); if (ACPI_SUCCESS(status)) { - info->valid |= ACPI_VALID_SXDS; + valid |= ACPI_VALID_SXDS; } } - /* Validate/Allocate/Clear caller buffer */ + /* + * Create a pointer to the string area of the return buffer. + * Point to the end of the base struct acpi_device_info structure. + */ + next_id_string = ACPI_CAST_PTR(char, info->compatible_id_list.ids); + if (cid_list) { - status = acpi_ut_initialize_buffer(buffer, size); - if (ACPI_FAILURE(status)) { - goto cleanup; + /* Point past the CID DEVICE_ID array */ + + next_id_string += + ((acpi_size) cid_list->count * + sizeof(struct acpica_device_id)); } - /* Populate the return buffer */ + /* + * Copy the HID, UID, and CIDs to the return buffer. The variable-length + * strings are copied to the reserved area at the end of the buffer. + * + * For HID and CID, check if the ID is a PCI Root Bridge. + */ + if (hid) { + next_id_string = acpi_ns_copy_device_id(&info->hardware_id, + hid, next_id_string); + + if (acpi_ut_is_pci_root_bridge(hid->string)) { + info->flags |= ACPI_PCI_ROOT_BRIDGE; + } + } - return_info = buffer->pointer; - ACPI_MEMCPY(return_info, info, sizeof(struct acpi_device_info)); + if (uid) { + next_id_string = acpi_ns_copy_device_id(&info->unique_id, + uid, next_id_string); + } if (cid_list) { - ACPI_MEMCPY(&return_info->compatibility_id, cid_list, - cid_list->size); + info->compatible_id_list.count = cid_list->count; + info->compatible_id_list.list_size = cid_list->list_size; + + /* Copy each CID */ + + for (i = 0; i < cid_list->count; i++) { + next_id_string = + acpi_ns_copy_device_id(&info->compatible_id_list. + ids[i], &cid_list->ids[i], + next_id_string); + + if (acpi_ut_is_pci_root_bridge(cid_list->ids[i].string)) { + info->flags |= ACPI_PCI_ROOT_BRIDGE; + } + } } + /* Copy the fixed-length data */ + + info->info_size = info_size; + info->type = type; + info->name = name; + info->param_count = param_count; + info->valid = valid; + + *return_buffer = info; + status = AE_OK; + cleanup: - ACPI_FREE(info); + if (hid) { + ACPI_FREE(hid); + } + if (uid) { + ACPI_FREE(uid); + } if (cid_list) { ACPI_FREE(cid_list); } diff --git a/drivers/acpi/acpica/uteval.c b/drivers/acpi/acpica/uteval.c index 006b16c26017..5503307b8bb7 100644 --- a/drivers/acpi/acpica/uteval.c +++ b/drivers/acpi/acpica/uteval.c @@ -44,19 +44,10 @@ #include #include "accommon.h" #include "acnamesp.h" -#include "acinterp.h" #define _COMPONENT ACPI_UTILITIES ACPI_MODULE_NAME("uteval") -/* Local prototypes */ -static void -acpi_ut_copy_id_string(char *destination, char *source, acpi_size max_length); - -static acpi_status -acpi_ut_translate_one_cid(union acpi_operand_object *obj_desc, - struct acpi_compatible_id *one_cid); - /* * Strings supported by the _OSI predefined (internal) method. * @@ -213,7 +204,7 @@ acpi_status acpi_osi_invalidate(char *interface) * RETURN: Status * * DESCRIPTION: Evaluates a namespace object and verifies the type of the - * return object. Common code that simplifies accessing objects + * return object. Common code that simplifies accessing objects * that have required return objects of fixed types. * * NOTE: Internal function, no parameter validation @@ -298,7 +289,7 @@ acpi_ut_evaluate_object(struct acpi_namespace_node *prefix_node, if ((acpi_gbl_enable_interpreter_slack) && (!expected_return_btypes)) { /* - * We received a return object, but one was not expected. This can + * We received a return object, but one was not expected. This can * happen frequently if the "implicit return" feature is enabled. * Just delete the return object and return AE_OK. */ @@ -340,12 +331,12 @@ acpi_ut_evaluate_object(struct acpi_namespace_node *prefix_node, * * PARAMETERS: object_name - Object name to be evaluated * device_node - Node for the device - * Address - Where the value is returned + * Value - Where the value is returned * * RETURN: Status * * DESCRIPTION: Evaluates a numeric namespace object for a selected device - * and stores result in *Address. + * and stores result in *Value. * * NOTE: Internal function, no parameter validation * @@ -354,7 +345,7 @@ acpi_ut_evaluate_object(struct acpi_namespace_node *prefix_node, acpi_status acpi_ut_evaluate_numeric_object(char *object_name, struct acpi_namespace_node *device_node, - acpi_integer * address) + acpi_integer *value) { union acpi_operand_object *obj_desc; acpi_status status; @@ -369,295 +360,7 @@ acpi_ut_evaluate_numeric_object(char *object_name, /* Get the returned Integer */ - *address = obj_desc->integer.value; - - /* On exit, we must delete the return object */ - - acpi_ut_remove_reference(obj_desc); - return_ACPI_STATUS(status); -} - -/******************************************************************************* - * - * FUNCTION: acpi_ut_copy_id_string - * - * PARAMETERS: Destination - Where to copy the string - * Source - Source string - * max_length - Length of the destination buffer - * - * RETURN: None - * - * DESCRIPTION: Copies an ID string for the _HID, _CID, and _UID methods. - * Performs removal of a leading asterisk if present -- workaround - * for a known issue on a bunch of machines. - * - ******************************************************************************/ - -static void -acpi_ut_copy_id_string(char *destination, char *source, acpi_size max_length) -{ - - /* - * Workaround for ID strings that have a leading asterisk. This construct - * is not allowed by the ACPI specification (ID strings must be - * alphanumeric), but enough existing machines have this embedded in their - * ID strings that the following code is useful. - */ - if (*source == '*') { - source++; - } - - /* Do the actual copy */ - - ACPI_STRNCPY(destination, source, max_length); -} - -/******************************************************************************* - * - * FUNCTION: acpi_ut_execute_HID - * - * PARAMETERS: device_node - Node for the device - * Hid - Where the HID is returned - * - * RETURN: Status - * - * DESCRIPTION: Executes the _HID control method that returns the hardware - * ID of the device. - * - * NOTE: Internal function, no parameter validation - * - ******************************************************************************/ - -acpi_status -acpi_ut_execute_HID(struct acpi_namespace_node *device_node, - struct acpica_device_id *hid) -{ - union acpi_operand_object *obj_desc; - acpi_status status; - - ACPI_FUNCTION_TRACE(ut_execute_HID); - - status = acpi_ut_evaluate_object(device_node, METHOD_NAME__HID, - ACPI_BTYPE_INTEGER | ACPI_BTYPE_STRING, - &obj_desc); - if (ACPI_FAILURE(status)) { - return_ACPI_STATUS(status); - } - - if (obj_desc->common.type == ACPI_TYPE_INTEGER) { - - /* Convert the Numeric HID to string */ - - acpi_ex_eisa_id_to_string((u32) obj_desc->integer.value, - hid->value); - } else { - /* Copy the String HID from the returned object */ - - acpi_ut_copy_id_string(hid->value, obj_desc->string.pointer, - sizeof(hid->value)); - } - - /* On exit, we must delete the return object */ - - acpi_ut_remove_reference(obj_desc); - return_ACPI_STATUS(status); -} - -/******************************************************************************* - * - * FUNCTION: acpi_ut_translate_one_cid - * - * PARAMETERS: obj_desc - _CID object, must be integer or string - * one_cid - Where the CID string is returned - * - * RETURN: Status - * - * DESCRIPTION: Return a numeric or string _CID value as a string. - * (Compatible ID) - * - * NOTE: Assumes a maximum _CID string length of - * ACPI_MAX_CID_LENGTH. - * - ******************************************************************************/ - -static acpi_status -acpi_ut_translate_one_cid(union acpi_operand_object *obj_desc, - struct acpi_compatible_id *one_cid) -{ - - switch (obj_desc->common.type) { - case ACPI_TYPE_INTEGER: - - /* Convert the Numeric CID to string */ - - acpi_ex_eisa_id_to_string((u32) obj_desc->integer.value, - one_cid->value); - return (AE_OK); - - case ACPI_TYPE_STRING: - - if (obj_desc->string.length > ACPI_MAX_CID_LENGTH) { - return (AE_AML_STRING_LIMIT); - } - - /* Copy the String CID from the returned object */ - - acpi_ut_copy_id_string(one_cid->value, obj_desc->string.pointer, - ACPI_MAX_CID_LENGTH); - return (AE_OK); - - default: - - return (AE_TYPE); - } -} - -/******************************************************************************* - * - * FUNCTION: acpi_ut_execute_CID - * - * PARAMETERS: device_node - Node for the device - * return_cid_list - Where the CID list is returned - * - * RETURN: Status - * - * DESCRIPTION: Executes the _CID control method that returns one or more - * compatible hardware IDs for the device. - * - * NOTE: Internal function, no parameter validation - * - ******************************************************************************/ - -acpi_status -acpi_ut_execute_CID(struct acpi_namespace_node * device_node, - struct acpi_compatible_id_list ** return_cid_list) -{ - union acpi_operand_object *obj_desc; - acpi_status status; - u32 count; - u32 size; - struct acpi_compatible_id_list *cid_list; - u32 i; - - ACPI_FUNCTION_TRACE(ut_execute_CID); - - /* Evaluate the _CID method for this device */ - - status = acpi_ut_evaluate_object(device_node, METHOD_NAME__CID, - ACPI_BTYPE_INTEGER | ACPI_BTYPE_STRING - | ACPI_BTYPE_PACKAGE, &obj_desc); - if (ACPI_FAILURE(status)) { - return_ACPI_STATUS(status); - } - - /* Get the number of _CIDs returned */ - - count = 1; - if (obj_desc->common.type == ACPI_TYPE_PACKAGE) { - count = obj_desc->package.count; - } - - /* Allocate a worst-case buffer for the _CIDs */ - - size = (((count - 1) * sizeof(struct acpi_compatible_id)) + - sizeof(struct acpi_compatible_id_list)); - - cid_list = ACPI_ALLOCATE_ZEROED((acpi_size) size); - if (!cid_list) { - return_ACPI_STATUS(AE_NO_MEMORY); - } - - /* Init CID list */ - - cid_list->count = count; - cid_list->size = size; - - /* - * A _CID can return either a single compatible ID or a package of - * compatible IDs. Each compatible ID can be one of the following: - * 1) Integer (32 bit compressed EISA ID) or - * 2) String (PCI ID format, e.g. "PCI\VEN_vvvv&DEV_dddd&SUBSYS_ssssssss") - */ - - /* The _CID object can be either a single CID or a package (list) of CIDs */ - - if (obj_desc->common.type == ACPI_TYPE_PACKAGE) { - - /* Translate each package element */ - - for (i = 0; i < count; i++) { - status = - acpi_ut_translate_one_cid(obj_desc->package. - elements[i], - &cid_list->id[i]); - if (ACPI_FAILURE(status)) { - break; - } - } - } else { - /* Only one CID, translate to a string */ - - status = acpi_ut_translate_one_cid(obj_desc, cid_list->id); - } - - /* Cleanup on error */ - - if (ACPI_FAILURE(status)) { - ACPI_FREE(cid_list); - } else { - *return_cid_list = cid_list; - } - - /* On exit, we must delete the _CID return object */ - - acpi_ut_remove_reference(obj_desc); - return_ACPI_STATUS(status); -} - -/******************************************************************************* - * - * FUNCTION: acpi_ut_execute_UID - * - * PARAMETERS: device_node - Node for the device - * Uid - Where the UID is returned - * - * RETURN: Status - * - * DESCRIPTION: Executes the _UID control method that returns the hardware - * ID of the device. - * - * NOTE: Internal function, no parameter validation - * - ******************************************************************************/ - -acpi_status -acpi_ut_execute_UID(struct acpi_namespace_node *device_node, - struct acpica_device_id *uid) -{ - union acpi_operand_object *obj_desc; - acpi_status status; - - ACPI_FUNCTION_TRACE(ut_execute_UID); - - status = acpi_ut_evaluate_object(device_node, METHOD_NAME__UID, - ACPI_BTYPE_INTEGER | ACPI_BTYPE_STRING, - &obj_desc); - if (ACPI_FAILURE(status)) { - return_ACPI_STATUS(status); - } - - if (obj_desc->common.type == ACPI_TYPE_INTEGER) { - - /* Convert the Numeric UID to string */ - - acpi_ex_unsigned_integer_to_string(obj_desc->integer.value, - uid->value); - } else { - /* Copy the String UID from the returned object */ - - acpi_ut_copy_id_string(uid->value, obj_desc->string.pointer, - sizeof(uid->value)); - } + *value = obj_desc->integer.value; /* On exit, we must delete the return object */ @@ -716,60 +419,64 @@ acpi_ut_execute_STA(struct acpi_namespace_node *device_node, u32 * flags) /******************************************************************************* * - * FUNCTION: acpi_ut_execute_Sxds + * FUNCTION: acpi_ut_execute_power_methods * * PARAMETERS: device_node - Node for the device - * Flags - Where the status flags are returned + * method_names - Array of power method names + * method_count - Number of methods to execute + * out_values - Where the power method values are returned * - * RETURN: Status + * RETURN: Status, out_values * - * DESCRIPTION: Executes _STA for selected device and stores results in - * *Flags. + * DESCRIPTION: Executes the specified power methods for the device and returns + * the result(s). * * NOTE: Internal function, no parameter validation * - ******************************************************************************/ +******************************************************************************/ acpi_status -acpi_ut_execute_sxds(struct acpi_namespace_node *device_node, u8 * highest) +acpi_ut_execute_power_methods(struct acpi_namespace_node *device_node, + const char **method_names, + u8 method_count, u8 *out_values) { union acpi_operand_object *obj_desc; acpi_status status; + acpi_status final_status = AE_NOT_FOUND; u32 i; - ACPI_FUNCTION_TRACE(ut_execute_sxds); + ACPI_FUNCTION_TRACE(ut_execute_power_methods); - for (i = 0; i < 4; i++) { - highest[i] = 0xFF; + for (i = 0; i < method_count; i++) { + /* + * Execute the power method (_sx_d or _sx_w). The only allowable + * return type is an Integer. + */ status = acpi_ut_evaluate_object(device_node, ACPI_CAST_PTR(char, - acpi_gbl_highest_dstate_names - [i]), + method_names[i]), ACPI_BTYPE_INTEGER, &obj_desc); - if (ACPI_FAILURE(status)) { - if (status != AE_NOT_FOUND) { - ACPI_DEBUG_PRINT((ACPI_DB_EXEC, - "%s on Device %4.4s, %s\n", - ACPI_CAST_PTR(char, - acpi_gbl_highest_dstate_names - [i]), - acpi_ut_get_node_name - (device_node), - acpi_format_exception - (status))); - - return_ACPI_STATUS(status); - } - } else { - /* Extract the Dstate value */ - - highest[i] = (u8) obj_desc->integer.value; + if (ACPI_SUCCESS(status)) { + out_values[i] = (u8)obj_desc->integer.value; /* Delete the return object */ acpi_ut_remove_reference(obj_desc); + final_status = AE_OK; /* At least one value is valid */ + continue; } + + out_values[i] = ACPI_UINT8_MAX; + if (status == AE_NOT_FOUND) { + continue; /* Ignore if not found */ + } + + ACPI_DEBUG_PRINT((ACPI_DB_EXEC, + "Failed %s on Device %4.4s, %s\n", + ACPI_CAST_PTR(char, method_names[i]), + acpi_ut_get_node_name(device_node), + acpi_format_exception(status))); } - return_ACPI_STATUS(AE_OK); + return_ACPI_STATUS(final_status); } diff --git a/drivers/acpi/acpica/utglobal.c b/drivers/acpi/acpica/utglobal.c index 59e46f257c02..ed7a33c67fbe 100644 --- a/drivers/acpi/acpica/utglobal.c +++ b/drivers/acpi/acpica/utglobal.c @@ -90,7 +90,15 @@ const char *acpi_gbl_sleep_state_names[ACPI_S_STATE_COUNT] = { "\\_S5_" }; -const char *acpi_gbl_highest_dstate_names[4] = { +const char *acpi_gbl_lowest_dstate_names[ACPI_NUM_sx_w_METHODS] = { + "_S0W", + "_S1W", + "_S2W", + "_S3W", + "_S4W" +}; + +const char *acpi_gbl_highest_dstate_names[ACPI_NUM_sx_d_METHODS] = { "_S1D", "_S2D", "_S3D", diff --git a/drivers/acpi/acpica/utids.c b/drivers/acpi/acpica/utids.c new file mode 100644 index 000000000000..52eaae404554 --- /dev/null +++ b/drivers/acpi/acpica/utids.c @@ -0,0 +1,382 @@ +/****************************************************************************** + * + * Module Name: utids - support for device IDs - HID, UID, CID + * + *****************************************************************************/ + +/* + * Copyright (C) 2000 - 2009, Intel Corp. + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions, and the following disclaimer, + * without modification. + * 2. Redistributions in binary form must reproduce at minimum a disclaimer + * substantially similar to the "NO WARRANTY" disclaimer below + * ("Disclaimer") and any redistribution must be conditioned upon + * including a substantially similar Disclaimer requirement for further + * binary redistribution. + * 3. Neither the names of the above-listed copyright holders nor the names + * of any contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * Alternatively, this software may be distributed under the terms of the + * GNU General Public License ("GPL") version 2 as published by the Free + * Software Foundation. + * + * NO WARRANTY + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTIBILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * HOLDERS OR CONTRIBUTORS BE LIABLE FOR SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, + * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING + * IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE + * POSSIBILITY OF SUCH DAMAGES. + */ + +#include +#include "accommon.h" +#include "acinterp.h" + +#define _COMPONENT ACPI_UTILITIES +ACPI_MODULE_NAME("utids") + +/* Local prototypes */ +static void acpi_ut_copy_id_string(char *destination, char *source); + +/******************************************************************************* + * + * FUNCTION: acpi_ut_copy_id_string + * + * PARAMETERS: Destination - Where to copy the string + * Source - Source string + * + * RETURN: None + * + * DESCRIPTION: Copies an ID string for the _HID, _CID, and _UID methods. + * Performs removal of a leading asterisk if present -- workaround + * for a known issue on a bunch of machines. + * + ******************************************************************************/ + +static void acpi_ut_copy_id_string(char *destination, char *source) +{ + + /* + * Workaround for ID strings that have a leading asterisk. This construct + * is not allowed by the ACPI specification (ID strings must be + * alphanumeric), but enough existing machines have this embedded in their + * ID strings that the following code is useful. + */ + if (*source == '*') { + source++; + } + + /* Do the actual copy */ + + ACPI_STRCPY(destination, source); +} + +/******************************************************************************* + * + * FUNCTION: acpi_ut_execute_HID + * + * PARAMETERS: device_node - Node for the device + * return_id - Where the string HID is returned + * + * RETURN: Status + * + * DESCRIPTION: Executes the _HID control method that returns the hardware + * ID of the device. The HID is either an 32-bit encoded EISAID + * Integer or a String. A string is always returned. An EISAID + * is converted to a string. + * + * NOTE: Internal function, no parameter validation + * + ******************************************************************************/ + +acpi_status +acpi_ut_execute_HID(struct acpi_namespace_node *device_node, + struct acpica_device_id **return_id) +{ + union acpi_operand_object *obj_desc; + struct acpica_device_id *hid; + u32 length; + acpi_status status; + + ACPI_FUNCTION_TRACE(ut_execute_HID); + + status = acpi_ut_evaluate_object(device_node, METHOD_NAME__HID, + ACPI_BTYPE_INTEGER | ACPI_BTYPE_STRING, + &obj_desc); + if (ACPI_FAILURE(status)) { + return_ACPI_STATUS(status); + } + + /* Get the size of the String to be returned, includes null terminator */ + + if (obj_desc->common.type == ACPI_TYPE_INTEGER) { + length = ACPI_EISAID_STRING_SIZE; + } else { + length = obj_desc->string.length + 1; + } + + /* Allocate a buffer for the HID */ + + hid = + ACPI_ALLOCATE_ZEROED(sizeof(struct acpica_device_id) + + (acpi_size) length); + if (!hid) { + status = AE_NO_MEMORY; + goto cleanup; + } + + /* Area for the string starts after DEVICE_ID struct */ + + hid->string = ACPI_ADD_PTR(char, hid, sizeof(struct acpica_device_id)); + + /* Convert EISAID to a string or simply copy existing string */ + + if (obj_desc->common.type == ACPI_TYPE_INTEGER) { + acpi_ex_eisa_id_to_string(hid->string, obj_desc->integer.value); + } else { + acpi_ut_copy_id_string(hid->string, obj_desc->string.pointer); + } + + hid->length = length; + *return_id = hid; + +cleanup: + + /* On exit, we must delete the return object */ + + acpi_ut_remove_reference(obj_desc); + return_ACPI_STATUS(status); +} + +/******************************************************************************* + * + * FUNCTION: acpi_ut_execute_UID + * + * PARAMETERS: device_node - Node for the device + * return_id - Where the string UID is returned + * + * RETURN: Status + * + * DESCRIPTION: Executes the _UID control method that returns the unique + * ID of the device. The UID is either a 64-bit Integer (NOT an + * EISAID) or a string. Always returns a string. A 64-bit integer + * is converted to a decimal string. + * + * NOTE: Internal function, no parameter validation + * + ******************************************************************************/ + +acpi_status +acpi_ut_execute_UID(struct acpi_namespace_node *device_node, + struct acpica_device_id **return_id) +{ + union acpi_operand_object *obj_desc; + struct acpica_device_id *uid; + u32 length; + acpi_status status; + + ACPI_FUNCTION_TRACE(ut_execute_UID); + + status = acpi_ut_evaluate_object(device_node, METHOD_NAME__UID, + ACPI_BTYPE_INTEGER | ACPI_BTYPE_STRING, + &obj_desc); + if (ACPI_FAILURE(status)) { + return_ACPI_STATUS(status); + } + + /* Get the size of the String to be returned, includes null terminator */ + + if (obj_desc->common.type == ACPI_TYPE_INTEGER) { + length = ACPI_MAX64_DECIMAL_DIGITS + 1; + } else { + length = obj_desc->string.length + 1; + } + + /* Allocate a buffer for the UID */ + + uid = + ACPI_ALLOCATE_ZEROED(sizeof(struct acpica_device_id) + + (acpi_size) length); + if (!uid) { + status = AE_NO_MEMORY; + goto cleanup; + } + + /* Area for the string starts after DEVICE_ID struct */ + + uid->string = ACPI_ADD_PTR(char, uid, sizeof(struct acpica_device_id)); + + /* Convert an Integer to string, or just copy an existing string */ + + if (obj_desc->common.type == ACPI_TYPE_INTEGER) { + acpi_ex_integer_to_string(uid->string, obj_desc->integer.value); + } else { + acpi_ut_copy_id_string(uid->string, obj_desc->string.pointer); + } + + uid->length = length; + *return_id = uid; + +cleanup: + + /* On exit, we must delete the return object */ + + acpi_ut_remove_reference(obj_desc); + return_ACPI_STATUS(status); +} + +/******************************************************************************* + * + * FUNCTION: acpi_ut_execute_CID + * + * PARAMETERS: device_node - Node for the device + * return_cid_list - Where the CID list is returned + * + * RETURN: Status, list of CID strings + * + * DESCRIPTION: Executes the _CID control method that returns one or more + * compatible hardware IDs for the device. + * + * NOTE: Internal function, no parameter validation + * + * A _CID method can return either a single compatible ID or a package of + * compatible IDs. Each compatible ID can be one of the following: + * 1) Integer (32 bit compressed EISA ID) or + * 2) String (PCI ID format, e.g. "PCI\VEN_vvvv&DEV_dddd&SUBSYS_ssssssss") + * + * The Integer CIDs are converted to string format by this function. + * + ******************************************************************************/ + +acpi_status +acpi_ut_execute_CID(struct acpi_namespace_node *device_node, + struct acpica_device_id_list **return_cid_list) +{ + union acpi_operand_object **cid_objects; + union acpi_operand_object *obj_desc; + struct acpica_device_id_list *cid_list; + char *next_id_string; + u32 string_area_size; + u32 length; + u32 cid_list_size; + acpi_status status; + u32 count; + u32 i; + + ACPI_FUNCTION_TRACE(ut_execute_CID); + + /* Evaluate the _CID method for this device */ + + status = acpi_ut_evaluate_object(device_node, METHOD_NAME__CID, + ACPI_BTYPE_INTEGER | ACPI_BTYPE_STRING + | ACPI_BTYPE_PACKAGE, &obj_desc); + if (ACPI_FAILURE(status)) { + return_ACPI_STATUS(status); + } + + /* + * Get the count and size of the returned _CIDs. _CID can return either + * a Package of Integers/Strings or a single Integer or String. + * Note: This section also validates that all CID elements are of the + * correct type (Integer or String). + */ + if (obj_desc->common.type == ACPI_TYPE_PACKAGE) { + count = obj_desc->package.count; + cid_objects = obj_desc->package.elements; + } else { /* Single Integer or String CID */ + + count = 1; + cid_objects = &obj_desc; + } + + string_area_size = 0; + for (i = 0; i < count; i++) { + + /* String lengths include null terminator */ + + switch (cid_objects[i]->common.type) { + case ACPI_TYPE_INTEGER: + string_area_size += ACPI_EISAID_STRING_SIZE; + break; + + case ACPI_TYPE_STRING: + string_area_size += cid_objects[i]->string.length + 1; + break; + + default: + status = AE_TYPE; + goto cleanup; + } + } + + /* + * Now that we know the length of the CIDs, allocate return buffer: + * 1) Size of the base structure + + * 2) Size of the CID DEVICE_ID array + + * 3) Size of the actual CID strings + */ + cid_list_size = sizeof(struct acpica_device_id_list) + + ((count - 1) * sizeof(struct acpica_device_id)) + string_area_size; + + cid_list = ACPI_ALLOCATE_ZEROED(cid_list_size); + if (!cid_list) { + status = AE_NO_MEMORY; + goto cleanup; + } + + /* Area for CID strings starts after the CID DEVICE_ID array */ + + next_id_string = ACPI_CAST_PTR(char, cid_list->ids) + + ((acpi_size) count * sizeof(struct acpica_device_id)); + + /* Copy/convert the CIDs to the return buffer */ + + for (i = 0; i < count; i++) { + if (cid_objects[i]->common.type == ACPI_TYPE_INTEGER) { + + /* Convert the Integer (EISAID) CID to a string */ + + acpi_ex_eisa_id_to_string(next_id_string, + cid_objects[i]->integer. + value); + length = ACPI_EISAID_STRING_SIZE; + } else { /* ACPI_TYPE_STRING */ + + /* Copy the String CID from the returned object */ + + acpi_ut_copy_id_string(next_id_string, + cid_objects[i]->string.pointer); + length = cid_objects[i]->string.length + 1; + } + + cid_list->ids[i].string = next_id_string; + cid_list->ids[i].length = length; + next_id_string += length; + } + + /* Finish the CID list */ + + cid_list->count = count; + cid_list->list_size = cid_list_size; + *return_cid_list = cid_list; + +cleanup: + + /* On exit, we must delete the _CID return object */ + + acpi_ut_remove_reference(obj_desc); + return_ACPI_STATUS(status); +} diff --git a/drivers/acpi/acpica/utmisc.c b/drivers/acpi/acpica/utmisc.c index fbe782348b0b..9cd65334ca75 100644 --- a/drivers/acpi/acpica/utmisc.c +++ b/drivers/acpi/acpica/utmisc.c @@ -118,6 +118,34 @@ const char *acpi_ut_validate_exception(acpi_status status) return (ACPI_CAST_PTR(const char, exception)); } +/******************************************************************************* + * + * FUNCTION: acpi_ut_is_pci_root_bridge + * + * PARAMETERS: Id - The HID/CID in string format + * + * RETURN: TRUE if the Id is a match for a PCI/PCI-Express Root Bridge + * + * DESCRIPTION: Determine if the input ID is a PCI Root Bridge ID. + * + ******************************************************************************/ + +u8 acpi_ut_is_pci_root_bridge(char *id) +{ + + /* + * Check if this is a PCI root bridge. + * ACPI 3.0+: check for a PCI Express root also. + */ + if (!(ACPI_STRCMP(id, + PCI_ROOT_HID_STRING)) || + !(ACPI_STRCMP(id, PCI_EXPRESS_ROOT_HID_STRING))) { + return (TRUE); + } + + return (FALSE); +} + /******************************************************************************* * * FUNCTION: acpi_ut_is_aml_table diff --git a/drivers/acpi/container.c b/drivers/acpi/container.c index fe0cdf83641a..2aee8c24dc56 100644 --- a/drivers/acpi/container.c +++ b/drivers/acpi/container.c @@ -200,20 +200,17 @@ container_walk_namespace_cb(acpi_handle handle, u32 lvl, void *context, void **rv) { char *hid = NULL; - struct acpi_buffer buffer = { ACPI_ALLOCATE_BUFFER, NULL }; struct acpi_device_info *info; acpi_status status; int *action = context; - - status = acpi_get_object_info(handle, &buffer); - if (ACPI_FAILURE(status) || !buffer.pointer) { + status = acpi_get_object_info(handle, &info); + if (ACPI_FAILURE(status)) { return AE_OK; } - info = buffer.pointer; if (info->valid & ACPI_VALID_HID) - hid = info->hardware_id.value; + hid = info->hardware_id.string; if (hid == NULL) { goto end; @@ -240,7 +237,7 @@ container_walk_namespace_cb(acpi_handle handle, } end: - kfree(buffer.pointer); + kfree(info); return AE_OK; } diff --git a/drivers/acpi/dock.c b/drivers/acpi/dock.c index efb959d6c8a9..39536b80bce7 100644 --- a/drivers/acpi/dock.c +++ b/drivers/acpi/dock.c @@ -231,18 +231,16 @@ static int is_ata(acpi_handle handle) static int is_battery(acpi_handle handle) { struct acpi_device_info *info; - struct acpi_buffer buffer = {ACPI_ALLOCATE_BUFFER, NULL}; int ret = 1; - if (!ACPI_SUCCESS(acpi_get_object_info(handle, &buffer))) + if (!ACPI_SUCCESS(acpi_get_object_info(handle, &info))) return 0; - info = buffer.pointer; if (!(info->valid & ACPI_VALID_HID)) ret = 0; else - ret = !strcmp("PNP0C0A", info->hardware_id.value); + ret = !strcmp("PNP0C0A", info->hardware_id.string); - kfree(buffer.pointer); + kfree(info); return ret; } diff --git a/drivers/acpi/glue.c b/drivers/acpi/glue.c index a8a5c29958c8..27a7072347ea 100644 --- a/drivers/acpi/glue.c +++ b/drivers/acpi/glue.c @@ -93,15 +93,13 @@ do_acpi_find_child(acpi_handle handle, u32 lvl, void *context, void **rv) { acpi_status status; struct acpi_device_info *info; - struct acpi_buffer buffer = { ACPI_ALLOCATE_BUFFER, NULL }; struct acpi_find_child *find = context; - status = acpi_get_object_info(handle, &buffer); + status = acpi_get_object_info(handle, &info); if (ACPI_SUCCESS(status)) { - info = buffer.pointer; if (info->address == find->address) find->handle = handle; - kfree(buffer.pointer); + kfree(info); } return AE_OK; } diff --git a/drivers/acpi/scan.c b/drivers/acpi/scan.c index 781435d7e369..0ab526de7c55 100644 --- a/drivers/acpi/scan.c +++ b/drivers/acpi/scan.c @@ -60,13 +60,13 @@ static int create_modalias(struct acpi_device *acpi_dev, char *modalias, } if (acpi_dev->flags.compatible_ids) { - struct acpi_compatible_id_list *cid_list; + struct acpica_device_id_list *cid_list; int i; cid_list = acpi_dev->pnp.cid_list; for (i = 0; i < cid_list->count; i++) { count = snprintf(&modalias[len], size, "%s:", - cid_list->id[i].value); + cid_list->ids[i].string); if (count < 0 || count >= size) { printk(KERN_ERR PREFIX "%s cid[%i] exceeds event buffer size", acpi_dev->pnp.device_name, i); @@ -287,14 +287,14 @@ int acpi_match_device_ids(struct acpi_device *device, } if (device->flags.compatible_ids) { - struct acpi_compatible_id_list *cid_list = device->pnp.cid_list; + struct acpica_device_id_list *cid_list = device->pnp.cid_list; int i; for (id = ids; id->id[0]; id++) { /* compare multiple _CID entries against driver ids */ for (i = 0; i < cid_list->count; i++) { if (!strcmp((char*)id->id, - cid_list->id[i].value)) + cid_list->ids[i].string)) return 0; } } @@ -999,33 +999,89 @@ static int acpi_dock_match(struct acpi_device *device) return acpi_get_handle(device->handle, "_DCK", &tmp); } +static struct acpica_device_id_list* +acpi_add_cid( + struct acpi_device_info *info, + struct acpica_device_id *new_cid) +{ + struct acpica_device_id_list *cid; + char *next_id_string; + acpi_size cid_length; + acpi_size new_cid_length; + u32 i; + + + /* Allocate new CID list with room for the new CID */ + + if (!new_cid) + new_cid_length = info->compatible_id_list.list_size; + else if (info->compatible_id_list.list_size) + new_cid_length = info->compatible_id_list.list_size + + new_cid->length + sizeof(struct acpica_device_id); + else + new_cid_length = sizeof(struct acpica_device_id_list) + new_cid->length; + + cid = ACPI_ALLOCATE_ZEROED(new_cid_length); + if (!cid) { + return NULL; + } + + cid->list_size = new_cid_length; + cid->count = info->compatible_id_list.count; + if (new_cid) + cid->count++; + next_id_string = (char *) cid->ids + (cid->count * sizeof(struct acpica_device_id)); + + /* Copy all existing CIDs */ + + for (i = 0; i < info->compatible_id_list.count; i++) { + cid_length = info->compatible_id_list.ids[i].length; + cid->ids[i].string = next_id_string; + cid->ids[i].length = cid_length; + + ACPI_MEMCPY(next_id_string, info->compatible_id_list.ids[i].string, + cid_length); + + next_id_string += cid_length; + } + + /* Append the new CID */ + + if (new_cid) { + cid->ids[i].string = next_id_string; + cid->ids[i].length = new_cid->length; + + ACPI_MEMCPY(next_id_string, new_cid->string, new_cid->length); + } + + return cid; +} + static void acpi_device_set_id(struct acpi_device *device, struct acpi_device *parent, acpi_handle handle, int type) { - struct acpi_device_info *info; - struct acpi_buffer buffer = { ACPI_ALLOCATE_BUFFER, NULL }; + struct acpi_device_info *info = NULL; char *hid = NULL; char *uid = NULL; - struct acpi_compatible_id_list *cid_list = NULL; - const char *cid_add = NULL; + struct acpica_device_id_list *cid_list = NULL; + char *cid_add = NULL; acpi_status status; switch (type) { case ACPI_BUS_TYPE_DEVICE: - status = acpi_get_object_info(handle, &buffer); + status = acpi_get_object_info(handle, &info); if (ACPI_FAILURE(status)) { printk(KERN_ERR PREFIX "%s: Error reading device info\n", __func__); return; } - info = buffer.pointer; if (info->valid & ACPI_VALID_HID) - hid = info->hardware_id.value; + hid = info->hardware_id.string; if (info->valid & ACPI_VALID_UID) - uid = info->unique_id.value; + uid = info->unique_id.string; if (info->valid & ACPI_VALID_CID) - cid_list = &info->compatibility_id; + cid_list = &info->compatible_id_list; if (info->valid & ACPI_VALID_ADR) { device->pnp.bus_address = info->address; device->flags.bus_address = 1; @@ -1076,55 +1132,44 @@ static void acpi_device_set_id(struct acpi_device *device, } if (hid) { - strcpy(device->pnp.hardware_id, hid); - device->flags.hardware_id = 1; - } + device->pnp.hardware_id = ACPI_ALLOCATE_ZEROED(strlen (hid) + 1); + if (device->pnp.hardware_id) { + strcpy(device->pnp.hardware_id, hid); + device->flags.hardware_id = 1; + } + } else + device->pnp.hardware_id = NULL; + if (uid) { - strcpy(device->pnp.unique_id, uid); - device->flags.unique_id = 1; - } + device->pnp.unique_id = ACPI_ALLOCATE_ZEROED(strlen (uid) + 1); + if (device->pnp.unique_id) { + strcpy(device->pnp.unique_id, uid); + device->flags.unique_id = 1; + } + } else + device->pnp.unique_id = NULL; + if (cid_list || cid_add) { - struct acpi_compatible_id_list *list; - int size = 0; - int count = 0; - - if (cid_list) { - size = cid_list->size; - } else if (cid_add) { - size = sizeof(struct acpi_compatible_id_list); - cid_list = ACPI_ALLOCATE_ZEROED((acpi_size) size); - if (!cid_list) { - printk(KERN_ERR "Memory allocation error\n"); - kfree(buffer.pointer); - return; - } else { - cid_list->count = 0; - cid_list->size = size; - } + struct acpica_device_id_list *list; + + if (cid_add) { + struct acpica_device_id cid; + cid.length = strlen (cid_add) + 1; + cid.string = cid_add; + + list = acpi_add_cid(info, &cid); + } else { + list = acpi_add_cid(info, NULL); } - if (cid_add) - size += sizeof(struct acpi_compatible_id); - list = kmalloc(size, GFP_KERNEL); if (list) { - if (cid_list) { - memcpy(list, cid_list, cid_list->size); - count = cid_list->count; - } - if (cid_add) { - strncpy(list->id[count].value, cid_add, - ACPI_MAX_CID_LENGTH); - count++; - device->flags.compatible_ids = 1; - } - list->size = size; - list->count = count; device->pnp.cid_list = list; - } else - printk(KERN_ERR PREFIX "Memory allocation error\n"); + if (cid_add) + device->flags.compatible_ids = 1; + } } - kfree(buffer.pointer); + kfree(info); } static int acpi_device_set_context(struct acpi_device *device, int type) diff --git a/drivers/char/agp/hp-agp.c b/drivers/char/agp/hp-agp.c index 8f3d4c184914..7bead4c816ca 100644 --- a/drivers/char/agp/hp-agp.c +++ b/drivers/char/agp/hp-agp.c @@ -478,7 +478,6 @@ zx1_gart_probe (acpi_handle obj, u32 depth, void *context, void **ret) { acpi_handle handle, parent; acpi_status status; - struct acpi_buffer buffer; struct acpi_device_info *info; u64 lba_hpa, sba_hpa, length; int match; @@ -490,13 +489,11 @@ zx1_gart_probe (acpi_handle obj, u32 depth, void *context, void **ret) /* Look for an enclosing IOC scope and find its CSR space */ handle = obj; do { - buffer.length = ACPI_ALLOCATE_LOCAL_BUFFER; - status = acpi_get_object_info(handle, &buffer); + status = acpi_get_object_info(handle, &info); if (ACPI_SUCCESS(status)) { /* TBD check _CID also */ - info = buffer.pointer; - info->hardware_id.value[sizeof(info->hardware_id)-1] = '\0'; - match = (strcmp(info->hardware_id.value, "HWP0001") == 0); + info->hardware_id.string[sizeof(info->hardware_id.length)-1] = '\0'; + match = (strcmp(info->hardware_id.string, "HWP0001") == 0); kfree(info); if (match) { status = hp_acpi_csr_space(handle, &sba_hpa, &length); diff --git a/drivers/ide/ide-acpi.c b/drivers/ide/ide-acpi.c index c509c9916464..c0cf45a11b93 100644 --- a/drivers/ide/ide-acpi.c +++ b/drivers/ide/ide-acpi.c @@ -114,8 +114,6 @@ static int ide_get_dev_handle(struct device *dev, acpi_handle *handle, unsigned int bus, devnum, func; acpi_integer addr; acpi_handle dev_handle; - struct acpi_buffer buffer = {.length = ACPI_ALLOCATE_BUFFER, - .pointer = NULL}; acpi_status status; struct acpi_device_info *dinfo = NULL; int ret = -ENODEV; @@ -134,12 +132,11 @@ static int ide_get_dev_handle(struct device *dev, acpi_handle *handle, goto err; } - status = acpi_get_object_info(dev_handle, &buffer); + status = acpi_get_object_info(dev_handle, &dinfo); if (ACPI_FAILURE(status)) { DEBPRINT("get_object_info for device failed\n"); goto err; } - dinfo = buffer.pointer; if (dinfo && (dinfo->valid & ACPI_VALID_ADR) && dinfo->address == addr) { *pcidevfn = addr; diff --git a/drivers/pci/hotplug/acpiphp_ibm.c b/drivers/pci/hotplug/acpiphp_ibm.c index 5befa7e379b7..a9d926b7d805 100644 --- a/drivers/pci/hotplug/acpiphp_ibm.c +++ b/drivers/pci/hotplug/acpiphp_ibm.c @@ -398,23 +398,21 @@ static acpi_status __init ibm_find_acpi_device(acpi_handle handle, acpi_handle *phandle = (acpi_handle *)context; acpi_status status; struct acpi_device_info *info; - struct acpi_buffer info_buffer = { ACPI_ALLOCATE_BUFFER, NULL }; int retval = 0; - status = acpi_get_object_info(handle, &info_buffer); + status = acpi_get_object_info(handle, &info); if (ACPI_FAILURE(status)) { err("%s: Failed to get device information status=0x%x\n", __func__, status); return retval; } - info = info_buffer.pointer; - info->hardware_id.value[sizeof(info->hardware_id.value) - 1] = '\0'; + info->hardware_id.string[sizeof(info->hardware_id.length) - 1] = '\0'; if (info->current_status && (info->valid & ACPI_VALID_HID) && - (!strcmp(info->hardware_id.value, IBM_HARDWARE_ID1) || - !strcmp(info->hardware_id.value, IBM_HARDWARE_ID2))) { + (!strcmp(info->hardware_id.string, IBM_HARDWARE_ID1) || + !strcmp(info->hardware_id.string, IBM_HARDWARE_ID2))) { dbg("found hardware: %s, handle: %p\n", - info->hardware_id.value, handle); + info->hardware_id.string, handle); *phandle = handle; /* returning non-zero causes the search to stop * and returns this value to the caller of diff --git a/drivers/platform/x86/sony-laptop.c b/drivers/platform/x86/sony-laptop.c index dafaa4a92df5..f9f68e0e7344 100644 --- a/drivers/platform/x86/sony-laptop.c +++ b/drivers/platform/x86/sony-laptop.c @@ -976,15 +976,12 @@ static acpi_status sony_walk_callback(acpi_handle handle, u32 level, void *context, void **return_value) { struct acpi_device_info *info; - struct acpi_buffer buffer = {ACPI_ALLOCATE_BUFFER, NULL}; - - if (ACPI_SUCCESS(acpi_get_object_info(handle, &buffer))) { - info = buffer.pointer; + if (ACPI_SUCCESS(acpi_get_object_info(handle, &info))) { printk(KERN_WARNING DRV_PFX "method: name: %4.4s, args %X\n", (char *)&info->name, info->param_count); - kfree(buffer.pointer); + kfree(info); } return AE_OK; diff --git a/drivers/pnp/pnpacpi/core.c b/drivers/pnp/pnpacpi/core.c index 9496494f340e..c07fdb94d665 100644 --- a/drivers/pnp/pnpacpi/core.c +++ b/drivers/pnp/pnpacpi/core.c @@ -194,13 +194,13 @@ static int __init pnpacpi_add_device(struct acpi_device *device) pnpacpi_parse_resource_option_data(dev); if (device->flags.compatible_ids) { - struct acpi_compatible_id_list *cid_list = device->pnp.cid_list; + struct acpica_device_id_list *cid_list = device->pnp.cid_list; int i; for (i = 0; i < cid_list->count; i++) { - if (!ispnpidacpi(cid_list->id[i].value)) + if (!ispnpidacpi(cid_list->ids[i].string)) continue; - pnp_add_id(dev, cid_list->id[i].value); + pnp_add_id(dev, cid_list->ids[i].string); } } diff --git a/include/acpi/acpi_bus.h b/include/acpi/acpi_bus.h index c65e4ce6c3af..b91420b52c6f 100644 --- a/include/acpi/acpi_bus.h +++ b/include/acpi/acpi_bus.h @@ -173,17 +173,15 @@ struct acpi_device_dir { typedef char acpi_bus_id[8]; typedef unsigned long acpi_bus_address; -typedef char acpi_hardware_id[15]; -typedef char acpi_unique_id[9]; typedef char acpi_device_name[40]; typedef char acpi_device_class[20]; struct acpi_device_pnp { acpi_bus_id bus_id; /* Object name */ acpi_bus_address bus_address; /* _ADR */ - acpi_hardware_id hardware_id; /* _HID */ - struct acpi_compatible_id_list *cid_list; /* _CIDs */ - acpi_unique_id unique_id; /* _UID */ + char *hardware_id; /* _HID */ + struct acpica_device_id_list *cid_list; /* _CIDs */ + char *unique_id; /* _UID */ acpi_device_name device_name; /* Driver-determined */ acpi_device_class device_class; /* " */ }; diff --git a/include/acpi/acpixf.h b/include/acpi/acpixf.h index b450a195319a..04904c7f1aa1 100644 --- a/include/acpi/acpixf.h +++ b/include/acpi/acpixf.h @@ -200,7 +200,8 @@ acpi_evaluate_object_typed(acpi_handle object, acpi_object_type return_type); acpi_status -acpi_get_object_info(acpi_handle handle, struct acpi_buffer *return_buffer); +acpi_get_object_info(acpi_handle handle, + struct acpi_device_info **return_buffer); acpi_status acpi_install_method(u8 *buffer); diff --git a/include/acpi/actypes.h b/include/acpi/actypes.h index 37ba576d06e8..7a4ff79e238c 100644 --- a/include/acpi/actypes.h +++ b/include/acpi/actypes.h @@ -338,7 +338,7 @@ typedef u32 acpi_physical_address; /* PM Timer ticks per second (HZ) */ -#define PM_TIMER_FREQUENCY 3579545 +#define PM_TIMER_FREQUENCY 3579545 /******************************************************************************* * @@ -969,38 +969,60 @@ acpi_status(*acpi_walk_callback) (acpi_handle obj_handle, #define ACPI_INTERRUPT_NOT_HANDLED 0x00 #define ACPI_INTERRUPT_HANDLED 0x01 -/* Length of _HID, _UID, _CID, and UUID values */ +/* Length of 32-bit EISAID values when converted back to a string */ + +#define ACPI_EISAID_STRING_SIZE 8 /* Includes null terminator */ + +/* Length of UUID (string) values */ -#define ACPI_DEVICE_ID_LENGTH 0x09 -#define ACPI_MAX_CID_LENGTH 48 #define ACPI_UUID_LENGTH 16 -/* Common string version of device HIDs and UIDs */ +/* Structures used for device/processor HID, UID, CID */ struct acpica_device_id { - char value[ACPI_DEVICE_ID_LENGTH]; + u32 length; /* Length of string + null */ + char *string; }; -/* Common string version of device CIDs */ - -struct acpi_compatible_id { - char value[ACPI_MAX_CID_LENGTH]; +struct acpica_device_id_list { + u32 count; /* Number of IDs in Ids array */ + u32 list_size; /* Size of list, including ID strings */ + struct acpica_device_id ids[1]; /* ID array */ }; -struct acpi_compatible_id_list { - u32 count; - u32 size; - struct acpi_compatible_id id[1]; +/* + * Structure returned from acpi_get_object_info. + * Optimized for both 32- and 64-bit builds + */ +struct acpi_device_info { + u32 info_size; /* Size of info, including ID strings */ + u32 name; /* ACPI object Name */ + acpi_object_type type; /* ACPI object Type */ + u8 param_count; /* If a method, required parameter count */ + u8 valid; /* Indicates which optional fields are valid */ + u8 flags; /* Miscellaneous info */ + u8 highest_dstates[4]; /* _sx_d values: 0xFF indicates not valid */ + u8 lowest_dstates[5]; /* _sx_w values: 0xFF indicates not valid */ + u32 current_status; /* _STA value */ + acpi_integer address; /* _ADR value */ + struct acpica_device_id hardware_id; /* _HID value */ + struct acpica_device_id unique_id; /* _UID value */ + struct acpica_device_id_list compatible_id_list; /* _CID list */ }; -/* Structure and flags for acpi_get_object_info */ +/* Values for Flags field above (acpi_get_object_info) */ + +#define ACPI_PCI_ROOT_BRIDGE 0x01 -#define ACPI_VALID_STA 0x0001 -#define ACPI_VALID_ADR 0x0002 -#define ACPI_VALID_HID 0x0004 -#define ACPI_VALID_UID 0x0008 -#define ACPI_VALID_CID 0x0010 -#define ACPI_VALID_SXDS 0x0020 +/* Flags for Valid field above (acpi_get_object_info) */ + +#define ACPI_VALID_STA 0x01 +#define ACPI_VALID_ADR 0x02 +#define ACPI_VALID_HID 0x04 +#define ACPI_VALID_UID 0x08 +#define ACPI_VALID_CID 0x10 +#define ACPI_VALID_SXDS 0x20 +#define ACPI_VALID_SXWS 0x40 /* Flags for _STA method */ @@ -1011,29 +1033,6 @@ struct acpi_compatible_id_list { #define ACPI_STA_DEVICE_OK 0x08 /* Synonym */ #define ACPI_STA_BATTERY_PRESENT 0x10 -#define ACPI_COMMON_OBJ_INFO \ - acpi_object_type type; /* ACPI object type */ \ - acpi_name name /* ACPI object Name */ - -struct acpi_obj_info_header { - ACPI_COMMON_OBJ_INFO; -}; - -/* Structure returned from Get Object Info */ - -struct acpi_device_info { - ACPI_COMMON_OBJ_INFO; - - u32 param_count; /* If a method, required parameter count */ - u32 valid; /* Indicates which fields below are valid */ - u32 current_status; /* _STA value */ - acpi_integer address; /* _ADR value if any */ - struct acpica_device_id hardware_id; /* _HID value if any */ - struct acpica_device_id unique_id; /* _UID value if any */ - u8 highest_dstates[4]; /* _sx_d values: 0xFF indicates not valid */ - struct acpi_compatible_id_list compatibility_id; /* List of _CIDs if any */ -}; - /* Context structs for address space handlers */ struct acpi_pci_id { -- cgit v1.2.3 From 6557a49a443a347d24aed58076365432ded30edc Mon Sep 17 00:00:00 2001 From: Lin Ming Date: Wed, 24 Jun 2009 11:32:04 +0800 Subject: ACPICA: ACPI 4.0: Interpreter support for IPMI. Adds support for IPMI which is similar to SMBus and uses a bi-directional data buffer. ACPICA BZ 773. http://acpica.org/bugzilla/show_bug.cgi?id=773 Signed-off-by: Lin Ming Signed-off-by: Bob Moore Signed-off-by: Len Brown --- drivers/acpi/acpica/acconfig.h | 3 +- drivers/acpi/acpica/exfield.c | 82 ++++++++++++++++++++++++++++-------------- drivers/acpi/acpica/exfldio.c | 7 ++-- include/acpi/actypes.h | 3 +- 4 files changed, 63 insertions(+), 32 deletions(-) (limited to 'include') diff --git a/drivers/acpi/acpica/acconfig.h b/drivers/acpi/acpica/acconfig.h index 9123d5a11627..8e679ef5b231 100644 --- a/drivers/acpi/acpica/acconfig.h +++ b/drivers/acpi/acpica/acconfig.h @@ -199,9 +199,10 @@ #define ACPI_RSDP_CHECKSUM_LENGTH 20 #define ACPI_RSDP_XCHECKSUM_LENGTH 36 -/* SMBus bidirectional buffer size */ +/* SMBus and IPMI bidirectional buffer size */ #define ACPI_SMBUS_BUFFER_SIZE 34 +#define ACPI_IPMI_BUFFER_SIZE 66 /* _sx_d and _sx_w control methods */ diff --git a/drivers/acpi/acpica/exfield.c b/drivers/acpi/acpica/exfield.c index 546dcdd86785..0b33d6c887b9 100644 --- a/drivers/acpi/acpica/exfield.c +++ b/drivers/acpi/acpica/exfield.c @@ -72,6 +72,7 @@ acpi_ex_read_data_from_field(struct acpi_walk_state *walk_state, union acpi_operand_object *buffer_desc; acpi_size length; void *buffer; + u32 function; ACPI_FUNCTION_TRACE_PTR(ex_read_data_from_field, obj_desc); @@ -97,13 +98,27 @@ acpi_ex_read_data_from_field(struct acpi_walk_state *walk_state, } } else if ((obj_desc->common.type == ACPI_TYPE_LOCAL_REGION_FIELD) && (obj_desc->field.region_obj->region.space_id == - ACPI_ADR_SPACE_SMBUS)) { + ACPI_ADR_SPACE_SMBUS + || obj_desc->field.region_obj->region.space_id == + ACPI_ADR_SPACE_IPMI)) { /* - * This is an SMBus read. We must create a buffer to hold the data - * and directly access the region handler. + * This is an SMBus or IPMI read. We must create a buffer to hold + * the data and then directly access the region handler. + * + * Note: Smbus protocol value is passed in upper 16-bits of Function */ - buffer_desc = - acpi_ut_create_buffer_object(ACPI_SMBUS_BUFFER_SIZE); + if (obj_desc->field.region_obj->region.space_id == + ACPI_ADR_SPACE_SMBUS) { + length = ACPI_SMBUS_BUFFER_SIZE; + function = + ACPI_READ | (obj_desc->field.attribute << 16); + } else { /* IPMI */ + + length = ACPI_IPMI_BUFFER_SIZE; + function = ACPI_READ; + } + + buffer_desc = acpi_ut_create_buffer_object(length); if (!buffer_desc) { return_ACPI_STATUS(AE_NO_MEMORY); } @@ -112,16 +127,13 @@ acpi_ex_read_data_from_field(struct acpi_walk_state *walk_state, acpi_ex_acquire_global_lock(obj_desc->common_field.field_flags); - /* - * Perform the read. - * Note: Smbus protocol value is passed in upper 16-bits of Function - */ + /* Call the region handler for the read */ + status = acpi_ex_access_region(obj_desc, 0, ACPI_CAST_PTR(acpi_integer, buffer_desc-> buffer.pointer), - ACPI_READ | (obj_desc->field. - attribute << 16)); + function); acpi_ex_release_global_lock(obj_desc->common_field.field_flags); goto exit; } @@ -212,6 +224,7 @@ acpi_ex_write_data_to_field(union acpi_operand_object *source_desc, u32 length; void *buffer; union acpi_operand_object *buffer_desc; + u32 function; ACPI_FUNCTION_TRACE_PTR(ex_write_data_to_field, obj_desc); @@ -234,39 +247,56 @@ acpi_ex_write_data_to_field(union acpi_operand_object *source_desc, } } else if ((obj_desc->common.type == ACPI_TYPE_LOCAL_REGION_FIELD) && (obj_desc->field.region_obj->region.space_id == - ACPI_ADR_SPACE_SMBUS)) { + ACPI_ADR_SPACE_SMBUS + || obj_desc->field.region_obj->region.space_id == + ACPI_ADR_SPACE_IPMI)) { /* - * This is an SMBus write. We will bypass the entire field mechanism - * and handoff the buffer directly to the handler. + * This is an SMBus or IPMI write. We will bypass the entire field + * mechanism and handoff the buffer directly to the handler. For + * these address spaces, the buffer is bi-directional; on a write, + * return data is returned in the same buffer. + * + * Source must be a buffer of sufficient size: + * ACPI_SMBUS_BUFFER_SIZE or ACPI_IPMI_BUFFER_SIZE. * - * Source must be a buffer of sufficient size (ACPI_SMBUS_BUFFER_SIZE). + * Note: SMBus protocol type is passed in upper 16-bits of Function */ if (source_desc->common.type != ACPI_TYPE_BUFFER) { ACPI_ERROR((AE_INFO, - "SMBus write requires Buffer, found type %s", + "SMBus or IPMI write requires Buffer, found type %s", acpi_ut_get_object_type_name(source_desc))); return_ACPI_STATUS(AE_AML_OPERAND_TYPE); } - if (source_desc->buffer.length < ACPI_SMBUS_BUFFER_SIZE) { + if (obj_desc->field.region_obj->region.space_id == + ACPI_ADR_SPACE_SMBUS) { + length = ACPI_SMBUS_BUFFER_SIZE; + function = + ACPI_WRITE | (obj_desc->field.attribute << 16); + } else { /* IPMI */ + + length = ACPI_IPMI_BUFFER_SIZE; + function = ACPI_WRITE; + } + + if (source_desc->buffer.length < length) { ACPI_ERROR((AE_INFO, - "SMBus write requires Buffer of length %X, found length %X", - ACPI_SMBUS_BUFFER_SIZE, - source_desc->buffer.length)); + "SMBus or IPMI write requires Buffer of length %X, found length %X", + length, source_desc->buffer.length)); return_ACPI_STATUS(AE_AML_BUFFER_LIMIT); } - buffer_desc = - acpi_ut_create_buffer_object(ACPI_SMBUS_BUFFER_SIZE); + /* Create the bi-directional buffer */ + + buffer_desc = acpi_ut_create_buffer_object(length); if (!buffer_desc) { return_ACPI_STATUS(AE_NO_MEMORY); } buffer = buffer_desc->buffer.pointer; - ACPI_MEMCPY(buffer, source_desc->buffer.pointer, - ACPI_SMBUS_BUFFER_SIZE); + ACPI_MEMCPY(buffer, source_desc->buffer.pointer, length); /* Lock entire transaction if requested */ @@ -275,12 +305,10 @@ acpi_ex_write_data_to_field(union acpi_operand_object *source_desc, /* * Perform the write (returns status and perhaps data in the * same buffer) - * Note: SMBus protocol type is passed in upper 16-bits of Function. */ status = acpi_ex_access_region(obj_desc, 0, (acpi_integer *) buffer, - ACPI_WRITE | (obj_desc->field. - attribute << 16)); + function); acpi_ex_release_global_lock(obj_desc->common_field.field_flags); *result_desc = buffer_desc; diff --git a/drivers/acpi/acpica/exfldio.c b/drivers/acpi/acpica/exfldio.c index 6687be167f5f..d7b3b418fb45 100644 --- a/drivers/acpi/acpica/exfldio.c +++ b/drivers/acpi/acpica/exfldio.c @@ -120,12 +120,13 @@ acpi_ex_setup_region(union acpi_operand_object *obj_desc, } /* - * Exit now for SMBus address space, it has a non-linear address space + * Exit now for SMBus or IPMI address space, it has a non-linear address space * and the request cannot be directly validated */ - if (rgn_desc->region.space_id == ACPI_ADR_SPACE_SMBUS) { + if (rgn_desc->region.space_id == ACPI_ADR_SPACE_SMBUS || + rgn_desc->region.space_id == ACPI_ADR_SPACE_IPMI) { - /* SMBus has a non-linear address space */ + /* SMBus or IPMI has a non-linear address space */ return_ACPI_STATUS(AE_OK); } diff --git a/include/acpi/actypes.h b/include/acpi/actypes.h index 7a4ff79e238c..4371805d2def 100644 --- a/include/acpi/actypes.h +++ b/include/acpi/actypes.h @@ -732,7 +732,8 @@ typedef u8 acpi_adr_space_type; #define ACPI_ADR_SPACE_SMBUS (acpi_adr_space_type) 4 #define ACPI_ADR_SPACE_CMOS (acpi_adr_space_type) 5 #define ACPI_ADR_SPACE_PCI_BAR_TARGET (acpi_adr_space_type) 6 -#define ACPI_ADR_SPACE_DATA_TABLE (acpi_adr_space_type) 7 +#define ACPI_ADR_SPACE_IPMI (acpi_adr_space_type) 7 +#define ACPI_ADR_SPACE_DATA_TABLE (acpi_adr_space_type) 8 #define ACPI_ADR_SPACE_FIXED_HARDWARE (acpi_adr_space_type) 127 /* -- cgit v1.2.3 From 8e4319c425077c4cc540696a5bb6c4d12f017dcd Mon Sep 17 00:00:00 2001 From: Bob Moore Date: Mon, 29 Jun 2009 13:43:27 +0800 Subject: ACPICA: Fix several acpi_attach_data problems Handler was never invoked. Now invoked if/when host node is deleted. Data object was not automatically deleted when host node was deleted. Interface to handler had an unused parameter, removed it. ACPICA BZ 778. http://acpica.org/bugzilla/show_bug.cgi?id=778 Signed-off-by: Bob Moore Signed-off-by: Lin Ming Signed-off-by: Len Brown --- drivers/acpi/acpica/acnamesp.h | 2 + drivers/acpi/acpica/nsalloc.c | 88 +++++++++++++++++++++++++++++------------- drivers/acpi/acpica/nsload.c | 3 +- drivers/acpi/bus.c | 2 +- drivers/acpi/glue.c | 2 +- drivers/acpi/scan.c | 2 +- include/acpi/acpi_bus.h | 4 +- include/acpi/actypes.h | 2 +- 8 files changed, 70 insertions(+), 35 deletions(-) (limited to 'include') diff --git a/drivers/acpi/acpica/acnamesp.h b/drivers/acpi/acpica/acnamesp.h index 94cdc2b8cb93..a78e02f62d5e 100644 --- a/drivers/acpi/acpica/acnamesp.h +++ b/drivers/acpi/acpica/acnamesp.h @@ -144,6 +144,8 @@ struct acpi_namespace_node *acpi_ns_create_node(u32 name); void acpi_ns_delete_node(struct acpi_namespace_node *node); +void acpi_ns_remove_node(struct acpi_namespace_node *node); + void acpi_ns_delete_namespace_subtree(struct acpi_namespace_node *parent_handle); diff --git a/drivers/acpi/acpica/nsalloc.c b/drivers/acpi/acpica/nsalloc.c index efc971ab7d65..8a58a1b85aa0 100644 --- a/drivers/acpi/acpica/nsalloc.c +++ b/drivers/acpi/acpica/nsalloc.c @@ -96,17 +96,68 @@ struct acpi_namespace_node *acpi_ns_create_node(u32 name) * * RETURN: None * - * DESCRIPTION: Delete a namespace node + * DESCRIPTION: Delete a namespace node. All node deletions must come through + * here. Detaches any attached objects, including any attached + * data. If a handler is associated with attached data, it is + * invoked before the node is deleted. * ******************************************************************************/ void acpi_ns_delete_node(struct acpi_namespace_node *node) +{ + union acpi_operand_object *obj_desc; + + ACPI_FUNCTION_NAME(ns_delete_node); + + /* Detach an object if there is one */ + + acpi_ns_detach_object(node); + + /* + * Delete an attached data object if present (an object that was created + * and attached via acpi_attach_data). Note: After any normal object is + * detached above, the only possible remaining object is a data object. + */ + obj_desc = node->object; + if (obj_desc && (obj_desc->common.type == ACPI_TYPE_LOCAL_DATA)) { + + /* Invoke the attached data deletion handler if present */ + + if (obj_desc->data.handler) { + obj_desc->data.handler(node, obj_desc->data.pointer); + } + + acpi_ut_remove_reference(obj_desc); + } + + /* Now we can delete the node */ + + (void)acpi_os_release_object(acpi_gbl_namespace_cache, node); + + ACPI_MEM_TRACKING(acpi_gbl_ns_node_list->total_freed++); + ACPI_DEBUG_PRINT((ACPI_DB_ALLOCATIONS, "Node %p, Remaining %X\n", + node, acpi_gbl_current_node_count)); +} + +/******************************************************************************* + * + * FUNCTION: acpi_ns_remove_node + * + * PARAMETERS: Node - Node to be removed/deleted + * + * RETURN: None + * + * DESCRIPTION: Remove (unlink) and delete a namespace node + * + ******************************************************************************/ + +void acpi_ns_remove_node(struct acpi_namespace_node *node) { struct acpi_namespace_node *parent_node; struct acpi_namespace_node *prev_node; struct acpi_namespace_node *next_node; - ACPI_FUNCTION_TRACE_PTR(ns_delete_node, node); + ACPI_FUNCTION_TRACE_PTR(ns_remove_node, node); parent_node = acpi_ns_get_parent_node(node); @@ -142,12 +193,9 @@ void acpi_ns_delete_node(struct acpi_namespace_node *node) } } - ACPI_MEM_TRACKING(acpi_gbl_ns_node_list->total_freed++); - - /* Detach an object if there is one, then delete the node */ + /* Delete the node and any attached objects */ - acpi_ns_detach_object(node); - (void)acpi_os_release_object(acpi_gbl_namespace_cache, node); + acpi_ns_delete_node(node); return_VOID; } @@ -273,25 +321,11 @@ void acpi_ns_delete_children(struct acpi_namespace_node *parent_node) parent_node, child_node)); } - /* Now we can free this child object */ - - ACPI_MEM_TRACKING(acpi_gbl_ns_node_list->total_freed++); - - ACPI_DEBUG_PRINT((ACPI_DB_ALLOCATIONS, - "Object %p, Remaining %X\n", child_node, - acpi_gbl_current_node_count)); - - /* Detach an object if there is one, then free the child node */ - - acpi_ns_detach_object(child_node); - - /* Now we can delete the node */ - - (void)acpi_os_release_object(acpi_gbl_namespace_cache, - child_node); - - /* And move on to the next child in the list */ - + /* + * Delete this child node and move on to the next child in the list. + * No need to unlink the node since we are deleting the entire branch. + */ + acpi_ns_delete_node(child_node); child_node = next_node; } while (!(flags & ANOBJ_END_OF_PEER_LIST)); @@ -433,7 +467,7 @@ void acpi_ns_delete_namespace_by_owner(acpi_owner_id owner_id) if (deletion_node) { acpi_ns_delete_children(deletion_node); - acpi_ns_delete_node(deletion_node); + acpi_ns_remove_node(deletion_node); deletion_node = NULL; } diff --git a/drivers/acpi/acpica/nsload.c b/drivers/acpi/acpica/nsload.c index dcd7a6adbbbc..a7234e60e985 100644 --- a/drivers/acpi/acpica/nsload.c +++ b/drivers/acpi/acpica/nsload.c @@ -270,8 +270,7 @@ static acpi_status acpi_ns_delete_subtree(acpi_handle start_handle) /* Now delete the starting object, and we are done */ - acpi_ns_delete_node(child_handle); - + acpi_ns_remove_node(child_handle); return_ACPI_STATUS(AE_OK); } diff --git a/drivers/acpi/bus.c b/drivers/acpi/bus.c index 2876fc70c3a9..620183f13e5e 100644 --- a/drivers/acpi/bus.c +++ b/drivers/acpi/bus.c @@ -141,7 +141,7 @@ int acpi_bus_get_status(struct acpi_device *device) EXPORT_SYMBOL(acpi_bus_get_status); void acpi_bus_private_data_handler(acpi_handle handle, - u32 function, void *context) + void *context) { return; } diff --git a/drivers/acpi/glue.c b/drivers/acpi/glue.c index 27a7072347ea..9a4ce33f137e 100644 --- a/drivers/acpi/glue.c +++ b/drivers/acpi/glue.c @@ -119,7 +119,7 @@ EXPORT_SYMBOL(acpi_get_child); /* Link ACPI devices with physical devices */ static void acpi_glue_data_handler(acpi_handle handle, - u32 function, void *context) + void *context) { /* we provide an empty handler */ } diff --git a/drivers/acpi/scan.c b/drivers/acpi/scan.c index 0ab526de7c55..9606af13d3b8 100644 --- a/drivers/acpi/scan.c +++ b/drivers/acpi/scan.c @@ -687,7 +687,7 @@ acpi_bus_get_ejd(acpi_handle handle, acpi_handle *ejd) } EXPORT_SYMBOL_GPL(acpi_bus_get_ejd); -void acpi_bus_data_handler(acpi_handle handle, u32 function, void *context) +void acpi_bus_data_handler(acpi_handle handle, void *context) { /* TBD */ diff --git a/include/acpi/acpi_bus.h b/include/acpi/acpi_bus.h index b91420b52c6f..6e83a68fbd7b 100644 --- a/include/acpi/acpi_bus.h +++ b/include/acpi/acpi_bus.h @@ -312,7 +312,7 @@ struct acpi_bus_event { extern struct kobject *acpi_kobj; extern int acpi_bus_generate_netlink_event(const char*, const char*, u8, int); -void acpi_bus_private_data_handler(acpi_handle, u32, void *); +void acpi_bus_private_data_handler(acpi_handle, void *); int acpi_bus_get_private_data(acpi_handle, void **); extern int acpi_notifier_call_chain(struct acpi_device *, u32, u32); extern int register_acpi_notifier(struct notifier_block *); @@ -325,7 +325,7 @@ extern void unregister_acpi_bus_notifier(struct notifier_block *nb); */ int acpi_bus_get_device(acpi_handle handle, struct acpi_device **device); -void acpi_bus_data_handler(acpi_handle handle, u32 function, void *context); +void acpi_bus_data_handler(acpi_handle handle, void *context); int acpi_bus_get_status(struct acpi_device *device); int acpi_bus_get_power(acpi_handle handle, int *state); int acpi_bus_set_power(acpi_handle handle, int state); diff --git a/include/acpi/actypes.h b/include/acpi/actypes.h index 4371805d2def..ef4601149f49 100644 --- a/include/acpi/actypes.h +++ b/include/acpi/actypes.h @@ -922,7 +922,7 @@ typedef void (*acpi_notify_handler) (acpi_handle device, u32 value, void *context); typedef -void (*acpi_object_handler) (acpi_handle object, u32 function, void *data); +void (*acpi_object_handler) (acpi_handle object, void *data); typedef acpi_status(*acpi_init_handler) (acpi_handle object, u32 function); -- cgit v1.2.3 From eb2289ba1ba994de25af0d94b5e80ba93d2c1c3c Mon Sep 17 00:00:00 2001 From: Bob Moore Date: Wed, 24 Jun 2009 13:42:00 +0800 Subject: ACPICA: ACPI 4.0: Changes for existing ACPI tables. FACS: new flag and new OspmFlags field. SRAT: x2APIC - add ClockDomain field to descriptor #2 Includes header and disassembler support. Signed-off-by: Bob Moore Signed-off-by: Lin Ming Signed-off-by: Len Brown --- include/acpi/actbl.h | 19 +++++++++++++------ include/acpi/actbl1.h | 8 +++++--- 2 files changed, 18 insertions(+), 9 deletions(-) (limited to 'include') diff --git a/include/acpi/actbl.h b/include/acpi/actbl.h index 222733d01f36..0649a5670026 100644 --- a/include/acpi/actbl.h +++ b/include/acpi/actbl.h @@ -161,17 +161,24 @@ struct acpi_table_facs { u32 flags; u64 xfirmware_waking_vector; /* 64-bit version of the Firmware Waking Vector (ACPI 2.0+) */ u8 version; /* Version of this table (ACPI 2.0+) */ - u8 reserved[31]; /* Reserved, must be zero */ + u8 reserved[3]; /* Reserved, must be zero */ + u32 ospm_flags; /* Flags to be set by OSPM (ACPI 4.0) */ + u8 reserved1[24]; /* Reserved, must be zero */ }; -/* Flag macros */ +/* global_lock flags */ + +#define ACPI_GLOCK_PENDING (1) /* 00: Pending global lock ownership */ +#define ACPI_GLOCK_OWNED (1<<1) /* 01: Global lock is owned */ + +/* Flags */ -#define ACPI_FACS_S4_BIOS_PRESENT (1) /* 00: S4BIOS support is present */ +#define ACPI_FACS_S4_BIOS_PRESENT (1) /* 00: S4BIOS support is present */ +#define ACPI_FACS_64BIT_WAKE (1<<1) /* 01: 64-bit wake vector supported (ACPI 4.0) */ -/* Global lock flags */ +/* ospm_flags */ -#define ACPI_GLOCK_PENDING 0x01 /* 00: Pending global lock ownership */ -#define ACPI_GLOCK_OWNED 0x02 /* 01: Global lock is owned */ +#define ACPI_FACS_64BIT_ENVIRONMENT (1) /* 00: 64-bit wake environment is required (ACPI 4.0) */ /******************************************************************************* * diff --git a/include/acpi/actbl1.h b/include/acpi/actbl1.h index 59ade0752473..ec36693f868c 100644 --- a/include/acpi/actbl1.h +++ b/include/acpi/actbl1.h @@ -1011,7 +1011,7 @@ struct acpi_madt_interrupt_source { #define ACPI_MADT_CPEI_OVERRIDE (1) -/* 9: Processor Local X2_APIC (07/2008) */ +/* 9: Processor Local X2APIC (ACPI 4.0) */ struct acpi_madt_local_x2apic { struct acpi_subtable_header header; @@ -1021,7 +1021,7 @@ struct acpi_madt_local_x2apic { u32 uid; /* ACPI processor UID */ }; -/* 10: Local X2APIC NMI (07/2008) */ +/* 10: Local X2APIC NMI (ACPI 4.0) */ struct acpi_madt_local_x2apic_nmi { struct acpi_subtable_header header; @@ -1211,7 +1211,7 @@ struct acpi_srat_mem_affinity { #define ACPI_SRAT_MEM_HOT_PLUGGABLE (1<<1) /* 01: Memory region is hot pluggable */ #define ACPI_SRAT_MEM_NON_VOLATILE (1<<2) /* 02: Memory region is non-volatile */ -/* 2: Processor Local X2_APIC Affinity (07/2008) */ +/* 2: Processor Local X2_APIC Affinity (ACPI 4.0) */ struct acpi_srat_x2apic_cpu_affinity { struct acpi_subtable_header header; @@ -1219,6 +1219,8 @@ struct acpi_srat_x2apic_cpu_affinity { u32 proximity_domain; u32 apic_id; u32 flags; + u32 clock_domain; + u32 reserved2; }; /* Flags for struct acpi_srat_cpu_affinity and struct acpi_srat_x2apic_cpu_affinity */ -- cgit v1.2.3 From 3ce804ed83827a7fd27190836f9421b29ac64512 Mon Sep 17 00:00:00 2001 From: Bob Moore Date: Thu, 25 Jun 2009 10:31:32 -0700 Subject: ACPICA: Update version to 20090625 Update version number. Signed-off-by: Bob Moore Signed-off-by: Lin Ming Signed-off-by: Len Brown --- include/acpi/acpixf.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/acpi/acpixf.h b/include/acpi/acpixf.h index 04904c7f1aa1..063e577e791e 100644 --- a/include/acpi/acpixf.h +++ b/include/acpi/acpixf.h @@ -47,7 +47,7 @@ /* Current ACPICA subsystem version in YYYYMMDD format */ -#define ACPI_CA_VERSION 0x20090521 +#define ACPI_CA_VERSION 0x20090625 #include "actypes.h" #include "actbl.h" -- cgit v1.2.3 From 36ce99c1dcab2978fc1900f19b431adedd8f99f6 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Thu, 27 Aug 2009 16:45:07 +0200 Subject: ALSA: Add debug module option Add debug module option to snd core. This controls the debug print level. When CONFIG_SND_DEBUG_VERBOSE is set, you can suppress the debug messages by giving or changing this parameter to a lower value. debug=0 means no debug messsages. As default, it's set to the verbose level 2. Since this option can be changed dynamically via sysfs file, you can suppress the verbose debug messages on the fly, which wasn't possible before. Signed-off-by: Takashi Iwai --- Documentation/sound/alsa/ALSA-Configuration.txt | 6 +++ include/sound/core.h | 38 ++++++-------- sound/core/misc.c | 67 ++++++++++++++----------- 3 files changed, 59 insertions(+), 52 deletions(-) (limited to 'include') diff --git a/Documentation/sound/alsa/ALSA-Configuration.txt b/Documentation/sound/alsa/ALSA-Configuration.txt index 4252697a95d6..8e5b3488b370 100644 --- a/Documentation/sound/alsa/ALSA-Configuration.txt +++ b/Documentation/sound/alsa/ALSA-Configuration.txt @@ -60,6 +60,12 @@ Prior to version 0.9.0rc4 options had a 'snd_' prefix. This was removed. slots - Reserve the slot index for the given driver. This option takes multiple strings. See "Module Autoloading Support" section for details. + debug - Specifies the debug message level + (0 = disable debug prints, 1 = normal debug messages, + 2 = verbose debug messages) + This option appears only when CONFIG_SND_DEBUG=y. + This option can be dynamically changed via sysfs + /sys/modules/snd/parameters/debug file. Module snd-pcm-oss ------------------ diff --git a/include/sound/core.h b/include/sound/core.h index 309cb9659a05..a89728db5584 100644 --- a/include/sound/core.h +++ b/include/sound/core.h @@ -340,18 +340,17 @@ unsigned int snd_dma_pointer(unsigned long dma, unsigned int size); struct resource; void release_and_free_resource(struct resource *res); -#ifdef CONFIG_SND_VERBOSE_PRINTK -void snd_verbose_printk(const char *file, int line, const char *format, ...) - __attribute__ ((format (printf, 3, 4))); -#endif -#if defined(CONFIG_SND_DEBUG) && defined(CONFIG_SND_VERBOSE_PRINTK) -void snd_verbose_printd(const char *file, int line, const char *format, ...) - __attribute__ ((format (printf, 3, 4))); -#endif - /* --- */ -#ifdef CONFIG_SND_VERBOSE_PRINTK +#if defined(CONFIG_SND_DEBUG) || defined(CONFIG_SND_VERBOSE_PRINTK) +void __snd_printk(unsigned int level, const char *file, int line, + const char *format, ...) + __attribute__ ((format (printf, 4, 5))); +#else +#define __snd_printk(level, file, line, format, args...) \ + prinkt(format, ##args) +#endif + /** * snd_printk - printk wrapper * @fmt: format string @@ -360,15 +359,9 @@ void snd_verbose_printd(const char *file, int line, const char *format, ...) * when configured with CONFIG_SND_VERBOSE_PRINTK. */ #define snd_printk(fmt, args...) \ - snd_verbose_printk(__FILE__, __LINE__, fmt ,##args) -#else -#define snd_printk(fmt, args...) \ - printk(fmt ,##args) -#endif + __snd_printk(0, __FILE__, __LINE__, fmt, ##args) #ifdef CONFIG_SND_DEBUG - -#ifdef CONFIG_SND_VERBOSE_PRINTK /** * snd_printd - debug printk * @fmt: format string @@ -377,11 +370,7 @@ void snd_verbose_printd(const char *file, int line, const char *format, ...) * Ignored when CONFIG_SND_DEBUG is not set. */ #define snd_printd(fmt, args...) \ - snd_verbose_printd(__FILE__, __LINE__, fmt ,##args) -#else -#define snd_printd(fmt, args...) \ - printk(fmt ,##args) -#endif + __snd_printk(1, __FILE__, __LINE__, fmt, ##args) /** * snd_BUG - give a BUG warning message and stack trace @@ -428,9 +417,10 @@ static inline int __snd_bug_on(int cond) * Works like snd_printk() for debugging purposes. * Ignored when CONFIG_SND_DEBUG_VERBOSE is not set. */ -#define snd_printdd(format, args...) snd_printk(format, ##args) +#define snd_printdd(format, args...) \ + __snd_printk(2, __FILE__, __LINE__, format, ##args) #else -#define snd_printdd(format, args...) /* nothing */ +#define snd_printdd(format, args...) do { } while (0) #endif diff --git a/sound/core/misc.c b/sound/core/misc.c index 1d29e678369e..23a032c6d487 100644 --- a/sound/core/misc.c +++ b/sound/core/misc.c @@ -24,6 +24,20 @@ #include #include +#ifdef CONFIG_SND_DEBUG + +#ifdef CONFIG_SND_DEBUG_VERBOSE +#define DEFAULT_DEBUG_LEVEL 2 +#else +#define DEFAULT_DEBUG_LEVEL 1 +#endif + +static int debug = DEFAULT_DEBUG_LEVEL; +module_param(debug, int, 0644); +MODULE_PARM_DESC(debug, "Debug level (0 = disable)"); + +#endif /* CONFIG_SND_DEBUG */ + void release_and_free_resource(struct resource *res) { if (res) { @@ -35,6 +49,7 @@ void release_and_free_resource(struct resource *res) EXPORT_SYMBOL(release_and_free_resource); #ifdef CONFIG_SND_VERBOSE_PRINTK +/* strip the leading path if the given path is absolute */ static const char *sanity_file_name(const char *path) { if (*path == '/') @@ -43,48 +58,44 @@ static const char *sanity_file_name(const char *path) return path; } -void snd_verbose_printk(const char *path, int line, const char *format, ...) +/* print file and line with a certain printk prefix */ +static int print_snd_pfx(unsigned int level, const char *path, int line, + const char *format) { const char *file = sanity_file_name(path); - va_list args; - - if (format[0] == '<' && format[1] >= '0' && format[1] <= '7' && format[2] == '>') { - char tmp[] = "<0>"; + char tmp[] = "<0>"; + const char *pfx = level ? KERN_DEBUG : KERN_DEFAULT; + int ret = 0; + + if (format[0] == '<' && format[2] == '>') { tmp[1] = format[1]; - printk("%sALSA %s:%d: ", tmp, file, line); - format += 3; - } else { - printk("ALSA %s:%d: ", file, line); + pfx = tmp; + ret = 1; } - va_start(args, format); - vprintk(format, args); - va_end(args); + printk("%sALSA %s:%d: ", pfx, file, line); + return ret; } - -EXPORT_SYMBOL(snd_verbose_printk); +#else +#define print_snd_pfx(level, path, line, format) 0 #endif -#if defined(CONFIG_SND_DEBUG) && defined(CONFIG_SND_VERBOSE_PRINTK) -void snd_verbose_printd(const char *path, int line, const char *format, ...) +#if defined(CONFIG_SND_DEBUG) || defined(CONFIG_SND_VERBOSE_PRINTK) +void __snd_printk(unsigned int level, const char *path, int line, + const char *format, ...) { - const char *file = sanity_file_name(path); va_list args; - if (format[0] == '<' && format[1] >= '0' && format[1] <= '7' && format[2] == '>') { - char tmp[] = "<0>"; - tmp[1] = format[1]; - printk("%sALSA %s:%d: ", tmp, file, line); - format += 3; - } else { - printk(KERN_DEBUG "ALSA %s:%d: ", file, line); - } +#ifdef CONFIG_SND_DEBUG + if (debug < level) + return; +#endif va_start(args, format); + if (print_snd_pfx(level, path, line, format)) + format += 3; /* skip the printk level-prefix */ vprintk(format, args); va_end(args); - } - -EXPORT_SYMBOL(snd_verbose_printd); +EXPORT_SYMBOL_GPL(__snd_printk); #endif #ifdef CONFIG_PCI -- cgit v1.2.3 From 6c347d43eea29221a8ebab9ff9cbe7a00cddac98 Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Thu, 27 Aug 2009 18:17:34 +0200 Subject: tracing: Undef TRACE_EVENT_FN between trace events headers inclusion The recent commit: tracing/events: fix the include file dependencies fixed a file dependency problem while including more than one trace event header file. This fix undefined TRACE_EVENT after an event header macro preprocessing in order to make tracepoint.h able to correctly declare the tracepoints necessary for the next event header file. But now we also need to undefine TRACE_EVENT_FN at the end of an event header file preprocessing for the same reason. This fixes the following build error: In file included from include/trace/events/napi.h:5, from net/core/net-traces.c:28: include/linux/tracepoint.h:285:1: warning: "TRACE_EVENT_FN" redefined In file included from include/trace/define_trace.h:61, from include/trace/events/skb.h:40, from net/core/net-traces.c:27: include/trace/ftrace.h:50:1: warning: this is the location of the previous definition In file included from include/trace/events/napi.h:5, from net/core/net-traces.c:28: include/linux/tracepoint.h:285:1: warning: "TRACE_EVENT_FN" redefined In file included from include/trace/define_trace.h:61, from include/trace/events/skb.h:40, from net/core/net-traces.c:27: include/trace/ftrace.h:50:1: warning: this is the location of the previous definition Reported-by: Ingo Molnar Signed-off-by: Frederic Weisbecker Cc: Masami Hiramatsu Cc: Xiao Guangrong Cc: Steven Rostedt Cc: Li Zefan LKML-Reference: <20090827161732.GA7618@nowhere> Signed-off-by: Ingo Molnar --- include/trace/define_trace.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include') diff --git a/include/trace/define_trace.h b/include/trace/define_trace.h index a89ed590597a..2a4b3bf74033 100644 --- a/include/trace/define_trace.h +++ b/include/trace/define_trace.h @@ -62,6 +62,7 @@ #endif #undef TRACE_EVENT +#undef TRACE_EVENT_FN #undef TRACE_HEADER_MULTI_READ /* Only undef what we defined in this file */ -- cgit v1.2.3 From a5fe1a03f7720b8da8364a1737e1e5a357904e99 Mon Sep 17 00:00:00 2001 From: Lin Ming Date: Thu, 13 Aug 2009 10:43:27 +0800 Subject: ACPICA: fix leak of acpi_os_validate_address http://bugzilla.kernel.org/show_bug.cgi?id=13620 If the dynamic region is created and added to resource list over and over again, it has the potential to be a memory leak by growing the list every time. This patch fixes the memory leak, as below 1) add a new field "count" to struct acpi_res_list. When inserting, if the region(addr, len) is already in the resource list, we just increase "count", otherwise, the region is inserted with count=1. When deleting, the "count" is decreased, if it's decreased to 0, the region is deleted from the resource list. With "count", the region with same address and length can only be inserted to the resource list once, so prevent potential memory leak. 2) add a new function acpi_os_invalidate_address, which is called when region is deleted. Signed-off-by: Lin Ming Signed-off-by: Len Brown --- drivers/acpi/acpica/utdelete.c | 6 +++ drivers/acpi/osl.c | 94 ++++++++++++++++++++++++++++++++++++++++-- include/acpi/acpiosxf.h | 3 ++ 3 files changed, 100 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/drivers/acpi/acpica/utdelete.c b/drivers/acpi/acpica/utdelete.c index bc1710315088..96e26e70c63d 100644 --- a/drivers/acpi/acpica/utdelete.c +++ b/drivers/acpi/acpica/utdelete.c @@ -215,6 +215,12 @@ static void acpi_ut_delete_internal_obj(union acpi_operand_object *object) ACPI_DEBUG_PRINT((ACPI_DB_ALLOCATIONS, "***** Region %p\n", object)); + /* Invalidate the region address/length via the host OS */ + + acpi_os_invalidate_address(object->region.space_id, + object->region.address, + (acpi_size) object->region.length); + second_desc = acpi_ns_get_secondary_object(object); if (second_desc) { /* diff --git a/drivers/acpi/osl.c b/drivers/acpi/osl.c index 5691f165a952..c5b4f1ed9b71 100644 --- a/drivers/acpi/osl.c +++ b/drivers/acpi/osl.c @@ -88,6 +88,7 @@ struct acpi_res_list { char name[5]; /* only can have a length of 4 chars, make use of this one instead of res->name, no need to kalloc then */ struct list_head resource_list; + int count; }; static LIST_HEAD(resource_list_head); @@ -1358,6 +1359,89 @@ acpi_os_validate_interface (char *interface) return AE_SUPPORT; } +static inline int acpi_res_list_add(struct acpi_res_list *res) +{ + struct acpi_res_list *res_list_elem; + + list_for_each_entry(res_list_elem, &resource_list_head, + resource_list) { + + if (res->resource_type == res_list_elem->resource_type && + res->start == res_list_elem->start && + res->end == res_list_elem->end) { + + /* + * The Region(addr,len) already exist in the list, + * just increase the count + */ + + res_list_elem->count++; + return 0; + } + } + + res->count = 1; + list_add(&res->resource_list, &resource_list_head); + return 1; +} + +static inline void acpi_res_list_del(struct acpi_res_list *res) +{ + struct acpi_res_list *res_list_elem; + + list_for_each_entry(res_list_elem, &resource_list_head, + resource_list) { + + if (res->resource_type == res_list_elem->resource_type && + res->start == res_list_elem->start && + res->end == res_list_elem->end) { + + /* + * If the res count is decreased to 0, + * remove and free it + */ + + if (--res_list_elem->count == 0) { + list_del(&res_list_elem->resource_list); + kfree(res_list_elem); + } + return; + } + } +} + +acpi_status +acpi_os_invalidate_address( + u8 space_id, + acpi_physical_address address, + acpi_size length) +{ + struct acpi_res_list res; + + switch (space_id) { + case ACPI_ADR_SPACE_SYSTEM_IO: + case ACPI_ADR_SPACE_SYSTEM_MEMORY: + /* Only interference checks against SystemIO and SytemMemory + are needed */ + res.start = address; + res.end = address + length - 1; + res.resource_type = space_id; + spin_lock(&acpi_res_lock); + acpi_res_list_del(&res); + spin_unlock(&acpi_res_lock); + break; + case ACPI_ADR_SPACE_PCI_CONFIG: + case ACPI_ADR_SPACE_EC: + case ACPI_ADR_SPACE_SMBUS: + case ACPI_ADR_SPACE_CMOS: + case ACPI_ADR_SPACE_PCI_BAR_TARGET: + case ACPI_ADR_SPACE_DATA_TABLE: + case ACPI_ADR_SPACE_FIXED_HARDWARE: + break; + } + return AE_OK; +} + /****************************************************************************** * * FUNCTION: acpi_os_validate_address @@ -1382,6 +1466,7 @@ acpi_os_validate_address ( char *name) { struct acpi_res_list *res; + int added; if (acpi_enforce_resources == ENFORCE_RESOURCES_NO) return AE_OK; @@ -1399,14 +1484,17 @@ acpi_os_validate_address ( res->end = address + length - 1; res->resource_type = space_id; spin_lock(&acpi_res_lock); - list_add(&res->resource_list, &resource_list_head); + added = acpi_res_list_add(res); spin_unlock(&acpi_res_lock); - pr_debug("Added %s resource: start: 0x%llx, end: 0x%llx, " - "name: %s\n", (space_id == ACPI_ADR_SPACE_SYSTEM_IO) + pr_debug("%s %s resource: start: 0x%llx, end: 0x%llx, " + "name: %s\n", added ? "Added" : "Already exist", + (space_id == ACPI_ADR_SPACE_SYSTEM_IO) ? "SystemIO" : "System Memory", (unsigned long long)res->start, (unsigned long long)res->end, res->name); + if (!added) + kfree(res); break; case ACPI_ADR_SPACE_PCI_CONFIG: case ACPI_ADR_SPACE_EC: diff --git a/include/acpi/acpiosxf.h b/include/acpi/acpiosxf.h index ab0b85cf21f3..eb0e7189075f 100644 --- a/include/acpi/acpiosxf.h +++ b/include/acpi/acpiosxf.h @@ -245,6 +245,9 @@ acpi_status acpi_osi_invalidate(char* interface); acpi_status acpi_os_validate_address(u8 space_id, acpi_physical_address address, acpi_size length, char *name); +acpi_status +acpi_os_invalidate_address(u8 space_id, acpi_physical_address address, + acpi_size length); u64 acpi_os_get_timer(void); -- cgit v1.2.3 From 5a53a7640a7af7acf904ed805c6fd1bf9fea829c Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Thu, 27 Aug 2009 21:04:12 +0200 Subject: ALSA: pcm - Increase protocol version Increase the PCM protocol version to indicate the drain ioctl behavior change. Signed-off-by: Takashi Iwai --- include/sound/asound.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/sound/asound.h b/include/sound/asound.h index 82aed3f47534..1f57bb92eb5a 100644 --- a/include/sound/asound.h +++ b/include/sound/asound.h @@ -138,7 +138,7 @@ struct snd_hwdep_dsp_image { * * *****************************************************************************/ -#define SNDRV_PCM_VERSION SNDRV_PROTOCOL_VERSION(2, 0, 9) +#define SNDRV_PCM_VERSION SNDRV_PROTOCOL_VERSION(2, 0, 10) typedef unsigned long snd_pcm_uframes_t; typedef signed long snd_pcm_sframes_t; -- cgit v1.2.3 From f726f30e32305a34a203ff975e60885aa7556c6a Mon Sep 17 00:00:00 2001 From: FUJITA Tomonori Date: Tue, 4 Aug 2009 19:08:24 +0000 Subject: dma: Add set_dma_mask hook to struct dma_map_ops POWERPC needs this hook. SPARC could use it too. Signed-off-by: FUJITA Tomonori Acked-by: Becky Bruce Signed-off-by: Benjamin Herrenschmidt --- include/linux/dma-mapping.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include') diff --git a/include/linux/dma-mapping.h b/include/linux/dma-mapping.h index c0f6c3cd788c..91b761846061 100644 --- a/include/linux/dma-mapping.h +++ b/include/linux/dma-mapping.h @@ -58,6 +58,7 @@ struct dma_map_ops { enum dma_data_direction dir); int (*mapping_error)(struct device *dev, dma_addr_t dma_addr); int (*dma_supported)(struct device *dev, u64 mask); + int (*set_dma_mask)(struct device *dev, u64 mask); int is_phys; }; -- cgit v1.2.3 From 77a53fd21870c726b670c0d8179294ac1ea33468 Mon Sep 17 00:00:00 2001 From: Dmitry Torokhov Date: Tue, 25 Aug 2009 19:24:13 -0700 Subject: Input: matrix-keypad - add function to build device keymap Signed-off-by: Dmitry Torokhov --- drivers/input/keyboard/matrix_keypad.c | 15 +++------------ drivers/input/keyboard/w90p910_keypad.c | 16 ++-------------- include/linux/input/matrix_keypad.h | 32 ++++++++++++++++++++++++++++++++ 3 files changed, 37 insertions(+), 26 deletions(-) (limited to 'include') diff --git a/drivers/input/keyboard/matrix_keypad.c b/drivers/input/keyboard/matrix_keypad.c index 541b981ff075..91cfe5170265 100644 --- a/drivers/input/keyboard/matrix_keypad.c +++ b/drivers/input/keyboard/matrix_keypad.c @@ -319,7 +319,6 @@ static int __devinit matrix_keypad_probe(struct platform_device *pdev) struct input_dev *input_dev; unsigned short *keycodes; unsigned int row_shift; - int i; int err; pdata = pdev->dev.platform_data; @@ -363,18 +362,10 @@ static int __devinit matrix_keypad_probe(struct platform_device *pdev) input_dev->keycode = keycodes; input_dev->keycodesize = sizeof(*keycodes); - input_dev->keycodemax = pdata->num_row_gpios << keypad->row_shift; - - for (i = 0; i < keymap_data->keymap_size; i++) { - unsigned int key = keymap_data->keymap[i]; - unsigned int row = KEY_ROW(key); - unsigned int col = KEY_COL(key); - unsigned short code = KEY_VAL(key); + input_dev->keycodemax = pdata->num_row_gpios << row_shift; - keycodes[MATRIX_SCAN_CODE(row, col, row_shift)] = code; - __set_bit(code, input_dev->keybit); - } - __clear_bit(KEY_RESERVED, input_dev->keybit); + matrix_keypad_build_keymap(keymap_data, row_shift, + input_dev->keycode, input_dev->keybit); input_set_capability(input_dev, EV_MSC, MSC_SCAN); input_set_drvdata(input_dev, keypad); diff --git a/drivers/input/keyboard/w90p910_keypad.c b/drivers/input/keyboard/w90p910_keypad.c index b8598ae124ee..2d03dd0f9e07 100644 --- a/drivers/input/keyboard/w90p910_keypad.c +++ b/drivers/input/keyboard/w90p910_keypad.c @@ -126,7 +126,6 @@ static int __devinit w90p910_keypad_probe(struct platform_device *pdev) struct resource *res; int irq; int error; - int i; if (!pdata) { dev_err(&pdev->dev, "no platform data defined\n"); @@ -197,19 +196,8 @@ static int __devinit w90p910_keypad_probe(struct platform_device *pdev) input_dev->evbit[0] = BIT_MASK(EV_KEY) | BIT_MASK(EV_REP); input_set_capability(input_dev, EV_MSC, MSC_SCAN); - for (i = 0; i < keymap_data->keymap_size; i++) { - unsigned int key = keymap_data->keymap[i]; - unsigned int row = KEY_ROW(key); - unsigned int col = KEY_COL(key); - unsigned short keycode = KEY_VAL(key); - unsigned int scancode = MATRIX_SCAN_CODE(row, col, - W90P910_ROW_SHIFT); - - keypad->keymap[scancode] = keycode; - __set_bit(keycode, input_dev->keybit); - } - __clear_bit(KEY_RESERVED, input_dev->keybit); - + matrix_keypad_build_keymap(keymap_data, W90P910_ROW_SHIFT, + input_dev->keycode, input_dev->keybit); error = request_irq(keypad->irq, w90p910_keypad_irq_handler, IRQF_DISABLED, pdev->name, keypad); diff --git a/include/linux/input/matrix_keypad.h b/include/linux/input/matrix_keypad.h index 15d5903af2dd..b3cd42d50e16 100644 --- a/include/linux/input/matrix_keypad.h +++ b/include/linux/input/matrix_keypad.h @@ -63,4 +63,36 @@ struct matrix_keypad_platform_data { bool wakeup; }; +/** + * matrix_keypad_build_keymap - convert platform keymap into matrix keymap + * @keymap_data: keymap supplied by the platform code + * @row_shift: number of bits to shift row value by to advance to the next + * line in the keymap + * @keymap: expanded version of keymap that is suitable for use by + * matrix keyboad driver + * @keybit: pointer to bitmap of keys supported by input device + * + * This function converts platform keymap (encoded with KEY() macro) into + * an array of keycodes that is suitable for using in a standard matrix + * keyboard driver that uses row and col as indices. + */ +static inline void +matrix_keypad_build_keymap(const struct matrix_keymap_data *keymap_data, + unsigned int row_shift, + unsigned short *keymap, unsigned long *keybit) +{ + int i; + + for (i = 0; i < keymap_data->keymap_size; i++) { + unsigned int key = keymap_data->keymap[i]; + unsigned int row = KEY_ROW(key); + unsigned int col = KEY_COL(key); + unsigned short code = KEY_VAL(key); + + keymap[MATRIX_SCAN_CODE(row, col, row_shift)] = code; + __set_bit(code, keybit); + } + __clear_bit(KEY_RESERVED, keybit); +} + #endif /* _MATRIX_KEYPAD_H */ -- cgit v1.2.3 From 9d8340687c524ce61e3c9c76758c4c81303acfc0 Mon Sep 17 00:00:00 2001 From: David Brownell Date: Tue, 25 Aug 2009 19:24:14 -0700 Subject: Input: add twl4030_keypad driver Add a driver for the keypad controller on TWL4030 family chips. These support up to an 8x8 key matrix. The TWL4030 multifunction chips are mostly used on OMAP3 (or OMAP 2430) based boards. [dtor@mail.ru: switch to matrix-keypad framework, fix changing keymap from userspace] Reviewed-by: Trilok Soni Signed-off-by: David Brownell Signed-off-by: Dmitry Torokhov --- drivers/input/keyboard/Kconfig | 11 + drivers/input/keyboard/Makefile | 1 + drivers/input/keyboard/twl4030_keypad.c | 480 ++++++++++++++++++++++++++++++++ include/linux/i2c/twl4030.h | 19 +- 4 files changed, 505 insertions(+), 6 deletions(-) create mode 100644 drivers/input/keyboard/twl4030_keypad.c (limited to 'include') diff --git a/drivers/input/keyboard/Kconfig b/drivers/input/keyboard/Kconfig index 50e407de8a78..3525c19be428 100644 --- a/drivers/input/keyboard/Kconfig +++ b/drivers/input/keyboard/Kconfig @@ -330,6 +330,17 @@ config KEYBOARD_OMAP To compile this driver as a module, choose M here: the module will be called omap-keypad. +config KEYBOARD_TWL4030 + tristate "TI TWL4030/TWL5030/TPS659x0 keypad support" + depends on TWL4030_CORE + help + Say Y here if your board use the keypad controller on + TWL4030 family chips. It's safe to say enable this + even on boards that don't use the keypad controller. + + To compile this driver as a module, choose M here: the + module will be called twl4030_keypad. + config KEYBOARD_TOSA tristate "Tosa keyboard" depends on MACH_TOSA diff --git a/drivers/input/keyboard/Makefile b/drivers/input/keyboard/Makefile index 152303029203..8a7a22b30266 100644 --- a/drivers/input/keyboard/Makefile +++ b/drivers/input/keyboard/Makefile @@ -30,5 +30,6 @@ obj-$(CONFIG_KEYBOARD_SPITZ) += spitzkbd.o obj-$(CONFIG_KEYBOARD_STOWAWAY) += stowaway.o obj-$(CONFIG_KEYBOARD_SUNKBD) += sunkbd.o obj-$(CONFIG_KEYBOARD_TOSA) += tosakbd.o +obj-$(CONFIG_KEYBOARD_TWL4030) += twl4030_keypad.o obj-$(CONFIG_KEYBOARD_XTKBD) += xtkbd.o obj-$(CONFIG_KEYBOARD_W90P910) += w90p910_keypad.o diff --git a/drivers/input/keyboard/twl4030_keypad.c b/drivers/input/keyboard/twl4030_keypad.c new file mode 100644 index 000000000000..9a2977c21696 --- /dev/null +++ b/drivers/input/keyboard/twl4030_keypad.c @@ -0,0 +1,480 @@ +/* + * twl4030_keypad.c - driver for 8x8 keypad controller in twl4030 chips + * + * Copyright (C) 2007 Texas Instruments, Inc. + * Copyright (C) 2008 Nokia Corporation + * + * Code re-written for 2430SDP by: + * Syed Mohammed Khasim + * + * Initial Code: + * Manjunatha G K + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#include +#include +#include +#include +#include +#include +#include + + +/* + * The TWL4030 family chips include a keypad controller that supports + * up to an 8x8 switch matrix. The controller can issue system wakeup + * events, since it uses only the always-on 32KiHz oscillator, and has + * an internal state machine that decodes pressed keys, including + * multi-key combinations. + * + * This driver lets boards define what keycodes they wish to report for + * which scancodes, as part of the "struct twl4030_keypad_data" used in + * the probe() routine. + * + * See the TPS65950 documentation; that's the general availability + * version of the TWL5030 second generation part. + */ +#define TWL4030_MAX_ROWS 8 /* TWL4030 hard limit */ +#define TWL4030_MAX_COLS 8 +#define TWL4030_ROW_SHIFT 3 +#define TWL4030_KEYMAP_SIZE (TWL4030_MAX_ROWS * TWL4030_MAX_COLS) + +struct twl4030_keypad { + unsigned short keymap[TWL4030_KEYMAP_SIZE]; + u16 kp_state[TWL4030_MAX_ROWS]; + unsigned n_rows; + unsigned n_cols; + unsigned irq; + + struct device *dbg_dev; + struct input_dev *input; +}; + +/*----------------------------------------------------------------------*/ + +/* arbitrary prescaler value 0..7 */ +#define PTV_PRESCALER 4 + +/* Register Offsets */ +#define KEYP_CTRL 0x00 +#define KEYP_DEB 0x01 +#define KEYP_LONG_KEY 0x02 +#define KEYP_LK_PTV 0x03 +#define KEYP_TIMEOUT_L 0x04 +#define KEYP_TIMEOUT_H 0x05 +#define KEYP_KBC 0x06 +#define KEYP_KBR 0x07 +#define KEYP_SMS 0x08 +#define KEYP_FULL_CODE_7_0 0x09 /* row 0 column status */ +#define KEYP_FULL_CODE_15_8 0x0a /* ... row 1 ... */ +#define KEYP_FULL_CODE_23_16 0x0b +#define KEYP_FULL_CODE_31_24 0x0c +#define KEYP_FULL_CODE_39_32 0x0d +#define KEYP_FULL_CODE_47_40 0x0e +#define KEYP_FULL_CODE_55_48 0x0f +#define KEYP_FULL_CODE_63_56 0x10 +#define KEYP_ISR1 0x11 +#define KEYP_IMR1 0x12 +#define KEYP_ISR2 0x13 +#define KEYP_IMR2 0x14 +#define KEYP_SIR 0x15 +#define KEYP_EDR 0x16 /* edge triggers */ +#define KEYP_SIH_CTRL 0x17 + +/* KEYP_CTRL_REG Fields */ +#define KEYP_CTRL_SOFT_NRST BIT(0) +#define KEYP_CTRL_SOFTMODEN BIT(1) +#define KEYP_CTRL_LK_EN BIT(2) +#define KEYP_CTRL_TOE_EN BIT(3) +#define KEYP_CTRL_TOLE_EN BIT(4) +#define KEYP_CTRL_RP_EN BIT(5) +#define KEYP_CTRL_KBD_ON BIT(6) + +/* KEYP_DEB, KEYP_LONG_KEY, KEYP_TIMEOUT_x*/ +#define KEYP_PERIOD_US(t, prescale) ((t) / (31 << (prescale + 1)) - 1) + +/* KEYP_LK_PTV_REG Fields */ +#define KEYP_LK_PTV_PTV_SHIFT 5 + +/* KEYP_{IMR,ISR,SIR} Fields */ +#define KEYP_IMR1_MIS BIT(3) +#define KEYP_IMR1_TO BIT(2) +#define KEYP_IMR1_LK BIT(1) +#define KEYP_IMR1_KP BIT(0) + +/* KEYP_EDR Fields */ +#define KEYP_EDR_KP_FALLING 0x01 +#define KEYP_EDR_KP_RISING 0x02 +#define KEYP_EDR_KP_BOTH 0x03 +#define KEYP_EDR_LK_FALLING 0x04 +#define KEYP_EDR_LK_RISING 0x08 +#define KEYP_EDR_TO_FALLING 0x10 +#define KEYP_EDR_TO_RISING 0x20 +#define KEYP_EDR_MIS_FALLING 0x40 +#define KEYP_EDR_MIS_RISING 0x80 + + +/*----------------------------------------------------------------------*/ + +static int twl4030_kpread(struct twl4030_keypad *kp, + u8 *data, u32 reg, u8 num_bytes) +{ + int ret = twl4030_i2c_read(TWL4030_MODULE_KEYPAD, data, reg, num_bytes); + + if (ret < 0) + dev_warn(kp->dbg_dev, + "Couldn't read TWL4030: %X - ret %d[%x]\n", + reg, ret, ret); + + return ret; +} + +static int twl4030_kpwrite_u8(struct twl4030_keypad *kp, u8 data, u32 reg) +{ + int ret = twl4030_i2c_write_u8(TWL4030_MODULE_KEYPAD, data, reg); + + if (ret < 0) + dev_warn(kp->dbg_dev, + "Could not write TWL4030: %X - ret %d[%x]\n", + reg, ret, ret); + + return ret; +} + +static inline u16 twl4030_col_xlate(struct twl4030_keypad *kp, u8 col) +{ + /* If all bits in a row are active for all coloumns then + * we have that row line connected to gnd. Mark this + * key on as if it was on matrix position n_cols (ie + * one higher than the size of the matrix). + */ + if (col == 0xFF) + return 1 << kp->n_cols; + else + return col & ((1 << kp->n_cols) - 1); +} + +static int twl4030_read_kp_matrix_state(struct twl4030_keypad *kp, u16 *state) +{ + u8 new_state[TWL4030_MAX_ROWS]; + int row; + int ret = twl4030_kpread(kp, new_state, + KEYP_FULL_CODE_7_0, kp->n_rows); + if (ret >= 0) + for (row = 0; row < kp->n_rows; row++) + state[row] = twl4030_col_xlate(kp, new_state[row]); + + return ret; +} + +static int twl4030_is_in_ghost_state(struct twl4030_keypad *kp, u16 *key_state) +{ + int i; + u16 check = 0; + + for (i = 0; i < kp->n_rows; i++) { + u16 col = key_state[i]; + + if ((col & check) && hweight16(col) > 1) + return 1; + + check |= col; + } + + return 0; +} + +static void twl4030_kp_scan(struct twl4030_keypad *kp, bool release_all) +{ + struct input_dev *input = kp->input; + u16 new_state[TWL4030_MAX_ROWS]; + int col, row; + + if (release_all) + memset(new_state, 0, sizeof(new_state)); + else { + /* check for any changes */ + int ret = twl4030_read_kp_matrix_state(kp, new_state); + + if (ret < 0) /* panic ... */ + return; + + if (twl4030_is_in_ghost_state(kp, new_state)) + return; + } + + /* check for changes and print those */ + for (row = 0; row < kp->n_rows; row++) { + int changed = new_state[row] ^ kp->kp_state[row]; + + if (!changed) + continue; + + for (col = 0; col < kp->n_cols; col++) { + int code; + + if (!(changed & (1 << col))) + continue; + + dev_dbg(kp->dbg_dev, "key [%d:%d] %s\n", row, col, + (new_state[row] & (1 << col)) ? + "press" : "release"); + + code = MATRIX_SCAN_CODE(row, col, TWL4030_ROW_SHIFT); + input_event(input, EV_MSC, MSC_SCAN, code); + input_report_key(input, kp->keymap[code], + new_state[row] & (1 << col)); + } + kp->kp_state[row] = new_state[row]; + } + input_sync(input); +} + +/* + * Keypad interrupt handler + */ +static irqreturn_t do_kp_irq(int irq, void *_kp) +{ + struct twl4030_keypad *kp = _kp; + u8 reg; + int ret; + +#ifdef CONFIG_LOCKDEP + /* WORKAROUND for lockdep forcing IRQF_DISABLED on us, which + * we don't want and can't tolerate. Although it might be + * friendlier not to borrow this thread context... + */ + local_irq_enable(); +#endif + + /* Read & Clear TWL4030 pending interrupt */ + ret = twl4030_kpread(kp, ®, KEYP_ISR1, 1); + + /* Release all keys if I2C has gone bad or + * the KEYP has gone to idle state */ + if (ret >= 0 && (reg & KEYP_IMR1_KP)) + twl4030_kp_scan(kp, false); + else + twl4030_kp_scan(kp, true); + + return IRQ_HANDLED; +} + +static int __devinit twl4030_kp_program(struct twl4030_keypad *kp) +{ + u8 reg; + int i; + + /* Enable controller, with hardware decoding but not autorepeat */ + reg = KEYP_CTRL_SOFT_NRST | KEYP_CTRL_SOFTMODEN + | KEYP_CTRL_TOE_EN | KEYP_CTRL_KBD_ON; + if (twl4030_kpwrite_u8(kp, reg, KEYP_CTRL) < 0) + return -EIO; + + /* NOTE: we could use sih_setup() here to package keypad + * event sources as four different IRQs ... but we don't. + */ + + /* Enable TO rising and KP rising and falling edge detection */ + reg = KEYP_EDR_KP_BOTH | KEYP_EDR_TO_RISING; + if (twl4030_kpwrite_u8(kp, reg, KEYP_EDR) < 0) + return -EIO; + + /* Set PTV prescaler Field */ + reg = (PTV_PRESCALER << KEYP_LK_PTV_PTV_SHIFT); + if (twl4030_kpwrite_u8(kp, reg, KEYP_LK_PTV) < 0) + return -EIO; + + /* Set key debounce time to 20 ms */ + i = KEYP_PERIOD_US(20000, PTV_PRESCALER); + if (twl4030_kpwrite_u8(kp, i, KEYP_DEB) < 0) + return -EIO; + + /* Set timeout period to 100 ms */ + i = KEYP_PERIOD_US(200000, PTV_PRESCALER); + if (twl4030_kpwrite_u8(kp, (i & 0xFF), KEYP_TIMEOUT_L) < 0) + return -EIO; + + if (twl4030_kpwrite_u8(kp, (i >> 8), KEYP_TIMEOUT_H) < 0) + return -EIO; + + /* + * Enable Clear-on-Read; disable remembering events that fire + * after the IRQ but before our handler acks (reads) them, + */ + reg = TWL4030_SIH_CTRL_COR_MASK | TWL4030_SIH_CTRL_PENDDIS_MASK; + if (twl4030_kpwrite_u8(kp, reg, KEYP_SIH_CTRL) < 0) + return -EIO; + + /* initialize key state; irqs update it from here on */ + if (twl4030_read_kp_matrix_state(kp, kp->kp_state) < 0) + return -EIO; + + return 0; +} + +/* + * Registers keypad device with input subsystem + * and configures TWL4030 keypad registers + */ +static int __devinit twl4030_kp_probe(struct platform_device *pdev) +{ + struct twl4030_keypad_data *pdata = pdev->dev.platform_data; + const struct matrix_keymap_data *keymap_data = pdata->keymap_data; + struct twl4030_keypad *kp; + struct input_dev *input; + u8 reg; + int error; + + if (!pdata || !pdata->rows || !pdata->cols || + pdata->rows > TWL4030_MAX_ROWS || pdata->cols > TWL4030_MAX_COLS) { + dev_err(&pdev->dev, "Invalid platform_data\n"); + return -EINVAL; + } + + kp = kzalloc(sizeof(*kp), GFP_KERNEL); + input = input_allocate_device(); + if (!kp || !input) { + error = -ENOMEM; + goto err1; + } + + /* Get the debug Device */ + kp->dbg_dev = &pdev->dev; + kp->input = input; + + kp->n_rows = pdata->rows; + kp->n_cols = pdata->cols; + kp->irq = platform_get_irq(pdev, 0); + + /* setup input device */ + __set_bit(EV_KEY, input->evbit); + + /* Enable auto repeat feature of Linux input subsystem */ + if (pdata->rep) + __set_bit(EV_REP, input->evbit); + + input_set_capability(input, EV_MSC, MSC_SCAN); + + input->name = "TWL4030 Keypad"; + input->phys = "twl4030_keypad/input0"; + input->dev.parent = &pdev->dev; + + input->id.bustype = BUS_HOST; + input->id.vendor = 0x0001; + input->id.product = 0x0001; + input->id.version = 0x0003; + + input->keycode = kp->keymap; + input->keycodesize = sizeof(kp->keymap[0]); + input->keycodemax = ARRAY_SIZE(kp->keymap); + + matrix_keypad_build_keymap(keymap_data, TWL4030_ROW_SHIFT, + input->keycode, input->keybit); + + error = input_register_device(input); + if (error) { + dev_err(kp->dbg_dev, + "Unable to register twl4030 keypad device\n"); + goto err1; + } + + error = twl4030_kp_program(kp); + if (error) + goto err2; + + /* + * This ISR will always execute in kernel thread context because of + * the need to access the TWL4030 over the I2C bus. + * + * NOTE: we assume this host is wired to TWL4040 INT1, not INT2 ... + */ + error = request_irq(kp->irq, do_kp_irq, 0, pdev->name, kp); + if (error) { + dev_info(kp->dbg_dev, "request_irq failed for irq no=%d\n", + kp->irq); + goto err3; + } + + /* Enable KP and TO interrupts now. */ + reg = (u8) ~(KEYP_IMR1_KP | KEYP_IMR1_TO); + if (twl4030_kpwrite_u8(kp, reg, KEYP_IMR1)) { + error = -EIO; + goto err4; + } + + platform_set_drvdata(pdev, kp); + return 0; + +err4: + /* mask all events - we don't care about the result */ + (void) twl4030_kpwrite_u8(kp, 0xff, KEYP_IMR1); +err3: + free_irq(kp->irq, NULL); +err2: + input_unregister_device(input); + input = NULL; +err1: + input_free_device(input); + kfree(kp); + return error; +} + +static int __devexit twl4030_kp_remove(struct platform_device *pdev) +{ + struct twl4030_keypad *kp = platform_get_drvdata(pdev); + + free_irq(kp->irq, kp); + input_unregister_device(kp->input); + platform_set_drvdata(pdev, NULL); + kfree(kp); + + return 0; +} + +/* + * NOTE: twl4030 are multi-function devices connected via I2C. + * So this device is a child of an I2C parent, thus it needs to + * support unplug/replug (which most platform devices don't). + */ + +static struct platform_driver twl4030_kp_driver = { + .probe = twl4030_kp_probe, + .remove = __devexit_p(twl4030_kp_remove), + .driver = { + .name = "twl4030_keypad", + .owner = THIS_MODULE, + }, +}; + +static int __init twl4030_kp_init(void) +{ + return platform_driver_register(&twl4030_kp_driver); +} +module_init(twl4030_kp_init); + +static void __exit twl4030_kp_exit(void) +{ + platform_driver_unregister(&twl4030_kp_driver); +} +module_exit(twl4030_kp_exit); + +MODULE_AUTHOR("Texas Instruments"); +MODULE_DESCRIPTION("TWL4030 Keypad Driver"); +MODULE_LICENSE("GPL"); +MODULE_ALIAS("platform:twl4030_keypad"); + diff --git a/include/linux/i2c/twl4030.h b/include/linux/i2c/twl4030.h index 0dc80ef24975..3fd21d7cb6bf 100644 --- a/include/linux/i2c/twl4030.h +++ b/include/linux/i2c/twl4030.h @@ -25,6 +25,9 @@ #ifndef __TWL4030_H_ #define __TWL4030_H_ +#include +#include + /* * Using the twl4030 core we address registers using a pair * { module id, relative register offset } @@ -302,13 +305,17 @@ struct twl4030_madc_platform_data { int irq_line; }; +/* Boards have uniqe mappings of {col, row} --> keycode. + * Column and row are 4 bits, but range only from 0..7. + * a PERSISTENT_KEY is "always on" and never reported. + */ +#define PERSISTENT_KEY(c, r) KEY((c), (r), KEY_RESERVED) + struct twl4030_keypad_data { - int rows; - int cols; - int *keymap; - int irq; - unsigned int keymapsize; - unsigned int rep:1; + const struct matrix_keymap_data *keymap_data; + unsigned rows; + unsigned cols; + bool rep; }; enum twl4030_usb_mode { -- cgit v1.2.3 From cf0baf16c3a3b3dd67ea3df346479032ab10e988 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Fri, 28 Aug 2009 07:22:05 +0200 Subject: ALSA: Fixed a typo of printk() Fixed a silly typo of printk() included in the previous patch... Signed-off-by: Takashi Iwai --- include/sound/core.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/sound/core.h b/include/sound/core.h index a89728db5584..1ec992b3f1d0 100644 --- a/include/sound/core.h +++ b/include/sound/core.h @@ -348,7 +348,7 @@ void __snd_printk(unsigned int level, const char *file, int line, __attribute__ ((format (printf, 4, 5))); #else #define __snd_printk(level, file, line, format, args...) \ - prinkt(format, ##args) + printk(format, ##args) #endif /** -- cgit v1.2.3 From 0dd7b74787eaf7858c6c573353a83c3e2766e674 Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Fri, 28 Aug 2009 00:50:06 +0200 Subject: tracing: Fix double CPP substitution in TRACE_EVENT_FN TRACE_EVENT_FN relays on TRACE_EVENT by reprocessing its parameters into the ftrace events CPP macro. This leads to a double substitution in some cases. For example, a bad consequence is a format always prefixed by "%s, %s\n" for every TRACE_EVENT_FN based events. Eg: cat /debug/tracing/events/syscalls/sys_enter/format [...] print fmt: "%s, %s\n", "\"NR %ld (%lx, %lx, %lx, %lx, %lx, %lx)\"",\ "REC->id, REC->args[0], REC->args[1], REC->args[2], REC->args[3],\ REC->args[4], REC->args[5]" This creates a failure in post-processing tools such as perf trace or trace-cmd. Then drop this double substitution and replace it by a new __cpparg() macro that relays CPP arguments containing commas. Signed-off-by: Frederic Weisbecker Cc: Josh Stone Cc: Li Zefan Cc: Steven Rostedt Cc: Jason Baron LKML-Reference: <1251413406-6704-1-git-send-email-fweisbec@gmail.com> Signed-off-by: Ingo Molnar --- include/trace/ftrace.h | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) (limited to 'include') diff --git a/include/trace/ftrace.h b/include/trace/ftrace.h index 360a77ad79e1..57c56a998ee6 100644 --- a/include/trace/ftrace.h +++ b/include/trace/ftrace.h @@ -45,14 +45,15 @@ }; \ static struct ftrace_event_call event_##name +#undef __cpparg +#define __cpparg(arg...) arg + /* Callbacks are meaningless to ftrace. */ #undef TRACE_EVENT_FN -#define TRACE_EVENT_FN(name, proto, args, tstruct, \ - assign, print, reg, unreg) \ - TRACE_EVENT(name, TP_PROTO(proto), TP_ARGS(args), \ - TP_STRUCT__entry(tstruct), \ - TP_fast_assign(assign), \ - TP_printk(print)) +#define TRACE_EVENT_FN(name, proto, args, tstruct, \ + assign, print, reg, unreg) \ + TRACE_EVENT(name, __cpparg(proto), __cpparg(args), \ + __cpparg(tstruct), __cpparg(assign), __cpparg(print)) \ #include TRACE_INCLUDE(TRACE_INCLUDE_FILE) -- cgit v1.2.3 From 468de9e54a900559b55aa939a4daeaea1915e572 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Thu, 27 Aug 2009 12:07:40 -0400 Subject: nfsd41: expand solo sequence check Compounds consisting of only a sequence operation don't need any additional caching beyond the sequence information we store in the slot entry. Fix nfsd4_is_solo_sequence to identify this case correctly. The additional check for a failed sequence in nfsd4_store_cache_entry() is redundant, since the nfsd4_is_solo_sequence call lower down catches this case. The final ce_cachethis set in nfsd4_sequence is also redundant. Signed-off-by: Andy Adamson Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 9 --------- include/linux/nfsd/xdr4.h | 2 +- 2 files changed, 1 insertion(+), 10 deletions(-) (limited to 'include') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 5f634d24861c..b44a2cfde6f1 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -991,16 +991,10 @@ nfsd4_store_cache_entry(struct nfsd4_compoundres *resp) { struct nfsd4_cache_entry *entry = &resp->cstate.slot->sl_cache_entry; struct svc_rqst *rqstp = resp->rqstp; - struct nfsd4_compoundargs *args = rqstp->rq_argp; - struct nfsd4_op *op = &args->ops[resp->opcnt]; struct kvec *resv = &rqstp->rq_res.head[0]; dprintk("--> %s entry %p\n", __func__, entry); - /* Don't cache a failed OP_SEQUENCE. */ - if (resp->opcnt == 1 && op->opnum == OP_SEQUENCE && resp->cstate.status) - return; - nfsd4_release_respages(entry->ce_respages, entry->ce_resused); entry->ce_opcnt = resp->opcnt; entry->ce_status = resp->cstate.status; @@ -1490,9 +1484,6 @@ nfsd4_sequence(struct svc_rqst *rqstp, slot->sl_inuse = true; slot->sl_seqid = seq->seqid; slot->sl_cache_entry.ce_cachethis = seq->cachethis; - /* Always set the cache entry cachethis for solo sequence */ - if (nfsd4_is_solo_sequence(resp)) - slot->sl_cache_entry.ce_cachethis = 1; cstate->slot = slot; cstate->session = session; diff --git a/include/linux/nfsd/xdr4.h b/include/linux/nfsd/xdr4.h index 5e4beb0deb80..3f716607c86d 100644 --- a/include/linux/nfsd/xdr4.h +++ b/include/linux/nfsd/xdr4.h @@ -467,7 +467,7 @@ struct nfsd4_compoundres { static inline bool nfsd4_is_solo_sequence(struct nfsd4_compoundres *resp) { struct nfsd4_compoundargs *args = resp->rqstp->rq_argp; - return args->opcnt == 1; + return resp->opcnt == 1 && args->ops[0].opnum == OP_SEQUENCE; } static inline bool nfsd4_not_cached(struct nfsd4_compoundres *resp) -- cgit v1.2.3 From 7285dd7fd375763bfb8ab1ac9cf3f1206f503c16 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Fri, 28 Aug 2009 20:25:24 +0200 Subject: clocksource: Resolve cpu hotplug dead lock with TSC unstable Martin Schwidefsky analyzed it: To register a clocksource the clocksource_mutex is acquired and if necessary timekeeping_notify is called to install the clocksource as the timekeeper clock. timekeeping_notify uses stop_machine which needs to take cpu_add_remove_lock mutex. Starting a new cpu is done with the cpu_add_remove_lock mutex held. native_cpu_up checks the tsc of the new cpu and if the tsc is no good clocksource_change_rating is called. Which needs the clocksource_mutex and the deadlock is complete. The solution is to replace the TSC via the clocksource watchdog mechanism. Mark the TSC as unstable and schedule the watchdog work so it gets removed in the watchdog thread context. Signed-off-by: Thomas Gleixner LKML-Reference: Cc: Martin Schwidefsky Cc: John Stultz --- arch/x86/kernel/tsc.c | 8 +++++--- include/linux/clocksource.h | 1 + kernel/time/clocksource.c | 33 ++++++++++++++++++++++++++++++--- 3 files changed, 36 insertions(+), 6 deletions(-) (limited to 'include') diff --git a/arch/x86/kernel/tsc.c b/arch/x86/kernel/tsc.c index 968425422c46..fc3672a303d6 100644 --- a/arch/x86/kernel/tsc.c +++ b/arch/x86/kernel/tsc.c @@ -767,12 +767,14 @@ void mark_tsc_unstable(char *reason) { if (!tsc_unstable) { tsc_unstable = 1; - printk("Marking TSC unstable due to %s\n", reason); + printk(KERN_INFO "Marking TSC unstable due to %s\n", reason); /* Change only the rating, when not registered */ if (clocksource_tsc.mult) - clocksource_change_rating(&clocksource_tsc, 0); - else + clocksource_mark_unstable(&clocksource_tsc); + else { + clocksource_tsc.flags |= CLOCK_SOURCE_UNSTABLE; clocksource_tsc.rating = 0; + } } } diff --git a/include/linux/clocksource.h b/include/linux/clocksource.h index 9ea40ff26f0e..83d2fbd81b93 100644 --- a/include/linux/clocksource.h +++ b/include/linux/clocksource.h @@ -277,6 +277,7 @@ extern struct clocksource* clocksource_get_next(void); extern void clocksource_change_rating(struct clocksource *cs, int rating); extern void clocksource_resume(void); extern struct clocksource * __init __weak clocksource_default_clock(void); +extern void clocksource_mark_unstable(struct clocksource *cs); #ifdef CONFIG_GENERIC_TIME_VSYSCALL extern void update_vsyscall(struct timespec *ts, struct clocksource *c); diff --git a/kernel/time/clocksource.c b/kernel/time/clocksource.c index e0c86ad6e9fb..a0af4ffcb6e5 100644 --- a/kernel/time/clocksource.c +++ b/kernel/time/clocksource.c @@ -149,15 +149,42 @@ static void clocksource_watchdog_work(struct work_struct *work) kthread_run(clocksource_watchdog_kthread, NULL, "kwatchdog"); } -static void clocksource_unstable(struct clocksource *cs, int64_t delta) +static void __clocksource_unstable(struct clocksource *cs) { - printk(KERN_WARNING "Clocksource %s unstable (delta = %Ld ns)\n", - cs->name, delta); cs->flags &= ~(CLOCK_SOURCE_VALID_FOR_HRES | CLOCK_SOURCE_WATCHDOG); cs->flags |= CLOCK_SOURCE_UNSTABLE; schedule_work(&watchdog_work); } +static void clocksource_unstable(struct clocksource *cs, int64_t delta) +{ + printk(KERN_WARNING "Clocksource %s unstable (delta = %Ld ns)\n", + cs->name, delta); + __clocksource_unstable(cs); +} + +/** + * clocksource_mark_unstable - mark clocksource unstable via watchdog + * @cs: clocksource to be marked unstable + * + * This function is called instead of clocksource_change_rating from + * cpu hotplug code to avoid a deadlock between the clocksource mutex + * and the cpu hotplug mutex. It defers the update of the clocksource + * to the watchdog thread. + */ +void clocksource_mark_unstable(struct clocksource *cs) +{ + unsigned long flags; + + spin_lock_irqsave(&watchdog_lock, flags); + if (!(cs->flags & CLOCK_SOURCE_UNSTABLE)) { + if (list_empty(&cs->wd_list)) + list_add(&cs->wd_list, &watchdog_list); + __clocksource_unstable(cs); + } + spin_unlock_irqrestore(&watchdog_lock, flags); +} + static void clocksource_watchdog(unsigned long data) { struct clocksource *cs; -- cgit v1.2.3 From 9e03fdfd05e733e1136d431973625b174029c5e6 Mon Sep 17 00:00:00 2001 From: Javier Cardona Date: Thu, 20 Aug 2009 09:21:45 -0700 Subject: mac80211: Update mesh config IE to 11s draft 3.02 The mesh config information element has changed significantly since draft 1.08 This patch brings it up to date. Thanks to Sam Leffler and Rui Paulo for identifying this. Signed-off-by: Javier Cardona Signed-off-by: John W. Linville --- include/linux/ieee80211.h | 2 +- net/mac80211/ieee80211_i.h | 4 ++++ net/mac80211/mesh.c | 49 ++++++++++++++++++++++++++++++++-------------- 3 files changed, 39 insertions(+), 16 deletions(-) (limited to 'include') diff --git a/include/linux/ieee80211.h b/include/linux/ieee80211.h index 21556a2d9e7e..52e15e079c61 100644 --- a/include/linux/ieee80211.h +++ b/include/linux/ieee80211.h @@ -115,7 +115,7 @@ #define IEEE80211_MAX_SSID_LEN 32 #define IEEE80211_MAX_MESH_ID_LEN 32 -#define IEEE80211_MESH_CONFIG_LEN 19 +#define IEEE80211_MESH_CONFIG_LEN 24 #define IEEE80211_QOS_CTL_LEN 2 #define IEEE80211_QOS_CTL_TID_MASK 0x000F diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h index 93e618a980d1..455cc7ade9f5 100644 --- a/net/mac80211/ieee80211_i.h +++ b/net/mac80211/ieee80211_i.h @@ -367,6 +367,10 @@ struct ieee80211_if_mesh { u8 mesh_pm_id[4]; /* Congestion Control Mode Identifier */ u8 mesh_cc_id[4]; + /* Synchronization Protocol Identifier */ + u8 mesh_sp_id[4]; + /* Authentication Protocol Identifier */ + u8 mesh_auth_id[4]; /* Local mesh Destination Sequence Number */ u32 dsn; /* Last used PREQ ID */ diff --git a/net/mac80211/mesh.c b/net/mac80211/mesh.c index 3185e18c8214..f7364e56f1ee 100644 --- a/net/mac80211/mesh.c +++ b/net/mac80211/mesh.c @@ -18,8 +18,11 @@ #define PP_OFFSET 1 /* Path Selection Protocol */ #define PM_OFFSET 5 /* Path Selection Metric */ #define CC_OFFSET 9 /* Congestion Control Mode */ -#define CAPAB_OFFSET 17 -#define ACCEPT_PLINKS 0x80 +#define SP_OFFSET 13 /* Synchronization Protocol */ +#define AUTH_OFFSET 17 /* Authentication Protocol */ +#define CAPAB_OFFSET 22 +#define CAPAB_ACCEPT_PLINKS 0x80 +#define CAPAB_FORWARDING 0x10 #define TMR_RUNNING_HK 0 #define TMR_RUNNING_MP 1 @@ -84,7 +87,9 @@ bool mesh_matches_local(struct ieee802_11_elems *ie, struct ieee80211_sub_if_dat memcmp(ifmsh->mesh_id, ie->mesh_id, ie->mesh_id_len) == 0 && memcmp(ifmsh->mesh_pp_id, ie->mesh_config + PP_OFFSET, 4) == 0 && memcmp(ifmsh->mesh_pm_id, ie->mesh_config + PM_OFFSET, 4) == 0 && - memcmp(ifmsh->mesh_cc_id, ie->mesh_config + CC_OFFSET, 4) == 0) + memcmp(ifmsh->mesh_cc_id, ie->mesh_config + CC_OFFSET, 4) == 0 && + memcmp(ifmsh->mesh_sp_id, ie->mesh_config + SP_OFFSET, 4) == 0 && + memcmp(ifmsh->mesh_auth_id, ie->mesh_config + AUTH_OFFSET, 4) == 0) return true; return false; @@ -97,7 +102,7 @@ bool mesh_matches_local(struct ieee802_11_elems *ie, struct ieee80211_sub_if_dat */ bool mesh_peer_accepts_plinks(struct ieee802_11_elems *ie) { - return (*(ie->mesh_config + CAPAB_OFFSET) & ACCEPT_PLINKS) != 0; + return (*(ie->mesh_config + CAPAB_OFFSET) & CAPAB_ACCEPT_PLINKS) != 0; } /** @@ -123,11 +128,18 @@ void mesh_accept_plinks_update(struct ieee80211_sub_if_data *sdata) void mesh_ids_set_default(struct ieee80211_if_mesh *sta) { - u8 def_id[4] = {0x00, 0x0F, 0xAC, 0xff}; - - memcpy(sta->mesh_pp_id, def_id, 4); - memcpy(sta->mesh_pm_id, def_id, 4); - memcpy(sta->mesh_cc_id, def_id, 4); + u8 oui[3] = {0x00, 0x0F, 0xAC}; + + memcpy(sta->mesh_pp_id, oui, sizeof(oui)); + memcpy(sta->mesh_pm_id, oui, sizeof(oui)); + memcpy(sta->mesh_cc_id, oui, sizeof(oui)); + memcpy(sta->mesh_sp_id, oui, sizeof(oui)); + memcpy(sta->mesh_auth_id, oui, sizeof(oui)); + sta->mesh_pp_id[sizeof(oui)] = 0; + sta->mesh_pm_id[sizeof(oui)] = 0; + sta->mesh_cc_id[sizeof(oui)] = 0xff; + sta->mesh_sp_id[sizeof(oui)] = 0xff; + sta->mesh_auth_id[sizeof(oui)] = 0x0; } int mesh_rmc_init(struct ieee80211_sub_if_data *sdata) @@ -245,7 +257,7 @@ void mesh_mgmt_ies_add(struct sk_buff *skb, struct ieee80211_sub_if_data *sdata) if (sdata->u.mesh.mesh_id_len) memcpy(pos, sdata->u.mesh.mesh_id, sdata->u.mesh.mesh_id_len); - pos = skb_put(skb, 21); + pos = skb_put(skb, 2 + IEEE80211_MESH_CONFIG_LEN); *pos++ = WLAN_EID_MESH_CONFIG; *pos++ = IEEE80211_MESH_CONFIG_LEN; /* Version */ @@ -263,15 +275,22 @@ void mesh_mgmt_ies_add(struct sk_buff *skb, struct ieee80211_sub_if_data *sdata) memcpy(pos, sdata->u.mesh.mesh_cc_id, 4); pos += 4; - /* Channel precedence: - * Not running simple channel unification protocol - */ - memset(pos, 0x00, 4); + /* Synchronization protocol identifier */ + memcpy(pos, sdata->u.mesh.mesh_sp_id, 4); pos += 4; + /* Authentication Protocol identifier */ + memcpy(pos, sdata->u.mesh.mesh_auth_id, 4); + pos += 4; + + /* Mesh Formation Info */ + memset(pos, 0x00, 1); + pos += 1; + /* Mesh capability */ sdata->u.mesh.accepting_plinks = mesh_plink_availables(sdata); - *pos++ = sdata->u.mesh.accepting_plinks ? ACCEPT_PLINKS : 0x00; + *pos = CAPAB_FORWARDING; + *pos++ |= sdata->u.mesh.accepting_plinks ? CAPAB_ACCEPT_PLINKS : 0x00; *pos++ = 0x00; return; -- cgit v1.2.3 From b0a4e7d8a291de63f35b04464de9ab4a83d38a7c Mon Sep 17 00:00:00 2001 From: "John W. Linville" Date: Thu, 20 Aug 2009 14:48:03 -0400 Subject: libipw: switch from ieee80211_* to libipw_* naming policy This eliminates the dual definition of ieee80211_channel (and possibly others), further clarifying who defines what and paving the way for inclusion of cfg80211.h. Signed-off-by: John W. Linville --- drivers/net/wireless/ipw2x00/ieee80211.h | 1087 -------------------------- drivers/net/wireless/ipw2x00/ipw2100.c | 198 ++--- drivers/net/wireless/ipw2x00/ipw2100.h | 14 +- drivers/net/wireless/ipw2x00/ipw2200.c | 902 ++++++++++----------- drivers/net/wireless/ipw2x00/ipw2200.h | 14 +- drivers/net/wireless/ipw2x00/libipw.h | 1087 ++++++++++++++++++++++++++ drivers/net/wireless/ipw2x00/libipw_geo.c | 80 +- drivers/net/wireless/ipw2x00/libipw_module.c | 80 +- drivers/net/wireless/ipw2x00/libipw_rx.c | 403 +++++----- drivers/net/wireless/ipw2x00/libipw_tx.c | 68 +- drivers/net/wireless/ipw2x00/libipw_wx.c | 92 +-- include/net/iw_handler.h | 6 +- 12 files changed, 2016 insertions(+), 2015 deletions(-) delete mode 100644 drivers/net/wireless/ipw2x00/ieee80211.h create mode 100644 drivers/net/wireless/ipw2x00/libipw.h (limited to 'include') diff --git a/drivers/net/wireless/ipw2x00/ieee80211.h b/drivers/net/wireless/ipw2x00/ieee80211.h deleted file mode 100644 index 70755c1336d5..000000000000 --- a/drivers/net/wireless/ipw2x00/ieee80211.h +++ /dev/null @@ -1,1087 +0,0 @@ -/* - * Merged with mainline ieee80211.h in Aug 2004. Original ieee802_11 - * remains copyright by the original authors - * - * Portions of the merged code are based on Host AP (software wireless - * LAN access point) driver for Intersil Prism2/2.5/3. - * - * Copyright (c) 2001-2002, SSH Communications Security Corp and Jouni Malinen - * - * Copyright (c) 2002-2003, Jouni Malinen - * - * Adaption to a generic IEEE 802.11 stack by James Ketrenos - * - * Copyright (c) 2004-2005, Intel Corporation - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License version 2 as - * published by the Free Software Foundation. See README and COPYING for - * more details. - * - * API Version History - * 1.0.x -- Initial version - * 1.1.x -- Added radiotap, QoS, TIM, ieee80211_geo APIs, - * various structure changes, and crypto API init method - */ -#ifndef IEEE80211_H -#define IEEE80211_H -#include /* ETH_ALEN */ -#include /* ARRAY_SIZE */ -#include -#include - -#include - -#define IEEE80211_VERSION "git-1.1.13" - -#define IEEE80211_DATA_LEN 2304 -/* Maximum size for the MA-UNITDATA primitive, 802.11 standard section - 6.2.1.1.2. - - The figure in section 7.1.2 suggests a body size of up to 2312 - bytes is allowed, which is a bit confusing, I suspect this - represents the 2304 bytes of real data, plus a possible 8 bytes of - WEP IV and ICV. (this interpretation suggested by Ramiro Barreiro) */ - -#define IEEE80211_1ADDR_LEN 10 -#define IEEE80211_2ADDR_LEN 16 -#define IEEE80211_3ADDR_LEN 24 -#define IEEE80211_4ADDR_LEN 30 -#define IEEE80211_FCS_LEN 4 -#define IEEE80211_HLEN (IEEE80211_4ADDR_LEN) -#define IEEE80211_FRAME_LEN (IEEE80211_DATA_LEN + IEEE80211_HLEN) - -#define MIN_FRAG_THRESHOLD 256U -#define MAX_FRAG_THRESHOLD 2346U - -/* QOS control */ -#define IEEE80211_QCTL_TID 0x000F - -/* debug macros */ - -#ifdef CONFIG_LIBIPW_DEBUG -extern u32 ieee80211_debug_level; -#define IEEE80211_DEBUG(level, fmt, args...) \ -do { if (ieee80211_debug_level & (level)) \ - printk(KERN_DEBUG "ieee80211: %c %s " fmt, \ - in_interrupt() ? 'I' : 'U', __func__ , ## args); } while (0) -static inline bool ieee80211_ratelimit_debug(u32 level) -{ - return (ieee80211_debug_level & level) && net_ratelimit(); -} -#else -#define IEEE80211_DEBUG(level, fmt, args...) do {} while (0) -static inline bool ieee80211_ratelimit_debug(u32 level) -{ - return false; -} -#endif /* CONFIG_LIBIPW_DEBUG */ - -/* - * To use the debug system: - * - * If you are defining a new debug classification, simply add it to the #define - * list here in the form of: - * - * #define IEEE80211_DL_xxxx VALUE - * - * shifting value to the left one bit from the previous entry. xxxx should be - * the name of the classification (for example, WEP) - * - * You then need to either add a IEEE80211_xxxx_DEBUG() macro definition for your - * classification, or use IEEE80211_DEBUG(IEEE80211_DL_xxxx, ...) whenever you want - * to send output to that classification. - * - * To add your debug level to the list of levels seen when you perform - * - * % cat /proc/net/ieee80211/debug_level - * - * you simply need to add your entry to the ieee80211_debug_level array. - * - * If you do not see debug_level in /proc/net/ieee80211 then you do not have - * CONFIG_LIBIPW_DEBUG defined in your kernel configuration - * - */ - -#define IEEE80211_DL_INFO (1<<0) -#define IEEE80211_DL_WX (1<<1) -#define IEEE80211_DL_SCAN (1<<2) -#define IEEE80211_DL_STATE (1<<3) -#define IEEE80211_DL_MGMT (1<<4) -#define IEEE80211_DL_FRAG (1<<5) -#define IEEE80211_DL_DROP (1<<7) - -#define IEEE80211_DL_TX (1<<8) -#define IEEE80211_DL_RX (1<<9) -#define IEEE80211_DL_QOS (1<<31) - -#define IEEE80211_ERROR(f, a...) printk(KERN_ERR "ieee80211: " f, ## a) -#define IEEE80211_WARNING(f, a...) printk(KERN_WARNING "ieee80211: " f, ## a) -#define IEEE80211_DEBUG_INFO(f, a...) IEEE80211_DEBUG(IEEE80211_DL_INFO, f, ## a) - -#define IEEE80211_DEBUG_WX(f, a...) IEEE80211_DEBUG(IEEE80211_DL_WX, f, ## a) -#define IEEE80211_DEBUG_SCAN(f, a...) IEEE80211_DEBUG(IEEE80211_DL_SCAN, f, ## a) -#define IEEE80211_DEBUG_STATE(f, a...) IEEE80211_DEBUG(IEEE80211_DL_STATE, f, ## a) -#define IEEE80211_DEBUG_MGMT(f, a...) IEEE80211_DEBUG(IEEE80211_DL_MGMT, f, ## a) -#define IEEE80211_DEBUG_FRAG(f, a...) IEEE80211_DEBUG(IEEE80211_DL_FRAG, f, ## a) -#define IEEE80211_DEBUG_DROP(f, a...) IEEE80211_DEBUG(IEEE80211_DL_DROP, f, ## a) -#define IEEE80211_DEBUG_TX(f, a...) IEEE80211_DEBUG(IEEE80211_DL_TX, f, ## a) -#define IEEE80211_DEBUG_RX(f, a...) IEEE80211_DEBUG(IEEE80211_DL_RX, f, ## a) -#define IEEE80211_DEBUG_QOS(f, a...) IEEE80211_DEBUG(IEEE80211_DL_QOS, f, ## a) -#include -#include /* ARPHRD_ETHER */ - -#ifndef WIRELESS_SPY -#define WIRELESS_SPY /* enable iwspy support */ -#endif -#include /* new driver API */ - -#define ETH_P_PREAUTH 0x88C7 /* IEEE 802.11i pre-authentication */ - -#ifndef ETH_P_80211_RAW -#define ETH_P_80211_RAW (ETH_P_ECONET + 1) -#endif - -/* IEEE 802.11 defines */ - -#define P80211_OUI_LEN 3 - -struct ieee80211_snap_hdr { - - u8 dsap; /* always 0xAA */ - u8 ssap; /* always 0xAA */ - u8 ctrl; /* always 0x03 */ - u8 oui[P80211_OUI_LEN]; /* organizational universal id */ - -} __attribute__ ((packed)); - -#define SNAP_SIZE sizeof(struct ieee80211_snap_hdr) - -#define WLAN_FC_GET_VERS(fc) ((fc) & IEEE80211_FCTL_VERS) -#define WLAN_FC_GET_TYPE(fc) ((fc) & IEEE80211_FCTL_FTYPE) -#define WLAN_FC_GET_STYPE(fc) ((fc) & IEEE80211_FCTL_STYPE) - -#define WLAN_GET_SEQ_FRAG(seq) ((seq) & IEEE80211_SCTL_FRAG) -#define WLAN_GET_SEQ_SEQ(seq) (((seq) & IEEE80211_SCTL_SEQ) >> 4) - -#define IEEE80211_STATMASK_SIGNAL (1<<0) -#define IEEE80211_STATMASK_RSSI (1<<1) -#define IEEE80211_STATMASK_NOISE (1<<2) -#define IEEE80211_STATMASK_RATE (1<<3) -#define IEEE80211_STATMASK_WEMASK 0x7 - -#define IEEE80211_CCK_MODULATION (1<<0) -#define IEEE80211_OFDM_MODULATION (1<<1) - -#define IEEE80211_24GHZ_BAND (1<<0) -#define IEEE80211_52GHZ_BAND (1<<1) - -#define IEEE80211_CCK_RATE_1MB 0x02 -#define IEEE80211_CCK_RATE_2MB 0x04 -#define IEEE80211_CCK_RATE_5MB 0x0B -#define IEEE80211_CCK_RATE_11MB 0x16 -#define IEEE80211_OFDM_RATE_6MB 0x0C -#define IEEE80211_OFDM_RATE_9MB 0x12 -#define IEEE80211_OFDM_RATE_12MB 0x18 -#define IEEE80211_OFDM_RATE_18MB 0x24 -#define IEEE80211_OFDM_RATE_24MB 0x30 -#define IEEE80211_OFDM_RATE_36MB 0x48 -#define IEEE80211_OFDM_RATE_48MB 0x60 -#define IEEE80211_OFDM_RATE_54MB 0x6C -#define IEEE80211_BASIC_RATE_MASK 0x80 - -#define IEEE80211_CCK_RATE_1MB_MASK (1<<0) -#define IEEE80211_CCK_RATE_2MB_MASK (1<<1) -#define IEEE80211_CCK_RATE_5MB_MASK (1<<2) -#define IEEE80211_CCK_RATE_11MB_MASK (1<<3) -#define IEEE80211_OFDM_RATE_6MB_MASK (1<<4) -#define IEEE80211_OFDM_RATE_9MB_MASK (1<<5) -#define IEEE80211_OFDM_RATE_12MB_MASK (1<<6) -#define IEEE80211_OFDM_RATE_18MB_MASK (1<<7) -#define IEEE80211_OFDM_RATE_24MB_MASK (1<<8) -#define IEEE80211_OFDM_RATE_36MB_MASK (1<<9) -#define IEEE80211_OFDM_RATE_48MB_MASK (1<<10) -#define IEEE80211_OFDM_RATE_54MB_MASK (1<<11) - -#define IEEE80211_CCK_RATES_MASK 0x0000000F -#define IEEE80211_CCK_BASIC_RATES_MASK (IEEE80211_CCK_RATE_1MB_MASK | \ - IEEE80211_CCK_RATE_2MB_MASK) -#define IEEE80211_CCK_DEFAULT_RATES_MASK (IEEE80211_CCK_BASIC_RATES_MASK | \ - IEEE80211_CCK_RATE_5MB_MASK | \ - IEEE80211_CCK_RATE_11MB_MASK) - -#define IEEE80211_OFDM_RATES_MASK 0x00000FF0 -#define IEEE80211_OFDM_BASIC_RATES_MASK (IEEE80211_OFDM_RATE_6MB_MASK | \ - IEEE80211_OFDM_RATE_12MB_MASK | \ - IEEE80211_OFDM_RATE_24MB_MASK) -#define IEEE80211_OFDM_DEFAULT_RATES_MASK (IEEE80211_OFDM_BASIC_RATES_MASK | \ - IEEE80211_OFDM_RATE_9MB_MASK | \ - IEEE80211_OFDM_RATE_18MB_MASK | \ - IEEE80211_OFDM_RATE_36MB_MASK | \ - IEEE80211_OFDM_RATE_48MB_MASK | \ - IEEE80211_OFDM_RATE_54MB_MASK) -#define IEEE80211_DEFAULT_RATES_MASK (IEEE80211_OFDM_DEFAULT_RATES_MASK | \ - IEEE80211_CCK_DEFAULT_RATES_MASK) - -#define IEEE80211_NUM_OFDM_RATES 8 -#define IEEE80211_NUM_CCK_RATES 4 -#define IEEE80211_OFDM_SHIFT_MASK_A 4 - -/* NOTE: This data is for statistical purposes; not all hardware provides this - * information for frames received. - * For ieee80211_rx_mgt, you need to set at least the 'len' parameter. - */ -struct ieee80211_rx_stats { - u32 mac_time; - s8 rssi; - u8 signal; - u8 noise; - u16 rate; /* in 100 kbps */ - u8 received_channel; - u8 control; - u8 mask; - u8 freq; - u16 len; - u64 tsf; - u32 beacon_time; -}; - -/* IEEE 802.11 requires that STA supports concurrent reception of at least - * three fragmented frames. This define can be increased to support more - * concurrent frames, but it should be noted that each entry can consume about - * 2 kB of RAM and increasing cache size will slow down frame reassembly. */ -#define IEEE80211_FRAG_CACHE_LEN 4 - -struct ieee80211_frag_entry { - unsigned long first_frag_time; - unsigned int seq; - unsigned int last_frag; - struct sk_buff *skb; - u8 src_addr[ETH_ALEN]; - u8 dst_addr[ETH_ALEN]; -}; - -struct ieee80211_stats { - unsigned int tx_unicast_frames; - unsigned int tx_multicast_frames; - unsigned int tx_fragments; - unsigned int tx_unicast_octets; - unsigned int tx_multicast_octets; - unsigned int tx_deferred_transmissions; - unsigned int tx_single_retry_frames; - unsigned int tx_multiple_retry_frames; - unsigned int tx_retry_limit_exceeded; - unsigned int tx_discards; - unsigned int rx_unicast_frames; - unsigned int rx_multicast_frames; - unsigned int rx_fragments; - unsigned int rx_unicast_octets; - unsigned int rx_multicast_octets; - unsigned int rx_fcs_errors; - unsigned int rx_discards_no_buffer; - unsigned int tx_discards_wrong_sa; - unsigned int rx_discards_undecryptable; - unsigned int rx_message_in_msg_fragments; - unsigned int rx_message_in_bad_msg_fragments; -}; - -struct ieee80211_device; - -#define SEC_KEY_1 (1<<0) -#define SEC_KEY_2 (1<<1) -#define SEC_KEY_3 (1<<2) -#define SEC_KEY_4 (1<<3) -#define SEC_ACTIVE_KEY (1<<4) -#define SEC_AUTH_MODE (1<<5) -#define SEC_UNICAST_GROUP (1<<6) -#define SEC_LEVEL (1<<7) -#define SEC_ENABLED (1<<8) -#define SEC_ENCRYPT (1<<9) - -#define SEC_LEVEL_0 0 /* None */ -#define SEC_LEVEL_1 1 /* WEP 40 and 104 bit */ -#define SEC_LEVEL_2 2 /* Level 1 + TKIP */ -#define SEC_LEVEL_2_CKIP 3 /* Level 1 + CKIP */ -#define SEC_LEVEL_3 4 /* Level 2 + CCMP */ - -#define SEC_ALG_NONE 0 -#define SEC_ALG_WEP 1 -#define SEC_ALG_TKIP 2 -#define SEC_ALG_CCMP 3 - -#define WEP_KEYS 4 -#define WEP_KEY_LEN 13 -#define SCM_KEY_LEN 32 -#define SCM_TEMPORAL_KEY_LENGTH 16 - -struct ieee80211_security { - u16 active_key:2, enabled:1, unicast_uses_group:1, encrypt:1; - u8 auth_mode; - u8 encode_alg[WEP_KEYS]; - u8 key_sizes[WEP_KEYS]; - u8 keys[WEP_KEYS][SCM_KEY_LEN]; - u8 level; - u16 flags; -} __attribute__ ((packed)); - -/* - - 802.11 data frame from AP - - ,-------------------------------------------------------------------. -Bytes | 2 | 2 | 6 | 6 | 6 | 2 | 0..2312 | 4 | - |------|------|---------|---------|---------|------|---------|------| -Desc. | ctrl | dura | DA/RA | TA | SA | Sequ | frame | fcs | - | | tion | (BSSID) | | | ence | data | | - `-------------------------------------------------------------------' - -Total: 28-2340 bytes - -*/ - -#define BEACON_PROBE_SSID_ID_POSITION 12 - -struct ieee80211_hdr_1addr { - __le16 frame_ctl; - __le16 duration_id; - u8 addr1[ETH_ALEN]; - u8 payload[0]; -} __attribute__ ((packed)); - -struct ieee80211_hdr_2addr { - __le16 frame_ctl; - __le16 duration_id; - u8 addr1[ETH_ALEN]; - u8 addr2[ETH_ALEN]; - u8 payload[0]; -} __attribute__ ((packed)); - -struct ieee80211_hdr_3addr { - __le16 frame_ctl; - __le16 duration_id; - u8 addr1[ETH_ALEN]; - u8 addr2[ETH_ALEN]; - u8 addr3[ETH_ALEN]; - __le16 seq_ctl; - u8 payload[0]; -} __attribute__ ((packed)); - -struct ieee80211_hdr_4addr { - __le16 frame_ctl; - __le16 duration_id; - u8 addr1[ETH_ALEN]; - u8 addr2[ETH_ALEN]; - u8 addr3[ETH_ALEN]; - __le16 seq_ctl; - u8 addr4[ETH_ALEN]; - u8 payload[0]; -} __attribute__ ((packed)); - -struct ieee80211_hdr_3addrqos { - __le16 frame_ctl; - __le16 duration_id; - u8 addr1[ETH_ALEN]; - u8 addr2[ETH_ALEN]; - u8 addr3[ETH_ALEN]; - __le16 seq_ctl; - u8 payload[0]; - __le16 qos_ctl; -} __attribute__ ((packed)); - -struct ieee80211_info_element { - u8 id; - u8 len; - u8 data[0]; -} __attribute__ ((packed)); - -/* - * These are the data types that can make up management packets - * - u16 auth_algorithm; - u16 auth_sequence; - u16 beacon_interval; - u16 capability; - u8 current_ap[ETH_ALEN]; - u16 listen_interval; - struct { - u16 association_id:14, reserved:2; - } __attribute__ ((packed)); - u32 time_stamp[2]; - u16 reason; - u16 status; -*/ - -struct ieee80211_auth { - struct ieee80211_hdr_3addr header; - __le16 algorithm; - __le16 transaction; - __le16 status; - /* challenge */ - struct ieee80211_info_element info_element[0]; -} __attribute__ ((packed)); - -struct ieee80211_channel_switch { - u8 id; - u8 len; - u8 mode; - u8 channel; - u8 count; -} __attribute__ ((packed)); - -struct ieee80211_action { - struct ieee80211_hdr_3addr header; - u8 category; - u8 action; - union { - struct ieee80211_action_exchange { - u8 token; - struct ieee80211_info_element info_element[0]; - } exchange; - struct ieee80211_channel_switch channel_switch; - - } format; -} __attribute__ ((packed)); - -struct ieee80211_disassoc { - struct ieee80211_hdr_3addr header; - __le16 reason; -} __attribute__ ((packed)); - -/* Alias deauth for disassoc */ -#define ieee80211_deauth ieee80211_disassoc - -struct ieee80211_probe_request { - struct ieee80211_hdr_3addr header; - /* SSID, supported rates */ - struct ieee80211_info_element info_element[0]; -} __attribute__ ((packed)); - -struct ieee80211_probe_response { - struct ieee80211_hdr_3addr header; - __le32 time_stamp[2]; - __le16 beacon_interval; - __le16 capability; - /* SSID, supported rates, FH params, DS params, - * CF params, IBSS params, TIM (if beacon), RSN */ - struct ieee80211_info_element info_element[0]; -} __attribute__ ((packed)); - -/* Alias beacon for probe_response */ -#define ieee80211_beacon ieee80211_probe_response - -struct ieee80211_assoc_request { - struct ieee80211_hdr_3addr header; - __le16 capability; - __le16 listen_interval; - /* SSID, supported rates, RSN */ - struct ieee80211_info_element info_element[0]; -} __attribute__ ((packed)); - -struct ieee80211_reassoc_request { - struct ieee80211_hdr_3addr header; - __le16 capability; - __le16 listen_interval; - u8 current_ap[ETH_ALEN]; - struct ieee80211_info_element info_element[0]; -} __attribute__ ((packed)); - -struct ieee80211_assoc_response { - struct ieee80211_hdr_3addr header; - __le16 capability; - __le16 status; - __le16 aid; - /* supported rates */ - struct ieee80211_info_element info_element[0]; -} __attribute__ ((packed)); - -struct ieee80211_txb { - u8 nr_frags; - u8 encrypted; - u8 rts_included; - u8 reserved; - u16 frag_size; - u16 payload_size; - struct sk_buff *fragments[0]; -}; - -/* SWEEP TABLE ENTRIES NUMBER */ -#define MAX_SWEEP_TAB_ENTRIES 42 -#define MAX_SWEEP_TAB_ENTRIES_PER_PACKET 7 -/* MAX_RATES_LENGTH needs to be 12. The spec says 8, and many APs - * only use 8, and then use extended rates for the remaining supported - * rates. Other APs, however, stick all of their supported rates on the - * main rates information element... */ -#define MAX_RATES_LENGTH ((u8)12) -#define MAX_RATES_EX_LENGTH ((u8)16) -#define MAX_NETWORK_COUNT 128 - -#define CRC_LENGTH 4U - -#define MAX_WPA_IE_LEN 64 - -#define NETWORK_HAS_OFDM (1<<1) -#define NETWORK_HAS_CCK (1<<2) - -/* QoS structure */ -#define NETWORK_HAS_QOS_PARAMETERS (1<<3) -#define NETWORK_HAS_QOS_INFORMATION (1<<4) -#define NETWORK_HAS_QOS_MASK (NETWORK_HAS_QOS_PARAMETERS | \ - NETWORK_HAS_QOS_INFORMATION) - -/* 802.11h */ -#define NETWORK_HAS_POWER_CONSTRAINT (1<<5) -#define NETWORK_HAS_CSA (1<<6) -#define NETWORK_HAS_QUIET (1<<7) -#define NETWORK_HAS_IBSS_DFS (1<<8) -#define NETWORK_HAS_TPC_REPORT (1<<9) - -#define NETWORK_HAS_ERP_VALUE (1<<10) - -#define QOS_QUEUE_NUM 4 -#define QOS_OUI_LEN 3 -#define QOS_OUI_TYPE 2 -#define QOS_ELEMENT_ID 221 -#define QOS_OUI_INFO_SUB_TYPE 0 -#define QOS_OUI_PARAM_SUB_TYPE 1 -#define QOS_VERSION_1 1 -#define QOS_AIFSN_MIN_VALUE 2 - -struct ieee80211_qos_information_element { - u8 elementID; - u8 length; - u8 qui[QOS_OUI_LEN]; - u8 qui_type; - u8 qui_subtype; - u8 version; - u8 ac_info; -} __attribute__ ((packed)); - -struct ieee80211_qos_ac_parameter { - u8 aci_aifsn; - u8 ecw_min_max; - __le16 tx_op_limit; -} __attribute__ ((packed)); - -struct ieee80211_qos_parameter_info { - struct ieee80211_qos_information_element info_element; - u8 reserved; - struct ieee80211_qos_ac_parameter ac_params_record[QOS_QUEUE_NUM]; -} __attribute__ ((packed)); - -struct ieee80211_qos_parameters { - __le16 cw_min[QOS_QUEUE_NUM]; - __le16 cw_max[QOS_QUEUE_NUM]; - u8 aifs[QOS_QUEUE_NUM]; - u8 flag[QOS_QUEUE_NUM]; - __le16 tx_op_limit[QOS_QUEUE_NUM]; -} __attribute__ ((packed)); - -struct ieee80211_qos_data { - struct ieee80211_qos_parameters parameters; - int active; - int supported; - u8 param_count; - u8 old_param_count; -}; - -struct ieee80211_tim_parameters { - u8 tim_count; - u8 tim_period; -} __attribute__ ((packed)); - -/*******************************************************/ - -enum { /* ieee80211_basic_report.map */ - IEEE80211_BASIC_MAP_BSS = (1 << 0), - IEEE80211_BASIC_MAP_OFDM = (1 << 1), - IEEE80211_BASIC_MAP_UNIDENTIFIED = (1 << 2), - IEEE80211_BASIC_MAP_RADAR = (1 << 3), - IEEE80211_BASIC_MAP_UNMEASURED = (1 << 4), - /* Bits 5-7 are reserved */ - -}; -struct ieee80211_basic_report { - u8 channel; - __le64 start_time; - __le16 duration; - u8 map; -} __attribute__ ((packed)); - -enum { /* ieee80211_measurement_request.mode */ - /* Bit 0 is reserved */ - IEEE80211_MEASUREMENT_ENABLE = (1 << 1), - IEEE80211_MEASUREMENT_REQUEST = (1 << 2), - IEEE80211_MEASUREMENT_REPORT = (1 << 3), - /* Bits 4-7 are reserved */ -}; - -enum { - IEEE80211_REPORT_BASIC = 0, /* required */ - IEEE80211_REPORT_CCA = 1, /* optional */ - IEEE80211_REPORT_RPI = 2, /* optional */ - /* 3-255 reserved */ -}; - -struct ieee80211_measurement_params { - u8 channel; - __le64 start_time; - __le16 duration; -} __attribute__ ((packed)); - -struct ieee80211_measurement_request { - struct ieee80211_info_element ie; - u8 token; - u8 mode; - u8 type; - struct ieee80211_measurement_params params[0]; -} __attribute__ ((packed)); - -struct ieee80211_measurement_report { - struct ieee80211_info_element ie; - u8 token; - u8 mode; - u8 type; - union { - struct ieee80211_basic_report basic[0]; - } u; -} __attribute__ ((packed)); - -struct ieee80211_tpc_report { - u8 transmit_power; - u8 link_margin; -} __attribute__ ((packed)); - -struct ieee80211_channel_map { - u8 channel; - u8 map; -} __attribute__ ((packed)); - -struct ieee80211_ibss_dfs { - struct ieee80211_info_element ie; - u8 owner[ETH_ALEN]; - u8 recovery_interval; - struct ieee80211_channel_map channel_map[0]; -}; - -struct ieee80211_csa { - u8 mode; - u8 channel; - u8 count; -} __attribute__ ((packed)); - -struct ieee80211_quiet { - u8 count; - u8 period; - u8 duration; - u8 offset; -} __attribute__ ((packed)); - -struct ieee80211_network { - /* These entries are used to identify a unique network */ - u8 bssid[ETH_ALEN]; - u8 channel; - /* Ensure null-terminated for any debug msgs */ - u8 ssid[IW_ESSID_MAX_SIZE + 1]; - u8 ssid_len; - - struct ieee80211_qos_data qos_data; - - /* These are network statistics */ - struct ieee80211_rx_stats stats; - u16 capability; - u8 rates[MAX_RATES_LENGTH]; - u8 rates_len; - u8 rates_ex[MAX_RATES_EX_LENGTH]; - u8 rates_ex_len; - unsigned long last_scanned; - u8 mode; - u32 flags; - u32 last_associate; - u32 time_stamp[2]; - u16 beacon_interval; - u16 listen_interval; - u16 atim_window; - u8 erp_value; - u8 wpa_ie[MAX_WPA_IE_LEN]; - size_t wpa_ie_len; - u8 rsn_ie[MAX_WPA_IE_LEN]; - size_t rsn_ie_len; - struct ieee80211_tim_parameters tim; - - /* 802.11h info */ - - /* Power Constraint - mandatory if spctrm mgmt required */ - u8 power_constraint; - - /* TPC Report - mandatory if spctrm mgmt required */ - struct ieee80211_tpc_report tpc_report; - - /* IBSS DFS - mandatory if spctrm mgmt required and IBSS - * NOTE: This is variable length and so must be allocated dynamically */ - struct ieee80211_ibss_dfs *ibss_dfs; - - /* Channel Switch Announcement - optional if spctrm mgmt required */ - struct ieee80211_csa csa; - - /* Quiet - optional if spctrm mgmt required */ - struct ieee80211_quiet quiet; - - struct list_head list; -}; - -enum ieee80211_state { - IEEE80211_UNINITIALIZED = 0, - IEEE80211_INITIALIZED, - IEEE80211_ASSOCIATING, - IEEE80211_ASSOCIATED, - IEEE80211_AUTHENTICATING, - IEEE80211_AUTHENTICATED, - IEEE80211_SHUTDOWN -}; - -#define DEFAULT_MAX_SCAN_AGE (15 * HZ) -#define DEFAULT_FTS 2346 - -#define CFG_IEEE80211_RESERVE_FCS (1<<0) -#define CFG_IEEE80211_COMPUTE_FCS (1<<1) -#define CFG_IEEE80211_RTS (1<<2) - -#define IEEE80211_24GHZ_MIN_CHANNEL 1 -#define IEEE80211_24GHZ_MAX_CHANNEL 14 -#define IEEE80211_24GHZ_CHANNELS (IEEE80211_24GHZ_MAX_CHANNEL - \ - IEEE80211_24GHZ_MIN_CHANNEL + 1) - -#define IEEE80211_52GHZ_MIN_CHANNEL 34 -#define IEEE80211_52GHZ_MAX_CHANNEL 165 -#define IEEE80211_52GHZ_CHANNELS (IEEE80211_52GHZ_MAX_CHANNEL - \ - IEEE80211_52GHZ_MIN_CHANNEL + 1) - -enum { - IEEE80211_CH_PASSIVE_ONLY = (1 << 0), - IEEE80211_CH_80211H_RULES = (1 << 1), - IEEE80211_CH_B_ONLY = (1 << 2), - IEEE80211_CH_NO_IBSS = (1 << 3), - IEEE80211_CH_UNIFORM_SPREADING = (1 << 4), - IEEE80211_CH_RADAR_DETECT = (1 << 5), - IEEE80211_CH_INVALID = (1 << 6), -}; - -struct ieee80211_channel { - u32 freq; /* in MHz */ - u8 channel; - u8 flags; - u8 max_power; /* in dBm */ -}; - -struct ieee80211_geo { - u8 name[4]; - u8 bg_channels; - u8 a_channels; - struct ieee80211_channel bg[IEEE80211_24GHZ_CHANNELS]; - struct ieee80211_channel a[IEEE80211_52GHZ_CHANNELS]; -}; - -struct ieee80211_device { - struct net_device *dev; - struct ieee80211_security sec; - - /* Bookkeeping structures */ - struct ieee80211_stats ieee_stats; - - struct ieee80211_geo geo; - - /* Probe / Beacon management */ - struct list_head network_free_list; - struct list_head network_list; - struct ieee80211_network *networks; - int scans; - int scan_age; - - int iw_mode; /* operating mode (IW_MODE_*) */ - struct iw_spy_data spy_data; /* iwspy support */ - - spinlock_t lock; - - int tx_headroom; /* Set to size of any additional room needed at front - * of allocated Tx SKBs */ - u32 config; - - /* WEP and other encryption related settings at the device level */ - int open_wep; /* Set to 1 to allow unencrypted frames */ - - int reset_on_keychange; /* Set to 1 if the HW needs to be reset on - * WEP key changes */ - - /* If the host performs {en,de}cryption, then set to 1 */ - int host_encrypt; - int host_encrypt_msdu; - int host_decrypt; - /* host performs multicast decryption */ - int host_mc_decrypt; - - /* host should strip IV and ICV from protected frames */ - /* meaningful only when hardware decryption is being used */ - int host_strip_iv_icv; - - int host_open_frag; - int host_build_iv; - int ieee802_1x; /* is IEEE 802.1X used */ - - /* WPA data */ - int wpa_enabled; - int drop_unencrypted; - int privacy_invoked; - size_t wpa_ie_len; - u8 *wpa_ie; - - struct lib80211_crypt_info crypt_info; - - int bcrx_sta_key; /* use individual keys to override default keys even - * with RX of broad/multicast frames */ - - /* Fragmentation structures */ - struct ieee80211_frag_entry frag_cache[IEEE80211_FRAG_CACHE_LEN]; - unsigned int frag_next_idx; - u16 fts; /* Fragmentation Threshold */ - u16 rts; /* RTS threshold */ - - /* Association info */ - u8 bssid[ETH_ALEN]; - - enum ieee80211_state state; - - int mode; /* A, B, G */ - int modulation; /* CCK, OFDM */ - int freq_band; /* 2.4Ghz, 5.2Ghz, Mixed */ - int abg_true; /* ABG flag */ - - int perfect_rssi; - int worst_rssi; - - u16 prev_seq_ctl; /* used to drop duplicate frames */ - - /* Callback functions */ - void (*set_security) (struct net_device * dev, - struct ieee80211_security * sec); - int (*hard_start_xmit) (struct ieee80211_txb * txb, - struct net_device * dev, int pri); - int (*reset_port) (struct net_device * dev); - int (*is_queue_full) (struct net_device * dev, int pri); - - int (*handle_management) (struct net_device * dev, - struct ieee80211_network * network, u16 type); - int (*is_qos_active) (struct net_device *dev, struct sk_buff *skb); - - /* Typical STA methods */ - int (*handle_auth) (struct net_device * dev, - struct ieee80211_auth * auth); - int (*handle_deauth) (struct net_device * dev, - struct ieee80211_deauth * auth); - int (*handle_action) (struct net_device * dev, - struct ieee80211_action * action, - struct ieee80211_rx_stats * stats); - int (*handle_disassoc) (struct net_device * dev, - struct ieee80211_disassoc * assoc); - int (*handle_beacon) (struct net_device * dev, - struct ieee80211_beacon * beacon, - struct ieee80211_network * network); - int (*handle_probe_response) (struct net_device * dev, - struct ieee80211_probe_response * resp, - struct ieee80211_network * network); - int (*handle_probe_request) (struct net_device * dev, - struct ieee80211_probe_request * req, - struct ieee80211_rx_stats * stats); - int (*handle_assoc_response) (struct net_device * dev, - struct ieee80211_assoc_response * resp, - struct ieee80211_network * network); - - /* Typical AP methods */ - int (*handle_assoc_request) (struct net_device * dev); - int (*handle_reassoc_request) (struct net_device * dev, - struct ieee80211_reassoc_request * req); - - /* This must be the last item so that it points to the data - * allocated beyond this structure by alloc_ieee80211 */ - u8 priv[0]; -}; - -#define IEEE_A (1<<0) -#define IEEE_B (1<<1) -#define IEEE_G (1<<2) -#define IEEE_MODE_MASK (IEEE_A|IEEE_B|IEEE_G) - -static inline void *ieee80211_priv(struct net_device *dev) -{ - return ((struct ieee80211_device *)netdev_priv(dev))->priv; -} - -static inline int ieee80211_is_valid_mode(struct ieee80211_device *ieee, - int mode) -{ - /* - * It is possible for both access points and our device to support - * combinations of modes, so as long as there is one valid combination - * of ap/device supported modes, then return success - * - */ - if ((mode & IEEE_A) && - (ieee->modulation & IEEE80211_OFDM_MODULATION) && - (ieee->freq_band & IEEE80211_52GHZ_BAND)) - return 1; - - if ((mode & IEEE_G) && - (ieee->modulation & IEEE80211_OFDM_MODULATION) && - (ieee->freq_band & IEEE80211_24GHZ_BAND)) - return 1; - - if ((mode & IEEE_B) && - (ieee->modulation & IEEE80211_CCK_MODULATION) && - (ieee->freq_band & IEEE80211_24GHZ_BAND)) - return 1; - - return 0; -} - -static inline int ieee80211_get_hdrlen(u16 fc) -{ - int hdrlen = IEEE80211_3ADDR_LEN; - u16 stype = WLAN_FC_GET_STYPE(fc); - - switch (WLAN_FC_GET_TYPE(fc)) { - case IEEE80211_FTYPE_DATA: - if ((fc & IEEE80211_FCTL_FROMDS) && (fc & IEEE80211_FCTL_TODS)) - hdrlen = IEEE80211_4ADDR_LEN; - if (stype & IEEE80211_STYPE_QOS_DATA) - hdrlen += 2; - break; - case IEEE80211_FTYPE_CTL: - switch (WLAN_FC_GET_STYPE(fc)) { - case IEEE80211_STYPE_CTS: - case IEEE80211_STYPE_ACK: - hdrlen = IEEE80211_1ADDR_LEN; - break; - default: - hdrlen = IEEE80211_2ADDR_LEN; - break; - } - break; - } - - return hdrlen; -} - -static inline u8 *ieee80211_get_payload(struct ieee80211_hdr *hdr) -{ - switch (ieee80211_get_hdrlen(le16_to_cpu(hdr->frame_control))) { - case IEEE80211_1ADDR_LEN: - return ((struct ieee80211_hdr_1addr *)hdr)->payload; - case IEEE80211_2ADDR_LEN: - return ((struct ieee80211_hdr_2addr *)hdr)->payload; - case IEEE80211_3ADDR_LEN: - return ((struct ieee80211_hdr_3addr *)hdr)->payload; - case IEEE80211_4ADDR_LEN: - return ((struct ieee80211_hdr_4addr *)hdr)->payload; - } - return NULL; -} - -static inline int ieee80211_is_ofdm_rate(u8 rate) -{ - switch (rate & ~IEEE80211_BASIC_RATE_MASK) { - case IEEE80211_OFDM_RATE_6MB: - case IEEE80211_OFDM_RATE_9MB: - case IEEE80211_OFDM_RATE_12MB: - case IEEE80211_OFDM_RATE_18MB: - case IEEE80211_OFDM_RATE_24MB: - case IEEE80211_OFDM_RATE_36MB: - case IEEE80211_OFDM_RATE_48MB: - case IEEE80211_OFDM_RATE_54MB: - return 1; - } - return 0; -} - -static inline int ieee80211_is_cck_rate(u8 rate) -{ - switch (rate & ~IEEE80211_BASIC_RATE_MASK) { - case IEEE80211_CCK_RATE_1MB: - case IEEE80211_CCK_RATE_2MB: - case IEEE80211_CCK_RATE_5MB: - case IEEE80211_CCK_RATE_11MB: - return 1; - } - return 0; -} - -/* ieee80211.c */ -extern void free_ieee80211(struct net_device *dev); -extern struct net_device *alloc_ieee80211(int sizeof_priv); -extern int ieee80211_change_mtu(struct net_device *dev, int new_mtu); - -extern void ieee80211_networks_age(struct ieee80211_device *ieee, - unsigned long age_secs); - -extern int ieee80211_set_encryption(struct ieee80211_device *ieee); - -/* ieee80211_tx.c */ -extern int ieee80211_xmit(struct sk_buff *skb, struct net_device *dev); -extern void ieee80211_txb_free(struct ieee80211_txb *); - -/* ieee80211_rx.c */ -extern void ieee80211_rx_any(struct ieee80211_device *ieee, - struct sk_buff *skb, struct ieee80211_rx_stats *stats); -extern int ieee80211_rx(struct ieee80211_device *ieee, struct sk_buff *skb, - struct ieee80211_rx_stats *rx_stats); -/* make sure to set stats->len */ -extern void ieee80211_rx_mgt(struct ieee80211_device *ieee, - struct ieee80211_hdr_4addr *header, - struct ieee80211_rx_stats *stats); -extern void ieee80211_network_reset(struct ieee80211_network *network); - -/* ieee80211_geo.c */ -extern const struct ieee80211_geo *ieee80211_get_geo(struct ieee80211_device - *ieee); -extern int ieee80211_set_geo(struct ieee80211_device *ieee, - const struct ieee80211_geo *geo); - -extern int ieee80211_is_valid_channel(struct ieee80211_device *ieee, - u8 channel); -extern int ieee80211_channel_to_index(struct ieee80211_device *ieee, - u8 channel); -extern u8 ieee80211_freq_to_channel(struct ieee80211_device *ieee, u32 freq); -extern u8 ieee80211_get_channel_flags(struct ieee80211_device *ieee, - u8 channel); -extern const struct ieee80211_channel *ieee80211_get_channel(struct - ieee80211_device - *ieee, u8 channel); -extern u32 ieee80211_channel_to_freq(struct ieee80211_device * ieee, - u8 channel); - -/* ieee80211_wx.c */ -extern int ieee80211_wx_get_scan(struct ieee80211_device *ieee, - struct iw_request_info *info, - union iwreq_data *wrqu, char *key); -extern int ieee80211_wx_set_encode(struct ieee80211_device *ieee, - struct iw_request_info *info, - union iwreq_data *wrqu, char *key); -extern int ieee80211_wx_get_encode(struct ieee80211_device *ieee, - struct iw_request_info *info, - union iwreq_data *wrqu, char *key); -extern int ieee80211_wx_set_encodeext(struct ieee80211_device *ieee, - struct iw_request_info *info, - union iwreq_data *wrqu, char *extra); -extern int ieee80211_wx_get_encodeext(struct ieee80211_device *ieee, - struct iw_request_info *info, - union iwreq_data *wrqu, char *extra); - -static inline void ieee80211_increment_scans(struct ieee80211_device *ieee) -{ - ieee->scans++; -} - -static inline int ieee80211_get_scans(struct ieee80211_device *ieee) -{ - return ieee->scans; -} - -#endif /* IEEE80211_H */ diff --git a/drivers/net/wireless/ipw2x00/ipw2100.c b/drivers/net/wireless/ipw2x00/ipw2100.c index dee50ed0897d..33bdb20e9f1e 100644 --- a/drivers/net/wireless/ipw2x00/ipw2100.c +++ b/drivers/net/wireless/ipw2x00/ipw2100.c @@ -1673,7 +1673,7 @@ static int ipw2100_start_scan(struct ipw2100_priv *priv) return err; } -static const struct ieee80211_geo ipw_geos[] = { +static const struct libipw_geo ipw_geos[] = { { /* Restricted */ "---", .bg_channels = 14, @@ -1694,7 +1694,7 @@ static int ipw2100_up(struct ipw2100_priv *priv, int deferred) /* Age scan list entries found before suspend */ if (priv->suspend_time) { - ieee80211_networks_age(priv->ieee, priv->suspend_time); + libipw_networks_age(priv->ieee, priv->suspend_time); priv->suspend_time = 0; } @@ -1752,11 +1752,11 @@ static int ipw2100_up(struct ipw2100_priv *priv, int deferred) } /* Initialize the geo */ - if (ieee80211_set_geo(priv->ieee, &ipw_geos[0])) { + if (libipw_set_geo(priv->ieee, &ipw_geos[0])) { printk(KERN_WARNING DRV_NAME "Could not set geo\n"); return 0; } - priv->ieee->freq_band = IEEE80211_24GHZ_BAND; + priv->ieee->freq_band = LIBIPW_24GHZ_BAND; lock = LOCK_NONE; if (ipw2100_set_ordinal(priv, IPW_ORD_PERS_DB_LOCK, &lock, &ord_len)) { @@ -1817,7 +1817,7 @@ static int ipw2100_up(struct ipw2100_priv *priv, int deferred) /* Called by register_netdev() */ static int ipw2100_net_init(struct net_device *dev) { - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); return ipw2100_up(priv, 1); } @@ -2340,8 +2340,8 @@ static u32 ipw2100_match_buf(struct ipw2100_priv *priv, u8 * in_buf, * * When packet is provided by the firmware, it contains the following: * - * . ieee80211_hdr - * . ieee80211_snap_hdr + * . libipw_hdr + * . libipw_snap_hdr * * The size of the constructed ethernet * @@ -2396,7 +2396,7 @@ static void ipw2100_corruption_detected(struct ipw2100_priv *priv, int i) } static void isr_rx(struct ipw2100_priv *priv, int i, - struct ieee80211_rx_stats *stats) + struct libipw_rx_stats *stats) { struct net_device *dev = priv->net_dev; struct ipw2100_status *status = &priv->status_queue.drv[i]; @@ -2435,13 +2435,13 @@ static void isr_rx(struct ipw2100_priv *priv, int i, #ifdef IPW2100_RX_DEBUG /* Make a copy of the frame so we can dump it to the logs if - * ieee80211_rx fails */ + * libipw_rx fails */ skb_copy_from_linear_data(packet->skb, packet_data, min_t(u32, status->frame_size, IPW_RX_NIC_BUFFER_LENGTH)); #endif - if (!ieee80211_rx(priv->ieee, packet->skb, stats)) { + if (!libipw_rx(priv->ieee, packet->skb, stats)) { #ifdef IPW2100_RX_DEBUG IPW_DEBUG_DROP("%s: Non consumed packet:\n", dev->name); @@ -2449,7 +2449,7 @@ static void isr_rx(struct ipw2100_priv *priv, int i, #endif dev->stats.rx_errors++; - /* ieee80211_rx failed, so it didn't free the SKB */ + /* libipw_rx failed, so it didn't free the SKB */ dev_kfree_skb_any(packet->skb); packet->skb = NULL; } @@ -2470,7 +2470,7 @@ static void isr_rx(struct ipw2100_priv *priv, int i, #ifdef CONFIG_IPW2100_MONITOR static void isr_rx_monitor(struct ipw2100_priv *priv, int i, - struct ieee80211_rx_stats *stats) + struct libipw_rx_stats *stats) { struct net_device *dev = priv->net_dev; struct ipw2100_status *status = &priv->status_queue.drv[i]; @@ -2528,10 +2528,10 @@ static void isr_rx_monitor(struct ipw2100_priv *priv, int i, skb_put(packet->skb, status->frame_size + sizeof(struct ipw_rt_hdr)); - if (!ieee80211_rx(priv->ieee, packet->skb, stats)) { + if (!libipw_rx(priv->ieee, packet->skb, stats)) { dev->stats.rx_errors++; - /* ieee80211_rx failed, so it didn't free the SKB */ + /* libipw_rx failed, so it didn't free the SKB */ dev_kfree_skb_any(packet->skb); packet->skb = NULL; } @@ -2615,7 +2615,7 @@ static void __ipw2100_rx_process(struct ipw2100_priv *priv) u16 frame_type; u32 r, w, i, s; struct ipw2100_rx *u; - struct ieee80211_rx_stats stats = { + struct libipw_rx_stats stats = { .mac_time = jiffies, }; @@ -2661,8 +2661,8 @@ static void __ipw2100_rx_process(struct ipw2100_priv *priv) stats.mask = 0; if (stats.rssi != 0) - stats.mask |= IEEE80211_STATMASK_RSSI; - stats.freq = IEEE80211_24GHZ_BAND; + stats.mask |= LIBIPW_STATMASK_RSSI; + stats.freq = LIBIPW_24GHZ_BAND; IPW_DEBUG_RX("%s: '%s' frame type received (%d).\n", priv->net_dev->name, frame_types[frame_type], @@ -2686,11 +2686,11 @@ static void __ipw2100_rx_process(struct ipw2100_priv *priv) break; } #endif - if (stats.len < sizeof(struct ieee80211_hdr_3addr)) + if (stats.len < sizeof(struct libipw_hdr_3addr)) break; switch (WLAN_FC_GET_TYPE(le16_to_cpu(u->rx_data.header.frame_ctl))) { case IEEE80211_FTYPE_MGMT: - ieee80211_rx_mgt(priv->ieee, + libipw_rx_mgt(priv->ieee, &u->rx_data.header, &stats); break; @@ -2884,7 +2884,7 @@ static int __ipw2100_tx_process(struct ipw2100_priv *priv) tbd->buf_length, PCI_DMA_TODEVICE); } - ieee80211_txb_free(packet->info.d_struct.txb); + libipw_txb_free(packet->info.d_struct.txb); packet->info.d_struct.txb = NULL; list_add_tail(element, &priv->tx_free_list); @@ -3028,7 +3028,7 @@ static void ipw2100_tx_send_data(struct ipw2100_priv *priv) int next = txq->next; int i = 0; struct ipw2100_data_header *ipw_hdr; - struct ieee80211_hdr_3addr *hdr; + struct libipw_hdr_3addr *hdr; while (!list_empty(&priv->tx_pend_list)) { /* if there isn't enough space in TBD queue, then @@ -3062,7 +3062,7 @@ static void ipw2100_tx_send_data(struct ipw2100_priv *priv) packet->index = txq->next; ipw_hdr = packet->info.d_struct.data; - hdr = (struct ieee80211_hdr_3addr *)packet->info.d_struct.txb-> + hdr = (struct libipw_hdr_3addr *)packet->info.d_struct.txb-> fragments[0]->data; if (priv->ieee->iw_mode == IW_MODE_INFRA) { @@ -3086,7 +3086,7 @@ static void ipw2100_tx_send_data(struct ipw2100_priv *priv) if (packet->info.d_struct.txb->nr_frags > 1) ipw_hdr->fragment_size = packet->info.d_struct.txb->frag_size - - IEEE80211_3ADDR_LEN; + LIBIPW_3ADDR_LEN; else ipw_hdr->fragment_size = 0; @@ -3119,13 +3119,13 @@ static void ipw2100_tx_send_data(struct ipw2100_priv *priv) IPW_BD_STATUS_TX_FRAME_NOT_LAST_FRAGMENT; tbd->buf_length = packet->info.d_struct.txb-> - fragments[i]->len - IEEE80211_3ADDR_LEN; + fragments[i]->len - LIBIPW_3ADDR_LEN; tbd->host_addr = pci_map_single(priv->pci_dev, packet->info.d_struct. txb->fragments[i]-> data + - IEEE80211_3ADDR_LEN, + LIBIPW_3ADDR_LEN, tbd->buf_length, PCI_DMA_TODEVICE); @@ -3330,10 +3330,10 @@ static irqreturn_t ipw2100_interrupt(int irq, void *data) return IRQ_NONE; } -static int ipw2100_tx(struct ieee80211_txb *txb, struct net_device *dev, +static int ipw2100_tx(struct libipw_txb *txb, struct net_device *dev, int pri) { - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); struct list_head *element; struct ipw2100_tx_packet *packet; unsigned long flags; @@ -4488,7 +4488,7 @@ static void ipw2100_tx_initialize(struct ipw2100_priv *priv) /* We simply drop any SKBs that have been queued for * transmit */ if (priv->tx_buffers[i].info.d_struct.txb) { - ieee80211_txb_free(priv->tx_buffers[i].info.d_struct. + libipw_txb_free(priv->tx_buffers[i].info.d_struct. txb); priv->tx_buffers[i].info.d_struct.txb = NULL; } @@ -4527,7 +4527,7 @@ static void ipw2100_tx_free(struct ipw2100_priv *priv) for (i = 0; i < TX_PENDED_QUEUE_LENGTH; i++) { if (priv->tx_buffers[i].info.d_struct.txb) { - ieee80211_txb_free(priv->tx_buffers[i].info.d_struct. + libipw_txb_free(priv->tx_buffers[i].info.d_struct. txb); priv->tx_buffers[i].info.d_struct.txb = NULL; } @@ -5558,9 +5558,9 @@ static void ipw2100_security_work(struct work_struct *work) } static void shim__set_security(struct net_device *dev, - struct ieee80211_security *sec) + struct libipw_security *sec) { - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); int i, force_update = 0; mutex_lock(&priv->action_mutex); @@ -5753,7 +5753,7 @@ static int ipw2100_adapter_setup(struct ipw2100_priv *priv) * method as well) to talk to the firmware */ static int ipw2100_set_address(struct net_device *dev, void *p) { - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); struct sockaddr *addr = p; int err = 0; @@ -5781,7 +5781,7 @@ static int ipw2100_set_address(struct net_device *dev, void *p) static int ipw2100_open(struct net_device *dev) { - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); unsigned long flags; IPW_DEBUG_INFO("dev->open\n"); @@ -5797,7 +5797,7 @@ static int ipw2100_open(struct net_device *dev) static int ipw2100_close(struct net_device *dev) { - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); unsigned long flags; struct list_head *element; struct ipw2100_tx_packet *packet; @@ -5818,7 +5818,7 @@ static int ipw2100_close(struct net_device *dev) list_del(element); DEC_STAT(&priv->tx_pend_stat); - ieee80211_txb_free(packet->info.d_struct.txb); + libipw_txb_free(packet->info.d_struct.txb); packet->info.d_struct.txb = NULL; list_add_tail(element, &priv->tx_free_list); @@ -5836,7 +5836,7 @@ static int ipw2100_close(struct net_device *dev) */ static void ipw2100_tx_timeout(struct net_device *dev) { - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); dev->stats.tx_errors++; @@ -5861,8 +5861,8 @@ static int ipw2100_wpa_enable(struct ipw2100_priv *priv, int value) static int ipw2100_wpa_set_auth_algs(struct ipw2100_priv *priv, int value) { - struct ieee80211_device *ieee = priv->ieee; - struct ieee80211_security sec = { + struct libipw_device *ieee = priv->ieee; + struct libipw_security sec = { .flags = SEC_AUTH_MODE, }; int ret = 0; @@ -5907,7 +5907,7 @@ static void ipw2100_wpa_assoc_frame(struct ipw2100_priv *priv, static void ipw_ethtool_get_drvinfo(struct net_device *dev, struct ethtool_drvinfo *info) { - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); char fw_ver[64], ucode_ver[64]; strcpy(info->driver, DRV_NAME); @@ -5924,7 +5924,7 @@ static void ipw_ethtool_get_drvinfo(struct net_device *dev, static u32 ipw2100_ethtool_get_link(struct net_device *dev) { - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); return (priv->status & STATUS_ASSOCIATED) ? 1 : 0; } @@ -6011,8 +6011,8 @@ static void ipw2100_irq_tasklet(struct ipw2100_priv *priv); static const struct net_device_ops ipw2100_netdev_ops = { .ndo_open = ipw2100_open, .ndo_stop = ipw2100_close, - .ndo_start_xmit = ieee80211_xmit, - .ndo_change_mtu = ieee80211_change_mtu, + .ndo_start_xmit = libipw_xmit, + .ndo_change_mtu = libipw_change_mtu, .ndo_init = ipw2100_net_init, .ndo_tx_timeout = ipw2100_tx_timeout, .ndo_set_mac_address = ipw2100_set_address, @@ -6032,7 +6032,7 @@ static struct net_device *ipw2100_alloc_device(struct pci_dev *pci_dev, dev = alloc_ieee80211(sizeof(struct ipw2100_priv)); if (!dev) return NULL; - priv = ieee80211_priv(dev); + priv = libipw_priv(dev); priv->ieee = netdev_priv(dev); priv->pci_dev = pci_dev; priv->net_dev = dev; @@ -6046,7 +6046,7 @@ static struct net_device *ipw2100_alloc_device(struct pci_dev *pci_dev, dev->netdev_ops = &ipw2100_netdev_ops; dev->ethtool_ops = &ipw2100_ethtool_ops; dev->wireless_handlers = &ipw2100_wx_handler_def; - priv->wireless_data.ieee80211 = priv->ieee; + priv->wireless_data.libipw = priv->ieee; dev->wireless_data = &priv->wireless_data; dev->watchdog_timeo = 3 * HZ; dev->irq = 0; @@ -6202,7 +6202,7 @@ static int ipw2100_pci_init_one(struct pci_dev *pci_dev, return err; } - priv = ieee80211_priv(dev); + priv = libipw_priv(dev); pci_set_master(pci_dev); pci_set_drvdata(pci_dev, priv); @@ -6629,7 +6629,7 @@ static int ipw2100_wx_get_name(struct net_device *dev, * This can be called at any time. No action lock required */ - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); if (!(priv->status & STATUS_ASSOCIATED)) strcpy(wrqu->name, "unassociated"); else @@ -6643,7 +6643,7 @@ static int ipw2100_wx_set_freq(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); struct iw_freq *fwrq = &wrqu->freq; int err = 0; @@ -6693,7 +6693,7 @@ static int ipw2100_wx_get_freq(struct net_device *dev, * This can be called at any time. No action lock required */ - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); wrqu->freq.e = 0; @@ -6714,7 +6714,7 @@ static int ipw2100_wx_set_mode(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); int err = 0; IPW_DEBUG_WX("SET Mode -> %d \n", wrqu->mode); @@ -6757,7 +6757,7 @@ static int ipw2100_wx_get_mode(struct net_device *dev, * This can be called at any time. No action lock required */ - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); wrqu->mode = priv->ieee->iw_mode; IPW_DEBUG_WX("GET Mode -> %d\n", wrqu->mode); @@ -6792,7 +6792,7 @@ static int ipw2100_wx_get_range(struct net_device *dev, * This can be called at any time. No action lock required */ - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); struct iw_range *range = (struct iw_range *)extra; u16 val; int i, level; @@ -6913,7 +6913,7 @@ static int ipw2100_wx_set_wap(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); int err = 0; static const unsigned char any[] = { @@ -6962,7 +6962,7 @@ static int ipw2100_wx_get_wap(struct net_device *dev, * This can be called at any time. No action lock required */ - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); /* If we are associated, trying to associate, or have a statically * configured BSSID then return that; otherwise return ANY */ @@ -6980,7 +6980,7 @@ static int ipw2100_wx_set_essid(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); char *essid = ""; /* ANY */ int length = 0; int err = 0; @@ -7035,7 +7035,7 @@ static int ipw2100_wx_get_essid(struct net_device *dev, * This can be called at any time. No action lock required */ - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); DECLARE_SSID_BUF(ssid); /* If we are associated, trying to associate, or have a statically @@ -7063,7 +7063,7 @@ static int ipw2100_wx_set_nick(struct net_device *dev, * This can be called at any time. No action lock required */ - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); if (wrqu->data.length > IW_ESSID_MAX_SIZE) return -E2BIG; @@ -7085,7 +7085,7 @@ static int ipw2100_wx_get_nick(struct net_device *dev, * This can be called at any time. No action lock required */ - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); wrqu->data.length = strlen(priv->nick); memcpy(extra, priv->nick, wrqu->data.length); @@ -7100,7 +7100,7 @@ static int ipw2100_wx_set_rate(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); u32 target_rate = wrqu->bitrate.value; u32 rate; int err = 0; @@ -7140,7 +7140,7 @@ static int ipw2100_wx_get_rate(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); int val; unsigned int len = sizeof(val); int err = 0; @@ -7192,7 +7192,7 @@ static int ipw2100_wx_set_rts(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); int value, err; /* Auto RTS not yet supported */ @@ -7231,7 +7231,7 @@ static int ipw2100_wx_get_rts(struct net_device *dev, * This can be called at any time. No action lock required */ - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); wrqu->rts.value = priv->rts_threshold & ~RTS_DISABLED; wrqu->rts.fixed = 1; /* no auto select */ @@ -7248,7 +7248,7 @@ static int ipw2100_wx_set_txpow(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); int err = 0, value; if (ipw_radio_kill_sw(priv, wrqu->txpower.disabled)) @@ -7293,7 +7293,7 @@ static int ipw2100_wx_get_txpow(struct net_device *dev, * This can be called at any time. No action lock required */ - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); wrqu->txpower.disabled = (priv->status & STATUS_RF_KILL_MASK) ? 1 : 0; @@ -7320,7 +7320,7 @@ static int ipw2100_wx_set_frag(struct net_device *dev, * This can be called at any time. No action lock required */ - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); if (!wrqu->frag.fixed) return -EINVAL; @@ -7350,7 +7350,7 @@ static int ipw2100_wx_get_frag(struct net_device *dev, * This can be called at any time. No action lock required */ - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); wrqu->frag.value = priv->frag_threshold & ~FRAG_DISABLED; wrqu->frag.fixed = 0; /* no auto select */ wrqu->frag.disabled = (priv->frag_threshold & FRAG_DISABLED) ? 1 : 0; @@ -7364,7 +7364,7 @@ static int ipw2100_wx_set_retry(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); int err = 0; if (wrqu->retry.flags & IW_RETRY_LIFETIME || wrqu->retry.disabled) @@ -7412,7 +7412,7 @@ static int ipw2100_wx_get_retry(struct net_device *dev, * This can be called at any time. No action lock required */ - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); wrqu->retry.disabled = 0; /* can't be disabled */ @@ -7440,7 +7440,7 @@ static int ipw2100_wx_set_scan(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); int err = 0; mutex_lock(&priv->action_mutex); @@ -7472,8 +7472,8 @@ static int ipw2100_wx_get_scan(struct net_device *dev, * This can be called at any time. No action lock required */ - struct ipw2100_priv *priv = ieee80211_priv(dev); - return ieee80211_wx_get_scan(priv->ieee, info, wrqu, extra); + struct ipw2100_priv *priv = libipw_priv(dev); + return libipw_wx_get_scan(priv->ieee, info, wrqu, extra); } /* @@ -7487,8 +7487,8 @@ static int ipw2100_wx_set_encode(struct net_device *dev, * No check of STATUS_INITIALIZED required */ - struct ipw2100_priv *priv = ieee80211_priv(dev); - return ieee80211_wx_set_encode(priv->ieee, info, wrqu, key); + struct ipw2100_priv *priv = libipw_priv(dev); + return libipw_wx_set_encode(priv->ieee, info, wrqu, key); } static int ipw2100_wx_get_encode(struct net_device *dev, @@ -7499,15 +7499,15 @@ static int ipw2100_wx_get_encode(struct net_device *dev, * This can be called at any time. No action lock required */ - struct ipw2100_priv *priv = ieee80211_priv(dev); - return ieee80211_wx_get_encode(priv->ieee, info, wrqu, key); + struct ipw2100_priv *priv = libipw_priv(dev); + return libipw_wx_get_encode(priv->ieee, info, wrqu, key); } static int ipw2100_wx_set_power(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); int err = 0; mutex_lock(&priv->action_mutex); @@ -7556,7 +7556,7 @@ static int ipw2100_wx_get_power(struct net_device *dev, * This can be called at any time. No action lock required */ - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); if (!(priv->power_mode & IPW_POWER_ENABLED)) wrqu->power.disabled = 1; @@ -7580,8 +7580,8 @@ static int ipw2100_wx_set_genie(struct net_device *dev, union iwreq_data *wrqu, char *extra) { - struct ipw2100_priv *priv = ieee80211_priv(dev); - struct ieee80211_device *ieee = priv->ieee; + struct ipw2100_priv *priv = libipw_priv(dev); + struct libipw_device *ieee = priv->ieee; u8 *buf; if (!ieee->wpa_enabled) @@ -7615,8 +7615,8 @@ static int ipw2100_wx_get_genie(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw2100_priv *priv = ieee80211_priv(dev); - struct ieee80211_device *ieee = priv->ieee; + struct ipw2100_priv *priv = libipw_priv(dev); + struct libipw_device *ieee = priv->ieee; if (ieee->wpa_ie_len == 0 || ieee->wpa_ie == NULL) { wrqu->data.length = 0; @@ -7637,8 +7637,8 @@ static int ipw2100_wx_set_auth(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw2100_priv *priv = ieee80211_priv(dev); - struct ieee80211_device *ieee = priv->ieee; + struct ipw2100_priv *priv = libipw_priv(dev); + struct libipw_device *ieee = priv->ieee; struct iw_param *param = &wrqu->param; struct lib80211_crypt_data *crypt; unsigned long flags; @@ -7682,7 +7682,7 @@ static int ipw2100_wx_set_auth(struct net_device *dev, * can use this to determine if the CAP_PRIVACY_ON bit should * be set. */ - struct ieee80211_security sec = { + struct libipw_security sec = { .flags = SEC_ENABLED, .enabled = param->value, }; @@ -7730,8 +7730,8 @@ static int ipw2100_wx_get_auth(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw2100_priv *priv = ieee80211_priv(dev); - struct ieee80211_device *ieee = priv->ieee; + struct ipw2100_priv *priv = libipw_priv(dev); + struct libipw_device *ieee = priv->ieee; struct lib80211_crypt_data *crypt; struct iw_param *param = &wrqu->param; int ret = 0; @@ -7792,8 +7792,8 @@ static int ipw2100_wx_set_encodeext(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw2100_priv *priv = ieee80211_priv(dev); - return ieee80211_wx_set_encodeext(priv->ieee, info, wrqu, extra); + struct ipw2100_priv *priv = libipw_priv(dev); + return libipw_wx_set_encodeext(priv->ieee, info, wrqu, extra); } /* SIOCGIWENCODEEXT */ @@ -7801,8 +7801,8 @@ static int ipw2100_wx_get_encodeext(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw2100_priv *priv = ieee80211_priv(dev); - return ieee80211_wx_get_encodeext(priv->ieee, info, wrqu, extra); + struct ipw2100_priv *priv = libipw_priv(dev); + return libipw_wx_get_encodeext(priv->ieee, info, wrqu, extra); } /* SIOCSIWMLME */ @@ -7810,7 +7810,7 @@ static int ipw2100_wx_set_mlme(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); struct iw_mlme *mlme = (struct iw_mlme *)extra; __le16 reason; @@ -7841,7 +7841,7 @@ static int ipw2100_wx_set_promisc(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); int *parms = (int *)extra; int enable = (parms[0] > 0); int err = 0; @@ -7872,7 +7872,7 @@ static int ipw2100_wx_reset(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); if (priv->status & STATUS_INITIALIZED) schedule_reset(priv); return 0; @@ -7884,7 +7884,7 @@ static int ipw2100_wx_set_powermode(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); int err = 0, mode = *(int *)extra; mutex_lock(&priv->action_mutex); @@ -7912,7 +7912,7 @@ static int ipw2100_wx_get_powermode(struct net_device *dev, * This can be called at any time. No action lock required */ - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); int level = IPW_POWER_LEVEL(priv->power_mode); s32 timeout, period; @@ -7948,7 +7948,7 @@ static int ipw2100_wx_set_preamble(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); int err, mode = *(int *)extra; mutex_lock(&priv->action_mutex); @@ -7981,7 +7981,7 @@ static int ipw2100_wx_get_preamble(struct net_device *dev, * This can be called at any time. No action lock required */ - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); if (priv->config & CFG_LONG_PREAMBLE) snprintf(wrqu->name, IFNAMSIZ, "long (1)"); @@ -7996,7 +7996,7 @@ static int ipw2100_wx_set_crc_check(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); int err, mode = *(int *)extra; mutex_lock(&priv->action_mutex); @@ -8028,7 +8028,7 @@ static int ipw2100_wx_get_crc_check(struct net_device *dev, * This can be called at any time. No action lock required */ - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); if (priv->config & CFG_CRC_CHECK) snprintf(wrqu->name, IFNAMSIZ, "CRC checked (1)"); @@ -8181,7 +8181,7 @@ static struct iw_statistics *ipw2100_wx_wireless_stats(struct net_device *dev) int beacon_qual; int quality; - struct ipw2100_priv *priv = ieee80211_priv(dev); + struct ipw2100_priv *priv = libipw_priv(dev); struct iw_statistics *wstats; u32 rssi, tx_retries, missed_beacons, tx_failures; u32 ord_len = sizeof(u32); diff --git a/drivers/net/wireless/ipw2x00/ipw2100.h b/drivers/net/wireless/ipw2x00/ipw2100.h index f183d951cd32..af175bdc9f29 100644 --- a/drivers/net/wireless/ipw2x00/ipw2100.h +++ b/drivers/net/wireless/ipw2x00/ipw2100.h @@ -46,7 +46,7 @@ #include #include -#include "ieee80211.h" +#include "libipw.h" struct ipw2100_priv; struct ipw2100_tx_packet; @@ -343,7 +343,7 @@ struct ipw2100_tx_packet { struct { /* DATA */ struct ipw2100_data_header *data; dma_addr_t data_phys; - struct ieee80211_txb *txb; + struct libipw_txb *txb; } d_struct; } info; int jiffy_start; @@ -492,7 +492,7 @@ struct ipw2100_priv { int stop_hang_check; /* Set 1 when shutting down to kill hang_check */ int stop_rf_kill; /* Set 1 when shutting down to kill rf_kill */ - struct ieee80211_device *ieee; + struct libipw_device *ieee; unsigned long status; unsigned long config; unsigned long capability; @@ -788,7 +788,7 @@ struct ipw2100_priv { #define IPW_CARD_DISABLE_PHY_OFF_COMPLETE_WAIT 100 // 100 milli #define IPW_PREPARE_POWER_DOWN_COMPLETE_WAIT 100 // 100 milli -#define IPW_HEADER_802_11_SIZE sizeof(struct ieee80211_hdr_3addr) +#define IPW_HEADER_802_11_SIZE sizeof(struct libipw_hdr_3addr) #define IPW_MAX_80211_PAYLOAD_SIZE 2304U #define IPW_MAX_802_11_PAYLOAD_LENGTH 2312 #define IPW_MAX_ACCEPTABLE_TX_FRAME_LENGTH 1536 @@ -803,13 +803,13 @@ struct ipw2100_priv { IPW_802_11_FCS_LENGTH) #define IPW_802_11_PAYLOAD_OFFSET \ - (sizeof(struct ieee80211_hdr_3addr) + \ - sizeof(struct ieee80211_snap_hdr)) + (sizeof(struct libipw_hdr_3addr) + \ + sizeof(struct libipw_snap_hdr)) struct ipw2100_rx { union { unsigned char payload[IPW_RX_NIC_BUFFER_LENGTH]; - struct ieee80211_hdr_4addr header; + struct libipw_hdr_4addr header; u32 status; struct ipw2100_notification notification; struct ipw2100_cmd_header command; diff --git a/drivers/net/wireless/ipw2x00/ipw2200.c b/drivers/net/wireless/ipw2x00/ipw2200.c index 8e18d5348350..3617e3c1e9ed 100644 --- a/drivers/net/wireless/ipw2x00/ipw2200.c +++ b/drivers/net/wireless/ipw2x00/ipw2200.c @@ -111,7 +111,7 @@ static int qos_no_ack_mask = 0; static int burst_duration_CCK = 0; static int burst_duration_OFDM = 0; -static struct ieee80211_qos_parameters def_qos_parameters_OFDM = { +static struct libipw_qos_parameters def_qos_parameters_OFDM = { {QOS_TX0_CW_MIN_OFDM, QOS_TX1_CW_MIN_OFDM, QOS_TX2_CW_MIN_OFDM, QOS_TX3_CW_MIN_OFDM}, {QOS_TX0_CW_MAX_OFDM, QOS_TX1_CW_MAX_OFDM, QOS_TX2_CW_MAX_OFDM, @@ -122,7 +122,7 @@ static struct ieee80211_qos_parameters def_qos_parameters_OFDM = { QOS_TX2_TXOP_LIMIT_OFDM, QOS_TX3_TXOP_LIMIT_OFDM} }; -static struct ieee80211_qos_parameters def_qos_parameters_CCK = { +static struct libipw_qos_parameters def_qos_parameters_CCK = { {QOS_TX0_CW_MIN_CCK, QOS_TX1_CW_MIN_CCK, QOS_TX2_CW_MIN_CCK, QOS_TX3_CW_MIN_CCK}, {QOS_TX0_CW_MAX_CCK, QOS_TX1_CW_MAX_CCK, QOS_TX2_CW_MAX_CCK, @@ -133,7 +133,7 @@ static struct ieee80211_qos_parameters def_qos_parameters_CCK = { QOS_TX3_TXOP_LIMIT_CCK} }; -static struct ieee80211_qos_parameters def_parameters_OFDM = { +static struct libipw_qos_parameters def_parameters_OFDM = { {DEF_TX0_CW_MIN_OFDM, DEF_TX1_CW_MIN_OFDM, DEF_TX2_CW_MIN_OFDM, DEF_TX3_CW_MIN_OFDM}, {DEF_TX0_CW_MAX_OFDM, DEF_TX1_CW_MAX_OFDM, DEF_TX2_CW_MAX_OFDM, @@ -144,7 +144,7 @@ static struct ieee80211_qos_parameters def_parameters_OFDM = { DEF_TX2_TXOP_LIMIT_OFDM, DEF_TX3_TXOP_LIMIT_OFDM} }; -static struct ieee80211_qos_parameters def_parameters_CCK = { +static struct libipw_qos_parameters def_parameters_CCK = { {DEF_TX0_CW_MIN_CCK, DEF_TX1_CW_MIN_CCK, DEF_TX2_CW_MIN_CCK, DEF_TX3_CW_MIN_CCK}, {DEF_TX0_CW_MAX_CCK, DEF_TX1_CW_MAX_CCK, DEF_TX2_CW_MAX_CCK, @@ -164,9 +164,9 @@ static int from_priority_to_tx_queue[] = { static u32 ipw_qos_get_burst_duration(struct ipw_priv *priv); -static int ipw_send_qos_params_command(struct ipw_priv *priv, struct ieee80211_qos_parameters +static int ipw_send_qos_params_command(struct ipw_priv *priv, struct libipw_qos_parameters *qos_param); -static int ipw_send_qos_info_command(struct ipw_priv *priv, struct ieee80211_qos_information_element +static int ipw_send_qos_info_command(struct ipw_priv *priv, struct libipw_qos_information_element *qos_param); #endif /* CONFIG_IPW2200_QOS */ @@ -1830,7 +1830,7 @@ static ssize_t store_speed_scan(struct device *d, struct device_attribute *attr, break; } - if (ieee80211_is_valid_channel(priv->ieee, channel)) + if (libipw_is_valid_channel(priv->ieee, channel)) priv->speed_scan[pos++] = channel; else IPW_WARNING("Skipping invalid channel request: %d\n", @@ -1882,7 +1882,7 @@ static ssize_t show_channels(struct device *d, char *buf) { struct ipw_priv *priv = dev_get_drvdata(d); - const struct ieee80211_geo *geo = ieee80211_get_geo(priv->ieee); + const struct libipw_geo *geo = libipw_get_geo(priv->ieee); int len = 0, i; len = sprintf(&buf[len], @@ -1892,14 +1892,14 @@ static ssize_t show_channels(struct device *d, for (i = 0; i < geo->bg_channels; i++) { len += sprintf(&buf[len], "%d: BSS%s%s, %s, Band %s.\n", geo->bg[i].channel, - geo->bg[i].flags & IEEE80211_CH_RADAR_DETECT ? + geo->bg[i].flags & LIBIPW_CH_RADAR_DETECT ? " (radar spectrum)" : "", - ((geo->bg[i].flags & IEEE80211_CH_NO_IBSS) || - (geo->bg[i].flags & IEEE80211_CH_RADAR_DETECT)) + ((geo->bg[i].flags & LIBIPW_CH_NO_IBSS) || + (geo->bg[i].flags & LIBIPW_CH_RADAR_DETECT)) ? "" : ", IBSS", - geo->bg[i].flags & IEEE80211_CH_PASSIVE_ONLY ? + geo->bg[i].flags & LIBIPW_CH_PASSIVE_ONLY ? "passive only" : "active/passive", - geo->bg[i].flags & IEEE80211_CH_B_ONLY ? + geo->bg[i].flags & LIBIPW_CH_B_ONLY ? "B" : "B/G"); } @@ -1909,12 +1909,12 @@ static ssize_t show_channels(struct device *d, for (i = 0; i < geo->a_channels; i++) { len += sprintf(&buf[len], "%d: BSS%s%s, %s.\n", geo->a[i].channel, - geo->a[i].flags & IEEE80211_CH_RADAR_DETECT ? + geo->a[i].flags & LIBIPW_CH_RADAR_DETECT ? " (radar spectrum)" : "", - ((geo->a[i].flags & IEEE80211_CH_NO_IBSS) || - (geo->a[i].flags & IEEE80211_CH_RADAR_DETECT)) + ((geo->a[i].flags & LIBIPW_CH_NO_IBSS) || + (geo->a[i].flags & LIBIPW_CH_RADAR_DETECT)) ? "" : ", IBSS", - geo->a[i].flags & IEEE80211_CH_PASSIVE_ONLY ? + geo->a[i].flags & LIBIPW_CH_PASSIVE_ONLY ? "passive only" : "active/passive"); } @@ -2429,7 +2429,7 @@ static int ipw_send_tx_power(struct ipw_priv *priv, struct ipw_tx_power *power) static int ipw_set_tx_power(struct ipw_priv *priv) { - const struct ieee80211_geo *geo = ieee80211_get_geo(priv->ieee); + const struct libipw_geo *geo = libipw_get_geo(priv->ieee); struct ipw_tx_power tx_power; s8 max_power; int i; @@ -2960,12 +2960,12 @@ static int ipw_fw_dma_wait(struct ipw_priv *priv) static void ipw_remove_current_network(struct ipw_priv *priv) { struct list_head *element, *safe; - struct ieee80211_network *network = NULL; + struct libipw_network *network = NULL; unsigned long flags; spin_lock_irqsave(&priv->ieee->lock, flags); list_for_each_safe(element, safe, &priv->ieee->network_list) { - network = list_entry(element, struct ieee80211_network, list); + network = list_entry(element, struct libipw_network, list); if (!memcmp(network->bssid, priv->bssid, ETH_ALEN)) { list_del(element); list_add_tail(&network->list, @@ -3751,7 +3751,7 @@ static void ipw_queue_tx_free_tfd(struct ipw_priv *priv, le16_to_cpu(bd->u.data.chunk_len[i]), PCI_DMA_TODEVICE); if (txq->txb[txq->q.last_used]) { - ieee80211_txb_free(txq->txb[txq->q.last_used]); + libipw_txb_free(txq->txb[txq->q.last_used]); txq->txb[txq->q.last_used] = NULL; } } @@ -4070,7 +4070,7 @@ static u32 ipw_get_max_rate(struct ipw_priv *priv) /* If currently associated in B mode, restrict the maximum * rate match to B rates */ if (priv->assoc_request.ieee_mode == IPW_B_MODE) - mask &= IEEE80211_CCK_RATES_MASK; + mask &= LIBIPW_CCK_RATES_MASK; /* TODO: Verify that the rate is supported by the current rates * list. */ @@ -4078,29 +4078,29 @@ static u32 ipw_get_max_rate(struct ipw_priv *priv) while (i && !(mask & i)) i >>= 1; switch (i) { - case IEEE80211_CCK_RATE_1MB_MASK: + case LIBIPW_CCK_RATE_1MB_MASK: return 1000000; - case IEEE80211_CCK_RATE_2MB_MASK: + case LIBIPW_CCK_RATE_2MB_MASK: return 2000000; - case IEEE80211_CCK_RATE_5MB_MASK: + case LIBIPW_CCK_RATE_5MB_MASK: return 5500000; - case IEEE80211_OFDM_RATE_6MB_MASK: + case LIBIPW_OFDM_RATE_6MB_MASK: return 6000000; - case IEEE80211_OFDM_RATE_9MB_MASK: + case LIBIPW_OFDM_RATE_9MB_MASK: return 9000000; - case IEEE80211_CCK_RATE_11MB_MASK: + case LIBIPW_CCK_RATE_11MB_MASK: return 11000000; - case IEEE80211_OFDM_RATE_12MB_MASK: + case LIBIPW_OFDM_RATE_12MB_MASK: return 12000000; - case IEEE80211_OFDM_RATE_18MB_MASK: + case LIBIPW_OFDM_RATE_18MB_MASK: return 18000000; - case IEEE80211_OFDM_RATE_24MB_MASK: + case LIBIPW_OFDM_RATE_24MB_MASK: return 24000000; - case IEEE80211_OFDM_RATE_36MB_MASK: + case LIBIPW_OFDM_RATE_36MB_MASK: return 36000000; - case IEEE80211_OFDM_RATE_48MB_MASK: + case LIBIPW_OFDM_RATE_48MB_MASK: return 48000000; - case IEEE80211_OFDM_RATE_54MB_MASK: + case LIBIPW_OFDM_RATE_54MB_MASK: return 54000000; } @@ -4466,11 +4466,11 @@ static void ipw_rx_notification(struct ipw_priv *priv, == IEEE80211_STYPE_ASSOC_RESP)) { if ((sizeof (struct - ieee80211_assoc_response) + libipw_assoc_response) <= size) && (size <= 2314)) { struct - ieee80211_rx_stats + libipw_rx_stats stats = { .len = size - 1, }; @@ -4478,10 +4478,10 @@ static void ipw_rx_notification(struct ipw_priv *priv, IPW_DEBUG_QOS ("QoS Associate " "size %d\n", size); - ieee80211_rx_mgt(priv-> + libipw_rx_mgt(priv-> ieee, (struct - ieee80211_hdr_4addr + libipw_hdr_4addr *) ¬if->u.raw, &stats); } @@ -4537,11 +4537,11 @@ static void ipw_rx_notification(struct ipw_priv *priv, case CMAS_INIT:{ if (priv->status & STATUS_AUTH) { struct - ieee80211_assoc_response + libipw_assoc_response *resp; resp = (struct - ieee80211_assoc_response + libipw_assoc_response *)¬if->u.raw; IPW_DEBUG(IPW_DL_NOTIF | IPW_DL_STATE | @@ -5227,33 +5227,33 @@ static struct ipw_rx_queue *ipw_rx_queue_alloc(struct ipw_priv *priv) static int ipw_is_rate_in_mask(struct ipw_priv *priv, int ieee_mode, u8 rate) { - rate &= ~IEEE80211_BASIC_RATE_MASK; + rate &= ~LIBIPW_BASIC_RATE_MASK; if (ieee_mode == IEEE_A) { switch (rate) { - case IEEE80211_OFDM_RATE_6MB: - return priv->rates_mask & IEEE80211_OFDM_RATE_6MB_MASK ? + case LIBIPW_OFDM_RATE_6MB: + return priv->rates_mask & LIBIPW_OFDM_RATE_6MB_MASK ? 1 : 0; - case IEEE80211_OFDM_RATE_9MB: - return priv->rates_mask & IEEE80211_OFDM_RATE_9MB_MASK ? + case LIBIPW_OFDM_RATE_9MB: + return priv->rates_mask & LIBIPW_OFDM_RATE_9MB_MASK ? 1 : 0; - case IEEE80211_OFDM_RATE_12MB: + case LIBIPW_OFDM_RATE_12MB: return priv-> - rates_mask & IEEE80211_OFDM_RATE_12MB_MASK ? 1 : 0; - case IEEE80211_OFDM_RATE_18MB: + rates_mask & LIBIPW_OFDM_RATE_12MB_MASK ? 1 : 0; + case LIBIPW_OFDM_RATE_18MB: return priv-> - rates_mask & IEEE80211_OFDM_RATE_18MB_MASK ? 1 : 0; - case IEEE80211_OFDM_RATE_24MB: + rates_mask & LIBIPW_OFDM_RATE_18MB_MASK ? 1 : 0; + case LIBIPW_OFDM_RATE_24MB: return priv-> - rates_mask & IEEE80211_OFDM_RATE_24MB_MASK ? 1 : 0; - case IEEE80211_OFDM_RATE_36MB: + rates_mask & LIBIPW_OFDM_RATE_24MB_MASK ? 1 : 0; + case LIBIPW_OFDM_RATE_36MB: return priv-> - rates_mask & IEEE80211_OFDM_RATE_36MB_MASK ? 1 : 0; - case IEEE80211_OFDM_RATE_48MB: + rates_mask & LIBIPW_OFDM_RATE_36MB_MASK ? 1 : 0; + case LIBIPW_OFDM_RATE_48MB: return priv-> - rates_mask & IEEE80211_OFDM_RATE_48MB_MASK ? 1 : 0; - case IEEE80211_OFDM_RATE_54MB: + rates_mask & LIBIPW_OFDM_RATE_48MB_MASK ? 1 : 0; + case LIBIPW_OFDM_RATE_54MB: return priv-> - rates_mask & IEEE80211_OFDM_RATE_54MB_MASK ? 1 : 0; + rates_mask & LIBIPW_OFDM_RATE_54MB_MASK ? 1 : 0; default: return 0; } @@ -5261,14 +5261,14 @@ static int ipw_is_rate_in_mask(struct ipw_priv *priv, int ieee_mode, u8 rate) /* B and G mixed */ switch (rate) { - case IEEE80211_CCK_RATE_1MB: - return priv->rates_mask & IEEE80211_CCK_RATE_1MB_MASK ? 1 : 0; - case IEEE80211_CCK_RATE_2MB: - return priv->rates_mask & IEEE80211_CCK_RATE_2MB_MASK ? 1 : 0; - case IEEE80211_CCK_RATE_5MB: - return priv->rates_mask & IEEE80211_CCK_RATE_5MB_MASK ? 1 : 0; - case IEEE80211_CCK_RATE_11MB: - return priv->rates_mask & IEEE80211_CCK_RATE_11MB_MASK ? 1 : 0; + case LIBIPW_CCK_RATE_1MB: + return priv->rates_mask & LIBIPW_CCK_RATE_1MB_MASK ? 1 : 0; + case LIBIPW_CCK_RATE_2MB: + return priv->rates_mask & LIBIPW_CCK_RATE_2MB_MASK ? 1 : 0; + case LIBIPW_CCK_RATE_5MB: + return priv->rates_mask & LIBIPW_CCK_RATE_5MB_MASK ? 1 : 0; + case LIBIPW_CCK_RATE_11MB: + return priv->rates_mask & LIBIPW_CCK_RATE_11MB_MASK ? 1 : 0; } /* If we are limited to B modulations, bail at this point */ @@ -5277,29 +5277,29 @@ static int ipw_is_rate_in_mask(struct ipw_priv *priv, int ieee_mode, u8 rate) /* G */ switch (rate) { - case IEEE80211_OFDM_RATE_6MB: - return priv->rates_mask & IEEE80211_OFDM_RATE_6MB_MASK ? 1 : 0; - case IEEE80211_OFDM_RATE_9MB: - return priv->rates_mask & IEEE80211_OFDM_RATE_9MB_MASK ? 1 : 0; - case IEEE80211_OFDM_RATE_12MB: - return priv->rates_mask & IEEE80211_OFDM_RATE_12MB_MASK ? 1 : 0; - case IEEE80211_OFDM_RATE_18MB: - return priv->rates_mask & IEEE80211_OFDM_RATE_18MB_MASK ? 1 : 0; - case IEEE80211_OFDM_RATE_24MB: - return priv->rates_mask & IEEE80211_OFDM_RATE_24MB_MASK ? 1 : 0; - case IEEE80211_OFDM_RATE_36MB: - return priv->rates_mask & IEEE80211_OFDM_RATE_36MB_MASK ? 1 : 0; - case IEEE80211_OFDM_RATE_48MB: - return priv->rates_mask & IEEE80211_OFDM_RATE_48MB_MASK ? 1 : 0; - case IEEE80211_OFDM_RATE_54MB: - return priv->rates_mask & IEEE80211_OFDM_RATE_54MB_MASK ? 1 : 0; + case LIBIPW_OFDM_RATE_6MB: + return priv->rates_mask & LIBIPW_OFDM_RATE_6MB_MASK ? 1 : 0; + case LIBIPW_OFDM_RATE_9MB: + return priv->rates_mask & LIBIPW_OFDM_RATE_9MB_MASK ? 1 : 0; + case LIBIPW_OFDM_RATE_12MB: + return priv->rates_mask & LIBIPW_OFDM_RATE_12MB_MASK ? 1 : 0; + case LIBIPW_OFDM_RATE_18MB: + return priv->rates_mask & LIBIPW_OFDM_RATE_18MB_MASK ? 1 : 0; + case LIBIPW_OFDM_RATE_24MB: + return priv->rates_mask & LIBIPW_OFDM_RATE_24MB_MASK ? 1 : 0; + case LIBIPW_OFDM_RATE_36MB: + return priv->rates_mask & LIBIPW_OFDM_RATE_36MB_MASK ? 1 : 0; + case LIBIPW_OFDM_RATE_48MB: + return priv->rates_mask & LIBIPW_OFDM_RATE_48MB_MASK ? 1 : 0; + case LIBIPW_OFDM_RATE_54MB: + return priv->rates_mask & LIBIPW_OFDM_RATE_54MB_MASK ? 1 : 0; } return 0; } static int ipw_compatible_rates(struct ipw_priv *priv, - const struct ieee80211_network *network, + const struct libipw_network *network, struct ipw_supported_rates *rates) { int num_rates, i; @@ -5311,7 +5311,7 @@ static int ipw_compatible_rates(struct ipw_priv *priv, if (!ipw_is_rate_in_mask(priv, network->mode, network->rates[i])) { - if (network->rates[i] & IEEE80211_BASIC_RATE_MASK) { + if (network->rates[i] & LIBIPW_BASIC_RATE_MASK) { IPW_DEBUG_SCAN("Adding masked mandatory " "rate %02X\n", network->rates[i]); @@ -5333,7 +5333,7 @@ static int ipw_compatible_rates(struct ipw_priv *priv, for (i = 0; i < num_rates; i++) { if (!ipw_is_rate_in_mask(priv, network->mode, network->rates_ex[i])) { - if (network->rates_ex[i] & IEEE80211_BASIC_RATE_MASK) { + if (network->rates_ex[i] & LIBIPW_BASIC_RATE_MASK) { IPW_DEBUG_SCAN("Adding masked mandatory " "rate %02X\n", network->rates_ex[i]); @@ -5369,73 +5369,73 @@ static void ipw_copy_rates(struct ipw_supported_rates *dest, static void ipw_add_cck_scan_rates(struct ipw_supported_rates *rates, u8 modulation, u32 rate_mask) { - u8 basic_mask = (IEEE80211_OFDM_MODULATION == modulation) ? - IEEE80211_BASIC_RATE_MASK : 0; + u8 basic_mask = (LIBIPW_OFDM_MODULATION == modulation) ? + LIBIPW_BASIC_RATE_MASK : 0; - if (rate_mask & IEEE80211_CCK_RATE_1MB_MASK) + if (rate_mask & LIBIPW_CCK_RATE_1MB_MASK) rates->supported_rates[rates->num_rates++] = - IEEE80211_BASIC_RATE_MASK | IEEE80211_CCK_RATE_1MB; + LIBIPW_BASIC_RATE_MASK | LIBIPW_CCK_RATE_1MB; - if (rate_mask & IEEE80211_CCK_RATE_2MB_MASK) + if (rate_mask & LIBIPW_CCK_RATE_2MB_MASK) rates->supported_rates[rates->num_rates++] = - IEEE80211_BASIC_RATE_MASK | IEEE80211_CCK_RATE_2MB; + LIBIPW_BASIC_RATE_MASK | LIBIPW_CCK_RATE_2MB; - if (rate_mask & IEEE80211_CCK_RATE_5MB_MASK) + if (rate_mask & LIBIPW_CCK_RATE_5MB_MASK) rates->supported_rates[rates->num_rates++] = basic_mask | - IEEE80211_CCK_RATE_5MB; + LIBIPW_CCK_RATE_5MB; - if (rate_mask & IEEE80211_CCK_RATE_11MB_MASK) + if (rate_mask & LIBIPW_CCK_RATE_11MB_MASK) rates->supported_rates[rates->num_rates++] = basic_mask | - IEEE80211_CCK_RATE_11MB; + LIBIPW_CCK_RATE_11MB; } static void ipw_add_ofdm_scan_rates(struct ipw_supported_rates *rates, u8 modulation, u32 rate_mask) { - u8 basic_mask = (IEEE80211_OFDM_MODULATION == modulation) ? - IEEE80211_BASIC_RATE_MASK : 0; + u8 basic_mask = (LIBIPW_OFDM_MODULATION == modulation) ? + LIBIPW_BASIC_RATE_MASK : 0; - if (rate_mask & IEEE80211_OFDM_RATE_6MB_MASK) + if (rate_mask & LIBIPW_OFDM_RATE_6MB_MASK) rates->supported_rates[rates->num_rates++] = basic_mask | - IEEE80211_OFDM_RATE_6MB; + LIBIPW_OFDM_RATE_6MB; - if (rate_mask & IEEE80211_OFDM_RATE_9MB_MASK) + if (rate_mask & LIBIPW_OFDM_RATE_9MB_MASK) rates->supported_rates[rates->num_rates++] = - IEEE80211_OFDM_RATE_9MB; + LIBIPW_OFDM_RATE_9MB; - if (rate_mask & IEEE80211_OFDM_RATE_12MB_MASK) + if (rate_mask & LIBIPW_OFDM_RATE_12MB_MASK) rates->supported_rates[rates->num_rates++] = basic_mask | - IEEE80211_OFDM_RATE_12MB; + LIBIPW_OFDM_RATE_12MB; - if (rate_mask & IEEE80211_OFDM_RATE_18MB_MASK) + if (rate_mask & LIBIPW_OFDM_RATE_18MB_MASK) rates->supported_rates[rates->num_rates++] = - IEEE80211_OFDM_RATE_18MB; + LIBIPW_OFDM_RATE_18MB; - if (rate_mask & IEEE80211_OFDM_RATE_24MB_MASK) + if (rate_mask & LIBIPW_OFDM_RATE_24MB_MASK) rates->supported_rates[rates->num_rates++] = basic_mask | - IEEE80211_OFDM_RATE_24MB; + LIBIPW_OFDM_RATE_24MB; - if (rate_mask & IEEE80211_OFDM_RATE_36MB_MASK) + if (rate_mask & LIBIPW_OFDM_RATE_36MB_MASK) rates->supported_rates[rates->num_rates++] = - IEEE80211_OFDM_RATE_36MB; + LIBIPW_OFDM_RATE_36MB; - if (rate_mask & IEEE80211_OFDM_RATE_48MB_MASK) + if (rate_mask & LIBIPW_OFDM_RATE_48MB_MASK) rates->supported_rates[rates->num_rates++] = - IEEE80211_OFDM_RATE_48MB; + LIBIPW_OFDM_RATE_48MB; - if (rate_mask & IEEE80211_OFDM_RATE_54MB_MASK) + if (rate_mask & LIBIPW_OFDM_RATE_54MB_MASK) rates->supported_rates[rates->num_rates++] = - IEEE80211_OFDM_RATE_54MB; + LIBIPW_OFDM_RATE_54MB; } struct ipw_network_match { - struct ieee80211_network *network; + struct libipw_network *network; struct ipw_supported_rates rates; }; static int ipw_find_adhoc_network(struct ipw_priv *priv, struct ipw_network_match *match, - struct ieee80211_network *network, + struct libipw_network *network, int roaming) { struct ipw_supported_rates rates; @@ -5556,7 +5556,7 @@ static int ipw_find_adhoc_network(struct ipw_priv *priv, } /* Filter out any incompatible freq / mode combinations */ - if (!ieee80211_is_valid_mode(priv->ieee, network->mode)) { + if (!libipw_is_valid_mode(priv->ieee, network->mode)) { IPW_DEBUG_MERGE("Network '%s (%pM)' excluded " "because of invalid frequency/mode " "combination.\n", @@ -5606,7 +5606,7 @@ static void ipw_merge_adhoc_network(struct work_struct *work) DECLARE_SSID_BUF(ssid); struct ipw_priv *priv = container_of(work, struct ipw_priv, merge_networks); - struct ieee80211_network *network = NULL; + struct libipw_network *network = NULL; struct ipw_network_match match = { .network = priv->assoc_network }; @@ -5648,7 +5648,7 @@ static void ipw_merge_adhoc_network(struct work_struct *work) static int ipw_best_network(struct ipw_priv *priv, struct ipw_network_match *match, - struct ieee80211_network *network, int roaming) + struct libipw_network *network, int roaming) { struct ipw_supported_rates rates; DECLARE_SSID_BUF(ssid); @@ -5782,7 +5782,7 @@ static int ipw_best_network(struct ipw_priv *priv, } /* Filter out any incompatible freq / mode combinations */ - if (!ieee80211_is_valid_mode(priv->ieee, network->mode)) { + if (!libipw_is_valid_mode(priv->ieee, network->mode)) { IPW_DEBUG_ASSOC("Network '%s (%pM)' excluded " "because of invalid frequency/mode " "combination.\n", @@ -5793,7 +5793,7 @@ static int ipw_best_network(struct ipw_priv *priv, } /* Filter out invalid channel in current GEO */ - if (!ieee80211_is_valid_channel(priv->ieee, network->channel)) { + if (!libipw_is_valid_channel(priv->ieee, network->channel)) { IPW_DEBUG_ASSOC("Network '%s (%pM)' excluded " "because of invalid channel in current GEO\n", print_ssid(ssid, network->ssid, @@ -5839,9 +5839,9 @@ static int ipw_best_network(struct ipw_priv *priv, } static void ipw_adhoc_create(struct ipw_priv *priv, - struct ieee80211_network *network) + struct libipw_network *network) { - const struct ieee80211_geo *geo = ieee80211_get_geo(priv->ieee); + const struct libipw_geo *geo = libipw_get_geo(priv->ieee); int i; /* @@ -5856,25 +5856,25 @@ static void ipw_adhoc_create(struct ipw_priv *priv, * FW fatal error. * */ - switch (ieee80211_is_valid_channel(priv->ieee, priv->channel)) { - case IEEE80211_52GHZ_BAND: + switch (libipw_is_valid_channel(priv->ieee, priv->channel)) { + case LIBIPW_52GHZ_BAND: network->mode = IEEE_A; - i = ieee80211_channel_to_index(priv->ieee, priv->channel); + i = libipw_channel_to_index(priv->ieee, priv->channel); BUG_ON(i == -1); - if (geo->a[i].flags & IEEE80211_CH_PASSIVE_ONLY) { + if (geo->a[i].flags & LIBIPW_CH_PASSIVE_ONLY) { IPW_WARNING("Overriding invalid channel\n"); priv->channel = geo->a[0].channel; } break; - case IEEE80211_24GHZ_BAND: + case LIBIPW_24GHZ_BAND: if (priv->ieee->mode & IEEE_G) network->mode = IEEE_G; else network->mode = IEEE_B; - i = ieee80211_channel_to_index(priv->ieee, priv->channel); + i = libipw_channel_to_index(priv->ieee, priv->channel); BUG_ON(i == -1); - if (geo->bg[i].flags & IEEE80211_CH_PASSIVE_ONLY) { + if (geo->bg[i].flags & LIBIPW_CH_PASSIVE_ONLY) { IPW_WARNING("Overriding invalid channel\n"); priv->channel = geo->bg[0].channel; } @@ -6110,9 +6110,9 @@ static void ipw_set_fixed_rate(struct ipw_priv *priv, int mode) * Tx rates */ switch (priv->ieee->freq_band) { - case IEEE80211_52GHZ_BAND: /* A only */ + case LIBIPW_52GHZ_BAND: /* A only */ /* IEEE_A */ - if (priv->rates_mask & ~IEEE80211_OFDM_RATES_MASK) { + if (priv->rates_mask & ~LIBIPW_OFDM_RATES_MASK) { /* Invalid fixed rate mask */ IPW_DEBUG_WX ("invalid fixed rate mask in ipw_set_fixed_rate\n"); @@ -6120,13 +6120,13 @@ static void ipw_set_fixed_rate(struct ipw_priv *priv, int mode) break; } - new_tx_rates >>= IEEE80211_OFDM_SHIFT_MASK_A; + new_tx_rates >>= LIBIPW_OFDM_SHIFT_MASK_A; break; default: /* 2.4Ghz or Mixed */ /* IEEE_B */ if (mode == IEEE_B) { - if (new_tx_rates & ~IEEE80211_CCK_RATES_MASK) { + if (new_tx_rates & ~LIBIPW_CCK_RATES_MASK) { /* Invalid fixed rate mask */ IPW_DEBUG_WX ("invalid fixed rate mask in ipw_set_fixed_rate\n"); @@ -6136,8 +6136,8 @@ static void ipw_set_fixed_rate(struct ipw_priv *priv, int mode) } /* IEEE_G */ - if (new_tx_rates & ~(IEEE80211_CCK_RATES_MASK | - IEEE80211_OFDM_RATES_MASK)) { + if (new_tx_rates & ~(LIBIPW_CCK_RATES_MASK | + LIBIPW_OFDM_RATES_MASK)) { /* Invalid fixed rate mask */ IPW_DEBUG_WX ("invalid fixed rate mask in ipw_set_fixed_rate\n"); @@ -6145,19 +6145,19 @@ static void ipw_set_fixed_rate(struct ipw_priv *priv, int mode) break; } - if (IEEE80211_OFDM_RATE_6MB_MASK & new_tx_rates) { - mask |= (IEEE80211_OFDM_RATE_6MB_MASK >> 1); - new_tx_rates &= ~IEEE80211_OFDM_RATE_6MB_MASK; + if (LIBIPW_OFDM_RATE_6MB_MASK & new_tx_rates) { + mask |= (LIBIPW_OFDM_RATE_6MB_MASK >> 1); + new_tx_rates &= ~LIBIPW_OFDM_RATE_6MB_MASK; } - if (IEEE80211_OFDM_RATE_9MB_MASK & new_tx_rates) { - mask |= (IEEE80211_OFDM_RATE_9MB_MASK >> 1); - new_tx_rates &= ~IEEE80211_OFDM_RATE_9MB_MASK; + if (LIBIPW_OFDM_RATE_9MB_MASK & new_tx_rates) { + mask |= (LIBIPW_OFDM_RATE_9MB_MASK >> 1); + new_tx_rates &= ~LIBIPW_OFDM_RATE_9MB_MASK; } - if (IEEE80211_OFDM_RATE_12MB_MASK & new_tx_rates) { - mask |= (IEEE80211_OFDM_RATE_12MB_MASK >> 1); - new_tx_rates &= ~IEEE80211_OFDM_RATE_12MB_MASK; + if (LIBIPW_OFDM_RATE_12MB_MASK & new_tx_rates) { + mask |= (LIBIPW_OFDM_RATE_12MB_MASK >> 1); + new_tx_rates &= ~LIBIPW_OFDM_RATE_12MB_MASK; } new_tx_rates |= mask; @@ -6190,12 +6190,12 @@ static void ipw_add_scan_channels(struct ipw_priv *priv, int scan_type) { int channel_index = 0; - const struct ieee80211_geo *geo; + const struct libipw_geo *geo; int i; - geo = ieee80211_get_geo(priv->ieee); + geo = libipw_get_geo(priv->ieee); - if (priv->ieee->freq_band & IEEE80211_52GHZ_BAND) { + if (priv->ieee->freq_band & LIBIPW_52GHZ_BAND) { int start = channel_index; for (i = 0; i < geo->a_channels; i++) { if ((priv->status & STATUS_ASSOCIATED) && @@ -6205,7 +6205,7 @@ static void ipw_add_scan_channels(struct ipw_priv *priv, scan->channels_list[channel_index] = geo->a[i].channel; ipw_set_scan_type(scan, channel_index, geo->a[i]. - flags & IEEE80211_CH_PASSIVE_ONLY ? + flags & LIBIPW_CH_PASSIVE_ONLY ? IPW_SCAN_PASSIVE_FULL_DWELL_SCAN : scan_type); } @@ -6217,11 +6217,11 @@ static void ipw_add_scan_channels(struct ipw_priv *priv, } } - if (priv->ieee->freq_band & IEEE80211_24GHZ_BAND) { + if (priv->ieee->freq_band & LIBIPW_24GHZ_BAND) { int start = channel_index; if (priv->config & CFG_SPEED_SCAN) { int index; - u8 channels[IEEE80211_24GHZ_CHANNELS] = { + u8 channels[LIBIPW_24GHZ_CHANNELS] = { /* nop out the list */ [0] = 0 }; @@ -6253,11 +6253,11 @@ static void ipw_add_scan_channels(struct ipw_priv *priv, channel_index++; scan->channels_list[channel_index] = channel; index = - ieee80211_channel_to_index(priv->ieee, channel); + libipw_channel_to_index(priv->ieee, channel); ipw_set_scan_type(scan, channel_index, geo->bg[index]. flags & - IEEE80211_CH_PASSIVE_ONLY ? + LIBIPW_CH_PASSIVE_ONLY ? IPW_SCAN_PASSIVE_FULL_DWELL_SCAN : scan_type); } @@ -6272,7 +6272,7 @@ static void ipw_add_scan_channels(struct ipw_priv *priv, ipw_set_scan_type(scan, channel_index, geo->bg[i]. flags & - IEEE80211_CH_PASSIVE_ONLY ? + LIBIPW_CH_PASSIVE_ONLY ? IPW_SCAN_PASSIVE_FULL_DWELL_SCAN : scan_type); } @@ -6339,7 +6339,7 @@ static int ipw_request_scan_helper(struct ipw_priv *priv, int type, int direct) } memset(&scan, 0, sizeof(scan)); - scan.full_scan_index = cpu_to_le32(ieee80211_get_scans(priv->ieee)); + scan.full_scan_index = cpu_to_le32(libipw_get_scans(priv->ieee)); if (type == IW_SCAN_TYPE_PASSIVE) { IPW_DEBUG_WX("use passive scanning\n"); @@ -6370,13 +6370,13 @@ static int ipw_request_scan_helper(struct ipw_priv *priv, int type, int direct) u8 channel; u8 band = 0; - switch (ieee80211_is_valid_channel(priv->ieee, priv->channel)) { - case IEEE80211_52GHZ_BAND: + switch (libipw_is_valid_channel(priv->ieee, priv->channel)) { + case LIBIPW_52GHZ_BAND: band = (u8) (IPW_A_MODE << 6) | 1; channel = priv->channel; break; - case IEEE80211_24GHZ_BAND: + case LIBIPW_24GHZ_BAND: band = (u8) (IPW_B_MODE << 6) | 1; channel = priv->channel; break; @@ -6497,8 +6497,8 @@ static int ipw_wpa_enable(struct ipw_priv *priv, int value) static int ipw_wpa_set_auth_algs(struct ipw_priv *priv, int value) { - struct ieee80211_device *ieee = priv->ieee; - struct ieee80211_security sec = { + struct libipw_device *ieee = priv->ieee; + struct libipw_security sec = { .flags = SEC_AUTH_MODE, }; int ret = 0; @@ -6548,8 +6548,8 @@ static int ipw_wx_set_genie(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); - struct ieee80211_device *ieee = priv->ieee; + struct ipw_priv *priv = libipw_priv(dev); + struct libipw_device *ieee = priv->ieee; u8 *buf; int err = 0; @@ -6584,8 +6584,8 @@ static int ipw_wx_get_genie(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); - struct ieee80211_device *ieee = priv->ieee; + struct ipw_priv *priv = libipw_priv(dev); + struct libipw_device *ieee = priv->ieee; int err = 0; if (ieee->wpa_ie_len == 0 || ieee->wpa_ie == NULL) { @@ -6627,8 +6627,8 @@ static int ipw_wx_set_auth(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); - struct ieee80211_device *ieee = priv->ieee; + struct ipw_priv *priv = libipw_priv(dev); + struct libipw_device *ieee = priv->ieee; struct iw_param *param = &wrqu->param; struct lib80211_crypt_data *crypt; unsigned long flags; @@ -6679,7 +6679,7 @@ static int ipw_wx_set_auth(struct net_device *dev, * can use this to determine if the CAP_PRIVACY_ON bit should * be set. */ - struct ieee80211_security sec = { + struct libipw_security sec = { .flags = SEC_ENABLED, .enabled = param->value, }; @@ -6727,8 +6727,8 @@ static int ipw_wx_get_auth(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); - struct ieee80211_device *ieee = priv->ieee; + struct ipw_priv *priv = libipw_priv(dev); + struct libipw_device *ieee = priv->ieee; struct lib80211_crypt_data *crypt; struct iw_param *param = &wrqu->param; int ret = 0; @@ -6786,7 +6786,7 @@ static int ipw_wx_set_encodeext(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); struct iw_encode_ext *ext = (struct iw_encode_ext *)extra; if (hwcrypto) { @@ -6808,7 +6808,7 @@ static int ipw_wx_set_encodeext(struct net_device *dev, } } - return ieee80211_wx_set_encodeext(priv->ieee, info, wrqu, extra); + return libipw_wx_set_encodeext(priv->ieee, info, wrqu, extra); } /* SIOCGIWENCODEEXT */ @@ -6816,8 +6816,8 @@ static int ipw_wx_get_encodeext(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); - return ieee80211_wx_get_encodeext(priv->ieee, info, wrqu, extra); + struct ipw_priv *priv = libipw_priv(dev); + return libipw_wx_get_encodeext(priv->ieee, info, wrqu, extra); } /* SIOCSIWMLME */ @@ -6825,7 +6825,7 @@ static int ipw_wx_set_mlme(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); struct iw_mlme *mlme = (struct iw_mlme *)extra; __le16 reason; @@ -6875,9 +6875,9 @@ static u8 ipw_qos_current_mode(struct ipw_priv * priv) */ static int ipw_qos_handle_probe_response(struct ipw_priv *priv, int active_network, - struct ieee80211_network *network) + struct libipw_network *network) { - u32 size = sizeof(struct ieee80211_qos_parameters); + u32 size = sizeof(struct libipw_qos_parameters); if (network->capability & WLAN_CAPABILITY_IBSS) network->qos_data.active = network->qos_data.supported; @@ -6935,12 +6935,12 @@ static int ipw_qos_handle_probe_response(struct ipw_priv *priv, * IPW_CMD_QOS_PARAMETERS and IPW_CMD_WME_INFO */ static int ipw_qos_activate(struct ipw_priv *priv, - struct ieee80211_qos_data *qos_network_data) + struct libipw_qos_data *qos_network_data) { int err; - struct ieee80211_qos_parameters qos_parameters[QOS_QOS_SETS]; - struct ieee80211_qos_parameters *active_one = NULL; - u32 size = sizeof(struct ieee80211_qos_parameters); + struct libipw_qos_parameters qos_parameters[QOS_QOS_SETS]; + struct libipw_qos_parameters *active_one = NULL; + u32 size = sizeof(struct libipw_qos_parameters); u32 burst_duration; int i; u8 type; @@ -7001,7 +7001,7 @@ static int ipw_qos_activate(struct ipw_priv *priv, IPW_DEBUG_QOS("QoS sending IPW_CMD_QOS_PARAMETERS\n"); err = ipw_send_qos_params_command(priv, - (struct ieee80211_qos_parameters *) + (struct libipw_qos_parameters *) &(qos_parameters[0])); if (err) IPW_DEBUG_QOS("QoS IPW_CMD_QOS_PARAMETERS failed\n"); @@ -7015,13 +7015,13 @@ static int ipw_qos_activate(struct ipw_priv *priv, static int ipw_qos_set_info_element(struct ipw_priv *priv) { int ret = 0; - struct ieee80211_qos_information_element qos_info; + struct libipw_qos_information_element qos_info; if (priv == NULL) return -1; qos_info.elementID = QOS_ELEMENT_ID; - qos_info.length = sizeof(struct ieee80211_qos_information_element) - 2; + qos_info.length = sizeof(struct libipw_qos_information_element) - 2; qos_info.version = QOS_VERSION_1; qos_info.ac_info = 0; @@ -7041,11 +7041,11 @@ static int ipw_qos_set_info_element(struct ipw_priv *priv) * Set the QoS parameter with the association request structure */ static int ipw_qos_association(struct ipw_priv *priv, - struct ieee80211_network *network) + struct libipw_network *network) { int err = 0; - struct ieee80211_qos_data *qos_data = NULL; - struct ieee80211_qos_data ibss_data = { + struct libipw_qos_data *qos_data = NULL; + struct libipw_qos_data ibss_data = { .supported = 1, .active = 1, }; @@ -7087,11 +7087,11 @@ static int ipw_qos_association(struct ipw_priv *priv, * setting */ static int ipw_qos_association_resp(struct ipw_priv *priv, - struct ieee80211_network *network) + struct libipw_network *network) { int ret = 0; unsigned long flags; - u32 size = sizeof(struct ieee80211_qos_parameters); + u32 size = sizeof(struct libipw_qos_parameters); int set_qos_param = 0; if ((priv == NULL) || (network == NULL) || @@ -7107,7 +7107,7 @@ static int ipw_qos_association_resp(struct ipw_priv *priv, spin_lock_irqsave(&priv->ieee->lock, flags); if (network->flags & NETWORK_HAS_QOS_PARAMETERS) { memcpy(&priv->assoc_network->qos_data, &network->qos_data, - sizeof(struct ieee80211_qos_data)); + sizeof(struct libipw_qos_data)); priv->assoc_network->qos_data.active = 1; if ((network->qos_data.old_param_count != network->qos_data.param_count)) { @@ -7143,7 +7143,7 @@ static u32 ipw_qos_get_burst_duration(struct ipw_priv *priv) if ((priv == NULL)) return 0; - if (!(priv->ieee->modulation & IEEE80211_OFDM_MODULATION)) + if (!(priv->ieee->modulation & LIBIPW_OFDM_MODULATION)) ret = priv->qos_data.burst_duration_CCK; else ret = priv->qos_data.burst_duration_OFDM; @@ -7195,8 +7195,8 @@ static int ipw_get_tx_queue_number(struct ipw_priv *priv, u16 priority) static int ipw_is_qos_active(struct net_device *dev, struct sk_buff *skb) { - struct ipw_priv *priv = ieee80211_priv(dev); - struct ieee80211_qos_data *qos_data = NULL; + struct ipw_priv *priv = libipw_priv(dev); + struct libipw_qos_data *qos_data = NULL; int active, supported; u8 *daddr = skb->data + ETH_ALEN; int unicast = !is_multicast_ether_addr(daddr); @@ -7260,10 +7260,10 @@ static void ipw_bg_qos_activate(struct work_struct *work) } static int ipw_handle_probe_response(struct net_device *dev, - struct ieee80211_probe_response *resp, - struct ieee80211_network *network) + struct libipw_probe_response *resp, + struct libipw_network *network) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); int active_network = ((priv->status & STATUS_ASSOCIATED) && (network == priv->assoc_network)); @@ -7273,10 +7273,10 @@ static int ipw_handle_probe_response(struct net_device *dev, } static int ipw_handle_beacon(struct net_device *dev, - struct ieee80211_beacon *resp, - struct ieee80211_network *network) + struct libipw_beacon *resp, + struct libipw_network *network) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); int active_network = ((priv->status & STATUS_ASSOCIATED) && (network == priv->assoc_network)); @@ -7286,22 +7286,22 @@ static int ipw_handle_beacon(struct net_device *dev, } static int ipw_handle_assoc_response(struct net_device *dev, - struct ieee80211_assoc_response *resp, - struct ieee80211_network *network) + struct libipw_assoc_response *resp, + struct libipw_network *network) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); ipw_qos_association_resp(priv, network); return 0; } -static int ipw_send_qos_params_command(struct ipw_priv *priv, struct ieee80211_qos_parameters +static int ipw_send_qos_params_command(struct ipw_priv *priv, struct libipw_qos_parameters *qos_param) { return ipw_send_cmd_pdu(priv, IPW_CMD_QOS_PARAMETERS, sizeof(*qos_param) * 3, qos_param); } -static int ipw_send_qos_info_command(struct ipw_priv *priv, struct ieee80211_qos_information_element +static int ipw_send_qos_info_command(struct ipw_priv *priv, struct libipw_qos_information_element *qos_param) { return ipw_send_cmd_pdu(priv, IPW_CMD_WME_INFO, sizeof(*qos_param), @@ -7311,7 +7311,7 @@ static int ipw_send_qos_info_command(struct ipw_priv *priv, struct ieee80211_qos #endif /* CONFIG_IPW2200_QOS */ static int ipw_associate_network(struct ipw_priv *priv, - struct ieee80211_network *network, + struct libipw_network *network, struct ipw_supported_rates *rates, int roaming) { int err; @@ -7493,7 +7493,7 @@ static int ipw_associate_network(struct ipw_priv *priv, static void ipw_roam(void *data) { struct ipw_priv *priv = data; - struct ieee80211_network *network = NULL; + struct libipw_network *network = NULL; struct ipw_network_match match = { .network = priv->assoc_network }; @@ -7568,7 +7568,7 @@ static int ipw_associate(void *data) { struct ipw_priv *priv = data; - struct ieee80211_network *network = NULL; + struct libipw_network *network = NULL; struct ipw_network_match match = { .network = NULL }; @@ -7622,8 +7622,8 @@ static int ipw_associate(void *data) priv->config & CFG_STATIC_CHANNEL) { /* Use oldest network if the free list is empty */ if (list_empty(&priv->ieee->network_free_list)) { - struct ieee80211_network *oldest = NULL; - struct ieee80211_network *target; + struct libipw_network *oldest = NULL; + struct libipw_network *target; list_for_each_entry(target, &priv->ieee->network_list, list) { if ((oldest == NULL) || @@ -7644,7 +7644,7 @@ static int ipw_associate(void *data) } element = priv->ieee->network_free_list.next; - network = list_entry(element, struct ieee80211_network, list); + network = list_entry(element, struct libipw_network, list); ipw_adhoc_create(priv, network); rates = &priv->rates; list_del(element); @@ -7700,18 +7700,18 @@ static void ipw_rebuild_decrypted_skb(struct ipw_priv *priv, switch (priv->ieee->sec.level) { case SEC_LEVEL_3: /* Remove CCMP HDR */ - memmove(skb->data + IEEE80211_3ADDR_LEN, - skb->data + IEEE80211_3ADDR_LEN + 8, - skb->len - IEEE80211_3ADDR_LEN - 8); + memmove(skb->data + LIBIPW_3ADDR_LEN, + skb->data + LIBIPW_3ADDR_LEN + 8, + skb->len - LIBIPW_3ADDR_LEN - 8); skb_trim(skb, skb->len - 16); /* CCMP_HDR_LEN + CCMP_MIC_LEN */ break; case SEC_LEVEL_2: break; case SEC_LEVEL_1: /* Remove IV */ - memmove(skb->data + IEEE80211_3ADDR_LEN, - skb->data + IEEE80211_3ADDR_LEN + 4, - skb->len - IEEE80211_3ADDR_LEN - 4); + memmove(skb->data + LIBIPW_3ADDR_LEN, + skb->data + LIBIPW_3ADDR_LEN + 4, + skb->len - LIBIPW_3ADDR_LEN - 4); skb_trim(skb, skb->len - 8); /* IV + ICV */ break; case SEC_LEVEL_0: @@ -7725,10 +7725,10 @@ static void ipw_rebuild_decrypted_skb(struct ipw_priv *priv, static void ipw_handle_data_packet(struct ipw_priv *priv, struct ipw_rx_mem_buffer *rxb, - struct ieee80211_rx_stats *stats) + struct libipw_rx_stats *stats) { struct net_device *dev = priv->net_dev; - struct ieee80211_hdr_4addr *hdr; + struct libipw_hdr_4addr *hdr; struct ipw_rx_packet *pkt = (struct ipw_rx_packet *)rxb->skb->data; /* We received data from the HW, so stop the watchdog */ @@ -7758,15 +7758,15 @@ static void ipw_handle_data_packet(struct ipw_priv *priv, IPW_DEBUG_RX("Rx packet of %d bytes.\n", rxb->skb->len); /* HW decrypt will not clear the WEP bit, MIC, PN, etc. */ - hdr = (struct ieee80211_hdr_4addr *)rxb->skb->data; + hdr = (struct libipw_hdr_4addr *)rxb->skb->data; if (priv->ieee->iw_mode != IW_MODE_MONITOR && (is_multicast_ether_addr(hdr->addr1) ? !priv->ieee->host_mc_decrypt : !priv->ieee->host_decrypt)) ipw_rebuild_decrypted_skb(priv, rxb->skb); - if (!ieee80211_rx(priv->ieee, rxb->skb, stats)) + if (!libipw_rx(priv->ieee, rxb->skb, stats)) dev->stats.rx_errors++; - else { /* ieee80211_rx succeeded, so it now owns the SKB */ + else { /* libipw_rx succeeded, so it now owns the SKB */ rxb->skb = NULL; __ipw_led_activity_on(priv); } @@ -7775,7 +7775,7 @@ static void ipw_handle_data_packet(struct ipw_priv *priv, #ifdef CONFIG_IPW2200_RADIOTAP static void ipw_handle_data_packet_monitor(struct ipw_priv *priv, struct ipw_rx_mem_buffer *rxb, - struct ieee80211_rx_stats *stats) + struct libipw_rx_stats *stats) { struct net_device *dev = priv->net_dev; struct ipw_rx_packet *pkt = (struct ipw_rx_packet *)rxb->skb->data; @@ -7921,9 +7921,9 @@ static void ipw_handle_data_packet_monitor(struct ipw_priv *priv, IPW_DEBUG_RX("Rx packet of %d bytes.\n", rxb->skb->len); - if (!ieee80211_rx(priv->ieee, rxb->skb, stats)) + if (!libipw_rx(priv->ieee, rxb->skb, stats)) dev->stats.rx_errors++; - else { /* ieee80211_rx succeeded, so it now owns the SKB */ + else { /* libipw_rx succeeded, so it now owns the SKB */ rxb->skb = NULL; /* no LED during capture */ } @@ -7931,28 +7931,28 @@ static void ipw_handle_data_packet_monitor(struct ipw_priv *priv, #endif #ifdef CONFIG_IPW2200_PROMISCUOUS -#define ieee80211_is_probe_response(fc) \ +#define libipw_is_probe_response(fc) \ ((fc & IEEE80211_FCTL_FTYPE) == IEEE80211_FTYPE_MGMT && \ (fc & IEEE80211_FCTL_STYPE) == IEEE80211_STYPE_PROBE_RESP ) -#define ieee80211_is_management(fc) \ +#define libipw_is_management(fc) \ ((fc & IEEE80211_FCTL_FTYPE) == IEEE80211_FTYPE_MGMT) -#define ieee80211_is_control(fc) \ +#define libipw_is_control(fc) \ ((fc & IEEE80211_FCTL_FTYPE) == IEEE80211_FTYPE_CTL) -#define ieee80211_is_data(fc) \ +#define libipw_is_data(fc) \ ((fc & IEEE80211_FCTL_FTYPE) == IEEE80211_FTYPE_DATA) -#define ieee80211_is_assoc_request(fc) \ +#define libipw_is_assoc_request(fc) \ ((fc & IEEE80211_FCTL_STYPE) == IEEE80211_STYPE_ASSOC_REQ) -#define ieee80211_is_reassoc_request(fc) \ +#define libipw_is_reassoc_request(fc) \ ((fc & IEEE80211_FCTL_STYPE) == IEEE80211_STYPE_REASSOC_REQ) static void ipw_handle_promiscuous_rx(struct ipw_priv *priv, struct ipw_rx_mem_buffer *rxb, - struct ieee80211_rx_stats *stats) + struct libipw_rx_stats *stats) { struct net_device *dev = priv->prom_net_dev; struct ipw_rx_packet *pkt = (struct ipw_rx_packet *)rxb->skb->data; @@ -8002,17 +8002,17 @@ static void ipw_handle_promiscuous_rx(struct ipw_priv *priv, } hdr = (void *)rxb->skb->data + IPW_RX_FRAME_SIZE; - if (ieee80211_is_management(le16_to_cpu(hdr->frame_control))) { + if (libipw_is_management(le16_to_cpu(hdr->frame_control))) { if (filter & IPW_PROM_NO_MGMT) return; if (filter & IPW_PROM_MGMT_HEADER_ONLY) hdr_only = 1; - } else if (ieee80211_is_control(le16_to_cpu(hdr->frame_control))) { + } else if (libipw_is_control(le16_to_cpu(hdr->frame_control))) { if (filter & IPW_PROM_NO_CTL) return; if (filter & IPW_PROM_CTL_HEADER_ONLY) hdr_only = 1; - } else if (ieee80211_is_data(le16_to_cpu(hdr->frame_control))) { + } else if (libipw_is_data(le16_to_cpu(hdr->frame_control))) { if (filter & IPW_PROM_NO_DATA) return; if (filter & IPW_PROM_DATA_HEADER_ONLY) @@ -8030,7 +8030,7 @@ static void ipw_handle_promiscuous_rx(struct ipw_priv *priv, ipw_rt = (void *)skb->data; if (hdr_only) - len = ieee80211_get_hdrlen(le16_to_cpu(hdr->frame_control)); + len = libipw_get_hdrlen(le16_to_cpu(hdr->frame_control)); memcpy(ipw_rt->payload, hdr, len); @@ -8127,7 +8127,7 @@ static void ipw_handle_promiscuous_rx(struct ipw_priv *priv, IPW_DEBUG_RX("Rx packet of %d bytes.\n", skb->len); - if (!ieee80211_rx(priv->prom_priv->ieee, skb, stats)) { + if (!libipw_rx(priv->prom_priv->ieee, skb, stats)) { dev->stats.rx_errors++; dev_kfree_skb_any(skb); } @@ -8135,7 +8135,7 @@ static void ipw_handle_promiscuous_rx(struct ipw_priv *priv, #endif static int is_network_packet(struct ipw_priv *priv, - struct ieee80211_hdr_4addr *header) + struct libipw_hdr_4addr *header) { /* Filter incoming packets to determine if they are targetted toward * this network, discarding packets coming from ourselves */ @@ -8173,7 +8173,7 @@ static int is_network_packet(struct ipw_priv *priv, #define IPW_PACKET_RETRY_TIME HZ static int is_duplicate_packet(struct ipw_priv *priv, - struct ieee80211_hdr_4addr *header) + struct libipw_hdr_4addr *header) { u16 sc = le16_to_cpu(header->seq_ctl); u16 seq = WLAN_GET_SEQ_SEQ(sc); @@ -8247,14 +8247,14 @@ static int is_duplicate_packet(struct ipw_priv *priv, static void ipw_handle_mgmt_packet(struct ipw_priv *priv, struct ipw_rx_mem_buffer *rxb, - struct ieee80211_rx_stats *stats) + struct libipw_rx_stats *stats) { struct sk_buff *skb = rxb->skb; struct ipw_rx_packet *pkt = (struct ipw_rx_packet *)skb->data; - struct ieee80211_hdr_4addr *header = (struct ieee80211_hdr_4addr *) + struct libipw_hdr_4addr *header = (struct libipw_hdr_4addr *) (skb->data + IPW_RX_FRAME_SIZE); - ieee80211_rx_mgt(priv->ieee, header, stats); + libipw_rx_mgt(priv->ieee, header, stats); if (priv->ieee->iw_mode == IW_MODE_ADHOC && ((WLAN_FC_GET_STYPE(le16_to_cpu(header->frame_ctl)) == @@ -8276,12 +8276,12 @@ static void ipw_handle_mgmt_packet(struct ipw_priv *priv, /* Advance past the ipw packet header to the 802.11 frame */ skb_pull(skb, IPW_RX_FRAME_SIZE); - /* Push the ieee80211_rx_stats before the 802.11 frame */ + /* Push the libipw_rx_stats before the 802.11 frame */ memcpy(skb_push(skb, sizeof(*stats)), stats, sizeof(*stats)); skb->dev = priv->ieee->dev; - /* Point raw at the ieee80211_stats */ + /* Point raw at the libipw_stats */ skb_reset_mac_header(skb); skb->pkt_type = PACKET_OTHERHOST; @@ -8301,7 +8301,7 @@ static void ipw_rx(struct ipw_priv *priv) { struct ipw_rx_mem_buffer *rxb; struct ipw_rx_packet *pkt; - struct ieee80211_hdr_4addr *header; + struct libipw_hdr_4addr *header; u32 r, w, i; u8 network_packet; u8 fill_rx = 0; @@ -8332,7 +8332,7 @@ static void ipw_rx(struct ipw_priv *priv) switch (pkt->header.message_type) { case RX_FRAME_TYPE: /* 802.11 frame */ { - struct ieee80211_rx_stats stats = { + struct libipw_rx_stats stats = { .rssi = pkt->u.frame.rssi_dbm - IPW_RSSI_TO_DBM, .signal = @@ -8347,19 +8347,19 @@ static void ipw_rx(struct ipw_priv *priv) .freq = (pkt->u.frame. control & (1 << 0)) ? - IEEE80211_24GHZ_BAND : - IEEE80211_52GHZ_BAND, + LIBIPW_24GHZ_BAND : + LIBIPW_52GHZ_BAND, .len = le16_to_cpu(pkt->u.frame.length), }; if (stats.rssi != 0) - stats.mask |= IEEE80211_STATMASK_RSSI; + stats.mask |= LIBIPW_STATMASK_RSSI; if (stats.signal != 0) - stats.mask |= IEEE80211_STATMASK_SIGNAL; + stats.mask |= LIBIPW_STATMASK_SIGNAL; if (stats.noise != 0) - stats.mask |= IEEE80211_STATMASK_NOISE; + stats.mask |= LIBIPW_STATMASK_NOISE; if (stats.rate != 0) - stats.mask |= IEEE80211_STATMASK_RATE; + stats.mask |= LIBIPW_STATMASK_RATE; priv->rx_packets++; @@ -8384,7 +8384,7 @@ static void ipw_rx(struct ipw_priv *priv) #endif header = - (struct ieee80211_hdr_4addr *)(rxb->skb-> + (struct libipw_hdr_4addr *)(rxb->skb-> data + IPW_RX_FRAME_SIZE); /* TODO: Check Ad-Hoc dest/source and make sure @@ -8407,7 +8407,7 @@ static void ipw_rx(struct ipw_priv *priv) le16_to_cpu(pkt->u.frame.length)); if (le16_to_cpu(pkt->u.frame.length) < - ieee80211_get_hdrlen(le16_to_cpu( + libipw_get_hdrlen(le16_to_cpu( header->frame_ctl))) { IPW_DEBUG_DROP ("Received packet is too small. " @@ -8592,9 +8592,9 @@ static int ipw_sw_reset(struct ipw_priv *priv, int option) ": Detected Intel PRO/Wireless 2915ABG Network " "Connection\n"); priv->ieee->abg_true = 1; - band = IEEE80211_52GHZ_BAND | IEEE80211_24GHZ_BAND; - modulation = IEEE80211_OFDM_MODULATION | - IEEE80211_CCK_MODULATION; + band = LIBIPW_52GHZ_BAND | LIBIPW_24GHZ_BAND; + modulation = LIBIPW_OFDM_MODULATION | + LIBIPW_CCK_MODULATION; priv->adapter = IPW_2915ABG; priv->ieee->mode = IEEE_A | IEEE_G | IEEE_B; } else { @@ -8604,9 +8604,9 @@ static int ipw_sw_reset(struct ipw_priv *priv, int option) "Connection\n"); priv->ieee->abg_true = 0; - band = IEEE80211_24GHZ_BAND; - modulation = IEEE80211_OFDM_MODULATION | - IEEE80211_CCK_MODULATION; + band = LIBIPW_24GHZ_BAND; + modulation = LIBIPW_OFDM_MODULATION | + LIBIPW_CCK_MODULATION; priv->adapter = IPW_2200BG; priv->ieee->mode = IEEE_G | IEEE_B; } @@ -8614,7 +8614,7 @@ static int ipw_sw_reset(struct ipw_priv *priv, int option) priv->ieee->freq_band = band; priv->ieee->modulation = modulation; - priv->rates_mask = IEEE80211_DEFAULT_RATES_MASK; + priv->rates_mask = LIBIPW_DEFAULT_RATES_MASK; priv->disassociate_threshold = IPW_MB_DISASSOCIATE_THRESHOLD_DEFAULT; priv->roaming_threshold = IPW_MB_ROAMING_THRESHOLD_DEFAULT; @@ -8644,7 +8644,7 @@ static int ipw_wx_get_name(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); mutex_lock(&priv->mutex); if (priv->status & STATUS_RF_KILL_MASK) strcpy(wrqu->name, "radio off"); @@ -8714,8 +8714,8 @@ static int ipw_wx_set_freq(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); - const struct ieee80211_geo *geo = ieee80211_get_geo(priv->ieee); + struct ipw_priv *priv = libipw_priv(dev); + const struct libipw_geo *geo = libipw_get_geo(priv->ieee); struct iw_freq *fwrq = &wrqu->freq; int ret = 0, i; u8 channel, flags; @@ -8730,23 +8730,23 @@ static int ipw_wx_set_freq(struct net_device *dev, } /* if setting by freq convert to channel */ if (fwrq->e == 1) { - channel = ieee80211_freq_to_channel(priv->ieee, fwrq->m); + channel = libipw_freq_to_channel(priv->ieee, fwrq->m); if (channel == 0) return -EINVAL; } else channel = fwrq->m; - if (!(band = ieee80211_is_valid_channel(priv->ieee, channel))) + if (!(band = libipw_is_valid_channel(priv->ieee, channel))) return -EINVAL; if (priv->ieee->iw_mode == IW_MODE_ADHOC) { - i = ieee80211_channel_to_index(priv->ieee, channel); + i = libipw_channel_to_index(priv->ieee, channel); if (i == -1) return -EINVAL; - flags = (band == IEEE80211_24GHZ_BAND) ? + flags = (band == LIBIPW_24GHZ_BAND) ? geo->bg[i].flags : geo->a[i].flags; - if (flags & IEEE80211_CH_PASSIVE_ONLY) { + if (flags & LIBIPW_CH_PASSIVE_ONLY) { IPW_DEBUG_WX("Invalid Ad-Hoc channel for 802.11a\n"); return -EINVAL; } @@ -8763,7 +8763,7 @@ static int ipw_wx_get_freq(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); wrqu->freq.e = 0; @@ -8774,16 +8774,16 @@ static int ipw_wx_get_freq(struct net_device *dev, priv->status & (STATUS_ASSOCIATING | STATUS_ASSOCIATED)) { int i; - i = ieee80211_channel_to_index(priv->ieee, priv->channel); + i = libipw_channel_to_index(priv->ieee, priv->channel); BUG_ON(i == -1); wrqu->freq.e = 1; - switch (ieee80211_is_valid_channel(priv->ieee, priv->channel)) { - case IEEE80211_52GHZ_BAND: + switch (libipw_is_valid_channel(priv->ieee, priv->channel)) { + case LIBIPW_52GHZ_BAND: wrqu->freq.m = priv->ieee->geo.a[i].freq * 100000; break; - case IEEE80211_24GHZ_BAND: + case LIBIPW_24GHZ_BAND: wrqu->freq.m = priv->ieee->geo.bg[i].freq * 100000; break; @@ -8802,7 +8802,7 @@ static int ipw_wx_set_mode(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); int err = 0; IPW_DEBUG_WX("Set MODE: %d\n", wrqu->mode); @@ -8854,7 +8854,7 @@ static int ipw_wx_get_mode(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); mutex_lock(&priv->mutex); wrqu->mode = priv->ieee->iw_mode; IPW_DEBUG_WX("Get MODE -> %d\n", wrqu->mode); @@ -8883,9 +8883,9 @@ static int ipw_wx_get_range(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); struct iw_range *range = (struct iw_range *)extra; - const struct ieee80211_geo *geo = ieee80211_get_geo(priv->ieee); + const struct libipw_geo *geo = libipw_get_geo(priv->ieee); int i = 0, j; wrqu->data.length = sizeof(*range); @@ -8929,7 +8929,7 @@ static int ipw_wx_get_range(struct net_device *dev, if (priv->ieee->mode & (IEEE_B | IEEE_G)) { for (j = 0; j < geo->bg_channels && i < IW_MAX_FREQUENCIES; j++) { if ((priv->ieee->iw_mode == IW_MODE_ADHOC) && - (geo->bg[j].flags & IEEE80211_CH_PASSIVE_ONLY)) + (geo->bg[j].flags & LIBIPW_CH_PASSIVE_ONLY)) continue; range->freq[i].i = geo->bg[j].channel; @@ -8942,7 +8942,7 @@ static int ipw_wx_get_range(struct net_device *dev, if (priv->ieee->mode & IEEE_A) { for (j = 0; j < geo->a_channels && i < IW_MAX_FREQUENCIES; j++) { if ((priv->ieee->iw_mode == IW_MODE_ADHOC) && - (geo->a[j].flags & IEEE80211_CH_PASSIVE_ONLY)) + (geo->a[j].flags & LIBIPW_CH_PASSIVE_ONLY)) continue; range->freq[i].i = geo->a[j].channel; @@ -8977,7 +8977,7 @@ static int ipw_wx_set_wap(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); static const unsigned char any[] = { 0xff, 0xff, 0xff, 0xff, 0xff, 0xff @@ -9026,7 +9026,7 @@ static int ipw_wx_get_wap(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); /* If we are associated, trying to associate, or have a statically * configured BSSID then return that; otherwise return ANY */ @@ -9048,7 +9048,7 @@ static int ipw_wx_set_essid(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); int length; DECLARE_SSID_BUF(ssid); @@ -9094,7 +9094,7 @@ static int ipw_wx_get_essid(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); DECLARE_SSID_BUF(ssid); /* If we are associated, trying to associate, or have a statically @@ -9120,7 +9120,7 @@ static int ipw_wx_set_nick(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); IPW_DEBUG_WX("Setting nick to '%s'\n", extra); if (wrqu->data.length > IW_ESSID_MAX_SIZE) @@ -9139,7 +9139,7 @@ static int ipw_wx_get_nick(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); IPW_DEBUG_WX("Getting nick\n"); mutex_lock(&priv->mutex); wrqu->data.length = strlen(priv->nick); @@ -9153,7 +9153,7 @@ static int ipw_wx_set_sens(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); int err = 0; IPW_DEBUG_WX("Setting roaming threshold to %d\n", wrqu->sens.value); @@ -9183,7 +9183,7 @@ static int ipw_wx_get_sens(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); mutex_lock(&priv->mutex); wrqu->sens.fixed = 1; wrqu->sens.value = priv->roaming_threshold; @@ -9200,7 +9200,7 @@ static int ipw_wx_set_rate(struct net_device *dev, union iwreq_data *wrqu, char *extra) { /* TODO: We should use semaphores or locks for access to priv */ - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); u32 target_rate = wrqu->bitrate.value; u32 fixed, mask; @@ -9210,7 +9210,7 @@ static int ipw_wx_set_rate(struct net_device *dev, if (target_rate == -1) { fixed = 0; - mask = IEEE80211_DEFAULT_RATES_MASK; + mask = LIBIPW_DEFAULT_RATES_MASK; /* Now we should reassociate */ goto apply; } @@ -9219,62 +9219,62 @@ static int ipw_wx_set_rate(struct net_device *dev, fixed = wrqu->bitrate.fixed; if (target_rate == 1000000 || !fixed) - mask |= IEEE80211_CCK_RATE_1MB_MASK; + mask |= LIBIPW_CCK_RATE_1MB_MASK; if (target_rate == 1000000) goto apply; if (target_rate == 2000000 || !fixed) - mask |= IEEE80211_CCK_RATE_2MB_MASK; + mask |= LIBIPW_CCK_RATE_2MB_MASK; if (target_rate == 2000000) goto apply; if (target_rate == 5500000 || !fixed) - mask |= IEEE80211_CCK_RATE_5MB_MASK; + mask |= LIBIPW_CCK_RATE_5MB_MASK; if (target_rate == 5500000) goto apply; if (target_rate == 6000000 || !fixed) - mask |= IEEE80211_OFDM_RATE_6MB_MASK; + mask |= LIBIPW_OFDM_RATE_6MB_MASK; if (target_rate == 6000000) goto apply; if (target_rate == 9000000 || !fixed) - mask |= IEEE80211_OFDM_RATE_9MB_MASK; + mask |= LIBIPW_OFDM_RATE_9MB_MASK; if (target_rate == 9000000) goto apply; if (target_rate == 11000000 || !fixed) - mask |= IEEE80211_CCK_RATE_11MB_MASK; + mask |= LIBIPW_CCK_RATE_11MB_MASK; if (target_rate == 11000000) goto apply; if (target_rate == 12000000 || !fixed) - mask |= IEEE80211_OFDM_RATE_12MB_MASK; + mask |= LIBIPW_OFDM_RATE_12MB_MASK; if (target_rate == 12000000) goto apply; if (target_rate == 18000000 || !fixed) - mask |= IEEE80211_OFDM_RATE_18MB_MASK; + mask |= LIBIPW_OFDM_RATE_18MB_MASK; if (target_rate == 18000000) goto apply; if (target_rate == 24000000 || !fixed) - mask |= IEEE80211_OFDM_RATE_24MB_MASK; + mask |= LIBIPW_OFDM_RATE_24MB_MASK; if (target_rate == 24000000) goto apply; if (target_rate == 36000000 || !fixed) - mask |= IEEE80211_OFDM_RATE_36MB_MASK; + mask |= LIBIPW_OFDM_RATE_36MB_MASK; if (target_rate == 36000000) goto apply; if (target_rate == 48000000 || !fixed) - mask |= IEEE80211_OFDM_RATE_48MB_MASK; + mask |= LIBIPW_OFDM_RATE_48MB_MASK; if (target_rate == 48000000) goto apply; if (target_rate == 54000000 || !fixed) - mask |= IEEE80211_OFDM_RATE_54MB_MASK; + mask |= LIBIPW_OFDM_RATE_54MB_MASK; if (target_rate == 54000000) goto apply; @@ -9285,7 +9285,7 @@ static int ipw_wx_set_rate(struct net_device *dev, IPW_DEBUG_WX("Setting rate mask to 0x%08X [%s]\n", mask, fixed ? "fixed" : "sub-rates"); mutex_lock(&priv->mutex); - if (mask == IEEE80211_DEFAULT_RATES_MASK) { + if (mask == LIBIPW_DEFAULT_RATES_MASK) { priv->config &= ~CFG_FIXED_RATE; ipw_set_fixed_rate(priv, priv->ieee->mode); } else @@ -9312,7 +9312,7 @@ static int ipw_wx_get_rate(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); mutex_lock(&priv->mutex); wrqu->bitrate.value = priv->last_rate; wrqu->bitrate.fixed = (priv->config & CFG_FIXED_RATE) ? 1 : 0; @@ -9325,7 +9325,7 @@ static int ipw_wx_set_rts(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); mutex_lock(&priv->mutex); if (wrqu->rts.disabled || !wrqu->rts.fixed) priv->rts_threshold = DEFAULT_RTS_THRESHOLD; @@ -9348,7 +9348,7 @@ static int ipw_wx_get_rts(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); mutex_lock(&priv->mutex); wrqu->rts.value = priv->rts_threshold; wrqu->rts.fixed = 0; /* no auto select */ @@ -9362,7 +9362,7 @@ static int ipw_wx_set_txpow(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); int err = 0; mutex_lock(&priv->mutex); @@ -9396,7 +9396,7 @@ static int ipw_wx_get_txpow(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); mutex_lock(&priv->mutex); wrqu->power.value = priv->tx_power; wrqu->power.fixed = 1; @@ -9414,7 +9414,7 @@ static int ipw_wx_set_frag(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); mutex_lock(&priv->mutex); if (wrqu->frag.disabled || !wrqu->frag.fixed) priv->ieee->fts = DEFAULT_FTS; @@ -9438,7 +9438,7 @@ static int ipw_wx_get_frag(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); mutex_lock(&priv->mutex); wrqu->frag.value = priv->ieee->fts; wrqu->frag.fixed = 0; /* no auto select */ @@ -9453,7 +9453,7 @@ static int ipw_wx_set_retry(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); if (wrqu->retry.flags & IW_RETRY_LIFETIME || wrqu->retry.disabled) return -EINVAL; @@ -9486,7 +9486,7 @@ static int ipw_wx_get_retry(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); mutex_lock(&priv->mutex); wrqu->retry.disabled = 0; @@ -9517,7 +9517,7 @@ static int ipw_wx_set_scan(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); struct iw_scan_req *req = (struct iw_scan_req *)extra; struct delayed_work *work = NULL; @@ -9553,20 +9553,20 @@ static int ipw_wx_get_scan(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); - return ieee80211_wx_get_scan(priv->ieee, info, wrqu, extra); + struct ipw_priv *priv = libipw_priv(dev); + return libipw_wx_get_scan(priv->ieee, info, wrqu, extra); } static int ipw_wx_set_encode(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *key) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); int ret; u32 cap = priv->capability; mutex_lock(&priv->mutex); - ret = ieee80211_wx_set_encode(priv->ieee, info, wrqu, key); + ret = libipw_wx_set_encode(priv->ieee, info, wrqu, key); /* In IBSS mode, we need to notify the firmware to update * the beacon info after we changed the capability. */ @@ -9583,15 +9583,15 @@ static int ipw_wx_get_encode(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *key) { - struct ipw_priv *priv = ieee80211_priv(dev); - return ieee80211_wx_get_encode(priv->ieee, info, wrqu, key); + struct ipw_priv *priv = libipw_priv(dev); + return libipw_wx_get_encode(priv->ieee, info, wrqu, key); } static int ipw_wx_set_power(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); int err; mutex_lock(&priv->mutex); if (wrqu->power.disabled) { @@ -9642,7 +9642,7 @@ static int ipw_wx_get_power(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); mutex_lock(&priv->mutex); if (!(priv->power_mode & IPW_POWER_ENABLED)) wrqu->power.disabled = 1; @@ -9659,7 +9659,7 @@ static int ipw_wx_set_powermode(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); int mode = *(int *)extra; int err; @@ -9685,7 +9685,7 @@ static int ipw_wx_get_powermode(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); int level = IPW_POWER_LEVEL(priv->power_mode); char *p = extra; @@ -9717,7 +9717,7 @@ static int ipw_wx_set_wireless_mode(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); int mode = *(int *)extra; u8 band = 0, modulation = 0; @@ -9729,8 +9729,8 @@ static int ipw_wx_set_wireless_mode(struct net_device *dev, if (priv->adapter == IPW_2915ABG) { priv->ieee->abg_true = 1; if (mode & IEEE_A) { - band |= IEEE80211_52GHZ_BAND; - modulation |= IEEE80211_OFDM_MODULATION; + band |= LIBIPW_52GHZ_BAND; + modulation |= LIBIPW_OFDM_MODULATION; } else priv->ieee->abg_true = 0; } else { @@ -9745,14 +9745,14 @@ static int ipw_wx_set_wireless_mode(struct net_device *dev, } if (mode & IEEE_B) { - band |= IEEE80211_24GHZ_BAND; - modulation |= IEEE80211_CCK_MODULATION; + band |= LIBIPW_24GHZ_BAND; + modulation |= LIBIPW_CCK_MODULATION; } else priv->ieee->abg_true = 0; if (mode & IEEE_G) { - band |= IEEE80211_24GHZ_BAND; - modulation |= IEEE80211_OFDM_MODULATION; + band |= LIBIPW_24GHZ_BAND; + modulation |= LIBIPW_OFDM_MODULATION; } else priv->ieee->abg_true = 0; @@ -9782,7 +9782,7 @@ static int ipw_wx_get_wireless_mode(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); mutex_lock(&priv->mutex); switch (priv->ieee->mode) { case IEEE_A: @@ -9823,7 +9823,7 @@ static int ipw_wx_set_preamble(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); int mode = *(int *)extra; mutex_lock(&priv->mutex); /* Switching from SHORT -> LONG requires a disassociation */ @@ -9856,7 +9856,7 @@ static int ipw_wx_get_preamble(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); mutex_lock(&priv->mutex); if (priv->config & CFG_PREAMBLE_LONG) snprintf(wrqu->name, IFNAMSIZ, "long (1)"); @@ -9871,7 +9871,7 @@ static int ipw_wx_set_monitor(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); int *parms = (int *)extra; int enable = (parms[0] > 0); mutex_lock(&priv->mutex); @@ -9905,7 +9905,7 @@ static int ipw_wx_reset(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); IPW_DEBUG_WX("RESET\n"); queue_work(priv->workqueue, &priv->adapter_restart); return 0; @@ -9915,7 +9915,7 @@ static int ipw_wx_sw_reset(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); union iwreq_data wrqu_sec = { .encoding = { .flags = IW_ENCODE_DISABLED, @@ -9938,7 +9938,7 @@ static int ipw_wx_sw_reset(struct net_device *dev, ipw_radio_kill_sw(priv, priv->status & STATUS_RF_KILL_SW); mutex_unlock(&priv->mutex); - ieee80211_wx_set_encode(priv->ieee, info, &wrqu_sec, NULL); + libipw_wx_set_encode(priv->ieee, info, &wrqu_sec, NULL); mutex_lock(&priv->mutex); if (!(priv->status & STATUS_RF_KILL_MASK)) { @@ -10083,7 +10083,7 @@ static struct iw_handler_def ipw_wx_handler_def = { */ static struct iw_statistics *ipw_get_wireless_stats(struct net_device *dev) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); struct iw_statistics *wstats; wstats = &priv->wstats; @@ -10164,13 +10164,13 @@ static int ipw_net_stop(struct net_device *dev) todo: modify to send one tfd per fragment instead of using chunking. otherwise -we need to heavily modify the ieee80211_skb_to_txb. +we need to heavily modify the libipw_skb_to_txb. */ -static int ipw_tx_skb(struct ipw_priv *priv, struct ieee80211_txb *txb, +static int ipw_tx_skb(struct ipw_priv *priv, struct libipw_txb *txb, int pri) { - struct ieee80211_hdr_3addrqos *hdr = (struct ieee80211_hdr_3addrqos *) + struct libipw_hdr_3addrqos *hdr = (struct libipw_hdr_3addrqos *) txb->fragments[0]->data; int i = 0; struct tfd_frame *tfd; @@ -10187,7 +10187,7 @@ static int ipw_tx_skb(struct ipw_priv *priv, struct ieee80211_txb *txb, if (!(priv->status & STATUS_ASSOCIATED)) goto drop; - hdr_len = ieee80211_get_hdrlen(le16_to_cpu(hdr->frame_ctl)); + hdr_len = libipw_get_hdrlen(le16_to_cpu(hdr->frame_ctl)); switch (priv->ieee->iw_mode) { case IW_MODE_ADHOC: unicast = !is_multicast_ether_addr(hdr->addr1); @@ -10356,13 +10356,13 @@ static int ipw_tx_skb(struct ipw_priv *priv, struct ieee80211_txb *txb, drop: IPW_DEBUG_DROP("Silently dropping Tx packet.\n"); - ieee80211_txb_free(txb); + libipw_txb_free(txb); return NETDEV_TX_OK; } static int ipw_net_is_queue_full(struct net_device *dev, int pri) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); #ifdef CONFIG_IPW2200_QOS int tx_id = ipw_get_tx_queue_number(priv, pri); struct clx2_tx_queue *txq = &priv->txq[tx_id]; @@ -10378,9 +10378,9 @@ static int ipw_net_is_queue_full(struct net_device *dev, int pri) #ifdef CONFIG_IPW2200_PROMISCUOUS static void ipw_handle_promiscuous_tx(struct ipw_priv *priv, - struct ieee80211_txb *txb) + struct libipw_txb *txb) { - struct ieee80211_rx_stats dummystats; + struct libipw_rx_stats dummystats; struct ieee80211_hdr *hdr; u8 n; u16 filter = priv->prom_priv->filter; @@ -10393,17 +10393,17 @@ static void ipw_handle_promiscuous_tx(struct ipw_priv *priv, /* Filtering of fragment chains is done agains the first fragment */ hdr = (void *)txb->fragments[0]->data; - if (ieee80211_is_management(le16_to_cpu(hdr->frame_control))) { + if (libipw_is_management(le16_to_cpu(hdr->frame_control))) { if (filter & IPW_PROM_NO_MGMT) return; if (filter & IPW_PROM_MGMT_HEADER_ONLY) hdr_only = 1; - } else if (ieee80211_is_control(le16_to_cpu(hdr->frame_control))) { + } else if (libipw_is_control(le16_to_cpu(hdr->frame_control))) { if (filter & IPW_PROM_NO_CTL) return; if (filter & IPW_PROM_CTL_HEADER_ONLY) hdr_only = 1; - } else if (ieee80211_is_data(le16_to_cpu(hdr->frame_control))) { + } else if (libipw_is_data(le16_to_cpu(hdr->frame_control))) { if (filter & IPW_PROM_NO_DATA) return; if (filter & IPW_PROM_DATA_HEADER_ONLY) @@ -10418,7 +10418,7 @@ static void ipw_handle_promiscuous_tx(struct ipw_priv *priv, if (hdr_only) { hdr = (void *)src->data; - len = ieee80211_get_hdrlen(le16_to_cpu(hdr->frame_control)); + len = libipw_get_hdrlen(le16_to_cpu(hdr->frame_control)); } else len = src->len; @@ -10452,16 +10452,16 @@ static void ipw_handle_promiscuous_tx(struct ipw_priv *priv, skb_copy_from_linear_data(src, skb_put(dst, len), len); - if (!ieee80211_rx(priv->prom_priv->ieee, dst, &dummystats)) + if (!libipw_rx(priv->prom_priv->ieee, dst, &dummystats)) dev_kfree_skb_any(dst); } } #endif -static int ipw_net_hard_start_xmit(struct ieee80211_txb *txb, +static int ipw_net_hard_start_xmit(struct libipw_txb *txb, struct net_device *dev, int pri) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); unsigned long flags; int ret; @@ -10488,7 +10488,7 @@ static void ipw_net_set_multicast_list(struct net_device *dev) static int ipw_net_set_mac_address(struct net_device *dev, void *p) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); struct sockaddr *addr = p; if (!is_valid_ether_addr(addr->sa_data)) @@ -10506,7 +10506,7 @@ static int ipw_net_set_mac_address(struct net_device *dev, void *p) static void ipw_ethtool_get_drvinfo(struct net_device *dev, struct ethtool_drvinfo *info) { - struct ipw_priv *p = ieee80211_priv(dev); + struct ipw_priv *p = libipw_priv(dev); char vers[64]; char date[32]; u32 len; @@ -10527,7 +10527,7 @@ static void ipw_ethtool_get_drvinfo(struct net_device *dev, static u32 ipw_ethtool_get_link(struct net_device *dev) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); return (priv->status & STATUS_ASSOCIATED) != 0; } @@ -10539,7 +10539,7 @@ static int ipw_ethtool_get_eeprom_len(struct net_device *dev) static int ipw_ethtool_get_eeprom(struct net_device *dev, struct ethtool_eeprom *eeprom, u8 * bytes) { - struct ipw_priv *p = ieee80211_priv(dev); + struct ipw_priv *p = libipw_priv(dev); if (eeprom->offset + eeprom->len > IPW_EEPROM_IMAGE_SIZE) return -EINVAL; @@ -10552,7 +10552,7 @@ static int ipw_ethtool_get_eeprom(struct net_device *dev, static int ipw_ethtool_set_eeprom(struct net_device *dev, struct ethtool_eeprom *eeprom, u8 * bytes) { - struct ipw_priv *p = ieee80211_priv(dev); + struct ipw_priv *p = libipw_priv(dev); int i; if (eeprom->offset + eeprom->len > IPW_EEPROM_IMAGE_SIZE) @@ -10768,9 +10768,9 @@ static int __devinit ipw_setup_deferred_work(struct ipw_priv *priv) } static void shim__set_security(struct net_device *dev, - struct ieee80211_security *sec) + struct libipw_security *sec) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); int i; for (i = 0; i < 4; i++) { if (sec->flags & (1 << i)) { @@ -10855,21 +10855,21 @@ static int init_supported_rates(struct ipw_priv *priv, memset(rates, 0, sizeof(*rates)); /* configure supported rates */ switch (priv->ieee->freq_band) { - case IEEE80211_52GHZ_BAND: + case LIBIPW_52GHZ_BAND: rates->ieee_mode = IPW_A_MODE; rates->purpose = IPW_RATE_CAPABILITIES; - ipw_add_ofdm_scan_rates(rates, IEEE80211_CCK_MODULATION, - IEEE80211_OFDM_DEFAULT_RATES_MASK); + ipw_add_ofdm_scan_rates(rates, LIBIPW_CCK_MODULATION, + LIBIPW_OFDM_DEFAULT_RATES_MASK); break; default: /* Mixed or 2.4Ghz */ rates->ieee_mode = IPW_G_MODE; rates->purpose = IPW_RATE_CAPABILITIES; - ipw_add_cck_scan_rates(rates, IEEE80211_CCK_MODULATION, - IEEE80211_CCK_DEFAULT_RATES_MASK); - if (priv->ieee->modulation & IEEE80211_OFDM_MODULATION) { - ipw_add_ofdm_scan_rates(rates, IEEE80211_CCK_MODULATION, - IEEE80211_OFDM_DEFAULT_RATES_MASK); + ipw_add_cck_scan_rates(rates, LIBIPW_CCK_MODULATION, + LIBIPW_CCK_DEFAULT_RATES_MASK); + if (priv->ieee->modulation & LIBIPW_OFDM_MODULATION) { + ipw_add_ofdm_scan_rates(rates, LIBIPW_CCK_MODULATION, + LIBIPW_OFDM_DEFAULT_RATES_MASK); } break; } @@ -10975,7 +10975,7 @@ static int ipw_config(struct ipw_priv *priv) * table. * */ -static const struct ieee80211_geo ipw_geos[] = { +static const struct libipw_geo ipw_geos[] = { { /* Restricted */ "---", .bg_channels = 11, @@ -10997,10 +10997,10 @@ static const struct ieee80211_geo ipw_geos[] = { {5200, 40}, {5220, 44}, {5240, 48}, - {5260, 52, IEEE80211_CH_PASSIVE_ONLY}, - {5280, 56, IEEE80211_CH_PASSIVE_ONLY}, - {5300, 60, IEEE80211_CH_PASSIVE_ONLY}, - {5320, 64, IEEE80211_CH_PASSIVE_ONLY}}, + {5260, 52, LIBIPW_CH_PASSIVE_ONLY}, + {5280, 56, LIBIPW_CH_PASSIVE_ONLY}, + {5300, 60, LIBIPW_CH_PASSIVE_ONLY}, + {5320, 64, LIBIPW_CH_PASSIVE_ONLY}}, }, { /* Rest of World */ @@ -11025,10 +11025,10 @@ static const struct ieee80211_geo ipw_geos[] = { {5200, 40}, {5220, 44}, {5240, 48}, - {5260, 52, IEEE80211_CH_PASSIVE_ONLY}, - {5280, 56, IEEE80211_CH_PASSIVE_ONLY}, - {5300, 60, IEEE80211_CH_PASSIVE_ONLY}, - {5320, 64, IEEE80211_CH_PASSIVE_ONLY}, + {5260, 52, LIBIPW_CH_PASSIVE_ONLY}, + {5280, 56, LIBIPW_CH_PASSIVE_ONLY}, + {5300, 60, LIBIPW_CH_PASSIVE_ONLY}, + {5320, 64, LIBIPW_CH_PASSIVE_ONLY}, {5745, 149}, {5765, 153}, {5785, 157}, @@ -11048,15 +11048,15 @@ static const struct ieee80211_geo ipw_geos[] = { {5200, 40}, {5220, 44}, {5240, 48}, - {5260, 52, IEEE80211_CH_PASSIVE_ONLY}, - {5280, 56, IEEE80211_CH_PASSIVE_ONLY}, - {5300, 60, IEEE80211_CH_PASSIVE_ONLY}, - {5320, 64, IEEE80211_CH_PASSIVE_ONLY}, - {5745, 149, IEEE80211_CH_PASSIVE_ONLY}, - {5765, 153, IEEE80211_CH_PASSIVE_ONLY}, - {5785, 157, IEEE80211_CH_PASSIVE_ONLY}, - {5805, 161, IEEE80211_CH_PASSIVE_ONLY}, - {5825, 165, IEEE80211_CH_PASSIVE_ONLY}}, + {5260, 52, LIBIPW_CH_PASSIVE_ONLY}, + {5280, 56, LIBIPW_CH_PASSIVE_ONLY}, + {5300, 60, LIBIPW_CH_PASSIVE_ONLY}, + {5320, 64, LIBIPW_CH_PASSIVE_ONLY}, + {5745, 149, LIBIPW_CH_PASSIVE_ONLY}, + {5765, 153, LIBIPW_CH_PASSIVE_ONLY}, + {5785, 157, LIBIPW_CH_PASSIVE_ONLY}, + {5805, 161, LIBIPW_CH_PASSIVE_ONLY}, + {5825, 165, LIBIPW_CH_PASSIVE_ONLY}}, }, { /* Custom Japan */ @@ -11093,21 +11093,21 @@ static const struct ieee80211_geo ipw_geos[] = { {5200, 40}, {5220, 44}, {5240, 48}, - {5260, 52, IEEE80211_CH_PASSIVE_ONLY}, - {5280, 56, IEEE80211_CH_PASSIVE_ONLY}, - {5300, 60, IEEE80211_CH_PASSIVE_ONLY}, - {5320, 64, IEEE80211_CH_PASSIVE_ONLY}, - {5500, 100, IEEE80211_CH_PASSIVE_ONLY}, - {5520, 104, IEEE80211_CH_PASSIVE_ONLY}, - {5540, 108, IEEE80211_CH_PASSIVE_ONLY}, - {5560, 112, IEEE80211_CH_PASSIVE_ONLY}, - {5580, 116, IEEE80211_CH_PASSIVE_ONLY}, - {5600, 120, IEEE80211_CH_PASSIVE_ONLY}, - {5620, 124, IEEE80211_CH_PASSIVE_ONLY}, - {5640, 128, IEEE80211_CH_PASSIVE_ONLY}, - {5660, 132, IEEE80211_CH_PASSIVE_ONLY}, - {5680, 136, IEEE80211_CH_PASSIVE_ONLY}, - {5700, 140, IEEE80211_CH_PASSIVE_ONLY}}, + {5260, 52, LIBIPW_CH_PASSIVE_ONLY}, + {5280, 56, LIBIPW_CH_PASSIVE_ONLY}, + {5300, 60, LIBIPW_CH_PASSIVE_ONLY}, + {5320, 64, LIBIPW_CH_PASSIVE_ONLY}, + {5500, 100, LIBIPW_CH_PASSIVE_ONLY}, + {5520, 104, LIBIPW_CH_PASSIVE_ONLY}, + {5540, 108, LIBIPW_CH_PASSIVE_ONLY}, + {5560, 112, LIBIPW_CH_PASSIVE_ONLY}, + {5580, 116, LIBIPW_CH_PASSIVE_ONLY}, + {5600, 120, LIBIPW_CH_PASSIVE_ONLY}, + {5620, 124, LIBIPW_CH_PASSIVE_ONLY}, + {5640, 128, LIBIPW_CH_PASSIVE_ONLY}, + {5660, 132, LIBIPW_CH_PASSIVE_ONLY}, + {5680, 136, LIBIPW_CH_PASSIVE_ONLY}, + {5700, 140, LIBIPW_CH_PASSIVE_ONLY}}, }, { /* Custom Japan */ @@ -11117,7 +11117,7 @@ static const struct ieee80211_geo ipw_geos[] = { {2427, 4}, {2432, 5}, {2437, 6}, {2442, 7}, {2447, 8}, {2452, 9}, {2457, 10}, {2462, 11}, {2467, 12}, - {2472, 13}, {2484, 14, IEEE80211_CH_B_ONLY}}, + {2472, 13}, {2484, 14, LIBIPW_CH_B_ONLY}}, .a_channels = 4, .a = {{5170, 34}, {5190, 38}, {5210, 42}, {5230, 46}}, @@ -11130,8 +11130,8 @@ static const struct ieee80211_geo ipw_geos[] = { {2427, 4}, {2432, 5}, {2437, 6}, {2442, 7}, {2447, 8}, {2452, 9}, {2457, 10}, {2462, 11}, {2467, 12}, - {2472, 13}, {2484, 14, IEEE80211_CH_B_ONLY | - IEEE80211_CH_PASSIVE_ONLY}}, + {2472, 13}, {2484, 14, LIBIPW_CH_B_ONLY | + LIBIPW_CH_PASSIVE_ONLY}}, }, { /* High Band */ @@ -11141,8 +11141,8 @@ static const struct ieee80211_geo ipw_geos[] = { {2427, 4}, {2432, 5}, {2437, 6}, {2442, 7}, {2447, 8}, {2452, 9}, {2457, 10}, {2462, 11}, - {2467, 12, IEEE80211_CH_PASSIVE_ONLY}, - {2472, 13, IEEE80211_CH_PASSIVE_ONLY}}, + {2467, 12, LIBIPW_CH_PASSIVE_ONLY}, + {2472, 13, LIBIPW_CH_PASSIVE_ONLY}}, .a_channels = 4, .a = {{5745, 149}, {5765, 153}, {5785, 157}, {5805, 161}}, @@ -11168,33 +11168,33 @@ static const struct ieee80211_geo ipw_geos[] = { {2427, 4}, {2432, 5}, {2437, 6}, {2442, 7}, {2447, 8}, {2452, 9}, {2457, 10}, {2462, 11}, - {2467, 12, IEEE80211_CH_PASSIVE_ONLY}, - {2472, 13, IEEE80211_CH_PASSIVE_ONLY}}, + {2467, 12, LIBIPW_CH_PASSIVE_ONLY}, + {2472, 13, LIBIPW_CH_PASSIVE_ONLY}}, .a_channels = 24, - .a = {{5180, 36, IEEE80211_CH_PASSIVE_ONLY}, - {5200, 40, IEEE80211_CH_PASSIVE_ONLY}, - {5220, 44, IEEE80211_CH_PASSIVE_ONLY}, - {5240, 48, IEEE80211_CH_PASSIVE_ONLY}, - {5260, 52, IEEE80211_CH_PASSIVE_ONLY}, - {5280, 56, IEEE80211_CH_PASSIVE_ONLY}, - {5300, 60, IEEE80211_CH_PASSIVE_ONLY}, - {5320, 64, IEEE80211_CH_PASSIVE_ONLY}, - {5500, 100, IEEE80211_CH_PASSIVE_ONLY}, - {5520, 104, IEEE80211_CH_PASSIVE_ONLY}, - {5540, 108, IEEE80211_CH_PASSIVE_ONLY}, - {5560, 112, IEEE80211_CH_PASSIVE_ONLY}, - {5580, 116, IEEE80211_CH_PASSIVE_ONLY}, - {5600, 120, IEEE80211_CH_PASSIVE_ONLY}, - {5620, 124, IEEE80211_CH_PASSIVE_ONLY}, - {5640, 128, IEEE80211_CH_PASSIVE_ONLY}, - {5660, 132, IEEE80211_CH_PASSIVE_ONLY}, - {5680, 136, IEEE80211_CH_PASSIVE_ONLY}, - {5700, 140, IEEE80211_CH_PASSIVE_ONLY}, - {5745, 149, IEEE80211_CH_PASSIVE_ONLY}, - {5765, 153, IEEE80211_CH_PASSIVE_ONLY}, - {5785, 157, IEEE80211_CH_PASSIVE_ONLY}, - {5805, 161, IEEE80211_CH_PASSIVE_ONLY}, - {5825, 165, IEEE80211_CH_PASSIVE_ONLY}}, + .a = {{5180, 36, LIBIPW_CH_PASSIVE_ONLY}, + {5200, 40, LIBIPW_CH_PASSIVE_ONLY}, + {5220, 44, LIBIPW_CH_PASSIVE_ONLY}, + {5240, 48, LIBIPW_CH_PASSIVE_ONLY}, + {5260, 52, LIBIPW_CH_PASSIVE_ONLY}, + {5280, 56, LIBIPW_CH_PASSIVE_ONLY}, + {5300, 60, LIBIPW_CH_PASSIVE_ONLY}, + {5320, 64, LIBIPW_CH_PASSIVE_ONLY}, + {5500, 100, LIBIPW_CH_PASSIVE_ONLY}, + {5520, 104, LIBIPW_CH_PASSIVE_ONLY}, + {5540, 108, LIBIPW_CH_PASSIVE_ONLY}, + {5560, 112, LIBIPW_CH_PASSIVE_ONLY}, + {5580, 116, LIBIPW_CH_PASSIVE_ONLY}, + {5600, 120, LIBIPW_CH_PASSIVE_ONLY}, + {5620, 124, LIBIPW_CH_PASSIVE_ONLY}, + {5640, 128, LIBIPW_CH_PASSIVE_ONLY}, + {5660, 132, LIBIPW_CH_PASSIVE_ONLY}, + {5680, 136, LIBIPW_CH_PASSIVE_ONLY}, + {5700, 140, LIBIPW_CH_PASSIVE_ONLY}, + {5745, 149, LIBIPW_CH_PASSIVE_ONLY}, + {5765, 153, LIBIPW_CH_PASSIVE_ONLY}, + {5785, 157, LIBIPW_CH_PASSIVE_ONLY}, + {5805, 161, LIBIPW_CH_PASSIVE_ONLY}, + {5825, 165, LIBIPW_CH_PASSIVE_ONLY}}, }, { /* Europe */ @@ -11205,19 +11205,19 @@ static const struct ieee80211_geo ipw_geos[] = { {2442, 7}, {2447, 8}, {2452, 9}, {2457, 10}, {2462, 11}}, .a_channels = 13, - .a = {{5180, 36, IEEE80211_CH_PASSIVE_ONLY}, - {5200, 40, IEEE80211_CH_PASSIVE_ONLY}, - {5220, 44, IEEE80211_CH_PASSIVE_ONLY}, - {5240, 48, IEEE80211_CH_PASSIVE_ONLY}, - {5260, 52, IEEE80211_CH_PASSIVE_ONLY}, - {5280, 56, IEEE80211_CH_PASSIVE_ONLY}, - {5300, 60, IEEE80211_CH_PASSIVE_ONLY}, - {5320, 64, IEEE80211_CH_PASSIVE_ONLY}, - {5745, 149, IEEE80211_CH_PASSIVE_ONLY}, - {5765, 153, IEEE80211_CH_PASSIVE_ONLY}, - {5785, 157, IEEE80211_CH_PASSIVE_ONLY}, - {5805, 161, IEEE80211_CH_PASSIVE_ONLY}, - {5825, 165, IEEE80211_CH_PASSIVE_ONLY}}, + .a = {{5180, 36, LIBIPW_CH_PASSIVE_ONLY}, + {5200, 40, LIBIPW_CH_PASSIVE_ONLY}, + {5220, 44, LIBIPW_CH_PASSIVE_ONLY}, + {5240, 48, LIBIPW_CH_PASSIVE_ONLY}, + {5260, 52, LIBIPW_CH_PASSIVE_ONLY}, + {5280, 56, LIBIPW_CH_PASSIVE_ONLY}, + {5300, 60, LIBIPW_CH_PASSIVE_ONLY}, + {5320, 64, LIBIPW_CH_PASSIVE_ONLY}, + {5745, 149, LIBIPW_CH_PASSIVE_ONLY}, + {5765, 153, LIBIPW_CH_PASSIVE_ONLY}, + {5785, 157, LIBIPW_CH_PASSIVE_ONLY}, + {5805, 161, LIBIPW_CH_PASSIVE_ONLY}, + {5825, 165, LIBIPW_CH_PASSIVE_ONLY}}, } }; @@ -11228,7 +11228,7 @@ static int ipw_up(struct ipw_priv *priv) /* Age scan list entries found before suspend */ if (priv->suspend_time) { - ieee80211_networks_age(priv->ieee, priv->suspend_time); + libipw_networks_age(priv->ieee, priv->suspend_time); priv->suspend_time = 0; } @@ -11273,7 +11273,7 @@ static int ipw_up(struct ipw_priv *priv) priv->eeprom[EEPROM_COUNTRY_CODE + 2]); j = 0; } - if (ieee80211_set_geo(priv->ieee, &ipw_geos[j])) { + if (libipw_set_geo(priv->ieee, &ipw_geos[j])) { IPW_WARNING("Could not set geography."); return 0; } @@ -11401,7 +11401,7 @@ static void ipw_bg_down(struct work_struct *work) /* Called by register_netdev() */ static int ipw_net_init(struct net_device *dev) { - struct ipw_priv *priv = ieee80211_priv(dev); + struct ipw_priv *priv = libipw_priv(dev); mutex_lock(&priv->mutex); if (ipw_up(priv)) { @@ -11480,7 +11480,7 @@ static struct attribute_group ipw_attribute_group = { #ifdef CONFIG_IPW2200_PROMISCUOUS static int ipw_prom_open(struct net_device *dev) { - struct ipw_prom_priv *prom_priv = ieee80211_priv(dev); + struct ipw_prom_priv *prom_priv = libipw_priv(dev); struct ipw_priv *priv = prom_priv->priv; IPW_DEBUG_INFO("prom dev->open\n"); @@ -11500,7 +11500,7 @@ static int ipw_prom_open(struct net_device *dev) static int ipw_prom_stop(struct net_device *dev) { - struct ipw_prom_priv *prom_priv = ieee80211_priv(dev); + struct ipw_prom_priv *prom_priv = libipw_priv(dev); struct ipw_priv *priv = prom_priv->priv; IPW_DEBUG_INFO("prom dev->stop\n"); @@ -11528,7 +11528,7 @@ static const struct net_device_ops ipw_prom_netdev_ops = { .ndo_open = ipw_prom_open, .ndo_stop = ipw_prom_stop, .ndo_start_xmit = ipw_prom_hard_start_xmit, - .ndo_change_mtu = ieee80211_change_mtu, + .ndo_change_mtu = libipw_change_mtu, .ndo_set_mac_address = eth_mac_addr, .ndo_validate_addr = eth_validate_addr, }; @@ -11544,7 +11544,7 @@ static int ipw_prom_alloc(struct ipw_priv *priv) if (priv->prom_net_dev == NULL) return -ENOMEM; - priv->prom_priv = ieee80211_priv(priv->prom_net_dev); + priv->prom_priv = libipw_priv(priv->prom_net_dev); priv->prom_priv->ieee = netdev_priv(priv->prom_net_dev); priv->prom_priv->priv = priv; @@ -11586,8 +11586,8 @@ static const struct net_device_ops ipw_netdev_ops = { .ndo_stop = ipw_net_stop, .ndo_set_multicast_list = ipw_net_set_multicast_list, .ndo_set_mac_address = ipw_net_set_mac_address, - .ndo_start_xmit = ieee80211_xmit, - .ndo_change_mtu = ieee80211_change_mtu, + .ndo_start_xmit = libipw_xmit, + .ndo_change_mtu = libipw_change_mtu, .ndo_validate_addr = eth_validate_addr, }; @@ -11607,7 +11607,7 @@ static int __devinit ipw_pci_probe(struct pci_dev *pdev, goto out; } - priv = ieee80211_priv(net_dev); + priv = libipw_priv(net_dev); priv->ieee = netdev_priv(net_dev); priv->net_dev = net_dev; diff --git a/drivers/net/wireless/ipw2x00/ipw2200.h b/drivers/net/wireless/ipw2x00/ipw2200.h index 05e8ccf01c5f..4448bad51b8a 100644 --- a/drivers/net/wireless/ipw2x00/ipw2200.h +++ b/drivers/net/wireless/ipw2x00/ipw2200.h @@ -55,7 +55,7 @@ #include -#include "ieee80211.h" +#include "libipw.h" /* Authentication and Association States */ enum connection_manager_assoc_states { @@ -365,8 +365,8 @@ enum connection_manager_assoc_states { /* QoS sturctures */ struct ipw_qos_info { int qos_enable; - struct ieee80211_qos_parameters *def_qos_parm_OFDM; - struct ieee80211_qos_parameters *def_qos_parm_CCK; + struct libipw_qos_parameters *def_qos_parm_OFDM; + struct libipw_qos_parameters *def_qos_parm_CCK; u32 burst_duration_CCK; u32 burst_duration_OFDM; u16 qos_no_ack_mask; @@ -534,7 +534,7 @@ typedef void destructor_func(const void *); struct clx2_tx_queue { struct clx2_queue q; struct tfd_frame *bd; - struct ieee80211_txb **txb; + struct libipw_txb **txb; }; /* @@ -1144,7 +1144,7 @@ enum ipw_prom_filter { struct ipw_priv; struct ipw_prom_priv { struct ipw_priv *priv; - struct ieee80211_device *ieee; + struct libipw_device *ieee; enum ipw_prom_filter filter; int tx_packets; int rx_packets; @@ -1175,7 +1175,7 @@ struct ipw_rt_hdr { struct ipw_priv { /* ieee device used by generic ieee processing code */ - struct ieee80211_device *ieee; + struct libipw_device *ieee; spinlock_t lock; spinlock_t irq_lock; @@ -1222,7 +1222,7 @@ struct ipw_priv { u32 roaming_threshold; struct ipw_associate assoc_request; - struct ieee80211_network *assoc_network; + struct libipw_network *assoc_network; unsigned long ts_scan_abort; struct ipw_supported_rates rates; diff --git a/drivers/net/wireless/ipw2x00/libipw.h b/drivers/net/wireless/ipw2x00/libipw.h new file mode 100644 index 000000000000..cefb94262cc2 --- /dev/null +++ b/drivers/net/wireless/ipw2x00/libipw.h @@ -0,0 +1,1087 @@ +/* + * Merged with mainline ieee80211.h in Aug 2004. Original ieee802_11 + * remains copyright by the original authors + * + * Portions of the merged code are based on Host AP (software wireless + * LAN access point) driver for Intersil Prism2/2.5/3. + * + * Copyright (c) 2001-2002, SSH Communications Security Corp and Jouni Malinen + * + * Copyright (c) 2002-2003, Jouni Malinen + * + * Adaption to a generic IEEE 802.11 stack by James Ketrenos + * + * Copyright (c) 2004-2005, Intel Corporation + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. See README and COPYING for + * more details. + * + * API Version History + * 1.0.x -- Initial version + * 1.1.x -- Added radiotap, QoS, TIM, libipw_geo APIs, + * various structure changes, and crypto API init method + */ +#ifndef LIBIPW_H +#define LIBIPW_H +#include /* ETH_ALEN */ +#include /* ARRAY_SIZE */ +#include +#include + +#include + +#define LIBIPW_VERSION "git-1.1.13" + +#define LIBIPW_DATA_LEN 2304 +/* Maximum size for the MA-UNITDATA primitive, 802.11 standard section + 6.2.1.1.2. + + The figure in section 7.1.2 suggests a body size of up to 2312 + bytes is allowed, which is a bit confusing, I suspect this + represents the 2304 bytes of real data, plus a possible 8 bytes of + WEP IV and ICV. (this interpretation suggested by Ramiro Barreiro) */ + +#define LIBIPW_1ADDR_LEN 10 +#define LIBIPW_2ADDR_LEN 16 +#define LIBIPW_3ADDR_LEN 24 +#define LIBIPW_4ADDR_LEN 30 +#define LIBIPW_FCS_LEN 4 +#define LIBIPW_HLEN (LIBIPW_4ADDR_LEN) +#define LIBIPW_FRAME_LEN (LIBIPW_DATA_LEN + LIBIPW_HLEN) + +#define MIN_FRAG_THRESHOLD 256U +#define MAX_FRAG_THRESHOLD 2346U + +/* QOS control */ +#define LIBIPW_QCTL_TID 0x000F + +/* debug macros */ + +#ifdef CONFIG_LIBIPW_DEBUG +extern u32 libipw_debug_level; +#define LIBIPW_DEBUG(level, fmt, args...) \ +do { if (libipw_debug_level & (level)) \ + printk(KERN_DEBUG "ieee80211: %c %s " fmt, \ + in_interrupt() ? 'I' : 'U', __func__ , ## args); } while (0) +static inline bool libipw_ratelimit_debug(u32 level) +{ + return (libipw_debug_level & level) && net_ratelimit(); +} +#else +#define LIBIPW_DEBUG(level, fmt, args...) do {} while (0) +static inline bool libipw_ratelimit_debug(u32 level) +{ + return false; +} +#endif /* CONFIG_LIBIPW_DEBUG */ + +/* + * To use the debug system: + * + * If you are defining a new debug classification, simply add it to the #define + * list here in the form of: + * + * #define LIBIPW_DL_xxxx VALUE + * + * shifting value to the left one bit from the previous entry. xxxx should be + * the name of the classification (for example, WEP) + * + * You then need to either add a LIBIPW_xxxx_DEBUG() macro definition for your + * classification, or use LIBIPW_DEBUG(LIBIPW_DL_xxxx, ...) whenever you want + * to send output to that classification. + * + * To add your debug level to the list of levels seen when you perform + * + * % cat /proc/net/ieee80211/debug_level + * + * you simply need to add your entry to the libipw_debug_level array. + * + * If you do not see debug_level in /proc/net/ieee80211 then you do not have + * CONFIG_LIBIPW_DEBUG defined in your kernel configuration + * + */ + +#define LIBIPW_DL_INFO (1<<0) +#define LIBIPW_DL_WX (1<<1) +#define LIBIPW_DL_SCAN (1<<2) +#define LIBIPW_DL_STATE (1<<3) +#define LIBIPW_DL_MGMT (1<<4) +#define LIBIPW_DL_FRAG (1<<5) +#define LIBIPW_DL_DROP (1<<7) + +#define LIBIPW_DL_TX (1<<8) +#define LIBIPW_DL_RX (1<<9) +#define LIBIPW_DL_QOS (1<<31) + +#define LIBIPW_ERROR(f, a...) printk(KERN_ERR "ieee80211: " f, ## a) +#define LIBIPW_WARNING(f, a...) printk(KERN_WARNING "ieee80211: " f, ## a) +#define LIBIPW_DEBUG_INFO(f, a...) LIBIPW_DEBUG(LIBIPW_DL_INFO, f, ## a) + +#define LIBIPW_DEBUG_WX(f, a...) LIBIPW_DEBUG(LIBIPW_DL_WX, f, ## a) +#define LIBIPW_DEBUG_SCAN(f, a...) LIBIPW_DEBUG(LIBIPW_DL_SCAN, f, ## a) +#define LIBIPW_DEBUG_STATE(f, a...) LIBIPW_DEBUG(LIBIPW_DL_STATE, f, ## a) +#define LIBIPW_DEBUG_MGMT(f, a...) LIBIPW_DEBUG(LIBIPW_DL_MGMT, f, ## a) +#define LIBIPW_DEBUG_FRAG(f, a...) LIBIPW_DEBUG(LIBIPW_DL_FRAG, f, ## a) +#define LIBIPW_DEBUG_DROP(f, a...) LIBIPW_DEBUG(LIBIPW_DL_DROP, f, ## a) +#define LIBIPW_DEBUG_TX(f, a...) LIBIPW_DEBUG(LIBIPW_DL_TX, f, ## a) +#define LIBIPW_DEBUG_RX(f, a...) LIBIPW_DEBUG(LIBIPW_DL_RX, f, ## a) +#define LIBIPW_DEBUG_QOS(f, a...) LIBIPW_DEBUG(LIBIPW_DL_QOS, f, ## a) +#include +#include /* ARPHRD_ETHER */ + +#ifndef WIRELESS_SPY +#define WIRELESS_SPY /* enable iwspy support */ +#endif +#include /* new driver API */ + +#define ETH_P_PREAUTH 0x88C7 /* IEEE 802.11i pre-authentication */ + +#ifndef ETH_P_80211_RAW +#define ETH_P_80211_RAW (ETH_P_ECONET + 1) +#endif + +/* IEEE 802.11 defines */ + +#define P80211_OUI_LEN 3 + +struct libipw_snap_hdr { + + u8 dsap; /* always 0xAA */ + u8 ssap; /* always 0xAA */ + u8 ctrl; /* always 0x03 */ + u8 oui[P80211_OUI_LEN]; /* organizational universal id */ + +} __attribute__ ((packed)); + +#define SNAP_SIZE sizeof(struct libipw_snap_hdr) + +#define WLAN_FC_GET_VERS(fc) ((fc) & IEEE80211_FCTL_VERS) +#define WLAN_FC_GET_TYPE(fc) ((fc) & IEEE80211_FCTL_FTYPE) +#define WLAN_FC_GET_STYPE(fc) ((fc) & IEEE80211_FCTL_STYPE) + +#define WLAN_GET_SEQ_FRAG(seq) ((seq) & IEEE80211_SCTL_FRAG) +#define WLAN_GET_SEQ_SEQ(seq) (((seq) & IEEE80211_SCTL_SEQ) >> 4) + +#define LIBIPW_STATMASK_SIGNAL (1<<0) +#define LIBIPW_STATMASK_RSSI (1<<1) +#define LIBIPW_STATMASK_NOISE (1<<2) +#define LIBIPW_STATMASK_RATE (1<<3) +#define LIBIPW_STATMASK_WEMASK 0x7 + +#define LIBIPW_CCK_MODULATION (1<<0) +#define LIBIPW_OFDM_MODULATION (1<<1) + +#define LIBIPW_24GHZ_BAND (1<<0) +#define LIBIPW_52GHZ_BAND (1<<1) + +#define LIBIPW_CCK_RATE_1MB 0x02 +#define LIBIPW_CCK_RATE_2MB 0x04 +#define LIBIPW_CCK_RATE_5MB 0x0B +#define LIBIPW_CCK_RATE_11MB 0x16 +#define LIBIPW_OFDM_RATE_6MB 0x0C +#define LIBIPW_OFDM_RATE_9MB 0x12 +#define LIBIPW_OFDM_RATE_12MB 0x18 +#define LIBIPW_OFDM_RATE_18MB 0x24 +#define LIBIPW_OFDM_RATE_24MB 0x30 +#define LIBIPW_OFDM_RATE_36MB 0x48 +#define LIBIPW_OFDM_RATE_48MB 0x60 +#define LIBIPW_OFDM_RATE_54MB 0x6C +#define LIBIPW_BASIC_RATE_MASK 0x80 + +#define LIBIPW_CCK_RATE_1MB_MASK (1<<0) +#define LIBIPW_CCK_RATE_2MB_MASK (1<<1) +#define LIBIPW_CCK_RATE_5MB_MASK (1<<2) +#define LIBIPW_CCK_RATE_11MB_MASK (1<<3) +#define LIBIPW_OFDM_RATE_6MB_MASK (1<<4) +#define LIBIPW_OFDM_RATE_9MB_MASK (1<<5) +#define LIBIPW_OFDM_RATE_12MB_MASK (1<<6) +#define LIBIPW_OFDM_RATE_18MB_MASK (1<<7) +#define LIBIPW_OFDM_RATE_24MB_MASK (1<<8) +#define LIBIPW_OFDM_RATE_36MB_MASK (1<<9) +#define LIBIPW_OFDM_RATE_48MB_MASK (1<<10) +#define LIBIPW_OFDM_RATE_54MB_MASK (1<<11) + +#define LIBIPW_CCK_RATES_MASK 0x0000000F +#define LIBIPW_CCK_BASIC_RATES_MASK (LIBIPW_CCK_RATE_1MB_MASK | \ + LIBIPW_CCK_RATE_2MB_MASK) +#define LIBIPW_CCK_DEFAULT_RATES_MASK (LIBIPW_CCK_BASIC_RATES_MASK | \ + LIBIPW_CCK_RATE_5MB_MASK | \ + LIBIPW_CCK_RATE_11MB_MASK) + +#define LIBIPW_OFDM_RATES_MASK 0x00000FF0 +#define LIBIPW_OFDM_BASIC_RATES_MASK (LIBIPW_OFDM_RATE_6MB_MASK | \ + LIBIPW_OFDM_RATE_12MB_MASK | \ + LIBIPW_OFDM_RATE_24MB_MASK) +#define LIBIPW_OFDM_DEFAULT_RATES_MASK (LIBIPW_OFDM_BASIC_RATES_MASK | \ + LIBIPW_OFDM_RATE_9MB_MASK | \ + LIBIPW_OFDM_RATE_18MB_MASK | \ + LIBIPW_OFDM_RATE_36MB_MASK | \ + LIBIPW_OFDM_RATE_48MB_MASK | \ + LIBIPW_OFDM_RATE_54MB_MASK) +#define LIBIPW_DEFAULT_RATES_MASK (LIBIPW_OFDM_DEFAULT_RATES_MASK | \ + LIBIPW_CCK_DEFAULT_RATES_MASK) + +#define LIBIPW_NUM_OFDM_RATES 8 +#define LIBIPW_NUM_CCK_RATES 4 +#define LIBIPW_OFDM_SHIFT_MASK_A 4 + +/* NOTE: This data is for statistical purposes; not all hardware provides this + * information for frames received. + * For libipw_rx_mgt, you need to set at least the 'len' parameter. + */ +struct libipw_rx_stats { + u32 mac_time; + s8 rssi; + u8 signal; + u8 noise; + u16 rate; /* in 100 kbps */ + u8 received_channel; + u8 control; + u8 mask; + u8 freq; + u16 len; + u64 tsf; + u32 beacon_time; +}; + +/* IEEE 802.11 requires that STA supports concurrent reception of at least + * three fragmented frames. This define can be increased to support more + * concurrent frames, but it should be noted that each entry can consume about + * 2 kB of RAM and increasing cache size will slow down frame reassembly. */ +#define LIBIPW_FRAG_CACHE_LEN 4 + +struct libipw_frag_entry { + unsigned long first_frag_time; + unsigned int seq; + unsigned int last_frag; + struct sk_buff *skb; + u8 src_addr[ETH_ALEN]; + u8 dst_addr[ETH_ALEN]; +}; + +struct libipw_stats { + unsigned int tx_unicast_frames; + unsigned int tx_multicast_frames; + unsigned int tx_fragments; + unsigned int tx_unicast_octets; + unsigned int tx_multicast_octets; + unsigned int tx_deferred_transmissions; + unsigned int tx_single_retry_frames; + unsigned int tx_multiple_retry_frames; + unsigned int tx_retry_limit_exceeded; + unsigned int tx_discards; + unsigned int rx_unicast_frames; + unsigned int rx_multicast_frames; + unsigned int rx_fragments; + unsigned int rx_unicast_octets; + unsigned int rx_multicast_octets; + unsigned int rx_fcs_errors; + unsigned int rx_discards_no_buffer; + unsigned int tx_discards_wrong_sa; + unsigned int rx_discards_undecryptable; + unsigned int rx_message_in_msg_fragments; + unsigned int rx_message_in_bad_msg_fragments; +}; + +struct libipw_device; + +#define SEC_KEY_1 (1<<0) +#define SEC_KEY_2 (1<<1) +#define SEC_KEY_3 (1<<2) +#define SEC_KEY_4 (1<<3) +#define SEC_ACTIVE_KEY (1<<4) +#define SEC_AUTH_MODE (1<<5) +#define SEC_UNICAST_GROUP (1<<6) +#define SEC_LEVEL (1<<7) +#define SEC_ENABLED (1<<8) +#define SEC_ENCRYPT (1<<9) + +#define SEC_LEVEL_0 0 /* None */ +#define SEC_LEVEL_1 1 /* WEP 40 and 104 bit */ +#define SEC_LEVEL_2 2 /* Level 1 + TKIP */ +#define SEC_LEVEL_2_CKIP 3 /* Level 1 + CKIP */ +#define SEC_LEVEL_3 4 /* Level 2 + CCMP */ + +#define SEC_ALG_NONE 0 +#define SEC_ALG_WEP 1 +#define SEC_ALG_TKIP 2 +#define SEC_ALG_CCMP 3 + +#define WEP_KEYS 4 +#define WEP_KEY_LEN 13 +#define SCM_KEY_LEN 32 +#define SCM_TEMPORAL_KEY_LENGTH 16 + +struct libipw_security { + u16 active_key:2, enabled:1, unicast_uses_group:1, encrypt:1; + u8 auth_mode; + u8 encode_alg[WEP_KEYS]; + u8 key_sizes[WEP_KEYS]; + u8 keys[WEP_KEYS][SCM_KEY_LEN]; + u8 level; + u16 flags; +} __attribute__ ((packed)); + +/* + + 802.11 data frame from AP + + ,-------------------------------------------------------------------. +Bytes | 2 | 2 | 6 | 6 | 6 | 2 | 0..2312 | 4 | + |------|------|---------|---------|---------|------|---------|------| +Desc. | ctrl | dura | DA/RA | TA | SA | Sequ | frame | fcs | + | | tion | (BSSID) | | | ence | data | | + `-------------------------------------------------------------------' + +Total: 28-2340 bytes + +*/ + +#define BEACON_PROBE_SSID_ID_POSITION 12 + +struct libipw_hdr_1addr { + __le16 frame_ctl; + __le16 duration_id; + u8 addr1[ETH_ALEN]; + u8 payload[0]; +} __attribute__ ((packed)); + +struct libipw_hdr_2addr { + __le16 frame_ctl; + __le16 duration_id; + u8 addr1[ETH_ALEN]; + u8 addr2[ETH_ALEN]; + u8 payload[0]; +} __attribute__ ((packed)); + +struct libipw_hdr_3addr { + __le16 frame_ctl; + __le16 duration_id; + u8 addr1[ETH_ALEN]; + u8 addr2[ETH_ALEN]; + u8 addr3[ETH_ALEN]; + __le16 seq_ctl; + u8 payload[0]; +} __attribute__ ((packed)); + +struct libipw_hdr_4addr { + __le16 frame_ctl; + __le16 duration_id; + u8 addr1[ETH_ALEN]; + u8 addr2[ETH_ALEN]; + u8 addr3[ETH_ALEN]; + __le16 seq_ctl; + u8 addr4[ETH_ALEN]; + u8 payload[0]; +} __attribute__ ((packed)); + +struct libipw_hdr_3addrqos { + __le16 frame_ctl; + __le16 duration_id; + u8 addr1[ETH_ALEN]; + u8 addr2[ETH_ALEN]; + u8 addr3[ETH_ALEN]; + __le16 seq_ctl; + u8 payload[0]; + __le16 qos_ctl; +} __attribute__ ((packed)); + +struct libipw_info_element { + u8 id; + u8 len; + u8 data[0]; +} __attribute__ ((packed)); + +/* + * These are the data types that can make up management packets + * + u16 auth_algorithm; + u16 auth_sequence; + u16 beacon_interval; + u16 capability; + u8 current_ap[ETH_ALEN]; + u16 listen_interval; + struct { + u16 association_id:14, reserved:2; + } __attribute__ ((packed)); + u32 time_stamp[2]; + u16 reason; + u16 status; +*/ + +struct libipw_auth { + struct libipw_hdr_3addr header; + __le16 algorithm; + __le16 transaction; + __le16 status; + /* challenge */ + struct libipw_info_element info_element[0]; +} __attribute__ ((packed)); + +struct libipw_channel_switch { + u8 id; + u8 len; + u8 mode; + u8 channel; + u8 count; +} __attribute__ ((packed)); + +struct libipw_action { + struct libipw_hdr_3addr header; + u8 category; + u8 action; + union { + struct libipw_action_exchange { + u8 token; + struct libipw_info_element info_element[0]; + } exchange; + struct libipw_channel_switch channel_switch; + + } format; +} __attribute__ ((packed)); + +struct libipw_disassoc { + struct libipw_hdr_3addr header; + __le16 reason; +} __attribute__ ((packed)); + +/* Alias deauth for disassoc */ +#define libipw_deauth libipw_disassoc + +struct libipw_probe_request { + struct libipw_hdr_3addr header; + /* SSID, supported rates */ + struct libipw_info_element info_element[0]; +} __attribute__ ((packed)); + +struct libipw_probe_response { + struct libipw_hdr_3addr header; + __le32 time_stamp[2]; + __le16 beacon_interval; + __le16 capability; + /* SSID, supported rates, FH params, DS params, + * CF params, IBSS params, TIM (if beacon), RSN */ + struct libipw_info_element info_element[0]; +} __attribute__ ((packed)); + +/* Alias beacon for probe_response */ +#define libipw_beacon libipw_probe_response + +struct libipw_assoc_request { + struct libipw_hdr_3addr header; + __le16 capability; + __le16 listen_interval; + /* SSID, supported rates, RSN */ + struct libipw_info_element info_element[0]; +} __attribute__ ((packed)); + +struct libipw_reassoc_request { + struct libipw_hdr_3addr header; + __le16 capability; + __le16 listen_interval; + u8 current_ap[ETH_ALEN]; + struct libipw_info_element info_element[0]; +} __attribute__ ((packed)); + +struct libipw_assoc_response { + struct libipw_hdr_3addr header; + __le16 capability; + __le16 status; + __le16 aid; + /* supported rates */ + struct libipw_info_element info_element[0]; +} __attribute__ ((packed)); + +struct libipw_txb { + u8 nr_frags; + u8 encrypted; + u8 rts_included; + u8 reserved; + u16 frag_size; + u16 payload_size; + struct sk_buff *fragments[0]; +}; + +/* SWEEP TABLE ENTRIES NUMBER */ +#define MAX_SWEEP_TAB_ENTRIES 42 +#define MAX_SWEEP_TAB_ENTRIES_PER_PACKET 7 +/* MAX_RATES_LENGTH needs to be 12. The spec says 8, and many APs + * only use 8, and then use extended rates for the remaining supported + * rates. Other APs, however, stick all of their supported rates on the + * main rates information element... */ +#define MAX_RATES_LENGTH ((u8)12) +#define MAX_RATES_EX_LENGTH ((u8)16) +#define MAX_NETWORK_COUNT 128 + +#define CRC_LENGTH 4U + +#define MAX_WPA_IE_LEN 64 + +#define NETWORK_HAS_OFDM (1<<1) +#define NETWORK_HAS_CCK (1<<2) + +/* QoS structure */ +#define NETWORK_HAS_QOS_PARAMETERS (1<<3) +#define NETWORK_HAS_QOS_INFORMATION (1<<4) +#define NETWORK_HAS_QOS_MASK (NETWORK_HAS_QOS_PARAMETERS | \ + NETWORK_HAS_QOS_INFORMATION) + +/* 802.11h */ +#define NETWORK_HAS_POWER_CONSTRAINT (1<<5) +#define NETWORK_HAS_CSA (1<<6) +#define NETWORK_HAS_QUIET (1<<7) +#define NETWORK_HAS_IBSS_DFS (1<<8) +#define NETWORK_HAS_TPC_REPORT (1<<9) + +#define NETWORK_HAS_ERP_VALUE (1<<10) + +#define QOS_QUEUE_NUM 4 +#define QOS_OUI_LEN 3 +#define QOS_OUI_TYPE 2 +#define QOS_ELEMENT_ID 221 +#define QOS_OUI_INFO_SUB_TYPE 0 +#define QOS_OUI_PARAM_SUB_TYPE 1 +#define QOS_VERSION_1 1 +#define QOS_AIFSN_MIN_VALUE 2 + +struct libipw_qos_information_element { + u8 elementID; + u8 length; + u8 qui[QOS_OUI_LEN]; + u8 qui_type; + u8 qui_subtype; + u8 version; + u8 ac_info; +} __attribute__ ((packed)); + +struct libipw_qos_ac_parameter { + u8 aci_aifsn; + u8 ecw_min_max; + __le16 tx_op_limit; +} __attribute__ ((packed)); + +struct libipw_qos_parameter_info { + struct libipw_qos_information_element info_element; + u8 reserved; + struct libipw_qos_ac_parameter ac_params_record[QOS_QUEUE_NUM]; +} __attribute__ ((packed)); + +struct libipw_qos_parameters { + __le16 cw_min[QOS_QUEUE_NUM]; + __le16 cw_max[QOS_QUEUE_NUM]; + u8 aifs[QOS_QUEUE_NUM]; + u8 flag[QOS_QUEUE_NUM]; + __le16 tx_op_limit[QOS_QUEUE_NUM]; +} __attribute__ ((packed)); + +struct libipw_qos_data { + struct libipw_qos_parameters parameters; + int active; + int supported; + u8 param_count; + u8 old_param_count; +}; + +struct libipw_tim_parameters { + u8 tim_count; + u8 tim_period; +} __attribute__ ((packed)); + +/*******************************************************/ + +enum { /* libipw_basic_report.map */ + LIBIPW_BASIC_MAP_BSS = (1 << 0), + LIBIPW_BASIC_MAP_OFDM = (1 << 1), + LIBIPW_BASIC_MAP_UNIDENTIFIED = (1 << 2), + LIBIPW_BASIC_MAP_RADAR = (1 << 3), + LIBIPW_BASIC_MAP_UNMEASURED = (1 << 4), + /* Bits 5-7 are reserved */ + +}; +struct libipw_basic_report { + u8 channel; + __le64 start_time; + __le16 duration; + u8 map; +} __attribute__ ((packed)); + +enum { /* libipw_measurement_request.mode */ + /* Bit 0 is reserved */ + LIBIPW_MEASUREMENT_ENABLE = (1 << 1), + LIBIPW_MEASUREMENT_REQUEST = (1 << 2), + LIBIPW_MEASUREMENT_REPORT = (1 << 3), + /* Bits 4-7 are reserved */ +}; + +enum { + LIBIPW_REPORT_BASIC = 0, /* required */ + LIBIPW_REPORT_CCA = 1, /* optional */ + LIBIPW_REPORT_RPI = 2, /* optional */ + /* 3-255 reserved */ +}; + +struct libipw_measurement_params { + u8 channel; + __le64 start_time; + __le16 duration; +} __attribute__ ((packed)); + +struct libipw_measurement_request { + struct libipw_info_element ie; + u8 token; + u8 mode; + u8 type; + struct libipw_measurement_params params[0]; +} __attribute__ ((packed)); + +struct libipw_measurement_report { + struct libipw_info_element ie; + u8 token; + u8 mode; + u8 type; + union { + struct libipw_basic_report basic[0]; + } u; +} __attribute__ ((packed)); + +struct libipw_tpc_report { + u8 transmit_power; + u8 link_margin; +} __attribute__ ((packed)); + +struct libipw_channel_map { + u8 channel; + u8 map; +} __attribute__ ((packed)); + +struct libipw_ibss_dfs { + struct libipw_info_element ie; + u8 owner[ETH_ALEN]; + u8 recovery_interval; + struct libipw_channel_map channel_map[0]; +}; + +struct libipw_csa { + u8 mode; + u8 channel; + u8 count; +} __attribute__ ((packed)); + +struct libipw_quiet { + u8 count; + u8 period; + u8 duration; + u8 offset; +} __attribute__ ((packed)); + +struct libipw_network { + /* These entries are used to identify a unique network */ + u8 bssid[ETH_ALEN]; + u8 channel; + /* Ensure null-terminated for any debug msgs */ + u8 ssid[IW_ESSID_MAX_SIZE + 1]; + u8 ssid_len; + + struct libipw_qos_data qos_data; + + /* These are network statistics */ + struct libipw_rx_stats stats; + u16 capability; + u8 rates[MAX_RATES_LENGTH]; + u8 rates_len; + u8 rates_ex[MAX_RATES_EX_LENGTH]; + u8 rates_ex_len; + unsigned long last_scanned; + u8 mode; + u32 flags; + u32 last_associate; + u32 time_stamp[2]; + u16 beacon_interval; + u16 listen_interval; + u16 atim_window; + u8 erp_value; + u8 wpa_ie[MAX_WPA_IE_LEN]; + size_t wpa_ie_len; + u8 rsn_ie[MAX_WPA_IE_LEN]; + size_t rsn_ie_len; + struct libipw_tim_parameters tim; + + /* 802.11h info */ + + /* Power Constraint - mandatory if spctrm mgmt required */ + u8 power_constraint; + + /* TPC Report - mandatory if spctrm mgmt required */ + struct libipw_tpc_report tpc_report; + + /* IBSS DFS - mandatory if spctrm mgmt required and IBSS + * NOTE: This is variable length and so must be allocated dynamically */ + struct libipw_ibss_dfs *ibss_dfs; + + /* Channel Switch Announcement - optional if spctrm mgmt required */ + struct libipw_csa csa; + + /* Quiet - optional if spctrm mgmt required */ + struct libipw_quiet quiet; + + struct list_head list; +}; + +enum libipw_state { + LIBIPW_UNINITIALIZED = 0, + LIBIPW_INITIALIZED, + LIBIPW_ASSOCIATING, + LIBIPW_ASSOCIATED, + LIBIPW_AUTHENTICATING, + LIBIPW_AUTHENTICATED, + LIBIPW_SHUTDOWN +}; + +#define DEFAULT_MAX_SCAN_AGE (15 * HZ) +#define DEFAULT_FTS 2346 + +#define CFG_LIBIPW_RESERVE_FCS (1<<0) +#define CFG_LIBIPW_COMPUTE_FCS (1<<1) +#define CFG_LIBIPW_RTS (1<<2) + +#define LIBIPW_24GHZ_MIN_CHANNEL 1 +#define LIBIPW_24GHZ_MAX_CHANNEL 14 +#define LIBIPW_24GHZ_CHANNELS (LIBIPW_24GHZ_MAX_CHANNEL - \ + LIBIPW_24GHZ_MIN_CHANNEL + 1) + +#define LIBIPW_52GHZ_MIN_CHANNEL 34 +#define LIBIPW_52GHZ_MAX_CHANNEL 165 +#define LIBIPW_52GHZ_CHANNELS (LIBIPW_52GHZ_MAX_CHANNEL - \ + LIBIPW_52GHZ_MIN_CHANNEL + 1) + +enum { + LIBIPW_CH_PASSIVE_ONLY = (1 << 0), + LIBIPW_CH_80211H_RULES = (1 << 1), + LIBIPW_CH_B_ONLY = (1 << 2), + LIBIPW_CH_NO_IBSS = (1 << 3), + LIBIPW_CH_UNIFORM_SPREADING = (1 << 4), + LIBIPW_CH_RADAR_DETECT = (1 << 5), + LIBIPW_CH_INVALID = (1 << 6), +}; + +struct libipw_channel { + u32 freq; /* in MHz */ + u8 channel; + u8 flags; + u8 max_power; /* in dBm */ +}; + +struct libipw_geo { + u8 name[4]; + u8 bg_channels; + u8 a_channels; + struct libipw_channel bg[LIBIPW_24GHZ_CHANNELS]; + struct libipw_channel a[LIBIPW_52GHZ_CHANNELS]; +}; + +struct libipw_device { + struct net_device *dev; + struct libipw_security sec; + + /* Bookkeeping structures */ + struct libipw_stats ieee_stats; + + struct libipw_geo geo; + + /* Probe / Beacon management */ + struct list_head network_free_list; + struct list_head network_list; + struct libipw_network *networks; + int scans; + int scan_age; + + int iw_mode; /* operating mode (IW_MODE_*) */ + struct iw_spy_data spy_data; /* iwspy support */ + + spinlock_t lock; + + int tx_headroom; /* Set to size of any additional room needed at front + * of allocated Tx SKBs */ + u32 config; + + /* WEP and other encryption related settings at the device level */ + int open_wep; /* Set to 1 to allow unencrypted frames */ + + int reset_on_keychange; /* Set to 1 if the HW needs to be reset on + * WEP key changes */ + + /* If the host performs {en,de}cryption, then set to 1 */ + int host_encrypt; + int host_encrypt_msdu; + int host_decrypt; + /* host performs multicast decryption */ + int host_mc_decrypt; + + /* host should strip IV and ICV from protected frames */ + /* meaningful only when hardware decryption is being used */ + int host_strip_iv_icv; + + int host_open_frag; + int host_build_iv; + int ieee802_1x; /* is IEEE 802.1X used */ + + /* WPA data */ + int wpa_enabled; + int drop_unencrypted; + int privacy_invoked; + size_t wpa_ie_len; + u8 *wpa_ie; + + struct lib80211_crypt_info crypt_info; + + int bcrx_sta_key; /* use individual keys to override default keys even + * with RX of broad/multicast frames */ + + /* Fragmentation structures */ + struct libipw_frag_entry frag_cache[LIBIPW_FRAG_CACHE_LEN]; + unsigned int frag_next_idx; + u16 fts; /* Fragmentation Threshold */ + u16 rts; /* RTS threshold */ + + /* Association info */ + u8 bssid[ETH_ALEN]; + + enum libipw_state state; + + int mode; /* A, B, G */ + int modulation; /* CCK, OFDM */ + int freq_band; /* 2.4Ghz, 5.2Ghz, Mixed */ + int abg_true; /* ABG flag */ + + int perfect_rssi; + int worst_rssi; + + u16 prev_seq_ctl; /* used to drop duplicate frames */ + + /* Callback functions */ + void (*set_security) (struct net_device * dev, + struct libipw_security * sec); + int (*hard_start_xmit) (struct libipw_txb * txb, + struct net_device * dev, int pri); + int (*reset_port) (struct net_device * dev); + int (*is_queue_full) (struct net_device * dev, int pri); + + int (*handle_management) (struct net_device * dev, + struct libipw_network * network, u16 type); + int (*is_qos_active) (struct net_device *dev, struct sk_buff *skb); + + /* Typical STA methods */ + int (*handle_auth) (struct net_device * dev, + struct libipw_auth * auth); + int (*handle_deauth) (struct net_device * dev, + struct libipw_deauth * auth); + int (*handle_action) (struct net_device * dev, + struct libipw_action * action, + struct libipw_rx_stats * stats); + int (*handle_disassoc) (struct net_device * dev, + struct libipw_disassoc * assoc); + int (*handle_beacon) (struct net_device * dev, + struct libipw_beacon * beacon, + struct libipw_network * network); + int (*handle_probe_response) (struct net_device * dev, + struct libipw_probe_response * resp, + struct libipw_network * network); + int (*handle_probe_request) (struct net_device * dev, + struct libipw_probe_request * req, + struct libipw_rx_stats * stats); + int (*handle_assoc_response) (struct net_device * dev, + struct libipw_assoc_response * resp, + struct libipw_network * network); + + /* Typical AP methods */ + int (*handle_assoc_request) (struct net_device * dev); + int (*handle_reassoc_request) (struct net_device * dev, + struct libipw_reassoc_request * req); + + /* This must be the last item so that it points to the data + * allocated beyond this structure by alloc_ieee80211 */ + u8 priv[0]; +}; + +#define IEEE_A (1<<0) +#define IEEE_B (1<<1) +#define IEEE_G (1<<2) +#define IEEE_MODE_MASK (IEEE_A|IEEE_B|IEEE_G) + +static inline void *libipw_priv(struct net_device *dev) +{ + return ((struct libipw_device *)netdev_priv(dev))->priv; +} + +static inline int libipw_is_valid_mode(struct libipw_device *ieee, + int mode) +{ + /* + * It is possible for both access points and our device to support + * combinations of modes, so as long as there is one valid combination + * of ap/device supported modes, then return success + * + */ + if ((mode & IEEE_A) && + (ieee->modulation & LIBIPW_OFDM_MODULATION) && + (ieee->freq_band & LIBIPW_52GHZ_BAND)) + return 1; + + if ((mode & IEEE_G) && + (ieee->modulation & LIBIPW_OFDM_MODULATION) && + (ieee->freq_band & LIBIPW_24GHZ_BAND)) + return 1; + + if ((mode & IEEE_B) && + (ieee->modulation & LIBIPW_CCK_MODULATION) && + (ieee->freq_band & LIBIPW_24GHZ_BAND)) + return 1; + + return 0; +} + +static inline int libipw_get_hdrlen(u16 fc) +{ + int hdrlen = LIBIPW_3ADDR_LEN; + u16 stype = WLAN_FC_GET_STYPE(fc); + + switch (WLAN_FC_GET_TYPE(fc)) { + case IEEE80211_FTYPE_DATA: + if ((fc & IEEE80211_FCTL_FROMDS) && (fc & IEEE80211_FCTL_TODS)) + hdrlen = LIBIPW_4ADDR_LEN; + if (stype & IEEE80211_STYPE_QOS_DATA) + hdrlen += 2; + break; + case IEEE80211_FTYPE_CTL: + switch (WLAN_FC_GET_STYPE(fc)) { + case IEEE80211_STYPE_CTS: + case IEEE80211_STYPE_ACK: + hdrlen = LIBIPW_1ADDR_LEN; + break; + default: + hdrlen = LIBIPW_2ADDR_LEN; + break; + } + break; + } + + return hdrlen; +} + +static inline u8 *libipw_get_payload(struct ieee80211_hdr *hdr) +{ + switch (libipw_get_hdrlen(le16_to_cpu(hdr->frame_control))) { + case LIBIPW_1ADDR_LEN: + return ((struct libipw_hdr_1addr *)hdr)->payload; + case LIBIPW_2ADDR_LEN: + return ((struct libipw_hdr_2addr *)hdr)->payload; + case LIBIPW_3ADDR_LEN: + return ((struct libipw_hdr_3addr *)hdr)->payload; + case LIBIPW_4ADDR_LEN: + return ((struct libipw_hdr_4addr *)hdr)->payload; + } + return NULL; +} + +static inline int libipw_is_ofdm_rate(u8 rate) +{ + switch (rate & ~LIBIPW_BASIC_RATE_MASK) { + case LIBIPW_OFDM_RATE_6MB: + case LIBIPW_OFDM_RATE_9MB: + case LIBIPW_OFDM_RATE_12MB: + case LIBIPW_OFDM_RATE_18MB: + case LIBIPW_OFDM_RATE_24MB: + case LIBIPW_OFDM_RATE_36MB: + case LIBIPW_OFDM_RATE_48MB: + case LIBIPW_OFDM_RATE_54MB: + return 1; + } + return 0; +} + +static inline int libipw_is_cck_rate(u8 rate) +{ + switch (rate & ~LIBIPW_BASIC_RATE_MASK) { + case LIBIPW_CCK_RATE_1MB: + case LIBIPW_CCK_RATE_2MB: + case LIBIPW_CCK_RATE_5MB: + case LIBIPW_CCK_RATE_11MB: + return 1; + } + return 0; +} + +/* ieee80211.c */ +extern void free_ieee80211(struct net_device *dev); +extern struct net_device *alloc_ieee80211(int sizeof_priv); +extern int libipw_change_mtu(struct net_device *dev, int new_mtu); + +extern void libipw_networks_age(struct libipw_device *ieee, + unsigned long age_secs); + +extern int libipw_set_encryption(struct libipw_device *ieee); + +/* libipw_tx.c */ +extern int libipw_xmit(struct sk_buff *skb, struct net_device *dev); +extern void libipw_txb_free(struct libipw_txb *); + +/* libipw_rx.c */ +extern void libipw_rx_any(struct libipw_device *ieee, + struct sk_buff *skb, struct libipw_rx_stats *stats); +extern int libipw_rx(struct libipw_device *ieee, struct sk_buff *skb, + struct libipw_rx_stats *rx_stats); +/* make sure to set stats->len */ +extern void libipw_rx_mgt(struct libipw_device *ieee, + struct libipw_hdr_4addr *header, + struct libipw_rx_stats *stats); +extern void libipw_network_reset(struct libipw_network *network); + +/* libipw_geo.c */ +extern const struct libipw_geo *libipw_get_geo(struct libipw_device + *ieee); +extern int libipw_set_geo(struct libipw_device *ieee, + const struct libipw_geo *geo); + +extern int libipw_is_valid_channel(struct libipw_device *ieee, + u8 channel); +extern int libipw_channel_to_index(struct libipw_device *ieee, + u8 channel); +extern u8 libipw_freq_to_channel(struct libipw_device *ieee, u32 freq); +extern u8 libipw_get_channel_flags(struct libipw_device *ieee, + u8 channel); +extern const struct libipw_channel *libipw_get_channel(struct + libipw_device + *ieee, u8 channel); +extern u32 libipw_channel_to_freq(struct libipw_device * ieee, + u8 channel); + +/* libipw_wx.c */ +extern int libipw_wx_get_scan(struct libipw_device *ieee, + struct iw_request_info *info, + union iwreq_data *wrqu, char *key); +extern int libipw_wx_set_encode(struct libipw_device *ieee, + struct iw_request_info *info, + union iwreq_data *wrqu, char *key); +extern int libipw_wx_get_encode(struct libipw_device *ieee, + struct iw_request_info *info, + union iwreq_data *wrqu, char *key); +extern int libipw_wx_set_encodeext(struct libipw_device *ieee, + struct iw_request_info *info, + union iwreq_data *wrqu, char *extra); +extern int libipw_wx_get_encodeext(struct libipw_device *ieee, + struct iw_request_info *info, + union iwreq_data *wrqu, char *extra); + +static inline void libipw_increment_scans(struct libipw_device *ieee) +{ + ieee->scans++; +} + +static inline int libipw_get_scans(struct libipw_device *ieee) +{ + return ieee->scans; +} + +#endif /* LIBIPW_H */ diff --git a/drivers/net/wireless/ipw2x00/libipw_geo.c b/drivers/net/wireless/ipw2x00/libipw_geo.c index 9dfbb8760f67..d04979ba2a5b 100644 --- a/drivers/net/wireless/ipw2x00/libipw_geo.c +++ b/drivers/net/wireless/ipw2x00/libipw_geo.c @@ -41,9 +41,9 @@ #include #include -#include "ieee80211.h" +#include "libipw.h" -int ieee80211_is_valid_channel(struct ieee80211_device *ieee, u8 channel) +int libipw_is_valid_channel(struct libipw_device *ieee, u8 channel) { int i; @@ -52,27 +52,27 @@ int ieee80211_is_valid_channel(struct ieee80211_device *ieee, u8 channel) if (ieee->geo.bg_channels == 0 && ieee->geo.a_channels == 0) return 0; - if (ieee->freq_band & IEEE80211_24GHZ_BAND) + if (ieee->freq_band & LIBIPW_24GHZ_BAND) for (i = 0; i < ieee->geo.bg_channels; i++) /* NOTE: If G mode is currently supported but * this is a B only channel, we don't see it * as valid. */ if ((ieee->geo.bg[i].channel == channel) && - !(ieee->geo.bg[i].flags & IEEE80211_CH_INVALID) && + !(ieee->geo.bg[i].flags & LIBIPW_CH_INVALID) && (!(ieee->mode & IEEE_G) || - !(ieee->geo.bg[i].flags & IEEE80211_CH_B_ONLY))) - return IEEE80211_24GHZ_BAND; + !(ieee->geo.bg[i].flags & LIBIPW_CH_B_ONLY))) + return LIBIPW_24GHZ_BAND; - if (ieee->freq_band & IEEE80211_52GHZ_BAND) + if (ieee->freq_band & LIBIPW_52GHZ_BAND) for (i = 0; i < ieee->geo.a_channels; i++) if ((ieee->geo.a[i].channel == channel) && - !(ieee->geo.a[i].flags & IEEE80211_CH_INVALID)) - return IEEE80211_52GHZ_BAND; + !(ieee->geo.a[i].flags & LIBIPW_CH_INVALID)) + return LIBIPW_52GHZ_BAND; return 0; } -int ieee80211_channel_to_index(struct ieee80211_device *ieee, u8 channel) +int libipw_channel_to_index(struct libipw_device *ieee, u8 channel) { int i; @@ -81,12 +81,12 @@ int ieee80211_channel_to_index(struct ieee80211_device *ieee, u8 channel) if (ieee->geo.bg_channels == 0 && ieee->geo.a_channels == 0) return -1; - if (ieee->freq_band & IEEE80211_24GHZ_BAND) + if (ieee->freq_band & LIBIPW_24GHZ_BAND) for (i = 0; i < ieee->geo.bg_channels; i++) if (ieee->geo.bg[i].channel == channel) return i; - if (ieee->freq_band & IEEE80211_52GHZ_BAND) + if (ieee->freq_band & LIBIPW_52GHZ_BAND) for (i = 0; i < ieee->geo.a_channels; i++) if (ieee->geo.a[i].channel == channel) return i; @@ -94,22 +94,22 @@ int ieee80211_channel_to_index(struct ieee80211_device *ieee, u8 channel) return -1; } -u32 ieee80211_channel_to_freq(struct ieee80211_device * ieee, u8 channel) +u32 libipw_channel_to_freq(struct libipw_device * ieee, u8 channel) { - const struct ieee80211_channel * ch; + const struct libipw_channel * ch; /* Driver needs to initialize the geography map before using * these helper functions */ if (ieee->geo.bg_channels == 0 && ieee->geo.a_channels == 0) return 0; - ch = ieee80211_get_channel(ieee, channel); + ch = libipw_get_channel(ieee, channel); if (!ch->channel) return 0; return ch->freq; } -u8 ieee80211_freq_to_channel(struct ieee80211_device * ieee, u32 freq) +u8 libipw_freq_to_channel(struct libipw_device * ieee, u32 freq) { int i; @@ -120,12 +120,12 @@ u8 ieee80211_freq_to_channel(struct ieee80211_device * ieee, u32 freq) freq /= 100000; - if (ieee->freq_band & IEEE80211_24GHZ_BAND) + if (ieee->freq_band & LIBIPW_24GHZ_BAND) for (i = 0; i < ieee->geo.bg_channels; i++) if (ieee->geo.bg[i].freq == freq) return ieee->geo.bg[i].channel; - if (ieee->freq_band & IEEE80211_52GHZ_BAND) + if (ieee->freq_band & LIBIPW_52GHZ_BAND) for (i = 0; i < ieee->geo.a_channels; i++) if (ieee->geo.a[i].freq == freq) return ieee->geo.a[i].channel; @@ -133,63 +133,63 @@ u8 ieee80211_freq_to_channel(struct ieee80211_device * ieee, u32 freq) return 0; } -int ieee80211_set_geo(struct ieee80211_device *ieee, - const struct ieee80211_geo *geo) +int libipw_set_geo(struct libipw_device *ieee, + const struct libipw_geo *geo) { memcpy(ieee->geo.name, geo->name, 3); ieee->geo.name[3] = '\0'; ieee->geo.bg_channels = geo->bg_channels; ieee->geo.a_channels = geo->a_channels; memcpy(ieee->geo.bg, geo->bg, geo->bg_channels * - sizeof(struct ieee80211_channel)); + sizeof(struct libipw_channel)); memcpy(ieee->geo.a, geo->a, ieee->geo.a_channels * - sizeof(struct ieee80211_channel)); + sizeof(struct libipw_channel)); return 0; } -const struct ieee80211_geo *ieee80211_get_geo(struct ieee80211_device *ieee) +const struct libipw_geo *libipw_get_geo(struct libipw_device *ieee) { return &ieee->geo; } -u8 ieee80211_get_channel_flags(struct ieee80211_device * ieee, u8 channel) +u8 libipw_get_channel_flags(struct libipw_device * ieee, u8 channel) { - int index = ieee80211_channel_to_index(ieee, channel); + int index = libipw_channel_to_index(ieee, channel); if (index == -1) - return IEEE80211_CH_INVALID; + return LIBIPW_CH_INVALID; - if (channel <= IEEE80211_24GHZ_CHANNELS) + if (channel <= LIBIPW_24GHZ_CHANNELS) return ieee->geo.bg[index].flags; return ieee->geo.a[index].flags; } -static const struct ieee80211_channel bad_channel = { +static const struct libipw_channel bad_channel = { .channel = 0, - .flags = IEEE80211_CH_INVALID, + .flags = LIBIPW_CH_INVALID, .max_power = 0, }; -const struct ieee80211_channel *ieee80211_get_channel(struct ieee80211_device +const struct libipw_channel *libipw_get_channel(struct libipw_device *ieee, u8 channel) { - int index = ieee80211_channel_to_index(ieee, channel); + int index = libipw_channel_to_index(ieee, channel); if (index == -1) return &bad_channel; - if (channel <= IEEE80211_24GHZ_CHANNELS) + if (channel <= LIBIPW_24GHZ_CHANNELS) return &ieee->geo.bg[index]; return &ieee->geo.a[index]; } -EXPORT_SYMBOL(ieee80211_get_channel); -EXPORT_SYMBOL(ieee80211_get_channel_flags); -EXPORT_SYMBOL(ieee80211_is_valid_channel); -EXPORT_SYMBOL(ieee80211_freq_to_channel); -EXPORT_SYMBOL(ieee80211_channel_to_freq); -EXPORT_SYMBOL(ieee80211_channel_to_index); -EXPORT_SYMBOL(ieee80211_set_geo); -EXPORT_SYMBOL(ieee80211_get_geo); +EXPORT_SYMBOL(libipw_get_channel); +EXPORT_SYMBOL(libipw_get_channel_flags); +EXPORT_SYMBOL(libipw_is_valid_channel); +EXPORT_SYMBOL(libipw_freq_to_channel); +EXPORT_SYMBOL(libipw_channel_to_freq); +EXPORT_SYMBOL(libipw_channel_to_index); +EXPORT_SYMBOL(libipw_set_geo); +EXPORT_SYMBOL(libipw_get_geo); diff --git a/drivers/net/wireless/ipw2x00/libipw_module.c b/drivers/net/wireless/ipw2x00/libipw_module.c index 8ce6e961c5da..00c3640b0f8a 100644 --- a/drivers/net/wireless/ipw2x00/libipw_module.c +++ b/drivers/net/wireless/ipw2x00/libipw_module.c @@ -50,11 +50,11 @@ #include #include -#include "ieee80211.h" +#include "libipw.h" #define DRV_DESCRIPTION "802.11 data/management/control stack" #define DRV_NAME "ieee80211" -#define DRV_VERSION IEEE80211_VERSION +#define DRV_VERSION LIBIPW_VERSION #define DRV_COPYRIGHT "Copyright (C) 2004-2005 Intel Corporation " MODULE_VERSION(DRV_VERSION); @@ -62,13 +62,13 @@ MODULE_DESCRIPTION(DRV_DESCRIPTION); MODULE_AUTHOR(DRV_COPYRIGHT); MODULE_LICENSE("GPL"); -static int ieee80211_networks_allocate(struct ieee80211_device *ieee) +static int libipw_networks_allocate(struct libipw_device *ieee) { if (ieee->networks) return 0; ieee->networks = - kzalloc(MAX_NETWORK_COUNT * sizeof(struct ieee80211_network), + kzalloc(MAX_NETWORK_COUNT * sizeof(struct libipw_network), GFP_KERNEL); if (!ieee->networks) { printk(KERN_WARNING "%s: Out of memory allocating beacons\n", @@ -79,7 +79,7 @@ static int ieee80211_networks_allocate(struct ieee80211_device *ieee) return 0; } -void ieee80211_network_reset(struct ieee80211_network *network) +void libipw_network_reset(struct libipw_network *network) { if (!network) return; @@ -90,7 +90,7 @@ void ieee80211_network_reset(struct ieee80211_network *network) } } -static inline void ieee80211_networks_free(struct ieee80211_device *ieee) +static inline void libipw_networks_free(struct libipw_device *ieee) { int i; @@ -105,10 +105,10 @@ static inline void ieee80211_networks_free(struct ieee80211_device *ieee) ieee->networks = NULL; } -void ieee80211_networks_age(struct ieee80211_device *ieee, +void libipw_networks_age(struct libipw_device *ieee, unsigned long age_secs) { - struct ieee80211_network *network = NULL; + struct libipw_network *network = NULL; unsigned long flags; unsigned long age_jiffies = msecs_to_jiffies(age_secs * MSEC_PER_SEC); @@ -118,9 +118,9 @@ void ieee80211_networks_age(struct ieee80211_device *ieee, } spin_unlock_irqrestore(&ieee->lock, flags); } -EXPORT_SYMBOL(ieee80211_networks_age); +EXPORT_SYMBOL(libipw_networks_age); -static void ieee80211_networks_initialize(struct ieee80211_device *ieee) +static void libipw_networks_initialize(struct libipw_device *ieee) { int i; @@ -131,38 +131,38 @@ static void ieee80211_networks_initialize(struct ieee80211_device *ieee) &ieee->network_free_list); } -int ieee80211_change_mtu(struct net_device *dev, int new_mtu) +int libipw_change_mtu(struct net_device *dev, int new_mtu) { - if ((new_mtu < 68) || (new_mtu > IEEE80211_DATA_LEN)) + if ((new_mtu < 68) || (new_mtu > LIBIPW_DATA_LEN)) return -EINVAL; dev->mtu = new_mtu; return 0; } -EXPORT_SYMBOL(ieee80211_change_mtu); +EXPORT_SYMBOL(libipw_change_mtu); struct net_device *alloc_ieee80211(int sizeof_priv) { - struct ieee80211_device *ieee; + struct libipw_device *ieee; struct net_device *dev; int err; - IEEE80211_DEBUG_INFO("Initializing...\n"); + LIBIPW_DEBUG_INFO("Initializing...\n"); - dev = alloc_etherdev(sizeof(struct ieee80211_device) + sizeof_priv); + dev = alloc_etherdev(sizeof(struct libipw_device) + sizeof_priv); if (!dev) { - IEEE80211_ERROR("Unable to allocate network device.\n"); + LIBIPW_ERROR("Unable to allocate network device.\n"); goto failed; } ieee = netdev_priv(dev); ieee->dev = dev; - err = ieee80211_networks_allocate(ieee); + err = libipw_networks_allocate(ieee); if (err) { - IEEE80211_ERROR("Unable to allocate beacon storage: %d\n", err); + LIBIPW_ERROR("Unable to allocate beacon storage: %d\n", err); goto failed_free_netdev; } - ieee80211_networks_initialize(ieee); + libipw_networks_initialize(ieee); /* Default fragmentation threshold is maximum payload size */ ieee->fts = DEFAULT_FTS; @@ -201,25 +201,25 @@ failed: void free_ieee80211(struct net_device *dev) { - struct ieee80211_device *ieee = netdev_priv(dev); + struct libipw_device *ieee = netdev_priv(dev); lib80211_crypt_info_free(&ieee->crypt_info); - ieee80211_networks_free(ieee); + libipw_networks_free(ieee); free_netdev(dev); } #ifdef CONFIG_LIBIPW_DEBUG static int debug = 0; -u32 ieee80211_debug_level = 0; -EXPORT_SYMBOL_GPL(ieee80211_debug_level); -static struct proc_dir_entry *ieee80211_proc = NULL; +u32 libipw_debug_level = 0; +EXPORT_SYMBOL_GPL(libipw_debug_level); +static struct proc_dir_entry *libipw_proc = NULL; static int show_debug_level(char *page, char **start, off_t offset, int count, int *eof, void *data) { - return snprintf(page, count, "0x%08X\n", ieee80211_debug_level); + return snprintf(page, count, "0x%08X\n", libipw_debug_level); } static int store_debug_level(struct file *file, const char __user * buffer, @@ -236,29 +236,29 @@ static int store_debug_level(struct file *file, const char __user * buffer, printk(KERN_INFO DRV_NAME ": %s is not in hex or decimal form.\n", buf); else - ieee80211_debug_level = val; + libipw_debug_level = val; return strnlen(buf, len); } #endif /* CONFIG_LIBIPW_DEBUG */ -static int __init ieee80211_init(void) +static int __init libipw_init(void) { #ifdef CONFIG_LIBIPW_DEBUG struct proc_dir_entry *e; - ieee80211_debug_level = debug; - ieee80211_proc = proc_mkdir(DRV_NAME, init_net.proc_net); - if (ieee80211_proc == NULL) { - IEEE80211_ERROR("Unable to create " DRV_NAME + libipw_debug_level = debug; + libipw_proc = proc_mkdir(DRV_NAME, init_net.proc_net); + if (libipw_proc == NULL) { + LIBIPW_ERROR("Unable to create " DRV_NAME " proc directory\n"); return -EIO; } e = create_proc_entry("debug_level", S_IFREG | S_IRUGO | S_IWUSR, - ieee80211_proc); + libipw_proc); if (!e) { remove_proc_entry(DRV_NAME, init_net.proc_net); - ieee80211_proc = NULL; + libipw_proc = NULL; return -EIO; } e->read_proc = show_debug_level; @@ -272,13 +272,13 @@ static int __init ieee80211_init(void) return 0; } -static void __exit ieee80211_exit(void) +static void __exit libipw_exit(void) { #ifdef CONFIG_LIBIPW_DEBUG - if (ieee80211_proc) { - remove_proc_entry("debug_level", ieee80211_proc); + if (libipw_proc) { + remove_proc_entry("debug_level", libipw_proc); remove_proc_entry(DRV_NAME, init_net.proc_net); - ieee80211_proc = NULL; + libipw_proc = NULL; } #endif /* CONFIG_LIBIPW_DEBUG */ } @@ -289,8 +289,8 @@ module_param(debug, int, 0444); MODULE_PARM_DESC(debug, "debug output mask"); #endif /* CONFIG_LIBIPW_DEBUG */ -module_exit(ieee80211_exit); -module_init(ieee80211_init); +module_exit(libipw_exit); +module_init(libipw_init); EXPORT_SYMBOL(alloc_ieee80211); EXPORT_SYMBOL(free_ieee80211); diff --git a/drivers/net/wireless/ipw2x00/libipw_rx.c b/drivers/net/wireless/ipw2x00/libipw_rx.c index dae4b8e4d8e9..282b1f7ff1e9 100644 --- a/drivers/net/wireless/ipw2x00/libipw_rx.c +++ b/drivers/net/wireless/ipw2x00/libipw_rx.c @@ -34,18 +34,18 @@ #include -#include "ieee80211.h" +#include "libipw.h" -static void ieee80211_monitor_rx(struct ieee80211_device *ieee, +static void libipw_monitor_rx(struct libipw_device *ieee, struct sk_buff *skb, - struct ieee80211_rx_stats *rx_stats) + struct libipw_rx_stats *rx_stats) { struct ieee80211_hdr *hdr = (struct ieee80211_hdr *)skb->data; u16 fc = le16_to_cpu(hdr->frame_control); skb->dev = ieee->dev; skb_reset_mac_header(skb); - skb_pull(skb, ieee80211_get_hdrlen(fc)); + skb_pull(skb, libipw_get_hdrlen(fc)); skb->pkt_type = PACKET_OTHERHOST; skb->protocol = htons(ETH_P_80211_RAW); memset(skb->cb, 0, sizeof(skb->cb)); @@ -53,22 +53,22 @@ static void ieee80211_monitor_rx(struct ieee80211_device *ieee, } /* Called only as a tasklet (software IRQ) */ -static struct ieee80211_frag_entry *ieee80211_frag_cache_find(struct - ieee80211_device +static struct libipw_frag_entry *libipw_frag_cache_find(struct + libipw_device *ieee, unsigned int seq, unsigned int frag, u8 * src, u8 * dst) { - struct ieee80211_frag_entry *entry; + struct libipw_frag_entry *entry; int i; - for (i = 0; i < IEEE80211_FRAG_CACHE_LEN; i++) { + for (i = 0; i < LIBIPW_FRAG_CACHE_LEN; i++) { entry = &ieee->frag_cache[i]; if (entry->skb != NULL && time_after(jiffies, entry->first_frag_time + 2 * HZ)) { - IEEE80211_DEBUG_FRAG("expiring fragment cache entry " + LIBIPW_DEBUG_FRAG("expiring fragment cache entry " "seq=%u last_frag=%u\n", entry->seq, entry->last_frag); dev_kfree_skb_any(entry->skb); @@ -86,13 +86,13 @@ static struct ieee80211_frag_entry *ieee80211_frag_cache_find(struct } /* Called only as a tasklet (software IRQ) */ -static struct sk_buff *ieee80211_frag_cache_get(struct ieee80211_device *ieee, - struct ieee80211_hdr_4addr *hdr) +static struct sk_buff *libipw_frag_cache_get(struct libipw_device *ieee, + struct libipw_hdr_4addr *hdr) { struct sk_buff *skb = NULL; u16 sc; unsigned int frag, seq; - struct ieee80211_frag_entry *entry; + struct libipw_frag_entry *entry; sc = le16_to_cpu(hdr->seq_ctl); frag = WLAN_GET_SEQ_FRAG(sc); @@ -101,7 +101,7 @@ static struct sk_buff *ieee80211_frag_cache_get(struct ieee80211_device *ieee, if (frag == 0) { /* Reserve enough space to fit maximum frame length */ skb = dev_alloc_skb(ieee->dev->mtu + - sizeof(struct ieee80211_hdr_4addr) + + sizeof(struct libipw_hdr_4addr) + 8 /* LLC */ + 2 /* alignment */ + 8 /* WEP */ + ETH_ALEN /* WDS */ ); @@ -110,7 +110,7 @@ static struct sk_buff *ieee80211_frag_cache_get(struct ieee80211_device *ieee, entry = &ieee->frag_cache[ieee->frag_next_idx]; ieee->frag_next_idx++; - if (ieee->frag_next_idx >= IEEE80211_FRAG_CACHE_LEN) + if (ieee->frag_next_idx >= LIBIPW_FRAG_CACHE_LEN) ieee->frag_next_idx = 0; if (entry->skb != NULL) @@ -125,7 +125,7 @@ static struct sk_buff *ieee80211_frag_cache_get(struct ieee80211_device *ieee, } else { /* received a fragment of a frame for which the head fragment * should have already been received */ - entry = ieee80211_frag_cache_find(ieee, seq, frag, hdr->addr2, + entry = libipw_frag_cache_find(ieee, seq, frag, hdr->addr2, hdr->addr1); if (entry != NULL) { entry->last_frag = frag; @@ -137,21 +137,21 @@ static struct sk_buff *ieee80211_frag_cache_get(struct ieee80211_device *ieee, } /* Called only as a tasklet (software IRQ) */ -static int ieee80211_frag_cache_invalidate(struct ieee80211_device *ieee, - struct ieee80211_hdr_4addr *hdr) +static int libipw_frag_cache_invalidate(struct libipw_device *ieee, + struct libipw_hdr_4addr *hdr) { u16 sc; unsigned int seq; - struct ieee80211_frag_entry *entry; + struct libipw_frag_entry *entry; sc = le16_to_cpu(hdr->seq_ctl); seq = WLAN_GET_SEQ_SEQ(sc); - entry = ieee80211_frag_cache_find(ieee, seq, -1, hdr->addr2, + entry = libipw_frag_cache_find(ieee, seq, -1, hdr->addr2, hdr->addr1); if (entry == NULL) { - IEEE80211_DEBUG_FRAG("could not invalidate fragment cache " + LIBIPW_DEBUG_FRAG("could not invalidate fragment cache " "entry (seq=%u)\n", seq); return -1; } @@ -161,14 +161,14 @@ static int ieee80211_frag_cache_invalidate(struct ieee80211_device *ieee, } #ifdef NOT_YET -/* ieee80211_rx_frame_mgtmt +/* libipw_rx_frame_mgtmt * * Responsible for handling management control frames * - * Called by ieee80211_rx */ + * Called by libipw_rx */ static int -ieee80211_rx_frame_mgmt(struct ieee80211_device *ieee, struct sk_buff *skb, - struct ieee80211_rx_stats *rx_stats, u16 type, +libipw_rx_frame_mgmt(struct libipw_device *ieee, struct sk_buff *skb, + struct libipw_rx_stats *rx_stats, u16 type, u16 stype) { if (ieee->iw_mode == IW_MODE_MASTER) { @@ -176,7 +176,7 @@ ieee80211_rx_frame_mgmt(struct ieee80211_device *ieee, struct sk_buff *skb, ieee->dev->name); return 0; /* - hostap_update_sta_ps(ieee, (struct hostap_ieee80211_hdr_4addr *) + hostap_update_sta_ps(ieee, (struct hostap_libipw_hdr_4addr *) skb->data);*/ } @@ -219,26 +219,27 @@ ieee80211_rx_frame_mgmt(struct ieee80211_device *ieee, struct sk_buff *skb, /* See IEEE 802.1H for LLC/SNAP encapsulation/decapsulation */ /* Ethernet-II snap header (RFC1042 for most EtherTypes) */ -static unsigned char rfc1042_header[] = { 0xaa, 0xaa, 0x03, 0x00, 0x00, 0x00 }; +static unsigned char libipw_rfc1042_header[] = + { 0xaa, 0xaa, 0x03, 0x00, 0x00, 0x00 }; /* Bridge-Tunnel header (for EtherTypes ETH_P_AARP and ETH_P_IPX) */ -static unsigned char bridge_tunnel_header[] = +static unsigned char libipw_bridge_tunnel_header[] = { 0xaa, 0xaa, 0x03, 0x00, 0x00, 0xf8 }; /* No encapsulation header if EtherType < 0x600 (=length) */ -/* Called by ieee80211_rx_frame_decrypt */ -static int ieee80211_is_eapol_frame(struct ieee80211_device *ieee, +/* Called by libipw_rx_frame_decrypt */ +static int libipw_is_eapol_frame(struct libipw_device *ieee, struct sk_buff *skb) { struct net_device *dev = ieee->dev; u16 fc, ethertype; - struct ieee80211_hdr_3addr *hdr; + struct libipw_hdr_3addr *hdr; u8 *pos; if (skb->len < 24) return 0; - hdr = (struct ieee80211_hdr_3addr *)skb->data; + hdr = (struct libipw_hdr_3addr *)skb->data; fc = le16_to_cpu(hdr->frame_ctl); /* check that the frame is unicast frame to us */ @@ -266,28 +267,28 @@ static int ieee80211_is_eapol_frame(struct ieee80211_device *ieee, return 0; } -/* Called only as a tasklet (software IRQ), by ieee80211_rx */ +/* Called only as a tasklet (software IRQ), by libipw_rx */ static int -ieee80211_rx_frame_decrypt(struct ieee80211_device *ieee, struct sk_buff *skb, +libipw_rx_frame_decrypt(struct libipw_device *ieee, struct sk_buff *skb, struct lib80211_crypt_data *crypt) { - struct ieee80211_hdr_3addr *hdr; + struct libipw_hdr_3addr *hdr; int res, hdrlen; if (crypt == NULL || crypt->ops->decrypt_mpdu == NULL) return 0; - hdr = (struct ieee80211_hdr_3addr *)skb->data; - hdrlen = ieee80211_get_hdrlen(le16_to_cpu(hdr->frame_ctl)); + hdr = (struct libipw_hdr_3addr *)skb->data; + hdrlen = libipw_get_hdrlen(le16_to_cpu(hdr->frame_ctl)); atomic_inc(&crypt->refcnt); res = crypt->ops->decrypt_mpdu(skb, hdrlen, crypt->priv); atomic_dec(&crypt->refcnt); if (res < 0) { - IEEE80211_DEBUG_DROP("decryption failed (SA=%pM) res=%d\n", + LIBIPW_DEBUG_DROP("decryption failed (SA=%pM) res=%d\n", hdr->addr2, res); if (res == -2) - IEEE80211_DEBUG_DROP("Decryption failed ICV " + LIBIPW_DEBUG_DROP("Decryption failed ICV " "mismatch (key %d)\n", skb->data[hdrlen + 3] >> 6); ieee->ieee_stats.rx_discards_undecryptable++; @@ -297,20 +298,20 @@ ieee80211_rx_frame_decrypt(struct ieee80211_device *ieee, struct sk_buff *skb, return res; } -/* Called only as a tasklet (software IRQ), by ieee80211_rx */ +/* Called only as a tasklet (software IRQ), by libipw_rx */ static int -ieee80211_rx_frame_decrypt_msdu(struct ieee80211_device *ieee, +libipw_rx_frame_decrypt_msdu(struct libipw_device *ieee, struct sk_buff *skb, int keyidx, struct lib80211_crypt_data *crypt) { - struct ieee80211_hdr_3addr *hdr; + struct libipw_hdr_3addr *hdr; int res, hdrlen; if (crypt == NULL || crypt->ops->decrypt_msdu == NULL) return 0; - hdr = (struct ieee80211_hdr_3addr *)skb->data; - hdrlen = ieee80211_get_hdrlen(le16_to_cpu(hdr->frame_ctl)); + hdr = (struct libipw_hdr_3addr *)skb->data; + hdrlen = libipw_get_hdrlen(le16_to_cpu(hdr->frame_ctl)); atomic_inc(&crypt->refcnt); res = crypt->ops->decrypt_msdu(skb, keyidx, hdrlen, crypt->priv); @@ -328,11 +329,11 @@ ieee80211_rx_frame_decrypt_msdu(struct ieee80211_device *ieee, /* All received frames are sent to this function. @skb contains the frame in * IEEE 802.11 format, i.e., in the format it was sent over air. * This function is called only as a tasklet (software IRQ). */ -int ieee80211_rx(struct ieee80211_device *ieee, struct sk_buff *skb, - struct ieee80211_rx_stats *rx_stats) +int libipw_rx(struct libipw_device *ieee, struct sk_buff *skb, + struct libipw_rx_stats *rx_stats) { struct net_device *dev = ieee->dev; - struct ieee80211_hdr_4addr *hdr; + struct libipw_hdr_4addr *hdr; size_t hdrlen; u16 fc, type, stype, sc; unsigned int frag; @@ -352,7 +353,7 @@ int ieee80211_rx(struct ieee80211_device *ieee, struct sk_buff *skb, int keyidx = 0; int can_be_decrypted = 0; - hdr = (struct ieee80211_hdr_4addr *)skb->data; + hdr = (struct libipw_hdr_4addr *)skb->data; if (skb->len < 10) { printk(KERN_INFO "%s: SKB length < 10\n", dev->name); goto rx_dropped; @@ -363,7 +364,7 @@ int ieee80211_rx(struct ieee80211_device *ieee, struct sk_buff *skb, stype = WLAN_FC_GET_STYPE(fc); sc = le16_to_cpu(hdr->seq_ctl); frag = WLAN_GET_SEQ_FRAG(sc); - hdrlen = ieee80211_get_hdrlen(fc); + hdrlen = libipw_get_hdrlen(fc); if (skb->len < hdrlen) { printk(KERN_INFO "%s: invalid SKB length %d\n", @@ -380,19 +381,19 @@ int ieee80211_rx(struct ieee80211_device *ieee, struct sk_buff *skb, struct iw_quality wstats; wstats.updated = 0; - if (rx_stats->mask & IEEE80211_STATMASK_RSSI) { + if (rx_stats->mask & LIBIPW_STATMASK_RSSI) { wstats.level = rx_stats->signal; wstats.updated |= IW_QUAL_LEVEL_UPDATED; } else wstats.updated |= IW_QUAL_LEVEL_INVALID; - if (rx_stats->mask & IEEE80211_STATMASK_NOISE) { + if (rx_stats->mask & LIBIPW_STATMASK_NOISE) { wstats.noise = rx_stats->noise; wstats.updated |= IW_QUAL_NOISE_UPDATED; } else wstats.updated |= IW_QUAL_NOISE_INVALID; - if (rx_stats->mask & IEEE80211_STATMASK_SIGNAL) { + if (rx_stats->mask & LIBIPW_STATMASK_SIGNAL) { wstats.qual = rx_stats->signal; wstats.updated |= IW_QUAL_QUAL_UPDATED; } else @@ -411,7 +412,7 @@ int ieee80211_rx(struct ieee80211_device *ieee, struct sk_buff *skb, if (ieee->iw_mode == IW_MODE_MONITOR) { dev->stats.rx_packets++; dev->stats.rx_bytes += skb->len; - ieee80211_monitor_rx(ieee, skb, rx_stats); + libipw_monitor_rx(ieee, skb, rx_stats); return 1; } @@ -457,7 +458,7 @@ int ieee80211_rx(struct ieee80211_device *ieee, struct sk_buff *skb, * frames from other than current BSS, so just drop the * frames silently instead of filling system log with * these reports. */ - IEEE80211_DEBUG_DROP("Decryption failed (not set)" + LIBIPW_DEBUG_DROP("Decryption failed (not set)" " (SA=%pM)\n", hdr->addr2); ieee->ieee_stats.rx_discards_undecryptable++; goto rx_dropped; @@ -475,7 +476,7 @@ int ieee80211_rx(struct ieee80211_device *ieee, struct sk_buff *skb, goto rx_dropped; } - if (ieee80211_rx_frame_mgmt(ieee, skb, rx_stats, type, stype)) + if (libipw_rx_frame_mgmt(ieee, skb, rx_stats, type, stype)) goto rx_dropped; else goto rx_exit; @@ -488,7 +489,7 @@ int ieee80211_rx(struct ieee80211_device *ieee, struct sk_buff *skb, ieee->prev_seq_ctl = sc; /* Data frame - extract src/dst addresses */ - if (skb->len < IEEE80211_3ADDR_LEN) + if (skb->len < LIBIPW_3ADDR_LEN) goto rx_dropped; switch (fc & (IEEE80211_FCTL_FROMDS | IEEE80211_FCTL_TODS)) { @@ -501,7 +502,7 @@ int ieee80211_rx(struct ieee80211_device *ieee, struct sk_buff *skb, memcpy(src, hdr->addr2, ETH_ALEN); break; case IEEE80211_FCTL_FROMDS | IEEE80211_FCTL_TODS: - if (skb->len < IEEE80211_4ADDR_LEN) + if (skb->len < LIBIPW_4ADDR_LEN) goto rx_dropped; memcpy(dst, hdr->addr3, ETH_ALEN); memcpy(src, hdr->addr4, ETH_ALEN); @@ -560,7 +561,7 @@ int ieee80211_rx(struct ieee80211_device *ieee, struct sk_buff *skb, stype != IEEE80211_STYPE_DATA_CFPOLL && stype != IEEE80211_STYPE_DATA_CFACKPOLL) { if (stype != IEEE80211_STYPE_NULLFUNC) - IEEE80211_DEBUG_DROP("RX: dropped data frame " + LIBIPW_DEBUG_DROP("RX: dropped data frame " "with no data (type=0x%02x, " "subtype=0x%02x, len=%d)\n", type, stype, skb->len); @@ -570,21 +571,21 @@ int ieee80211_rx(struct ieee80211_device *ieee, struct sk_buff *skb, /* skb: hdr + (possibly fragmented, possibly encrypted) payload */ if ((fc & IEEE80211_FCTL_PROTECTED) && can_be_decrypted && - (keyidx = ieee80211_rx_frame_decrypt(ieee, skb, crypt)) < 0) + (keyidx = libipw_rx_frame_decrypt(ieee, skb, crypt)) < 0) goto rx_dropped; - hdr = (struct ieee80211_hdr_4addr *)skb->data; + hdr = (struct libipw_hdr_4addr *)skb->data; /* skb: hdr + (possibly fragmented) plaintext payload */ // PR: FIXME: hostap has additional conditions in the "if" below: // ieee->host_decrypt && (fc & IEEE80211_FCTL_PROTECTED) && if ((frag != 0) || (fc & IEEE80211_FCTL_MOREFRAGS)) { int flen; - struct sk_buff *frag_skb = ieee80211_frag_cache_get(ieee, hdr); - IEEE80211_DEBUG_FRAG("Rx Fragment received (%u)\n", frag); + struct sk_buff *frag_skb = libipw_frag_cache_get(ieee, hdr); + LIBIPW_DEBUG_FRAG("Rx Fragment received (%u)\n", frag); if (!frag_skb) { - IEEE80211_DEBUG(IEEE80211_DL_RX | IEEE80211_DL_FRAG, + LIBIPW_DEBUG(LIBIPW_DL_RX | LIBIPW_DL_FRAG, "Rx cannot get skb from fragment " "cache (morefrag=%d seq=%u frag=%u)\n", (fc & IEEE80211_FCTL_MOREFRAGS) != 0, @@ -600,7 +601,7 @@ int ieee80211_rx(struct ieee80211_device *ieee, struct sk_buff *skb, printk(KERN_WARNING "%s: host decrypted and " "reassembled frame did not fit skb\n", dev->name); - ieee80211_frag_cache_invalidate(ieee, hdr); + libipw_frag_cache_invalidate(ieee, hdr); goto rx_dropped; } @@ -627,24 +628,24 @@ int ieee80211_rx(struct ieee80211_device *ieee, struct sk_buff *skb, /* this was the last fragment and the frame will be * delivered, so remove skb from fragment cache */ skb = frag_skb; - hdr = (struct ieee80211_hdr_4addr *)skb->data; - ieee80211_frag_cache_invalidate(ieee, hdr); + hdr = (struct libipw_hdr_4addr *)skb->data; + libipw_frag_cache_invalidate(ieee, hdr); } /* skb: hdr + (possible reassembled) full MSDU payload; possibly still * encrypted/authenticated */ if ((fc & IEEE80211_FCTL_PROTECTED) && can_be_decrypted && - ieee80211_rx_frame_decrypt_msdu(ieee, skb, keyidx, crypt)) + libipw_rx_frame_decrypt_msdu(ieee, skb, keyidx, crypt)) goto rx_dropped; - hdr = (struct ieee80211_hdr_4addr *)skb->data; + hdr = (struct libipw_hdr_4addr *)skb->data; if (crypt && !(fc & IEEE80211_FCTL_PROTECTED) && !ieee->open_wep) { if ( /*ieee->ieee802_1x && */ - ieee80211_is_eapol_frame(ieee, skb)) { + libipw_is_eapol_frame(ieee, skb)) { /* pass unencrypted EAPOL frames even if encryption is * configured */ } else { - IEEE80211_DEBUG_DROP("encryption configured, but RX " + LIBIPW_DEBUG_DROP("encryption configured, but RX " "frame not encrypted (SA=%pM)\n", hdr->addr2); goto rx_dropped; @@ -652,8 +653,8 @@ int ieee80211_rx(struct ieee80211_device *ieee, struct sk_buff *skb, } if (crypt && !(fc & IEEE80211_FCTL_PROTECTED) && !ieee->open_wep && - !ieee80211_is_eapol_frame(ieee, skb)) { - IEEE80211_DEBUG_DROP("dropped unencrypted RX data " + !libipw_is_eapol_frame(ieee, skb)) { + LIBIPW_DEBUG_DROP("dropped unencrypted RX data " "frame from %pM (drop_unencrypted=1)\n", hdr->addr2); goto rx_dropped; @@ -736,9 +737,9 @@ int ieee80211_rx(struct ieee80211_device *ieee, struct sk_buff *skb, /* convert hdr + possible LLC headers into Ethernet header */ if (skb->len - hdrlen >= 8 && - ((memcmp(payload, rfc1042_header, SNAP_SIZE) == 0 && + ((memcmp(payload, libipw_rfc1042_header, SNAP_SIZE) == 0 && ethertype != ETH_P_AARP && ethertype != ETH_P_IPX) || - memcmp(payload, bridge_tunnel_header, SNAP_SIZE) == 0)) { + memcmp(payload, libipw_bridge_tunnel_header, SNAP_SIZE) == 0)) { /* remove RFC1042 or Bridge-Tunnel encapsulation and * replace EtherType */ skb_pull(skb, hdrlen + SNAP_SIZE); @@ -807,7 +808,7 @@ int ieee80211_rx(struct ieee80211_device *ieee, struct sk_buff *skb, /* netif_rx always succeeds, but it might drop * the packet. If it drops the packet, we log that * in our stats. */ - IEEE80211_DEBUG_DROP + LIBIPW_DEBUG_DROP ("RX: netif_rx dropped the packet\n"); dev->stats.rx_dropped++; } @@ -829,18 +830,18 @@ int ieee80211_rx(struct ieee80211_device *ieee, struct sk_buff *skb, return 0; } -/* Filter out unrelated packets, call ieee80211_rx[_mgt] +/* Filter out unrelated packets, call libipw_rx[_mgt] * This function takes over the skb, it should not be used again after calling * this function. */ -void ieee80211_rx_any(struct ieee80211_device *ieee, - struct sk_buff *skb, struct ieee80211_rx_stats *stats) +void libipw_rx_any(struct libipw_device *ieee, + struct sk_buff *skb, struct libipw_rx_stats *stats) { - struct ieee80211_hdr_4addr *hdr; + struct libipw_hdr_4addr *hdr; int is_packet_for_us; u16 fc; if (ieee->iw_mode == IW_MODE_MONITOR) { - if (!ieee80211_rx(ieee, skb, stats)) + if (!libipw_rx(ieee, skb, stats)) dev_kfree_skb_irq(skb); return; } @@ -848,7 +849,7 @@ void ieee80211_rx_any(struct ieee80211_device *ieee, if (skb->len < sizeof(struct ieee80211_hdr)) goto drop_free; - hdr = (struct ieee80211_hdr_4addr *)skb->data; + hdr = (struct libipw_hdr_4addr *)skb->data; fc = le16_to_cpu(hdr->frame_ctl); if ((fc & IEEE80211_FCTL_VERS) != 0) @@ -856,9 +857,9 @@ void ieee80211_rx_any(struct ieee80211_device *ieee, switch (fc & IEEE80211_FCTL_FTYPE) { case IEEE80211_FTYPE_MGMT: - if (skb->len < sizeof(struct ieee80211_hdr_3addr)) + if (skb->len < sizeof(struct libipw_hdr_3addr)) goto drop_free; - ieee80211_rx_mgt(ieee, hdr, stats); + libipw_rx_mgt(ieee, hdr, stats); dev_kfree_skb_irq(skb); return; case IEEE80211_FTYPE_DATA: @@ -910,7 +911,7 @@ void ieee80211_rx_any(struct ieee80211_device *ieee, } if (is_packet_for_us) - if (!ieee80211_rx(ieee, skb, stats)) + if (!libipw_rx(ieee, skb, stats)) dev_kfree_skb_irq(skb); return; @@ -928,7 +929,7 @@ static u8 qos_oui[QOS_OUI_LEN] = { 0x00, 0x50, 0xF2 }; * Make ther structure we read from the beacon packet has * the right values */ -static int ieee80211_verify_qos_info(struct ieee80211_qos_information_element +static int libipw_verify_qos_info(struct libipw_qos_information_element *info_element, int sub_type) { @@ -947,12 +948,12 @@ static int ieee80211_verify_qos_info(struct ieee80211_qos_information_element /* * Parse a QoS parameter element */ -static int ieee80211_read_qos_param_element(struct ieee80211_qos_parameter_info - *element_param, struct ieee80211_info_element +static int libipw_read_qos_param_element(struct libipw_qos_parameter_info + *element_param, struct libipw_info_element *info_element) { int ret = 0; - u16 size = sizeof(struct ieee80211_qos_parameter_info) - 2; + u16 size = sizeof(struct libipw_qos_parameter_info) - 2; if ((info_element == NULL) || (element_param == NULL)) return -1; @@ -965,7 +966,7 @@ static int ieee80211_read_qos_param_element(struct ieee80211_qos_parameter_info } else ret = -1; if (ret == 0) - ret = ieee80211_verify_qos_info(&element_param->info_element, + ret = libipw_verify_qos_info(&element_param->info_element, QOS_OUI_PARAM_SUB_TYPE); return ret; } @@ -973,13 +974,13 @@ static int ieee80211_read_qos_param_element(struct ieee80211_qos_parameter_info /* * Parse a QoS information element */ -static int ieee80211_read_qos_info_element(struct - ieee80211_qos_information_element - *element_info, struct ieee80211_info_element +static int libipw_read_qos_info_element(struct + libipw_qos_information_element + *element_info, struct libipw_info_element *info_element) { int ret = 0; - u16 size = sizeof(struct ieee80211_qos_information_element) - 2; + u16 size = sizeof(struct libipw_qos_information_element) - 2; if (element_info == NULL) return -1; @@ -995,7 +996,7 @@ static int ieee80211_read_qos_info_element(struct ret = -1; if (ret == 0) - ret = ieee80211_verify_qos_info(element_info, + ret = libipw_verify_qos_info(element_info, QOS_OUI_INFO_SUB_TYPE); return ret; } @@ -1003,15 +1004,15 @@ static int ieee80211_read_qos_info_element(struct /* * Write QoS parameters from the ac parameters. */ -static int ieee80211_qos_convert_ac_to_parameters(struct - ieee80211_qos_parameter_info +static int libipw_qos_convert_ac_to_parameters(struct + libipw_qos_parameter_info *param_elm, struct - ieee80211_qos_parameters + libipw_qos_parameters *qos_param) { int rc = 0; int i; - struct ieee80211_qos_ac_parameter *ac_params; + struct libipw_qos_ac_parameter *ac_params; u32 txop; u8 cw_min; u8 cw_max; @@ -1042,27 +1043,27 @@ static int ieee80211_qos_convert_ac_to_parameters(struct * parameters element. check the information element length to decide * which type to read */ -static int ieee80211_parse_qos_info_param_IE(struct ieee80211_info_element +static int libipw_parse_qos_info_param_IE(struct libipw_info_element *info_element, - struct ieee80211_network *network) + struct libipw_network *network) { int rc = 0; - struct ieee80211_qos_parameters *qos_param = NULL; - struct ieee80211_qos_information_element qos_info_element; + struct libipw_qos_parameters *qos_param = NULL; + struct libipw_qos_information_element qos_info_element; - rc = ieee80211_read_qos_info_element(&qos_info_element, info_element); + rc = libipw_read_qos_info_element(&qos_info_element, info_element); if (rc == 0) { network->qos_data.param_count = qos_info_element.ac_info & 0x0F; network->flags |= NETWORK_HAS_QOS_INFORMATION; } else { - struct ieee80211_qos_parameter_info param_element; + struct libipw_qos_parameter_info param_element; - rc = ieee80211_read_qos_param_element(¶m_element, + rc = libipw_read_qos_param_element(¶m_element, info_element); if (rc == 0) { qos_param = &(network->qos_data.parameters); - ieee80211_qos_convert_ac_to_parameters(¶m_element, + libipw_qos_convert_ac_to_parameters(¶m_element, qos_param); network->flags |= NETWORK_HAS_QOS_PARAMETERS; network->qos_data.param_count = @@ -1071,7 +1072,7 @@ static int ieee80211_parse_qos_info_param_IE(struct ieee80211_info_element } if (rc == 0) { - IEEE80211_DEBUG_QOS("QoS is supported\n"); + LIBIPW_DEBUG_QOS("QoS is supported\n"); network->qos_data.supported = 1; } return rc; @@ -1116,9 +1117,9 @@ static const char *get_info_element_string(u16 id) } #endif -static int ieee80211_parse_info_param(struct ieee80211_info_element +static int libipw_parse_info_param(struct libipw_info_element *info_element, u16 length, - struct ieee80211_network *network) + struct libipw_network *network) { DECLARE_SSID_BUF(ssid); u8 i; @@ -1129,7 +1130,7 @@ static int ieee80211_parse_info_param(struct ieee80211_info_element while (length >= sizeof(*info_element)) { if (sizeof(*info_element) + info_element->len > length) { - IEEE80211_DEBUG_MGMT("Info elem: parse failed: " + LIBIPW_DEBUG_MGMT("Info elem: parse failed: " "info_element->len + 2 > left : " "info_element->len+2=%zd left=%d, id=%d.\n", info_element->len + @@ -1151,7 +1152,7 @@ static int ieee80211_parse_info_param(struct ieee80211_info_element memset(network->ssid + network->ssid_len, 0, IW_ESSID_MAX_SIZE - network->ssid_len); - IEEE80211_DEBUG_MGMT("WLAN_EID_SSID: '%s' len=%d.\n", + LIBIPW_DEBUG_MGMT("WLAN_EID_SSID: '%s' len=%d.\n", print_ssid(ssid, network->ssid, network->ssid_len), network->ssid_len); @@ -1170,17 +1171,17 @@ static int ieee80211_parse_info_param(struct ieee80211_info_element (p - rates_str), "%02X ", network->rates[i]); #endif - if (ieee80211_is_ofdm_rate + if (libipw_is_ofdm_rate (info_element->data[i])) { network->flags |= NETWORK_HAS_OFDM; if (info_element->data[i] & - IEEE80211_BASIC_RATE_MASK) + LIBIPW_BASIC_RATE_MASK) network->flags &= ~NETWORK_HAS_CCK; } } - IEEE80211_DEBUG_MGMT("WLAN_EID_SUPP_RATES: '%s' (%d)\n", + LIBIPW_DEBUG_MGMT("WLAN_EID_SUPP_RATES: '%s' (%d)\n", rates_str, network->rates_len); break; @@ -1197,61 +1198,61 @@ static int ieee80211_parse_info_param(struct ieee80211_info_element (p - rates_str), "%02X ", network->rates[i]); #endif - if (ieee80211_is_ofdm_rate + if (libipw_is_ofdm_rate (info_element->data[i])) { network->flags |= NETWORK_HAS_OFDM; if (info_element->data[i] & - IEEE80211_BASIC_RATE_MASK) + LIBIPW_BASIC_RATE_MASK) network->flags &= ~NETWORK_HAS_CCK; } } - IEEE80211_DEBUG_MGMT("WLAN_EID_EXT_SUPP_RATES: '%s' (%d)\n", + LIBIPW_DEBUG_MGMT("WLAN_EID_EXT_SUPP_RATES: '%s' (%d)\n", rates_str, network->rates_ex_len); break; case WLAN_EID_DS_PARAMS: - IEEE80211_DEBUG_MGMT("WLAN_EID_DS_PARAMS: %d\n", + LIBIPW_DEBUG_MGMT("WLAN_EID_DS_PARAMS: %d\n", info_element->data[0]); network->channel = info_element->data[0]; break; case WLAN_EID_FH_PARAMS: - IEEE80211_DEBUG_MGMT("WLAN_EID_FH_PARAMS: ignored\n"); + LIBIPW_DEBUG_MGMT("WLAN_EID_FH_PARAMS: ignored\n"); break; case WLAN_EID_CF_PARAMS: - IEEE80211_DEBUG_MGMT("WLAN_EID_CF_PARAMS: ignored\n"); + LIBIPW_DEBUG_MGMT("WLAN_EID_CF_PARAMS: ignored\n"); break; case WLAN_EID_TIM: network->tim.tim_count = info_element->data[0]; network->tim.tim_period = info_element->data[1]; - IEEE80211_DEBUG_MGMT("WLAN_EID_TIM: partially ignored\n"); + LIBIPW_DEBUG_MGMT("WLAN_EID_TIM: partially ignored\n"); break; case WLAN_EID_ERP_INFO: network->erp_value = info_element->data[0]; network->flags |= NETWORK_HAS_ERP_VALUE; - IEEE80211_DEBUG_MGMT("MFIE_TYPE_ERP_SET: %d\n", + LIBIPW_DEBUG_MGMT("MFIE_TYPE_ERP_SET: %d\n", network->erp_value); break; case WLAN_EID_IBSS_PARAMS: network->atim_window = info_element->data[0]; - IEEE80211_DEBUG_MGMT("WLAN_EID_IBSS_PARAMS: %d\n", + LIBIPW_DEBUG_MGMT("WLAN_EID_IBSS_PARAMS: %d\n", network->atim_window); break; case WLAN_EID_CHALLENGE: - IEEE80211_DEBUG_MGMT("WLAN_EID_CHALLENGE: ignored\n"); + LIBIPW_DEBUG_MGMT("WLAN_EID_CHALLENGE: ignored\n"); break; case WLAN_EID_GENERIC: - IEEE80211_DEBUG_MGMT("WLAN_EID_GENERIC: %d bytes\n", + LIBIPW_DEBUG_MGMT("WLAN_EID_GENERIC: %d bytes\n", info_element->len); - if (!ieee80211_parse_qos_info_param_IE(info_element, + if (!libipw_parse_qos_info_param_IE(info_element, network)) break; @@ -1268,7 +1269,7 @@ static int ieee80211_parse_info_param(struct ieee80211_info_element break; case WLAN_EID_RSN: - IEEE80211_DEBUG_MGMT("WLAN_EID_RSN: %d bytes\n", + LIBIPW_DEBUG_MGMT("WLAN_EID_RSN: %d bytes\n", info_element->len); network->rsn_ie_len = min(info_element->len + 2, MAX_WPA_IE_LEN); @@ -1318,7 +1319,7 @@ static int ieee80211_parse_info_param(struct ieee80211_info_element break; default: - IEEE80211_DEBUG_MGMT + LIBIPW_DEBUG_MGMT ("Unsupported info element: %s (%d)\n", get_info_element_string(info_element->id), info_element->id); @@ -1327,20 +1328,20 @@ static int ieee80211_parse_info_param(struct ieee80211_info_element length -= sizeof(*info_element) + info_element->len; info_element = - (struct ieee80211_info_element *)&info_element-> + (struct libipw_info_element *)&info_element-> data[info_element->len]; } return 0; } -static int ieee80211_handle_assoc_resp(struct ieee80211_device *ieee, struct ieee80211_assoc_response - *frame, struct ieee80211_rx_stats *stats) +static int libipw_handle_assoc_resp(struct libipw_device *ieee, struct libipw_assoc_response + *frame, struct libipw_rx_stats *stats) { - struct ieee80211_network network_resp = { + struct libipw_network network_resp = { .ibss_dfs = NULL, }; - struct ieee80211_network *network = &network_resp; + struct libipw_network *network = &network_resp; struct net_device *dev = ieee->dev; network->flags = 0; @@ -1361,7 +1362,7 @@ static int ieee80211_handle_assoc_resp(struct ieee80211_device *ieee, struct iee network->erp_value = (network->capability & WLAN_CAPABILITY_IBSS) ? 0x3 : 0x0; - if (stats->freq == IEEE80211_52GHZ_BAND) { + if (stats->freq == LIBIPW_52GHZ_BAND) { /* for A band (No DS info) */ network->channel = stats->received_channel; } else @@ -1370,12 +1371,12 @@ static int ieee80211_handle_assoc_resp(struct ieee80211_device *ieee, struct iee network->wpa_ie_len = 0; network->rsn_ie_len = 0; - if (ieee80211_parse_info_param + if (libipw_parse_info_param (frame->info_element, stats->len - sizeof(*frame), network)) return 1; network->mode = 0; - if (stats->freq == IEEE80211_52GHZ_BAND) + if (stats->freq == LIBIPW_52GHZ_BAND) network->mode = IEEE_A; else { if (network->flags & NETWORK_HAS_OFDM) @@ -1394,10 +1395,10 @@ static int ieee80211_handle_assoc_resp(struct ieee80211_device *ieee, struct iee /***************************************************/ -static int ieee80211_network_init(struct ieee80211_device *ieee, struct ieee80211_probe_response +static int libipw_network_init(struct libipw_device *ieee, struct libipw_probe_response *beacon, - struct ieee80211_network *network, - struct ieee80211_rx_stats *stats) + struct libipw_network *network, + struct libipw_rx_stats *stats) { DECLARE_SSID_BUF(ssid); @@ -1423,7 +1424,7 @@ static int ieee80211_network_init(struct ieee80211_device *ieee, struct ieee8021 network->erp_value = (network->capability & WLAN_CAPABILITY_IBSS) ? 0x3 : 0x0; - if (stats->freq == IEEE80211_52GHZ_BAND) { + if (stats->freq == LIBIPW_52GHZ_BAND) { /* for A band (No DS info) */ network->channel = stats->received_channel; } else @@ -1432,12 +1433,12 @@ static int ieee80211_network_init(struct ieee80211_device *ieee, struct ieee8021 network->wpa_ie_len = 0; network->rsn_ie_len = 0; - if (ieee80211_parse_info_param + if (libipw_parse_info_param (beacon->info_element, stats->len - sizeof(*beacon), network)) return 1; network->mode = 0; - if (stats->freq == IEEE80211_52GHZ_BAND) + if (stats->freq == LIBIPW_52GHZ_BAND) network->mode = IEEE_A; else { if (network->flags & NETWORK_HAS_OFDM) @@ -1447,7 +1448,7 @@ static int ieee80211_network_init(struct ieee80211_device *ieee, struct ieee8021 } if (network->mode == 0) { - IEEE80211_DEBUG_SCAN("Filtered out '%s (%pM)' " + LIBIPW_DEBUG_SCAN("Filtered out '%s (%pM)' " "network.\n", print_ssid(ssid, network->ssid, network->ssid_len), @@ -1460,8 +1461,8 @@ static int ieee80211_network_init(struct ieee80211_device *ieee, struct ieee8021 return 0; } -static inline int is_same_network(struct ieee80211_network *src, - struct ieee80211_network *dst) +static inline int is_same_network(struct libipw_network *src, + struct libipw_network *dst) { /* A network is only a duplicate if the channel, BSSID, and ESSID * all match. We treat all with the same BSSID and channel @@ -1472,13 +1473,13 @@ static inline int is_same_network(struct ieee80211_network *src, !memcmp(src->ssid, dst->ssid, src->ssid_len)); } -static void update_network(struct ieee80211_network *dst, - struct ieee80211_network *src) +static void update_network(struct libipw_network *dst, + struct libipw_network *src) { int qos_active; u8 old_param; - ieee80211_network_reset(dst); + libipw_network_reset(dst); dst->ibss_dfs = src->ibss_dfs; /* We only update the statistics if they were created by receiving @@ -1488,9 +1489,9 @@ static void update_network(struct ieee80211_network *dst, * down the signal level of an AP. */ if (dst->channel == src->stats.received_channel) memcpy(&dst->stats, &src->stats, - sizeof(struct ieee80211_rx_stats)); + sizeof(struct libipw_rx_stats)); else - IEEE80211_DEBUG_SCAN("Network %pM info received " + LIBIPW_DEBUG_SCAN("Network %pM info received " "off channel (%d vs. %d)\n", src->bssid, dst->channel, src->stats.received_channel); @@ -1521,7 +1522,7 @@ static void update_network(struct ieee80211_network *dst, old_param = dst->qos_data.old_param_count; if (dst->flags & NETWORK_HAS_QOS_MASK) memcpy(&dst->qos_data, &src->qos_data, - sizeof(struct ieee80211_qos_data)); + sizeof(struct libipw_qos_data)); else { dst->qos_data.supported = src->qos_data.supported; dst->qos_data.param_count = src->qos_data.param_count; @@ -1529,11 +1530,11 @@ static void update_network(struct ieee80211_network *dst, if (dst->qos_data.supported == 1) { if (dst->ssid_len) - IEEE80211_DEBUG_QOS + LIBIPW_DEBUG_QOS ("QoS the network %s is QoS supported\n", dst->ssid); else - IEEE80211_DEBUG_QOS + LIBIPW_DEBUG_QOS ("QoS the network is QoS supported\n"); } dst->qos_data.active = qos_active; @@ -1547,25 +1548,25 @@ static inline int is_beacon(__le16 fc) return (WLAN_FC_GET_STYPE(le16_to_cpu(fc)) == IEEE80211_STYPE_BEACON); } -static void ieee80211_process_probe_response(struct ieee80211_device +static void libipw_process_probe_response(struct libipw_device *ieee, struct - ieee80211_probe_response - *beacon, struct ieee80211_rx_stats + libipw_probe_response + *beacon, struct libipw_rx_stats *stats) { struct net_device *dev = ieee->dev; - struct ieee80211_network network = { + struct libipw_network network = { .ibss_dfs = NULL, }; - struct ieee80211_network *target; - struct ieee80211_network *oldest = NULL; + struct libipw_network *target; + struct libipw_network *oldest = NULL; #ifdef CONFIG_LIBIPW_DEBUG - struct ieee80211_info_element *info_element = beacon->info_element; + struct libipw_info_element *info_element = beacon->info_element; #endif unsigned long flags; DECLARE_SSID_BUF(ssid); - IEEE80211_DEBUG_SCAN("'%s' (%pM" + LIBIPW_DEBUG_SCAN("'%s' (%pM" "): %c%c%c%c %c%c%c%c-%c%c%c%c %c%c%c%c\n", print_ssid(ssid, info_element->data, info_element->len), beacon->header.addr3, @@ -1586,8 +1587,8 @@ static void ieee80211_process_probe_response(struct ieee80211_device (beacon->capability & cpu_to_le16(1 << 0x1)) ? '1' : '0', (beacon->capability & cpu_to_le16(1 << 0x0)) ? '1' : '0'); - if (ieee80211_network_init(ieee, beacon, &network, stats)) { - IEEE80211_DEBUG_SCAN("Dropped '%s' (%pM) via %s.\n", + if (libipw_network_init(ieee, beacon, &network, stats)) { + LIBIPW_DEBUG_SCAN("Dropped '%s' (%pM) via %s.\n", print_ssid(ssid, info_element->data, info_element->len), beacon->header.addr3, @@ -1624,21 +1625,21 @@ static void ieee80211_process_probe_response(struct ieee80211_device /* If there are no more slots, expire the oldest */ list_del(&oldest->list); target = oldest; - IEEE80211_DEBUG_SCAN("Expired '%s' (%pM) from " + LIBIPW_DEBUG_SCAN("Expired '%s' (%pM) from " "network list.\n", print_ssid(ssid, target->ssid, target->ssid_len), target->bssid); - ieee80211_network_reset(target); + libipw_network_reset(target); } else { /* Otherwise just pull from the free list */ target = list_entry(ieee->network_free_list.next, - struct ieee80211_network, list); + struct libipw_network, list); list_del(ieee->network_free_list.next); } #ifdef CONFIG_LIBIPW_DEBUG - IEEE80211_DEBUG_SCAN("Adding '%s' (%pM) via %s.\n", + LIBIPW_DEBUG_SCAN("Adding '%s' (%pM) via %s.\n", print_ssid(ssid, network.ssid, network.ssid_len), network.bssid, @@ -1649,7 +1650,7 @@ static void ieee80211_process_probe_response(struct ieee80211_device network.ibss_dfs = NULL; list_add_tail(&target->list, &ieee->network_list); } else { - IEEE80211_DEBUG_SCAN("Updating '%s' (%pM) via %s.\n", + LIBIPW_DEBUG_SCAN("Updating '%s' (%pM) via %s.\n", print_ssid(ssid, target->ssid, target->ssid_len), target->bssid, @@ -1670,121 +1671,121 @@ static void ieee80211_process_probe_response(struct ieee80211_device } } -void ieee80211_rx_mgt(struct ieee80211_device *ieee, - struct ieee80211_hdr_4addr *header, - struct ieee80211_rx_stats *stats) +void libipw_rx_mgt(struct libipw_device *ieee, + struct libipw_hdr_4addr *header, + struct libipw_rx_stats *stats) { switch (WLAN_FC_GET_STYPE(le16_to_cpu(header->frame_ctl))) { case IEEE80211_STYPE_ASSOC_RESP: - IEEE80211_DEBUG_MGMT("received ASSOCIATION RESPONSE (%d)\n", + LIBIPW_DEBUG_MGMT("received ASSOCIATION RESPONSE (%d)\n", WLAN_FC_GET_STYPE(le16_to_cpu (header->frame_ctl))); - ieee80211_handle_assoc_resp(ieee, - (struct ieee80211_assoc_response *) + libipw_handle_assoc_resp(ieee, + (struct libipw_assoc_response *) header, stats); break; case IEEE80211_STYPE_REASSOC_RESP: - IEEE80211_DEBUG_MGMT("received REASSOCIATION RESPONSE (%d)\n", + LIBIPW_DEBUG_MGMT("received REASSOCIATION RESPONSE (%d)\n", WLAN_FC_GET_STYPE(le16_to_cpu (header->frame_ctl))); break; case IEEE80211_STYPE_PROBE_REQ: - IEEE80211_DEBUG_MGMT("received auth (%d)\n", + LIBIPW_DEBUG_MGMT("received auth (%d)\n", WLAN_FC_GET_STYPE(le16_to_cpu (header->frame_ctl))); if (ieee->handle_probe_request != NULL) ieee->handle_probe_request(ieee->dev, (struct - ieee80211_probe_request *) + libipw_probe_request *) header, stats); break; case IEEE80211_STYPE_PROBE_RESP: - IEEE80211_DEBUG_MGMT("received PROBE RESPONSE (%d)\n", + LIBIPW_DEBUG_MGMT("received PROBE RESPONSE (%d)\n", WLAN_FC_GET_STYPE(le16_to_cpu (header->frame_ctl))); - IEEE80211_DEBUG_SCAN("Probe response\n"); - ieee80211_process_probe_response(ieee, + LIBIPW_DEBUG_SCAN("Probe response\n"); + libipw_process_probe_response(ieee, (struct - ieee80211_probe_response *) + libipw_probe_response *) header, stats); break; case IEEE80211_STYPE_BEACON: - IEEE80211_DEBUG_MGMT("received BEACON (%d)\n", + LIBIPW_DEBUG_MGMT("received BEACON (%d)\n", WLAN_FC_GET_STYPE(le16_to_cpu (header->frame_ctl))); - IEEE80211_DEBUG_SCAN("Beacon\n"); - ieee80211_process_probe_response(ieee, + LIBIPW_DEBUG_SCAN("Beacon\n"); + libipw_process_probe_response(ieee, (struct - ieee80211_probe_response *) + libipw_probe_response *) header, stats); break; case IEEE80211_STYPE_AUTH: - IEEE80211_DEBUG_MGMT("received auth (%d)\n", + LIBIPW_DEBUG_MGMT("received auth (%d)\n", WLAN_FC_GET_STYPE(le16_to_cpu (header->frame_ctl))); if (ieee->handle_auth != NULL) ieee->handle_auth(ieee->dev, - (struct ieee80211_auth *)header); + (struct libipw_auth *)header); break; case IEEE80211_STYPE_DISASSOC: if (ieee->handle_disassoc != NULL) ieee->handle_disassoc(ieee->dev, - (struct ieee80211_disassoc *) + (struct libipw_disassoc *) header); break; case IEEE80211_STYPE_ACTION: - IEEE80211_DEBUG_MGMT("ACTION\n"); + LIBIPW_DEBUG_MGMT("ACTION\n"); if (ieee->handle_action) ieee->handle_action(ieee->dev, - (struct ieee80211_action *) + (struct libipw_action *) header, stats); break; case IEEE80211_STYPE_REASSOC_REQ: - IEEE80211_DEBUG_MGMT("received reassoc (%d)\n", + LIBIPW_DEBUG_MGMT("received reassoc (%d)\n", WLAN_FC_GET_STYPE(le16_to_cpu (header->frame_ctl))); - IEEE80211_DEBUG_MGMT("%s: IEEE80211_REASSOC_REQ received\n", + LIBIPW_DEBUG_MGMT("%s: LIBIPW_REASSOC_REQ received\n", ieee->dev->name); if (ieee->handle_reassoc_request != NULL) ieee->handle_reassoc_request(ieee->dev, - (struct ieee80211_reassoc_request *) + (struct libipw_reassoc_request *) header); break; case IEEE80211_STYPE_ASSOC_REQ: - IEEE80211_DEBUG_MGMT("received assoc (%d)\n", + LIBIPW_DEBUG_MGMT("received assoc (%d)\n", WLAN_FC_GET_STYPE(le16_to_cpu (header->frame_ctl))); - IEEE80211_DEBUG_MGMT("%s: IEEE80211_ASSOC_REQ received\n", + LIBIPW_DEBUG_MGMT("%s: LIBIPW_ASSOC_REQ received\n", ieee->dev->name); if (ieee->handle_assoc_request != NULL) ieee->handle_assoc_request(ieee->dev); break; case IEEE80211_STYPE_DEAUTH: - IEEE80211_DEBUG_MGMT("DEAUTH\n"); + LIBIPW_DEBUG_MGMT("DEAUTH\n"); if (ieee->handle_deauth != NULL) ieee->handle_deauth(ieee->dev, - (struct ieee80211_deauth *) + (struct libipw_deauth *) header); break; default: - IEEE80211_DEBUG_MGMT("received UNKNOWN (%d)\n", + LIBIPW_DEBUG_MGMT("received UNKNOWN (%d)\n", WLAN_FC_GET_STYPE(le16_to_cpu (header->frame_ctl))); - IEEE80211_DEBUG_MGMT("%s: Unknown management packet: %d\n", + LIBIPW_DEBUG_MGMT("%s: Unknown management packet: %d\n", ieee->dev->name, WLAN_FC_GET_STYPE(le16_to_cpu (header->frame_ctl))); @@ -1792,6 +1793,6 @@ void ieee80211_rx_mgt(struct ieee80211_device *ieee, } } -EXPORT_SYMBOL_GPL(ieee80211_rx_any); -EXPORT_SYMBOL(ieee80211_rx_mgt); -EXPORT_SYMBOL(ieee80211_rx); +EXPORT_SYMBOL_GPL(libipw_rx_any); +EXPORT_SYMBOL(libipw_rx_mgt); +EXPORT_SYMBOL(libipw_rx); diff --git a/drivers/net/wireless/ipw2x00/libipw_tx.c b/drivers/net/wireless/ipw2x00/libipw_tx.c index 2e8f84fb29fa..ce1855f32b7d 100644 --- a/drivers/net/wireless/ipw2x00/libipw_tx.c +++ b/drivers/net/wireless/ipw2x00/libipw_tx.c @@ -41,7 +41,7 @@ #include #include -#include "ieee80211.h" +#include "libipw.h" /* @@ -126,12 +126,12 @@ payload of each frame is reduced to 492 bytes. static u8 P802_1H_OUI[P80211_OUI_LEN] = { 0x00, 0x00, 0xf8 }; static u8 RFC1042_OUI[P80211_OUI_LEN] = { 0x00, 0x00, 0x00 }; -static int ieee80211_copy_snap(u8 * data, __be16 h_proto) +static int libipw_copy_snap(u8 * data, __be16 h_proto) { - struct ieee80211_snap_hdr *snap; + struct libipw_snap_hdr *snap; u8 *oui; - snap = (struct ieee80211_snap_hdr *)data; + snap = (struct libipw_snap_hdr *)data; snap->dsap = 0xaa; snap->ssap = 0xaa; snap->ctrl = 0x03; @@ -149,7 +149,7 @@ static int ieee80211_copy_snap(u8 * data, __be16 h_proto) return SNAP_SIZE + sizeof(u16); } -static int ieee80211_encrypt_fragment(struct ieee80211_device *ieee, +static int libipw_encrypt_fragment(struct libipw_device *ieee, struct sk_buff *frag, int hdr_len) { struct lib80211_crypt_data *crypt = @@ -177,7 +177,7 @@ static int ieee80211_encrypt_fragment(struct ieee80211_device *ieee, return 0; } -void ieee80211_txb_free(struct ieee80211_txb *txb) +void libipw_txb_free(struct libipw_txb *txb) { int i; if (unlikely(!txb)) @@ -188,17 +188,17 @@ void ieee80211_txb_free(struct ieee80211_txb *txb) kfree(txb); } -static struct ieee80211_txb *ieee80211_alloc_txb(int nr_frags, int txb_size, +static struct libipw_txb *libipw_alloc_txb(int nr_frags, int txb_size, int headroom, gfp_t gfp_mask) { - struct ieee80211_txb *txb; + struct libipw_txb *txb; int i; - txb = kmalloc(sizeof(struct ieee80211_txb) + (sizeof(u8 *) * nr_frags), + txb = kmalloc(sizeof(struct libipw_txb) + (sizeof(u8 *) * nr_frags), gfp_mask); if (!txb) return NULL; - memset(txb, 0, sizeof(struct ieee80211_txb)); + memset(txb, 0, sizeof(struct libipw_txb)); txb->nr_frags = nr_frags; txb->frag_size = txb_size; @@ -220,7 +220,7 @@ static struct ieee80211_txb *ieee80211_alloc_txb(int nr_frags, int txb_size, return txb; } -static int ieee80211_classify(struct sk_buff *skb) +static int libipw_classify(struct sk_buff *skb) { struct ethhdr *eth; struct iphdr *ip; @@ -252,11 +252,11 @@ static int ieee80211_classify(struct sk_buff *skb) /* Incoming skb is converted to a txb which consists of * a block of 802.11 fragment packets (stored as skbs) */ -int ieee80211_xmit(struct sk_buff *skb, struct net_device *dev) +int libipw_xmit(struct sk_buff *skb, struct net_device *dev) { - struct ieee80211_device *ieee = netdev_priv(dev); - struct ieee80211_txb *txb = NULL; - struct ieee80211_hdr_3addrqos *frag_hdr; + struct libipw_device *ieee = netdev_priv(dev); + struct libipw_txb *txb = NULL; + struct libipw_hdr_3addrqos *frag_hdr; int i, bytes_per_frag, nr_frags, bytes_last_frag, frag_size, rts_required; unsigned long flags; @@ -264,7 +264,7 @@ int ieee80211_xmit(struct sk_buff *skb, struct net_device *dev) __be16 ether_type; int bytes, fc, hdr_len; struct sk_buff *skb_frag; - struct ieee80211_hdr_3addrqos header = {/* Ensure zero initialized */ + struct libipw_hdr_3addrqos header = {/* Ensure zero initialized */ .duration_id = 0, .seq_ctl = 0, .qos_ctl = 0 @@ -331,14 +331,14 @@ int ieee80211_xmit(struct sk_buff *skb, struct net_device *dev) memcpy(header.addr2, src, ETH_ALEN); memcpy(header.addr3, ieee->bssid, ETH_ALEN); } - hdr_len = IEEE80211_3ADDR_LEN; + hdr_len = LIBIPW_3ADDR_LEN; if (ieee->is_qos_active && ieee->is_qos_active(dev, skb)) { fc |= IEEE80211_STYPE_QOS_DATA; hdr_len += 2; - skb->priority = ieee80211_classify(skb); - header.qos_ctl |= cpu_to_le16(skb->priority & IEEE80211_QCTL_TID); + skb->priority = libipw_classify(skb); + header.qos_ctl |= cpu_to_le16(skb->priority & LIBIPW_QCTL_TID); } header.frame_ctl = cpu_to_le16(fc); @@ -362,12 +362,12 @@ int ieee80211_xmit(struct sk_buff *skb, struct net_device *dev) skb_reserve(skb_new, crypt->ops->extra_msdu_prefix_len); memcpy(skb_put(skb_new, hdr_len), &header, hdr_len); snapped = 1; - ieee80211_copy_snap(skb_put(skb_new, SNAP_SIZE + sizeof(u16)), + libipw_copy_snap(skb_put(skb_new, SNAP_SIZE + sizeof(u16)), ether_type); skb_copy_from_linear_data(skb, skb_put(skb_new, skb->len), skb->len); res = crypt->ops->encrypt_msdu(skb_new, hdr_len, crypt->priv); if (res < 0) { - IEEE80211_ERROR("msdu encryption failed\n"); + LIBIPW_ERROR("msdu encryption failed\n"); dev_kfree_skb_any(skb_new); goto failed; } @@ -393,8 +393,8 @@ int ieee80211_xmit(struct sk_buff *skb, struct net_device *dev) * for it when determining the amount of payload space. */ bytes_per_frag = frag_size - hdr_len; if (ieee->config & - (CFG_IEEE80211_COMPUTE_FCS | CFG_IEEE80211_RESERVE_FCS)) - bytes_per_frag -= IEEE80211_FCS_LEN; + (CFG_LIBIPW_COMPUTE_FCS | CFG_LIBIPW_RESERVE_FCS)) + bytes_per_frag -= LIBIPW_FCS_LEN; /* Each fragment may need to have room for encryptiong * pre/postfix */ @@ -417,14 +417,14 @@ int ieee80211_xmit(struct sk_buff *skb, struct net_device *dev) } rts_required = (frag_size > ieee->rts - && ieee->config & CFG_IEEE80211_RTS); + && ieee->config & CFG_LIBIPW_RTS); if (rts_required) nr_frags++; /* When we allocate the TXB we allocate enough space for the reserve * and full fragment bytes (bytes_per_frag doesn't include prefix, * postfix, header, FCS, etc.) */ - txb = ieee80211_alloc_txb(nr_frags, frag_size, + txb = libipw_alloc_txb(nr_frags, frag_size, ieee->tx_headroom, GFP_ATOMIC); if (unlikely(!txb)) { printk(KERN_WARNING "%s: Could not allocate TXB\n", @@ -441,7 +441,7 @@ int ieee80211_xmit(struct sk_buff *skb, struct net_device *dev) if (rts_required) { skb_frag = txb->fragments[0]; frag_hdr = - (struct ieee80211_hdr_3addrqos *)skb_put(skb_frag, hdr_len); + (struct libipw_hdr_3addrqos *)skb_put(skb_frag, hdr_len); /* * Set header frame_ctl to the RTS. @@ -456,7 +456,7 @@ int ieee80211_xmit(struct sk_buff *skb, struct net_device *dev) header.frame_ctl = cpu_to_le16(fc); if (ieee->config & - (CFG_IEEE80211_COMPUTE_FCS | CFG_IEEE80211_RESERVE_FCS)) + (CFG_LIBIPW_COMPUTE_FCS | CFG_LIBIPW_RESERVE_FCS)) skb_put(skb_frag, 4); txb->rts_included = 1; @@ -472,7 +472,7 @@ int ieee80211_xmit(struct sk_buff *skb, struct net_device *dev) crypt->ops->extra_mpdu_prefix_len); frag_hdr = - (struct ieee80211_hdr_3addrqos *)skb_put(skb_frag, hdr_len); + (struct libipw_hdr_3addrqos *)skb_put(skb_frag, hdr_len); memcpy(frag_hdr, &header, hdr_len); /* If this is not the last fragment, then add the MOREFRAGS @@ -487,7 +487,7 @@ int ieee80211_xmit(struct sk_buff *skb, struct net_device *dev) } if (i == 0 && !snapped) { - ieee80211_copy_snap(skb_put + libipw_copy_snap(skb_put (skb_frag, SNAP_SIZE + sizeof(u16)), ether_type); bytes -= SNAP_SIZE + sizeof(u16); @@ -501,7 +501,7 @@ int ieee80211_xmit(struct sk_buff *skb, struct net_device *dev) /* Encryption routine will move the header forward in order * to insert the IV between the header and the payload */ if (host_encrypt) - ieee80211_encrypt_fragment(ieee, skb_frag, hdr_len); + libipw_encrypt_fragment(ieee, skb_frag, hdr_len); else if (host_build_iv) { atomic_inc(&crypt->refcnt); if (crypt->ops->build_iv) @@ -513,7 +513,7 @@ int ieee80211_xmit(struct sk_buff *skb, struct net_device *dev) } if (ieee->config & - (CFG_IEEE80211_COMPUTE_FCS | CFG_IEEE80211_RESERVE_FCS)) + (CFG_LIBIPW_COMPUTE_FCS | CFG_LIBIPW_RESERVE_FCS)) skb_put(skb_frag, 4); } @@ -530,7 +530,7 @@ int ieee80211_xmit(struct sk_buff *skb, struct net_device *dev) return NETDEV_TX_OK; } - ieee80211_txb_free(txb); + libipw_txb_free(txb); } return NETDEV_TX_OK; @@ -541,6 +541,6 @@ int ieee80211_xmit(struct sk_buff *skb, struct net_device *dev) dev->stats.tx_errors++; return NETDEV_TX_BUSY; } -EXPORT_SYMBOL(ieee80211_xmit); +EXPORT_SYMBOL(libipw_xmit); -EXPORT_SYMBOL(ieee80211_txb_free); +EXPORT_SYMBOL(libipw_txb_free); diff --git a/drivers/net/wireless/ipw2x00/libipw_wx.c b/drivers/net/wireless/ipw2x00/libipw_wx.c index 3c0812db030a..f79ce57ebe68 100644 --- a/drivers/net/wireless/ipw2x00/libipw_wx.c +++ b/drivers/net/wireless/ipw2x00/libipw_wx.c @@ -37,9 +37,9 @@ #include #include -#include "ieee80211.h" +#include "libipw.h" -static const char *ieee80211_modes[] = { +static const char *libipw_modes[] = { "?", "a", "b", "ab", "g", "ag", "bg", "abg" }; @@ -54,9 +54,9 @@ static inline unsigned int elapsed_jiffies_msecs(unsigned long start) } #define MAX_CUSTOM_LEN 64 -static char *ieee80211_translate_scan(struct ieee80211_device *ieee, +static char *libipw_translate_scan(struct libipw_device *ieee, char *start, char *stop, - struct ieee80211_network *network, + struct libipw_network *network, struct iw_request_info *info) { char custom[MAX_CUSTOM_LEN]; @@ -84,7 +84,7 @@ static char *ieee80211_translate_scan(struct ieee80211_device *ieee, /* Add the protocol name */ iwe.cmd = SIOCGIWNAME; snprintf(iwe.u.name, IFNAMSIZ, "IEEE 802.11%s", - ieee80211_modes[network->mode]); + libipw_modes[network->mode]); start = iwe_stream_add_event(info, start, stop, &iwe, IW_EV_CHAR_LEN); /* Add mode */ @@ -102,7 +102,7 @@ static char *ieee80211_translate_scan(struct ieee80211_device *ieee, /* Add channel and frequency */ /* Note : userspace automatically computes channel using iwrange */ iwe.cmd = SIOCGIWFREQ; - iwe.u.freq.m = ieee80211_channel_to_freq(ieee, network->channel); + iwe.u.freq.m = libipw_channel_to_freq(ieee, network->channel); iwe.u.freq.e = 6; iwe.u.freq.i = 0; start = iwe_stream_add_event(info, start, stop, &iwe, IW_EV_FREQ_LEN); @@ -155,7 +155,7 @@ static char *ieee80211_translate_scan(struct ieee80211_device *ieee, iwe.u.qual.updated = IW_QUAL_QUAL_UPDATED | IW_QUAL_LEVEL_UPDATED | IW_QUAL_NOISE_UPDATED; - if (!(network->stats.mask & IEEE80211_STATMASK_RSSI)) { + if (!(network->stats.mask & LIBIPW_STATMASK_RSSI)) { iwe.u.qual.updated |= IW_QUAL_QUAL_INVALID | IW_QUAL_LEVEL_INVALID; iwe.u.qual.qual = 0; @@ -180,14 +180,14 @@ static char *ieee80211_translate_scan(struct ieee80211_device *ieee, iwe.u.qual.qual = 0; } - if (!(network->stats.mask & IEEE80211_STATMASK_NOISE)) { + if (!(network->stats.mask & LIBIPW_STATMASK_NOISE)) { iwe.u.qual.updated |= IW_QUAL_NOISE_INVALID; iwe.u.qual.noise = 0; } else { iwe.u.qual.noise = network->stats.noise; } - if (!(network->stats.mask & IEEE80211_STATMASK_SIGNAL)) { + if (!(network->stats.mask & LIBIPW_STATMASK_SIGNAL)) { iwe.u.qual.updated |= IW_QUAL_LEVEL_INVALID; iwe.u.qual.level = 0; } else { @@ -237,14 +237,14 @@ static char *ieee80211_translate_scan(struct ieee80211_device *ieee, p = custom; p += snprintf(p, MAX_CUSTOM_LEN - (p - custom), " Channel flags: "); - if (ieee80211_get_channel_flags(ieee, network->channel) & - IEEE80211_CH_INVALID) { + if (libipw_get_channel_flags(ieee, network->channel) & + LIBIPW_CH_INVALID) { iwe.cmd = IWEVCUSTOM; p += snprintf(p, MAX_CUSTOM_LEN - (p - custom), "INVALID "); } - if (ieee80211_get_channel_flags(ieee, network->channel) & - IEEE80211_CH_RADAR_DETECT) { + if (libipw_get_channel_flags(ieee, network->channel) & + LIBIPW_CH_RADAR_DETECT) { iwe.cmd = IWEVCUSTOM; p += snprintf(p, MAX_CUSTOM_LEN - (p - custom), "DFS "); } @@ -259,11 +259,11 @@ static char *ieee80211_translate_scan(struct ieee80211_device *ieee, #define SCAN_ITEM_SIZE 128 -int ieee80211_wx_get_scan(struct ieee80211_device *ieee, +int libipw_wx_get_scan(struct libipw_device *ieee, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { - struct ieee80211_network *network; + struct libipw_network *network; unsigned long flags; int err = 0; @@ -272,7 +272,7 @@ int ieee80211_wx_get_scan(struct ieee80211_device *ieee, int i = 0; DECLARE_SSID_BUF(ssid); - IEEE80211_DEBUG_WX("Getting scan\n"); + LIBIPW_DEBUG_WX("Getting scan\n"); spin_lock_irqsave(&ieee->lock, flags); @@ -285,10 +285,10 @@ int ieee80211_wx_get_scan(struct ieee80211_device *ieee, if (ieee->scan_age == 0 || time_after(network->last_scanned + ieee->scan_age, jiffies)) - ev = ieee80211_translate_scan(ieee, ev, stop, network, + ev = libipw_translate_scan(ieee, ev, stop, network, info); else { - IEEE80211_DEBUG_SCAN("Not showing network '%s (" + LIBIPW_DEBUG_SCAN("Not showing network '%s (" "%pM)' due to age (%ums).\n", print_ssid(ssid, network->ssid, network->ssid_len), @@ -303,18 +303,18 @@ int ieee80211_wx_get_scan(struct ieee80211_device *ieee, wrqu->data.length = ev - extra; wrqu->data.flags = 0; - IEEE80211_DEBUG_WX("exit: %d networks returned.\n", i); + LIBIPW_DEBUG_WX("exit: %d networks returned.\n", i); return err; } -int ieee80211_wx_set_encode(struct ieee80211_device *ieee, +int libipw_wx_set_encode(struct libipw_device *ieee, struct iw_request_info *info, union iwreq_data *wrqu, char *keybuf) { struct iw_point *erq = &(wrqu->encoding); struct net_device *dev = ieee->dev; - struct ieee80211_security sec = { + struct libipw_security sec = { .flags = 0 }; int i, key, key_provided, len; @@ -322,7 +322,7 @@ int ieee80211_wx_set_encode(struct ieee80211_device *ieee, int host_crypto = ieee->host_encrypt || ieee->host_decrypt || ieee->host_build_iv; DECLARE_SSID_BUF(ssid); - IEEE80211_DEBUG_WX("SET_ENCODE\n"); + LIBIPW_DEBUG_WX("SET_ENCODE\n"); key = erq->flags & IW_ENCODE_INDEX; if (key) { @@ -335,18 +335,18 @@ int ieee80211_wx_set_encode(struct ieee80211_device *ieee, key = ieee->crypt_info.tx_keyidx; } - IEEE80211_DEBUG_WX("Key: %d [%s]\n", key, key_provided ? + LIBIPW_DEBUG_WX("Key: %d [%s]\n", key, key_provided ? "provided" : "default"); crypt = &ieee->crypt_info.crypt[key]; if (erq->flags & IW_ENCODE_DISABLED) { if (key_provided && *crypt) { - IEEE80211_DEBUG_WX("Disabling encryption on key %d.\n", + LIBIPW_DEBUG_WX("Disabling encryption on key %d.\n", key); lib80211_crypt_delayed_deinit(&ieee->crypt_info, crypt); } else - IEEE80211_DEBUG_WX("Disabling encryption.\n"); + LIBIPW_DEBUG_WX("Disabling encryption.\n"); /* Check all the keys to see if any are still configured, * and if no key index was provided, de-init them all */ @@ -410,7 +410,7 @@ int ieee80211_wx_set_encode(struct ieee80211_device *ieee, /* If a new key was provided, set it up */ if (erq->length > 0) { -#ifdef CONFIG_IEEE80211_DEBUG +#ifdef CONFIG_LIBIPW_DEBUG DECLARE_SSID_BUF(ssid); #endif @@ -419,7 +419,7 @@ int ieee80211_wx_set_encode(struct ieee80211_device *ieee, if (len > erq->length) memset(sec.keys[key] + erq->length, 0, len - erq->length); - IEEE80211_DEBUG_WX("Setting key %d to '%s' (%d:%d bytes)\n", + LIBIPW_DEBUG_WX("Setting key %d to '%s' (%d:%d bytes)\n", key, print_ssid(ssid, sec.keys[key], len), erq->length, len); sec.key_sizes[key] = len; @@ -438,7 +438,7 @@ int ieee80211_wx_set_encode(struct ieee80211_device *ieee, NULL, (*crypt)->priv); if (len == 0) { /* Set a default key of all 0 */ - IEEE80211_DEBUG_WX("Setting key %d to all " + LIBIPW_DEBUG_WX("Setting key %d to all " "zero.\n", key); memset(sec.keys[key], 0, 13); (*crypt)->ops->set_key(sec.keys[key], 13, NULL, @@ -449,7 +449,7 @@ int ieee80211_wx_set_encode(struct ieee80211_device *ieee, } /* No key data - just set the default TX key index */ if (key_provided) { - IEEE80211_DEBUG_WX("Setting key %d to default Tx " + LIBIPW_DEBUG_WX("Setting key %d to default Tx " "key.\n", key); ieee->crypt_info.tx_keyidx = key; sec.active_key = key; @@ -461,7 +461,7 @@ int ieee80211_wx_set_encode(struct ieee80211_device *ieee, sec.auth_mode = ieee->open_wep ? WLAN_AUTH_OPEN : WLAN_AUTH_SHARED_KEY; sec.flags |= SEC_AUTH_MODE; - IEEE80211_DEBUG_WX("Auth: %s\n", + LIBIPW_DEBUG_WX("Auth: %s\n", sec.auth_mode == WLAN_AUTH_OPEN ? "OPEN" : "SHARED KEY"); } @@ -490,16 +490,16 @@ int ieee80211_wx_set_encode(struct ieee80211_device *ieee, return 0; } -int ieee80211_wx_get_encode(struct ieee80211_device *ieee, +int libipw_wx_get_encode(struct libipw_device *ieee, struct iw_request_info *info, union iwreq_data *wrqu, char *keybuf) { struct iw_point *erq = &(wrqu->encoding); int len, key; struct lib80211_crypt_data *crypt; - struct ieee80211_security *sec = &ieee->sec; + struct libipw_security *sec = &ieee->sec; - IEEE80211_DEBUG_WX("GET_ENCODE\n"); + LIBIPW_DEBUG_WX("GET_ENCODE\n"); key = erq->flags & IW_ENCODE_INDEX; if (key) { @@ -532,7 +532,7 @@ int ieee80211_wx_get_encode(struct ieee80211_device *ieee, return 0; } -int ieee80211_wx_set_encodeext(struct ieee80211_device *ieee, +int libipw_wx_set_encodeext(struct libipw_device *ieee, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { @@ -545,7 +545,7 @@ int ieee80211_wx_set_encodeext(struct ieee80211_device *ieee, struct lib80211_crypto_ops *ops; struct lib80211_crypt_data **crypt; - struct ieee80211_security sec = { + struct libipw_security sec = { .flags = 0, }; @@ -611,7 +611,7 @@ int ieee80211_wx_set_encodeext(struct ieee80211_device *ieee, module = "lib80211_crypt_ccmp"; break; default: - IEEE80211_DEBUG_WX("%s: unknown crypto alg %d\n", + LIBIPW_DEBUG_WX("%s: unknown crypto alg %d\n", dev->name, ext->alg); ret = -EINVAL; goto done; @@ -623,7 +623,7 @@ int ieee80211_wx_set_encodeext(struct ieee80211_device *ieee, ops = lib80211_get_crypto_ops(alg); } if (ops == NULL) { - IEEE80211_DEBUG_WX("%s: unknown crypto alg %d\n", + LIBIPW_DEBUG_WX("%s: unknown crypto alg %d\n", dev->name, ext->alg); ret = -EINVAL; goto done; @@ -653,7 +653,7 @@ int ieee80211_wx_set_encodeext(struct ieee80211_device *ieee, if (ext->key_len > 0 && (*crypt)->ops->set_key && (*crypt)->ops->set_key(ext->key, ext->key_len, ext->rx_seq, (*crypt)->priv) < 0) { - IEEE80211_DEBUG_WX("%s: key setting failed\n", dev->name); + LIBIPW_DEBUG_WX("%s: key setting failed\n", dev->name); ret = -EINVAL; goto done; } @@ -700,20 +700,20 @@ int ieee80211_wx_set_encodeext(struct ieee80211_device *ieee, if (ieee->reset_on_keychange && ieee->iw_mode != IW_MODE_INFRA && ieee->reset_port && ieee->reset_port(dev)) { - IEEE80211_DEBUG_WX("%s: reset_port failed\n", dev->name); + LIBIPW_DEBUG_WX("%s: reset_port failed\n", dev->name); return -EINVAL; } return ret; } -int ieee80211_wx_get_encodeext(struct ieee80211_device *ieee, +int libipw_wx_get_encodeext(struct libipw_device *ieee, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { struct iw_point *encoding = &wrqu->encoding; struct iw_encode_ext *ext = (struct iw_encode_ext *)extra; - struct ieee80211_security *sec = &ieee->sec; + struct libipw_security *sec = &ieee->sec; int idx, max_key_len; max_key_len = encoding->length - sizeof(*ext); @@ -763,9 +763,9 @@ int ieee80211_wx_get_encodeext(struct ieee80211_device *ieee, return 0; } -EXPORT_SYMBOL(ieee80211_wx_set_encodeext); -EXPORT_SYMBOL(ieee80211_wx_get_encodeext); +EXPORT_SYMBOL(libipw_wx_set_encodeext); +EXPORT_SYMBOL(libipw_wx_get_encodeext); -EXPORT_SYMBOL(ieee80211_wx_get_scan); -EXPORT_SYMBOL(ieee80211_wx_set_encode); -EXPORT_SYMBOL(ieee80211_wx_get_encode); +EXPORT_SYMBOL(libipw_wx_get_scan); +EXPORT_SYMBOL(libipw_wx_set_encode); +EXPORT_SYMBOL(libipw_wx_get_encode); diff --git a/include/net/iw_handler.h b/include/net/iw_handler.h index 2b3fbbb8669e..e9054a283fde 100644 --- a/include/net/iw_handler.h +++ b/include/net/iw_handler.h @@ -416,13 +416,13 @@ struct iw_spy_data * data (i.e. valid as long as struct net_device exist, same locking rules). */ /* Forward declaration */ -struct ieee80211_device; +struct libipw_device; /* The struct */ struct iw_public_data { /* Driver enhanced spy support */ struct iw_spy_data * spy_data; - /* Structure managed by the in-kernel IEEE 802.11 layer */ - struct ieee80211_device * ieee80211; + /* Legacy structure managed by the ipw2x00-specific IEEE 802.11 layer */ + struct libipw_device * libipw; }; /**************************** PROTOTYPES ****************************/ -- cgit v1.2.3 From 103bf9f7d35849bce52ad412e4da5063b0716969 Mon Sep 17 00:00:00 2001 From: "John W. Linville" Date: Thu, 20 Aug 2009 16:34:15 -0400 Subject: mac80211: remove ieee80211_rx namespace hack With the libipw naming scheme change, it is no longer necessary for mac80211 to avoid the ieee80211_rx name clash. Reported-by: Johannes Berg Signed-off-by: John W. Linville --- include/net/mac80211.h | 11 +---------- net/mac80211/rx.c | 4 ++-- 2 files changed, 3 insertions(+), 12 deletions(-) (limited to 'include') diff --git a/include/net/mac80211.h b/include/net/mac80211.h index aac84d7bd46e..466859b285e1 100644 --- a/include/net/mac80211.h +++ b/include/net/mac80211.h @@ -1657,12 +1657,6 @@ void ieee80211_free_hw(struct ieee80211_hw *hw); */ void ieee80211_restart_hw(struct ieee80211_hw *hw); -/* - * trick to avoid symbol clashes with the ieee80211 subsystem, - * use the inline below instead - */ -void __ieee80211_rx(struct ieee80211_hw *hw, struct sk_buff *skb); - /** * ieee80211_rx - receive frame * @@ -1678,10 +1672,7 @@ void __ieee80211_rx(struct ieee80211_hw *hw, struct sk_buff *skb); * @hw: the hardware this frame came in on * @skb: the buffer to receive, owned by mac80211 after this call */ -static inline void ieee80211_rx(struct ieee80211_hw *hw, struct sk_buff *skb) -{ - __ieee80211_rx(hw, skb); -} +void ieee80211_rx(struct ieee80211_hw *hw, struct sk_buff *skb); /** * ieee80211_rx_irqsafe - receive frame diff --git a/net/mac80211/rx.c b/net/mac80211/rx.c index 7065fd7e7ba2..dff2239db6e2 100644 --- a/net/mac80211/rx.c +++ b/net/mac80211/rx.c @@ -2440,7 +2440,7 @@ static u8 ieee80211_rx_reorder_ampdu(struct ieee80211_local *local, * This is the receive path handler. It is called by a low level driver when an * 802.11 MPDU is received from the hardware. */ -void __ieee80211_rx(struct ieee80211_hw *hw, struct sk_buff *skb) +void ieee80211_rx(struct ieee80211_hw *hw, struct sk_buff *skb) { struct ieee80211_local *local = hw_to_local(hw); struct ieee80211_rate *rate = NULL; @@ -2523,7 +2523,7 @@ void __ieee80211_rx(struct ieee80211_hw *hw, struct sk_buff *skb) rcu_read_unlock(); } -EXPORT_SYMBOL(__ieee80211_rx); +EXPORT_SYMBOL(ieee80211_rx); /* This is a version of the rx handler that can be called from hard irq * context. Post the skb on the queue and schedule the tasklet */ -- cgit v1.2.3 From 06e4da268c0e8f3b8408403d65e47d2885a78ff2 Mon Sep 17 00:00:00 2001 From: Gábor Stefanik Date: Wed, 26 Aug 2009 20:51:26 +0200 Subject: ssb: Implement PMU LDO control and use it in b43 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Implement the "PMU LDO set voltage" and "PMU LDO PA ref enable" functions, and use them during LP-PHY baseband init in b43. Signed-off-by: Gábor Stefanik Signed-off-by: John W. Linville --- drivers/net/wireless/b43/phy_lp.c | 10 +--- drivers/ssb/driver_chipcommon_pmu.c | 94 +++++++++++++++++++++++++++++++ include/linux/ssb/ssb_driver_chipcommon.h | 10 ++++ 3 files changed, 107 insertions(+), 7 deletions(-) (limited to 'include') diff --git a/drivers/net/wireless/b43/phy_lp.c b/drivers/net/wireless/b43/phy_lp.c index 1a57d3390e92..80f245c04042 100644 --- a/drivers/net/wireless/b43/phy_lp.c +++ b/drivers/net/wireless/b43/phy_lp.c @@ -234,19 +234,15 @@ static void lpphy_baseband_rev0_1_init(struct b43_wldev *dev) if ((bus->sprom.boardflags_lo & B43_BFL_FEM) && ((b43_current_band(dev->wl) == IEEE80211_BAND_5GHZ) || (bus->sprom.boardflags_hi & B43_BFH_PAREF))) { - /* TODO: - * Set the LDO voltage to 0x0028 - FIXME: What is this? - * Call sb_pmu_set_ldo_voltage with 4 and the LDO voltage - * as arguments - * Call sb_pmu_paref_ldo_enable with argument TRUE - */ + ssb_pmu_set_ldo_voltage(&bus->chipco, LDO_PAREF, 0x28); + ssb_pmu_set_ldo_paref(&bus->chipco, true); if (dev->phy.rev == 0) { b43_phy_maskset(dev, B43_LPPHY_LP_RF_SIGNAL_LUT, 0xFFCF, 0x0010); } b43_lptab_write(dev, B43_LPTAB16(11, 7), 60); } else { - //TODO: Call ssb_pmu_paref_ldo_enable with argument FALSE + ssb_pmu_set_ldo_paref(&bus->chipco, false); b43_phy_maskset(dev, B43_LPPHY_LP_RF_SIGNAL_LUT, 0xFFCF, 0x0020); b43_lptab_write(dev, B43_LPTAB16(11, 7), 100); diff --git a/drivers/ssb/driver_chipcommon_pmu.c b/drivers/ssb/driver_chipcommon_pmu.c index 4aaddeec55a2..64abd11f6fbb 100644 --- a/drivers/ssb/driver_chipcommon_pmu.c +++ b/drivers/ssb/driver_chipcommon_pmu.c @@ -28,6 +28,21 @@ static void ssb_chipco_pll_write(struct ssb_chipcommon *cc, chipco_write32(cc, SSB_CHIPCO_PLLCTL_DATA, value); } +static void ssb_chipco_regctl_maskset(struct ssb_chipcommon *cc, + u32 offset, u32 mask, u32 set) +{ + u32 value; + + chipco_read32(cc, SSB_CHIPCO_REGCTL_ADDR); + chipco_write32(cc, SSB_CHIPCO_REGCTL_ADDR, offset); + chipco_read32(cc, SSB_CHIPCO_REGCTL_ADDR); + value = chipco_read32(cc, SSB_CHIPCO_REGCTL_DATA); + value &= mask; + value |= set; + chipco_write32(cc, SSB_CHIPCO_REGCTL_DATA, value); + chipco_read32(cc, SSB_CHIPCO_REGCTL_DATA); +} + struct pmu0_plltab_entry { u16 freq; /* Crystal frequency in kHz.*/ u8 xf; /* Crystal frequency value for PMU control */ @@ -506,3 +521,82 @@ void ssb_pmu_init(struct ssb_chipcommon *cc) ssb_pmu_pll_init(cc); ssb_pmu_resources_init(cc); } + +void ssb_pmu_set_ldo_voltage(struct ssb_chipcommon *cc, + enum ssb_pmu_ldo_volt_id id, u32 voltage) +{ + struct ssb_bus *bus = cc->dev->bus; + u32 addr, shift, mask; + + switch (bus->chip_id) { + case 0x4328: + case 0x5354: + switch (id) { + case LDO_VOLT1: + addr = 2; + shift = 25; + mask = 0xF; + break; + case LDO_VOLT2: + addr = 3; + shift = 1; + mask = 0xF; + break; + case LDO_VOLT3: + addr = 3; + shift = 9; + mask = 0xF; + break; + case LDO_PAREF: + addr = 3; + shift = 17; + mask = 0x3F; + break; + default: + SSB_WARN_ON(1); + return; + } + break; + case 0x4312: + if (SSB_WARN_ON(id != LDO_PAREF)) + return; + addr = 0; + shift = 21; + mask = 0x3F; + break; + default: + return; + } + + ssb_chipco_regctl_maskset(cc, addr, ~(mask << shift), + (voltage & mask) << shift); +} + +void ssb_pmu_set_ldo_paref(struct ssb_chipcommon *cc, bool on) +{ + struct ssb_bus *bus = cc->dev->bus; + int ldo; + + switch (bus->chip_id) { + case 0x4312: + ldo = SSB_PMURES_4312_PA_REF_LDO; + break; + case 0x4328: + ldo = SSB_PMURES_4328_PA_REF_LDO; + break; + case 0x5354: + ldo = SSB_PMURES_5354_PA_REF_LDO; + break; + default: + return; + } + + if (on) + chipco_set32(cc, SSB_CHIPCO_PMU_MINRES_MSK, 1 << ldo); + else + chipco_mask32(cc, SSB_CHIPCO_PMU_MINRES_MSK, ~(1 << ldo)); + chipco_read32(cc, SSB_CHIPCO_PMU_MINRES_MSK); //SPEC FIXME found via mmiotrace - dummy read? +} + +EXPORT_SYMBOL(ssb_pmu_set_ldo_voltage); +EXPORT_SYMBOL(ssb_pmu_set_ldo_paref); diff --git a/include/linux/ssb/ssb_driver_chipcommon.h b/include/linux/ssb/ssb_driver_chipcommon.h index d3b1d18922f2..4e27acf0a92f 100644 --- a/include/linux/ssb/ssb_driver_chipcommon.h +++ b/include/linux/ssb/ssb_driver_chipcommon.h @@ -629,5 +629,15 @@ extern int ssb_chipco_serial_init(struct ssb_chipcommon *cc, /* PMU support */ extern void ssb_pmu_init(struct ssb_chipcommon *cc); +enum ssb_pmu_ldo_volt_id { + LDO_PAREF = 0, + LDO_VOLT1, + LDO_VOLT2, + LDO_VOLT3, +}; + +void ssb_pmu_set_ldo_voltage(struct ssb_chipcommon *cc, + enum ssb_pmu_ldo_volt_id id, u32 voltage); +void ssb_pmu_set_ldo_paref(struct ssb_chipcommon *cc, bool on); #endif /* LINUX_SSB_CHIPCO_H_ */ -- cgit v1.2.3 From b24aad44438d5bc21cbbfb94a99d9bf710d8295b Mon Sep 17 00:00:00 2001 From: Bob Moore Date: Fri, 24 Jul 2009 13:30:17 +0800 Subject: ACPICA: Split large ACPI table header Split out the non-acpi-defined ACPI tables into the existing (but empty) actbl2.h file. Preparation for new ACPI 4.0 tables. Signed-off-by: Bob Moore Signed-off-by: Lin Ming Signed-off-by: Len Brown --- include/acpi/actbl.h | 35 +-- include/acpi/actbl1.h | 553 +---------------------------------------------- include/acpi/actbl2.h | 585 ++++++++++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 611 insertions(+), 562 deletions(-) create mode 100644 include/acpi/actbl2.h (limited to 'include') diff --git a/include/acpi/actbl.h b/include/acpi/actbl.h index 0649a5670026..55fcfc6725b2 100644 --- a/include/acpi/actbl.h +++ b/include/acpi/actbl.h @@ -44,6 +44,19 @@ #ifndef __ACTBL_H__ #define __ACTBL_H__ +/******************************************************************************* + * + * Fundamental ACPI tables + * + * This file contains definitions for the ACPI tables that are directly consumed + * by ACPICA. All other tables are consumed by the OS-dependent ACPI-related + * device drivers and other OS support code. + * + * The RSDP and FACS do not use the common ACPI table header. All other ACPI + * tables use the header. + * + ******************************************************************************/ + /* * Values for description table header signatures. Useful because they make * it more difficult to inadvertently type in the wrong signature. @@ -65,11 +78,6 @@ #pragma pack(1) /* - * These are the ACPI tables that are directly consumed by the subsystem. - * - * The RSDP and FACS do not use the common ACPI table header. All other ACPI - * tables use the header. - * * Note about bitfields: The u8 type is used for bitfields in ACPI tables. * This is the only type that is even remotely portable. Anything else is not * portable, so do not use any other bitfield types. @@ -77,9 +85,8 @@ /******************************************************************************* * - * ACPI Table Header. This common header is used by all tables except the - * RSDP and FACS. The define is used for direct inclusion of header into - * other ACPI tables + * Master ACPI Table Header. This common header is used by all ACPI tables + * except the RSDP and FACS. * ******************************************************************************/ @@ -95,13 +102,16 @@ struct acpi_table_header { u32 asl_compiler_revision; /* ASL compiler version */ }; -/* +/******************************************************************************* + * * GAS - Generic Address Structure (ACPI 2.0+) * * Note: Since this structure is used in the ACPI tables, it is byte aligned. - * If misalignment is not supported, access to the Address field must be - * performed with care. - */ + * If misaliged access is not supported by the hardware, accesses to the + * 64-bit Address field must be performed with care. + * + ******************************************************************************/ + struct acpi_generic_address { u8 space_id; /* Address space where struct or register exists */ u8 bit_width; /* Size in bits of given register */ @@ -325,5 +335,6 @@ struct acpi_table_desc { */ #include +#include #endif /* __ACTBL_H__ */ diff --git a/include/acpi/actbl1.h b/include/acpi/actbl1.h index ec36693f868c..582af1fcb8f5 100644 --- a/include/acpi/actbl1.h +++ b/include/acpi/actbl1.h @@ -46,41 +46,29 @@ /******************************************************************************* * - * Additional ACPI Tables + * Additional ACPI Tables (1) * * These tables are not consumed directly by the ACPICA subsystem, but are * included here to support device drivers and the AML disassembler. * + * The tables in this file are fully defined within the ACPI specification. + * ******************************************************************************/ /* * Values for description table header signatures. Useful because they make * it more difficult to inadvertently type in the wrong signature. */ -#define ACPI_SIG_ASF "ASF!" /* Alert Standard Format table */ #define ACPI_SIG_BERT "BERT" /* Boot Error Record Table */ -#define ACPI_SIG_BOOT "BOOT" /* Simple Boot Flag Table */ #define ACPI_SIG_CPEP "CPEP" /* Corrected Platform Error Polling table */ -#define ACPI_SIG_DBGP "DBGP" /* Debug Port table */ -#define ACPI_SIG_DMAR "DMAR" /* DMA Remapping table */ #define ACPI_SIG_ECDT "ECDT" /* Embedded Controller Boot Resources Table */ #define ACPI_SIG_EINJ "EINJ" /* Error Injection table */ #define ACPI_SIG_ERST "ERST" /* Error Record Serialization Table */ #define ACPI_SIG_HEST "HEST" /* Hardware Error Source Table */ -#define ACPI_SIG_HPET "HPET" /* High Precision Event Timer table */ -#define ACPI_SIG_IBFT "IBFT" /* i_sCSI Boot Firmware Table */ #define ACPI_SIG_MADT "APIC" /* Multiple APIC Description Table */ -#define ACPI_SIG_MCFG "MCFG" /* PCI Memory Mapped Configuration table */ #define ACPI_SIG_SBST "SBST" /* Smart Battery Specification Table */ -#define ACPI_SIG_SLIC "SLIC" /* Software Licensing Description Table */ #define ACPI_SIG_SLIT "SLIT" /* System Locality Distance Information Table */ -#define ACPI_SIG_SPCR "SPCR" /* Serial Port Console Redirection table */ -#define ACPI_SIG_SPMI "SPMI" /* Server Platform Management Interface table */ #define ACPI_SIG_SRAT "SRAT" /* System Resource Affinity Table */ -#define ACPI_SIG_TCPA "TCPA" /* Trusted Computing Platform Alliance table */ -#define ACPI_SIG_UEFI "UEFI" /* Uefi Boot Optimization Table */ -#define ACPI_SIG_WDAT "WDAT" /* Watchdog Action Table */ -#define ACPI_SIG_WDRT "WDRT" /* Watchdog Resource Table */ /* * All tables must be byte-packed to match the ACPI specification, since @@ -113,115 +101,6 @@ struct acpi_whea_header { u64 mask; /* Bitmask required for this register instruction */ }; -/******************************************************************************* - * - * ASF - Alert Standard Format table (Signature "ASF!") - * - * Conforms to the Alert Standard Format Specification V2.0, 23 April 2003 - * - ******************************************************************************/ - -struct acpi_table_asf { - struct acpi_table_header header; /* Common ACPI table header */ -}; - -/* ASF subtable header */ - -struct acpi_asf_header { - u8 type; - u8 reserved; - u16 length; -}; - -/* Values for Type field above */ - -enum acpi_asf_type { - ACPI_ASF_TYPE_INFO = 0, - ACPI_ASF_TYPE_ALERT = 1, - ACPI_ASF_TYPE_CONTROL = 2, - ACPI_ASF_TYPE_BOOT = 3, - ACPI_ASF_TYPE_ADDRESS = 4, - ACPI_ASF_TYPE_RESERVED = 5 -}; - -/* - * ASF subtables - */ - -/* 0: ASF Information */ - -struct acpi_asf_info { - struct acpi_asf_header header; - u8 min_reset_value; - u8 min_poll_interval; - u16 system_id; - u32 mfg_id; - u8 flags; - u8 reserved2[3]; -}; - -/* 1: ASF Alerts */ - -struct acpi_asf_alert { - struct acpi_asf_header header; - u8 assert_mask; - u8 deassert_mask; - u8 alerts; - u8 data_length; -}; - -struct acpi_asf_alert_data { - u8 address; - u8 command; - u8 mask; - u8 value; - u8 sensor_type; - u8 type; - u8 offset; - u8 source_type; - u8 severity; - u8 sensor_number; - u8 entity; - u8 instance; -}; - -/* 2: ASF Remote Control */ - -struct acpi_asf_remote { - struct acpi_asf_header header; - u8 controls; - u8 data_length; - u16 reserved2; -}; - -struct acpi_asf_control_data { - u8 function; - u8 address; - u8 command; - u8 value; -}; - -/* 3: ASF RMCP Boot Options */ - -struct acpi_asf_rmcp { - struct acpi_asf_header header; - u8 capabilities[7]; - u8 completion_code; - u32 enterprise_id; - u8 command; - u16 parameter; - u16 boot_options; - u16 oem_parameters; -}; - -/* 4: ASF Address */ - -struct acpi_asf_address { - struct acpi_asf_header header; - u8 eprom_address; - u8 devices; -}; - /******************************************************************************* * * BERT - Boot Error Record Table @@ -251,18 +130,6 @@ struct acpi_bert_region { #define ACPI_BERT_MULTIPLE_UNCORRECTABLE (4) #define ACPI_BERT_MULTIPLE_CORRECTABLE (8) -/******************************************************************************* - * - * BOOT - Simple Boot Flag Table - * - ******************************************************************************/ - -struct acpi_table_boot { - struct acpi_table_header header; /* Common ACPI table header */ - u8 cmos_index; /* Index in CMOS RAM for the boot register */ - u8 reserved[3]; -}; - /******************************************************************************* * * CPEP - Corrected Platform Error Polling table @@ -284,123 +151,6 @@ struct acpi_cpep_polling { u32 interval; /* Polling interval (msec) */ }; -/******************************************************************************* - * - * DBGP - Debug Port table - * - ******************************************************************************/ - -struct acpi_table_dbgp { - struct acpi_table_header header; /* Common ACPI table header */ - u8 type; /* 0=full 16550, 1=subset of 16550 */ - u8 reserved[3]; - struct acpi_generic_address debug_port; -}; - -/******************************************************************************* - * - * DMAR - DMA Remapping table - * From "Intel Virtualization Technology for Directed I/O", Sept. 2007 - * - ******************************************************************************/ - -struct acpi_table_dmar { - struct acpi_table_header header; /* Common ACPI table header */ - u8 width; /* Host Address Width */ - u8 flags; - u8 reserved[10]; -}; - -/* Flags */ - -#define ACPI_DMAR_INTR_REMAP (1) - -/* DMAR subtable header */ - -struct acpi_dmar_header { - u16 type; - u16 length; -}; - -/* Values for subtable type in struct acpi_dmar_header */ - -enum acpi_dmar_type { - ACPI_DMAR_TYPE_HARDWARE_UNIT = 0, - ACPI_DMAR_TYPE_RESERVED_MEMORY = 1, - ACPI_DMAR_TYPE_ATSR = 2, - ACPI_DMAR_TYPE_RESERVED = 3 /* 3 and greater are reserved */ -}; - -struct acpi_dmar_device_scope { - u8 entry_type; - u8 length; - u16 reserved; - u8 enumeration_id; - u8 bus; -}; - -/* Values for entry_type in struct acpi_dmar_device_scope */ - -enum acpi_dmar_scope_type { - ACPI_DMAR_SCOPE_TYPE_NOT_USED = 0, - ACPI_DMAR_SCOPE_TYPE_ENDPOINT = 1, - ACPI_DMAR_SCOPE_TYPE_BRIDGE = 2, - ACPI_DMAR_SCOPE_TYPE_IOAPIC = 3, - ACPI_DMAR_SCOPE_TYPE_HPET = 4, - ACPI_DMAR_SCOPE_TYPE_RESERVED = 5 /* 5 and greater are reserved */ -}; - -struct acpi_dmar_pci_path { - u8 dev; - u8 fn; -}; - -/* - * DMAR Sub-tables, correspond to Type in struct acpi_dmar_header - */ - -/* 0: Hardware Unit Definition */ - -struct acpi_dmar_hardware_unit { - struct acpi_dmar_header header; - u8 flags; - u8 reserved; - u16 segment; - u64 address; /* Register Base Address */ -}; - -/* Flags */ - -#define ACPI_DMAR_INCLUDE_ALL (1) - -/* 1: Reserved Memory Defininition */ - -struct acpi_dmar_reserved_memory { - struct acpi_dmar_header header; - u16 reserved; - u16 segment; - u64 base_address; /* 4_k aligned base address */ - u64 end_address; /* 4_k aligned limit address */ -}; - -/* Flags */ - -#define ACPI_DMAR_ALLOW_ALL (1) - - -/* 2: Root Port ATS Capability Reporting Structure */ - -struct acpi_dmar_atsr { - struct acpi_dmar_header header; - u8 flags; - u8 reserved; - u16 segment; -}; - -/* Flags */ - -#define ACPI_DMAR_ALL_PORTS (1) - /******************************************************************************* * * ECDT - Embedded Controller Boot Resources Table @@ -762,119 +512,6 @@ struct acpi_hest_generic { u32 error_status_block_length; }; -/******************************************************************************* - * - * HPET - High Precision Event Timer table - * - ******************************************************************************/ - -struct acpi_table_hpet { - struct acpi_table_header header; /* Common ACPI table header */ - u32 id; /* Hardware ID of event timer block */ - struct acpi_generic_address address; /* Address of event timer block */ - u8 sequence; /* HPET sequence number */ - u16 minimum_tick; /* Main counter min tick, periodic mode */ - u8 flags; -}; - -/*! Flags */ - -#define ACPI_HPET_PAGE_PROTECT (1) /* 00: No page protection */ -#define ACPI_HPET_PAGE_PROTECT_4 (1<<1) /* 01: 4KB page protected */ -#define ACPI_HPET_PAGE_PROTECT_64 (1<<2) /* 02: 64KB page protected */ - -/*! [End] no source code translation !*/ - -/******************************************************************************* - * - * IBFT - Boot Firmware Table - * - ******************************************************************************/ - -struct acpi_table_ibft { - struct acpi_table_header header; /* Common ACPI table header */ - u8 reserved[12]; -}; - -/* IBFT common subtable header */ - -struct acpi_ibft_header { - u8 type; - u8 version; - u16 length; - u8 index; - u8 flags; -}; - -/* Values for Type field above */ - -enum acpi_ibft_type { - ACPI_IBFT_TYPE_NOT_USED = 0, - ACPI_IBFT_TYPE_CONTROL = 1, - ACPI_IBFT_TYPE_INITIATOR = 2, - ACPI_IBFT_TYPE_NIC = 3, - ACPI_IBFT_TYPE_TARGET = 4, - ACPI_IBFT_TYPE_EXTENSIONS = 5, - ACPI_IBFT_TYPE_RESERVED = 6 /* 6 and greater are reserved */ -}; - -/* IBFT subtables */ - -struct acpi_ibft_control { - struct acpi_ibft_header header; - u16 extensions; - u16 initiator_offset; - u16 nic0_offset; - u16 target0_offset; - u16 nic1_offset; - u16 target1_offset; -}; - -struct acpi_ibft_initiator { - struct acpi_ibft_header header; - u8 sns_server[16]; - u8 slp_server[16]; - u8 primary_server[16]; - u8 secondary_server[16]; - u16 name_length; - u16 name_offset; -}; - -struct acpi_ibft_nic { - struct acpi_ibft_header header; - u8 ip_address[16]; - u8 subnet_mask_prefix; - u8 origin; - u8 gateway[16]; - u8 primary_dns[16]; - u8 secondary_dns[16]; - u8 dhcp[16]; - u16 vlan; - u8 mac_address[6]; - u16 pci_address; - u16 name_length; - u16 name_offset; -}; - -struct acpi_ibft_target { - struct acpi_ibft_header header; - u8 target_ip_address[16]; - u16 target_ip_socket; - u8 target_boot_lun[8]; - u8 chap_type; - u8 nic_association; - u16 target_name_length; - u16 target_name_offset; - u16 chap_name_length; - u16 chap_name_offset; - u16 chap_secret_length; - u16 chap_secret_offset; - u16 reverse_chap_name_length; - u16 reverse_chap_name_offset; - u16 reverse_chap_secret_length; - u16 reverse_chap_secret_offset; -}; - /******************************************************************************* * * MADT - Multiple APIC Description Table @@ -1056,27 +693,6 @@ struct acpi_madt_local_x2apic_nmi { #define ACPI_MADT_TRIGGER_RESERVED (2<<2) #define ACPI_MADT_TRIGGER_LEVEL (3<<2) -/******************************************************************************* - * - * MCFG - PCI Memory Mapped Configuration table and sub-table - * - ******************************************************************************/ - -struct acpi_table_mcfg { - struct acpi_table_header header; /* Common ACPI table header */ - u8 reserved[8]; -}; - -/* Subtable */ - -struct acpi_mcfg_allocation { - u64 address; /* Base address, processor-relative */ - u16 pci_segment; /* PCI segment group number */ - u8 start_bus_number; /* Starting PCI Bus number */ - u8 end_bus_number; /* Final PCI Bus number */ - u32 reserved; -}; - /******************************************************************************* * * SBST - Smart Battery Specification Table @@ -1102,59 +718,6 @@ struct acpi_table_slit { u8 entry[1]; /* Real size = localities^2 */ }; -/******************************************************************************* - * - * SPCR - Serial Port Console Redirection table - * - ******************************************************************************/ - -struct acpi_table_spcr { - struct acpi_table_header header; /* Common ACPI table header */ - u8 interface_type; /* 0=full 16550, 1=subset of 16550 */ - u8 reserved[3]; - struct acpi_generic_address serial_port; - u8 interrupt_type; - u8 pc_interrupt; - u32 interrupt; - u8 baud_rate; - u8 parity; - u8 stop_bits; - u8 flow_control; - u8 terminal_type; - u8 reserved1; - u16 pci_device_id; - u16 pci_vendor_id; - u8 pci_bus; - u8 pci_device; - u8 pci_function; - u32 pci_flags; - u8 pci_segment; - u32 reserved2; -}; - -/******************************************************************************* - * - * SPMI - Server Platform Management Interface table - * - ******************************************************************************/ - -struct acpi_table_spmi { - struct acpi_table_header header; /* Common ACPI table header */ - u8 reserved; - u8 interface_type; - u16 spec_revision; /* Version of IPMI */ - u8 interrupt_type; - u8 gpe_number; /* GPE assigned */ - u8 reserved1; - u8 pci_device_flag; - u32 interrupt; - struct acpi_generic_address ipmi_register; - u8 pci_segment; - u8 pci_bus; - u8 pci_device; - u8 pci_function; -}; - /******************************************************************************* * * SRAT - System Resource Affinity Table @@ -1227,116 +790,6 @@ struct acpi_srat_x2apic_cpu_affinity { #define ACPI_SRAT_CPU_ENABLED (1) /* 00: Use affinity structure */ -/******************************************************************************* - * - * TCPA - Trusted Computing Platform Alliance table - * - ******************************************************************************/ - -struct acpi_table_tcpa { - struct acpi_table_header header; /* Common ACPI table header */ - u16 reserved; - u32 max_log_length; /* Maximum length for the event log area */ - u64 log_address; /* Address of the event log area */ -}; - -/******************************************************************************* - * - * UEFI - UEFI Boot optimization Table - * - ******************************************************************************/ - -struct acpi_table_uefi { - struct acpi_table_header header; /* Common ACPI table header */ - u8 identifier[16]; /* UUID identifier */ - u16 data_offset; /* Offset of remaining data in table */ - u8 data; -}; - -/******************************************************************************* - * - * WDAT - Watchdog Action Table - * - ******************************************************************************/ - -struct acpi_table_wdat { - struct acpi_table_header header; /* Common ACPI table header */ - u32 header_length; /* Watchdog Header Length */ - u16 pci_segment; /* PCI Segment number */ - u8 pci_bus; /* PCI Bus number */ - u8 pci_device; /* PCI Device number */ - u8 pci_function; /* PCI Function number */ - u8 reserved[3]; - u32 timer_period; /* Period of one timer count (msec) */ - u32 max_count; /* Maximum counter value supported */ - u32 min_count; /* Minimum counter value */ - u8 flags; - u8 reserved2[3]; - u32 entries; /* Number of watchdog entries that follow */ -}; - -/* WDAT Instruction Entries (actions) */ - -struct acpi_wdat_entry { - struct acpi_whea_header whea_header; /* Common header for WHEA tables */ -}; - -/* Values for Action field above */ - -enum acpi_wdat_actions { - ACPI_WDAT_RESET = 1, - ACPI_WDAT_GET_CURRENT_COUNTDOWN = 4, - ACPI_WDAT_GET_COUNTDOWN = 5, - ACPI_WDAT_SET_COUNTDOWN = 6, - ACPI_WDAT_GET_RUNNING_STATE = 8, - ACPI_WDAT_SET_RUNNING_STATE = 9, - ACPI_WDAT_GET_STOPPED_STATE = 10, - ACPI_WDAT_SET_STOPPED_STATE = 11, - ACPI_WDAT_GET_REBOOT = 16, - ACPI_WDAT_SET_REBOOT = 17, - ACPI_WDAT_GET_SHUTDOWN = 18, - ACPI_WDAT_SET_SHUTDOWN = 19, - ACPI_WDAT_GET_STATUS = 32, - ACPI_WDAT_SET_STATUS = 33, - ACPI_WDAT_ACTION_RESERVED = 34 /* 34 and greater are reserved */ -}; - -/* Values for Instruction field above */ - -enum acpi_wdat_instructions { - ACPI_WDAT_READ_VALUE = 0, - ACPI_WDAT_READ_COUNTDOWN = 1, - ACPI_WDAT_WRITE_VALUE = 2, - ACPI_WDAT_WRITE_COUNTDOWN = 3, - ACPI_WDAT_INSTRUCTION_RESERVED = 4, /* 4 and greater are reserved */ - ACPI_WDAT_PRESERVE_REGISTER = 0x80 /* Except for this value */ -}; - -/******************************************************************************* - * - * WDRT - Watchdog Resource Table - * - ******************************************************************************/ - -struct acpi_table_wdrt { - struct acpi_table_header header; /* Common ACPI table header */ - u32 header_length; /* Watchdog Header Length */ - u8 pci_segment; /* PCI Segment number */ - u8 pci_bus; /* PCI Bus number */ - u8 pci_device; /* PCI Device number */ - u8 pci_function; /* PCI Function number */ - u32 timer_period; /* Period of one timer count (msec) */ - u32 max_count; /* Maximum counter value supported */ - u32 min_count; /* Minimum counter value */ - u8 flags; - u8 reserved[3]; - u32 entries; /* Number of watchdog entries that follow */ -}; - -/* Flags */ - -#define ACPI_WDRT_TIMER_ENABLED (1) /* 00: Timer enabled */ - /* Reset to default packing */ #pragma pack() diff --git a/include/acpi/actbl2.h b/include/acpi/actbl2.h new file mode 100644 index 000000000000..b271aba0e524 --- /dev/null +++ b/include/acpi/actbl2.h @@ -0,0 +1,585 @@ +#ifndef __ACTBL2_H__ +#define __ACTBL2_H__ + +/******************************************************************************* + * + * Additional ACPI Tables (2) + * + * These tables are not consumed directly by the ACPICA subsystem, but are + * included here to support device drivers and the AML disassembler. + * + * The tables in this file are defined by third-party specifications, and are + * not defined directly by the ACPI specification itself. + * + ******************************************************************************/ + +/* + * Values for description table header signatures. Useful because they make + * it more difficult to inadvertently type in the wrong signature. + */ +#define ACPI_SIG_ASF "ASF!" /* Alert Standard Format table */ +#define ACPI_SIG_BOOT "BOOT" /* Simple Boot Flag Table */ +#define ACPI_SIG_DBGP "DBGP" /* Debug Port table */ +#define ACPI_SIG_DMAR "DMAR" /* DMA Remapping table */ +#define ACPI_SIG_HPET "HPET" /* High Precision Event Timer table */ +#define ACPI_SIG_IBFT "IBFT" /* i_sCSI Boot Firmware Table */ +#define ACPI_SIG_MCFG "MCFG" /* PCI Memory Mapped Configuration table */ +#define ACPI_SIG_SLIC "SLIC" /* Software Licensing Description Table */ +#define ACPI_SIG_SPCR "SPCR" /* Serial Port Console Redirection table */ +#define ACPI_SIG_SPMI "SPMI" /* Server Platform Management Interface table */ +#define ACPI_SIG_TCPA "TCPA" /* Trusted Computing Platform Alliance table */ +#define ACPI_SIG_UEFI "UEFI" /* Uefi Boot Optimization Table */ +#define ACPI_SIG_WDAT "WDAT" /* Watchdog Action Table */ +#define ACPI_SIG_WDRT "WDRT" /* Watchdog Resource Table */ + +/* + * All tables must be byte-packed to match the ACPI specification, since + * the tables are provided by the system BIOS. + */ +#pragma pack(1) + +/* + * Note about bitfields: The u8 type is used for bitfields in ACPI tables. + * This is the only type that is even remotely portable. Anything else is not + * portable, so do not use any other bitfield types. + */ + +/******************************************************************************* + * + * ASF - Alert Standard Format table (Signature "ASF!") + * + * Conforms to the Alert Standard Format Specification V2.0, 23 April 2003 + * + ******************************************************************************/ + +struct acpi_table_asf { + struct acpi_table_header header; /* Common ACPI table header */ +}; + +/* ASF subtable header */ + +struct acpi_asf_header { + u8 type; + u8 reserved; + u16 length; +}; + +/* Values for Type field above */ + +enum acpi_asf_type { + ACPI_ASF_TYPE_INFO = 0, + ACPI_ASF_TYPE_ALERT = 1, + ACPI_ASF_TYPE_CONTROL = 2, + ACPI_ASF_TYPE_BOOT = 3, + ACPI_ASF_TYPE_ADDRESS = 4, + ACPI_ASF_TYPE_RESERVED = 5 +}; + +/* + * ASF subtables + */ + +/* 0: ASF Information */ + +struct acpi_asf_info { + struct acpi_asf_header header; + u8 min_reset_value; + u8 min_poll_interval; + u16 system_id; + u32 mfg_id; + u8 flags; + u8 reserved2[3]; +}; + +/* 1: ASF Alerts */ + +struct acpi_asf_alert { + struct acpi_asf_header header; + u8 assert_mask; + u8 deassert_mask; + u8 alerts; + u8 data_length; +}; + +struct acpi_asf_alert_data { + u8 address; + u8 command; + u8 mask; + u8 value; + u8 sensor_type; + u8 type; + u8 offset; + u8 source_type; + u8 severity; + u8 sensor_number; + u8 entity; + u8 instance; +}; + +/* 2: ASF Remote Control */ + +struct acpi_asf_remote { + struct acpi_asf_header header; + u8 controls; + u8 data_length; + u16 reserved2; +}; + +struct acpi_asf_control_data { + u8 function; + u8 address; + u8 command; + u8 value; +}; + +/* 3: ASF RMCP Boot Options */ + +struct acpi_asf_rmcp { + struct acpi_asf_header header; + u8 capabilities[7]; + u8 completion_code; + u32 enterprise_id; + u8 command; + u16 parameter; + u16 boot_options; + u16 oem_parameters; +}; + +/* 4: ASF Address */ + +struct acpi_asf_address { + struct acpi_asf_header header; + u8 eprom_address; + u8 devices; +}; + +/******************************************************************************* + * + * BOOT - Simple Boot Flag Table + * + ******************************************************************************/ + +struct acpi_table_boot { + struct acpi_table_header header; /* Common ACPI table header */ + u8 cmos_index; /* Index in CMOS RAM for the boot register */ + u8 reserved[3]; +}; + +/******************************************************************************* + * + * DBGP - Debug Port table + * + ******************************************************************************/ + +struct acpi_table_dbgp { + struct acpi_table_header header; /* Common ACPI table header */ + u8 type; /* 0=full 16550, 1=subset of 16550 */ + u8 reserved[3]; + struct acpi_generic_address debug_port; +}; + +/******************************************************************************* + * + * DMAR - DMA Remapping table + * From "Intel Virtualization Technology for Directed I/O", Sept. 2007 + * + ******************************************************************************/ + +struct acpi_table_dmar { + struct acpi_table_header header; /* Common ACPI table header */ + u8 width; /* Host Address Width */ + u8 flags; + u8 reserved[10]; +}; + +/* Flags */ + +#define ACPI_DMAR_INTR_REMAP (1) + +/* DMAR subtable header */ + +struct acpi_dmar_header { + u16 type; + u16 length; +}; + +/* Values for subtable type in struct acpi_dmar_header */ + +enum acpi_dmar_type { + ACPI_DMAR_TYPE_HARDWARE_UNIT = 0, + ACPI_DMAR_TYPE_RESERVED_MEMORY = 1, + ACPI_DMAR_TYPE_ATSR = 2, + ACPI_DMAR_TYPE_RESERVED = 3 /* 3 and greater are reserved */ +}; + +struct acpi_dmar_device_scope { + u8 entry_type; + u8 length; + u16 reserved; + u8 enumeration_id; + u8 bus; +}; + +/* Values for entry_type in struct acpi_dmar_device_scope */ + +enum acpi_dmar_scope_type { + ACPI_DMAR_SCOPE_TYPE_NOT_USED = 0, + ACPI_DMAR_SCOPE_TYPE_ENDPOINT = 1, + ACPI_DMAR_SCOPE_TYPE_BRIDGE = 2, + ACPI_DMAR_SCOPE_TYPE_IOAPIC = 3, + ACPI_DMAR_SCOPE_TYPE_HPET = 4, + ACPI_DMAR_SCOPE_TYPE_RESERVED = 5 /* 5 and greater are reserved */ +}; + +struct acpi_dmar_pci_path { + u8 dev; + u8 fn; +}; + +/* + * DMAR Sub-tables, correspond to Type in struct acpi_dmar_header + */ + +/* 0: Hardware Unit Definition */ + +struct acpi_dmar_hardware_unit { + struct acpi_dmar_header header; + u8 flags; + u8 reserved; + u16 segment; + u64 address; /* Register Base Address */ +}; + +/* Flags */ + +#define ACPI_DMAR_INCLUDE_ALL (1) + +/* 1: Reserved Memory Defininition */ + +struct acpi_dmar_reserved_memory { + struct acpi_dmar_header header; + u16 reserved; + u16 segment; + u64 base_address; /* 4_k aligned base address */ + u64 end_address; /* 4_k aligned limit address */ +}; + +/* Flags */ + +#define ACPI_DMAR_ALLOW_ALL (1) + +/* 2: Root Port ATS Capability Reporting Structure */ + +struct acpi_dmar_atsr { + struct acpi_dmar_header header; + u8 flags; + u8 reserved; + u16 segment; +}; + +/* Flags */ + +#define ACPI_DMAR_ALL_PORTS (1) + +/******************************************************************************* + * + * HPET - High Precision Event Timer table + * + ******************************************************************************/ + +struct acpi_table_hpet { + struct acpi_table_header header; /* Common ACPI table header */ + u32 id; /* Hardware ID of event timer block */ + struct acpi_generic_address address; /* Address of event timer block */ + u8 sequence; /* HPET sequence number */ + u16 minimum_tick; /* Main counter min tick, periodic mode */ + u8 flags; +}; + +/*! Flags */ + +#define ACPI_HPET_PAGE_PROTECT (1) /* 00: No page protection */ +#define ACPI_HPET_PAGE_PROTECT_4 (1<<1) /* 01: 4KB page protected */ +#define ACPI_HPET_PAGE_PROTECT_64 (1<<2) /* 02: 64KB page protected */ + +/*! [End] no source code translation !*/ + +/******************************************************************************* + * + * IBFT - Boot Firmware Table + * + ******************************************************************************/ + +struct acpi_table_ibft { + struct acpi_table_header header; /* Common ACPI table header */ + u8 reserved[12]; +}; + +/* IBFT common subtable header */ + +struct acpi_ibft_header { + u8 type; + u8 version; + u16 length; + u8 index; + u8 flags; +}; + +/* Values for Type field above */ + +enum acpi_ibft_type { + ACPI_IBFT_TYPE_NOT_USED = 0, + ACPI_IBFT_TYPE_CONTROL = 1, + ACPI_IBFT_TYPE_INITIATOR = 2, + ACPI_IBFT_TYPE_NIC = 3, + ACPI_IBFT_TYPE_TARGET = 4, + ACPI_IBFT_TYPE_EXTENSIONS = 5, + ACPI_IBFT_TYPE_RESERVED = 6 /* 6 and greater are reserved */ +}; + +/* IBFT subtables */ + +struct acpi_ibft_control { + struct acpi_ibft_header header; + u16 extensions; + u16 initiator_offset; + u16 nic0_offset; + u16 target0_offset; + u16 nic1_offset; + u16 target1_offset; +}; + +struct acpi_ibft_initiator { + struct acpi_ibft_header header; + u8 sns_server[16]; + u8 slp_server[16]; + u8 primary_server[16]; + u8 secondary_server[16]; + u16 name_length; + u16 name_offset; +}; + +struct acpi_ibft_nic { + struct acpi_ibft_header header; + u8 ip_address[16]; + u8 subnet_mask_prefix; + u8 origin; + u8 gateway[16]; + u8 primary_dns[16]; + u8 secondary_dns[16]; + u8 dhcp[16]; + u16 vlan; + u8 mac_address[6]; + u16 pci_address; + u16 name_length; + u16 name_offset; +}; + +struct acpi_ibft_target { + struct acpi_ibft_header header; + u8 target_ip_address[16]; + u16 target_ip_socket; + u8 target_boot_lun[8]; + u8 chap_type; + u8 nic_association; + u16 target_name_length; + u16 target_name_offset; + u16 chap_name_length; + u16 chap_name_offset; + u16 chap_secret_length; + u16 chap_secret_offset; + u16 reverse_chap_name_length; + u16 reverse_chap_name_offset; + u16 reverse_chap_secret_length; + u16 reverse_chap_secret_offset; +}; + +/******************************************************************************* + * + * MCFG - PCI Memory Mapped Configuration table and sub-table + * + ******************************************************************************/ + +struct acpi_table_mcfg { + struct acpi_table_header header; /* Common ACPI table header */ + u8 reserved[8]; +}; + +/* Subtable */ + +struct acpi_mcfg_allocation { + u64 address; /* Base address, processor-relative */ + u16 pci_segment; /* PCI segment group number */ + u8 start_bus_number; /* Starting PCI Bus number */ + u8 end_bus_number; /* Final PCI Bus number */ + u32 reserved; +}; + +/******************************************************************************* + * + * SPCR - Serial Port Console Redirection table + * + ******************************************************************************/ + +struct acpi_table_spcr { + struct acpi_table_header header; /* Common ACPI table header */ + u8 interface_type; /* 0=full 16550, 1=subset of 16550 */ + u8 reserved[3]; + struct acpi_generic_address serial_port; + u8 interrupt_type; + u8 pc_interrupt; + u32 interrupt; + u8 baud_rate; + u8 parity; + u8 stop_bits; + u8 flow_control; + u8 terminal_type; + u8 reserved1; + u16 pci_device_id; + u16 pci_vendor_id; + u8 pci_bus; + u8 pci_device; + u8 pci_function; + u32 pci_flags; + u8 pci_segment; + u32 reserved2; +}; + +/******************************************************************************* + * + * SPMI - Server Platform Management Interface table + * + ******************************************************************************/ + +struct acpi_table_spmi { + struct acpi_table_header header; /* Common ACPI table header */ + u8 reserved; + u8 interface_type; + u16 spec_revision; /* Version of IPMI */ + u8 interrupt_type; + u8 gpe_number; /* GPE assigned */ + u8 reserved1; + u8 pci_device_flag; + u32 interrupt; + struct acpi_generic_address ipmi_register; + u8 pci_segment; + u8 pci_bus; + u8 pci_device; + u8 pci_function; +}; + +/******************************************************************************* + * + * TCPA - Trusted Computing Platform Alliance table + * + ******************************************************************************/ + +struct acpi_table_tcpa { + struct acpi_table_header header; /* Common ACPI table header */ + u16 reserved; + u32 max_log_length; /* Maximum length for the event log area */ + u64 log_address; /* Address of the event log area */ +}; + +/******************************************************************************* + * + * UEFI - UEFI Boot optimization Table + * + ******************************************************************************/ + +struct acpi_table_uefi { + struct acpi_table_header header; /* Common ACPI table header */ + u8 identifier[16]; /* UUID identifier */ + u16 data_offset; /* Offset of remaining data in table */ + u8 data; +}; + +/******************************************************************************* + * + * WDAT - Watchdog Action Table + * + ******************************************************************************/ + +struct acpi_table_wdat { + struct acpi_table_header header; /* Common ACPI table header */ + u32 header_length; /* Watchdog Header Length */ + u16 pci_segment; /* PCI Segment number */ + u8 pci_bus; /* PCI Bus number */ + u8 pci_device; /* PCI Device number */ + u8 pci_function; /* PCI Function number */ + u8 reserved[3]; + u32 timer_period; /* Period of one timer count (msec) */ + u32 max_count; /* Maximum counter value supported */ + u32 min_count; /* Minimum counter value */ + u8 flags; + u8 reserved2[3]; + u32 entries; /* Number of watchdog entries that follow */ +}; + +/* WDAT Instruction Entries (actions) */ + +struct acpi_wdat_entry { + struct acpi_whea_header whea_header; /* Common header for WHEA tables */ +}; + +/* Values for Action field above */ + +enum acpi_wdat_actions { + ACPI_WDAT_RESET = 1, + ACPI_WDAT_GET_CURRENT_COUNTDOWN = 4, + ACPI_WDAT_GET_COUNTDOWN = 5, + ACPI_WDAT_SET_COUNTDOWN = 6, + ACPI_WDAT_GET_RUNNING_STATE = 8, + ACPI_WDAT_SET_RUNNING_STATE = 9, + ACPI_WDAT_GET_STOPPED_STATE = 10, + ACPI_WDAT_SET_STOPPED_STATE = 11, + ACPI_WDAT_GET_REBOOT = 16, + ACPI_WDAT_SET_REBOOT = 17, + ACPI_WDAT_GET_SHUTDOWN = 18, + ACPI_WDAT_SET_SHUTDOWN = 19, + ACPI_WDAT_GET_STATUS = 32, + ACPI_WDAT_SET_STATUS = 33, + ACPI_WDAT_ACTION_RESERVED = 34 /* 34 and greater are reserved */ +}; + +/* Values for Instruction field above */ + +enum acpi_wdat_instructions { + ACPI_WDAT_READ_VALUE = 0, + ACPI_WDAT_READ_COUNTDOWN = 1, + ACPI_WDAT_WRITE_VALUE = 2, + ACPI_WDAT_WRITE_COUNTDOWN = 3, + ACPI_WDAT_INSTRUCTION_RESERVED = 4, /* 4 and greater are reserved */ + ACPI_WDAT_PRESERVE_REGISTER = 0x80 /* Except for this value */ +}; + +/******************************************************************************* + * + * WDRT - Watchdog Resource Table + * + ******************************************************************************/ + +struct acpi_table_wdrt { + struct acpi_table_header header; /* Common ACPI table header */ + u32 header_length; /* Watchdog Header Length */ + u8 pci_segment; /* PCI Segment number */ + u8 pci_bus; /* PCI Bus number */ + u8 pci_device; /* PCI Device number */ + u8 pci_function; /* PCI Function number */ + u32 timer_period; /* Period of one timer count (msec) */ + u32 max_count; /* Maximum counter value supported */ + u32 min_count; /* Minimum counter value */ + u8 flags; + u8 reserved[3]; + u32 entries; /* Number of watchdog entries that follow */ +}; + +/* Flags */ + +#define ACPI_WDRT_TIMER_ENABLED (1) /* 00: Timer enabled */ + +/* Reset to default packing */ + +#pragma pack() + +#endif /* __ACTBL2_H__ */ -- cgit v1.2.3 From 6e2d5ebd0d36199920676fdceaff4f4bfe66297b Mon Sep 17 00:00:00 2001 From: Bob Moore Date: Mon, 27 Jul 2009 10:53:00 +0800 Subject: ACPICA: ACPI 4: Update headers for new and changed ACPI tables. Add IVRS,MSCT,UEFI,WAET,WDAT. Updated several existing tables for ACPI 4.0-related changes. Added document references for all tables not defined in ACPI spec. Signed-off-by: Bob Moore Signed-off-by: Lin Ming Signed-off-by: Len Brown --- include/acpi/actbl.h | 30 +++-- include/acpi/actbl1.h | 339 +++++++++++++++++++++++++++++++++----------------- include/acpi/actbl2.h | 339 +++++++++++++++++++++++++++++++++++++++++++++----- 3 files changed, 557 insertions(+), 151 deletions(-) (limited to 'include') diff --git a/include/acpi/actbl.h b/include/acpi/actbl.h index 55fcfc6725b2..1b6587952604 100644 --- a/include/acpi/actbl.h +++ b/include/acpi/actbl.h @@ -58,8 +58,9 @@ ******************************************************************************/ /* - * Values for description table header signatures. Useful because they make - * it more difficult to inadvertently type in the wrong signature. + * Values for description table header signatures for tables defined in this + * file. Useful because they make it more difficult to inadvertently type in + * the wrong signature. */ #define ACPI_SIG_DSDT "DSDT" /* Differentiated System Description Table */ #define ACPI_SIG_FADT "FACP" /* Fixed ACPI Description Table */ @@ -123,6 +124,7 @@ struct acpi_generic_address { /******************************************************************************* * * RSDP - Root System Description Pointer (Signature is "RSD PTR ") + * Version 2 * ******************************************************************************/ @@ -143,6 +145,7 @@ struct acpi_table_rsdp { /******************************************************************************* * * RSDT/XSDT - Root System Description Tables + * Version 1 (both) * ******************************************************************************/ @@ -176,23 +179,24 @@ struct acpi_table_facs { u8 reserved1[24]; /* Reserved, must be zero */ }; -/* global_lock flags */ +/* Masks for global_lock flag field above */ #define ACPI_GLOCK_PENDING (1) /* 00: Pending global lock ownership */ #define ACPI_GLOCK_OWNED (1<<1) /* 01: Global lock is owned */ -/* Flags */ +/* Masks for Flags field above */ #define ACPI_FACS_S4_BIOS_PRESENT (1) /* 00: S4BIOS support is present */ #define ACPI_FACS_64BIT_WAKE (1<<1) /* 01: 64-bit wake vector supported (ACPI 4.0) */ -/* ospm_flags */ +/* Masks for ospm_flags field above */ #define ACPI_FACS_64BIT_ENVIRONMENT (1) /* 00: 64-bit wake environment is required (ACPI 4.0) */ /******************************************************************************* * * FADT - Fixed ACPI Description Table (Signature "FACP") + * Version 4 * ******************************************************************************/ @@ -253,7 +257,7 @@ struct acpi_table_fadt { struct acpi_generic_address xgpe1_block; /* 64-bit Extended General Purpose Event 1 Reg Blk address */ }; -/* FADT Boot Architecture Flags (boot_flags) */ +/* Masks for FADT Boot Architecture Flags (boot_flags) */ #define ACPI_FADT_LEGACY_DEVICES (1) /* 00: [V2] System has LPC or ISA bus devices */ #define ACPI_FADT_8042 (1<<1) /* 01: [V3] System has an 8042 controller on port 60/64 */ @@ -263,7 +267,7 @@ struct acpi_table_fadt { #define FADT2_REVISION_ID 3 -/* FADT flags */ +/* Masks for FADT flags */ #define ACPI_FADT_WBINVD (1) /* 00: [V1] The wbinvd instruction works properly */ #define ACPI_FADT_WBINVD_FLUSH (1<<1) /* 01: [V1] wbinvd flushes but does not invalidate caches */ @@ -286,7 +290,7 @@ struct acpi_table_fadt { #define ACPI_FADT_APIC_CLUSTER (1<<18) /* 18: [V4] All local APICs must use cluster model (ACPI 3.0) */ #define ACPI_FADT_APIC_PHYSICAL (1<<19) /* 19: [V4] All local x_aPICs must use physical dest mode (ACPI 3.0) */ -/* FADT Prefered Power Management Profiles */ +/* Values for preferred_profile (Prefered Power Management Profiles) */ enum acpi_prefered_pm_profiles { PM_UNSPECIFIED = 0, @@ -304,14 +308,16 @@ enum acpi_prefered_pm_profiles { #define ACPI_FADT_OFFSET(f) (u8) ACPI_OFFSET (struct acpi_table_fadt, f) +/* + * Internal table-related structures + */ union acpi_name_union { u32 integer; char ascii[4]; }; -/* - * Internal ACPI Table Descriptor. One per ACPI table - */ +/* Internal ACPI Table Descriptor. One per ACPI table. */ + struct acpi_table_desc { acpi_physical_address address; struct acpi_table_header *pointer; @@ -321,7 +327,7 @@ struct acpi_table_desc { u8 flags; }; -/* Flags for above */ +/* Masks for Flags field above */ #define ACPI_TABLE_ORIGIN_UNKNOWN (0) #define ACPI_TABLE_ORIGIN_MAPPED (1) diff --git a/include/acpi/actbl1.h b/include/acpi/actbl1.h index 582af1fcb8f5..0417f2abc44b 100644 --- a/include/acpi/actbl1.h +++ b/include/acpi/actbl1.h @@ -56,8 +56,9 @@ ******************************************************************************/ /* - * Values for description table header signatures. Useful because they make - * it more difficult to inadvertently type in the wrong signature. + * Values for description table header signatures for tables defined in this + * file. Useful because they make it more difficult to inadvertently type in + * the wrong signature. */ #define ACPI_SIG_BERT "BERT" /* Boot Error Record Table */ #define ACPI_SIG_CPEP "CPEP" /* Corrected Platform Error Polling table */ @@ -66,6 +67,7 @@ #define ACPI_SIG_ERST "ERST" /* Error Record Serialization Table */ #define ACPI_SIG_HEST "HEST" /* Hardware Error Source Table */ #define ACPI_SIG_MADT "APIC" /* Multiple APIC Description Table */ +#define ACPI_SIG_MSCT "MSCT" /* Maximum System Characteristics Table */ #define ACPI_SIG_SBST "SBST" /* Smart Battery Specification Table */ #define ACPI_SIG_SLIT "SLIT" /* System Locality Distance Information Table */ #define ACPI_SIG_SRAT "SRAT" /* System Resource Affinity Table */ @@ -82,14 +84,20 @@ * portable, so do not use any other bitfield types. */ -/* Common Subtable header (used in MADT, SRAT, etc.) */ +/******************************************************************************* + * + * Common subtable headers + * + ******************************************************************************/ + +/* Generic subtable header (used in MADT, SRAT, etc.) */ struct acpi_subtable_header { u8 type; u8 length; }; -/* Common Subtable header for WHEA tables (EINJ, ERST, WDAT) */ +/* Subtable header for WHEA tables (EINJ, ERST, WDAT) */ struct acpi_whea_header { u8 action; @@ -103,7 +111,8 @@ struct acpi_whea_header { /******************************************************************************* * - * BERT - Boot Error Record Table + * BERT - Boot Error Record Table (ACPI 4.0) + * Version 1 * ******************************************************************************/ @@ -113,26 +122,43 @@ struct acpi_table_bert { u64 address; /* Physical addresss of the error region */ }; -/* Boot Error Region */ +/* Boot Error Region (not a subtable, pointed to by Address field above) */ struct acpi_bert_region { - u32 block_status; - u32 raw_data_offset; - u32 raw_data_length; - u32 data_length; - u32 error_severity; + u32 block_status; /* Type of error information */ + u32 raw_data_offset; /* Offset to raw error data */ + u32 raw_data_length; /* Length of raw error data */ + u32 data_length; /* Length of generic error data */ + u32 error_severity; /* Severity code */ }; -/* block_status Flags */ +/* Values for block_status flags above */ #define ACPI_BERT_UNCORRECTABLE (1) -#define ACPI_BERT_CORRECTABLE (2) -#define ACPI_BERT_MULTIPLE_UNCORRECTABLE (4) -#define ACPI_BERT_MULTIPLE_CORRECTABLE (8) +#define ACPI_BERT_CORRECTABLE (1<<1) +#define ACPI_BERT_MULTIPLE_UNCORRECTABLE (1<<2) +#define ACPI_BERT_MULTIPLE_CORRECTABLE (1<<3) +#define ACPI_BERT_ERROR_ENTRY_COUNT (0xFF<<4) /* 8 bits, error count */ + +/* Values for error_severity above */ + +enum acpi_bert_error_severity { + ACPI_BERT_ERROR_CORRECTABLE = 0, + ACPI_BERT_ERROR_FATAL = 1, + ACPI_BERT_ERROR_CORRECTED = 2, + ACPI_BERT_ERROR_NONE = 3, + ACPI_BERT_ERROR_RESERVED = 4 /* 4 and greater are reserved */ +}; + +/* + * Note: The generic error data that follows the error_severity field above + * uses the struct acpi_hest_generic_data defined under the HEST table below + */ /******************************************************************************* * - * CPEP - Corrected Platform Error Polling table + * CPEP - Corrected Platform Error Polling table (ACPI 4.0) + * Version 1 * ******************************************************************************/ @@ -144,8 +170,7 @@ struct acpi_table_cpep { /* Subtable */ struct acpi_cpep_polling { - u8 type; - u8 length; + struct acpi_subtable_header header; u8 id; /* Processor ID */ u8 eid; /* Processor EID */ u32 interval; /* Polling interval (msec) */ @@ -154,6 +179,7 @@ struct acpi_cpep_polling { /******************************************************************************* * * ECDT - Embedded Controller Boot Resources Table + * Version 1 * ******************************************************************************/ @@ -168,14 +194,16 @@ struct acpi_table_ecdt { /******************************************************************************* * - * EINJ - Error Injection Table + * EINJ - Error Injection Table (ACPI 4.0) + * Version 1 * ******************************************************************************/ struct acpi_table_einj { struct acpi_table_header header; /* Common ACPI table header */ u32 header_length; - u32 reserved; + u8 flags; + u8 reserved[3]; u32 entries; }; @@ -185,6 +213,10 @@ struct acpi_einj_entry { struct acpi_whea_header whea_header; /* Common header for WHEA tables */ }; +/* Masks for Flags field above */ + +#define ACPI_EINJ_PRESERVE (1) + /* Values for Action field above */ enum acpi_einj_actions { @@ -220,9 +252,34 @@ struct acpi_einj_trigger { u32 entry_count; }; +/* Command status return values */ + +enum acpi_einj_command_status { + ACPI_EINJ_SUCCESS = 0, + ACPI_EINJ_FAILURE = 1, + ACPI_EINJ_INVALID_ACCESS = 2, + ACPI_EINJ_STATUS_RESERVED = 3 /* 3 and greater are reserved */ +}; + +/* Error types returned from ACPI_EINJ_GET_ERROR_TYPE (bitfield) */ + +#define ACPI_EINJ_PROCESSOR_CORRECTABLE (1) +#define ACPI_EINJ_PROCESSOR_UNCORRECTABLE (1<<1) +#define ACPI_EINJ_PROCESSOR_FATAL (1<<2) +#define ACPI_EINJ_MEMORY_CORRECTABLE (1<<3) +#define ACPI_EINJ_MEMORY_UNCORRECTABLE (1<<4) +#define ACPI_EINJ_MEMORY_FATAL (1<<5) +#define ACPI_EINJ_PCIX_CORRECTABLE (1<<6) +#define ACPI_EINJ_PCIX_UNCORRECTABLE (1<<7) +#define ACPI_EINJ_PCIX_FATAL (1<<8) +#define ACPI_EINJ_PLATFORM_CORRECTABLE (1<<9) +#define ACPI_EINJ_PLATFORM_UNCORRECTABLE (1<<10) +#define ACPI_EINJ_PLATFORM_FATAL (1<<11) + /******************************************************************************* * - * ERST - Error Record Serialization Table + * ERST - Error Record Serialization Table (ACPI 4.0) + * Version 1 * ******************************************************************************/ @@ -239,19 +296,23 @@ struct acpi_erst_entry { struct acpi_whea_header whea_header; /* Common header for WHEA tables */ }; +/* Masks for Flags field above */ + +#define ACPI_ERST_PRESERVE (1) + /* Values for Action field above */ enum acpi_erst_actions { - ACPI_ERST_BEGIN_WRITE_OPERATION = 0, - ACPI_ERST_BEGIN_READ_OPERATION = 1, - ACPI_ERST_BETGIN_CLEAR_OPERATION = 2, - ACPI_ERST_END_OPERATION = 3, + ACPI_ERST_BEGIN_WRITE = 0, + ACPI_ERST_BEGIN_READ = 1, + ACPI_ERST_BEGIN_CLEAR = 2, + ACPI_ERST_END = 3, ACPI_ERST_SET_RECORD_OFFSET = 4, ACPI_ERST_EXECUTE_OPERATION = 5, ACPI_ERST_CHECK_BUSY_STATUS = 6, ACPI_ERST_GET_COMMAND_STATUS = 7, - ACPI_ERST_GET_RECORD_IDENTIFIER = 8, - ACPI_ERST_SET_RECORD_IDENTIFIER = 9, + ACPI_ERST_GET_RECORD_ID = 8, + ACPI_ERST_SET_RECORD_ID = 9, ACPI_ERST_GET_RECORD_COUNT = 10, ACPI_ERST_BEGIN_DUMMY_WRIITE = 11, ACPI_ERST_NOT_USED = 12, @@ -286,9 +347,29 @@ enum acpi_erst_instructions { ACPI_ERST_INSTRUCTION_RESERVED = 19 /* 19 and greater are reserved */ }; +/* Command status return values */ + +enum acpi_erst_command_status { + ACPI_ERST_SUCESS = 0, + ACPI_ERST_NO_SPACE = 1, + ACPI_ERST_NOT_AVAILABLE = 2, + ACPI_ERST_FAILURE = 3, + ACPI_ERST_RECORD_EMPTY = 4, + ACPI_ERST_NOT_FOUND = 5, + ACPI_ERST_STATUS_RESERVED = 6 /* 6 and greater are reserved */ +}; + +/* Error Record Serialization Information */ + +struct acpi_erst_info { + u16 signature; /* Should be "ER" */ + u8 data[48]; +}; + /******************************************************************************* * - * HEST - Hardware Error Source Table + * HEST - Hardware Error Source Table (ACPI 4.0) + * Version 1 * ******************************************************************************/ @@ -301,70 +382,49 @@ struct acpi_table_hest { struct acpi_hest_header { u16 type; + u16 source_id; }; /* Values for Type field above for subtables */ enum acpi_hest_types { - ACPI_HEST_TYPE_XPF_MACHINE_CHECK = 0, - ACPI_HEST_TYPE_XPF_CORRECTED_MACHINE_CHECK = 1, - ACPI_HEST_TYPE_XPF_UNUSED = 2, - ACPI_HEST_TYPE_XPF_NON_MASKABLE_INTERRUPT = 3, - ACPI_HEST_TYPE_IPF_CORRECTED_MACHINE_CHECK = 4, - ACPI_HEST_TYPE_IPF_CORRECTED_PLATFORM_ERROR = 5, + ACPI_HEST_TYPE_IA32_CHECK = 0, + ACPI_HEST_TYPE_IA32_CORRECTED_CHECK = 1, + ACPI_HEST_TYPE_IA32_NMI = 2, + ACPI_HEST_TYPE_NOT_USED3 = 3, + ACPI_HEST_TYPE_NOT_USED4 = 4, + ACPI_HEST_TYPE_NOT_USED5 = 5, ACPI_HEST_TYPE_AER_ROOT_PORT = 6, ACPI_HEST_TYPE_AER_ENDPOINT = 7, ACPI_HEST_TYPE_AER_BRIDGE = 8, - ACPI_HEST_TYPE_GENERIC_HARDWARE_ERROR_SOURCE = 9, + ACPI_HEST_TYPE_GENERIC_ERROR = 9, ACPI_HEST_TYPE_RESERVED = 10 /* 10 and greater are reserved */ }; /* - * HEST Sub-subtables + * HEST substructures contained in subtables */ -/* XPF Machine Check Error Bank */ - -struct acpi_hest_xpf_error_bank { +/* + * IA32 Error Bank(s) - Follows the struct acpi_hest_ia_machine_check and + * struct acpi_hest_ia_corrected structures. + */ +struct acpi_hest_ia_error_bank { u8 bank_number; u8 clear_status_on_init; u8 status_format; - u8 config_write_enable; + u8 reserved; u32 control_register; - u64 control_init_data; + u64 control_data; u32 status_register; u32 address_register; u32 misc_register; }; -/* Generic Error Status */ - -struct acpi_hest_generic_status { - u32 block_status; - u32 raw_data_offset; - u32 raw_data_length; - u32 data_length; - u32 error_severity; -}; - -/* Generic Error Data */ - -struct acpi_hest_generic_data { - u8 section_type[16]; - u32 error_severity; - u16 revision; - u8 validation_bits; - u8 flags; - u32 error_data_length; - u8 fru_id[16]; - u8 fru_text[20]; -}; - -/* Common HEST structure for PCI/AER types below (6,7,8) */ +/* Common HEST sub-structure for PCI/AER structures below (6,7,8) */ struct acpi_hest_aer_common { - u16 source_id; - u16 config_write_enable; + u16 reserved1; u8 flags; u8 enabled; u32 records_to_pre_allocate; @@ -373,13 +433,18 @@ struct acpi_hest_aer_common { u16 device; u16 function; u16 device_control; - u16 reserved; + u16 reserved2; u32 uncorrectable_error_mask; u32 uncorrectable_error_severity; u32 correctable_error_mask; u32 advanced_error_capabilities; }; +/* Masks for HEST Flags fields */ + +#define ACPI_HEST_FIRMWARE_FIRST (1) +#define ACPI_HEST_GLOBAL (1<<1) + /* Hardware Error Notification */ struct acpi_hest_notify { @@ -405,71 +470,59 @@ enum acpi_hest_notify_types { ACPI_HEST_NOTIFY_RESERVED = 5 /* 5 and greater are reserved */ }; +/* Values for config_write_enable bitfield above */ + +#define ACPI_HEST_TYPE (1) +#define ACPI_HEST_POLL_INTERVAL (1<<1) +#define ACPI_HEST_POLL_THRESHOLD_VALUE (1<<2) +#define ACPI_HEST_POLL_THRESHOLD_WINDOW (1<<3) +#define ACPI_HEST_ERR_THRESHOLD_VALUE (1<<4) +#define ACPI_HEST_ERR_THRESHOLD_WINDOW (1<<5) + /* * HEST subtables - * - * From WHEA Design Document, 16 May 2007. - * Note: There is no subtable type 2 in this version of the document, - * and there are two different subtable type 3s. */ - /* 0: XPF Machine Check Exception */ +/* 0: IA32 Machine Check Exception */ -struct acpi_hest_xpf_machine_check { +struct acpi_hest_ia_machine_check { struct acpi_hest_header header; - u16 source_id; - u16 config_write_enable; + u16 reserved1; u8 flags; - u8 reserved1; + u8 enabled; u32 records_to_pre_allocate; u32 max_sections_per_record; u64 global_capability_data; u64 global_control_data; u8 num_hardware_banks; - u8 reserved2[7]; + u8 reserved3[7]; }; -/* 1: XPF Corrected Machine Check */ +/* 1: IA32 Corrected Machine Check */ -struct acpi_table_hest_xpf_corrected { +struct acpi_table_hest_ia_corrected { struct acpi_hest_header header; - u16 source_id; - u16 config_write_enable; + u16 reserved1; u8 flags; u8 enabled; u32 records_to_pre_allocate; u32 max_sections_per_record; struct acpi_hest_notify notify; u8 num_hardware_banks; - u8 reserved[3]; + u8 reserved2[3]; }; -/* 3: XPF Non-Maskable Interrupt */ +/* 2: IA32 Non-Maskable Interrupt */ -struct acpi_hest_xpf_nmi { +struct acpi_hest_ia_nmi { struct acpi_hest_header header; - u16 source_id; u32 reserved; u32 records_to_pre_allocate; u32 max_sections_per_record; u32 max_raw_data_length; }; -/* 4: IPF Corrected Machine Check */ - -struct acpi_hest_ipf_corrected { - struct acpi_hest_header header; - u8 enabled; - u8 reserved; -}; - -/* 5: IPF Corrected Platform Error */ - -struct acpi_hest_ipf_corrected_platform { - struct acpi_hest_header header; - u8 enabled; - u8 reserved; -}; +/* 3,4,5: Not used */ /* 6: PCI Express Root Port AER */ @@ -491,30 +544,61 @@ struct acpi_hest_aer { struct acpi_hest_aer_bridge { struct acpi_hest_header header; struct acpi_hest_aer_common aer; - u32 secondary_uncorrectable_error_mask; - u32 secondary_uncorrectable_error_severity; - u32 secondary_advanced_capabilities; + u32 second_uncorrectable_error_mask; + u32 second_uncorrectable_error_severity; + u32 second_advanced_capabilities; }; /* 9: Generic Hardware Error Source */ struct acpi_hest_generic { struct acpi_hest_header header; - u16 source_id; u16 related_source_id; - u8 config_write_enable; + u8 reserved; u8 enabled; u32 records_to_pre_allocate; u32 max_sections_per_record; u32 max_raw_data_length; struct acpi_generic_address error_status_address; struct acpi_hest_notify notify; - u32 error_status_block_length; + u32 error_block_length; +}; + +/* Generic Error Status block */ + +struct acpi_hest_generic_status { + u32 block_status; + u32 raw_data_offset; + u32 raw_data_length; + u32 data_length; + u32 error_severity; +}; + +/* Values for block_status flags above */ + +#define ACPI_HEST_UNCORRECTABLE (1) +#define ACPI_HEST_CORRECTABLE (1<<1) +#define ACPI_HEST_MULTIPLE_UNCORRECTABLE (1<<2) +#define ACPI_HEST_MULTIPLE_CORRECTABLE (1<<3) +#define ACPI_HEST_ERROR_ENTRY_COUNT (0xFF<<4) /* 8 bits, error count */ + +/* Generic Error Data entry */ + +struct acpi_hest_generic_data { + u8 section_type[16]; + u32 error_severity; + u16 revision; + u8 validation_bits; + u8 flags; + u32 error_data_length; + u8 fru_id[16]; + u8 fru_text[20]; }; /******************************************************************************* * * MADT - Multiple APIC Description Table + * Version 3 * ******************************************************************************/ @@ -524,16 +608,16 @@ struct acpi_table_madt { u32 flags; }; -/* Flags */ +/* Masks for Flags field above */ -#define ACPI_MADT_PCAT_COMPAT (1) /* 00: System also has dual 8259s */ +#define ACPI_MADT_PCAT_COMPAT (1) /* 00: System also has dual 8259s */ /* Values for PCATCompat flag */ #define ACPI_MADT_DUAL_PIC 0 #define ACPI_MADT_MULTIPLE_APIC 1 -/* Values for subtable type in struct acpi_subtable_header */ +/* Values for MADT subtable type in struct acpi_subtable_header */ enum acpi_madt_type { ACPI_MADT_TYPE_LOCAL_APIC = 0, @@ -644,7 +728,7 @@ struct acpi_madt_interrupt_source { u32 flags; /* Interrupt Source Flags */ }; -/* Flags field above */ +/* Masks for Flags field above */ #define ACPI_MADT_CPEI_OVERRIDE (1) @@ -693,9 +777,36 @@ struct acpi_madt_local_x2apic_nmi { #define ACPI_MADT_TRIGGER_RESERVED (2<<2) #define ACPI_MADT_TRIGGER_LEVEL (3<<2) +/******************************************************************************* + * + * MSCT - Maximum System Characteristics Table (ACPI 4.0) + * Version 1 + * + ******************************************************************************/ + +struct acpi_table_msct { + struct acpi_table_header header; /* Common ACPI table header */ + u32 proximity_offset; /* Location of proximity info struct(s) */ + u32 max_proximity_domains; /* Max number of proximity domains */ + u32 max_clock_domains; /* Max number of clock domains */ + u64 max_address; /* Max physical address in system */ +}; + +/* Subtable - Maximum Proximity Domain Information. Version 1 */ + +struct acpi_msct_proximity { + u8 revision; + u8 length; + u32 range_start; /* Start of domain range */ + u32 range_end; /* End of domain range */ + u32 processor_capacity; + u64 memory_capacity; /* In bytes */ +}; + /******************************************************************************* * * SBST - Smart Battery Specification Table + * Version 1 * ******************************************************************************/ @@ -709,6 +820,7 @@ struct acpi_table_sbst { /******************************************************************************* * * SLIT - System Locality Distance Information Table + * Version 1 * ******************************************************************************/ @@ -721,6 +833,7 @@ struct acpi_table_slit { /******************************************************************************* * * SRAT - System Resource Affinity Table + * Version 3 * ******************************************************************************/ @@ -755,6 +868,10 @@ struct acpi_srat_cpu_affinity { u32 reserved; /* Reserved, must be zero */ }; +/* Flags */ + +#define ACPI_SRAT_CPU_USE_AFFINITY (1) /* 00: Use affinity structure */ + /* 1: Memory Affinity */ struct acpi_srat_mem_affinity { diff --git a/include/acpi/actbl2.h b/include/acpi/actbl2.h index b271aba0e524..6f3dce9991e1 100644 --- a/include/acpi/actbl2.h +++ b/include/acpi/actbl2.h @@ -14,8 +14,9 @@ ******************************************************************************/ /* - * Values for description table header signatures. Useful because they make - * it more difficult to inadvertently type in the wrong signature. + * Values for description table header signatures for tables defined in this + * file. Useful because they make it more difficult to inadvertently type in + * the wrong signature. */ #define ACPI_SIG_ASF "ASF!" /* Alert Standard Format table */ #define ACPI_SIG_BOOT "BOOT" /* Simple Boot Flag Table */ @@ -23,12 +24,14 @@ #define ACPI_SIG_DMAR "DMAR" /* DMA Remapping table */ #define ACPI_SIG_HPET "HPET" /* High Precision Event Timer table */ #define ACPI_SIG_IBFT "IBFT" /* i_sCSI Boot Firmware Table */ +#define ACPI_SIG_IVRS "IVRS" /* I/O Virtualization Reporting Structure */ #define ACPI_SIG_MCFG "MCFG" /* PCI Memory Mapped Configuration table */ #define ACPI_SIG_SLIC "SLIC" /* Software Licensing Description Table */ #define ACPI_SIG_SPCR "SPCR" /* Serial Port Console Redirection table */ #define ACPI_SIG_SPMI "SPMI" /* Server Platform Management Interface table */ #define ACPI_SIG_TCPA "TCPA" /* Trusted Computing Platform Alliance table */ #define ACPI_SIG_UEFI "UEFI" /* Uefi Boot Optimization Table */ +#define ACPI_SIG_WAET "WAET" /* Windows ACPI Emulated devices Table */ #define ACPI_SIG_WDAT "WDAT" /* Watchdog Action Table */ #define ACPI_SIG_WDRT "WDRT" /* Watchdog Resource Table */ @@ -47,6 +50,7 @@ /******************************************************************************* * * ASF - Alert Standard Format table (Signature "ASF!") + * Revision 0x10 * * Conforms to the Alert Standard Format Specification V2.0, 23 April 2003 * @@ -91,6 +95,10 @@ struct acpi_asf_info { u8 reserved2[3]; }; +/* Masks for Flags field above */ + +#define ACPI_ASF_SMBUS_PROTOCOLS (1) + /* 1: ASF Alerts */ struct acpi_asf_alert { @@ -156,6 +164,9 @@ struct acpi_asf_address { /******************************************************************************* * * BOOT - Simple Boot Flag Table + * Version 1 + * + * Conforms to the "Simple Boot Flag Specification", Version 2.1 * ******************************************************************************/ @@ -168,6 +179,9 @@ struct acpi_table_boot { /******************************************************************************* * * DBGP - Debug Port table + * Version 1 + * + * Conforms to the "Debug Port Specification", Version 1.00, 2/9/2000 * ******************************************************************************/ @@ -181,7 +195,10 @@ struct acpi_table_dbgp { /******************************************************************************* * * DMAR - DMA Remapping table - * From "Intel Virtualization Technology for Directed I/O", Sept. 2007 + * Version 1 + * + * Conforms to "Intel Virtualization Technology for Directed I/O", + * Version 1.2, Sept. 2008 * ******************************************************************************/ @@ -192,7 +209,7 @@ struct acpi_table_dmar { u8 reserved[10]; }; -/* Flags */ +/* Masks for Flags field above */ #define ACPI_DMAR_INTR_REMAP (1) @@ -209,9 +226,12 @@ enum acpi_dmar_type { ACPI_DMAR_TYPE_HARDWARE_UNIT = 0, ACPI_DMAR_TYPE_RESERVED_MEMORY = 1, ACPI_DMAR_TYPE_ATSR = 2, - ACPI_DMAR_TYPE_RESERVED = 3 /* 3 and greater are reserved */ + ACPI_DMAR_HARDWARE_AFFINITY = 3, + ACPI_DMAR_TYPE_RESERVED = 4 /* 4 and greater are reserved */ }; +/* DMAR Device Scope structure */ + struct acpi_dmar_device_scope { u8 entry_type; u8 length; @@ -250,7 +270,7 @@ struct acpi_dmar_hardware_unit { u64 address; /* Register Base Address */ }; -/* Flags */ +/* Masks for Flags field above */ #define ACPI_DMAR_INCLUDE_ALL (1) @@ -264,7 +284,7 @@ struct acpi_dmar_reserved_memory { u64 end_address; /* 4_k aligned limit address */ }; -/* Flags */ +/* Masks for Flags field above */ #define ACPI_DMAR_ALLOW_ALL (1) @@ -277,13 +297,26 @@ struct acpi_dmar_atsr { u16 segment; }; -/* Flags */ +/* Masks for Flags field above */ #define ACPI_DMAR_ALL_PORTS (1) +/* 3: Remapping Hardware Static Affinity Structure */ + +struct acpi_dmar_rhsa { + struct acpi_dmar_header header; + u32 reserved; + u64 base_address; + u32 proximity_domain; +}; + /******************************************************************************* * * HPET - High Precision Event Timer table + * Version 1 + * + * Conforms to "IA-PC HPET (High Precision Event Timers) Specification", + * Version 1.0a, October 2004 * ******************************************************************************/ @@ -296,17 +329,28 @@ struct acpi_table_hpet { u8 flags; }; -/*! Flags */ +/* Masks for Flags field above */ -#define ACPI_HPET_PAGE_PROTECT (1) /* 00: No page protection */ -#define ACPI_HPET_PAGE_PROTECT_4 (1<<1) /* 01: 4KB page protected */ -#define ACPI_HPET_PAGE_PROTECT_64 (1<<2) /* 02: 64KB page protected */ +#define ACPI_HPET_PAGE_PROTECT_MASK (3) -/*! [End] no source code translation !*/ +/* Values for Page Protect flags */ + +enum acpi_hpet_page_protect { + ACPI_HPET_NO_PAGE_PROTECT = 0, + ACPI_HPET_PAGE_PROTECT4 = 1, + ACPI_HPET_PAGE_PROTECT64 = 2 +}; /******************************************************************************* * * IBFT - Boot Firmware Table + * Version 1 + * + * Conforms to "iSCSI Boot Firmware Table (iBFT) as Defined in ACPI 3.0b + * Specification", Version 1.01, March 1, 2007 + * + * Note: It appears that this table is not intended to appear in the RSDT/XSDT. + * Therefore, it is not currently supported by the disassembler. * ******************************************************************************/ @@ -394,9 +438,184 @@ struct acpi_ibft_target { u16 reverse_chap_secret_offset; }; +/******************************************************************************* + * + * IVRS - I/O Virtualization Reporting Structure + * Version 1 + * + * Conforms to "AMD I/O Virtualization Technology (IOMMU) Specification", + * Revision 1.26, February 2009. + * + ******************************************************************************/ + +struct acpi_table_ivrs { + struct acpi_table_header header; /* Common ACPI table header */ + u32 info; /* Common virtualization info */ + u64 reserved; +}; + +/* Values for Info field above */ + +#define ACPI_IVRS_PHYSICAL_SIZE 0x00007F00 /* 7 bits, physical address size */ +#define ACPI_IVRS_VIRTUAL_SIZE 0x003F8000 /* 7 bits, virtual address size */ +#define ACPI_IVRS_ATS_RESERVED 0x00400000 /* ATS address translation range reserved */ + +/* IVRS subtable header */ + +struct acpi_ivrs_header { + u8 type; /* Subtable type */ + u8 flags; + u16 length; /* Subtable length */ + u16 device_id; /* ID of IOMMU */ +}; + +/* Values for subtable Type above */ + +enum acpi_ivrs_type { + ACPI_IVRS_TYPE_HARDWARE = 0x10, + ACPI_IVRS_TYPE_MEMORY1 = 0x20, + ACPI_IVRS_TYPE_MEMORY2 = 0x21, + ACPI_IVRS_TYPE_MEMORY3 = 0x22 +}; + +/* Masks for Flags field above for IVHD subtable */ + +#define ACPI_IVHD_TT_ENABLE (1) +#define ACPI_IVHD_PASS_PW (1<<1) +#define ACPI_IVHD_RES_PASS_PW (1<<2) +#define ACPI_IVHD_ISOC (1<<3) +#define ACPI_IVHD_IOTLB (1<<4) + +/* Masks for Flags field above for IVMD subtable */ + +#define ACPI_IVMD_UNITY (1) +#define ACPI_IVMD_READ (1<<1) +#define ACPI_IVMD_WRITE (1<<2) +#define ACPI_IVMD_EXCLUSION_RANGE (1<<3) + +/* + * IVRS subtables, correspond to Type in struct acpi_ivrs_header + */ + +/* 0x10: I/O Virtualization Hardware Definition Block (IVHD) */ + +struct acpi_ivrs_hardware { + struct acpi_ivrs_header header; + u16 capability_offset; /* Offset for IOMMU control fields */ + u64 base_address; /* IOMMU control registers */ + u16 pci_segment_group; + u16 info; /* MSI number and unit ID */ + u32 reserved; +}; + +/* Masks for Info field above */ + +#define ACPI_IVHD_MSI_NUMBER_MASK 0x001F /* 5 bits, MSI message number */ +#define ACPI_IVHD_UNIT_ID_MASK 0x1F00 /* 5 bits, unit_iD */ + +/* + * Device Entries for IVHD subtable, appear after struct acpi_ivrs_hardware structure. + * Upper two bits of the Type field are the (encoded) length of the structure. + * Currently, only 4 and 8 byte entries are defined. 16 and 32 byte entries + * are reserved for future use but not defined. + */ +struct acpi_ivrs_de_header { + u8 type; + u16 id; + u8 data_setting; +}; + +/* Length of device entry is in the top two bits of Type field above */ + +#define ACPI_IVHD_ENTRY_LENGTH 0xC0 + +/* Values for device entry Type field above */ + +enum acpi_ivrs_device_entry_type { + /* 4-byte device entries, all use struct acpi_ivrs_device4 */ + + ACPI_IVRS_TYPE_PAD4 = 0, + ACPI_IVRS_TYPE_ALL = 1, + ACPI_IVRS_TYPE_SELECT = 2, + ACPI_IVRS_TYPE_START = 3, + ACPI_IVRS_TYPE_END = 4, + + /* 8-byte device entries */ + + ACPI_IVRS_TYPE_PAD8 = 64, + ACPI_IVRS_TYPE_NOT_USED = 65, + ACPI_IVRS_TYPE_ALIAS_SELECT = 66, /* Uses struct acpi_ivrs_device8a */ + ACPI_IVRS_TYPE_ALIAS_START = 67, /* Uses struct acpi_ivrs_device8a */ + ACPI_IVRS_TYPE_EXT_SELECT = 70, /* Uses struct acpi_ivrs_device8b */ + ACPI_IVRS_TYPE_EXT_START = 71, /* Uses struct acpi_ivrs_device8b */ + ACPI_IVRS_TYPE_SPECIAL = 72 /* Uses struct acpi_ivrs_device8c */ +}; + +/* Values for Data field above */ + +#define ACPI_IVHD_INIT_PASS (1) +#define ACPI_IVHD_EINT_PASS (1<<1) +#define ACPI_IVHD_NMI_PASS (1<<2) +#define ACPI_IVHD_SYSTEM_MGMT (3<<4) +#define ACPI_IVHD_LINT0_PASS (1<<6) +#define ACPI_IVHD_LINT1_PASS (1<<7) + +/* Types 0-4: 4-byte device entry */ + +struct acpi_ivrs_device4 { + struct acpi_ivrs_de_header header; +}; + +/* Types 66-67: 8-byte device entry */ + +struct acpi_ivrs_device8a { + struct acpi_ivrs_de_header header; + u8 reserved1; + u16 used_id; + u8 reserved2; +}; + +/* Types 70-71: 8-byte device entry */ + +struct acpi_ivrs_device8b { + struct acpi_ivrs_de_header header; + u32 extended_data; +}; + +/* Values for extended_data above */ + +#define ACPI_IVHD_ATS_DISABLED (1<<31) + +/* Type 72: 8-byte device entry */ + +struct acpi_ivrs_device8c { + struct acpi_ivrs_de_header header; + u8 handle; + u16 used_id; + u8 variety; +}; + +/* Values for Variety field above */ + +#define ACPI_IVHD_IOAPIC 1 +#define ACPI_IVHD_HPET 2 + +/* 0x20, 0x21, 0x22: I/O Virtualization Memory Definition Block (IVMD) */ + +struct acpi_ivrs_memory { + struct acpi_ivrs_header header; + u16 aux_data; + u64 reserved; + u64 start_address; + u64 memory_length; +}; + /******************************************************************************* * * MCFG - PCI Memory Mapped Configuration table and sub-table + * Version 1 + * + * Conforms to "PCI Firmware Specification", Revision 3.0, June 20, 2005 * ******************************************************************************/ @@ -418,6 +637,10 @@ struct acpi_mcfg_allocation { /******************************************************************************* * * SPCR - Serial Port Console Redirection table + * Version 1 + * + * Conforms to "Serial Port Console Redirection Table", + * Version 1.00, January 11, 2002 * ******************************************************************************/ @@ -445,16 +668,25 @@ struct acpi_table_spcr { u32 reserved2; }; +/* Masks for pci_flags field above */ + +#define ACPI_SPCR_DO_NOT_DISABLE (1) + /******************************************************************************* * * SPMI - Server Platform Management Interface table + * Version 5 + * + * Conforms to "Intelligent Platform Management Interface Specification + * Second Generation v2.0", Document Revision 1.0, February 12, 2004 with + * June 12, 2009 markup. * ******************************************************************************/ struct acpi_table_spmi { struct acpi_table_header header; /* Common ACPI table header */ - u8 reserved; u8 interface_type; + u8 reserved; /* Must be 1 */ u16 spec_revision; /* Version of IPMI */ u8 interrupt_type; u8 gpe_number; /* GPE assigned */ @@ -466,11 +698,27 @@ struct acpi_table_spmi { u8 pci_bus; u8 pci_device; u8 pci_function; + u8 reserved2; +}; + +/* Values for interface_type above */ + +enum acpi_spmi_interface_types { + ACPI_SPMI_NOT_USED = 0, + ACPI_SPMI_KEYBOARD = 1, + ACPI_SPMI_SMI = 2, + ACPI_SPMI_BLOCK_TRANSFER = 3, + ACPI_SPMI_SMBUS = 4, + ACPI_SPMI_RESERVED = 5 /* 5 and above are reserved */ }; /******************************************************************************* * * TCPA - Trusted Computing Platform Alliance table + * Version 1 + * + * Conforms to "TCG PC Specific Implementation Specification", + * Version 1.1, August 18, 2003 * ******************************************************************************/ @@ -484,6 +732,10 @@ struct acpi_table_tcpa { /******************************************************************************* * * UEFI - UEFI Boot optimization Table + * Version 1 + * + * Conforms to "Unified Extensible Firmware Interface Specification", + * Version 2.3, May 8, 2009 * ******************************************************************************/ @@ -491,12 +743,34 @@ struct acpi_table_uefi { struct acpi_table_header header; /* Common ACPI table header */ u8 identifier[16]; /* UUID identifier */ u16 data_offset; /* Offset of remaining data in table */ - u8 data; }; +/******************************************************************************* + * + * WAET - Windows ACPI Emulated devices Table + * Version 1 + * + * Conforms to "Windows ACPI Emulated Devices Table", version 1.0, April 6, 2009 + * + ******************************************************************************/ + +struct acpi_table_waet { + struct acpi_table_header header; /* Common ACPI table header */ + u32 flags; +}; + +/* Masks for Flags field above */ + +#define ACPI_WAET_RTC_NO_ACK (1) /* RTC requires no int acknowledge */ +#define ACPI_WAET_TIMER_ONE_READ (1<<1) /* PM timer requires only one read */ + /******************************************************************************* * * WDAT - Watchdog Action Table + * Version 1 + * + * Conforms to "Hardware Watchdog Timers Design Specification", + * Copyright 2006 Microsoft Corporation. * ******************************************************************************/ @@ -516,10 +790,20 @@ struct acpi_table_wdat { u32 entries; /* Number of watchdog entries that follow */ }; +/* Masks for Flags field above */ + +#define ACPI_WDAT_ENABLED (1) +#define ACPI_WDAT_STOPPED 0x80 + /* WDAT Instruction Entries (actions) */ struct acpi_wdat_entry { - struct acpi_whea_header whea_header; /* Common header for WHEA tables */ + u8 action; + u8 instruction; + u16 reserved; + struct acpi_generic_address register_region; + u32 value; /* Value used with Read/Write register */ + u32 mask; /* Bitmask required for this register instruction */ }; /* Values for Action field above */ @@ -556,28 +840,27 @@ enum acpi_wdat_instructions { /******************************************************************************* * * WDRT - Watchdog Resource Table + * Version 1 + * + * Conforms to "Watchdog Timer Hardware Requirements for Windows Server 2003", + * Version 1.01, August 28, 2006 * ******************************************************************************/ struct acpi_table_wdrt { struct acpi_table_header header; /* Common ACPI table header */ - u32 header_length; /* Watchdog Header Length */ - u8 pci_segment; /* PCI Segment number */ + struct acpi_generic_address control_register; + struct acpi_generic_address count_register; + u16 pci_device_id; + u16 pci_vendor_id; u8 pci_bus; /* PCI Bus number */ u8 pci_device; /* PCI Device number */ u8 pci_function; /* PCI Function number */ - u32 timer_period; /* Period of one timer count (msec) */ - u32 max_count; /* Maximum counter value supported */ - u32 min_count; /* Minimum counter value */ - u8 flags; - u8 reserved[3]; - u32 entries; /* Number of watchdog entries that follow */ + u8 pci_segment; /* PCI Segment number */ + u16 max_count; /* Maximum counter value supported */ + u8 units; }; -/* Flags */ - -#define ACPI_WDRT_TIMER_ENABLED (1) /* 00: Timer enabled */ - /* Reset to default packing */ #pragma pack() -- cgit v1.2.3 From c276e3884163355464a76e60ed9e272b52b4acc2 Mon Sep 17 00:00:00 2001 From: Bob Moore Date: Mon, 27 Jul 2009 14:55:02 +0800 Subject: ACPICA: Update definitions for HEST table Eliminate duplicated code in disassembler. Shorten identifiers that were too long. Signed-off-by: Bob Moore Signed-off-by: Lin Ming Signed-off-by: Len Brown --- include/acpi/actbl1.h | 24 ++++++++++++------------ 1 file changed, 12 insertions(+), 12 deletions(-) (limited to 'include') diff --git a/include/acpi/actbl1.h b/include/acpi/actbl1.h index 0417f2abc44b..34b10c06bcfd 100644 --- a/include/acpi/actbl1.h +++ b/include/acpi/actbl1.h @@ -427,17 +427,17 @@ struct acpi_hest_aer_common { u16 reserved1; u8 flags; u8 enabled; - u32 records_to_pre_allocate; + u32 records_to_preallocate; u32 max_sections_per_record; u32 bus; u16 device; u16 function; u16 device_control; u16 reserved2; - u32 uncorrectable_error_mask; - u32 uncorrectable_error_severity; - u32 correctable_error_mask; - u32 advanced_error_capabilities; + u32 uncorrectable_mask; + u32 uncorrectable_severity; + u32 correctable_mask; + u32 advanced_capabilities; }; /* Masks for HEST Flags fields */ @@ -490,7 +490,7 @@ struct acpi_hest_ia_machine_check { u16 reserved1; u8 flags; u8 enabled; - u32 records_to_pre_allocate; + u32 records_to_preallocate; u32 max_sections_per_record; u64 global_capability_data; u64 global_control_data; @@ -505,7 +505,7 @@ struct acpi_table_hest_ia_corrected { u16 reserved1; u8 flags; u8 enabled; - u32 records_to_pre_allocate; + u32 records_to_preallocate; u32 max_sections_per_record; struct acpi_hest_notify notify; u8 num_hardware_banks; @@ -517,7 +517,7 @@ struct acpi_table_hest_ia_corrected { struct acpi_hest_ia_nmi { struct acpi_hest_header header; u32 reserved; - u32 records_to_pre_allocate; + u32 records_to_preallocate; u32 max_sections_per_record; u32 max_raw_data_length; }; @@ -544,9 +544,9 @@ struct acpi_hest_aer { struct acpi_hest_aer_bridge { struct acpi_hest_header header; struct acpi_hest_aer_common aer; - u32 second_uncorrectable_error_mask; - u32 second_uncorrectable_error_severity; - u32 second_advanced_capabilities; + u32 uncorrectable_mask2; + u32 uncorrectable_severity2; + u32 advanced_capabilities2; }; /* 9: Generic Hardware Error Source */ @@ -556,7 +556,7 @@ struct acpi_hest_generic { u16 related_source_id; u8 reserved; u8 enabled; - u32 records_to_pre_allocate; + u32 records_to_preallocate; u32 max_sections_per_record; u32 max_raw_data_length; struct acpi_generic_address error_status_address; -- cgit v1.2.3 From 1872bbc94b2d092ece22a8fbf1c3e81f0fba0052 Mon Sep 17 00:00:00 2001 From: Bob Moore Date: Thu, 13 Aug 2009 13:31:00 +0800 Subject: ACPICA: Fix typo for HEST ACPI table Problem with the name of one of the subtables. Signed-off-by: Bob Moore Signed-off-by: Lin Ming Signed-off-by: Len Brown --- include/acpi/actbl1.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/acpi/actbl1.h b/include/acpi/actbl1.h index 34b10c06bcfd..0b9b430b092b 100644 --- a/include/acpi/actbl1.h +++ b/include/acpi/actbl1.h @@ -500,7 +500,7 @@ struct acpi_hest_ia_machine_check { /* 1: IA32 Corrected Machine Check */ -struct acpi_table_hest_ia_corrected { +struct acpi_hest_ia_corrected { struct acpi_hest_header header; u16 reserved1; u8 flags; -- cgit v1.2.3 From 49ae80c9944401222e47108883c486b5a5a24006 Mon Sep 17 00:00:00 2001 From: Bob Moore Date: Thu, 13 Aug 2009 13:43:12 +0800 Subject: ACPICA: Update version to 20090730 Version 20090730. Signed-off-by: Bob Moore Signed-off-by: Lin Ming Signed-off-by: Len Brown --- include/acpi/acpixf.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/acpi/acpixf.h b/include/acpi/acpixf.h index 063e577e791e..f3b358b7432f 100644 --- a/include/acpi/acpixf.h +++ b/include/acpi/acpixf.h @@ -47,7 +47,7 @@ /* Current ACPICA subsystem version in YYYYMMDD format */ -#define ACPI_CA_VERSION 0x20090625 +#define ACPI_CA_VERSION 0x20090730 #include "actypes.h" #include "actbl.h" -- cgit v1.2.3 From a192a9580bcc41692be1f36b77c3b681827f566a Mon Sep 17 00:00:00 2001 From: Len Brown Date: Tue, 28 Jul 2009 16:45:54 -0400 Subject: ACPI: Move definition of PREFIX from acpi_bus.h to internal..h Linux/ACPI core files using internal.h all PREFIX "ACPI: ", however, not all ACPI drivers use/want it -- and they should not have to #undef PREFIX to define their own. Add GPL commment to internal.h while we are there. This does not change any actual console output, asside from a whitespace fix. Signed-off-by: Len Brown --- arch/x86/pci/mmconfig-shared.c | 2 ++ drivers/acpi/ac.c | 2 ++ drivers/acpi/battery.c | 2 ++ drivers/acpi/blacklist.c | 2 ++ drivers/acpi/button.c | 2 ++ drivers/acpi/cm_sbs.c | 2 ++ drivers/acpi/container.c | 2 ++ drivers/acpi/dock.c | 2 ++ drivers/acpi/ec.c | 1 - drivers/acpi/event.c | 2 ++ drivers/acpi/fan.c | 2 ++ drivers/acpi/glue.c | 2 ++ drivers/acpi/internal.h | 22 +++++++++++++++++++++- drivers/acpi/numa.c | 2 ++ drivers/acpi/pci_irq.c | 2 ++ drivers/acpi/pci_link.c | 2 ++ drivers/acpi/pci_root.c | 2 ++ drivers/acpi/power.c | 2 ++ drivers/acpi/processor_core.c | 2 ++ drivers/acpi/processor_idle.c | 2 ++ drivers/acpi/processor_perflib.c | 2 ++ drivers/acpi/processor_thermal.c | 2 ++ drivers/acpi/processor_throttling.c | 2 ++ drivers/acpi/sbs.c | 2 ++ drivers/acpi/sbshc.c | 2 ++ drivers/acpi/system.c | 2 ++ drivers/acpi/thermal.c | 2 ++ drivers/acpi/utils.c | 2 ++ drivers/acpi/video.c | 2 ++ drivers/acpi/video_detect.c | 2 ++ drivers/pci/dmar.c | 3 +-- drivers/platform/x86/fujitsu-laptop.c | 4 ++-- drivers/platform/x86/wmi.c | 1 - include/acpi/acpi_bus.h | 2 -- 34 files changed, 80 insertions(+), 9 deletions(-) (limited to 'include') diff --git a/arch/x86/pci/mmconfig-shared.c b/arch/x86/pci/mmconfig-shared.c index 712443ec6d43..81d3466765ca 100644 --- a/arch/x86/pci/mmconfig-shared.c +++ b/arch/x86/pci/mmconfig-shared.c @@ -18,6 +18,8 @@ #include #include +#define PREFIX "ACPI: " + /* aperture is up to 256MB but BIOS may reserve less */ #define MMCONFIG_APER_MIN (2 * 1024*1024) #define MMCONFIG_APER_MAX (256 * 1024*1024) diff --git a/drivers/acpi/ac.c b/drivers/acpi/ac.c index 0df8fcb687d6..98b9690b0159 100644 --- a/drivers/acpi/ac.c +++ b/drivers/acpi/ac.c @@ -37,6 +37,8 @@ #include #include +#define PREFIX "ACPI: " + #define ACPI_AC_CLASS "ac_adapter" #define ACPI_AC_DEVICE_NAME "AC Adapter" #define ACPI_AC_FILE_STATE "state" diff --git a/drivers/acpi/battery.c b/drivers/acpi/battery.c index 58b4517ce712..f8c3d1bb6969 100644 --- a/drivers/acpi/battery.c +++ b/drivers/acpi/battery.c @@ -45,6 +45,8 @@ #include #endif +#define PREFIX "ACPI: " + #define ACPI_BATTERY_VALUE_UNKNOWN 0xFFFFFFFF #define ACPI_BATTERY_CLASS "battery" diff --git a/drivers/acpi/blacklist.c b/drivers/acpi/blacklist.c index f6baa77deefb..19152ea2b104 100644 --- a/drivers/acpi/blacklist.c +++ b/drivers/acpi/blacklist.c @@ -34,6 +34,8 @@ #include #include +#include "internal.h" + enum acpi_blacklist_predicates { all_versions, less_than_or_equal, diff --git a/drivers/acpi/button.c b/drivers/acpi/button.c index 9195deba9d94..d295bdccc09c 100644 --- a/drivers/acpi/button.c +++ b/drivers/acpi/button.c @@ -33,6 +33,8 @@ #include #include +#define PREFIX "ACPI: " + #define ACPI_BUTTON_CLASS "button" #define ACPI_BUTTON_FILE_INFO "info" #define ACPI_BUTTON_FILE_STATE "state" diff --git a/drivers/acpi/cm_sbs.c b/drivers/acpi/cm_sbs.c index 332fe4b21708..6c9ee68e46fb 100644 --- a/drivers/acpi/cm_sbs.c +++ b/drivers/acpi/cm_sbs.c @@ -28,6 +28,8 @@ #include #include +#define PREFIX "ACPI: " + ACPI_MODULE_NAME("cm_sbs"); #define ACPI_AC_CLASS "ac_adapter" #define ACPI_BATTERY_CLASS "battery" diff --git a/drivers/acpi/container.c b/drivers/acpi/container.c index fe0cdf83641a..5f2c3c00a315 100644 --- a/drivers/acpi/container.c +++ b/drivers/acpi/container.c @@ -35,6 +35,8 @@ #include #include +#define PREFIX "ACPI: " + #define ACPI_CONTAINER_DEVICE_NAME "ACPI container device" #define ACPI_CONTAINER_CLASS "container" diff --git a/drivers/acpi/dock.c b/drivers/acpi/dock.c index efb959d6c8a9..9a855669ff12 100644 --- a/drivers/acpi/dock.c +++ b/drivers/acpi/dock.c @@ -33,6 +33,8 @@ #include #include +#define PREFIX "ACPI: " + #define ACPI_DOCK_DRIVER_DESCRIPTION "ACPI Dock Station Driver" ACPI_MODULE_NAME("dock"); diff --git a/drivers/acpi/ec.c b/drivers/acpi/ec.c index 391f331674c7..5180f0f1dd02 100644 --- a/drivers/acpi/ec.c +++ b/drivers/acpi/ec.c @@ -47,7 +47,6 @@ #define ACPI_EC_DEVICE_NAME "Embedded Controller" #define ACPI_EC_FILE_INFO "info" -#undef PREFIX #define PREFIX "ACPI: EC: " /* EC status register */ diff --git a/drivers/acpi/event.c b/drivers/acpi/event.c index aeb7e5fb4a04..c511071bfd79 100644 --- a/drivers/acpi/event.c +++ b/drivers/acpi/event.c @@ -14,6 +14,8 @@ #include #include +#include "internal.h" + #define _COMPONENT ACPI_SYSTEM_COMPONENT ACPI_MODULE_NAME("event"); diff --git a/drivers/acpi/fan.c b/drivers/acpi/fan.c index 53698ea08371..f419849a0d3f 100644 --- a/drivers/acpi/fan.c +++ b/drivers/acpi/fan.c @@ -34,6 +34,8 @@ #include #include +#define PREFIX "ACPI: " + #define ACPI_FAN_CLASS "fan" #define ACPI_FAN_FILE_STATE "state" diff --git a/drivers/acpi/glue.c b/drivers/acpi/glue.c index a8a5c29958c8..dc36a448de43 100644 --- a/drivers/acpi/glue.c +++ b/drivers/acpi/glue.c @@ -12,6 +12,8 @@ #include #include +#include "internal.h" + #define ACPI_GLUE_DEBUG 0 #if ACPI_GLUE_DEBUG #define DBG(x...) printk(PREFIX x) diff --git a/drivers/acpi/internal.h b/drivers/acpi/internal.h index 11a69b53004e..074cf8682d52 100644 --- a/drivers/acpi/internal.h +++ b/drivers/acpi/internal.h @@ -1,4 +1,24 @@ -/* For use by Linux/ACPI infrastructure, not drivers */ +/* + * acpi/internal.h + * For use by Linux/ACPI infrastructure, not drivers + * + * Copyright (c) 2009, Intel Corporation. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program; if not, write to the Free Software Foundation, Inc., + * 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA. + */ + +#define PREFIX "ACPI: " int init_acpi_device_notify(void); int acpi_scan_init(void); diff --git a/drivers/acpi/numa.c b/drivers/acpi/numa.c index d440ccd27d91..202dd0c976a3 100644 --- a/drivers/acpi/numa.c +++ b/drivers/acpi/numa.c @@ -30,6 +30,8 @@ #include #include +#define PREFIX "ACPI: " + #define ACPI_NUMA 0x80000000 #define _COMPONENT ACPI_NUMA ACPI_MODULE_NAME("numa"); diff --git a/drivers/acpi/pci_irq.c b/drivers/acpi/pci_irq.c index b794eb88ab90..843699ed93f2 100644 --- a/drivers/acpi/pci_irq.c +++ b/drivers/acpi/pci_irq.c @@ -40,6 +40,8 @@ #include #include +#define PREFIX "ACPI: " + #define _COMPONENT ACPI_PCI_COMPONENT ACPI_MODULE_NAME("pci_irq"); diff --git a/drivers/acpi/pci_link.c b/drivers/acpi/pci_link.c index 16e0f9d3d17c..394ae89409c2 100644 --- a/drivers/acpi/pci_link.c +++ b/drivers/acpi/pci_link.c @@ -43,6 +43,8 @@ #include #include +#define PREFIX "ACPI: " + #define _COMPONENT ACPI_PCI_COMPONENT ACPI_MODULE_NAME("pci_link"); #define ACPI_PCI_LINK_CLASS "pci_irq_routing" diff --git a/drivers/acpi/pci_root.c b/drivers/acpi/pci_root.c index 55b5b90c2a44..dee916707a7d 100644 --- a/drivers/acpi/pci_root.c +++ b/drivers/acpi/pci_root.c @@ -36,6 +36,8 @@ #include #include +#define PREFIX "ACPI: " + #define _COMPONENT ACPI_PCI_COMPONENT ACPI_MODULE_NAME("pci_root"); #define ACPI_PCI_ROOT_CLASS "pci_bridge" diff --git a/drivers/acpi/power.c b/drivers/acpi/power.c index d74365d4a6e7..e86603f37dee 100644 --- a/drivers/acpi/power.c +++ b/drivers/acpi/power.c @@ -44,6 +44,8 @@ #include #include +#define PREFIX "ACPI: " + #define _COMPONENT ACPI_POWER_COMPONENT ACPI_MODULE_NAME("power"); #define ACPI_POWER_CLASS "power_resource" diff --git a/drivers/acpi/processor_core.c b/drivers/acpi/processor_core.c index 2cc4b3033872..b4a1ab297e7b 100644 --- a/drivers/acpi/processor_core.c +++ b/drivers/acpi/processor_core.c @@ -59,6 +59,8 @@ #include #include +#define PREFIX "ACPI: " + #define ACPI_PROCESSOR_CLASS "processor" #define ACPI_PROCESSOR_DEVICE_NAME "Processor" #define ACPI_PROCESSOR_FILE_INFO "info" diff --git a/drivers/acpi/processor_idle.c b/drivers/acpi/processor_idle.c index 66393d5c4c7c..22aab1fc9b45 100644 --- a/drivers/acpi/processor_idle.c +++ b/drivers/acpi/processor_idle.c @@ -60,6 +60,8 @@ #include #include +#define PREFIX "ACPI: " + #define ACPI_PROCESSOR_CLASS "processor" #define _COMPONENT ACPI_PROCESSOR_COMPONENT ACPI_MODULE_NAME("processor_idle"); diff --git a/drivers/acpi/processor_perflib.c b/drivers/acpi/processor_perflib.c index 60e543d3234e..11088cf10319 100644 --- a/drivers/acpi/processor_perflib.c +++ b/drivers/acpi/processor_perflib.c @@ -39,6 +39,8 @@ #include #include +#define PREFIX "ACPI: " + #define ACPI_PROCESSOR_CLASS "processor" #define ACPI_PROCESSOR_FILE_PERFORMANCE "performance" #define _COMPONENT ACPI_PROCESSOR_COMPONENT diff --git a/drivers/acpi/processor_thermal.c b/drivers/acpi/processor_thermal.c index 31adda1099e0..3e3181c0efc3 100644 --- a/drivers/acpi/processor_thermal.c +++ b/drivers/acpi/processor_thermal.c @@ -40,6 +40,8 @@ #include #include +#define PREFIX "ACPI: " + #define ACPI_PROCESSOR_CLASS "processor" #define _COMPONENT ACPI_PROCESSOR_COMPONENT ACPI_MODULE_NAME("processor_thermal"); diff --git a/drivers/acpi/processor_throttling.c b/drivers/acpi/processor_throttling.c index ae39797aab55..b366b9c13d4d 100644 --- a/drivers/acpi/processor_throttling.c +++ b/drivers/acpi/processor_throttling.c @@ -41,6 +41,8 @@ #include #include +#define PREFIX "ACPI: " + #define ACPI_PROCESSOR_CLASS "processor" #define _COMPONENT ACPI_PROCESSOR_COMPONENT ACPI_MODULE_NAME("processor_throttling"); diff --git a/drivers/acpi/sbs.c b/drivers/acpi/sbs.c index 4b214b74ebaa..52b9db8afc20 100644 --- a/drivers/acpi/sbs.c +++ b/drivers/acpi/sbs.c @@ -46,6 +46,8 @@ #include "sbshc.h" +#define PREFIX "ACPI: " + #define ACPI_SBS_CLASS "sbs" #define ACPI_AC_CLASS "ac_adapter" #define ACPI_BATTERY_CLASS "battery" diff --git a/drivers/acpi/sbshc.c b/drivers/acpi/sbshc.c index 0619734895b2..d9339806df45 100644 --- a/drivers/acpi/sbshc.c +++ b/drivers/acpi/sbshc.c @@ -15,6 +15,8 @@ #include #include "sbshc.h" +#define PREFIX "ACPI: " + #define ACPI_SMB_HC_CLASS "smbus_host_controller" #define ACPI_SMB_HC_DEVICE_NAME "ACPI SMBus HC" diff --git a/drivers/acpi/system.c b/drivers/acpi/system.c index 9c61ab2177cf..d11282975f35 100644 --- a/drivers/acpi/system.c +++ b/drivers/acpi/system.c @@ -31,6 +31,8 @@ #include +#define PREFIX "ACPI: " + #define _COMPONENT ACPI_SYSTEM_COMPONENT ACPI_MODULE_NAME("system"); diff --git a/drivers/acpi/thermal.c b/drivers/acpi/thermal.c index 564ea1424288..65f67815902a 100644 --- a/drivers/acpi/thermal.c +++ b/drivers/acpi/thermal.c @@ -47,6 +47,8 @@ #include #include +#define PREFIX "ACPI: " + #define ACPI_THERMAL_CLASS "thermal_zone" #define ACPI_THERMAL_DEVICE_NAME "Thermal Zone" #define ACPI_THERMAL_FILE_STATE "state" diff --git a/drivers/acpi/utils.c b/drivers/acpi/utils.c index f844941089bb..811fec10462b 100644 --- a/drivers/acpi/utils.c +++ b/drivers/acpi/utils.c @@ -30,6 +30,8 @@ #include #include +#include "internal.h" + #define _COMPONENT ACPI_BUS_COMPONENT ACPI_MODULE_NAME("utils"); diff --git a/drivers/acpi/video.c b/drivers/acpi/video.c index 8851315ce858..a0fa3946b507 100644 --- a/drivers/acpi/video.c +++ b/drivers/acpi/video.c @@ -44,6 +44,8 @@ #include #include +#define PREFIX "ACPI: " + #define ACPI_VIDEO_CLASS "video" #define ACPI_VIDEO_BUS_NAME "Video Bus" #define ACPI_VIDEO_DEVICE_NAME "Video Device" diff --git a/drivers/acpi/video_detect.c b/drivers/acpi/video_detect.c index 7cd2b63435ea..7032f25da9b5 100644 --- a/drivers/acpi/video_detect.c +++ b/drivers/acpi/video_detect.c @@ -38,6 +38,8 @@ #include #include +#define PREFIX "ACPI: " + ACPI_MODULE_NAME("video"); #define _COMPONENT ACPI_VIDEO_COMPONENT diff --git a/drivers/pci/dmar.c b/drivers/pci/dmar.c index 7b287cb38b7a..998f02d2ba42 100644 --- a/drivers/pci/dmar.c +++ b/drivers/pci/dmar.c @@ -34,8 +34,7 @@ #include #include -#undef PREFIX -#define PREFIX "DMAR:" +#define PREFIX "DMAR: " /* No locks are needed as DMA remapping hardware unit * list is constructed at boot time and hotplug of diff --git a/drivers/platform/x86/fujitsu-laptop.c b/drivers/platform/x86/fujitsu-laptop.c index 218b9a16ac3f..eabddc9c192b 100644 --- a/drivers/platform/x86/fujitsu-laptop.c +++ b/drivers/platform/x86/fujitsu-laptop.c @@ -700,7 +700,7 @@ static int acpi_fujitsu_add(struct acpi_device *device) goto end; } - printk(KERN_INFO PREFIX "%s [%s] (%s)\n", + printk(KERN_INFO "ACPI: %s [%s] (%s)\n", acpi_device_name(device), acpi_device_bid(device), !device->power.state ? "on" : "off"); @@ -874,7 +874,7 @@ static int acpi_fujitsu_hotkey_add(struct acpi_device *device) goto end; } - printk(KERN_INFO PREFIX "%s [%s] (%s)\n", + printk(KERN_INFO "ACPI: %s [%s] (%s)\n", acpi_device_name(device), acpi_device_bid(device), !device->power.state ? "on" : "off"); diff --git a/drivers/platform/x86/wmi.c b/drivers/platform/x86/wmi.c index f215a5919192..177f8d767df4 100644 --- a/drivers/platform/x86/wmi.c +++ b/drivers/platform/x86/wmi.c @@ -42,7 +42,6 @@ MODULE_LICENSE("GPL"); #define ACPI_WMI_CLASS "wmi" -#undef PREFIX #define PREFIX "ACPI: WMI: " static DEFINE_MUTEX(wmi_data_lock); diff --git a/include/acpi/acpi_bus.h b/include/acpi/acpi_bus.h index c65e4ce6c3af..f485107ddc43 100644 --- a/include/acpi/acpi_bus.h +++ b/include/acpi/acpi_bus.h @@ -30,8 +30,6 @@ #include -#define PREFIX "ACPI: " - /* TBD: Make dynamic */ #define ACPI_MAX_HANDLES 10 struct acpi_handle_list { -- cgit v1.2.3 From e55a5999ffcf72dc4d43d73618957964cb87065a Mon Sep 17 00:00:00 2001 From: Feng Tang Date: Tue, 28 Jul 2009 17:41:53 +0800 Subject: ACPI: Handle CONFIG_ACPI=n better from linux/acpi.h linux/acpi.h is the top level header for interfacing with the ACPI sub-system, so acpi_disabled should be up there instead of down in asm/acpi.h -- particularly since asm/acpi.h doesn't exist for all architectures. Same story for acpi_table_parse(), which is a top-level API to Linux/ACPI. This is necessary for building some code that used to always depend on CONFIG_ACPI=y, but will soon also need to build with CONFIG_ACPI=n. Signed-off-by: Feng Tang Signed-off-by: Len Brown --- arch/x86/include/asm/acpi.h | 1 - include/linux/acpi.h | 11 ++++++++++- 2 files changed, 10 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/arch/x86/include/asm/acpi.h b/arch/x86/include/asm/acpi.h index 20d1465a2ab0..4518dc500903 100644 --- a/arch/x86/include/asm/acpi.h +++ b/arch/x86/include/asm/acpi.h @@ -144,7 +144,6 @@ static inline unsigned int acpi_processor_cstate_check(unsigned int max_cstate) #else /* !CONFIG_ACPI */ -#define acpi_disabled 1 #define acpi_lapic 0 #define acpi_ioapic 0 static inline void acpi_noirq_set(void) { } diff --git a/include/linux/acpi.h b/include/linux/acpi.h index 34321cfffeab..3fce811bf9ac 100644 --- a/include/linux/acpi.h +++ b/include/linux/acpi.h @@ -292,7 +292,10 @@ void __init acpi_s4_no_nvs(void); extern acpi_status acpi_pci_osc_control_set(acpi_handle handle, u32 flags); extern void acpi_early_init(void); -#else /* CONFIG_ACPI */ +#else /* !CONFIG_ACPI */ + +#define acpi_disabled 1 + static inline void acpi_early_init(void) { } static inline int early_acpi_boot_init(void) @@ -331,5 +334,11 @@ static inline int acpi_check_mem_region(resource_size_t start, return 0; } +struct acpi_table_header; +static inline int acpi_table_parse(char *id, + int (*handler)(struct acpi_table_header *)) +{ + return -1; +} #endif /* !CONFIG_ACPI */ #endif /*_LINUX_ACPI_H*/ -- cgit v1.2.3 From 117a9ac777f8034d4675b821172d2ff71f6ec47a Mon Sep 17 00:00:00 2001 From: Feng Tang Date: Fri, 14 Aug 2009 15:10:24 -0400 Subject: SFI: create linux/sfi.h include/linux/include/sfi.h defines everything that customers of SFI need to know in order to use the SFI suport in the kernel. The primary API is sfi_table_parse(), where a driver or another part of the kernel can supply a handler to parse the named table. sfi.h also includes the currently defined table signatures and table formats. Signed-off-by: Feng Tang Signed-off-by: Len Brown --- include/linux/sfi.h | 206 ++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 206 insertions(+) create mode 100644 include/linux/sfi.h (limited to 'include') diff --git a/include/linux/sfi.h b/include/linux/sfi.h new file mode 100644 index 000000000000..9a6f7607174e --- /dev/null +++ b/include/linux/sfi.h @@ -0,0 +1,206 @@ +/* sfi.h Simple Firmware Interface */ + +/* + + This file is provided under a dual BSD/GPLv2 license. When using or + redistributing this file, you may do so under either license. + + GPL LICENSE SUMMARY + + Copyright(c) 2009 Intel Corporation. All rights reserved. + + This program is free software; you can redistribute it and/or modify + it under the terms of version 2 of the GNU General Public License as + published by the Free Software Foundation. + + This program is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software + Foundation, Inc., 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA. + The full GNU General Public License is included in this distribution + in the file called LICENSE.GPL. + + BSD LICENSE + + Copyright(c) 2009 Intel Corporation. All rights reserved. + + Redistribution and use in source and binary forms, with or without + modification, are permitted provided that the following conditions + are met: + + * Redistributions of source code must retain the above copyright + notice, this list of conditions and the following disclaimer. + * Redistributions in binary form must reproduce the above copyright + notice, this list of conditions and the following disclaimer in + the documentation and/or other materials provided with the + distribution. + * Neither the name of Intel Corporation nor the names of its + contributors may be used to endorse or promote products derived + from this software without specific prior written permission. + + THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + +*/ + +#ifndef _LINUX_SFI_H +#define _LINUX_SFI_H + +/* Table signatures reserved by the SFI specification */ +#define SFI_SIG_SYST "SYST" +#define SFI_SIG_FREQ "FREQ" +#define SFI_SIG_IDLE "IDLE" +#define SFI_SIG_CPUS "CPUS" +#define SFI_SIG_MTMR "MTMR" +#define SFI_SIG_MRTC "MRTC" +#define SFI_SIG_MMAP "MMAP" +#define SFI_SIG_APIC "APIC" +#define SFI_SIG_XSDT "XSDT" +#define SFI_SIG_WAKE "WAKE" +#define SFI_SIG_SPIB "SPIB" +#define SFI_SIG_I2CB "I2CB" +#define SFI_SIG_GPEM "GPEM" + +#define SFI_SIGNATURE_SIZE 4 +#define SFI_OEM_ID_SIZE 6 +#define SFI_OEM_TABLE_ID_SIZE 8 + +#define SFI_SYST_SEARCH_BEGIN 0x000E0000 +#define SFI_SYST_SEARCH_END 0x000FFFFF + +#define SFI_GET_NUM_ENTRIES(ptable, entry_type) \ + ((ptable->header.len - sizeof(struct sfi_table_header)) / \ + (sizeof(entry_type))) +/* + * Table structures must be byte-packed to match the SFI specification, + * as they are provided by the BIOS. + */ +struct sfi_table_header { + char sig[SFI_SIGNATURE_SIZE]; + u32 len; + u8 rev; + u8 csum; + char oem_id[SFI_OEM_ID_SIZE]; + char oem_table_id[SFI_OEM_TABLE_ID_SIZE]; +} __packed; + +struct sfi_table_simple { + struct sfi_table_header header; + u64 pentry[1]; +} __packed; + +/* Comply with UEFI spec 2.1 */ +struct sfi_mem_entry { + u32 type; + u64 phys_start; + u64 virt_start; + u64 pages; + u64 attrib; +} __packed; + +struct sfi_cpu_table_entry { + u32 apic_id; +} __packed; + +struct sfi_cstate_table_entry { + u32 hint; /* MWAIT hint */ + u32 latency; /* latency in ms */ +} __packed; + +struct sfi_apic_table_entry { + u64 phys_addr; /* phy base addr for APIC reg */ +} __packed; + +struct sfi_freq_table_entry { + u32 freq_mhz; /* in MHZ */ + u32 latency; /* transition latency in ms */ + u32 ctrl_val; /* value to write to PERF_CTL */ +} __packed; + +struct sfi_wake_table_entry { + u64 phys_addr; /* pointer to where the wake vector locates */ +} __packed; + +struct sfi_timer_table_entry { + u64 phys_addr; /* phy base addr for the timer */ + u32 freq_hz; /* in HZ */ + u32 irq; +} __packed; + +struct sfi_rtc_table_entry { + u64 phys_addr; /* phy base addr for the RTC */ + u32 irq; +} __packed; + +struct sfi_spi_table_entry { + u16 host_num; /* attached to host 0, 1...*/ + u16 cs; /* chip select */ + u16 irq_info; + char name[16]; + u8 dev_info[10]; +} __packed; + +struct sfi_i2c_table_entry { + u16 host_num; + u16 addr; /* slave addr */ + u16 irq_info; + char name[16]; + u8 dev_info[10]; +} __packed; + +struct sfi_gpe_table_entry { + u16 logical_id; /* logical id */ + u16 phys_id; /* physical GPE id */ +} __packed; + + +typedef int (*sfi_table_handler) (struct sfi_table_header *table); + +#ifdef CONFIG_SFI +extern void __init sfi_init(void); +extern int __init sfi_platform_init(void); +extern void __init sfi_init_late(void); +extern int sfi_table_parse(char *signature, char *oem_id, char *oem_table_id, + sfi_table_handler handler); + +extern int sfi_disabled; +static inline void disable_sfi(void) +{ + sfi_disabled = 1; +} + +#else /* !CONFIG_SFI */ + +static inline void sfi_init(void) +{ +} + +static inline void sfi_init_late(void) +{ +} + +#define sfi_disabled 0 + +static inline int sfi_table_parse(char *signature, char *oem_id, + char *oem_table_id, + sfi_table_handler handler) +{ + return -1; +} + +#endif /* !CONFIG_SFI */ + +#endif /*_LINUX_SFI_H*/ -- cgit v1.2.3 From 13e82d023c4c3f13ab1e665cbb917a7ebba8935c Mon Sep 17 00:00:00 2001 From: Feng Tang Date: Fri, 14 Aug 2009 15:17:53 -0400 Subject: SFI: add capability to parse ACPI tables Extend SFI to access standard ACPI tables. (eg. the PCI MCFG) using sfi_acpi_table_parse(). Note that this is _not_ a hybrid ACPI + SFI mode. The platform boots in either ACPI mode or SFI mode. SFI runs only with acpi_disabled=1, which can be set at build-time via CONFIG_ACPI=n, or at boot time by the failure to find ACPI platform support. So this extension simply allows SFI-platforms to re-use existing standard table formats that happen to be defined to live in ACPI envelopes. Signed-off-by: Feng Tang Signed-off-by: Len Brown --- drivers/sfi/sfi_acpi.c | 175 +++++++++++++++++++++++++++++++++++++++++++++++ include/linux/sfi_acpi.h | 93 +++++++++++++++++++++++++ 2 files changed, 268 insertions(+) create mode 100644 drivers/sfi/sfi_acpi.c create mode 100644 include/linux/sfi_acpi.h (limited to 'include') diff --git a/drivers/sfi/sfi_acpi.c b/drivers/sfi/sfi_acpi.c new file mode 100644 index 000000000000..34aba30eb84b --- /dev/null +++ b/drivers/sfi/sfi_acpi.c @@ -0,0 +1,175 @@ +/* sfi_acpi.c Simple Firmware Interface - ACPI extensions */ + +/* + + This file is provided under a dual BSD/GPLv2 license. When using or + redistributing this file, you may do so under either license. + + GPL LICENSE SUMMARY + + Copyright(c) 2009 Intel Corporation. All rights reserved. + + This program is free software; you can redistribute it and/or modify + it under the terms of version 2 of the GNU General Public License as + published by the Free Software Foundation. + + This program is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software + Foundation, Inc., 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA. + The full GNU General Public License is included in this distribution + in the file called LICENSE.GPL. + + BSD LICENSE + + Copyright(c) 2009 Intel Corporation. All rights reserved. + + Redistribution and use in source and binary forms, with or without + modification, are permitted provided that the following conditions + are met: + + * Redistributions of source code must retain the above copyright + notice, this list of conditions and the following disclaimer. + * Redistributions in binary form must reproduce the above copyright + notice, this list of conditions and the following disclaimer in + the documentation and/or other materials provided with the + distribution. + * Neither the name of Intel Corporation nor the names of its + contributors may be used to endorse or promote products derived + from this software without specific prior written permission. + + THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + +*/ + +#define KMSG_COMPONENT "SFI" +#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt + +#include +#include + +#include +#include "sfi_core.h" + +/* + * SFI can access ACPI-defined tables via an optional ACPI XSDT. + * + * This allows re-use, and avoids re-definition, of standard tables. + * For example, the "MCFG" table is defined by PCI, reserved by ACPI, + * and is expected to be present many SFI-only systems. + */ + +static struct acpi_table_xsdt *xsdt_va __read_mostly; + +#define XSDT_GET_NUM_ENTRIES(ptable, entry_type) \ + ((ptable->header.length - sizeof(struct acpi_table_header)) / \ + (sizeof(entry_type))) + +static inline struct sfi_table_header *acpi_to_sfi_th( + struct acpi_table_header *th) +{ + return (struct sfi_table_header *)th; +} + +static inline struct acpi_table_header *sfi_to_acpi_th( + struct sfi_table_header *th) +{ + return (struct acpi_table_header *)th; +} + +/* + * sfi_acpi_parse_xsdt() + * + * Parse the ACPI XSDT for later access by sfi_acpi_table_parse(). + */ +static int __init sfi_acpi_parse_xsdt(struct sfi_table_header *th) +{ + struct sfi_table_key key = SFI_ANY_KEY; + int tbl_cnt, i; + void *ret; + + xsdt_va = (struct acpi_table_xsdt *)th; + tbl_cnt = XSDT_GET_NUM_ENTRIES(xsdt_va, u64); + for (i = 0; i < tbl_cnt; i++) { + ret = sfi_check_table(xsdt_va->table_offset_entry[i], &key); + if (IS_ERR(ret)) { + disable_sfi(); + return -1; + } + } + + return 0; +} + +int __init sfi_acpi_init(void) +{ + struct sfi_table_key xsdt_key = { .sig = SFI_SIG_XSDT }; + + sfi_table_parse(SFI_SIG_XSDT, NULL, NULL, sfi_acpi_parse_xsdt); + + /* Only call the get_table to keep the table mapped */ + xsdt_va = (struct acpi_table_xsdt *)sfi_get_table(&xsdt_key); + return 0; +} + +static struct acpi_table_header *sfi_acpi_get_table(struct sfi_table_key *key) +{ + u32 tbl_cnt, i; + void *ret; + + tbl_cnt = XSDT_GET_NUM_ENTRIES(xsdt_va, u64); + for (i = 0; i < tbl_cnt; i++) { + ret = sfi_check_table(xsdt_va->table_offset_entry[i], key); + if (!IS_ERR(ret) && ret) + return sfi_to_acpi_th(ret); + } + + return NULL; +} + +static void sfi_acpi_put_table(struct acpi_table_header *table) +{ + sfi_put_table(acpi_to_sfi_th(table)); +} + +/* + * sfi_acpi_table_parse() + * + * Find specified table in XSDT, run handler on it and return its return value + */ +int sfi_acpi_table_parse(char *signature, char *oem_id, char *oem_table_id, + int(*handler)(struct acpi_table_header *)) +{ + struct acpi_table_header *table = NULL; + struct sfi_table_key key; + int ret = 0; + + if (sfi_disabled) + return -1; + + key.sig = signature; + key.oem_id = oem_id; + key.oem_table_id = oem_table_id; + + table = sfi_acpi_get_table(&key); + if (!table) + return -EINVAL; + + ret = handler(table); + sfi_acpi_put_table(table); + return ret; +} diff --git a/include/linux/sfi_acpi.h b/include/linux/sfi_acpi.h new file mode 100644 index 000000000000..c4a5a8cd4469 --- /dev/null +++ b/include/linux/sfi_acpi.h @@ -0,0 +1,93 @@ +/* sfi.h Simple Firmware Interface */ + +/* + + This file is provided under a dual BSD/GPLv2 license. When using or + redistributing this file, you may do so under either license. + + GPL LICENSE SUMMARY + + Copyright(c) 2009 Intel Corporation. All rights reserved. + + This program is free software; you can redistribute it and/or modify + it under the terms of version 2 of the GNU General Public License as + published by the Free Software Foundation. + + This program is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software + Foundation, Inc., 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA. + The full GNU General Public License is included in this distribution + in the file called LICENSE.GPL. + + BSD LICENSE + + Copyright(c) 2009 Intel Corporation. All rights reserved. + + Redistribution and use in source and binary forms, with or without + modification, are permitted provided that the following conditions + are met: + + * Redistributions of source code must retain the above copyright + notice, this list of conditions and the following disclaimer. + * Redistributions in binary form must reproduce the above copyright + notice, this list of conditions and the following disclaimer in + the documentation and/or other materials provided with the + distribution. + * Neither the name of Intel Corporation nor the names of its + contributors may be used to endorse or promote products derived + from this software without specific prior written permission. + + THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + +*/ + +#ifndef _LINUX_SFI_ACPI_H +#define _LINUX_SFI_ACPI_H + +#ifdef CONFIG_SFI +#include /* struct acpi_table_header */ + +extern int sfi_acpi_table_parse(char *signature, char *oem_id, + char *oem_table_id, + int (*handler)(struct acpi_table_header *)); + +static inline int acpi_sfi_table_parse(char *signature, + int (*handler)(struct acpi_table_header *)) +{ + if (!acpi_table_parse(signature, handler)) + return 0; + + return sfi_acpi_table_parse(signature, NULL, NULL, handler); +} +#else /* !CONFIG_SFI */ + +static inline int sfi_acpi_table_parse(char *signature, char *oem_id, + char *oem_table_id, + int (*handler)(struct acpi_table_header *)) +{ + return -1; +} + +static inline int acpi_sfi_table_parse(char *signature, + int (*handler)(struct acpi_table_header *)) +{ + return acpi_table_parse(signature, handler); +} +#endif /* !CONFIG_SFI */ + +#endif /*_LINUX_SFI_ACPI_H*/ -- cgit v1.2.3 From df19a6267705456f463871ae2aabc44299909d2a Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Fri, 28 Aug 2009 23:48:54 -0700 Subject: tcp: keepalive cleanups Introduce keepalive_probes(tp) helper, and use it, like keepalive_time_when(tp) and keepalive_intvl_when(tp) Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- include/net/tcp.h | 5 +++++ net/ipv4/tcp.c | 6 +++--- net/ipv4/tcp_timer.c | 3 +-- 3 files changed, 9 insertions(+), 5 deletions(-) (limited to 'include') diff --git a/include/net/tcp.h b/include/net/tcp.h index 88af84306471..cbb2a4889fc9 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -1007,6 +1007,11 @@ static inline int keepalive_time_when(const struct tcp_sock *tp) return tp->keepalive_time ? : sysctl_tcp_keepalive_time; } +static inline int keepalive_probes(const struct tcp_sock *tp) +{ + return tp->keepalive_probes ? : sysctl_tcp_keepalive_probes; +} + static inline int tcp_fin_time(const struct sock *sk) { int fin_timeout = tcp_sk(sk)->linger2 ? : sysctl_tcp_fin_timeout; diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c index 91145244ea63..59f69a6c5863 100644 --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@ -2336,13 +2336,13 @@ static int do_tcp_getsockopt(struct sock *sk, int level, val = !!(tp->nonagle&TCP_NAGLE_CORK); break; case TCP_KEEPIDLE: - val = (tp->keepalive_time ? : sysctl_tcp_keepalive_time) / HZ; + val = keepalive_time_when(tp) / HZ; break; case TCP_KEEPINTVL: - val = (tp->keepalive_intvl ? : sysctl_tcp_keepalive_intvl) / HZ; + val = keepalive_intvl_when(tp) / HZ; break; case TCP_KEEPCNT: - val = tp->keepalive_probes ? : sysctl_tcp_keepalive_probes; + val = keepalive_probes(tp); break; case TCP_SYNCNT: val = icsk->icsk_syn_retries ? : sysctl_tcp_syn_retries; diff --git a/net/ipv4/tcp_timer.c b/net/ipv4/tcp_timer.c index b144a26359bc..c520fb6e06d9 100644 --- a/net/ipv4/tcp_timer.c +++ b/net/ipv4/tcp_timer.c @@ -499,8 +499,7 @@ static void tcp_keepalive_timer (unsigned long data) elapsed = tcp_time_stamp - tp->rcv_tstamp; if (elapsed >= keepalive_time_when(tp)) { - if ((!tp->keepalive_probes && icsk->icsk_probes_out >= sysctl_tcp_keepalive_probes) || - (tp->keepalive_probes && icsk->icsk_probes_out >= tp->keepalive_probes)) { + if (icsk->icsk_probes_out >= keepalive_probes(tp)) { tcp_send_active_reset(sk, GFP_ATOMIC); tcp_write_err(sk); goto out; -- cgit v1.2.3 From 0c7d400fafaeab6014504a6a6249f01bac7f7db4 Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Sat, 29 Aug 2009 20:44:04 +1000 Subject: crypto: skcipher - Fix skcipher_dequeue_givcrypt NULL test As struct skcipher_givcrypt_request includes struct crypto_request at a non-zero offset, testing for NULL after converting the pointer returned by crypto_dequeue_request does not work. This can result in IPsec crashes when the queue is depleted. This patch fixes it by doing the pointer conversion only when the return value is non-NULL. In particular, we create a new function __crypto_dequeue_request that does the pointer conversion. Reported-by: Brad Bosch Signed-off-by: Herbert Xu --- crypto/algapi.c | 11 +++++++++-- include/crypto/algapi.h | 1 + include/crypto/internal/skcipher.h | 4 ++-- 3 files changed, 12 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/crypto/algapi.c b/crypto/algapi.c index 56c62e2858d5..df0863d56995 100644 --- a/crypto/algapi.c +++ b/crypto/algapi.c @@ -692,7 +692,7 @@ out: } EXPORT_SYMBOL_GPL(crypto_enqueue_request); -struct crypto_async_request *crypto_dequeue_request(struct crypto_queue *queue) +void *__crypto_dequeue_request(struct crypto_queue *queue, unsigned int offset) { struct list_head *request; @@ -707,7 +707,14 @@ struct crypto_async_request *crypto_dequeue_request(struct crypto_queue *queue) request = queue->list.next; list_del(request); - return list_entry(request, struct crypto_async_request, list); + return (char *)list_entry(request, struct crypto_async_request, list) - + offset; +} +EXPORT_SYMBOL_GPL(__crypto_dequeue_request); + +struct crypto_async_request *crypto_dequeue_request(struct crypto_queue *queue) +{ + return __crypto_dequeue_request(queue, 0); } EXPORT_SYMBOL_GPL(crypto_dequeue_request); diff --git a/include/crypto/algapi.h b/include/crypto/algapi.h index 010545436efa..5a2bd1cc9656 100644 --- a/include/crypto/algapi.h +++ b/include/crypto/algapi.h @@ -137,6 +137,7 @@ struct crypto_instance *crypto_alloc_instance(const char *name, void crypto_init_queue(struct crypto_queue *queue, unsigned int max_qlen); int crypto_enqueue_request(struct crypto_queue *queue, struct crypto_async_request *request); +void *__crypto_dequeue_request(struct crypto_queue *queue, unsigned int offset); struct crypto_async_request *crypto_dequeue_request(struct crypto_queue *queue); int crypto_tfm_in_queue(struct crypto_queue *queue, struct crypto_tfm *tfm); diff --git a/include/crypto/internal/skcipher.h b/include/crypto/internal/skcipher.h index 2ba42cd7d6aa..3a748a6bf772 100644 --- a/include/crypto/internal/skcipher.h +++ b/include/crypto/internal/skcipher.h @@ -79,8 +79,8 @@ static inline int skcipher_enqueue_givcrypt( static inline struct skcipher_givcrypt_request *skcipher_dequeue_givcrypt( struct crypto_queue *queue) { - return container_of(ablkcipher_dequeue_request(queue), - struct skcipher_givcrypt_request, creq); + return __crypto_dequeue_request( + queue, offsetof(struct skcipher_givcrypt_request, creq.base)); } static inline void *skcipher_givcrypt_reqctx( -- cgit v1.2.3 From 2b022e3d4bf9885f781221c59d86283a2cdfc2ed Mon Sep 17 00:00:00 2001 From: Xiao Guangrong Date: Mon, 10 Aug 2009 10:48:59 +0800 Subject: timers: Add tracepoints for timer_list timers Add tracepoints which cover the timer life cycle. The tracepoints are integrated with the already existing debug_object debug points as far as possible. Based on patches from Mathieu: http://marc.info/?l=linux-kernel&m=123791201816247&w=2 and Anton: http://marc.info/?l=linux-kernel&m=124331396919301&w=2 [ tglx: Fixed timeout value in timer_start tracepoint, massaged comments and made the printk's more readable ] Signed-off-by: Xiao Guangrong Cc: Anton Blanchard Cc: Steven Rostedt Cc: Frederic Weisbecker Cc: Mathieu Desnoyers Cc: Peter Zijlstra Cc: KOSAKI Motohiro Cc: Zhaolei LKML-Reference: <4A7F8A9B.3040201@cn.fujitsu.com> Signed-off-by: Thomas Gleixner --- include/trace/events/timer.h | 137 +++++++++++++++++++++++++++++++++++++++++++ kernel/timer.c | 32 ++++++++-- 2 files changed, 165 insertions(+), 4 deletions(-) create mode 100644 include/trace/events/timer.h (limited to 'include') diff --git a/include/trace/events/timer.h b/include/trace/events/timer.h new file mode 100644 index 000000000000..725892a93b49 --- /dev/null +++ b/include/trace/events/timer.h @@ -0,0 +1,137 @@ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM timer + +#if !defined(_TRACE_TIMER_H) || defined(TRACE_HEADER_MULTI_READ) +#define _TRACE_TIMER_H + +#include +#include + +/** + * timer_init - called when the timer is initialized + * @timer: pointer to struct timer_list + */ +TRACE_EVENT(timer_init, + + TP_PROTO(struct timer_list *timer), + + TP_ARGS(timer), + + TP_STRUCT__entry( + __field( void *, timer ) + ), + + TP_fast_assign( + __entry->timer = timer; + ), + + TP_printk("timer %p", __entry->timer) +); + +/** + * timer_start - called when the timer is started + * @timer: pointer to struct timer_list + * @expires: the timers expiry time + */ +TRACE_EVENT(timer_start, + + TP_PROTO(struct timer_list *timer, unsigned long expires), + + TP_ARGS(timer, expires), + + TP_STRUCT__entry( + __field( void *, timer ) + __field( void *, function ) + __field( unsigned long, expires ) + __field( unsigned long, now ) + ), + + TP_fast_assign( + __entry->timer = timer; + __entry->function = timer->function; + __entry->expires = expires; + __entry->now = jiffies; + ), + + TP_printk("timer %p: func %pf, expires %lu, timeout %ld", + __entry->timer, __entry->function, __entry->expires, + (long)__entry->expires - __entry->now) +); + +/** + * timer_expire_entry - called immediately before the timer callback + * @timer: pointer to struct timer_list + * + * Allows to determine the timer latency. + */ +TRACE_EVENT(timer_expire_entry, + + TP_PROTO(struct timer_list *timer), + + TP_ARGS(timer), + + TP_STRUCT__entry( + __field( void *, timer ) + __field( unsigned long, now ) + ), + + TP_fast_assign( + __entry->timer = timer; + __entry->now = jiffies; + ), + + TP_printk("timer %p: now %lu", __entry->timer, __entry->now) +); + +/** + * timer_expire_exit - called immediately after the timer callback returns + * @timer: pointer to struct timer_list + * + * When used in combination with the timer_expire_entry tracepoint we can + * determine the runtime of the timer callback function. + * + * NOTE: Do NOT derefernce timer in TP_fast_assign. The pointer might + * be invalid. We solely track the pointer. + */ +TRACE_EVENT(timer_expire_exit, + + TP_PROTO(struct timer_list *timer), + + TP_ARGS(timer), + + TP_STRUCT__entry( + __field(void *, timer ) + ), + + TP_fast_assign( + __entry->timer = timer; + ), + + TP_printk("timer %p", __entry->timer) +); + +/** + * timer_cancel - called when the timer is canceled + * @timer: pointer to struct timer_list + */ +TRACE_EVENT(timer_cancel, + + TP_PROTO(struct timer_list *timer), + + TP_ARGS(timer), + + TP_STRUCT__entry( + __field( void *, timer ) + ), + + TP_fast_assign( + __entry->timer = timer; + ), + + TP_printk("timer %p", __entry->timer) +); + +#endif /* _TRACE_TIMER_H */ + +/* This part must be outside protection */ +#include diff --git a/kernel/timer.c b/kernel/timer.c index 8e92be654dad..a7352b00703c 100644 --- a/kernel/timer.c +++ b/kernel/timer.c @@ -46,6 +46,9 @@ #include #include +#define CREATE_TRACE_POINTS +#include + u64 jiffies_64 __cacheline_aligned_in_smp = INITIAL_JIFFIES; EXPORT_SYMBOL(jiffies_64); @@ -521,6 +524,25 @@ static inline void debug_timer_activate(struct timer_list *timer) { } static inline void debug_timer_deactivate(struct timer_list *timer) { } #endif +static inline void debug_init(struct timer_list *timer) +{ + debug_timer_init(timer); + trace_timer_init(timer); +} + +static inline void +debug_activate(struct timer_list *timer, unsigned long expires) +{ + debug_timer_activate(timer); + trace_timer_start(timer, expires); +} + +static inline void debug_deactivate(struct timer_list *timer) +{ + debug_timer_deactivate(timer); + trace_timer_cancel(timer); +} + static void __init_timer(struct timer_list *timer, const char *name, struct lock_class_key *key) @@ -549,7 +571,7 @@ void init_timer_key(struct timer_list *timer, const char *name, struct lock_class_key *key) { - debug_timer_init(timer); + debug_init(timer); __init_timer(timer, name, key); } EXPORT_SYMBOL(init_timer_key); @@ -568,7 +590,7 @@ static inline void detach_timer(struct timer_list *timer, { struct list_head *entry = &timer->entry; - debug_timer_deactivate(timer); + debug_deactivate(timer); __list_del(entry->prev, entry->next); if (clear_pending) @@ -632,7 +654,7 @@ __mod_timer(struct timer_list *timer, unsigned long expires, goto out_unlock; } - debug_timer_activate(timer); + debug_activate(timer, expires); new_base = __get_cpu_var(tvec_bases); @@ -787,7 +809,7 @@ void add_timer_on(struct timer_list *timer, int cpu) BUG_ON(timer_pending(timer) || !timer->function); spin_lock_irqsave(&base->lock, flags); timer_set_base(timer, base); - debug_timer_activate(timer); + debug_activate(timer, timer->expires); if (time_before(timer->expires, base->next_timer) && !tbase_get_deferrable(timer->base)) base->next_timer = timer->expires; @@ -1000,7 +1022,9 @@ static inline void __run_timers(struct tvec_base *base) */ lock_map_acquire(&lockdep_map); + trace_timer_expire_entry(timer); fn(data); + trace_timer_expire_exit(timer); lock_map_release(&lockdep_map); -- cgit v1.2.3 From c6a2a1770245f654f35f60e1458d4356680f9519 Mon Sep 17 00:00:00 2001 From: Xiao Guangrong Date: Mon, 10 Aug 2009 10:51:23 +0800 Subject: hrtimer: Add tracepoint for hrtimers Add tracepoints which cover the life cycle of a hrtimer. The tracepoints are integrated with the already existing debug_object debug points as far as possible. [ tglx: Fixed comments, made output conistent, easier to read and parse. Fixed output for 32bit archs which do not use the scalar representation of ktime_t. Hand current time to trace_hrtimer_expiry_entry instead of calling get_time() inside of the trace assignment. ] Signed-off-by: Xiao Guangrong Cc: Steven Rostedt Cc: Frederic Weisbecker Cc: Mathieu Desnoyers Cc: Anton Blanchard Cc: Peter Zijlstra Cc: KOSAKI Motohiro Cc: Zhaolei LKML-Reference: <4A7F8B2B.5020908@cn.fujitsu.com> Signed-off-by: Thomas Gleixner --- include/trace/events/timer.h | 139 +++++++++++++++++++++++++++++++++++++++++++ kernel/hrtimer.c | 40 ++++++++++--- 2 files changed, 171 insertions(+), 8 deletions(-) (limited to 'include') diff --git a/include/trace/events/timer.h b/include/trace/events/timer.h index 725892a93b49..df3c07fa0cb8 100644 --- a/include/trace/events/timer.h +++ b/include/trace/events/timer.h @@ -5,6 +5,7 @@ #define _TRACE_TIMER_H #include +#include #include /** @@ -131,6 +132,144 @@ TRACE_EVENT(timer_cancel, TP_printk("timer %p", __entry->timer) ); +/** + * hrtimer_init - called when the hrtimer is initialized + * @timer: pointer to struct hrtimer + * @clockid: the hrtimers clock + * @mode: the hrtimers mode + */ +TRACE_EVENT(hrtimer_init, + + TP_PROTO(struct hrtimer *timer, clockid_t clockid, + enum hrtimer_mode mode), + + TP_ARGS(timer, clockid, mode), + + TP_STRUCT__entry( + __field( void *, timer ) + __field( clockid_t, clockid ) + __field( enum hrtimer_mode, mode ) + ), + + TP_fast_assign( + __entry->timer = timer; + __entry->clockid = clockid; + __entry->mode = mode; + ), + + TP_printk("hrtimer %p, clockid %s, mode %s", __entry->timer, + __entry->clockid == CLOCK_REALTIME ? + "CLOCK_REALTIME" : "CLOCK_MONOTONIC", + __entry->mode == HRTIMER_MODE_ABS ? + "HRTIMER_MODE_ABS" : "HRTIMER_MODE_REL") +); + +/** + * hrtimer_start - called when the hrtimer is started + * @timer: pointer to struct hrtimer + */ +TRACE_EVENT(hrtimer_start, + + TP_PROTO(struct hrtimer *timer), + + TP_ARGS(timer), + + TP_STRUCT__entry( + __field( void *, timer ) + __field( void *, function ) + __field( s64, expires ) + __field( s64, softexpires ) + ), + + TP_fast_assign( + __entry->timer = timer; + __entry->function = timer->function; + __entry->expires = hrtimer_get_expires(timer).tv64; + __entry->softexpires = hrtimer_get_softexpires(timer).tv64; + ), + + TP_printk("hrtimer %p, func %pf, expires %llu, softexpires %llu", + __entry->timer, __entry->function, + (unsigned long long)ktime_to_ns((ktime_t) { + .tv64 = __entry->expires }), + (unsigned long long)ktime_to_ns((ktime_t) { + .tv64 = __entry->softexpires })) +); + +/** + * htimmer_expire_entry - called immediately before the hrtimer callback + * @timer: pointer to struct hrtimer + * @now: pointer to variable which contains current time of the + * timers base. + * + * Allows to determine the timer latency. + */ +TRACE_EVENT(hrtimer_expire_entry, + + TP_PROTO(struct hrtimer *timer, ktime_t *now), + + TP_ARGS(timer, now), + + TP_STRUCT__entry( + __field( void *, timer ) + __field( s64, now ) + ), + + TP_fast_assign( + __entry->timer = timer; + __entry->now = now->tv64; + ), + + TP_printk("hrtimer %p, now %llu", __entry->timer, + (unsigned long long)ktime_to_ns((ktime_t) { + .tv64 = __entry->now })) + ); + +/** + * hrtimer_expire_exit - called immediately after the hrtimer callback returns + * @timer: pointer to struct hrtimer + * + * When used in combination with the hrtimer_expire_entry tracepoint we can + * determine the runtime of the callback function. + */ +TRACE_EVENT(hrtimer_expire_exit, + + TP_PROTO(struct hrtimer *timer), + + TP_ARGS(timer), + + TP_STRUCT__entry( + __field( void *, timer ) + ), + + TP_fast_assign( + __entry->timer = timer; + ), + + TP_printk("hrtimer %p", __entry->timer) +); + +/** + * hrtimer_cancel - called when the hrtimer is canceled + * @timer: pointer to struct hrtimer + */ +TRACE_EVENT(hrtimer_cancel, + + TP_PROTO(struct hrtimer *timer), + + TP_ARGS(timer), + + TP_STRUCT__entry( + __field( void *, timer ) + ), + + TP_fast_assign( + __entry->timer = timer; + ), + + TP_printk("hrtimer %p", __entry->timer) +); + #endif /* _TRACE_TIMER_H */ /* This part must be outside protection */ diff --git a/kernel/hrtimer.c b/kernel/hrtimer.c index e2f91ecc01a8..b44d1b07377b 100644 --- a/kernel/hrtimer.c +++ b/kernel/hrtimer.c @@ -48,6 +48,8 @@ #include +#include + /* * The timer bases: * @@ -441,6 +443,26 @@ static inline void debug_hrtimer_activate(struct hrtimer *timer) { } static inline void debug_hrtimer_deactivate(struct hrtimer *timer) { } #endif +static inline void +debug_init(struct hrtimer *timer, clockid_t clockid, + enum hrtimer_mode mode) +{ + debug_hrtimer_init(timer); + trace_hrtimer_init(timer, clockid, mode); +} + +static inline void debug_activate(struct hrtimer *timer) +{ + debug_hrtimer_activate(timer); + trace_hrtimer_start(timer); +} + +static inline void debug_deactivate(struct hrtimer *timer) +{ + debug_hrtimer_deactivate(timer); + trace_hrtimer_cancel(timer); +} + /* High resolution timer related functions */ #ifdef CONFIG_HIGH_RES_TIMERS @@ -797,7 +819,7 @@ static int enqueue_hrtimer(struct hrtimer *timer, struct hrtimer *entry; int leftmost = 1; - debug_hrtimer_activate(timer); + debug_activate(timer); /* * Find the right place in the rbtree: @@ -883,7 +905,7 @@ remove_hrtimer(struct hrtimer *timer, struct hrtimer_clock_base *base) * reprogramming happens in the interrupt handler. This is a * rare case and less expensive than a smp call. */ - debug_hrtimer_deactivate(timer); + debug_deactivate(timer); timer_stats_hrtimer_clear_start_info(timer); reprogram = base->cpu_base == &__get_cpu_var(hrtimer_bases); __remove_hrtimer(timer, base, HRTIMER_STATE_INACTIVE, @@ -1116,7 +1138,7 @@ static void __hrtimer_init(struct hrtimer *timer, clockid_t clock_id, void hrtimer_init(struct hrtimer *timer, clockid_t clock_id, enum hrtimer_mode mode) { - debug_hrtimer_init(timer); + debug_init(timer, clock_id, mode); __hrtimer_init(timer, clock_id, mode); } EXPORT_SYMBOL_GPL(hrtimer_init); @@ -1140,7 +1162,7 @@ int hrtimer_get_res(const clockid_t which_clock, struct timespec *tp) } EXPORT_SYMBOL_GPL(hrtimer_get_res); -static void __run_hrtimer(struct hrtimer *timer) +static void __run_hrtimer(struct hrtimer *timer, ktime_t *now) { struct hrtimer_clock_base *base = timer->base; struct hrtimer_cpu_base *cpu_base = base->cpu_base; @@ -1149,7 +1171,7 @@ static void __run_hrtimer(struct hrtimer *timer) WARN_ON(!irqs_disabled()); - debug_hrtimer_deactivate(timer); + debug_deactivate(timer); __remove_hrtimer(timer, base, HRTIMER_STATE_CALLBACK, 0); timer_stats_account_hrtimer(timer); fn = timer->function; @@ -1160,7 +1182,9 @@ static void __run_hrtimer(struct hrtimer *timer) * the timer base. */ spin_unlock(&cpu_base->lock); + trace_hrtimer_expire_entry(timer, now); restart = fn(timer); + trace_hrtimer_expire_exit(timer); spin_lock(&cpu_base->lock); /* @@ -1271,7 +1295,7 @@ void hrtimer_interrupt(struct clock_event_device *dev) break; } - __run_hrtimer(timer); + __run_hrtimer(timer, &basenow); } base++; } @@ -1393,7 +1417,7 @@ void hrtimer_run_queues(void) hrtimer_get_expires_tv64(timer)) break; - __run_hrtimer(timer); + __run_hrtimer(timer, &base->softirq_time); } spin_unlock(&cpu_base->lock); } @@ -1569,7 +1593,7 @@ static void migrate_hrtimer_list(struct hrtimer_clock_base *old_base, while ((node = rb_first(&old_base->active))) { timer = rb_entry(node, struct hrtimer, node); BUG_ON(hrtimer_callback_running(timer)); - debug_hrtimer_deactivate(timer); + debug_deactivate(timer); /* * Mark it as STATE_MIGRATE not INACTIVE otherwise the -- cgit v1.2.3 From 3f0a525ebf4b8ef041a332bbe4a73aee94bb064b Mon Sep 17 00:00:00 2001 From: Xiao Guangrong Date: Mon, 10 Aug 2009 10:52:30 +0800 Subject: itimers: Add tracepoints for itimer Add tracepoints for all itimer variants: ITIMER_REAL, ITIMER_VIRTUAL and ITIMER_PROF. [ tglx: Fixed comments and made the output more readable, parseable and consistent. Replaced pid_vnr by pid_nr because the hrtimer callback can happen in any namespace ] Signed-off-by: Xiao Guangrong Cc: Steven Rostedt Cc: Frederic Weisbecker Cc: Mathieu Desnoyers Cc: Anton Blanchard Cc: Peter Zijlstra Cc: KOSAKI Motohiro Cc: Zhaolei LKML-Reference: <4A7F8B6E.2010109@cn.fujitsu.com> Signed-off-by: Thomas Gleixner --- include/trace/events/timer.h | 66 ++++++++++++++++++++++++++++++++++++++++++++ kernel/itimer.c | 5 ++++ kernel/posix-cpu-timers.c | 7 ++++- 3 files changed, 77 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/trace/events/timer.h b/include/trace/events/timer.h index df3c07fa0cb8..1844c48d640e 100644 --- a/include/trace/events/timer.h +++ b/include/trace/events/timer.h @@ -270,6 +270,72 @@ TRACE_EVENT(hrtimer_cancel, TP_printk("hrtimer %p", __entry->timer) ); +/** + * itimer_state - called when itimer is started or canceled + * @which: name of the interval timer + * @value: the itimers value, itimer is canceled if value->it_value is + * zero, otherwise it is started + * @expires: the itimers expiry time + */ +TRACE_EVENT(itimer_state, + + TP_PROTO(int which, const struct itimerval *const value, + cputime_t expires), + + TP_ARGS(which, value, expires), + + TP_STRUCT__entry( + __field( int, which ) + __field( cputime_t, expires ) + __field( long, value_sec ) + __field( long, value_usec ) + __field( long, interval_sec ) + __field( long, interval_usec ) + ), + + TP_fast_assign( + __entry->which = which; + __entry->expires = expires; + __entry->value_sec = value->it_value.tv_sec; + __entry->value_usec = value->it_value.tv_usec; + __entry->interval_sec = value->it_interval.tv_sec; + __entry->interval_usec = value->it_interval.tv_usec; + ), + + TP_printk("which %d, expires %lu, it_value %lu.%lu, it_interval %lu.%lu", + __entry->which, __entry->expires, + __entry->value_sec, __entry->value_usec, + __entry->interval_sec, __entry->interval_usec) +); + +/** + * itimer_expire - called when itimer expires + * @which: type of the interval timer + * @pid: pid of the process which owns the timer + * @now: current time, used to calculate the latency of itimer + */ +TRACE_EVENT(itimer_expire, + + TP_PROTO(int which, struct pid *pid, cputime_t now), + + TP_ARGS(which, pid, now), + + TP_STRUCT__entry( + __field( int , which ) + __field( pid_t, pid ) + __field( cputime_t, now ) + ), + + TP_fast_assign( + __entry->which = which; + __entry->now = now; + __entry->pid = pid_nr(pid); + ), + + TP_printk("which %d, pid %d, now %lu", __entry->which, + (int) __entry->pid, __entry->now) +); + #endif /* _TRACE_TIMER_H */ /* This part must be outside protection */ diff --git a/kernel/itimer.c b/kernel/itimer.c index 8078a32d3b10..b03451ede528 100644 --- a/kernel/itimer.c +++ b/kernel/itimer.c @@ -12,6 +12,7 @@ #include #include #include +#include #include @@ -122,6 +123,7 @@ enum hrtimer_restart it_real_fn(struct hrtimer *timer) struct signal_struct *sig = container_of(timer, struct signal_struct, real_timer); + trace_itimer_expire(ITIMER_REAL, sig->leader_pid, 0); kill_pid_info(SIGALRM, SEND_SIG_PRIV, sig->leader_pid); return HRTIMER_NORESTART; @@ -166,6 +168,8 @@ static void set_cpu_itimer(struct task_struct *tsk, unsigned int clock_id, } it->expires = nval; it->incr = ninterval; + trace_itimer_state(clock_id == CPUCLOCK_VIRT ? + ITIMER_VIRTUAL : ITIMER_PROF, value, nval); spin_unlock_irq(&tsk->sighand->siglock); @@ -217,6 +221,7 @@ again: } else tsk->signal->it_real_incr.tv64 = 0; + trace_itimer_state(ITIMER_REAL, value, 0); spin_unlock_irq(&tsk->sighand->siglock); break; case ITIMER_VIRTUAL: diff --git a/kernel/posix-cpu-timers.c b/kernel/posix-cpu-timers.c index 12161f74744e..5c9dc228747b 100644 --- a/kernel/posix-cpu-timers.c +++ b/kernel/posix-cpu-timers.c @@ -8,6 +8,7 @@ #include #include #include +#include /* * Called after updating RLIMIT_CPU to set timer expiration if necessary. @@ -1090,9 +1091,13 @@ static void check_cpu_itimer(struct task_struct *tsk, struct cpu_itimer *it, cputime_one_jiffy); it->error -= onecputick; } - } else + } else { it->expires = cputime_zero; + } + trace_itimer_expire(signo == SIGPROF ? + ITIMER_PROF : ITIMER_VIRTUAL, + tsk->signal->leader_pid, cur_time); __group_send_sig_info(signo, SEND_SIG_PRIV, tsk); } -- cgit v1.2.3 From dd5d19bafd90d33043a4a14b2e2d98612caa293c Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Thu, 27 Aug 2009 14:58:16 -0700 Subject: rcu: Create rcutree plugins to handle hotplug CPU for multi-level trees When offlining CPUs from a multi-level tree, there is the possibility of offlining the last CPU from a given node when there are preempted RCU read-side critical sections that started life on one of the CPUs on that node. In this case, the corresponding tasks will be enqueued via the task_struct's rcu_node_entry list_head onto one of the rcu_node's blocked_tasks[] lists. These tasks need to be moved somewhere else so that they will prevent the current grace period from ending. That somewhere is the root rcu_node. Signed-off-by: Paul E. McKenney Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <20090827215816.GA30472@linux.vnet.ibm.com> Signed-off-by: Ingo Molnar --- include/linux/init_task.h | 2 +- include/linux/sched.h | 4 +-- kernel/rcutree.c | 2 ++ kernel/rcutree_plugin.h | 69 +++++++++++++++++++++++++++++++++++++++++++---- 4 files changed, 69 insertions(+), 8 deletions(-) (limited to 'include') diff --git a/include/linux/init_task.h b/include/linux/init_task.h index 79d4baee31b6..9e7f2e8fc66e 100644 --- a/include/linux/init_task.h +++ b/include/linux/init_task.h @@ -98,7 +98,7 @@ extern struct group_info init_groups; #define INIT_TASK_RCU_PREEMPT(tsk) \ .rcu_read_lock_nesting = 0, \ .rcu_read_unlock_special = 0, \ - .rcu_blocked_cpu = -1, \ + .rcu_blocked_node = NULL, \ .rcu_node_entry = LIST_HEAD_INIT(tsk.rcu_node_entry), #else #define INIT_TASK_RCU_PREEMPT(tsk) diff --git a/include/linux/sched.h b/include/linux/sched.h index bfca26d63b13..3fe03151a8e6 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -1208,7 +1208,7 @@ struct task_struct { #ifdef CONFIG_TREE_PREEMPT_RCU int rcu_read_lock_nesting; char rcu_read_unlock_special; - int rcu_blocked_cpu; + void *rcu_blocked_node; struct list_head rcu_node_entry; #endif /* #ifdef CONFIG_TREE_PREEMPT_RCU */ @@ -1735,7 +1735,7 @@ static inline void rcu_copy_process(struct task_struct *p) { p->rcu_read_lock_nesting = 0; p->rcu_read_unlock_special = 0; - p->rcu_blocked_cpu = -1; + p->rcu_blocked_node = NULL; INIT_LIST_HEAD(&p->rcu_node_entry); } diff --git a/kernel/rcutree.c b/kernel/rcutree.c index fee6316a8673..d903e2f2b840 100644 --- a/kernel/rcutree.c +++ b/kernel/rcutree.c @@ -81,6 +81,7 @@ struct rcu_state rcu_bh_state = RCU_STATE_INITIALIZER(rcu_bh_state); DEFINE_PER_CPU(struct rcu_data, rcu_bh_data); extern long rcu_batches_completed_sched(void); +static struct rcu_node *rcu_get_root(struct rcu_state *rsp); static void cpu_quiet_msk(unsigned long mask, struct rcu_state *rsp, struct rcu_node *rnp, unsigned long flags); static void cpu_quiet_msk_finish(struct rcu_state *rsp, unsigned long flags); @@ -876,6 +877,7 @@ static void __rcu_offline_cpu(int cpu, struct rcu_state *rsp) spin_unlock(&rnp->lock); /* irqs remain disabled. */ break; } + rcu_preempt_offline_tasks(rsp, rnp); mask = rnp->grpmask; spin_unlock(&rnp->lock); /* irqs remain disabled. */ rnp = rnp->parent; diff --git a/kernel/rcutree_plugin.h b/kernel/rcutree_plugin.h index 201334cdc200..04343bee646d 100644 --- a/kernel/rcutree_plugin.h +++ b/kernel/rcutree_plugin.h @@ -92,7 +92,7 @@ static void rcu_preempt_qs(int cpu) rnp = rdp->mynode; spin_lock(&rnp->lock); t->rcu_read_unlock_special |= RCU_READ_UNLOCK_BLOCKED; - t->rcu_blocked_cpu = cpu; + t->rcu_blocked_node = (void *)rnp; /* * If this CPU has already checked in, then this task @@ -170,12 +170,21 @@ static void rcu_read_unlock_special(struct task_struct *t) if (special & RCU_READ_UNLOCK_BLOCKED) { t->rcu_read_unlock_special &= ~RCU_READ_UNLOCK_BLOCKED; - /* Remove this task from the list it blocked on. */ - rnp = rcu_preempt_state.rda[t->rcu_blocked_cpu]->mynode; - spin_lock(&rnp->lock); + /* + * Remove this task from the list it blocked on. The + * task can migrate while we acquire the lock, but at + * most one time. So at most two passes through loop. + */ + for (;;) { + rnp = (struct rcu_node *)t->rcu_blocked_node; + spin_lock(&rnp->lock); + if (rnp == (struct rcu_node *)t->rcu_blocked_node) + break; + spin_unlock(&rnp->lock); + } empty = list_empty(&rnp->blocked_tasks[rnp->gpnum & 0x1]); list_del_init(&t->rcu_node_entry); - t->rcu_blocked_cpu = -1; + t->rcu_blocked_node = NULL; /* * If this was the last task on the current list, and if @@ -261,6 +270,47 @@ static int rcu_preempted_readers(struct rcu_node *rnp) #ifdef CONFIG_HOTPLUG_CPU +/* + * Handle tasklist migration for case in which all CPUs covered by the + * specified rcu_node have gone offline. Move them up to the root + * rcu_node. The reason for not just moving them to the immediate + * parent is to remove the need for rcu_read_unlock_special() to + * make more than two attempts to acquire the target rcu_node's lock. + * + * The caller must hold rnp->lock with irqs disabled. + */ +static void rcu_preempt_offline_tasks(struct rcu_state *rsp, + struct rcu_node *rnp) +{ + int i; + struct list_head *lp; + struct list_head *lp_root; + struct rcu_node *rnp_root = rcu_get_root(rsp); + struct task_struct *tp; + + if (rnp == rnp_root) + return; /* Shouldn't happen: at least one CPU online. */ + + /* + * Move tasks up to root rcu_node. Rely on the fact that the + * root rcu_node can be at most one ahead of the rest of the + * rcu_nodes in terms of gp_num value. This fact allows us to + * move the blocked_tasks[] array directly, element by element. + */ + for (i = 0; i < 2; i++) { + lp = &rnp->blocked_tasks[i]; + lp_root = &rnp_root->blocked_tasks[i]; + while (!list_empty(lp)) { + tp = list_entry(lp->next, typeof(*tp), rcu_node_entry); + spin_lock(&rnp_root->lock); /* irqs already disabled */ + list_del(&tp->rcu_node_entry); + tp->rcu_blocked_node = rnp_root; + list_add(&tp->rcu_node_entry, lp_root); + spin_unlock(&rnp_root->lock); /* irqs remain disabled */ + } + } +} + /* * Do CPU-offline processing for preemptable RCU. */ @@ -409,6 +459,15 @@ static int rcu_preempted_readers(struct rcu_node *rnp) #ifdef CONFIG_HOTPLUG_CPU +/* + * Because preemptable RCU does not exist, it never needs to migrate + * tasks that were blocked within RCU read-side critical sections. + */ +static void rcu_preempt_offline_tasks(struct rcu_state *rsp, + struct rcu_node *rnp) +{ +} + /* * Because preemptable RCU does not exist, it never needs CPU-offline * processing. -- cgit v1.2.3 From 868489660dabc0c28087cca3dbc1adbbc398c6fe Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Thu, 27 Aug 2009 15:00:12 -0700 Subject: rcu: Changes from reviews: avoid casts, fix/add warnings, improve comments Changes suggested by review comments from Josh Triplett and Mathieu Desnoyers. Signed-off-by: Paul E. McKenney Acked-by: Josh Triplett Acked-by: Mathieu Desnoyers Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <20090827220012.GA30525@linux.vnet.ibm.com> Signed-off-by: Ingo Molnar --- include/linux/sched.h | 4 +++- kernel/rcutree.c | 13 ++++++------- kernel/rcutree.h | 2 ++ kernel/rcutree_plugin.h | 10 ++++++---- 4 files changed, 17 insertions(+), 12 deletions(-) (limited to 'include') diff --git a/include/linux/sched.h b/include/linux/sched.h index 3fe03151a8e6..855fd0d3f174 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -1163,6 +1163,8 @@ struct sched_rt_entity { #endif }; +struct rcu_node; + struct task_struct { volatile long state; /* -1 unrunnable, 0 runnable, >0 stopped */ void *stack; @@ -1208,7 +1210,7 @@ struct task_struct { #ifdef CONFIG_TREE_PREEMPT_RCU int rcu_read_lock_nesting; char rcu_read_unlock_special; - void *rcu_blocked_node; + struct rcu_node *rcu_blocked_node; struct list_head rcu_node_entry; #endif /* #ifdef CONFIG_TREE_PREEMPT_RCU */ diff --git a/kernel/rcutree.c b/kernel/rcutree.c index d903e2f2b840..71bc79791cd9 100644 --- a/kernel/rcutree.c +++ b/kernel/rcutree.c @@ -229,7 +229,6 @@ static int rcu_implicit_offline_qs(struct rcu_data *rdp) #endif /* #ifdef CONFIG_SMP */ #ifdef CONFIG_NO_HZ -static DEFINE_RATELIMIT_STATE(rcu_rs, 10 * HZ, 5); /** * rcu_enter_nohz - inform RCU that current CPU is entering nohz @@ -249,7 +248,7 @@ void rcu_enter_nohz(void) rdtp = &__get_cpu_var(rcu_dynticks); rdtp->dynticks++; rdtp->dynticks_nesting--; - WARN_ON_RATELIMIT(rdtp->dynticks & 0x1, &rcu_rs); + WARN_ON_ONCE(rdtp->dynticks & 0x1); local_irq_restore(flags); } @@ -268,7 +267,7 @@ void rcu_exit_nohz(void) rdtp = &__get_cpu_var(rcu_dynticks); rdtp->dynticks++; rdtp->dynticks_nesting++; - WARN_ON_RATELIMIT(!(rdtp->dynticks & 0x1), &rcu_rs); + WARN_ON_ONCE(!(rdtp->dynticks & 0x1)); local_irq_restore(flags); smp_mb(); /* CPUs seeing ++ must see later RCU read-side crit sects */ } @@ -287,7 +286,7 @@ void rcu_nmi_enter(void) if (rdtp->dynticks & 0x1) return; rdtp->dynticks_nmi++; - WARN_ON_RATELIMIT(!(rdtp->dynticks_nmi & 0x1), &rcu_rs); + WARN_ON_ONCE(!(rdtp->dynticks_nmi & 0x1)); smp_mb(); /* CPUs seeing ++ must see later RCU read-side crit sects */ } @@ -306,7 +305,7 @@ void rcu_nmi_exit(void) return; smp_mb(); /* CPUs seeing ++ must see prior RCU read-side crit sects */ rdtp->dynticks_nmi++; - WARN_ON_RATELIMIT(rdtp->dynticks_nmi & 0x1, &rcu_rs); + WARN_ON_ONCE(rdtp->dynticks_nmi & 0x1); } /** @@ -322,7 +321,7 @@ void rcu_irq_enter(void) if (rdtp->dynticks_nesting++) return; rdtp->dynticks++; - WARN_ON_RATELIMIT(!(rdtp->dynticks & 0x1), &rcu_rs); + WARN_ON_ONCE(!(rdtp->dynticks & 0x1)); smp_mb(); /* CPUs seeing ++ must see later RCU read-side crit sects */ } @@ -341,7 +340,7 @@ void rcu_irq_exit(void) return; smp_mb(); /* CPUs seeing ++ must see prior RCU read-side crit sects */ rdtp->dynticks++; - WARN_ON_RATELIMIT(rdtp->dynticks & 0x1, &rcu_rs); + WARN_ON_ONCE(rdtp->dynticks & 0x1); /* If the interrupt queued a callback, get out of dyntick mode. */ if (__get_cpu_var(rcu_sched_data).nxtlist || diff --git a/kernel/rcutree.h b/kernel/rcutree.h index ca560364d8cd..bf8a6f9f134d 100644 --- a/kernel/rcutree.h +++ b/kernel/rcutree.h @@ -81,6 +81,8 @@ struct rcu_dynticks { struct rcu_node { spinlock_t lock; long gpnum; /* Current grace period for this node. */ + /* This will either be equal to or one */ + /* behind the root rcu_node's gpnum. */ unsigned long qsmask; /* CPUs or groups that need to switch in */ /* order for current grace period to proceed.*/ unsigned long qsmaskinit; diff --git a/kernel/rcutree_plugin.h b/kernel/rcutree_plugin.h index 04343bee646d..47789369ea59 100644 --- a/kernel/rcutree_plugin.h +++ b/kernel/rcutree_plugin.h @@ -92,7 +92,7 @@ static void rcu_preempt_qs(int cpu) rnp = rdp->mynode; spin_lock(&rnp->lock); t->rcu_read_unlock_special |= RCU_READ_UNLOCK_BLOCKED; - t->rcu_blocked_node = (void *)rnp; + t->rcu_blocked_node = rnp; /* * If this CPU has already checked in, then this task @@ -176,9 +176,9 @@ static void rcu_read_unlock_special(struct task_struct *t) * most one time. So at most two passes through loop. */ for (;;) { - rnp = (struct rcu_node *)t->rcu_blocked_node; + rnp = t->rcu_blocked_node; spin_lock(&rnp->lock); - if (rnp == (struct rcu_node *)t->rcu_blocked_node) + if (rnp == t->rcu_blocked_node) break; spin_unlock(&rnp->lock); } @@ -288,8 +288,10 @@ static void rcu_preempt_offline_tasks(struct rcu_state *rsp, struct rcu_node *rnp_root = rcu_get_root(rsp); struct task_struct *tp; - if (rnp == rnp_root) + if (rnp == rnp_root) { + WARN_ONCE(1, "Last CPU thought to be offlined?"); return; /* Shouldn't happen: at least one CPU online. */ + } /* * Move tasks up to root rcu_node. Rely on the fact that the -- cgit v1.2.3 From 5bfb5b51382f4bd01d9ced11503264d2cc74fe41 Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Wed, 26 Aug 2009 16:20:48 -0700 Subject: irq: Add irq_node() primitive ... to return irq_desc node info without #ifdefs at the callsites. Signed-off-by: Yinghai Lu Cc: Andrew Morton Cc: Jesse Barnes LKML-Reference: <4A95C350.8060308@kernel.org> Signed-off-by: Ingo Molnar --- include/linux/irqnr.h | 6 ++++++ 1 file changed, 6 insertions(+) (limited to 'include') diff --git a/include/linux/irqnr.h b/include/linux/irqnr.h index ec87b212ff7d..7bf89bc8cbca 100644 --- a/include/linux/irqnr.h +++ b/include/linux/irqnr.h @@ -41,6 +41,12 @@ extern struct irq_desc *irq_to_desc(unsigned int irq); ; \ else +#ifdef CONFIG_SMP +#define irq_node(irq) (irq_to_desc(irq)->node) +#else +#define irq_node(irq) 0 +#endif + #endif /* CONFIG_GENERIC_HARDIRQS */ #define for_each_irq_nr(irq) \ -- cgit v1.2.3 From 138d15692bf76841f252d4b836a535cf5f9154e9 Mon Sep 17 00:00:00 2001 From: Alexey Starikovskiy Date: Fri, 28 Aug 2009 23:29:38 +0400 Subject: ACPICA: Don't switch task then not allowed Signed-off-by: Alexey Starikovskiy Signed-off-by: Len Brown --- include/acpi/platform/aclinux.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/include/acpi/platform/aclinux.h b/include/acpi/platform/aclinux.h index fcb8e4b159b1..9d7febde10a1 100644 --- a/include/acpi/platform/aclinux.h +++ b/include/acpi/platform/aclinux.h @@ -149,10 +149,10 @@ static inline void *acpi_os_acquire_object(acpi_cache_t * cache) #define ACPI_FREE(a) kfree(a) /* Used within ACPICA to show where it is safe to preempt execution */ - +#include #define ACPI_PREEMPTION_POINT() \ do { \ - if (!irqs_disabled()) \ + if (!in_atomic_preempt_off()) \ cond_resched(); \ } while (0) -- cgit v1.2.3 From 2befdcea96fcd9a13e94373c66ea1dd7365d2a74 Mon Sep 17 00:00:00 2001 From: Matt Carlson Date: Fri, 28 Aug 2009 12:28:45 +0000 Subject: tg3: Add new 5785 10/100 only device ID This patch adds a new device ID for those 5785 devices that will only use 10/100 phys. Signed-off-by: Matt Carlson Signed-off-by: David S. Miller --- drivers/net/tg3.c | 3 ++- drivers/net/tg3.h | 2 ++ include/linux/pci_ids.h | 1 - 3 files changed, 4 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/drivers/net/tg3.c b/drivers/net/tg3.c index ab3159ef4c56..d43b30b80f1e 100644 --- a/drivers/net/tg3.c +++ b/drivers/net/tg3.c @@ -219,7 +219,8 @@ static struct pci_device_id tg3_pci_tbl[] = { {PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, PCI_DEVICE_ID_TIGON3_5761E)}, {PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, TG3PCI_DEVICE_TIGON3_5761S)}, {PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, TG3PCI_DEVICE_TIGON3_5761SE)}, - {PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, PCI_DEVICE_ID_TIGON3_5785)}, + {PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, TG3PCI_DEVICE_TIGON3_5785_G)}, + {PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, TG3PCI_DEVICE_TIGON3_5785_F)}, {PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, TG3PCI_DEVICE_TIGON3_57780)}, {PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, TG3PCI_DEVICE_TIGON3_57760)}, {PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, TG3PCI_DEVICE_TIGON3_57790)}, diff --git a/drivers/net/tg3.h b/drivers/net/tg3.h index 60b12ab79334..1c9495df67ce 100644 --- a/drivers/net/tg3.h +++ b/drivers/net/tg3.h @@ -44,6 +44,8 @@ #define TG3PCI_DEVICE_TIGON3_57760 0x1690 #define TG3PCI_DEVICE_TIGON3_57790 0x1694 #define TG3PCI_DEVICE_TIGON3_57788 0x1691 +#define TG3PCI_DEVICE_TIGON3_5785_G 0x1699 /* GPHY */ +#define TG3PCI_DEVICE_TIGON3_5785_F 0x16a0 /* 10/100 only */ /* 0x04 --> 0x64 unused */ #define TG3PCI_MSI_DATA 0x00000064 /* 0x66 --> 0x68 unused */ diff --git a/include/linux/pci_ids.h b/include/linux/pci_ids.h index fdc3110c90c0..85492546d33d 100644 --- a/include/linux/pci_ids.h +++ b/include/linux/pci_ids.h @@ -2064,7 +2064,6 @@ #define PCI_DEVICE_ID_TIGON3_5787M 0x1693 #define PCI_DEVICE_ID_TIGON3_5782 0x1696 #define PCI_DEVICE_ID_TIGON3_5784 0x1698 -#define PCI_DEVICE_ID_TIGON3_5785 0x1699 #define PCI_DEVICE_ID_TIGON3_5786 0x169a #define PCI_DEVICE_ID_TIGON3_5787 0x169b #define PCI_DEVICE_ID_TIGON3_5788 0x169c -- cgit v1.2.3 From ad283ea4a3ce82cda2efe33163748a397b31b1eb Mon Sep 17 00:00:00 2001 From: Dan Williams Date: Sat, 29 Aug 2009 19:09:26 -0700 Subject: async_tx: add sum check flags Replace the flat zero_sum_result with a collection of flags to contain the P (xor) zero-sum result, and the soon to be utilized Q (raid6 reed solomon syndrome) zero-sum result. Use the SUM_CHECK_ namespace instead of DMA_ since these flags will be used on non-dma-zero-sum enabled platforms. Reviewed-by: Andre Noll Acked-by: Maciej Sosnowski Signed-off-by: Dan Williams --- arch/arm/include/asm/hardware/iop3xx-adma.h | 5 +++-- arch/arm/mach-iop13xx/include/mach/adma.h | 12 +++++++----- crypto/async_tx/async_xor.c | 4 ++-- drivers/md/raid5.c | 2 +- drivers/md/raid5.h | 5 +++-- include/linux/async_tx.h | 2 +- include/linux/dmaengine.h | 21 ++++++++++++++++++++- 7 files changed, 37 insertions(+), 14 deletions(-) (limited to 'include') diff --git a/arch/arm/include/asm/hardware/iop3xx-adma.h b/arch/arm/include/asm/hardware/iop3xx-adma.h index 83e6ba338e2c..26eefea02314 100644 --- a/arch/arm/include/asm/hardware/iop3xx-adma.h +++ b/arch/arm/include/asm/hardware/iop3xx-adma.h @@ -756,13 +756,14 @@ static inline void iop_desc_set_block_fill_val(struct iop_adma_desc_slot *desc, hw_desc->src[0] = val; } -static inline int iop_desc_get_zero_result(struct iop_adma_desc_slot *desc) +static inline enum sum_check_flags +iop_desc_get_zero_result(struct iop_adma_desc_slot *desc) { struct iop3xx_desc_aau *hw_desc = desc->hw_desc; struct iop3xx_aau_desc_ctrl desc_ctrl = hw_desc->desc_ctrl_field; iop_paranoia(!(desc_ctrl.tx_complete && desc_ctrl.zero_result_en)); - return desc_ctrl.zero_result_err; + return desc_ctrl.zero_result_err << SUM_CHECK_P; } static inline void iop_chan_append(struct iop_adma_chan *chan) diff --git a/arch/arm/mach-iop13xx/include/mach/adma.h b/arch/arm/mach-iop13xx/include/mach/adma.h index 5722e86f2174..1cd31df8924d 100644 --- a/arch/arm/mach-iop13xx/include/mach/adma.h +++ b/arch/arm/mach-iop13xx/include/mach/adma.h @@ -428,18 +428,20 @@ static inline void iop_desc_set_block_fill_val(struct iop_adma_desc_slot *desc, hw_desc->block_fill_data = val; } -static inline int iop_desc_get_zero_result(struct iop_adma_desc_slot *desc) +static inline enum sum_check_flags +iop_desc_get_zero_result(struct iop_adma_desc_slot *desc) { struct iop13xx_adma_desc_hw *hw_desc = desc->hw_desc; struct iop13xx_adma_desc_ctrl desc_ctrl = hw_desc->desc_ctrl_field; struct iop13xx_adma_byte_count byte_count = hw_desc->byte_count_field; + enum sum_check_flags flags; BUG_ON(!(byte_count.tx_complete && desc_ctrl.zero_result)); - if (desc_ctrl.pq_xfer_en) - return byte_count.zero_result_err_q; - else - return byte_count.zero_result_err; + flags = byte_count.zero_result_err_q << SUM_CHECK_Q; + flags |= byte_count.zero_result_err << SUM_CHECK_P; + + return flags; } static inline void iop_chan_append(struct iop_adma_chan *chan) diff --git a/crypto/async_tx/async_xor.c b/crypto/async_tx/async_xor.c index 1e96c4df7061..78fb7780272a 100644 --- a/crypto/async_tx/async_xor.c +++ b/crypto/async_tx/async_xor.c @@ -246,7 +246,7 @@ static int page_is_zero(struct page *p, unsigned int offset, size_t len) */ struct dma_async_tx_descriptor * async_xor_val(struct page *dest, struct page **src_list, unsigned int offset, - int src_cnt, size_t len, u32 *result, + int src_cnt, size_t len, enum sum_check_flags *result, struct async_submit_ctl *submit) { struct dma_chan *chan = async_tx_find_channel(submit, DMA_XOR_VAL, @@ -304,7 +304,7 @@ async_xor_val(struct page *dest, struct page **src_list, unsigned int offset, async_tx_quiesce(&tx); - *result = page_is_zero(dest, offset, len) ? 0 : 1; + *result = !page_is_zero(dest, offset, len) << SUM_CHECK_P; async_tx_sync_epilog(submit); submit->flags = flags_orig; diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c index 7727954cf726..1f2a266f3cf7 100644 --- a/drivers/md/raid5.c +++ b/drivers/md/raid5.c @@ -2590,7 +2590,7 @@ static void handle_parity_checks5(raid5_conf_t *conf, struct stripe_head *sh, * we are done. Otherwise update the mismatch count and repair * parity if !MD_RECOVERY_CHECK */ - if (sh->ops.zero_sum_result == 0) + if ((sh->ops.zero_sum_result & SUM_CHECK_P_RESULT) == 0) /* parity is correct (on disc, * not in buffer any more) */ diff --git a/drivers/md/raid5.h b/drivers/md/raid5.h index e7baabffee86..75f2c6c4cf90 100644 --- a/drivers/md/raid5.h +++ b/drivers/md/raid5.h @@ -2,6 +2,7 @@ #define _RAID5_H #include +#include /* * @@ -215,8 +216,8 @@ struct stripe_head { * @target - STRIPE_OP_COMPUTE_BLK target */ struct stripe_operations { - int target; - u32 zero_sum_result; + int target; + enum sum_check_flags zero_sum_result; } ops; struct r5dev { struct bio req; diff --git a/include/linux/async_tx.h b/include/linux/async_tx.h index 00cfb637ddf2..3d21a2517518 100644 --- a/include/linux/async_tx.h +++ b/include/linux/async_tx.h @@ -148,7 +148,7 @@ async_xor(struct page *dest, struct page **src_list, unsigned int offset, struct dma_async_tx_descriptor * async_xor_val(struct page *dest, struct page **src_list, unsigned int offset, - int src_cnt, size_t len, u32 *result, + int src_cnt, size_t len, enum sum_check_flags *result, struct async_submit_ctl *submit); struct dma_async_tx_descriptor * diff --git a/include/linux/dmaengine.h b/include/linux/dmaengine.h index 6768727d00d7..02447afcebad 100644 --- a/include/linux/dmaengine.h +++ b/include/linux/dmaengine.h @@ -86,6 +86,25 @@ enum dma_ctrl_flags { DMA_COMPL_SKIP_DEST_UNMAP = (1 << 3), }; +/** + * enum sum_check_bits - bit position of pq_check_flags + */ +enum sum_check_bits { + SUM_CHECK_P = 0, + SUM_CHECK_Q = 1, +}; + +/** + * enum pq_check_flags - result of async_{xor,pq}_zero_sum operations + * @SUM_CHECK_P_RESULT - 1 if xor zero sum error, 0 otherwise + * @SUM_CHECK_Q_RESULT - 1 if reed-solomon zero sum error, 0 otherwise + */ +enum sum_check_flags { + SUM_CHECK_P_RESULT = (1 << SUM_CHECK_P), + SUM_CHECK_Q_RESULT = (1 << SUM_CHECK_Q), +}; + + /** * dma_cap_mask_t - capabilities bitmap modeled after cpumask_t. * See linux/cpumask.h @@ -245,7 +264,7 @@ struct dma_device { unsigned int src_cnt, size_t len, unsigned long flags); struct dma_async_tx_descriptor *(*device_prep_dma_xor_val)( struct dma_chan *chan, dma_addr_t *src, unsigned int src_cnt, - size_t len, u32 *result, unsigned long flags); + size_t len, enum sum_check_flags *result, unsigned long flags); struct dma_async_tx_descriptor *(*device_prep_dma_memset)( struct dma_chan *chan, dma_addr_t dest, int value, size_t len, unsigned long flags); -- cgit v1.2.3 From 95475e57113c66aac7583925736ed2e2d58c990d Mon Sep 17 00:00:00 2001 From: Dan Williams Date: Tue, 14 Jul 2009 12:19:02 -0700 Subject: async_tx: remove walk of tx->parent chain in dma_wait_for_async_tx We currently walk the parent chain when waiting for a given tx to complete however this walk may race with the driver cleanup routine. The routines in async_raid6_recov.c may fall back to the synchronous path at any point so we need to be prepared to call async_tx_quiesce() (which calls dma_wait_for_async_tx). To remove the ->parent walk we guarantee that every time a dependency is attached ->issue_pending() is invoked, then we can simply poll the initial descriptor until completion. This also allows for a lighter weight 'issue pending' implementation as there is no longer a requirement to iterate through all the channels' ->issue_pending() routines as long as operations have been submitted in an ordered chain. async_tx_issue_pending() is added for this case. Signed-off-by: Dan Williams --- crypto/async_tx/async_tx.c | 13 +++++++------ drivers/dma/dmaengine.c | 45 ++++++++++----------------------------------- include/linux/async_tx.h | 23 +++++++++++++++++++++++ 3 files changed, 40 insertions(+), 41 deletions(-) (limited to 'include') diff --git a/crypto/async_tx/async_tx.c b/crypto/async_tx/async_tx.c index 6e37ad3f4417..60615fedcf5e 100644 --- a/crypto/async_tx/async_tx.c +++ b/crypto/async_tx/async_tx.c @@ -77,8 +77,8 @@ static void async_tx_channel_switch(struct dma_async_tx_descriptor *depend_tx, struct dma_async_tx_descriptor *tx) { - struct dma_chan *chan; - struct dma_device *device; + struct dma_chan *chan = depend_tx->chan; + struct dma_device *device = chan->device; struct dma_async_tx_descriptor *intr_tx = (void *) ~0; /* first check to see if we can still append to depend_tx */ @@ -90,11 +90,11 @@ async_tx_channel_switch(struct dma_async_tx_descriptor *depend_tx, } spin_unlock_bh(&depend_tx->lock); - if (!intr_tx) + /* attached dependency, flush the parent channel */ + if (!intr_tx) { + device->device_issue_pending(chan); return; - - chan = depend_tx->chan; - device = chan->device; + } /* see if we can schedule an interrupt * otherwise poll for completion @@ -128,6 +128,7 @@ async_tx_channel_switch(struct dma_async_tx_descriptor *depend_tx, intr_tx->tx_submit(intr_tx); async_tx_ack(intr_tx); } + device->device_issue_pending(chan); } else { if (dma_wait_for_async_tx(depend_tx) == DMA_ERROR) panic("%s: DMA_ERROR waiting for depend_tx\n", diff --git a/drivers/dma/dmaengine.c b/drivers/dma/dmaengine.c index 6781e8f3c064..e002e0e0d055 100644 --- a/drivers/dma/dmaengine.c +++ b/drivers/dma/dmaengine.c @@ -934,49 +934,24 @@ EXPORT_SYMBOL(dma_async_tx_descriptor_init); /* dma_wait_for_async_tx - spin wait for a transaction to complete * @tx: in-flight transaction to wait on - * - * This routine assumes that tx was obtained from a call to async_memcpy, - * async_xor, async_memset, etc which ensures that tx is "in-flight" (prepped - * and submitted). Walking the parent chain is only meant to cover for DMA - * drivers that do not implement the DMA_INTERRUPT capability and may race with - * the driver's descriptor cleanup routine. */ enum dma_status dma_wait_for_async_tx(struct dma_async_tx_descriptor *tx) { - enum dma_status status; - struct dma_async_tx_descriptor *iter; - struct dma_async_tx_descriptor *parent; + unsigned long dma_sync_wait_timeout = jiffies + msecs_to_jiffies(5000); if (!tx) return DMA_SUCCESS; - WARN_ONCE(tx->parent, "%s: speculatively walking dependency chain for" - " %s\n", __func__, dma_chan_name(tx->chan)); - - /* poll through the dependency chain, return when tx is complete */ - do { - iter = tx; - - /* find the root of the unsubmitted dependency chain */ - do { - parent = iter->parent; - if (!parent) - break; - else - iter = parent; - } while (parent); - - /* there is a small window for ->parent == NULL and - * ->cookie == -EBUSY - */ - while (iter->cookie == -EBUSY) - cpu_relax(); - - status = dma_sync_wait(iter->chan, iter->cookie); - } while (status == DMA_IN_PROGRESS || (iter != tx)); - - return status; + while (tx->cookie == -EBUSY) { + if (time_after_eq(jiffies, dma_sync_wait_timeout)) { + pr_err("%s timeout waiting for descriptor submission\n", + __func__); + return DMA_ERROR; + } + cpu_relax(); + } + return dma_sync_wait(tx->chan, tx->cookie); } EXPORT_SYMBOL_GPL(dma_wait_for_async_tx); diff --git a/include/linux/async_tx.h b/include/linux/async_tx.h index 3d21a2517518..12a2efcbd565 100644 --- a/include/linux/async_tx.h +++ b/include/linux/async_tx.h @@ -83,6 +83,24 @@ struct async_submit_ctl { #ifdef CONFIG_DMA_ENGINE #define async_tx_issue_pending_all dma_issue_pending_all + +/** + * async_tx_issue_pending - send pending descriptor to the hardware channel + * @tx: descriptor handle to retrieve hardware context + * + * Note: any dependent operations will have already been issued by + * async_tx_channel_switch, or (in the case of no channel switch) will + * be already pending on this channel. + */ +static inline void async_tx_issue_pending(struct dma_async_tx_descriptor *tx) +{ + if (likely(tx)) { + struct dma_chan *chan = tx->chan; + struct dma_device *dma = chan->device; + + dma->device_issue_pending(chan); + } +} #ifdef CONFIG_ARCH_HAS_ASYNC_TX_FIND_CHANNEL #include #else @@ -98,6 +116,11 @@ static inline void async_tx_issue_pending_all(void) do { } while (0); } +static inline void async_tx_issue_pending(struct dma_async_tx_descriptor *tx) +{ + do { } while (0); +} + static inline struct dma_chan * async_tx_find_channel(struct async_submit_ctl *submit, enum dma_transaction_type tx_type, struct page **dst, -- cgit v1.2.3 From b2f46fd8ef3dff2ab30f31126833f78b7480283a Mon Sep 17 00:00:00 2001 From: Dan Williams Date: Tue, 14 Jul 2009 12:20:36 -0700 Subject: async_tx: add support for asynchronous GF multiplication [ Based on an original patch by Yuri Tikhonov ] This adds support for doing asynchronous GF multiplication by adding two additional functions to the async_tx API: async_gen_syndrome() does simultaneous XOR and Galois field multiplication of sources. async_syndrome_val() validates the given source buffers against known P and Q values. When a request is made to run async_pq against more than the hardware maximum number of supported sources we need to reuse the previous generated P and Q values as sources into the next operation. Care must be taken to remove Q from P' and P from Q'. For example to perform a 5 source pq op with hardware that only supports 4 sources at a time the following approach is taken: p, q = PQ(src0, src1, src2, src3, COEF({01}, {02}, {04}, {08})) p', q' = PQ(p, q, q, src4, COEF({00}, {01}, {00}, {10})) p' = p + q + q + src4 = p + src4 q' = {00}*p + {01}*q + {00}*q + {10}*src4 = q + {10}*src4 Note: 4 is the minimum acceptable maxpq otherwise we punt to synchronous-software path. The DMA_PREP_CONTINUE flag indicates to the driver to reuse p and q as sources (in the above manner) and fill the remaining slots up to maxpq with the new sources/coefficients. Note1: Some devices have native support for P+Q continuation and can skip this extra work. Devices with this capability can advertise it with dma_set_maxpq. It is up to each driver how to handle the DMA_PREP_CONTINUE flag. Note2: The api supports disabling the generation of P when generating Q, this is ignored by the synchronous path but is implemented by some dma devices to save unnecessary writes. In this case the continuation algorithm is simplified to only reuse Q as a source. Cc: H. Peter Anvin Cc: David Woodhouse Signed-off-by: Yuri Tikhonov Signed-off-by: Ilya Yanok Reviewed-by: Andre Noll Acked-by: Maciej Sosnowski Signed-off-by: Dan Williams --- Documentation/crypto/async-tx-api.txt | 3 + arch/arm/mach-iop13xx/setup.c | 2 +- crypto/async_tx/Kconfig | 4 + crypto/async_tx/Makefile | 1 + crypto/async_tx/async_pq.c | 388 ++++++++++++++++++++++++++++++++++ crypto/async_tx/async_xor.c | 2 +- drivers/dma/dmaengine.c | 4 + drivers/dma/iop-adma.c | 2 +- include/linux/async_tx.h | 9 + include/linux/dmaengine.h | 87 +++++++- 10 files changed, 493 insertions(+), 9 deletions(-) create mode 100644 crypto/async_tx/async_pq.c (limited to 'include') diff --git a/Documentation/crypto/async-tx-api.txt b/Documentation/crypto/async-tx-api.txt index 6b15e488c0e7..0e48e054d69a 100644 --- a/Documentation/crypto/async-tx-api.txt +++ b/Documentation/crypto/async-tx-api.txt @@ -64,6 +64,9 @@ xor - xor a series of source buffers and write the result to a xor_val - xor a series of source buffers and set a flag if the result is zero. The implementation attempts to prevent writes to memory +pq - generate the p+q (raid6 syndrome) from a series of source buffers +pq_val - validate that a p and or q buffer are in sync with a given series of + sources 3.3 Descriptor management: The return value is non-NULL and points to a 'descriptor' when the operation diff --git a/arch/arm/mach-iop13xx/setup.c b/arch/arm/mach-iop13xx/setup.c index 9800228b71d3..2e7ca0d75f8a 100644 --- a/arch/arm/mach-iop13xx/setup.c +++ b/arch/arm/mach-iop13xx/setup.c @@ -506,7 +506,7 @@ void __init iop13xx_platform_init(void) dma_cap_set(DMA_MEMSET, plat_data->cap_mask); dma_cap_set(DMA_MEMCPY_CRC32C, plat_data->cap_mask); dma_cap_set(DMA_INTERRUPT, plat_data->cap_mask); - dma_cap_set(DMA_PQ_XOR, plat_data->cap_mask); + dma_cap_set(DMA_PQ, plat_data->cap_mask); dma_cap_set(DMA_PQ_UPDATE, plat_data->cap_mask); dma_cap_set(DMA_PQ_VAL, plat_data->cap_mask); break; diff --git a/crypto/async_tx/Kconfig b/crypto/async_tx/Kconfig index d8fb39145986..cb6d7314f198 100644 --- a/crypto/async_tx/Kconfig +++ b/crypto/async_tx/Kconfig @@ -14,3 +14,7 @@ config ASYNC_MEMSET tristate select ASYNC_CORE +config ASYNC_PQ + tristate + select ASYNC_CORE + diff --git a/crypto/async_tx/Makefile b/crypto/async_tx/Makefile index 27baa7d52fbc..1b9926588259 100644 --- a/crypto/async_tx/Makefile +++ b/crypto/async_tx/Makefile @@ -2,3 +2,4 @@ obj-$(CONFIG_ASYNC_CORE) += async_tx.o obj-$(CONFIG_ASYNC_MEMCPY) += async_memcpy.o obj-$(CONFIG_ASYNC_MEMSET) += async_memset.o obj-$(CONFIG_ASYNC_XOR) += async_xor.o +obj-$(CONFIG_ASYNC_PQ) += async_pq.o diff --git a/crypto/async_tx/async_pq.c b/crypto/async_tx/async_pq.c new file mode 100644 index 000000000000..108b21efb499 --- /dev/null +++ b/crypto/async_tx/async_pq.c @@ -0,0 +1,388 @@ +/* + * Copyright(c) 2007 Yuri Tikhonov + * Copyright(c) 2009 Intel Corporation + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the Free + * Software Foundation; either version 2 of the License, or (at your option) + * any later version. + * + * This program is distributed in the hope that it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program; if not, write to the Free Software Foundation, Inc., 59 + * Temple Place - Suite 330, Boston, MA 02111-1307, USA. + * + * The full GNU General Public License is included in this distribution in the + * file called COPYING. + */ +#include +#include +#include +#include +#include + +/** + * scribble - space to hold throwaway P buffer for synchronous gen_syndrome + */ +static struct page *scribble; + +static bool is_raid6_zero_block(struct page *p) +{ + return p == (void *) raid6_empty_zero_page; +} + +/* the struct page *blocks[] parameter passed to async_gen_syndrome() + * and async_syndrome_val() contains the 'P' destination address at + * blocks[disks-2] and the 'Q' destination address at blocks[disks-1] + * + * note: these are macros as they are used as lvalues + */ +#define P(b, d) (b[d-2]) +#define Q(b, d) (b[d-1]) + +/** + * do_async_gen_syndrome - asynchronously calculate P and/or Q + */ +static __async_inline struct dma_async_tx_descriptor * +do_async_gen_syndrome(struct dma_chan *chan, struct page **blocks, + const unsigned char *scfs, unsigned int offset, int disks, + size_t len, dma_addr_t *dma_src, + struct async_submit_ctl *submit) +{ + struct dma_async_tx_descriptor *tx = NULL; + struct dma_device *dma = chan->device; + enum dma_ctrl_flags dma_flags = 0; + enum async_tx_flags flags_orig = submit->flags; + dma_async_tx_callback cb_fn_orig = submit->cb_fn; + dma_async_tx_callback cb_param_orig = submit->cb_param; + int src_cnt = disks - 2; + unsigned char coefs[src_cnt]; + unsigned short pq_src_cnt; + dma_addr_t dma_dest[2]; + int src_off = 0; + int idx; + int i; + + /* DMAs use destinations as sources, so use BIDIRECTIONAL mapping */ + if (P(blocks, disks)) + dma_dest[0] = dma_map_page(dma->dev, P(blocks, disks), offset, + len, DMA_BIDIRECTIONAL); + else + dma_flags |= DMA_PREP_PQ_DISABLE_P; + if (Q(blocks, disks)) + dma_dest[1] = dma_map_page(dma->dev, Q(blocks, disks), offset, + len, DMA_BIDIRECTIONAL); + else + dma_flags |= DMA_PREP_PQ_DISABLE_Q; + + /* convert source addresses being careful to collapse 'empty' + * sources and update the coefficients accordingly + */ + for (i = 0, idx = 0; i < src_cnt; i++) { + if (is_raid6_zero_block(blocks[i])) + continue; + dma_src[idx] = dma_map_page(dma->dev, blocks[i], offset, len, + DMA_TO_DEVICE); + coefs[idx] = scfs[i]; + idx++; + } + src_cnt = idx; + + while (src_cnt > 0) { + submit->flags = flags_orig; + pq_src_cnt = min(src_cnt, dma_maxpq(dma, dma_flags)); + /* if we are submitting additional pqs, leave the chain open, + * clear the callback parameters, and leave the destination + * buffers mapped + */ + if (src_cnt > pq_src_cnt) { + submit->flags &= ~ASYNC_TX_ACK; + dma_flags |= DMA_COMPL_SKIP_DEST_UNMAP; + submit->cb_fn = NULL; + submit->cb_param = NULL; + } else { + dma_flags &= ~DMA_COMPL_SKIP_DEST_UNMAP; + submit->cb_fn = cb_fn_orig; + submit->cb_param = cb_param_orig; + if (cb_fn_orig) + dma_flags |= DMA_PREP_INTERRUPT; + } + + /* Since we have clobbered the src_list we are committed + * to doing this asynchronously. Drivers force forward + * progress in case they can not provide a descriptor + */ + for (;;) { + tx = dma->device_prep_dma_pq(chan, dma_dest, + &dma_src[src_off], + pq_src_cnt, + &coefs[src_off], len, + dma_flags); + if (likely(tx)) + break; + async_tx_quiesce(&submit->depend_tx); + dma_async_issue_pending(chan); + } + + async_tx_submit(chan, tx, submit); + submit->depend_tx = tx; + + /* drop completed sources */ + src_cnt -= pq_src_cnt; + src_off += pq_src_cnt; + + dma_flags |= DMA_PREP_CONTINUE; + } + + return tx; +} + +/** + * do_sync_gen_syndrome - synchronously calculate a raid6 syndrome + */ +static void +do_sync_gen_syndrome(struct page **blocks, unsigned int offset, int disks, + size_t len, struct async_submit_ctl *submit) +{ + void **srcs; + int i; + + if (submit->scribble) + srcs = submit->scribble; + else + srcs = (void **) blocks; + + for (i = 0; i < disks; i++) { + if (is_raid6_zero_block(blocks[i])) { + BUG_ON(i > disks - 3); /* P or Q can't be zero */ + srcs[i] = blocks[i]; + } else + srcs[i] = page_address(blocks[i]) + offset; + } + raid6_call.gen_syndrome(disks, len, srcs); + async_tx_sync_epilog(submit); +} + +/** + * async_gen_syndrome - asynchronously calculate a raid6 syndrome + * @blocks: source blocks from idx 0..disks-3, P @ disks-2 and Q @ disks-1 + * @offset: common offset into each block (src and dest) to start transaction + * @disks: number of blocks (including missing P or Q, see below) + * @len: length of operation in bytes + * @submit: submission/completion modifiers + * + * General note: This routine assumes a field of GF(2^8) with a + * primitive polynomial of 0x11d and a generator of {02}. + * + * 'disks' note: callers can optionally omit either P or Q (but not + * both) from the calculation by setting blocks[disks-2] or + * blocks[disks-1] to NULL. When P or Q is omitted 'len' must be <= + * PAGE_SIZE as a temporary buffer of this size is used in the + * synchronous path. 'disks' always accounts for both destination + * buffers. + * + * 'blocks' note: if submit->scribble is NULL then the contents of + * 'blocks' may be overridden + */ +struct dma_async_tx_descriptor * +async_gen_syndrome(struct page **blocks, unsigned int offset, int disks, + size_t len, struct async_submit_ctl *submit) +{ + int src_cnt = disks - 2; + struct dma_chan *chan = async_tx_find_channel(submit, DMA_PQ, + &P(blocks, disks), 2, + blocks, src_cnt, len); + struct dma_device *device = chan ? chan->device : NULL; + dma_addr_t *dma_src = NULL; + + BUG_ON(disks > 255 || !(P(blocks, disks) || Q(blocks, disks))); + + if (submit->scribble) + dma_src = submit->scribble; + else if (sizeof(dma_addr_t) <= sizeof(struct page *)) + dma_src = (dma_addr_t *) blocks; + + if (dma_src && device && + (src_cnt <= dma_maxpq(device, 0) || + dma_maxpq(device, DMA_PREP_CONTINUE) > 0)) { + /* run the p+q asynchronously */ + pr_debug("%s: (async) disks: %d len: %zu\n", + __func__, disks, len); + return do_async_gen_syndrome(chan, blocks, raid6_gfexp, offset, + disks, len, dma_src, submit); + } + + /* run the pq synchronously */ + pr_debug("%s: (sync) disks: %d len: %zu\n", __func__, disks, len); + + /* wait for any prerequisite operations */ + async_tx_quiesce(&submit->depend_tx); + + if (!P(blocks, disks)) { + P(blocks, disks) = scribble; + BUG_ON(len + offset > PAGE_SIZE); + } + if (!Q(blocks, disks)) { + Q(blocks, disks) = scribble; + BUG_ON(len + offset > PAGE_SIZE); + } + do_sync_gen_syndrome(blocks, offset, disks, len, submit); + + return NULL; +} +EXPORT_SYMBOL_GPL(async_gen_syndrome); + +/** + * async_syndrome_val - asynchronously validate a raid6 syndrome + * @blocks: source blocks from idx 0..disks-3, P @ disks-2 and Q @ disks-1 + * @offset: common offset into each block (src and dest) to start transaction + * @disks: number of blocks (including missing P or Q, see below) + * @len: length of operation in bytes + * @pqres: on val failure SUM_CHECK_P_RESULT and/or SUM_CHECK_Q_RESULT are set + * @spare: temporary result buffer for the synchronous case + * @submit: submission / completion modifiers + * + * The same notes from async_gen_syndrome apply to the 'blocks', + * and 'disks' parameters of this routine. The synchronous path + * requires a temporary result buffer and submit->scribble to be + * specified. + */ +struct dma_async_tx_descriptor * +async_syndrome_val(struct page **blocks, unsigned int offset, int disks, + size_t len, enum sum_check_flags *pqres, struct page *spare, + struct async_submit_ctl *submit) +{ + struct dma_chan *chan = async_tx_find_channel(submit, DMA_PQ_VAL, + NULL, 0, blocks, disks, + len); + struct dma_device *device = chan ? chan->device : NULL; + struct dma_async_tx_descriptor *tx; + enum dma_ctrl_flags dma_flags = submit->cb_fn ? DMA_PREP_INTERRUPT : 0; + dma_addr_t *dma_src = NULL; + + BUG_ON(disks < 4); + + if (submit->scribble) + dma_src = submit->scribble; + else if (sizeof(dma_addr_t) <= sizeof(struct page *)) + dma_src = (dma_addr_t *) blocks; + + if (dma_src && device && disks <= dma_maxpq(device, 0)) { + struct device *dev = device->dev; + dma_addr_t *pq = &dma_src[disks-2]; + int i; + + pr_debug("%s: (async) disks: %d len: %zu\n", + __func__, disks, len); + if (!P(blocks, disks)) + dma_flags |= DMA_PREP_PQ_DISABLE_P; + if (!Q(blocks, disks)) + dma_flags |= DMA_PREP_PQ_DISABLE_Q; + for (i = 0; i < disks; i++) + if (likely(blocks[i])) { + BUG_ON(is_raid6_zero_block(blocks[i])); + dma_src[i] = dma_map_page(dev, blocks[i], + offset, len, + DMA_TO_DEVICE); + } + + for (;;) { + tx = device->device_prep_dma_pq_val(chan, pq, dma_src, + disks - 2, + raid6_gfexp, + len, pqres, + dma_flags); + if (likely(tx)) + break; + async_tx_quiesce(&submit->depend_tx); + dma_async_issue_pending(chan); + } + async_tx_submit(chan, tx, submit); + + return tx; + } else { + struct page *p_src = P(blocks, disks); + struct page *q_src = Q(blocks, disks); + enum async_tx_flags flags_orig = submit->flags; + dma_async_tx_callback cb_fn_orig = submit->cb_fn; + void *scribble = submit->scribble; + void *cb_param_orig = submit->cb_param; + void *p, *q, *s; + + pr_debug("%s: (sync) disks: %d len: %zu\n", + __func__, disks, len); + + /* caller must provide a temporary result buffer and + * allow the input parameters to be preserved + */ + BUG_ON(!spare || !scribble); + + /* wait for any prerequisite operations */ + async_tx_quiesce(&submit->depend_tx); + + /* recompute p and/or q into the temporary buffer and then + * check to see the result matches the current value + */ + tx = NULL; + *pqres = 0; + if (p_src) { + init_async_submit(submit, ASYNC_TX_XOR_ZERO_DST, NULL, + NULL, NULL, scribble); + tx = async_xor(spare, blocks, offset, disks-2, len, submit); + async_tx_quiesce(&tx); + p = page_address(p_src) + offset; + s = page_address(spare) + offset; + *pqres |= !!memcmp(p, s, len) << SUM_CHECK_P; + } + + if (q_src) { + P(blocks, disks) = NULL; + Q(blocks, disks) = spare; + init_async_submit(submit, 0, NULL, NULL, NULL, scribble); + tx = async_gen_syndrome(blocks, offset, disks, len, submit); + async_tx_quiesce(&tx); + q = page_address(q_src) + offset; + s = page_address(spare) + offset; + *pqres |= !!memcmp(q, s, len) << SUM_CHECK_Q; + } + + /* restore P, Q and submit */ + P(blocks, disks) = p_src; + Q(blocks, disks) = q_src; + + submit->cb_fn = cb_fn_orig; + submit->cb_param = cb_param_orig; + submit->flags = flags_orig; + async_tx_sync_epilog(submit); + + return NULL; + } +} +EXPORT_SYMBOL_GPL(async_syndrome_val); + +static int __init async_pq_init(void) +{ + scribble = alloc_page(GFP_KERNEL); + + if (scribble) + return 0; + + pr_err("%s: failed to allocate required spare page\n", __func__); + + return -ENOMEM; +} + +static void __exit async_pq_exit(void) +{ + put_page(scribble); +} + +module_init(async_pq_init); +module_exit(async_pq_exit); + +MODULE_DESCRIPTION("asynchronous raid6 syndrome generation/validation"); +MODULE_LICENSE("GPL"); diff --git a/crypto/async_tx/async_xor.c b/crypto/async_tx/async_xor.c index 78fb7780272a..56b5f98da463 100644 --- a/crypto/async_tx/async_xor.c +++ b/crypto/async_tx/async_xor.c @@ -62,7 +62,7 @@ do_async_xor(struct dma_chan *chan, struct page *dest, struct page **src_list, while (src_cnt) { submit->flags = flags_orig; dma_flags = 0; - xor_src_cnt = min(src_cnt, dma->max_xor); + xor_src_cnt = min(src_cnt, (int)dma->max_xor); /* if we are submitting additional xors, leave the chain open, * clear the callback parameters, and leave the destination * buffer mapped diff --git a/drivers/dma/dmaengine.c b/drivers/dma/dmaengine.c index e002e0e0d055..cd5673d3043b 100644 --- a/drivers/dma/dmaengine.c +++ b/drivers/dma/dmaengine.c @@ -646,6 +646,10 @@ int dma_async_device_register(struct dma_device *device) !device->device_prep_dma_xor); BUG_ON(dma_has_cap(DMA_XOR_VAL, device->cap_mask) && !device->device_prep_dma_xor_val); + BUG_ON(dma_has_cap(DMA_PQ, device->cap_mask) && + !device->device_prep_dma_pq); + BUG_ON(dma_has_cap(DMA_PQ_VAL, device->cap_mask) && + !device->device_prep_dma_pq_val); BUG_ON(dma_has_cap(DMA_MEMSET, device->cap_mask) && !device->device_prep_dma_memset); BUG_ON(dma_has_cap(DMA_INTERRUPT, device->cap_mask) && diff --git a/drivers/dma/iop-adma.c b/drivers/dma/iop-adma.c index 6ff79a672699..4496bc606662 100644 --- a/drivers/dma/iop-adma.c +++ b/drivers/dma/iop-adma.c @@ -1257,7 +1257,7 @@ static int __devinit iop_adma_probe(struct platform_device *pdev) dev_printk(KERN_INFO, &pdev->dev, "Intel(R) IOP: " "( %s%s%s%s%s%s%s%s%s%s)\n", - dma_has_cap(DMA_PQ_XOR, dma_dev->cap_mask) ? "pq_xor " : "", + dma_has_cap(DMA_PQ, dma_dev->cap_mask) ? "pq " : "", dma_has_cap(DMA_PQ_UPDATE, dma_dev->cap_mask) ? "pq_update " : "", dma_has_cap(DMA_PQ_VAL, dma_dev->cap_mask) ? "pq_val " : "", dma_has_cap(DMA_XOR, dma_dev->cap_mask) ? "xor " : "", diff --git a/include/linux/async_tx.h b/include/linux/async_tx.h index 12a2efcbd565..e6ce5f004f98 100644 --- a/include/linux/async_tx.h +++ b/include/linux/async_tx.h @@ -185,5 +185,14 @@ async_memset(struct page *dest, int val, unsigned int offset, struct dma_async_tx_descriptor *async_trigger_callback(struct async_submit_ctl *submit); +struct dma_async_tx_descriptor * +async_gen_syndrome(struct page **blocks, unsigned int offset, int src_cnt, + size_t len, struct async_submit_ctl *submit); + +struct dma_async_tx_descriptor * +async_syndrome_val(struct page **blocks, unsigned int offset, int src_cnt, + size_t len, enum sum_check_flags *pqres, struct page *spare, + struct async_submit_ctl *submit); + void async_tx_quiesce(struct dma_async_tx_descriptor **tx); #endif /* _ASYNC_TX_H_ */ diff --git a/include/linux/dmaengine.h b/include/linux/dmaengine.h index 02447afcebad..ce010cd991d2 100644 --- a/include/linux/dmaengine.h +++ b/include/linux/dmaengine.h @@ -52,7 +52,7 @@ enum dma_status { enum dma_transaction_type { DMA_MEMCPY, DMA_XOR, - DMA_PQ_XOR, + DMA_PQ, DMA_DUAL_XOR, DMA_PQ_UPDATE, DMA_XOR_VAL, @@ -70,20 +70,28 @@ enum dma_transaction_type { /** * enum dma_ctrl_flags - DMA flags to augment operation preparation, - * control completion, and communicate status. + * control completion, and communicate status. * @DMA_PREP_INTERRUPT - trigger an interrupt (callback) upon completion of - * this transaction + * this transaction * @DMA_CTRL_ACK - the descriptor cannot be reused until the client - * acknowledges receipt, i.e. has has a chance to establish any - * dependency chains + * acknowledges receipt, i.e. has has a chance to establish any dependency + * chains * @DMA_COMPL_SKIP_SRC_UNMAP - set to disable dma-unmapping the source buffer(s) * @DMA_COMPL_SKIP_DEST_UNMAP - set to disable dma-unmapping the destination(s) + * @DMA_PREP_PQ_DISABLE_P - prevent generation of P while generating Q + * @DMA_PREP_PQ_DISABLE_Q - prevent generation of Q while generating P + * @DMA_PREP_CONTINUE - indicate to a driver that it is reusing buffers as + * sources that were the result of a previous operation, in the case of a PQ + * operation it continues the calculation with new sources */ enum dma_ctrl_flags { DMA_PREP_INTERRUPT = (1 << 0), DMA_CTRL_ACK = (1 << 1), DMA_COMPL_SKIP_SRC_UNMAP = (1 << 2), DMA_COMPL_SKIP_DEST_UNMAP = (1 << 3), + DMA_PREP_PQ_DISABLE_P = (1 << 4), + DMA_PREP_PQ_DISABLE_Q = (1 << 5), + DMA_PREP_CONTINUE = (1 << 6), }; /** @@ -226,6 +234,7 @@ struct dma_async_tx_descriptor { * @global_node: list_head for global dma_device_list * @cap_mask: one or more dma_capability flags * @max_xor: maximum number of xor sources, 0 if no capability + * @max_pq: maximum number of PQ sources and PQ-continue capability * @dev_id: unique device ID * @dev: struct device reference for dma mapping api * @device_alloc_chan_resources: allocate resources and return the @@ -234,6 +243,8 @@ struct dma_async_tx_descriptor { * @device_prep_dma_memcpy: prepares a memcpy operation * @device_prep_dma_xor: prepares a xor operation * @device_prep_dma_xor_val: prepares a xor validation operation + * @device_prep_dma_pq: prepares a pq operation + * @device_prep_dma_pq_val: prepares a pqzero_sum operation * @device_prep_dma_memset: prepares a memset operation * @device_prep_dma_interrupt: prepares an end of chain interrupt operation * @device_prep_slave_sg: prepares a slave dma operation @@ -248,7 +259,9 @@ struct dma_device { struct list_head channels; struct list_head global_node; dma_cap_mask_t cap_mask; - int max_xor; + unsigned short max_xor; + unsigned short max_pq; + #define DMA_HAS_PQ_CONTINUE (1 << 15) int dev_id; struct device *dev; @@ -265,6 +278,14 @@ struct dma_device { struct dma_async_tx_descriptor *(*device_prep_dma_xor_val)( struct dma_chan *chan, dma_addr_t *src, unsigned int src_cnt, size_t len, enum sum_check_flags *result, unsigned long flags); + struct dma_async_tx_descriptor *(*device_prep_dma_pq)( + struct dma_chan *chan, dma_addr_t *dst, dma_addr_t *src, + unsigned int src_cnt, const unsigned char *scf, + size_t len, unsigned long flags); + struct dma_async_tx_descriptor *(*device_prep_dma_pq_val)( + struct dma_chan *chan, dma_addr_t *pq, dma_addr_t *src, + unsigned int src_cnt, const unsigned char *scf, size_t len, + enum sum_check_flags *pqres, unsigned long flags); struct dma_async_tx_descriptor *(*device_prep_dma_memset)( struct dma_chan *chan, dma_addr_t dest, int value, size_t len, unsigned long flags); @@ -283,6 +304,60 @@ struct dma_device { void (*device_issue_pending)(struct dma_chan *chan); }; +static inline void +dma_set_maxpq(struct dma_device *dma, int maxpq, int has_pq_continue) +{ + dma->max_pq = maxpq; + if (has_pq_continue) + dma->max_pq |= DMA_HAS_PQ_CONTINUE; +} + +static inline bool dmaf_continue(enum dma_ctrl_flags flags) +{ + return (flags & DMA_PREP_CONTINUE) == DMA_PREP_CONTINUE; +} + +static inline bool dmaf_p_disabled_continue(enum dma_ctrl_flags flags) +{ + enum dma_ctrl_flags mask = DMA_PREP_CONTINUE | DMA_PREP_PQ_DISABLE_P; + + return (flags & mask) == mask; +} + +static inline bool dma_dev_has_pq_continue(struct dma_device *dma) +{ + return (dma->max_pq & DMA_HAS_PQ_CONTINUE) == DMA_HAS_PQ_CONTINUE; +} + +static unsigned short dma_dev_to_maxpq(struct dma_device *dma) +{ + return dma->max_pq & ~DMA_HAS_PQ_CONTINUE; +} + +/* dma_maxpq - reduce maxpq in the face of continued operations + * @dma - dma device with PQ capability + * @flags - to check if DMA_PREP_CONTINUE and DMA_PREP_PQ_DISABLE_P are set + * + * When an engine does not support native continuation we need 3 extra + * source slots to reuse P and Q with the following coefficients: + * 1/ {00} * P : remove P from Q', but use it as a source for P' + * 2/ {01} * Q : use Q to continue Q' calculation + * 3/ {00} * Q : subtract Q from P' to cancel (2) + * + * In the case where P is disabled we only need 1 extra source: + * 1/ {01} * Q : use Q to continue Q' calculation + */ +static inline int dma_maxpq(struct dma_device *dma, enum dma_ctrl_flags flags) +{ + if (dma_dev_has_pq_continue(dma) || !dmaf_continue(flags)) + return dma_dev_to_maxpq(dma); + else if (dmaf_p_disabled_continue(flags)) + return dma_dev_to_maxpq(dma) - 1; + else if (dmaf_continue(flags)) + return dma_dev_to_maxpq(dma) - 3; + BUG(); +} + /* --- public DMA engine API --- */ #ifdef CONFIG_DMA_ENGINE -- cgit v1.2.3 From 0a82a6239beecc95db6e05fe43ee62d16b381d38 Mon Sep 17 00:00:00 2001 From: Dan Williams Date: Tue, 14 Jul 2009 12:20:37 -0700 Subject: async_tx: add support for asynchronous RAID6 recovery operations async_raid6_2data_recov() recovers two data disk failures async_raid6_datap_recov() recovers a data disk and the P disk These routines are a port of the synchronous versions found in drivers/md/raid6recov.c. The primary difference is breaking out the xor operations into separate calls to async_xor. Two helper routines are introduced to perform scalar multiplication where needed. async_sum_product() multiplies two sources by scalar coefficients and then sums (xor) the result. async_mult() simply multiplies a single source by a scalar. This implemention also includes, in contrast to the original synchronous-only code, special case handling for the 4-disk and 5-disk array cases. In these situations the default N-disk algorithm will present 0-source or 1-source operations to dma devices. To cover for dma devices where the minimum source count is 2 we implement 4-disk and 5-disk handling in the recovery code. [ Impact: asynchronous raid6 recovery routines for 2data and datap cases ] Cc: Yuri Tikhonov Cc: Ilya Yanok Cc: H. Peter Anvin Cc: David Woodhouse Reviewed-by: Andre Noll Acked-by: Maciej Sosnowski Signed-off-by: Dan Williams --- Documentation/crypto/async-tx-api.txt | 4 + crypto/async_tx/Kconfig | 5 + crypto/async_tx/Makefile | 1 + crypto/async_tx/async_raid6_recov.c | 448 ++++++++++++++++++++++++++++++++++ include/linux/async_tx.h | 8 + 5 files changed, 466 insertions(+) create mode 100644 crypto/async_tx/async_raid6_recov.c (limited to 'include') diff --git a/Documentation/crypto/async-tx-api.txt b/Documentation/crypto/async-tx-api.txt index 0e48e054d69a..ba046b8fa92f 100644 --- a/Documentation/crypto/async-tx-api.txt +++ b/Documentation/crypto/async-tx-api.txt @@ -67,6 +67,10 @@ xor_val - xor a series of source buffers and set a flag if the pq - generate the p+q (raid6 syndrome) from a series of source buffers pq_val - validate that a p and or q buffer are in sync with a given series of sources +datap - (raid6_datap_recov) recover a raid6 data block and the p block + from the given sources +2data - (raid6_2data_recov) recover 2 raid6 data blocks from the given + sources 3.3 Descriptor management: The return value is non-NULL and points to a 'descriptor' when the operation diff --git a/crypto/async_tx/Kconfig b/crypto/async_tx/Kconfig index cb6d7314f198..e5aeb2b79e6f 100644 --- a/crypto/async_tx/Kconfig +++ b/crypto/async_tx/Kconfig @@ -18,3 +18,8 @@ config ASYNC_PQ tristate select ASYNC_CORE +config ASYNC_RAID6_RECOV + tristate + select ASYNC_CORE + select ASYNC_PQ + diff --git a/crypto/async_tx/Makefile b/crypto/async_tx/Makefile index 1b9926588259..9a1a76811b80 100644 --- a/crypto/async_tx/Makefile +++ b/crypto/async_tx/Makefile @@ -3,3 +3,4 @@ obj-$(CONFIG_ASYNC_MEMCPY) += async_memcpy.o obj-$(CONFIG_ASYNC_MEMSET) += async_memset.o obj-$(CONFIG_ASYNC_XOR) += async_xor.o obj-$(CONFIG_ASYNC_PQ) += async_pq.o +obj-$(CONFIG_ASYNC_RAID6_RECOV) += async_raid6_recov.o diff --git a/crypto/async_tx/async_raid6_recov.c b/crypto/async_tx/async_raid6_recov.c new file mode 100644 index 000000000000..0c14d48c9896 --- /dev/null +++ b/crypto/async_tx/async_raid6_recov.c @@ -0,0 +1,448 @@ +/* + * Asynchronous RAID-6 recovery calculations ASYNC_TX API. + * Copyright(c) 2009 Intel Corporation + * + * based on raid6recov.c: + * Copyright 2002 H. Peter Anvin + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the Free + * Software Foundation; either version 2 of the License, or (at your option) + * any later version. + * + * This program is distributed in the hope that it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program; if not, write to the Free Software Foundation, Inc., 51 + * Franklin St - Fifth Floor, Boston, MA 02110-1301 USA. + * + */ +#include +#include +#include +#include +#include + +static struct dma_async_tx_descriptor * +async_sum_product(struct page *dest, struct page **srcs, unsigned char *coef, + size_t len, struct async_submit_ctl *submit) +{ + struct dma_chan *chan = async_tx_find_channel(submit, DMA_PQ, + &dest, 1, srcs, 2, len); + struct dma_device *dma = chan ? chan->device : NULL; + const u8 *amul, *bmul; + u8 ax, bx; + u8 *a, *b, *c; + + if (dma) { + dma_addr_t dma_dest[2]; + dma_addr_t dma_src[2]; + struct device *dev = dma->dev; + struct dma_async_tx_descriptor *tx; + enum dma_ctrl_flags dma_flags = DMA_PREP_PQ_DISABLE_P; + + dma_dest[1] = dma_map_page(dev, dest, 0, len, DMA_BIDIRECTIONAL); + dma_src[0] = dma_map_page(dev, srcs[0], 0, len, DMA_TO_DEVICE); + dma_src[1] = dma_map_page(dev, srcs[1], 0, len, DMA_TO_DEVICE); + tx = dma->device_prep_dma_pq(chan, dma_dest, dma_src, 2, coef, + len, dma_flags); + if (tx) { + async_tx_submit(chan, tx, submit); + return tx; + } + } + + /* run the operation synchronously */ + async_tx_quiesce(&submit->depend_tx); + amul = raid6_gfmul[coef[0]]; + bmul = raid6_gfmul[coef[1]]; + a = page_address(srcs[0]); + b = page_address(srcs[1]); + c = page_address(dest); + + while (len--) { + ax = amul[*a++]; + bx = bmul[*b++]; + *c++ = ax ^ bx; + } + + return NULL; +} + +static struct dma_async_tx_descriptor * +async_mult(struct page *dest, struct page *src, u8 coef, size_t len, + struct async_submit_ctl *submit) +{ + struct dma_chan *chan = async_tx_find_channel(submit, DMA_PQ, + &dest, 1, &src, 1, len); + struct dma_device *dma = chan ? chan->device : NULL; + const u8 *qmul; /* Q multiplier table */ + u8 *d, *s; + + if (dma) { + dma_addr_t dma_dest[2]; + dma_addr_t dma_src[1]; + struct device *dev = dma->dev; + struct dma_async_tx_descriptor *tx; + enum dma_ctrl_flags dma_flags = DMA_PREP_PQ_DISABLE_P; + + dma_dest[1] = dma_map_page(dev, dest, 0, len, DMA_BIDIRECTIONAL); + dma_src[0] = dma_map_page(dev, src, 0, len, DMA_TO_DEVICE); + tx = dma->device_prep_dma_pq(chan, dma_dest, dma_src, 1, &coef, + len, dma_flags); + if (tx) { + async_tx_submit(chan, tx, submit); + return tx; + } + } + + /* no channel available, or failed to allocate a descriptor, so + * perform the operation synchronously + */ + async_tx_quiesce(&submit->depend_tx); + qmul = raid6_gfmul[coef]; + d = page_address(dest); + s = page_address(src); + + while (len--) + *d++ = qmul[*s++]; + + return NULL; +} + +static struct dma_async_tx_descriptor * +__2data_recov_4(size_t bytes, int faila, int failb, struct page **blocks, + struct async_submit_ctl *submit) +{ + struct dma_async_tx_descriptor *tx = NULL; + struct page *p, *q, *a, *b; + struct page *srcs[2]; + unsigned char coef[2]; + enum async_tx_flags flags = submit->flags; + dma_async_tx_callback cb_fn = submit->cb_fn; + void *cb_param = submit->cb_param; + void *scribble = submit->scribble; + + p = blocks[4-2]; + q = blocks[4-1]; + + a = blocks[faila]; + b = blocks[failb]; + + /* in the 4 disk case P + Pxy == P and Q + Qxy == Q */ + /* Dx = A*(P+Pxy) + B*(Q+Qxy) */ + srcs[0] = p; + srcs[1] = q; + coef[0] = raid6_gfexi[failb-faila]; + coef[1] = raid6_gfinv[raid6_gfexp[faila]^raid6_gfexp[failb]]; + init_async_submit(submit, 0, tx, NULL, NULL, scribble); + tx = async_sum_product(b, srcs, coef, bytes, submit); + + /* Dy = P+Pxy+Dx */ + srcs[0] = p; + srcs[1] = b; + init_async_submit(submit, flags | ASYNC_TX_XOR_ZERO_DST, tx, cb_fn, + cb_param, scribble); + tx = async_xor(a, srcs, 0, 2, bytes, submit); + + return tx; + +} + +static struct dma_async_tx_descriptor * +__2data_recov_5(size_t bytes, int faila, int failb, struct page **blocks, + struct async_submit_ctl *submit) +{ + struct dma_async_tx_descriptor *tx = NULL; + struct page *p, *q, *g, *dp, *dq; + struct page *srcs[2]; + unsigned char coef[2]; + enum async_tx_flags flags = submit->flags; + dma_async_tx_callback cb_fn = submit->cb_fn; + void *cb_param = submit->cb_param; + void *scribble = submit->scribble; + int uninitialized_var(good); + int i; + + for (i = 0; i < 3; i++) { + if (i == faila || i == failb) + continue; + else { + good = i; + break; + } + } + BUG_ON(i >= 3); + + p = blocks[5-2]; + q = blocks[5-1]; + g = blocks[good]; + + /* Compute syndrome with zero for the missing data pages + * Use the dead data pages as temporary storage for delta p and + * delta q + */ + dp = blocks[faila]; + dq = blocks[failb]; + + init_async_submit(submit, 0, tx, NULL, NULL, scribble); + tx = async_memcpy(dp, g, 0, 0, bytes, submit); + init_async_submit(submit, 0, tx, NULL, NULL, scribble); + tx = async_mult(dq, g, raid6_gfexp[good], bytes, submit); + + /* compute P + Pxy */ + srcs[0] = dp; + srcs[1] = p; + init_async_submit(submit, ASYNC_TX_XOR_DROP_DST, tx, NULL, NULL, + scribble); + tx = async_xor(dp, srcs, 0, 2, bytes, submit); + + /* compute Q + Qxy */ + srcs[0] = dq; + srcs[1] = q; + init_async_submit(submit, ASYNC_TX_XOR_DROP_DST, tx, NULL, NULL, + scribble); + tx = async_xor(dq, srcs, 0, 2, bytes, submit); + + /* Dx = A*(P+Pxy) + B*(Q+Qxy) */ + srcs[0] = dp; + srcs[1] = dq; + coef[0] = raid6_gfexi[failb-faila]; + coef[1] = raid6_gfinv[raid6_gfexp[faila]^raid6_gfexp[failb]]; + init_async_submit(submit, 0, tx, NULL, NULL, scribble); + tx = async_sum_product(dq, srcs, coef, bytes, submit); + + /* Dy = P+Pxy+Dx */ + srcs[0] = dp; + srcs[1] = dq; + init_async_submit(submit, flags | ASYNC_TX_XOR_DROP_DST, tx, cb_fn, + cb_param, scribble); + tx = async_xor(dp, srcs, 0, 2, bytes, submit); + + return tx; +} + +static struct dma_async_tx_descriptor * +__2data_recov_n(int disks, size_t bytes, int faila, int failb, + struct page **blocks, struct async_submit_ctl *submit) +{ + struct dma_async_tx_descriptor *tx = NULL; + struct page *p, *q, *dp, *dq; + struct page *srcs[2]; + unsigned char coef[2]; + enum async_tx_flags flags = submit->flags; + dma_async_tx_callback cb_fn = submit->cb_fn; + void *cb_param = submit->cb_param; + void *scribble = submit->scribble; + + p = blocks[disks-2]; + q = blocks[disks-1]; + + /* Compute syndrome with zero for the missing data pages + * Use the dead data pages as temporary storage for + * delta p and delta q + */ + dp = blocks[faila]; + blocks[faila] = (void *)raid6_empty_zero_page; + blocks[disks-2] = dp; + dq = blocks[failb]; + blocks[failb] = (void *)raid6_empty_zero_page; + blocks[disks-1] = dq; + + init_async_submit(submit, 0, tx, NULL, NULL, scribble); + tx = async_gen_syndrome(blocks, 0, disks, bytes, submit); + + /* Restore pointer table */ + blocks[faila] = dp; + blocks[failb] = dq; + blocks[disks-2] = p; + blocks[disks-1] = q; + + /* compute P + Pxy */ + srcs[0] = dp; + srcs[1] = p; + init_async_submit(submit, ASYNC_TX_XOR_DROP_DST, tx, NULL, NULL, + scribble); + tx = async_xor(dp, srcs, 0, 2, bytes, submit); + + /* compute Q + Qxy */ + srcs[0] = dq; + srcs[1] = q; + init_async_submit(submit, ASYNC_TX_XOR_DROP_DST, tx, NULL, NULL, + scribble); + tx = async_xor(dq, srcs, 0, 2, bytes, submit); + + /* Dx = A*(P+Pxy) + B*(Q+Qxy) */ + srcs[0] = dp; + srcs[1] = dq; + coef[0] = raid6_gfexi[failb-faila]; + coef[1] = raid6_gfinv[raid6_gfexp[faila]^raid6_gfexp[failb]]; + init_async_submit(submit, 0, tx, NULL, NULL, scribble); + tx = async_sum_product(dq, srcs, coef, bytes, submit); + + /* Dy = P+Pxy+Dx */ + srcs[0] = dp; + srcs[1] = dq; + init_async_submit(submit, flags | ASYNC_TX_XOR_DROP_DST, tx, cb_fn, + cb_param, scribble); + tx = async_xor(dp, srcs, 0, 2, bytes, submit); + + return tx; +} + +/** + * async_raid6_2data_recov - asynchronously calculate two missing data blocks + * @disks: number of disks in the RAID-6 array + * @bytes: block size + * @faila: first failed drive index + * @failb: second failed drive index + * @blocks: array of source pointers where the last two entries are p and q + * @submit: submission/completion modifiers + */ +struct dma_async_tx_descriptor * +async_raid6_2data_recov(int disks, size_t bytes, int faila, int failb, + struct page **blocks, struct async_submit_ctl *submit) +{ + BUG_ON(faila == failb); + if (failb < faila) + swap(faila, failb); + + pr_debug("%s: disks: %d len: %zu\n", __func__, disks, bytes); + + /* we need to preserve the contents of 'blocks' for the async + * case, so punt to synchronous if a scribble buffer is not available + */ + if (!submit->scribble) { + void **ptrs = (void **) blocks; + int i; + + async_tx_quiesce(&submit->depend_tx); + for (i = 0; i < disks; i++) + ptrs[i] = page_address(blocks[i]); + + raid6_2data_recov(disks, bytes, faila, failb, ptrs); + + async_tx_sync_epilog(submit); + + return NULL; + } + + switch (disks) { + case 4: + /* dma devices do not uniformly understand a zero source pq + * operation (in contrast to the synchronous case), so + * explicitly handle the 4 disk special case + */ + return __2data_recov_4(bytes, faila, failb, blocks, submit); + case 5: + /* dma devices do not uniformly understand a single + * source pq operation (in contrast to the synchronous + * case), so explicitly handle the 5 disk special case + */ + return __2data_recov_5(bytes, faila, failb, blocks, submit); + default: + return __2data_recov_n(disks, bytes, faila, failb, blocks, submit); + } +} +EXPORT_SYMBOL_GPL(async_raid6_2data_recov); + +/** + * async_raid6_datap_recov - asynchronously calculate a data and the 'p' block + * @disks: number of disks in the RAID-6 array + * @bytes: block size + * @faila: failed drive index + * @blocks: array of source pointers where the last two entries are p and q + * @submit: submission/completion modifiers + */ +struct dma_async_tx_descriptor * +async_raid6_datap_recov(int disks, size_t bytes, int faila, + struct page **blocks, struct async_submit_ctl *submit) +{ + struct dma_async_tx_descriptor *tx = NULL; + struct page *p, *q, *dq; + u8 coef; + enum async_tx_flags flags = submit->flags; + dma_async_tx_callback cb_fn = submit->cb_fn; + void *cb_param = submit->cb_param; + void *scribble = submit->scribble; + struct page *srcs[2]; + + pr_debug("%s: disks: %d len: %zu\n", __func__, disks, bytes); + + /* we need to preserve the contents of 'blocks' for the async + * case, so punt to synchronous if a scribble buffer is not available + */ + if (!scribble) { + void **ptrs = (void **) blocks; + int i; + + async_tx_quiesce(&submit->depend_tx); + for (i = 0; i < disks; i++) + ptrs[i] = page_address(blocks[i]); + + raid6_datap_recov(disks, bytes, faila, ptrs); + + async_tx_sync_epilog(submit); + + return NULL; + } + + p = blocks[disks-2]; + q = blocks[disks-1]; + + /* Compute syndrome with zero for the missing data page + * Use the dead data page as temporary storage for delta q + */ + dq = blocks[faila]; + blocks[faila] = (void *)raid6_empty_zero_page; + blocks[disks-1] = dq; + + /* in the 4 disk case we only need to perform a single source + * multiplication + */ + if (disks == 4) { + int good = faila == 0 ? 1 : 0; + struct page *g = blocks[good]; + + init_async_submit(submit, 0, tx, NULL, NULL, scribble); + tx = async_memcpy(p, g, 0, 0, bytes, submit); + + init_async_submit(submit, 0, tx, NULL, NULL, scribble); + tx = async_mult(dq, g, raid6_gfexp[good], bytes, submit); + } else { + init_async_submit(submit, 0, tx, NULL, NULL, scribble); + tx = async_gen_syndrome(blocks, 0, disks, bytes, submit); + } + + /* Restore pointer table */ + blocks[faila] = dq; + blocks[disks-1] = q; + + /* calculate g^{-faila} */ + coef = raid6_gfinv[raid6_gfexp[faila]]; + + srcs[0] = dq; + srcs[1] = q; + init_async_submit(submit, ASYNC_TX_XOR_DROP_DST, tx, NULL, NULL, + scribble); + tx = async_xor(dq, srcs, 0, 2, bytes, submit); + + init_async_submit(submit, 0, tx, NULL, NULL, scribble); + tx = async_mult(dq, dq, coef, bytes, submit); + + srcs[0] = p; + srcs[1] = dq; + init_async_submit(submit, flags | ASYNC_TX_XOR_DROP_DST, tx, cb_fn, + cb_param, scribble); + tx = async_xor(p, srcs, 0, 2, bytes, submit); + + return tx; +} +EXPORT_SYMBOL_GPL(async_raid6_datap_recov); + +MODULE_AUTHOR("Dan Williams "); +MODULE_DESCRIPTION("asynchronous RAID-6 recovery api"); +MODULE_LICENSE("GPL"); diff --git a/include/linux/async_tx.h b/include/linux/async_tx.h index e6ce5f004f98..866e61c4e2e0 100644 --- a/include/linux/async_tx.h +++ b/include/linux/async_tx.h @@ -194,5 +194,13 @@ async_syndrome_val(struct page **blocks, unsigned int offset, int src_cnt, size_t len, enum sum_check_flags *pqres, struct page *spare, struct async_submit_ctl *submit); +struct dma_async_tx_descriptor * +async_raid6_2data_recov(int src_num, size_t bytes, int faila, int failb, + struct page **ptrs, struct async_submit_ctl *submit); + +struct dma_async_tx_descriptor * +async_raid6_datap_recov(int src_num, size_t bytes, int faila, + struct page **ptrs, struct async_submit_ctl *submit); + void async_tx_quiesce(struct dma_async_tx_descriptor **tx); #endif /* _ASYNC_TX_H_ */ -- cgit v1.2.3 From acdfcd04d9df7d084ff752f82afad6ed4ad5f363 Mon Sep 17 00:00:00 2001 From: Aaro Koskinen Date: Fri, 28 Aug 2009 14:28:54 +0300 Subject: SLUB: fix ARCH_KMALLOC_MINALIGN cases 64 and 256 If the minalign is 64 bytes, then the 96 byte cache should not be created because it would conflict with the 128 byte cache. If the minalign is 256 bytes, patching the size_index table should not result in a buffer overrun. The calculation "(i - 1) / 8" used to access size_index[] is moved to a separate function as suggested by Christoph Lameter. Acked-by: Christoph Lameter Signed-off-by: Aaro Koskinen Signed-off-by: Pekka Enberg --- include/linux/slub_def.h | 6 ++---- mm/slub.c | 30 ++++++++++++++++++++++++------ 2 files changed, 26 insertions(+), 10 deletions(-) (limited to 'include') diff --git a/include/linux/slub_def.h b/include/linux/slub_def.h index 4dcbc2c71491..aa5d4a69d461 100644 --- a/include/linux/slub_def.h +++ b/include/linux/slub_def.h @@ -152,12 +152,10 @@ static __always_inline int kmalloc_index(size_t size) if (size <= KMALLOC_MIN_SIZE) return KMALLOC_SHIFT_LOW; -#if KMALLOC_MIN_SIZE <= 64 - if (size > 64 && size <= 96) + if (KMALLOC_MIN_SIZE <= 32 && size > 64 && size <= 96) return 1; - if (size > 128 && size <= 192) + if (KMALLOC_MIN_SIZE <= 64 && size > 128 && size <= 192) return 2; -#endif if (size <= 8) return 3; if (size <= 16) return 4; if (size <= 32) return 5; diff --git a/mm/slub.c b/mm/slub.c index e16c9fb1f48b..be493bd63c31 100644 --- a/mm/slub.c +++ b/mm/slub.c @@ -2825,6 +2825,11 @@ static s8 size_index[24] = { 2 /* 192 */ }; +static inline int size_index_elem(size_t bytes) +{ + return (bytes - 1) / 8; +} + static struct kmem_cache *get_slab(size_t size, gfp_t flags) { int index; @@ -2833,7 +2838,7 @@ static struct kmem_cache *get_slab(size_t size, gfp_t flags) if (!size) return ZERO_SIZE_PTR; - index = size_index[(size - 1) / 8]; + index = size_index[size_index_elem(size)]; } else index = fls(size - 1); @@ -3188,10 +3193,12 @@ void __init kmem_cache_init(void) slab_state = PARTIAL; /* Caches that are not of the two-to-the-power-of size */ - if (KMALLOC_MIN_SIZE <= 64) { + if (KMALLOC_MIN_SIZE <= 32) { create_kmalloc_cache(&kmalloc_caches[1], "kmalloc-96", 96, GFP_NOWAIT); caches++; + } + if (KMALLOC_MIN_SIZE <= 64) { create_kmalloc_cache(&kmalloc_caches[2], "kmalloc-192", 192, GFP_NOWAIT); caches++; @@ -3218,17 +3225,28 @@ void __init kmem_cache_init(void) BUILD_BUG_ON(KMALLOC_MIN_SIZE > 256 || (KMALLOC_MIN_SIZE & (KMALLOC_MIN_SIZE - 1))); - for (i = 8; i < KMALLOC_MIN_SIZE; i += 8) - size_index[(i - 1) / 8] = KMALLOC_SHIFT_LOW; + for (i = 8; i < KMALLOC_MIN_SIZE; i += 8) { + int elem = size_index_elem(i); + if (elem >= ARRAY_SIZE(size_index)) + break; + size_index[elem] = KMALLOC_SHIFT_LOW; + } - if (KMALLOC_MIN_SIZE == 128) { + if (KMALLOC_MIN_SIZE == 64) { + /* + * The 96 byte size cache is not used if the alignment + * is 64 byte. + */ + for (i = 64 + 8; i <= 96; i += 8) + size_index[size_index_elem(i)] = 7; + } else if (KMALLOC_MIN_SIZE == 128) { /* * The 192 byte sized cache is not used if the alignment * is 128 byte. Redirect kmalloc to use the 256 byte cache * instead. */ for (i = 128 + 8; i <= 192; i += 8) - size_index[(i - 1) / 8] = 8; + size_index[size_index_elem(i)] = 8; } slab_state = UP; -- cgit v1.2.3 From e500011ffa191d662ac64d4ada6a5187b3180e16 Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Sun, 30 Aug 2009 13:19:12 -0700 Subject: timers: Drop a function prototype Drop prototype for non-existent next_timer_interrupt() function. Signed-off-by: Randy Dunlap Cc: akpm LKML-Reference: <4A9ADEC0.70306@oracle.com> Signed-off-by: Thomas Gleixner --- include/linux/timer.h | 5 ----- 1 file changed, 5 deletions(-) (limited to 'include') diff --git a/include/linux/timer.h b/include/linux/timer.h index be62ec2ebea5..a2d1eb6cb3f0 100644 --- a/include/linux/timer.h +++ b/include/linux/timer.h @@ -173,11 +173,6 @@ extern int mod_timer_pinned(struct timer_list *timer, unsigned long expires); */ #define NEXT_TIMER_MAX_DELTA ((1UL << 30) - 1) -/* - * Return when the next timer-wheel timeout occurs (in absolute jiffies), - * locks the timer base: - */ -extern unsigned long next_timer_interrupt(void); /* * Return when the next timer-wheel timeout occurs (in absolute jiffies), * locks the timer base and does the comparison against the given -- cgit v1.2.3 From f380ef86916904e4b79f7bec599deb51057b2d0c Mon Sep 17 00:00:00 2001 From: Maarten Maathuis Date: Wed, 19 Aug 2009 00:56:44 +0200 Subject: drm/crtc_helper: place drm_helper_encoder_in_use() in the header file - The symbol was already exported. Signed-off-by: Maarten Maathuis Signed-off-by: Dave Airlie --- include/drm/drm_crtc_helper.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include') diff --git a/include/drm/drm_crtc_helper.h b/include/drm/drm_crtc_helper.h index 6769ff6c1bc0..e44a4f87303c 100644 --- a/include/drm/drm_crtc_helper.h +++ b/include/drm/drm_crtc_helper.h @@ -98,6 +98,7 @@ extern bool drm_crtc_helper_set_mode(struct drm_crtc *crtc, int x, int y, struct drm_framebuffer *old_fb); extern bool drm_helper_crtc_in_use(struct drm_crtc *crtc); +extern bool drm_helper_encoder_in_use(struct drm_encoder *encoder); extern void drm_helper_connector_dpms(struct drm_connector *connector, int mode); -- cgit v1.2.3 From 785b93ef8c309730c2de84ce9c229e40e2d01480 Mon Sep 17 00:00:00 2001 From: Dave Airlie Date: Fri, 28 Aug 2009 15:46:53 +1000 Subject: drm/kms: move driver specific fb common code to helper functions (v2) Initially I always meant this code to be shared, but things ran away from me before I got to it. This refactors the i915 and radeon kms fbdev interaction layers out into generic helpers + driver specific pieces. It moves all the panic/sysrq enhancements to the core file, and stores a linked list of kernel fbs. This could possibly be improved to only store the fb which has fbcon on it for panics etc. radeon retains some specific codes used for a big endian workaround. changes: fix oops in v1 fix freeing path for crtc_info Reviewed-by: Jesse Barnes Signed-off-by: Dave Airlie --- drivers/gpu/drm/Makefile | 3 +- drivers/gpu/drm/drm_fb_helper.c | 697 ++++++++++++++++++++++++++++++ drivers/gpu/drm/i915/i915_dma.c | 3 +- drivers/gpu/drm/i915/intel_display.c | 12 - drivers/gpu/drm/i915/intel_drv.h | 3 - drivers/gpu/drm/i915/intel_fb.c | 737 ++------------------------------ drivers/gpu/drm/radeon/radeon_display.c | 5 +- drivers/gpu/drm/radeon/radeon_fb.c | 670 ++++------------------------- drivers/gpu/drm/radeon/radeon_mode.h | 2 - include/drm/drm_crtc.h | 2 + include/drm/drm_fb_helper.h | 82 ++++ 11 files changed, 907 insertions(+), 1309 deletions(-) create mode 100644 drivers/gpu/drm/drm_fb_helper.c create mode 100644 include/drm/drm_fb_helper.h (limited to 'include') diff --git a/drivers/gpu/drm/Makefile b/drivers/gpu/drm/Makefile index 5f0aec4f082a..99071684de25 100644 --- a/drivers/gpu/drm/Makefile +++ b/drivers/gpu/drm/Makefile @@ -11,7 +11,8 @@ drm-y := drm_auth.o drm_bufs.o drm_cache.o \ drm_agpsupport.o drm_scatter.o ati_pcigart.o drm_pci.o \ drm_sysfs.o drm_hashtab.o drm_sman.o drm_mm.o \ drm_crtc.o drm_crtc_helper.o drm_modes.o drm_edid.o \ - drm_info.o drm_debugfs.o drm_encoder_slave.o + drm_info.o drm_debugfs.o drm_encoder_slave.o \ + drm_fb_helper.o drm-$(CONFIG_COMPAT) += drm_ioc32.o diff --git a/drivers/gpu/drm/drm_fb_helper.c b/drivers/gpu/drm/drm_fb_helper.c new file mode 100644 index 000000000000..d6ffea74a502 --- /dev/null +++ b/drivers/gpu/drm/drm_fb_helper.c @@ -0,0 +1,697 @@ +/* + * Copyright (c) 2006-2009 Red Hat Inc. + * Copyright (c) 2006-2008 Intel Corporation + * Copyright (c) 2007 Dave Airlie + * + * DRM framebuffer helper functions + * + * Permission to use, copy, modify, distribute, and sell this software and its + * documentation for any purpose is hereby granted without fee, provided that + * the above copyright notice appear in all copies and that both that copyright + * notice and this permission notice appear in supporting documentation, and + * that the name of the copyright holders not be used in advertising or + * publicity pertaining to distribution of the software without specific, + * written prior permission. The copyright holders make no representations + * about the suitability of this software for any purpose. It is provided "as + * is" without express or implied warranty. + * + * THE COPYRIGHT HOLDERS DISCLAIM ALL WARRANTIES WITH REGARD TO THIS SOFTWARE, + * INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS, IN NO + * EVENT SHALL THE COPYRIGHT HOLDERS BE LIABLE FOR ANY SPECIAL, INDIRECT OR + * CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, + * DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER + * TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE + * OF THIS SOFTWARE. + * + * Authors: + * Dave Airlie + * Jesse Barnes + */ +#include +#include +#include "drmP.h" +#include "drm_crtc.h" +#include "drm_fb_helper.h" +#include "drm_crtc_helper.h" + +static LIST_HEAD(kernel_fb_helper_list); + +bool drm_fb_helper_force_kernel_mode(void) +{ + int i = 0; + bool ret, error = false; + struct drm_fb_helper *helper; + + if (list_empty(&kernel_fb_helper_list)) + return false; + + list_for_each_entry(helper, &kernel_fb_helper_list, kernel_fb_list) { + for (i = 0; i < helper->crtc_count; i++) { + struct drm_mode_set *mode_set = &helper->crtc_info[i].mode_set; + ret = drm_crtc_helper_set_config(mode_set); + if (ret) + error = true; + } + } + return error; +} + +int drm_fb_helper_panic(struct notifier_block *n, unsigned long ununsed, + void *panic_str) +{ + DRM_ERROR("panic occurred, switching back to text console\n"); + return drm_fb_helper_force_kernel_mode(); + return 0; +} +EXPORT_SYMBOL(drm_fb_helper_panic); + +static struct notifier_block paniced = { + .notifier_call = drm_fb_helper_panic, +}; + +/** + * drm_fb_helper_restore - restore the framebuffer console (kernel) config + * + * Restore's the kernel's fbcon mode, used for lastclose & panic paths. + */ +void drm_fb_helper_restore(void) +{ + bool ret; + ret = drm_fb_helper_force_kernel_mode(); + if (ret == true) + DRM_ERROR("Failed to restore crtc configuration\n"); +} +EXPORT_SYMBOL(drm_fb_helper_restore); + +static void drm_fb_helper_restore_work_fn(struct work_struct *ignored) +{ + drm_fb_helper_restore(); +} +static DECLARE_WORK(drm_fb_helper_restore_work, drm_fb_helper_restore_work_fn); + +static void drm_fb_helper_sysrq(int dummy1, struct tty_struct *dummy3) +{ + schedule_work(&drm_fb_helper_restore_work); +} + +static struct sysrq_key_op sysrq_drm_fb_helper_restore_op = { + .handler = drm_fb_helper_sysrq, + .help_msg = "force-fb(V)", + .action_msg = "Restore framebuffer console", +}; + +static void drm_fb_helper_on(struct fb_info *info) +{ + struct drm_fb_helper *fb_helper = info->par; + struct drm_device *dev = fb_helper->dev; + struct drm_crtc *crtc; + struct drm_encoder *encoder; + int i; + + /* + * For each CRTC in this fb, turn the crtc on then, + * find all associated encoders and turn them on. + */ + list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) { + struct drm_crtc_helper_funcs *crtc_funcs = crtc->helper_private; + + for (i = 0; i < fb_helper->crtc_count; i++) { + if (crtc->base.id == fb_helper->crtc_info[i].crtc_id) + break; + } + + mutex_lock(&dev->mode_config.mutex); + crtc_funcs->dpms(crtc, DRM_MODE_DPMS_ON); + mutex_unlock(&dev->mode_config.mutex); + + /* Found a CRTC on this fb, now find encoders */ + list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) { + if (encoder->crtc == crtc) { + struct drm_encoder_helper_funcs *encoder_funcs; + + encoder_funcs = encoder->helper_private; + mutex_lock(&dev->mode_config.mutex); + encoder_funcs->dpms(encoder, DRM_MODE_DPMS_ON); + mutex_unlock(&dev->mode_config.mutex); + } + } + } +} + +static void drm_fb_helper_off(struct fb_info *info, int dpms_mode) +{ + struct drm_fb_helper *fb_helper = info->par; + struct drm_device *dev = fb_helper->dev; + struct drm_crtc *crtc; + struct drm_encoder *encoder; + int i; + + /* + * For each CRTC in this fb, find all associated encoders + * and turn them off, then turn off the CRTC. + */ + list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) { + struct drm_crtc_helper_funcs *crtc_funcs = crtc->helper_private; + + for (i = 0; i < fb_helper->crtc_count; i++) { + if (crtc->base.id == fb_helper->crtc_info[i].crtc_id) + break; + } + + /* Found a CRTC on this fb, now find encoders */ + list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) { + if (encoder->crtc == crtc) { + struct drm_encoder_helper_funcs *encoder_funcs; + + encoder_funcs = encoder->helper_private; + mutex_lock(&dev->mode_config.mutex); + encoder_funcs->dpms(encoder, dpms_mode); + mutex_unlock(&dev->mode_config.mutex); + } + } + if (dpms_mode == DRM_MODE_DPMS_OFF) { + mutex_lock(&dev->mode_config.mutex); + crtc_funcs->dpms(crtc, dpms_mode); + mutex_unlock(&dev->mode_config.mutex); + } + } +} + +int drm_fb_helper_blank(int blank, struct fb_info *info) +{ + switch (blank) { + case FB_BLANK_UNBLANK: + drm_fb_helper_on(info); + break; + case FB_BLANK_NORMAL: + drm_fb_helper_off(info, DRM_MODE_DPMS_STANDBY); + break; + case FB_BLANK_HSYNC_SUSPEND: + drm_fb_helper_off(info, DRM_MODE_DPMS_STANDBY); + break; + case FB_BLANK_VSYNC_SUSPEND: + drm_fb_helper_off(info, DRM_MODE_DPMS_SUSPEND); + break; + case FB_BLANK_POWERDOWN: + drm_fb_helper_off(info, DRM_MODE_DPMS_OFF); + break; + } + return 0; +} +EXPORT_SYMBOL(drm_fb_helper_blank); + +static void drm_fb_helper_crtc_free(struct drm_fb_helper *helper) +{ + int i; + + for (i = 0; i < helper->crtc_count; i++) + kfree(helper->crtc_info[i].mode_set.connectors); + kfree(helper->crtc_info); +} + +int drm_fb_helper_init_crtc_count(struct drm_fb_helper *helper, int crtc_count, int max_conn_count) +{ + struct drm_device *dev = helper->dev; + struct drm_crtc *crtc; + int ret = 0; + int i; + + helper->crtc_info = kcalloc(crtc_count, sizeof(struct drm_fb_helper_crtc), GFP_KERNEL); + if (!helper->crtc_info) + return -ENOMEM; + + helper->crtc_count = crtc_count; + + for (i = 0; i < crtc_count; i++) { + helper->crtc_info[i].mode_set.connectors = + kcalloc(max_conn_count, + sizeof(struct drm_connector *), + GFP_KERNEL); + + if (!helper->crtc_info[i].mode_set.connectors) { + ret = -ENOMEM; + goto out_free; + } + helper->crtc_info[i].mode_set.num_connectors = 0; + } + + i = 0; + list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) { + helper->crtc_info[i].crtc_id = crtc->base.id; + helper->crtc_info[i].mode_set.crtc = crtc; + i++; + } + helper->conn_limit = max_conn_count; + return 0; +out_free: + drm_fb_helper_crtc_free(helper); + return -ENOMEM; +} +EXPORT_SYMBOL(drm_fb_helper_init_crtc_count); + +int drm_fb_helper_setcolreg(unsigned regno, + unsigned red, + unsigned green, + unsigned blue, + unsigned transp, + struct fb_info *info) +{ + struct drm_fb_helper *fb_helper = info->par; + struct drm_device *dev = fb_helper->dev; + struct drm_crtc *crtc; + int i; + + list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) { + struct drm_framebuffer *fb = fb_helper->fb; + + for (i = 0; i < fb_helper->crtc_count; i++) { + if (crtc->base.id == fb_helper->crtc_info[i].crtc_id) + break; + } + if (i == fb_helper->crtc_count) + continue; + + if (regno > 255) + return 1; + + if (fb->depth == 8) { + fb_helper->funcs->gamma_set(crtc, red, green, blue, regno); + return 0; + } + + if (regno < 16) { + switch (fb->depth) { + case 15: + fb->pseudo_palette[regno] = ((red & 0xf800) >> 1) | + ((green & 0xf800) >> 6) | + ((blue & 0xf800) >> 11); + break; + case 16: + fb->pseudo_palette[regno] = (red & 0xf800) | + ((green & 0xfc00) >> 5) | + ((blue & 0xf800) >> 11); + break; + case 24: + case 32: + fb->pseudo_palette[regno] = + (((red >> 8) & 0xff) << info->var.red.offset) | + (((green >> 8) & 0xff) << info->var.green.offset) | + (((blue >> 8) & 0xff) << info->var.blue.offset); + break; + } + } + } + return 0; +} +EXPORT_SYMBOL(drm_fb_helper_setcolreg); + +int drm_fb_helper_check_var(struct fb_var_screeninfo *var, + struct fb_info *info) +{ + struct drm_fb_helper *fb_helper = info->par; + struct drm_framebuffer *fb = fb_helper->fb; + int depth; + + if (var->pixclock == -1 || !var->pixclock) + return -EINVAL; + + /* Need to resize the fb object !!! */ + if (var->xres > fb->width || var->yres > fb->height) { + DRM_ERROR("Requested width/height is greater than current fb " + "object %dx%d > %dx%d\n", var->xres, var->yres, + fb->width, fb->height); + DRM_ERROR("Need resizing code.\n"); + return -EINVAL; + } + + switch (var->bits_per_pixel) { + case 16: + depth = (var->green.length == 6) ? 16 : 15; + break; + case 32: + depth = (var->transp.length > 0) ? 32 : 24; + break; + default: + depth = var->bits_per_pixel; + break; + } + + switch (depth) { + case 8: + var->red.offset = 0; + var->green.offset = 0; + var->blue.offset = 0; + var->red.length = 8; + var->green.length = 8; + var->blue.length = 8; + var->transp.length = 0; + var->transp.offset = 0; + break; + case 15: + var->red.offset = 10; + var->green.offset = 5; + var->blue.offset = 0; + var->red.length = 5; + var->green.length = 5; + var->blue.length = 5; + var->transp.length = 1; + var->transp.offset = 15; + break; + case 16: + var->red.offset = 11; + var->green.offset = 5; + var->blue.offset = 0; + var->red.length = 5; + var->green.length = 6; + var->blue.length = 5; + var->transp.length = 0; + var->transp.offset = 0; + break; + case 24: + var->red.offset = 16; + var->green.offset = 8; + var->blue.offset = 0; + var->red.length = 8; + var->green.length = 8; + var->blue.length = 8; + var->transp.length = 0; + var->transp.offset = 0; + break; + case 32: + var->red.offset = 16; + var->green.offset = 8; + var->blue.offset = 0; + var->red.length = 8; + var->green.length = 8; + var->blue.length = 8; + var->transp.length = 8; + var->transp.offset = 24; + break; + default: + return -EINVAL; + } + return 0; +} +EXPORT_SYMBOL(drm_fb_helper_check_var); + +/* this will let fbcon do the mode init */ +int drm_fb_helper_set_par(struct fb_info *info) +{ + struct drm_fb_helper *fb_helper = info->par; + struct drm_device *dev = fb_helper->dev; + struct fb_var_screeninfo *var = &info->var; + struct drm_crtc *crtc; + int ret; + int i; + + if (var->pixclock != -1) { + DRM_ERROR("PIXEL CLCOK SET\n"); + return -EINVAL; + } + + list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) { + + for (i = 0; i < fb_helper->crtc_count; i++) { + if (crtc->base.id == fb_helper->crtc_info[i].crtc_id) + break; + } + if (i == fb_helper->crtc_count) + continue; + + if (crtc->fb == fb_helper->crtc_info[i].mode_set.fb) { + mutex_lock(&dev->mode_config.mutex); + ret = crtc->funcs->set_config(&fb_helper->crtc_info->mode_set); + mutex_unlock(&dev->mode_config.mutex); + if (ret) + return ret; + } + } + return 0; +} +EXPORT_SYMBOL(drm_fb_helper_set_par); + +int drm_fb_helper_pan_display(struct fb_var_screeninfo *var, + struct fb_info *info) +{ + struct drm_fb_helper *fb_helper = info->par; + struct drm_device *dev = fb_helper->dev; + struct drm_mode_set *modeset; + struct drm_crtc *crtc; + int ret = 0; + int i; + + list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) { + for (i = 0; i < fb_helper->crtc_count; i++) { + if (crtc->base.id == fb_helper->crtc_info[i].crtc_id) + break; + } + + if (i == fb_helper->crtc_count) + continue; + + modeset = &fb_helper->crtc_info[i].mode_set; + + modeset->x = var->xoffset; + modeset->y = var->yoffset; + + if (modeset->num_connectors) { + mutex_lock(&dev->mode_config.mutex); + ret = crtc->funcs->set_config(modeset); + mutex_unlock(&dev->mode_config.mutex); + if (!ret) { + info->var.xoffset = var->xoffset; + info->var.yoffset = var->yoffset; + } + } + } + return ret; +} +EXPORT_SYMBOL(drm_fb_helper_pan_display); + +int drm_fb_helper_single_fb_probe(struct drm_device *dev, + int (*fb_create)(struct drm_device *dev, + uint32_t fb_width, + uint32_t fb_height, + uint32_t surface_width, + uint32_t surface_height, + struct drm_framebuffer **fb_ptr)) +{ + struct drm_crtc *crtc; + struct drm_connector *connector; + unsigned int fb_width = (unsigned)-1, fb_height = (unsigned)-1; + unsigned int surface_width = 0, surface_height = 0; + int new_fb = 0; + int crtc_count = 0; + int ret, i, conn_count = 0; + struct fb_info *info; + struct drm_framebuffer *fb; + struct drm_mode_set *modeset = NULL; + struct drm_fb_helper *fb_helper; + + /* first up get a count of crtcs now in use and new min/maxes width/heights */ + list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) { + if (drm_helper_crtc_in_use(crtc)) { + if (crtc->desired_mode) { + if (crtc->desired_mode->hdisplay < fb_width) + fb_width = crtc->desired_mode->hdisplay; + + if (crtc->desired_mode->vdisplay < fb_height) + fb_height = crtc->desired_mode->vdisplay; + + if (crtc->desired_mode->hdisplay > surface_width) + surface_width = crtc->desired_mode->hdisplay; + + if (crtc->desired_mode->vdisplay > surface_height) + surface_height = crtc->desired_mode->vdisplay; + } + crtc_count++; + } + } + + if (crtc_count == 0 || fb_width == -1 || fb_height == -1) { + /* hmm everyone went away - assume VGA cable just fell out + and will come back later. */ + return 0; + } + + /* do we have an fb already? */ + if (list_empty(&dev->mode_config.fb_kernel_list)) { + ret = (*fb_create)(dev, fb_width, fb_height, surface_width, + surface_height, &fb); + if (ret) + return -EINVAL; + new_fb = 1; + } else { + fb = list_first_entry(&dev->mode_config.fb_kernel_list, + struct drm_framebuffer, filp_head); + + /* if someone hotplugs something bigger than we have already allocated, we are pwned. + As really we can't resize an fbdev that is in the wild currently due to fbdev + not really being designed for the lower layers moving stuff around under it. + - so in the grand style of things - punt. */ + if ((fb->width < surface_width) || + (fb->height < surface_height)) { + DRM_ERROR("Framebuffer not large enough to scale console onto.\n"); + return -EINVAL; + } + } + + info = fb->fbdev; + fb_helper = info->par; + + crtc_count = 0; + /* okay we need to setup new connector sets in the crtcs */ + list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) { + modeset = &fb_helper->crtc_info[crtc_count].mode_set; + modeset->fb = fb; + conn_count = 0; + list_for_each_entry(connector, &dev->mode_config.connector_list, head) { + if (connector->encoder) + if (connector->encoder->crtc == modeset->crtc) { + modeset->connectors[conn_count] = connector; + conn_count++; + if (conn_count > fb_helper->conn_limit) + BUG(); + } + } + + for (i = conn_count; i < fb_helper->conn_limit; i++) + modeset->connectors[i] = NULL; + + modeset->crtc = crtc; + crtc_count++; + + modeset->num_connectors = conn_count; + if (modeset->crtc->desired_mode) { + if (modeset->mode) + drm_mode_destroy(dev, modeset->mode); + modeset->mode = drm_mode_duplicate(dev, + modeset->crtc->desired_mode); + } + } + fb_helper->crtc_count = crtc_count; + fb_helper->fb = fb; + + if (new_fb) { + info->var.pixclock = -1; + if (register_framebuffer(info) < 0) + return -EINVAL; + } else { + drm_fb_helper_set_par(info); + } + printk(KERN_INFO "fb%d: %s frame buffer device\n", info->node, + info->fix.id); + + /* Switch back to kernel console on panic */ + /* multi card linked list maybe */ + if (list_empty(&kernel_fb_helper_list)) { + printk(KERN_INFO "registered panic notifier\n"); + atomic_notifier_chain_register(&panic_notifier_list, + &paniced); + register_sysrq_key('v', &sysrq_drm_fb_helper_restore_op); + } + list_add(&fb_helper->kernel_fb_list, &kernel_fb_helper_list); + return 0; +} +EXPORT_SYMBOL(drm_fb_helper_single_fb_probe); + +void drm_fb_helper_free(struct drm_fb_helper *helper) +{ + list_del(&helper->kernel_fb_list); + if (list_empty(&kernel_fb_helper_list)) { + printk(KERN_INFO "unregistered panic notifier\n"); + atomic_notifier_chain_unregister(&panic_notifier_list, + &paniced); + unregister_sysrq_key('v', &sysrq_drm_fb_helper_restore_op); + } + drm_fb_helper_crtc_free(helper); +} +EXPORT_SYMBOL(drm_fb_helper_free); + +void drm_fb_helper_fill_fix(struct fb_info *info, uint32_t pitch) +{ + info->fix.type = FB_TYPE_PACKED_PIXELS; + info->fix.visual = FB_VISUAL_TRUECOLOR; + info->fix.type_aux = 0; + info->fix.xpanstep = 1; /* doing it in hw */ + info->fix.ypanstep = 1; /* doing it in hw */ + info->fix.ywrapstep = 0; + info->fix.accel = FB_ACCEL_I830; + info->fix.type_aux = 0; + + info->fix.line_length = pitch; + return; +} +EXPORT_SYMBOL(drm_fb_helper_fill_fix); + +void drm_fb_helper_fill_var(struct fb_info *info, struct drm_framebuffer *fb, + uint32_t fb_width, uint32_t fb_height) +{ + info->pseudo_palette = fb->pseudo_palette; + info->var.xres_virtual = fb->width; + info->var.yres_virtual = fb->height; + info->var.bits_per_pixel = fb->bits_per_pixel; + info->var.xoffset = 0; + info->var.yoffset = 0; + info->var.activate = FB_ACTIVATE_NOW; + info->var.height = -1; + info->var.width = -1; + + switch (fb->depth) { + case 8: + info->var.red.offset = 0; + info->var.green.offset = 0; + info->var.blue.offset = 0; + info->var.red.length = 8; /* 8bit DAC */ + info->var.green.length = 8; + info->var.blue.length = 8; + info->var.transp.offset = 0; + info->var.transp.length = 0; + break; + case 15: + info->var.red.offset = 10; + info->var.green.offset = 5; + info->var.blue.offset = 0; + info->var.red.length = 5; + info->var.green.length = 5; + info->var.blue.length = 5; + info->var.transp.offset = 15; + info->var.transp.length = 1; + break; + case 16: + info->var.red.offset = 11; + info->var.green.offset = 5; + info->var.blue.offset = 0; + info->var.red.length = 5; + info->var.green.length = 6; + info->var.blue.length = 5; + info->var.transp.offset = 0; + break; + case 24: + info->var.red.offset = 16; + info->var.green.offset = 8; + info->var.blue.offset = 0; + info->var.red.length = 8; + info->var.green.length = 8; + info->var.blue.length = 8; + info->var.transp.offset = 0; + info->var.transp.length = 0; + break; + case 32: + info->var.red.offset = 16; + info->var.green.offset = 8; + info->var.blue.offset = 0; + info->var.red.length = 8; + info->var.green.length = 8; + info->var.blue.length = 8; + info->var.transp.offset = 24; + info->var.transp.length = 8; + break; + default: + break; + } + + info->var.xres = fb_width; + info->var.yres = fb_height; +} +EXPORT_SYMBOL(drm_fb_helper_fill_var); diff --git a/drivers/gpu/drm/i915/i915_dma.c b/drivers/gpu/drm/i915/i915_dma.c index 544d889b9b16..c628c3671394 100644 --- a/drivers/gpu/drm/i915/i915_dma.c +++ b/drivers/gpu/drm/i915/i915_dma.c @@ -29,6 +29,7 @@ #include "drmP.h" #include "drm.h" #include "drm_crtc_helper.h" +#include "drm_fb_helper.h" #include "intel_drv.h" #include "i915_drm.h" #include "i915_drv.h" @@ -1347,7 +1348,7 @@ void i915_driver_lastclose(struct drm_device * dev) drm_i915_private_t *dev_priv = dev->dev_private; if (!dev_priv || drm_core_check_feature(dev, DRIVER_MODESET)) { - intelfb_restore(); + drm_fb_helper_restore(); return; } diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index d6fce2133413..5fb7a4f4a427 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -3060,8 +3060,6 @@ static void intel_crtc_destroy(struct drm_crtc *crtc) { struct intel_crtc *intel_crtc = to_intel_crtc(crtc); - if (intel_crtc->mode_set.mode) - drm_mode_destroy(crtc->dev, intel_crtc->mode_set.mode); drm_crtc_cleanup(crtc); kfree(intel_crtc); } @@ -3107,16 +3105,6 @@ static void intel_crtc_init(struct drm_device *dev, int pipe) intel_crtc->cursor_addr = 0; intel_crtc->dpms_mode = DRM_MODE_DPMS_OFF; drm_crtc_helper_add(&intel_crtc->base, &intel_helper_funcs); - - intel_crtc->mode_set.crtc = &intel_crtc->base; - intel_crtc->mode_set.connectors = (struct drm_connector **)(intel_crtc + 1); - intel_crtc->mode_set.num_connectors = 0; - - if (i915_fbpercrtc) { - - - - } } int intel_get_pipe_from_crtc_id(struct drm_device *dev, void *data, diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index d6f92ea1b553..38910f8f30ed 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -96,9 +96,6 @@ struct intel_crtc { uint32_t cursor_addr; u8 lut_r[256], lut_g[256], lut_b[256]; int dpms_mode; - struct intel_framebuffer *fbdev_fb; - /* a mode_set for fbdev users on this crtc */ - struct drm_mode_set mode_set; }; #define to_intel_crtc(x) container_of(x, struct intel_crtc, base) diff --git a/drivers/gpu/drm/i915/intel_fb.c b/drivers/gpu/drm/i915/intel_fb.c index 1d30802e773e..3041530c3673 100644 --- a/drivers/gpu/drm/i915/intel_fb.c +++ b/drivers/gpu/drm/i915/intel_fb.c @@ -39,339 +39,34 @@ #include "drmP.h" #include "drm.h" #include "drm_crtc.h" +#include "drm_fb_helper.h" #include "intel_drv.h" #include "i915_drm.h" #include "i915_drv.h" struct intelfb_par { - struct drm_device *dev; - struct drm_display_mode *our_mode; + struct drm_fb_helper helper; struct intel_framebuffer *intel_fb; - int crtc_count; - /* crtc currently bound to this */ - uint32_t crtc_ids[2]; + struct drm_display_mode *our_mode; }; -static int intelfb_setcolreg(unsigned regno, unsigned red, unsigned green, - unsigned blue, unsigned transp, - struct fb_info *info) -{ - struct intelfb_par *par = info->par; - struct drm_device *dev = par->dev; - struct drm_crtc *crtc; - int i; - - list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) { - struct intel_crtc *intel_crtc = to_intel_crtc(crtc); - struct drm_mode_set *modeset = &intel_crtc->mode_set; - struct drm_framebuffer *fb = modeset->fb; - - for (i = 0; i < par->crtc_count; i++) - if (crtc->base.id == par->crtc_ids[i]) - break; - - if (i == par->crtc_count) - continue; - - - if (regno > 255) - return 1; - - if (fb->depth == 8) { - intel_crtc_fb_gamma_set(crtc, red, green, blue, regno); - return 0; - } - - if (regno < 16) { - switch (fb->depth) { - case 15: - fb->pseudo_palette[regno] = ((red & 0xf800) >> 1) | - ((green & 0xf800) >> 6) | - ((blue & 0xf800) >> 11); - break; - case 16: - fb->pseudo_palette[regno] = (red & 0xf800) | - ((green & 0xfc00) >> 5) | - ((blue & 0xf800) >> 11); - break; - case 24: - case 32: - fb->pseudo_palette[regno] = ((red & 0xff00) << 8) | - (green & 0xff00) | - ((blue & 0xff00) >> 8); - break; - } - } - } - return 0; -} - -static int intelfb_check_var(struct fb_var_screeninfo *var, - struct fb_info *info) -{ - struct intelfb_par *par = info->par; - struct intel_framebuffer *intel_fb = par->intel_fb; - struct drm_framebuffer *fb = &intel_fb->base; - int depth; - - if (var->pixclock == -1 || !var->pixclock) - return -EINVAL; - - /* Need to resize the fb object !!! */ - if (var->xres > fb->width || var->yres > fb->height) { - DRM_ERROR("Requested width/height is greater than current fb object %dx%d > %dx%d\n",var->xres,var->yres,fb->width,fb->height); - DRM_ERROR("Need resizing code.\n"); - return -EINVAL; - } - - switch (var->bits_per_pixel) { - case 16: - depth = (var->green.length == 6) ? 16 : 15; - break; - case 32: - depth = (var->transp.length > 0) ? 32 : 24; - break; - default: - depth = var->bits_per_pixel; - break; - } - - switch (depth) { - case 8: - var->red.offset = 0; - var->green.offset = 0; - var->blue.offset = 0; - var->red.length = 8; - var->green.length = 8; - var->blue.length = 8; - var->transp.length = 0; - var->transp.offset = 0; - break; - case 15: - var->red.offset = 10; - var->green.offset = 5; - var->blue.offset = 0; - var->red.length = 5; - var->green.length = 5; - var->blue.length = 5; - var->transp.length = 1; - var->transp.offset = 15; - break; - case 16: - var->red.offset = 11; - var->green.offset = 5; - var->blue.offset = 0; - var->red.length = 5; - var->green.length = 6; - var->blue.length = 5; - var->transp.length = 0; - var->transp.offset = 0; - break; - case 24: - var->red.offset = 16; - var->green.offset = 8; - var->blue.offset = 0; - var->red.length = 8; - var->green.length = 8; - var->blue.length = 8; - var->transp.length = 0; - var->transp.offset = 0; - break; - case 32: - var->red.offset = 16; - var->green.offset = 8; - var->blue.offset = 0; - var->red.length = 8; - var->green.length = 8; - var->blue.length = 8; - var->transp.length = 8; - var->transp.offset = 24; - break; - default: - return -EINVAL; - } - - return 0; -} - -/* this will let fbcon do the mode init */ -/* FIXME: take mode config lock? */ -static int intelfb_set_par(struct fb_info *info) -{ - struct intelfb_par *par = info->par; - struct drm_device *dev = par->dev; - struct fb_var_screeninfo *var = &info->var; - int i; - - DRM_DEBUG("%d %d\n", var->xres, var->pixclock); - - if (var->pixclock != -1) { - - DRM_ERROR("PIXEL CLOCK SET\n"); - return -EINVAL; - } else { - struct drm_crtc *crtc; - int ret; - - list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) { - struct intel_crtc *intel_crtc = to_intel_crtc(crtc); - - for (i = 0; i < par->crtc_count; i++) - if (crtc->base.id == par->crtc_ids[i]) - break; - - if (i == par->crtc_count) - continue; - - if (crtc->fb == intel_crtc->mode_set.fb) { - mutex_lock(&dev->mode_config.mutex); - ret = crtc->funcs->set_config(&intel_crtc->mode_set); - mutex_unlock(&dev->mode_config.mutex); - if (ret) - return ret; - } - } - return 0; - } -} - -static int intelfb_pan_display(struct fb_var_screeninfo *var, - struct fb_info *info) -{ - struct intelfb_par *par = info->par; - struct drm_device *dev = par->dev; - struct drm_mode_set *modeset; - struct drm_crtc *crtc; - struct intel_crtc *intel_crtc; - int ret = 0; - int i; - - list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) { - for (i = 0; i < par->crtc_count; i++) - if (crtc->base.id == par->crtc_ids[i]) - break; - - if (i == par->crtc_count) - continue; - - intel_crtc = to_intel_crtc(crtc); - modeset = &intel_crtc->mode_set; - - modeset->x = var->xoffset; - modeset->y = var->yoffset; - - if (modeset->num_connectors) { - mutex_lock(&dev->mode_config.mutex); - ret = crtc->funcs->set_config(modeset); - mutex_unlock(&dev->mode_config.mutex); - if (!ret) { - info->var.xoffset = var->xoffset; - info->var.yoffset = var->yoffset; - } - } - } - - return ret; -} - -static void intelfb_on(struct fb_info *info) -{ - struct intelfb_par *par = info->par; - struct drm_device *dev = par->dev; - struct drm_crtc *crtc; - struct drm_encoder *encoder; - int i; - - /* - * For each CRTC in this fb, find all associated encoders - * and turn them off, then turn off the CRTC. - */ - list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) { - struct drm_crtc_helper_funcs *crtc_funcs = crtc->helper_private; - - for (i = 0; i < par->crtc_count; i++) - if (crtc->base.id == par->crtc_ids[i]) - break; - - crtc_funcs->dpms(crtc, DRM_MODE_DPMS_ON); - - /* Found a CRTC on this fb, now find encoders */ - list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) { - if (encoder->crtc == crtc) { - struct drm_encoder_helper_funcs *encoder_funcs; - encoder_funcs = encoder->helper_private; - encoder_funcs->dpms(encoder, DRM_MODE_DPMS_ON); - } - } - } -} - -static void intelfb_off(struct fb_info *info, int dpms_mode) -{ - struct intelfb_par *par = info->par; - struct drm_device *dev = par->dev; - struct drm_crtc *crtc; - struct drm_encoder *encoder; - int i; - - /* - * For each CRTC in this fb, find all associated encoders - * and turn them off, then turn off the CRTC. - */ - list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) { - struct drm_crtc_helper_funcs *crtc_funcs = crtc->helper_private; - - for (i = 0; i < par->crtc_count; i++) - if (crtc->base.id == par->crtc_ids[i]) - break; - - /* Found a CRTC on this fb, now find encoders */ - list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) { - if (encoder->crtc == crtc) { - struct drm_encoder_helper_funcs *encoder_funcs; - encoder_funcs = encoder->helper_private; - encoder_funcs->dpms(encoder, dpms_mode); - } - } - if (dpms_mode == DRM_MODE_DPMS_OFF) - crtc_funcs->dpms(crtc, dpms_mode); - } -} - -static int intelfb_blank(int blank, struct fb_info *info) -{ - switch (blank) { - case FB_BLANK_UNBLANK: - intelfb_on(info); - break; - case FB_BLANK_NORMAL: - intelfb_off(info, DRM_MODE_DPMS_STANDBY); - break; - case FB_BLANK_HSYNC_SUSPEND: - intelfb_off(info, DRM_MODE_DPMS_STANDBY); - break; - case FB_BLANK_VSYNC_SUSPEND: - intelfb_off(info, DRM_MODE_DPMS_SUSPEND); - break; - case FB_BLANK_POWERDOWN: - intelfb_off(info, DRM_MODE_DPMS_OFF); - break; - } - return 0; -} - static struct fb_ops intelfb_ops = { .owner = THIS_MODULE, - .fb_check_var = intelfb_check_var, - .fb_set_par = intelfb_set_par, - .fb_setcolreg = intelfb_setcolreg, + .fb_check_var = drm_fb_helper_check_var, + .fb_set_par = drm_fb_helper_set_par, + .fb_setcolreg = drm_fb_helper_setcolreg, .fb_fillrect = cfb_fillrect, .fb_copyarea = cfb_copyarea, .fb_imageblit = cfb_imageblit, - .fb_pan_display = intelfb_pan_display, - .fb_blank = intelfb_blank, + .fb_pan_display = drm_fb_helper_pan_display, + .fb_blank = drm_fb_helper_blank, }; +static struct drm_fb_helper_funcs intel_fb_helper_funcs = { + .gamma_set = intel_crtc_fb_gamma_set, +}; + + /** * Curretly it is assumed that the old framebuffer is reused. * @@ -412,25 +107,10 @@ int intelfb_resize(struct drm_device *dev, struct drm_crtc *crtc) } EXPORT_SYMBOL(intelfb_resize); -static struct drm_mode_set kernelfb_mode; - -static int intelfb_panic(struct notifier_block *n, unsigned long ununsed, - void *panic_str) -{ - DRM_ERROR("panic occurred, switching back to text console\n"); - - intelfb_restore(); - return 0; -} - -static struct notifier_block paniced = { - .notifier_call = intelfb_panic, -}; - static int intelfb_create(struct drm_device *dev, uint32_t fb_width, uint32_t fb_height, uint32_t surface_width, uint32_t surface_height, - struct intel_framebuffer **intel_fb_p) + struct drm_framebuffer **fb_p) { struct fb_info *info; struct intelfb_par *par; @@ -479,7 +159,7 @@ static int intelfb_create(struct drm_device *dev, uint32_t fb_width, list_add(&fb->filp_head, &dev->mode_config.fb_kernel_list); intel_fb = to_intel_framebuffer(fb); - *intel_fb_p = intel_fb; + *fb_p = fb; info = framebuffer_alloc(sizeof(struct intelfb_par), device); if (!info) { @@ -489,21 +169,19 @@ static int intelfb_create(struct drm_device *dev, uint32_t fb_width, par = info->par; + par->helper.funcs = &intel_fb_helper_funcs; + par->helper.dev = dev; + ret = drm_fb_helper_init_crtc_count(&par->helper, 2, + INTELFB_CONN_LIMIT); + if (ret) + goto out_unref; + strcpy(info->fix.id, "inteldrmfb"); - info->fix.type = FB_TYPE_PACKED_PIXELS; - info->fix.visual = FB_VISUAL_TRUECOLOR; - info->fix.type_aux = 0; - info->fix.xpanstep = 1; /* doing it in hw */ - info->fix.ypanstep = 1; /* doing it in hw */ - info->fix.ywrapstep = 0; - info->fix.accel = FB_ACCEL_I830; - info->fix.type_aux = 0; info->flags = FBINFO_DEFAULT; info->fbops = &intelfb_ops; - info->fix.line_length = fb->pitch; /* setup aperture base/size for vesafb takeover */ info->aperture_base = dev->mode_config.fb_base; @@ -527,18 +205,8 @@ static int intelfb_create(struct drm_device *dev, uint32_t fb_width, // memset(info->screen_base, 0, size); - info->pseudo_palette = fb->pseudo_palette; - info->var.xres_virtual = fb->width; - info->var.yres_virtual = fb->height; - info->var.bits_per_pixel = fb->bits_per_pixel; - info->var.xoffset = 0; - info->var.yoffset = 0; - info->var.activate = FB_ACTIVATE_NOW; - info->var.height = -1; - info->var.width = -1; - - info->var.xres = fb_width; - info->var.yres = fb_height; + drm_fb_helper_fill_fix(info, fb->depth); + drm_fb_helper_fill_var(info, fb, fb_width, fb_height); /* FIXME: we really shouldn't expose mmio space at all */ info->fix.mmio_start = pci_resource_start(dev->pdev, mmio_bar); @@ -550,64 +218,9 @@ static int intelfb_create(struct drm_device *dev, uint32_t fb_width, info->pixmap.flags = FB_PIXMAP_SYSTEM; info->pixmap.scan_align = 1; - switch(fb->depth) { - case 8: - info->var.red.offset = 0; - info->var.green.offset = 0; - info->var.blue.offset = 0; - info->var.red.length = 8; /* 8bit DAC */ - info->var.green.length = 8; - info->var.blue.length = 8; - info->var.transp.offset = 0; - info->var.transp.length = 0; - break; - case 15: - info->var.red.offset = 10; - info->var.green.offset = 5; - info->var.blue.offset = 0; - info->var.red.length = 5; - info->var.green.length = 5; - info->var.blue.length = 5; - info->var.transp.offset = 15; - info->var.transp.length = 1; - break; - case 16: - info->var.red.offset = 11; - info->var.green.offset = 5; - info->var.blue.offset = 0; - info->var.red.length = 5; - info->var.green.length = 6; - info->var.blue.length = 5; - info->var.transp.offset = 0; - break; - case 24: - info->var.red.offset = 16; - info->var.green.offset = 8; - info->var.blue.offset = 0; - info->var.red.length = 8; - info->var.green.length = 8; - info->var.blue.length = 8; - info->var.transp.offset = 0; - info->var.transp.length = 0; - break; - case 32: - info->var.red.offset = 16; - info->var.green.offset = 8; - info->var.blue.offset = 0; - info->var.red.length = 8; - info->var.green.length = 8; - info->var.blue.length = 8; - info->var.transp.offset = 24; - info->var.transp.length = 8; - break; - default: - break; - } - fb->fbdev = info; par->intel_fb = intel_fb; - par->dev = dev; /* To allow resizeing without swapping buffers */ DRM_DEBUG("allocated %dx%d fb: 0x%08x, bo %p\n", intel_fb->base.width, @@ -625,307 +238,12 @@ out: return ret; } -static int intelfb_multi_fb_probe_crtc(struct drm_device *dev, struct drm_crtc *crtc) -{ - struct intel_crtc *intel_crtc = to_intel_crtc(crtc); - struct intel_framebuffer *intel_fb; - struct drm_framebuffer *fb; - struct drm_connector *connector; - struct fb_info *info; - struct intelfb_par *par; - struct drm_mode_set *modeset; - unsigned int width, height; - int new_fb = 0; - int ret, i, conn_count; - - if (!drm_helper_crtc_in_use(crtc)) - return 0; - - if (!crtc->desired_mode) - return 0; - - width = crtc->desired_mode->hdisplay; - height = crtc->desired_mode->vdisplay; - - /* is there an fb bound to this crtc already */ - if (!intel_crtc->mode_set.fb) { - ret = intelfb_create(dev, width, height, width, height, &intel_fb); - if (ret) - return -EINVAL; - new_fb = 1; - } else { - fb = intel_crtc->mode_set.fb; - intel_fb = to_intel_framebuffer(fb); - if ((intel_fb->base.width < width) || (intel_fb->base.height < height)) - return -EINVAL; - } - - info = intel_fb->base.fbdev; - par = info->par; - - modeset = &intel_crtc->mode_set; - modeset->fb = &intel_fb->base; - conn_count = 0; - list_for_each_entry(connector, &dev->mode_config.connector_list, head) { - if (connector->encoder) - if (connector->encoder->crtc == modeset->crtc) { - modeset->connectors[conn_count] = connector; - conn_count++; - if (conn_count > INTELFB_CONN_LIMIT) - BUG(); - } - } - - for (i = conn_count; i < INTELFB_CONN_LIMIT; i++) - modeset->connectors[i] = NULL; - - par->crtc_ids[0] = crtc->base.id; - - modeset->num_connectors = conn_count; - if (modeset->crtc->desired_mode) { - if (modeset->mode) - drm_mode_destroy(dev, modeset->mode); - modeset->mode = drm_mode_duplicate(dev, - modeset->crtc->desired_mode); - } - - par->crtc_count = 1; - - if (new_fb) { - info->var.pixclock = -1; - if (register_framebuffer(info) < 0) - return -EINVAL; - } else - intelfb_set_par(info); - - DRM_INFO("fb%d: %s frame buffer device\n", info->node, - info->fix.id); - - /* Switch back to kernel console on panic */ - kernelfb_mode = *modeset; - atomic_notifier_chain_register(&panic_notifier_list, &paniced); - DRM_DEBUG("registered panic notifier\n"); - - return 0; -} - -static int intelfb_multi_fb_probe(struct drm_device *dev) -{ - - struct drm_crtc *crtc; - int ret = 0; - - list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) { - ret = intelfb_multi_fb_probe_crtc(dev, crtc); - if (ret) - return ret; - } - return ret; -} - -static int intelfb_single_fb_probe(struct drm_device *dev) -{ - struct drm_crtc *crtc; - struct drm_connector *connector; - unsigned int fb_width = (unsigned)-1, fb_height = (unsigned)-1; - unsigned int surface_width = 0, surface_height = 0; - int new_fb = 0; - int crtc_count = 0; - int ret, i, conn_count = 0; - struct intel_framebuffer *intel_fb; - struct fb_info *info; - struct intelfb_par *par; - struct drm_mode_set *modeset = NULL; - - DRM_DEBUG("\n"); - - /* Get a count of crtcs now in use and new min/maxes width/heights */ - list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) { - if (!drm_helper_crtc_in_use(crtc)) - continue; - - crtc_count++; - if (!crtc->desired_mode) - continue; - - /* Smallest mode determines console size... */ - if (crtc->desired_mode->hdisplay < fb_width) - fb_width = crtc->desired_mode->hdisplay; - - if (crtc->desired_mode->vdisplay < fb_height) - fb_height = crtc->desired_mode->vdisplay; - - /* ... but largest for memory allocation dimensions */ - if (crtc->desired_mode->hdisplay > surface_width) - surface_width = crtc->desired_mode->hdisplay; - - if (crtc->desired_mode->vdisplay > surface_height) - surface_height = crtc->desired_mode->vdisplay; - } - - if (crtc_count == 0 || fb_width == -1 || fb_height == -1) { - /* hmm everyone went away - assume VGA cable just fell out - and will come back later. */ - DRM_DEBUG("no CRTCs available?\n"); - return 0; - } - -//fail - /* Find the fb for our new config */ - if (list_empty(&dev->mode_config.fb_kernel_list)) { - DRM_DEBUG("creating new fb (console size %dx%d, " - "buffer size %dx%d)\n", fb_width, fb_height, - surface_width, surface_height); - ret = intelfb_create(dev, fb_width, fb_height, surface_width, - surface_height, &intel_fb); - if (ret) - return -EINVAL; - new_fb = 1; - } else { - struct drm_framebuffer *fb; - - fb = list_first_entry(&dev->mode_config.fb_kernel_list, - struct drm_framebuffer, filp_head); - intel_fb = to_intel_framebuffer(fb); - - /* if someone hotplugs something bigger than we have already - * allocated, we are pwned. As really we can't resize an - * fbdev that is in the wild currently due to fbdev not really - * being designed for the lower layers moving stuff around - * under it. - * - so in the grand style of things - punt. - */ - if ((fb->width < surface_width) || - (fb->height < surface_height)) { - DRM_ERROR("fb not large enough for console\n"); - return -EINVAL; - } - } -// fail - - info = intel_fb->base.fbdev; - par = info->par; - - crtc_count = 0; - /* - * For each CRTC, set up the connector list for the CRTC's mode - * set configuration. - */ - list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) { - struct intel_crtc *intel_crtc = to_intel_crtc(crtc); - - modeset = &intel_crtc->mode_set; - modeset->fb = &intel_fb->base; - conn_count = 0; - list_for_each_entry(connector, &dev->mode_config.connector_list, - head) { - if (!connector->encoder) - continue; - - if(connector->encoder->crtc == modeset->crtc) { - modeset->connectors[conn_count++] = connector; - if (conn_count > INTELFB_CONN_LIMIT) - BUG(); - } - } - - /* Zero out remaining connector pointers */ - for (i = conn_count; i < INTELFB_CONN_LIMIT; i++) - modeset->connectors[i] = NULL; - - par->crtc_ids[crtc_count++] = crtc->base.id; - - modeset->num_connectors = conn_count; - if (modeset->crtc->desired_mode) { - if (modeset->mode) - drm_mode_destroy(dev, modeset->mode); - modeset->mode = drm_mode_duplicate(dev, - modeset->crtc->desired_mode); - } - } - par->crtc_count = crtc_count; - - if (new_fb) { - info->var.pixclock = -1; - if (register_framebuffer(info) < 0) - return -EINVAL; - } else - intelfb_set_par(info); - - DRM_INFO("fb%d: %s frame buffer device\n", info->node, - info->fix.id); - - /* Switch back to kernel console on panic */ - kernelfb_mode = *modeset; - atomic_notifier_chain_register(&panic_notifier_list, &paniced); - DRM_DEBUG("registered panic notifier\n"); - - return 0; -} - -/** - * intelfb_restore - restore the framebuffer console (kernel) config - * - * Restore's the kernel's fbcon mode, used for lastclose & panic paths. - */ -void intelfb_restore(void) -{ - int ret; - if ((ret = drm_crtc_helper_set_config(&kernelfb_mode)) != 0) { - DRM_ERROR("Failed to restore crtc configuration: %d\n", - ret); - } -} - -static void intelfb_restore_work_fn(struct work_struct *ignored) -{ - intelfb_restore(); -} -static DECLARE_WORK(intelfb_restore_work, intelfb_restore_work_fn); - -static void intelfb_sysrq(int dummy1, struct tty_struct *dummy3) -{ - schedule_work(&intelfb_restore_work); -} - -static struct sysrq_key_op sysrq_intelfb_restore_op = { - .handler = intelfb_sysrq, - .help_msg = "force-fb(V)", - .action_msg = "Restore framebuffer console", -}; - int intelfb_probe(struct drm_device *dev) { int ret; DRM_DEBUG("\n"); - - /* something has changed in the lower levels of hell - deal with it - here */ - - /* two modes : a) 1 fb to rule all crtcs. - b) one fb per crtc. - two actions 1) new connected device - 2) device removed. - case a/1 : if the fb surface isn't big enough - resize the surface fb. - if the fb size isn't big enough - resize fb into surface. - if everything big enough configure the new crtc/etc. - case a/2 : undo the configuration - possibly resize down the fb to fit the new configuration. - case b/1 : see if it is on a new crtc - setup a new fb and add it. - case b/2 : teardown the new fb. - */ - - /* mode a first */ - /* search for an fb */ - if (i915_fbpercrtc == 1) { - ret = intelfb_multi_fb_probe(dev); - } else { - ret = intelfb_single_fb_probe(dev); - } - - register_sysrq_key('v', &sysrq_intelfb_restore_op); - + ret = drm_fb_helper_single_fb_probe(dev, intelfb_create); return ret; } EXPORT_SYMBOL(intelfb_probe); @@ -940,13 +258,14 @@ int intelfb_remove(struct drm_device *dev, struct drm_framebuffer *fb) info = fb->fbdev; if (info) { + struct intelfb_par *par = info->par; unregister_framebuffer(info); iounmap(info->screen_base); + if (info->par) + drm_fb_helper_free(&par->helper); framebuffer_release(info); } - atomic_notifier_chain_unregister(&panic_notifier_list, &paniced); - memset(&kernelfb_mode, 0, sizeof(struct drm_mode_set)); return 0; } EXPORT_SYMBOL(intelfb_remove); diff --git a/drivers/gpu/drm/radeon/radeon_display.c b/drivers/gpu/drm/radeon/radeon_display.c index a8fa1bb84cf7..af035605d147 100644 --- a/drivers/gpu/drm/radeon/radeon_display.c +++ b/drivers/gpu/drm/radeon/radeon_display.c @@ -158,9 +158,6 @@ static void radeon_crtc_destroy(struct drm_crtc *crtc) { struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc); - if (radeon_crtc->mode_set.mode) { - drm_mode_destroy(crtc->dev, radeon_crtc->mode_set.mode); - } drm_crtc_cleanup(crtc); kfree(radeon_crtc); } @@ -189,9 +186,11 @@ static void radeon_crtc_init(struct drm_device *dev, int index) radeon_crtc->crtc_id = index; rdev->mode_info.crtcs[index] = radeon_crtc; +#if 0 radeon_crtc->mode_set.crtc = &radeon_crtc->base; radeon_crtc->mode_set.connectors = (struct drm_connector **)(radeon_crtc + 1); radeon_crtc->mode_set.num_connectors = 0; +#endif for (i = 0; i < 256; i++) { radeon_crtc->lut_r[i] = i << 2; diff --git a/drivers/gpu/drm/radeon/radeon_fb.c b/drivers/gpu/drm/radeon/radeon_fb.c index ec383edf5f38..ebb58959f418 100644 --- a/drivers/gpu/drm/radeon/radeon_fb.c +++ b/drivers/gpu/drm/radeon/radeon_fb.c @@ -28,15 +28,7 @@ */ #include -#include -#include -#include -#include -#include -#include -#include #include -#include #include "drmP.h" #include "drm.h" @@ -45,375 +37,86 @@ #include "radeon_drm.h" #include "radeon.h" +#include "drm_fb_helper.h" + struct radeon_fb_device { - struct radeon_device *rdev; - struct drm_display_mode *mode; + struct drm_fb_helper helper; struct radeon_framebuffer *rfb; - int crtc_count; - /* crtc currently bound to this */ - uint32_t crtc_ids[2]; + struct radeon_device *rdev; }; -static int radeonfb_setcolreg(unsigned regno, - unsigned red, - unsigned green, - unsigned blue, - unsigned transp, - struct fb_info *info) -{ - struct radeon_fb_device *rfbdev = info->par; - struct drm_device *dev = rfbdev->rdev->ddev; - struct drm_crtc *crtc; - int i; - - list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) { - struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc); - struct drm_mode_set *modeset = &radeon_crtc->mode_set; - struct drm_framebuffer *fb = modeset->fb; - - for (i = 0; i < rfbdev->crtc_count; i++) { - if (crtc->base.id == rfbdev->crtc_ids[i]) { - break; - } - } - if (i == rfbdev->crtc_count) { - continue; - } - if (regno > 255) { - return 1; - } - if (fb->depth == 8) { - radeon_crtc_fb_gamma_set(crtc, red, green, blue, regno); - return 0; - } - - if (regno < 16) { - switch (fb->depth) { - case 15: - fb->pseudo_palette[regno] = ((red & 0xf800) >> 1) | - ((green & 0xf800) >> 6) | - ((blue & 0xf800) >> 11); - break; - case 16: - fb->pseudo_palette[regno] = (red & 0xf800) | - ((green & 0xfc00) >> 5) | - ((blue & 0xf800) >> 11); - break; - case 24: - case 32: - fb->pseudo_palette[regno] = - (((red >> 8) & 0xff) << info->var.red.offset) | - (((green >> 8) & 0xff) << info->var.green.offset) | - (((blue >> 8) & 0xff) << info->var.blue.offset); - break; - } - } - } - return 0; -} - -static int radeonfb_check_var(struct fb_var_screeninfo *var, - struct fb_info *info) +static int radeon_fb_check_var(struct fb_var_screeninfo *var, + struct fb_info *info) { - struct radeon_fb_device *rfbdev = info->par; - struct radeon_framebuffer *rfb = rfbdev->rfb; - struct drm_framebuffer *fb = &rfb->base; - int depth; - - if (var->pixclock == -1 || !var->pixclock) { - return -EINVAL; - } - /* Need to resize the fb object !!! */ - if (var->xres > fb->width || var->yres > fb->height) { - DRM_ERROR("Requested width/height is greater than current fb " - "object %dx%d > %dx%d\n", var->xres, var->yres, - fb->width, fb->height); - DRM_ERROR("Need resizing code.\n"); - return -EINVAL; - } - - switch (var->bits_per_pixel) { - case 16: - depth = (var->green.length == 6) ? 16 : 15; - break; - case 32: - depth = (var->transp.length > 0) ? 32 : 24; - break; - default: - depth = var->bits_per_pixel; - break; - } - - switch (depth) { - case 8: - var->red.offset = 0; - var->green.offset = 0; - var->blue.offset = 0; - var->red.length = 8; - var->green.length = 8; - var->blue.length = 8; - var->transp.length = 0; - var->transp.offset = 0; - break; -#ifdef __LITTLE_ENDIAN - case 15: - var->red.offset = 10; - var->green.offset = 5; - var->blue.offset = 0; - var->red.length = 5; - var->green.length = 5; - var->blue.length = 5; - var->transp.length = 1; - var->transp.offset = 15; - break; - case 16: - var->red.offset = 11; - var->green.offset = 5; - var->blue.offset = 0; - var->red.length = 5; - var->green.length = 6; - var->blue.length = 5; - var->transp.length = 0; - var->transp.offset = 0; - break; - case 24: - var->red.offset = 16; - var->green.offset = 8; - var->blue.offset = 0; - var->red.length = 8; - var->green.length = 8; - var->blue.length = 8; - var->transp.length = 0; - var->transp.offset = 0; - break; - case 32: - var->red.offset = 16; - var->green.offset = 8; - var->blue.offset = 0; - var->red.length = 8; - var->green.length = 8; - var->blue.length = 8; - var->transp.length = 8; - var->transp.offset = 24; - break; -#else - case 24: - var->red.offset = 8; - var->green.offset = 16; - var->blue.offset = 24; - var->red.length = 8; - var->green.length = 8; - var->blue.length = 8; - var->transp.length = 0; - var->transp.offset = 0; - break; - case 32: - var->red.offset = 8; - var->green.offset = 16; - var->blue.offset = 24; - var->red.length = 8; - var->green.length = 8; - var->blue.length = 8; - var->transp.length = 8; - var->transp.offset = 0; - break; -#endif - default: - return -EINVAL; - } - return 0; -} - -/* this will let fbcon do the mode init */ -static int radeonfb_set_par(struct fb_info *info) -{ - struct radeon_fb_device *rfbdev = info->par; - struct drm_device *dev = rfbdev->rdev->ddev; - struct fb_var_screeninfo *var = &info->var; - struct drm_crtc *crtc; int ret; - int i; - - if (var->pixclock != -1) { - DRM_ERROR("PIXEL CLCOK SET\n"); - return -EINVAL; - } - - list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) { - struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc); - - for (i = 0; i < rfbdev->crtc_count; i++) { - if (crtc->base.id == rfbdev->crtc_ids[i]) { - break; - } - } - if (i == rfbdev->crtc_count) { - continue; - } - if (crtc->fb == radeon_crtc->mode_set.fb) { - mutex_lock(&dev->mode_config.mutex); - ret = crtc->funcs->set_config(&radeon_crtc->mode_set); - mutex_unlock(&dev->mode_config.mutex); - if (ret) { - return ret; - } - } - } - return 0; -} - -static int radeonfb_pan_display(struct fb_var_screeninfo *var, - struct fb_info *info) -{ - struct radeon_fb_device *rfbdev = info->par; - struct drm_device *dev = rfbdev->rdev->ddev; - struct drm_mode_set *modeset; - struct drm_crtc *crtc; - struct radeon_crtc *radeon_crtc; - int ret = 0; - int i; - - list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) { - for (i = 0; i < rfbdev->crtc_count; i++) { - if (crtc->base.id == rfbdev->crtc_ids[i]) { - break; - } - } - - if (i == rfbdev->crtc_count) { - continue; - } - - radeon_crtc = to_radeon_crtc(crtc); - modeset = &radeon_crtc->mode_set; - - modeset->x = var->xoffset; - modeset->y = var->yoffset; - - if (modeset->num_connectors) { - mutex_lock(&dev->mode_config.mutex); - ret = crtc->funcs->set_config(modeset); - mutex_unlock(&dev->mode_config.mutex); - if (!ret) { - info->var.xoffset = var->xoffset; - info->var.yoffset = var->yoffset; - } + ret = drm_fb_helper_check_var(var, info); + if (ret) + return ret; + + /* big endian override for radeon endian workaround */ +#ifdef __BIG_ENDIAN + { + int depth; + switch (var->bits_per_pixel) { + case 16: + depth = (var->green.length == 6) ? 16 : 15; + break; + case 32: + depth = (var->transp.length > 0) ? 32 : 24; + break; + default: + depth = var->bits_per_pixel; + break; } - } - return ret; -} - -static void radeonfb_on(struct fb_info *info) -{ - struct radeon_fb_device *rfbdev = info->par; - struct drm_device *dev = rfbdev->rdev->ddev; - struct drm_crtc *crtc; - struct drm_encoder *encoder; - int i; - - /* - * For each CRTC in this fb, find all associated encoders - * and turn them off, then turn off the CRTC. - */ - list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) { - struct drm_crtc_helper_funcs *crtc_funcs = crtc->helper_private; - - for (i = 0; i < rfbdev->crtc_count; i++) { - if (crtc->base.id == rfbdev->crtc_ids[i]) { - break; - } - } - - mutex_lock(&dev->mode_config.mutex); - crtc_funcs->dpms(crtc, DRM_MODE_DPMS_ON); - mutex_unlock(&dev->mode_config.mutex); - - /* Found a CRTC on this fb, now find encoders */ - list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) { - if (encoder->crtc == crtc) { - struct drm_encoder_helper_funcs *encoder_funcs; - - encoder_funcs = encoder->helper_private; - mutex_lock(&dev->mode_config.mutex); - encoder_funcs->dpms(encoder, DRM_MODE_DPMS_ON); - mutex_unlock(&dev->mode_config.mutex); - } - } - } -} - -static void radeonfb_off(struct fb_info *info, int dpms_mode) -{ - struct radeon_fb_device *rfbdev = info->par; - struct drm_device *dev = rfbdev->rdev->ddev; - struct drm_crtc *crtc; - struct drm_encoder *encoder; - int i; - - /* - * For each CRTC in this fb, find all associated encoders - * and turn them off, then turn off the CRTC. - */ - list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) { - struct drm_crtc_helper_funcs *crtc_funcs = crtc->helper_private; - - for (i = 0; i < rfbdev->crtc_count; i++) { - if (crtc->base.id == rfbdev->crtc_ids[i]) { - break; - } - } - - /* Found a CRTC on this fb, now find encoders */ - list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) { - if (encoder->crtc == crtc) { - struct drm_encoder_helper_funcs *encoder_funcs; - - encoder_funcs = encoder->helper_private; - mutex_lock(&dev->mode_config.mutex); - encoder_funcs->dpms(encoder, dpms_mode); - mutex_unlock(&dev->mode_config.mutex); - } - } - if (dpms_mode == DRM_MODE_DPMS_OFF) { - mutex_lock(&dev->mode_config.mutex); - crtc_funcs->dpms(crtc, dpms_mode); - mutex_unlock(&dev->mode_config.mutex); + switch (depth) { + case 8: + var->red.offset = 0; + var->green.offset = 0; + var->blue.offset = 0; + var->red.length = 8; + var->green.length = 8; + var->blue.length = 8; + var->transp.length = 0; + var->transp.offset = 0; + break; + case 24: + var->red.offset = 8; + var->green.offset = 16; + var->blue.offset = 24; + var->red.length = 8; + var->green.length = 8; + var->blue.length = 8; + var->transp.length = 0; + var->transp.offset = 0; + break; + case 32: + var->red.offset = 8; + var->green.offset = 16; + var->blue.offset = 24; + var->red.length = 8; + var->green.length = 8; + var->blue.length = 8; + var->transp.length = 8; + var->transp.offset = 0; + break; + default: + return -EINVAL; } } -} - -int radeonfb_blank(int blank, struct fb_info *info) -{ - switch (blank) { - case FB_BLANK_UNBLANK: - radeonfb_on(info); - break; - case FB_BLANK_NORMAL: - radeonfb_off(info, DRM_MODE_DPMS_STANDBY); - break; - case FB_BLANK_HSYNC_SUSPEND: - radeonfb_off(info, DRM_MODE_DPMS_STANDBY); - break; - case FB_BLANK_VSYNC_SUSPEND: - radeonfb_off(info, DRM_MODE_DPMS_SUSPEND); - break; - case FB_BLANK_POWERDOWN: - radeonfb_off(info, DRM_MODE_DPMS_OFF); - break; - } +#endif return 0; } static struct fb_ops radeonfb_ops = { .owner = THIS_MODULE, - .fb_check_var = radeonfb_check_var, - .fb_set_par = radeonfb_set_par, - .fb_setcolreg = radeonfb_setcolreg, + .fb_check_var = radeon_fb_check_var, + .fb_set_par = drm_fb_helper_set_par, + .fb_setcolreg = drm_fb_helper_setcolreg, .fb_fillrect = cfb_fillrect, .fb_copyarea = cfb_copyarea, .fb_imageblit = cfb_imageblit, - .fb_pan_display = radeonfb_pan_display, - .fb_blank = radeonfb_blank, + .fb_pan_display = drm_fb_helper_pan_display, + .fb_blank = drm_fb_helper_blank, }; /** @@ -456,21 +159,6 @@ int radeonfb_resize(struct drm_device *dev, struct drm_crtc *crtc) } EXPORT_SYMBOL(radeonfb_resize); -static struct drm_mode_set panic_mode; - -int radeonfb_panic(struct notifier_block *n, unsigned long ununsed, - void *panic_str) -{ - DRM_ERROR("panic occurred, switching back to text console\n"); - drm_crtc_helper_set_config(&panic_mode); - return 0; -} -EXPORT_SYMBOL(radeonfb_panic); - -static struct notifier_block paniced = { - .notifier_call = radeonfb_panic, -}; - static int radeon_align_pitch(struct radeon_device *rdev, int width, int bpp, bool tiled) { int aligned = width; @@ -495,11 +183,16 @@ static int radeon_align_pitch(struct radeon_device *rdev, int width, int bpp, bo return aligned; } -int radeonfb_create(struct radeon_device *rdev, +static struct drm_fb_helper_funcs radeon_fb_helper_funcs = { + .gamma_set = radeon_crtc_fb_gamma_set, +}; + +int radeonfb_create(struct drm_device *dev, uint32_t fb_width, uint32_t fb_height, uint32_t surface_width, uint32_t surface_height, - struct radeon_framebuffer **rfb_p) + struct drm_framebuffer **fb_p) { + struct radeon_device *rdev = dev->dev_private; struct fb_info *info; struct radeon_fb_device *rfbdev; struct drm_framebuffer *fb = NULL; @@ -554,8 +247,8 @@ int radeonfb_create(struct radeon_device *rdev, list_add(&fb->filp_head, &rdev->ddev->mode_config.fb_kernel_list); + *fb_p = fb; rfb = to_radeon_framebuffer(fb); - *rfb_p = rfb; rdev->fbdev_rfb = rfb; rdev->fbdev_robj = robj; @@ -564,7 +257,14 @@ int radeonfb_create(struct radeon_device *rdev, ret = -ENOMEM; goto out_unref; } + rfbdev = info->par; + rfbdev->helper.funcs = &radeon_fb_helper_funcs; + rfbdev->helper.dev = dev; + ret = drm_fb_helper_init_crtc_count(&rfbdev->helper, 2, + RADEONFB_CONN_LIMIT); + if (ret) + goto out_unref; if (fb_tiled) radeon_object_check_tiling(robj, 0, 0); @@ -577,33 +277,19 @@ int radeonfb_create(struct radeon_device *rdev, memset_io(fbptr, 0, aligned_size); strcpy(info->fix.id, "radeondrmfb"); - info->fix.type = FB_TYPE_PACKED_PIXELS; - info->fix.visual = FB_VISUAL_TRUECOLOR; - info->fix.type_aux = 0; - info->fix.xpanstep = 1; /* doing it in hw */ - info->fix.ypanstep = 1; /* doing it in hw */ - info->fix.ywrapstep = 0; - info->fix.accel = FB_ACCEL_NONE; - info->fix.type_aux = 0; + + drm_fb_helper_fill_fix(info, fb->pitch); + info->flags = FBINFO_DEFAULT; info->fbops = &radeonfb_ops; - info->fix.line_length = fb->pitch; + tmp = fb_gpuaddr - rdev->mc.vram_location; info->fix.smem_start = rdev->mc.aper_base + tmp; info->fix.smem_len = size; info->screen_base = fbptr; info->screen_size = size; - info->pseudo_palette = fb->pseudo_palette; - info->var.xres_virtual = fb->width; - info->var.yres_virtual = fb->height; - info->var.bits_per_pixel = fb->bits_per_pixel; - info->var.xoffset = 0; - info->var.yoffset = 0; - info->var.activate = FB_ACTIVATE_NOW; - info->var.height = -1; - info->var.width = -1; - info->var.xres = fb_width; - info->var.yres = fb_height; + + drm_fb_helper_fill_var(info, fb, fb_width, fb_height); /* setup aperture base/size for vesafb takeover */ info->aperture_base = rdev->ddev->mode_config.fb_base; @@ -626,6 +312,9 @@ int radeonfb_create(struct radeon_device *rdev, DRM_INFO("fb depth is %d\n", fb->depth); DRM_INFO(" pitch is %d\n", fb->pitch); +#ifdef __BIG_ENDIAN + /* fill var sets defaults for this stuff - override + on big endian */ switch (fb->depth) { case 8: info->var.red.offset = 0; @@ -637,47 +326,6 @@ int radeonfb_create(struct radeon_device *rdev, info->var.transp.offset = 0; info->var.transp.length = 0; break; -#ifdef __LITTLE_ENDIAN - case 15: - info->var.red.offset = 10; - info->var.green.offset = 5; - info->var.blue.offset = 0; - info->var.red.length = 5; - info->var.green.length = 5; - info->var.blue.length = 5; - info->var.transp.offset = 15; - info->var.transp.length = 1; - break; - case 16: - info->var.red.offset = 11; - info->var.green.offset = 5; - info->var.blue.offset = 0; - info->var.red.length = 5; - info->var.green.length = 6; - info->var.blue.length = 5; - info->var.transp.offset = 0; - break; - case 24: - info->var.red.offset = 16; - info->var.green.offset = 8; - info->var.blue.offset = 0; - info->var.red.length = 8; - info->var.green.length = 8; - info->var.blue.length = 8; - info->var.transp.offset = 0; - info->var.transp.length = 0; - break; - case 32: - info->var.red.offset = 16; - info->var.green.offset = 8; - info->var.blue.offset = 0; - info->var.red.length = 8; - info->var.green.length = 8; - info->var.blue.length = 8; - info->var.transp.offset = 24; - info->var.transp.length = 8; - break; -#else case 24: info->var.red.offset = 8; info->var.green.offset = 16; @@ -699,9 +347,9 @@ int radeonfb_create(struct radeon_device *rdev, info->var.transp.length = 8; break; default: -#endif break; } +#endif fb->fbdev = info; rfbdev->rfb = rfb; @@ -726,145 +374,10 @@ out: return ret; } -static int radeonfb_single_fb_probe(struct radeon_device *rdev) -{ - struct drm_crtc *crtc; - struct drm_connector *connector; - unsigned int fb_width = (unsigned)-1, fb_height = (unsigned)-1; - unsigned int surface_width = 0, surface_height = 0; - int new_fb = 0; - int crtc_count = 0; - int ret, i, conn_count = 0; - struct radeon_framebuffer *rfb; - struct fb_info *info; - struct radeon_fb_device *rfbdev; - struct drm_mode_set *modeset = NULL; - - /* first up get a count of crtcs now in use and new min/maxes width/heights */ - list_for_each_entry(crtc, &rdev->ddev->mode_config.crtc_list, head) { - if (drm_helper_crtc_in_use(crtc)) { - if (crtc->desired_mode) { - if (crtc->desired_mode->hdisplay < fb_width) - fb_width = crtc->desired_mode->hdisplay; - - if (crtc->desired_mode->vdisplay < fb_height) - fb_height = crtc->desired_mode->vdisplay; - - if (crtc->desired_mode->hdisplay > surface_width) - surface_width = crtc->desired_mode->hdisplay; - - if (crtc->desired_mode->vdisplay > surface_height) - surface_height = crtc->desired_mode->vdisplay; - } - crtc_count++; - } - } - - if (crtc_count == 0 || fb_width == -1 || fb_height == -1) { - /* hmm everyone went away - assume VGA cable just fell out - and will come back later. */ - return 0; - } - - /* do we have an fb already? */ - if (list_empty(&rdev->ddev->mode_config.fb_kernel_list)) { - /* create an fb if we don't have one */ - ret = radeonfb_create(rdev, fb_width, fb_height, surface_width, surface_height, &rfb); - if (ret) { - return -EINVAL; - } - new_fb = 1; - } else { - struct drm_framebuffer *fb; - fb = list_first_entry(&rdev->ddev->mode_config.fb_kernel_list, struct drm_framebuffer, filp_head); - rfb = to_radeon_framebuffer(fb); - - /* if someone hotplugs something bigger than we have already allocated, we are pwned. - As really we can't resize an fbdev that is in the wild currently due to fbdev - not really being designed for the lower layers moving stuff around under it. - - so in the grand style of things - punt. */ - if ((fb->width < surface_width) || (fb->height < surface_height)) { - DRM_ERROR("Framebuffer not large enough to scale console onto.\n"); - return -EINVAL; - } - } - - info = rfb->base.fbdev; - rdev->fbdev_info = info; - rfbdev = info->par; - - crtc_count = 0; - /* okay we need to setup new connector sets in the crtcs */ - list_for_each_entry(crtc, &rdev->ddev->mode_config.crtc_list, head) { - struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc); - modeset = &radeon_crtc->mode_set; - modeset->fb = &rfb->base; - conn_count = 0; - list_for_each_entry(connector, &rdev->ddev->mode_config.connector_list, head) { - if (connector->encoder) - if (connector->encoder->crtc == modeset->crtc) { - modeset->connectors[conn_count] = connector; - conn_count++; - if (conn_count > RADEONFB_CONN_LIMIT) - BUG(); - } - } - - for (i = conn_count; i < RADEONFB_CONN_LIMIT; i++) - modeset->connectors[i] = NULL; - - - rfbdev->crtc_ids[crtc_count++] = crtc->base.id; - - modeset->num_connectors = conn_count; - if (modeset->crtc->desired_mode) { - if (modeset->mode) { - drm_mode_destroy(rdev->ddev, modeset->mode); - } - modeset->mode = drm_mode_duplicate(rdev->ddev, - modeset->crtc->desired_mode); - } - } - rfbdev->crtc_count = crtc_count; - - if (new_fb) { - info->var.pixclock = -1; - if (register_framebuffer(info) < 0) - return -EINVAL; - } else { - radeonfb_set_par(info); - } - printk(KERN_INFO "fb%d: %s frame buffer device\n", info->node, - info->fix.id); - - /* Switch back to kernel console on panic */ - panic_mode = *modeset; - atomic_notifier_chain_register(&panic_notifier_list, &paniced); - printk(KERN_INFO "registered panic notifier\n"); - - return 0; -} - int radeonfb_probe(struct drm_device *dev) { int ret; - - /* something has changed in the lower levels of hell - deal with it - here */ - - /* two modes : a) 1 fb to rule all crtcs. - b) one fb per crtc. - two actions 1) new connected device - 2) device removed. - case a/1 : if the fb surface isn't big enough - resize the surface fb. - if the fb size isn't big enough - resize fb into surface. - if everything big enough configure the new crtc/etc. - case a/2 : undo the configuration - possibly resize down the fb to fit the new configuration. - case b/1 : see if it is on a new crtc - setup a new fb and add it. - case b/2 : teardown the new fb. - */ - ret = radeonfb_single_fb_probe(dev->dev_private); + ret = drm_fb_helper_single_fb_probe(dev, &radeonfb_create); return ret; } EXPORT_SYMBOL(radeonfb_probe); @@ -880,16 +393,17 @@ int radeonfb_remove(struct drm_device *dev, struct drm_framebuffer *fb) } info = fb->fbdev; if (info) { + struct radeon_fb_device *rfbdev = info->par; robj = rfb->obj->driver_private; unregister_framebuffer(info); radeon_object_kunmap(robj); radeon_object_unpin(robj); + drm_fb_helper_free(&rfbdev->helper); framebuffer_release(info); } printk(KERN_INFO "unregistered panic notifier\n"); - atomic_notifier_chain_unregister(&panic_notifier_list, &paniced); - memset(&panic_mode, 0, sizeof(struct drm_mode_set)); + return 0; } EXPORT_SYMBOL(radeonfb_remove); diff --git a/drivers/gpu/drm/radeon/radeon_mode.h b/drivers/gpu/drm/radeon/radeon_mode.h index 3b09a1f2d8f9..20e9509a7130 100644 --- a/drivers/gpu/drm/radeon/radeon_mode.h +++ b/drivers/gpu/drm/radeon/radeon_mode.h @@ -195,8 +195,6 @@ struct radeon_crtc { bool enabled; bool can_tile; uint32_t crtc_offset; - struct radeon_framebuffer *fbdev_fb; - struct drm_mode_set mode_set; struct drm_gem_object *cursor_bo; uint64_t cursor_addr; int cursor_width; diff --git a/include/drm/drm_crtc.h b/include/drm/drm_crtc.h index db92a83f8ca9..b0427a70fcbd 100644 --- a/include/drm/drm_crtc.h +++ b/include/drm/drm_crtc.h @@ -259,6 +259,8 @@ struct drm_framebuffer { void *fbdev; u32 pseudo_palette[17]; struct list_head filp_head; + /* if you are using the helper */ + void *helper_private; }; struct drm_property_blob { diff --git a/include/drm/drm_fb_helper.h b/include/drm/drm_fb_helper.h new file mode 100644 index 000000000000..88fffbdfa26f --- /dev/null +++ b/include/drm/drm_fb_helper.h @@ -0,0 +1,82 @@ +/* + * Copyright (c) 2006-2009 Red Hat Inc. + * Copyright (c) 2006-2008 Intel Corporation + * Copyright (c) 2007 Dave Airlie + * + * DRM framebuffer helper functions + * + * Permission to use, copy, modify, distribute, and sell this software and its + * documentation for any purpose is hereby granted without fee, provided that + * the above copyright notice appear in all copies and that both that copyright + * notice and this permission notice appear in supporting documentation, and + * that the name of the copyright holders not be used in advertising or + * publicity pertaining to distribution of the software without specific, + * written prior permission. The copyright holders make no representations + * about the suitability of this software for any purpose. It is provided "as + * is" without express or implied warranty. + * + * THE COPYRIGHT HOLDERS DISCLAIM ALL WARRANTIES WITH REGARD TO THIS SOFTWARE, + * INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS, IN NO + * EVENT SHALL THE COPYRIGHT HOLDERS BE LIABLE FOR ANY SPECIAL, INDIRECT OR + * CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, + * DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER + * TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE + * OF THIS SOFTWARE. + * + * Authors: + * Dave Airlie + * Jesse Barnes + */ +#ifndef DRM_FB_HELPER_H +#define DRM_FB_HELPER_H + +struct drm_fb_helper_crtc { + uint32_t crtc_id; + struct drm_mode_set mode_set; +}; + +struct drm_fb_helper_funcs { + void (*gamma_set)(struct drm_crtc *crtc, u16 red, u16 green, + u16 blue, int regno); +}; + +struct drm_fb_helper { + struct drm_framebuffer *fb; + struct drm_device *dev; + struct drm_display_mode *mode; + int crtc_count; + struct drm_fb_helper_crtc *crtc_info; + struct drm_fb_helper_funcs *funcs; + int conn_limit; + struct list_head kernel_fb_list; +}; + +int drm_fb_helper_single_fb_probe(struct drm_device *dev, + int (*fb_create)(struct drm_device *dev, + uint32_t fb_width, + uint32_t fb_height, + uint32_t surface_width, + uint32_t surface_height, + struct drm_framebuffer **fb_ptr)); +int drm_fb_helper_init_crtc_count(struct drm_fb_helper *helper, int crtc_count, + int max_conn); +void drm_fb_helper_free(struct drm_fb_helper *helper); +int drm_fb_helper_blank(int blank, struct fb_info *info); +int drm_fb_helper_pan_display(struct fb_var_screeninfo *var, + struct fb_info *info); +int drm_fb_helper_set_par(struct fb_info *info); +int drm_fb_helper_check_var(struct fb_var_screeninfo *var, + struct fb_info *info); +int drm_fb_helper_setcolreg(unsigned regno, + unsigned red, + unsigned green, + unsigned blue, + unsigned transp, + struct fb_info *info); + +void drm_fb_helper_restore(void); +void drm_fb_helper_fill_var(struct fb_info *info, struct drm_framebuffer *fb, + uint32_t fb_width, uint32_t fb_height); +void drm_fb_helper_fill_fix(struct fb_info *info, uint32_t pitch); + +#endif -- cgit v1.2.3 From 7b3d3e4fc685a7d7e0b4c207ce24dfbab5689eb0 Mon Sep 17 00:00:00 2001 From: Krishna Kumar Date: Sat, 29 Aug 2009 20:21:21 +0000 Subject: netdevice: Consolidate to use existing macros where available. Patch compiled and 32 simultaneous netperf testing ran fine. Signed-off-by: Krishna Kumar Signed-off-by: David S. Miller --- include/linux/netdevice.h | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) (limited to 'include') diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index 9192cdf5bd2c..60d3aac49ed4 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -1257,7 +1257,7 @@ static inline void netif_tx_wake_queue(struct netdev_queue *dev_queue) { #ifdef CONFIG_NETPOLL_TRAP if (netpoll_trap()) { - clear_bit(__QUEUE_STATE_XOFF, &dev_queue->state); + netif_tx_start_queue(dev_queue); return; } #endif @@ -1363,7 +1363,8 @@ static inline int netif_running(const struct net_device *dev) static inline void netif_start_subqueue(struct net_device *dev, u16 queue_index) { struct netdev_queue *txq = netdev_get_tx_queue(dev, queue_index); - clear_bit(__QUEUE_STATE_XOFF, &txq->state); + + netif_tx_start_queue(txq); } /** @@ -1380,7 +1381,7 @@ static inline void netif_stop_subqueue(struct net_device *dev, u16 queue_index) if (netpoll_trap()) return; #endif - set_bit(__QUEUE_STATE_XOFF, &txq->state); + netif_tx_stop_queue(txq); } /** @@ -1394,7 +1395,8 @@ static inline int __netif_subqueue_stopped(const struct net_device *dev, u16 queue_index) { struct netdev_queue *txq = netdev_get_tx_queue(dev, queue_index); - return test_bit(__QUEUE_STATE_XOFF, &txq->state); + + return netif_tx_queue_stopped(txq); } static inline int netif_subqueue_stopped(const struct net_device *dev, @@ -1746,8 +1748,7 @@ static inline void netif_tx_unlock(struct net_device *dev) * force a schedule. */ clear_bit(__QUEUE_STATE_FROZEN, &txq->state); - if (!test_bit(__QUEUE_STATE_XOFF, &txq->state)) - __netif_schedule(txq->qdisc); + netif_schedule_queue(txq); } spin_unlock(&dev->tx_global_lock); } -- cgit v1.2.3 From 8e254c1d183f0225ad21f9049641529e56cce4da Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Mon, 31 Aug 2009 16:49:41 +0800 Subject: tracing/filters: Defer pred allocation init_preds() allocates about 5392 bytes of memory (on x86_32) for a TRACE_EVENT. With my config, at system boot total memory occupied is: 5392 * (642 + 15) == 3459KB 642 == cat available_events | wc -l 15 == number of dirs in events/ftrace That's quite a lot, so we'd better defer memory allocation util it's needed, that's when filter is used. Signed-off-by: Li Zefan Cc: Steven Rostedt Cc: Frederic Weisbecker Cc: Tom Zanussi Cc: Masami Hiramatsu LKML-Reference: <4A9B8EA5.6020700@cn.fujitsu.com> Signed-off-by: Ingo Molnar --- include/linux/ftrace_event.h | 1 - include/linux/syscalls.h | 2 -- include/trace/ftrace.h | 1 - kernel/trace/trace_events_filter.c | 50 +++++++++++++++++++++++++++++++------- kernel/trace/trace_export.c | 1 - 5 files changed, 41 insertions(+), 14 deletions(-) (limited to 'include') diff --git a/include/linux/ftrace_event.h b/include/linux/ftrace_event.h index ace2da9e0a0d..755480484eb6 100644 --- a/include/linux/ftrace_event.h +++ b/include/linux/ftrace_event.h @@ -133,7 +133,6 @@ struct ftrace_event_call { #define MAX_FILTER_PRED 32 #define MAX_FILTER_STR_VAL 128 -extern int init_preds(struct ftrace_event_call *call); extern void destroy_preds(struct ftrace_event_call *call); extern int filter_match_preds(struct ftrace_event_call *call, void *rec); extern int filter_current_check_discard(struct ftrace_event_call *call, diff --git a/include/linux/syscalls.h b/include/linux/syscalls.h index f124c8995555..a8e37821cc60 100644 --- a/include/linux/syscalls.h +++ b/include/linux/syscalls.h @@ -177,7 +177,6 @@ static void prof_sysexit_disable_##sname(struct ftrace_event_call *event_call) \ event_enter_##sname.id = id; \ set_syscall_enter_id(num, id); \ INIT_LIST_HEAD(&event_enter_##sname.fields); \ - init_preds(&event_enter_##sname); \ return 0; \ } \ TRACE_SYS_ENTER_PROFILE(sname); \ @@ -214,7 +213,6 @@ static void prof_sysexit_disable_##sname(struct ftrace_event_call *event_call) \ event_exit_##sname.id = id; \ set_syscall_exit_id(num, id); \ INIT_LIST_HEAD(&event_exit_##sname.fields); \ - init_preds(&event_exit_##sname); \ return 0; \ } \ TRACE_SYS_EXIT_PROFILE(sname); \ diff --git a/include/trace/ftrace.h b/include/trace/ftrace.h index 57c56a998ee6..bfbc842600a1 100644 --- a/include/trace/ftrace.h +++ b/include/trace/ftrace.h @@ -622,7 +622,6 @@ static int ftrace_raw_init_event_##call(void) \ return -ENODEV; \ event_##call.id = id; \ INIT_LIST_HEAD(&event_##call.fields); \ - init_preds(&event_##call); \ return 0; \ } \ \ diff --git a/kernel/trace/trace_events_filter.c b/kernel/trace/trace_events_filter.c index 9f03082c81d8..c6b2edfb7fe9 100644 --- a/kernel/trace/trace_events_filter.c +++ b/kernel/trace/trace_events_filter.c @@ -309,7 +309,7 @@ void print_event_filter(struct ftrace_event_call *call, struct trace_seq *s) struct event_filter *filter = call->filter; mutex_lock(&event_mutex); - if (filter->filter_string) + if (filter && filter->filter_string) trace_seq_printf(s, "%s\n", filter->filter_string); else trace_seq_printf(s, "none\n"); @@ -322,7 +322,7 @@ void print_subsystem_event_filter(struct event_subsystem *system, struct event_filter *filter = system->filter; mutex_lock(&event_mutex); - if (filter->filter_string) + if (filter && filter->filter_string) trace_seq_printf(s, "%s\n", filter->filter_string); else trace_seq_printf(s, "none\n"); @@ -390,6 +390,9 @@ void destroy_preds(struct ftrace_event_call *call) struct event_filter *filter = call->filter; int i; + if (!filter) + return; + for (i = 0; i < MAX_FILTER_PRED; i++) { if (filter->preds[i]) filter_free_pred(filter->preds[i]); @@ -400,7 +403,7 @@ void destroy_preds(struct ftrace_event_call *call) call->filter = NULL; } -int init_preds(struct ftrace_event_call *call) +static int init_preds(struct ftrace_event_call *call) { struct event_filter *filter; struct filter_pred *pred; @@ -410,7 +413,6 @@ int init_preds(struct ftrace_event_call *call) if (!call->filter) return -ENOMEM; - call->filter_active = 0; filter->n_preds = 0; filter->preds = kzalloc(MAX_FILTER_PRED * sizeof(pred), GFP_KERNEL); @@ -432,7 +434,28 @@ oom: return -ENOMEM; } -EXPORT_SYMBOL_GPL(init_preds); + +static int init_subsystem_preds(struct event_subsystem *system) +{ + struct ftrace_event_call *call; + int err; + + list_for_each_entry(call, &ftrace_events, list) { + if (!call->define_fields) + continue; + + if (strcmp(call->system, system->name) != 0) + continue; + + if (!call->filter) { + err = init_preds(call); + if (err) + return err; + } + } + + return 0; +} enum { FILTER_DISABLE_ALL, @@ -449,6 +472,9 @@ static void filter_free_subsystem_preds(struct event_subsystem *system, if (!call->define_fields) continue; + if (strcmp(call->system, system->name) != 0) + continue; + if (flag == FILTER_INIT_NO_RESET) { call->filter->no_reset = false; continue; @@ -457,10 +483,8 @@ static void filter_free_subsystem_preds(struct event_subsystem *system, if (flag == FILTER_SKIP_NO_RESET && call->filter->no_reset) continue; - if (!strcmp(call->system, system->name)) { - filter_disable_preds(call); - remove_filter_string(call->filter); - } + filter_disable_preds(call); + remove_filter_string(call->filter); } } @@ -1094,6 +1118,10 @@ int apply_event_filter(struct ftrace_event_call *call, char *filter_string) mutex_lock(&event_mutex); + err = init_preds(call); + if (err) + goto out_unlock; + if (!strcmp(strstrip(filter_string), "0")) { filter_disable_preds(call); remove_filter_string(call->filter); @@ -1139,6 +1167,10 @@ int apply_subsystem_event_filter(struct event_subsystem *system, mutex_lock(&event_mutex); + err = init_subsystem_preds(system); + if (err) + goto out_unlock; + if (!strcmp(strstrip(filter_string), "0")) { filter_free_subsystem_preds(system, FILTER_DISABLE_ALL); remove_filter_string(system->filter); diff --git a/kernel/trace/trace_export.c b/kernel/trace/trace_export.c index 029a91f42287..df1bf6e48bb9 100644 --- a/kernel/trace/trace_export.c +++ b/kernel/trace/trace_export.c @@ -135,7 +135,6 @@ __attribute__((section("_ftrace_events"))) event_##call = { \ static int ftrace_raw_init_event_##call(void) \ { \ INIT_LIST_HEAD(&event_##call.fields); \ - init_preds(&event_##call); \ return 0; \ } \ -- cgit v1.2.3 From 69d0ee7377eef808e34ba5542b554ec97244b871 Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Mon, 31 Aug 2009 14:43:36 +0200 Subject: locking: Move spinlock function bodies to header file Move spinlock function bodies to header file by creating a static inline version of each variant. Use the inline version on the out-of-line code. This shouldn't make any difference besides that the spinlock code can now be used to generate inlined spinlock code. Signed-off-by: Heiko Carstens Acked-by: Arnd Bergmann Acked-by: Peter Zijlstra Cc: Nick Piggin Cc: Martin Schwidefsky Cc: Horst Hartmann Cc: Christian Ehrhardt Cc: Andrew Morton Cc: Linus Torvalds Cc: David Miller Cc: Benjamin Herrenschmidt Cc: Paul Mackerras Cc: Geert Uytterhoeven Cc: Roman Zippel Cc: LKML-Reference: <20090831124417.859022429@de.ibm.com> Signed-off-by: Ingo Molnar --- include/linux/spinlock.h | 18 +-- include/linux/spinlock_api_smp.h | 263 +++++++++++++++++++++++++++++++++++++++ kernel/spinlock.c | 174 +++++--------------------- 3 files changed, 300 insertions(+), 155 deletions(-) (limited to 'include') diff --git a/include/linux/spinlock.h b/include/linux/spinlock.h index 4be57ab03478..da76a06556bd 100644 --- a/include/linux/spinlock.h +++ b/include/linux/spinlock.h @@ -143,15 +143,6 @@ static inline void smp_mb__after_lock(void) { smp_mb(); } */ #define spin_unlock_wait(lock) __raw_spin_unlock_wait(&(lock)->raw_lock) -/* - * Pull the _spin_*()/_read_*()/_write_*() functions/declarations: - */ -#if defined(CONFIG_SMP) || defined(CONFIG_DEBUG_SPINLOCK) -# include -#else -# include -#endif - #ifdef CONFIG_DEBUG_SPINLOCK extern void _raw_spin_lock(spinlock_t *lock); #define _raw_spin_lock_flags(lock, flags) _raw_spin_lock(lock) @@ -380,4 +371,13 @@ extern int _atomic_dec_and_lock(atomic_t *atomic, spinlock_t *lock); */ #define spin_can_lock(lock) (!spin_is_locked(lock)) +/* + * Pull the _spin_*()/_read_*()/_write_*() functions/declarations: + */ +#if defined(CONFIG_SMP) || defined(CONFIG_DEBUG_SPINLOCK) +# include +#else +# include +#endif + #endif /* __LINUX_SPINLOCK_H */ diff --git a/include/linux/spinlock_api_smp.h b/include/linux/spinlock_api_smp.h index d79845d034b5..6b108f5fb149 100644 --- a/include/linux/spinlock_api_smp.h +++ b/include/linux/spinlock_api_smp.h @@ -60,4 +60,267 @@ void __lockfunc _read_unlock_irqrestore(rwlock_t *lock, unsigned long flags) void __lockfunc _write_unlock_irqrestore(rwlock_t *lock, unsigned long flags) __releases(lock); +static inline int __spin_trylock(spinlock_t *lock) +{ + preempt_disable(); + if (_raw_spin_trylock(lock)) { + spin_acquire(&lock->dep_map, 0, 1, _RET_IP_); + return 1; + } + preempt_enable(); + return 0; +} + +static inline int __read_trylock(rwlock_t *lock) +{ + preempt_disable(); + if (_raw_read_trylock(lock)) { + rwlock_acquire_read(&lock->dep_map, 0, 1, _RET_IP_); + return 1; + } + preempt_enable(); + return 0; +} + +static inline int __write_trylock(rwlock_t *lock) +{ + preempt_disable(); + if (_raw_write_trylock(lock)) { + rwlock_acquire(&lock->dep_map, 0, 1, _RET_IP_); + return 1; + } + preempt_enable(); + return 0; +} + +/* + * If lockdep is enabled then we use the non-preemption spin-ops + * even on CONFIG_PREEMPT, because lockdep assumes that interrupts are + * not re-enabled during lock-acquire (which the preempt-spin-ops do): + */ +#if !defined(CONFIG_GENERIC_LOCKBREAK) || defined(CONFIG_DEBUG_LOCK_ALLOC) + +static inline void __read_lock(rwlock_t *lock) +{ + preempt_disable(); + rwlock_acquire_read(&lock->dep_map, 0, 0, _RET_IP_); + LOCK_CONTENDED(lock, _raw_read_trylock, _raw_read_lock); +} + +static inline unsigned long __spin_lock_irqsave(spinlock_t *lock) +{ + unsigned long flags; + + local_irq_save(flags); + preempt_disable(); + spin_acquire(&lock->dep_map, 0, 0, _RET_IP_); + /* + * On lockdep we dont want the hand-coded irq-enable of + * _raw_spin_lock_flags() code, because lockdep assumes + * that interrupts are not re-enabled during lock-acquire: + */ +#ifdef CONFIG_LOCKDEP + LOCK_CONTENDED(lock, _raw_spin_trylock, _raw_spin_lock); +#else + _raw_spin_lock_flags(lock, &flags); +#endif + return flags; +} + +static inline void __spin_lock_irq(spinlock_t *lock) +{ + local_irq_disable(); + preempt_disable(); + spin_acquire(&lock->dep_map, 0, 0, _RET_IP_); + LOCK_CONTENDED(lock, _raw_spin_trylock, _raw_spin_lock); +} + +static inline void __spin_lock_bh(spinlock_t *lock) +{ + local_bh_disable(); + preempt_disable(); + spin_acquire(&lock->dep_map, 0, 0, _RET_IP_); + LOCK_CONTENDED(lock, _raw_spin_trylock, _raw_spin_lock); +} + +static inline unsigned long __read_lock_irqsave(rwlock_t *lock) +{ + unsigned long flags; + + local_irq_save(flags); + preempt_disable(); + rwlock_acquire_read(&lock->dep_map, 0, 0, _RET_IP_); + LOCK_CONTENDED_FLAGS(lock, _raw_read_trylock, _raw_read_lock, + _raw_read_lock_flags, &flags); + return flags; +} + +static inline void __read_lock_irq(rwlock_t *lock) +{ + local_irq_disable(); + preempt_disable(); + rwlock_acquire_read(&lock->dep_map, 0, 0, _RET_IP_); + LOCK_CONTENDED(lock, _raw_read_trylock, _raw_read_lock); +} + +static inline void __read_lock_bh(rwlock_t *lock) +{ + local_bh_disable(); + preempt_disable(); + rwlock_acquire_read(&lock->dep_map, 0, 0, _RET_IP_); + LOCK_CONTENDED(lock, _raw_read_trylock, _raw_read_lock); +} + +static inline unsigned long __write_lock_irqsave(rwlock_t *lock) +{ + unsigned long flags; + + local_irq_save(flags); + preempt_disable(); + rwlock_acquire(&lock->dep_map, 0, 0, _RET_IP_); + LOCK_CONTENDED_FLAGS(lock, _raw_write_trylock, _raw_write_lock, + _raw_write_lock_flags, &flags); + return flags; +} + +static inline void __write_lock_irq(rwlock_t *lock) +{ + local_irq_disable(); + preempt_disable(); + rwlock_acquire(&lock->dep_map, 0, 0, _RET_IP_); + LOCK_CONTENDED(lock, _raw_write_trylock, _raw_write_lock); +} + +static inline void __write_lock_bh(rwlock_t *lock) +{ + local_bh_disable(); + preempt_disable(); + rwlock_acquire(&lock->dep_map, 0, 0, _RET_IP_); + LOCK_CONTENDED(lock, _raw_write_trylock, _raw_write_lock); +} + +static inline void __spin_lock(spinlock_t *lock) +{ + preempt_disable(); + spin_acquire(&lock->dep_map, 0, 0, _RET_IP_); + LOCK_CONTENDED(lock, _raw_spin_trylock, _raw_spin_lock); +} + +static inline void __write_lock(rwlock_t *lock) +{ + preempt_disable(); + rwlock_acquire(&lock->dep_map, 0, 0, _RET_IP_); + LOCK_CONTENDED(lock, _raw_write_trylock, _raw_write_lock); +} + +#endif /* CONFIG_PREEMPT */ + +static inline void __spin_unlock(spinlock_t *lock) +{ + spin_release(&lock->dep_map, 1, _RET_IP_); + _raw_spin_unlock(lock); + preempt_enable(); +} + +static inline void __write_unlock(rwlock_t *lock) +{ + rwlock_release(&lock->dep_map, 1, _RET_IP_); + _raw_write_unlock(lock); + preempt_enable(); +} + +static inline void __read_unlock(rwlock_t *lock) +{ + rwlock_release(&lock->dep_map, 1, _RET_IP_); + _raw_read_unlock(lock); + preempt_enable(); +} + +static inline void __spin_unlock_irqrestore(spinlock_t *lock, + unsigned long flags) +{ + spin_release(&lock->dep_map, 1, _RET_IP_); + _raw_spin_unlock(lock); + local_irq_restore(flags); + preempt_enable(); +} + +static inline void __spin_unlock_irq(spinlock_t *lock) +{ + spin_release(&lock->dep_map, 1, _RET_IP_); + _raw_spin_unlock(lock); + local_irq_enable(); + preempt_enable(); +} + +static inline void __spin_unlock_bh(spinlock_t *lock) +{ + spin_release(&lock->dep_map, 1, _RET_IP_); + _raw_spin_unlock(lock); + preempt_enable_no_resched(); + local_bh_enable_ip((unsigned long)__builtin_return_address(0)); +} + +static inline void __read_unlock_irqrestore(rwlock_t *lock, unsigned long flags) +{ + rwlock_release(&lock->dep_map, 1, _RET_IP_); + _raw_read_unlock(lock); + local_irq_restore(flags); + preempt_enable(); +} + +static inline void __read_unlock_irq(rwlock_t *lock) +{ + rwlock_release(&lock->dep_map, 1, _RET_IP_); + _raw_read_unlock(lock); + local_irq_enable(); + preempt_enable(); +} + +static inline void __read_unlock_bh(rwlock_t *lock) +{ + rwlock_release(&lock->dep_map, 1, _RET_IP_); + _raw_read_unlock(lock); + preempt_enable_no_resched(); + local_bh_enable_ip((unsigned long)__builtin_return_address(0)); +} + +static inline void __write_unlock_irqrestore(rwlock_t *lock, + unsigned long flags) +{ + rwlock_release(&lock->dep_map, 1, _RET_IP_); + _raw_write_unlock(lock); + local_irq_restore(flags); + preempt_enable(); +} + +static inline void __write_unlock_irq(rwlock_t *lock) +{ + rwlock_release(&lock->dep_map, 1, _RET_IP_); + _raw_write_unlock(lock); + local_irq_enable(); + preempt_enable(); +} + +static inline void __write_unlock_bh(rwlock_t *lock) +{ + rwlock_release(&lock->dep_map, 1, _RET_IP_); + _raw_write_unlock(lock); + preempt_enable_no_resched(); + local_bh_enable_ip((unsigned long)__builtin_return_address(0)); +} + +static inline int __spin_trylock_bh(spinlock_t *lock) +{ + local_bh_disable(); + preempt_disable(); + if (_raw_spin_trylock(lock)) { + spin_acquire(&lock->dep_map, 0, 1, _RET_IP_); + return 1; + } + preempt_enable_no_resched(); + local_bh_enable_ip((unsigned long)__builtin_return_address(0)); + return 0; +} + #endif /* __LINUX_SPINLOCK_API_SMP_H */ diff --git a/kernel/spinlock.c b/kernel/spinlock.c index 7932653c4ebd..2c000f5c070b 100644 --- a/kernel/spinlock.c +++ b/kernel/spinlock.c @@ -23,40 +23,19 @@ int __lockfunc _spin_trylock(spinlock_t *lock) { - preempt_disable(); - if (_raw_spin_trylock(lock)) { - spin_acquire(&lock->dep_map, 0, 1, _RET_IP_); - return 1; - } - - preempt_enable(); - return 0; + return __spin_trylock(lock); } EXPORT_SYMBOL(_spin_trylock); int __lockfunc _read_trylock(rwlock_t *lock) { - preempt_disable(); - if (_raw_read_trylock(lock)) { - rwlock_acquire_read(&lock->dep_map, 0, 1, _RET_IP_); - return 1; - } - - preempt_enable(); - return 0; + return __read_trylock(lock); } EXPORT_SYMBOL(_read_trylock); int __lockfunc _write_trylock(rwlock_t *lock) { - preempt_disable(); - if (_raw_write_trylock(lock)) { - rwlock_acquire(&lock->dep_map, 0, 1, _RET_IP_); - return 1; - } - - preempt_enable(); - return 0; + return __write_trylock(lock); } EXPORT_SYMBOL(_write_trylock); @@ -69,129 +48,74 @@ EXPORT_SYMBOL(_write_trylock); void __lockfunc _read_lock(rwlock_t *lock) { - preempt_disable(); - rwlock_acquire_read(&lock->dep_map, 0, 0, _RET_IP_); - LOCK_CONTENDED(lock, _raw_read_trylock, _raw_read_lock); + __read_lock(lock); } EXPORT_SYMBOL(_read_lock); unsigned long __lockfunc _spin_lock_irqsave(spinlock_t *lock) { - unsigned long flags; - - local_irq_save(flags); - preempt_disable(); - spin_acquire(&lock->dep_map, 0, 0, _RET_IP_); - /* - * On lockdep we dont want the hand-coded irq-enable of - * _raw_spin_lock_flags() code, because lockdep assumes - * that interrupts are not re-enabled during lock-acquire: - */ -#ifdef CONFIG_LOCKDEP - LOCK_CONTENDED(lock, _raw_spin_trylock, _raw_spin_lock); -#else - _raw_spin_lock_flags(lock, &flags); -#endif - return flags; + return __spin_lock_irqsave(lock); } EXPORT_SYMBOL(_spin_lock_irqsave); void __lockfunc _spin_lock_irq(spinlock_t *lock) { - local_irq_disable(); - preempt_disable(); - spin_acquire(&lock->dep_map, 0, 0, _RET_IP_); - LOCK_CONTENDED(lock, _raw_spin_trylock, _raw_spin_lock); + __spin_lock_irq(lock); } EXPORT_SYMBOL(_spin_lock_irq); void __lockfunc _spin_lock_bh(spinlock_t *lock) { - local_bh_disable(); - preempt_disable(); - spin_acquire(&lock->dep_map, 0, 0, _RET_IP_); - LOCK_CONTENDED(lock, _raw_spin_trylock, _raw_spin_lock); + __spin_lock_bh(lock); } EXPORT_SYMBOL(_spin_lock_bh); unsigned long __lockfunc _read_lock_irqsave(rwlock_t *lock) { - unsigned long flags; - - local_irq_save(flags); - preempt_disable(); - rwlock_acquire_read(&lock->dep_map, 0, 0, _RET_IP_); - LOCK_CONTENDED_FLAGS(lock, _raw_read_trylock, _raw_read_lock, - _raw_read_lock_flags, &flags); - return flags; + return __read_lock_irqsave(lock); } EXPORT_SYMBOL(_read_lock_irqsave); void __lockfunc _read_lock_irq(rwlock_t *lock) { - local_irq_disable(); - preempt_disable(); - rwlock_acquire_read(&lock->dep_map, 0, 0, _RET_IP_); - LOCK_CONTENDED(lock, _raw_read_trylock, _raw_read_lock); + __read_lock_irq(lock); } EXPORT_SYMBOL(_read_lock_irq); void __lockfunc _read_lock_bh(rwlock_t *lock) { - local_bh_disable(); - preempt_disable(); - rwlock_acquire_read(&lock->dep_map, 0, 0, _RET_IP_); - LOCK_CONTENDED(lock, _raw_read_trylock, _raw_read_lock); + __read_lock_bh(lock); } EXPORT_SYMBOL(_read_lock_bh); unsigned long __lockfunc _write_lock_irqsave(rwlock_t *lock) { - unsigned long flags; - - local_irq_save(flags); - preempt_disable(); - rwlock_acquire(&lock->dep_map, 0, 0, _RET_IP_); - LOCK_CONTENDED_FLAGS(lock, _raw_write_trylock, _raw_write_lock, - _raw_write_lock_flags, &flags); - return flags; + return __write_lock_irqsave(lock); } EXPORT_SYMBOL(_write_lock_irqsave); void __lockfunc _write_lock_irq(rwlock_t *lock) { - local_irq_disable(); - preempt_disable(); - rwlock_acquire(&lock->dep_map, 0, 0, _RET_IP_); - LOCK_CONTENDED(lock, _raw_write_trylock, _raw_write_lock); + __write_lock_irq(lock); } EXPORT_SYMBOL(_write_lock_irq); void __lockfunc _write_lock_bh(rwlock_t *lock) { - local_bh_disable(); - preempt_disable(); - rwlock_acquire(&lock->dep_map, 0, 0, _RET_IP_); - LOCK_CONTENDED(lock, _raw_write_trylock, _raw_write_lock); + __write_lock_bh(lock); } EXPORT_SYMBOL(_write_lock_bh); void __lockfunc _spin_lock(spinlock_t *lock) { - preempt_disable(); - spin_acquire(&lock->dep_map, 0, 0, _RET_IP_); - LOCK_CONTENDED(lock, _raw_spin_trylock, _raw_spin_lock); + __spin_lock(lock); } - EXPORT_SYMBOL(_spin_lock); void __lockfunc _write_lock(rwlock_t *lock) { - preempt_disable(); - rwlock_acquire(&lock->dep_map, 0, 0, _RET_IP_); - LOCK_CONTENDED(lock, _raw_write_trylock, _raw_write_lock); + __write_lock(lock); } - EXPORT_SYMBOL(_write_lock); #else /* CONFIG_PREEMPT: */ @@ -320,121 +244,79 @@ EXPORT_SYMBOL(_spin_lock_nest_lock); void __lockfunc _spin_unlock(spinlock_t *lock) { - spin_release(&lock->dep_map, 1, _RET_IP_); - _raw_spin_unlock(lock); - preempt_enable(); + __spin_unlock(lock); } EXPORT_SYMBOL(_spin_unlock); void __lockfunc _write_unlock(rwlock_t *lock) { - rwlock_release(&lock->dep_map, 1, _RET_IP_); - _raw_write_unlock(lock); - preempt_enable(); + __write_unlock(lock); } EXPORT_SYMBOL(_write_unlock); void __lockfunc _read_unlock(rwlock_t *lock) { - rwlock_release(&lock->dep_map, 1, _RET_IP_); - _raw_read_unlock(lock); - preempt_enable(); + __read_unlock(lock); } EXPORT_SYMBOL(_read_unlock); void __lockfunc _spin_unlock_irqrestore(spinlock_t *lock, unsigned long flags) { - spin_release(&lock->dep_map, 1, _RET_IP_); - _raw_spin_unlock(lock); - local_irq_restore(flags); - preempt_enable(); + __spin_unlock_irqrestore(lock, flags); } EXPORT_SYMBOL(_spin_unlock_irqrestore); void __lockfunc _spin_unlock_irq(spinlock_t *lock) { - spin_release(&lock->dep_map, 1, _RET_IP_); - _raw_spin_unlock(lock); - local_irq_enable(); - preempt_enable(); + __spin_unlock_irq(lock); } EXPORT_SYMBOL(_spin_unlock_irq); void __lockfunc _spin_unlock_bh(spinlock_t *lock) { - spin_release(&lock->dep_map, 1, _RET_IP_); - _raw_spin_unlock(lock); - preempt_enable_no_resched(); - local_bh_enable_ip((unsigned long)__builtin_return_address(0)); + __spin_unlock_bh(lock); } EXPORT_SYMBOL(_spin_unlock_bh); void __lockfunc _read_unlock_irqrestore(rwlock_t *lock, unsigned long flags) { - rwlock_release(&lock->dep_map, 1, _RET_IP_); - _raw_read_unlock(lock); - local_irq_restore(flags); - preempt_enable(); + __read_unlock_irqrestore(lock, flags); } EXPORT_SYMBOL(_read_unlock_irqrestore); void __lockfunc _read_unlock_irq(rwlock_t *lock) { - rwlock_release(&lock->dep_map, 1, _RET_IP_); - _raw_read_unlock(lock); - local_irq_enable(); - preempt_enable(); + __read_unlock_irq(lock); } EXPORT_SYMBOL(_read_unlock_irq); void __lockfunc _read_unlock_bh(rwlock_t *lock) { - rwlock_release(&lock->dep_map, 1, _RET_IP_); - _raw_read_unlock(lock); - preempt_enable_no_resched(); - local_bh_enable_ip((unsigned long)__builtin_return_address(0)); + __read_unlock_bh(lock); } EXPORT_SYMBOL(_read_unlock_bh); void __lockfunc _write_unlock_irqrestore(rwlock_t *lock, unsigned long flags) { - rwlock_release(&lock->dep_map, 1, _RET_IP_); - _raw_write_unlock(lock); - local_irq_restore(flags); - preempt_enable(); + __write_unlock_irqrestore(lock, flags); } EXPORT_SYMBOL(_write_unlock_irqrestore); void __lockfunc _write_unlock_irq(rwlock_t *lock) { - rwlock_release(&lock->dep_map, 1, _RET_IP_); - _raw_write_unlock(lock); - local_irq_enable(); - preempt_enable(); + __write_unlock_irq(lock); } EXPORT_SYMBOL(_write_unlock_irq); void __lockfunc _write_unlock_bh(rwlock_t *lock) { - rwlock_release(&lock->dep_map, 1, _RET_IP_); - _raw_write_unlock(lock); - preempt_enable_no_resched(); - local_bh_enable_ip((unsigned long)__builtin_return_address(0)); + __write_unlock_bh(lock); } EXPORT_SYMBOL(_write_unlock_bh); int __lockfunc _spin_trylock_bh(spinlock_t *lock) { - local_bh_disable(); - preempt_disable(); - if (_raw_spin_trylock(lock)) { - spin_acquire(&lock->dep_map, 0, 1, _RET_IP_); - return 1; - } - - preempt_enable_no_resched(); - local_bh_enable_ip((unsigned long)__builtin_return_address(0)); - return 0; + return __spin_trylock_bh(lock); } EXPORT_SYMBOL(_spin_trylock_bh); -- cgit v1.2.3 From 892a7c67c12da63fa4b51728bbe5b982356a090a Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Mon, 31 Aug 2009 14:43:37 +0200 Subject: locking: Allow arch-inlined spinlocks This allows an architecture to specify per lock variant if the locking code should be kept out-of-line or inlined. If an architecure wants out-of-line locking code no change is needed. To force inlining of e.g. spin_lock() the line: #define __always_inline__spin_lock needs to be added to arch/<...>/include/asm/spinlock.h If CONFIG_DEBUG_SPINLOCK or CONFIG_GENERIC_LOCKBREAK are defined the per architecture defines are (partly) ignored and still out-of-line spinlock code will be generated. Signed-off-by: Heiko Carstens Acked-by: Peter Zijlstra Cc: Arnd Bergmann Cc: Nick Piggin Cc: Martin Schwidefsky Cc: Horst Hartmann Cc: Christian Ehrhardt Cc: Andrew Morton Cc: Linus Torvalds Cc: David Miller Cc: Benjamin Herrenschmidt Cc: Paul Mackerras Cc: Geert Uytterhoeven Cc: Roman Zippel Cc: LKML-Reference: <20090831124418.375299024@de.ibm.com> Signed-off-by: Ingo Molnar --- include/linux/spinlock_api_smp.h | 119 +++++++++++++++++++++++++++++++++++++++ kernel/spinlock.c | 56 ++++++++++++++++++ 2 files changed, 175 insertions(+) (limited to 'include') diff --git a/include/linux/spinlock_api_smp.h b/include/linux/spinlock_api_smp.h index 6b108f5fb149..1a411e3fab95 100644 --- a/include/linux/spinlock_api_smp.h +++ b/include/linux/spinlock_api_smp.h @@ -60,6 +60,125 @@ void __lockfunc _read_unlock_irqrestore(rwlock_t *lock, unsigned long flags) void __lockfunc _write_unlock_irqrestore(rwlock_t *lock, unsigned long flags) __releases(lock); +#ifndef CONFIG_DEBUG_SPINLOCK +#ifndef CONFIG_GENERIC_LOCKBREAK + +#ifdef __always_inline__spin_lock +#define _spin_lock(lock) __spin_lock(lock) +#endif + +#ifdef __always_inline__read_lock +#define _read_lock(lock) __read_lock(lock) +#endif + +#ifdef __always_inline__write_lock +#define _write_lock(lock) __write_lock(lock) +#endif + +#ifdef __always_inline__spin_lock_bh +#define _spin_lock_bh(lock) __spin_lock_bh(lock) +#endif + +#ifdef __always_inline__read_lock_bh +#define _read_lock_bh(lock) __read_lock_bh(lock) +#endif + +#ifdef __always_inline__write_lock_bh +#define _write_lock_bh(lock) __write_lock_bh(lock) +#endif + +#ifdef __always_inline__spin_lock_irq +#define _spin_lock_irq(lock) __spin_lock_irq(lock) +#endif + +#ifdef __always_inline__read_lock_irq +#define _read_lock_irq(lock) __read_lock_irq(lock) +#endif + +#ifdef __always_inline__write_lock_irq +#define _write_lock_irq(lock) __write_lock_irq(lock) +#endif + +#ifdef __always_inline__spin_lock_irqsave +#define _spin_lock_irqsave(lock) __spin_lock_irqsave(lock) +#endif + +#ifdef __always_inline__read_lock_irqsave +#define _read_lock_irqsave(lock) __read_lock_irqsave(lock) +#endif + +#ifdef __always_inline__write_lock_irqsave +#define _write_lock_irqsave(lock) __write_lock_irqsave(lock) +#endif + +#endif /* !CONFIG_GENERIC_LOCKBREAK */ + +#ifdef __always_inline__spin_trylock +#define _spin_trylock(lock) __spin_trylock(lock) +#endif + +#ifdef __always_inline__read_trylock +#define _read_trylock(lock) __read_trylock(lock) +#endif + +#ifdef __always_inline__write_trylock +#define _write_trylock(lock) __write_trylock(lock) +#endif + +#ifdef __always_inline__spin_trylock_bh +#define _spin_trylock_bh(lock) __spin_trylock_bh(lock) +#endif + +#ifdef __always_inline__spin_unlock +#define _spin_unlock(lock) __spin_unlock(lock) +#endif + +#ifdef __always_inline__read_unlock +#define _read_unlock(lock) __read_unlock(lock) +#endif + +#ifdef __always_inline__write_unlock +#define _write_unlock(lock) __write_unlock(lock) +#endif + +#ifdef __always_inline__spin_unlock_bh +#define _spin_unlock_bh(lock) __spin_unlock_bh(lock) +#endif + +#ifdef __always_inline__read_unlock_bh +#define _read_unlock_bh(lock) __read_unlock_bh(lock) +#endif + +#ifdef __always_inline__write_unlock_bh +#define _write_unlock_bh(lock) __write_unlock_bh(lock) +#endif + +#ifdef __always_inline__spin_unlock_irq +#define _spin_unlock_irq(lock) __spin_unlock_irq(lock) +#endif + +#ifdef __always_inline__read_unlock_irq +#define _read_unlock_irq(lock) __read_unlock_irq(lock) +#endif + +#ifdef __always_inline__write_unlock_irq +#define _write_unlock_irq(lock) __write_unlock_irq(lock) +#endif + +#ifdef __always_inline__spin_unlock_irqrestore +#define _spin_unlock_irqrestore(lock, flags) __spin_unlock_irqrestore(lock, flags) +#endif + +#ifdef __always_inline__read_unlock_irqrestore +#define _read_unlock_irqrestore(lock, flags) __read_unlock_irqrestore(lock, flags) +#endif + +#ifdef __always_inline__write_unlock_irqrestore +#define _write_unlock_irqrestore(lock, flags) __write_unlock_irqrestore(lock, flags) +#endif + +#endif /* CONFIG_DEBUG_SPINLOCK */ + static inline int __spin_trylock(spinlock_t *lock) { preempt_disable(); diff --git a/kernel/spinlock.c b/kernel/spinlock.c index 2c000f5c070b..5ddab730cb2f 100644 --- a/kernel/spinlock.c +++ b/kernel/spinlock.c @@ -21,23 +21,29 @@ #include #include +#ifndef _spin_trylock int __lockfunc _spin_trylock(spinlock_t *lock) { return __spin_trylock(lock); } EXPORT_SYMBOL(_spin_trylock); +#endif +#ifndef _read_trylock int __lockfunc _read_trylock(rwlock_t *lock) { return __read_trylock(lock); } EXPORT_SYMBOL(_read_trylock); +#endif +#ifndef _write_trylock int __lockfunc _write_trylock(rwlock_t *lock) { return __write_trylock(lock); } EXPORT_SYMBOL(_write_trylock); +#endif /* * If lockdep is enabled then we use the non-preemption spin-ops @@ -46,77 +52,101 @@ EXPORT_SYMBOL(_write_trylock); */ #if !defined(CONFIG_GENERIC_LOCKBREAK) || defined(CONFIG_DEBUG_LOCK_ALLOC) +#ifndef _read_lock void __lockfunc _read_lock(rwlock_t *lock) { __read_lock(lock); } EXPORT_SYMBOL(_read_lock); +#endif +#ifndef _spin_lock_irqsave unsigned long __lockfunc _spin_lock_irqsave(spinlock_t *lock) { return __spin_lock_irqsave(lock); } EXPORT_SYMBOL(_spin_lock_irqsave); +#endif +#ifndef _spin_lock_irq void __lockfunc _spin_lock_irq(spinlock_t *lock) { __spin_lock_irq(lock); } EXPORT_SYMBOL(_spin_lock_irq); +#endif +#ifndef _spin_lock_bh void __lockfunc _spin_lock_bh(spinlock_t *lock) { __spin_lock_bh(lock); } EXPORT_SYMBOL(_spin_lock_bh); +#endif +#ifndef _read_lock_irqsave unsigned long __lockfunc _read_lock_irqsave(rwlock_t *lock) { return __read_lock_irqsave(lock); } EXPORT_SYMBOL(_read_lock_irqsave); +#endif +#ifndef _read_lock_irq void __lockfunc _read_lock_irq(rwlock_t *lock) { __read_lock_irq(lock); } EXPORT_SYMBOL(_read_lock_irq); +#endif +#ifndef _read_lock_bh void __lockfunc _read_lock_bh(rwlock_t *lock) { __read_lock_bh(lock); } EXPORT_SYMBOL(_read_lock_bh); +#endif +#ifndef _write_lock_irqsave unsigned long __lockfunc _write_lock_irqsave(rwlock_t *lock) { return __write_lock_irqsave(lock); } EXPORT_SYMBOL(_write_lock_irqsave); +#endif +#ifndef _write_lock_irq void __lockfunc _write_lock_irq(rwlock_t *lock) { __write_lock_irq(lock); } EXPORT_SYMBOL(_write_lock_irq); +#endif +#ifndef _write_lock_bh void __lockfunc _write_lock_bh(rwlock_t *lock) { __write_lock_bh(lock); } EXPORT_SYMBOL(_write_lock_bh); +#endif +#ifndef _spin_lock void __lockfunc _spin_lock(spinlock_t *lock) { __spin_lock(lock); } EXPORT_SYMBOL(_spin_lock); +#endif +#ifndef _write_lock void __lockfunc _write_lock(rwlock_t *lock) { __write_lock(lock); } EXPORT_SYMBOL(_write_lock); +#endif #else /* CONFIG_PREEMPT: */ @@ -242,83 +272,109 @@ EXPORT_SYMBOL(_spin_lock_nest_lock); #endif +#ifndef _spin_unlock void __lockfunc _spin_unlock(spinlock_t *lock) { __spin_unlock(lock); } EXPORT_SYMBOL(_spin_unlock); +#endif +#ifndef _write_unlock void __lockfunc _write_unlock(rwlock_t *lock) { __write_unlock(lock); } EXPORT_SYMBOL(_write_unlock); +#endif +#ifndef _read_unlock void __lockfunc _read_unlock(rwlock_t *lock) { __read_unlock(lock); } EXPORT_SYMBOL(_read_unlock); +#endif +#ifndef _spin_unlock_irqrestore void __lockfunc _spin_unlock_irqrestore(spinlock_t *lock, unsigned long flags) { __spin_unlock_irqrestore(lock, flags); } EXPORT_SYMBOL(_spin_unlock_irqrestore); +#endif +#ifndef _spin_unlock_irq void __lockfunc _spin_unlock_irq(spinlock_t *lock) { __spin_unlock_irq(lock); } EXPORT_SYMBOL(_spin_unlock_irq); +#endif +#ifndef _spin_unlock_bh void __lockfunc _spin_unlock_bh(spinlock_t *lock) { __spin_unlock_bh(lock); } EXPORT_SYMBOL(_spin_unlock_bh); +#endif +#ifndef _read_unlock_irqrestore void __lockfunc _read_unlock_irqrestore(rwlock_t *lock, unsigned long flags) { __read_unlock_irqrestore(lock, flags); } EXPORT_SYMBOL(_read_unlock_irqrestore); +#endif +#ifndef _read_unlock_irq void __lockfunc _read_unlock_irq(rwlock_t *lock) { __read_unlock_irq(lock); } EXPORT_SYMBOL(_read_unlock_irq); +#endif +#ifndef _read_unlock_bh void __lockfunc _read_unlock_bh(rwlock_t *lock) { __read_unlock_bh(lock); } EXPORT_SYMBOL(_read_unlock_bh); +#endif +#ifndef _write_unlock_irqrestore void __lockfunc _write_unlock_irqrestore(rwlock_t *lock, unsigned long flags) { __write_unlock_irqrestore(lock, flags); } EXPORT_SYMBOL(_write_unlock_irqrestore); +#endif +#ifndef _write_unlock_irq void __lockfunc _write_unlock_irq(rwlock_t *lock) { __write_unlock_irq(lock); } EXPORT_SYMBOL(_write_unlock_irq); +#endif +#ifndef _write_unlock_bh void __lockfunc _write_unlock_bh(rwlock_t *lock) { __write_unlock_bh(lock); } EXPORT_SYMBOL(_write_unlock_bh); +#endif +#ifndef _spin_trylock_bh int __lockfunc _spin_trylock_bh(spinlock_t *lock) { return __spin_trylock_bh(lock); } EXPORT_SYMBOL(_spin_trylock_bh); +#endif notrace int in_lock_functions(unsigned long addr) { -- cgit v1.2.3 From bb7bed082500179519c7caf0678ba3bed9752658 Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Mon, 31 Aug 2009 14:43:38 +0200 Subject: locking: Simplify spinlock inlining For !DEBUG_SPINLOCK && !PREEMPT && SMP the spin_unlock() functions were always inlined by using special defines which would call the __raw* functions. The out-of-line variants for these functions would be generated anyway. Use the new per unlock/locking variant mechanism to force inlining of the unlock functions like before. This is not a functional change, we just get rid of one additional way to force inlining. Signed-off-by: Heiko Carstens Acked-by: Peter Zijlstra Cc: Arnd Bergmann Cc: Nick Piggin Cc: Martin Schwidefsky Cc: Horst Hartmann Cc: Christian Ehrhardt Cc: Andrew Morton Cc: Linus Torvalds Cc: David Miller Cc: Benjamin Herrenschmidt Cc: Paul Mackerras Cc: Geert Uytterhoeven Cc: Roman Zippel Cc: LKML-Reference: <20090831124418.848735034@de.ibm.com> Signed-off-by: Ingo Molnar --- include/linux/spinlock.h | 46 ++++++---------------------------------- include/linux/spinlock_api_smp.h | 12 +++++++++++ 2 files changed, 18 insertions(+), 40 deletions(-) (limited to 'include') diff --git a/include/linux/spinlock.h b/include/linux/spinlock.h index da76a06556bd..f0ca7a7a1757 100644 --- a/include/linux/spinlock.h +++ b/include/linux/spinlock.h @@ -259,50 +259,16 @@ static inline void smp_mb__after_lock(void) { smp_mb(); } #define spin_lock_irq(lock) _spin_lock_irq(lock) #define spin_lock_bh(lock) _spin_lock_bh(lock) - #define read_lock_irq(lock) _read_lock_irq(lock) #define read_lock_bh(lock) _read_lock_bh(lock) - #define write_lock_irq(lock) _write_lock_irq(lock) #define write_lock_bh(lock) _write_lock_bh(lock) - -/* - * We inline the unlock functions in the nondebug case: - */ -#if defined(CONFIG_DEBUG_SPINLOCK) || defined(CONFIG_PREEMPT) || \ - !defined(CONFIG_SMP) -# define spin_unlock(lock) _spin_unlock(lock) -# define read_unlock(lock) _read_unlock(lock) -# define write_unlock(lock) _write_unlock(lock) -# define spin_unlock_irq(lock) _spin_unlock_irq(lock) -# define read_unlock_irq(lock) _read_unlock_irq(lock) -# define write_unlock_irq(lock) _write_unlock_irq(lock) -#else -# define spin_unlock(lock) \ - do {__raw_spin_unlock(&(lock)->raw_lock); __release(lock); } while (0) -# define read_unlock(lock) \ - do {__raw_read_unlock(&(lock)->raw_lock); __release(lock); } while (0) -# define write_unlock(lock) \ - do {__raw_write_unlock(&(lock)->raw_lock); __release(lock); } while (0) -# define spin_unlock_irq(lock) \ -do { \ - __raw_spin_unlock(&(lock)->raw_lock); \ - __release(lock); \ - local_irq_enable(); \ -} while (0) -# define read_unlock_irq(lock) \ -do { \ - __raw_read_unlock(&(lock)->raw_lock); \ - __release(lock); \ - local_irq_enable(); \ -} while (0) -# define write_unlock_irq(lock) \ -do { \ - __raw_write_unlock(&(lock)->raw_lock); \ - __release(lock); \ - local_irq_enable(); \ -} while (0) -#endif +#define spin_unlock(lock) _spin_unlock(lock) +#define read_unlock(lock) _read_unlock(lock) +#define write_unlock(lock) _write_unlock(lock) +#define spin_unlock_irq(lock) _spin_unlock_irq(lock) +#define read_unlock_irq(lock) _read_unlock_irq(lock) +#define write_unlock_irq(lock) _write_unlock_irq(lock) #define spin_unlock_irqrestore(lock, flags) \ do { \ diff --git a/include/linux/spinlock_api_smp.h b/include/linux/spinlock_api_smp.h index 1a411e3fab95..7a7e18fc2415 100644 --- a/include/linux/spinlock_api_smp.h +++ b/include/linux/spinlock_api_smp.h @@ -60,6 +60,18 @@ void __lockfunc _read_unlock_irqrestore(rwlock_t *lock, unsigned long flags) void __lockfunc _write_unlock_irqrestore(rwlock_t *lock, unsigned long flags) __releases(lock); +/* + * We inline the unlock functions in the nondebug case: + */ +#if !defined(CONFIG_DEBUG_SPINLOCK) && !defined(CONFIG_PREEMPT) +#define __always_inline__spin_unlock +#define __always_inline__read_unlock +#define __always_inline__write_unlock +#define __always_inline__spin_unlock_irq +#define __always_inline__read_unlock_irq +#define __always_inline__write_unlock_irq +#endif + #ifndef CONFIG_DEBUG_SPINLOCK #ifndef CONFIG_GENERIC_LOCKBREAK -- cgit v1.2.3 From 98a56ab382079f777e261e14512cbd4fb2107af4 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Thu, 17 Sep 2009 08:48:28 -0400 Subject: ext4: Fix spelling typo in the trace format for trace_ext4_da_writepages() Signed-off-by: "Theodore Ts'o" --- include/trace/events/ext4.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/trace/events/ext4.h b/include/trace/events/ext4.h index 8d433c4e3709..15051d2d1219 100644 --- a/include/trace/events/ext4.h +++ b/include/trace/events/ext4.h @@ -243,7 +243,7 @@ TRACE_EVENT(ext4_da_writepages, __entry->range_cyclic = wbc->range_cyclic; ), - TP_printk("dev %s ino %lu nr_t_write %ld pages_skipped %ld range_start %llu range_end %llu nonblocking %d for_kupdate %d for_reclaim %d range_cyclic %d", + TP_printk("dev %s ino %lu nr_to_write %ld pages_skipped %ld range_start %llu range_end %llu nonblocking %d for_kupdate %d for_reclaim %d range_cyclic %d", jbd2_dev_to_name(__entry->dev), __entry->ino, __entry->nr_to_write, __entry->pages_skipped, __entry->range_start, __entry->range_end, __entry->nonblocking, -- cgit v1.2.3 From 2b980dbd77d229eb60588802162c9659726b11f4 Mon Sep 17 00:00:00 2001 From: Paul Moore Date: Fri, 28 Aug 2009 18:12:43 -0400 Subject: lsm: Add hooks to the TUN driver The TUN driver lacks any LSM hooks which makes it difficult for LSM modules, such as SELinux, to enforce access controls on network traffic generated by TUN users; this is particularly problematic for virtualization apps such as QEMU and KVM. This patch adds three new LSM hooks designed to control the creation and attachment of TUN devices, the hooks are: * security_tun_dev_create() Provides access control for the creation of new TUN devices * security_tun_dev_post_create() Provides the ability to create the necessary socket LSM state for newly created TUN devices * security_tun_dev_attach() Provides access control for attaching to existing, persistent TUN devices and the ability to update the TUN device's socket LSM state as necessary Signed-off-by: Paul Moore Acked-by: Eric Paris Acked-by: Serge Hallyn Acked-by: David S. Miller Signed-off-by: James Morris --- drivers/net/tun.c | 22 +++++++++++++++------- include/linux/security.h | 31 +++++++++++++++++++++++++++++++ security/capability.c | 19 +++++++++++++++++++ security/security.c | 18 ++++++++++++++++++ 4 files changed, 83 insertions(+), 7 deletions(-) (limited to 'include') diff --git a/drivers/net/tun.c b/drivers/net/tun.c index 42b6c6319bc2..87214a257d2a 100644 --- a/drivers/net/tun.c +++ b/drivers/net/tun.c @@ -130,17 +130,10 @@ static inline struct tun_sock *tun_sk(struct sock *sk) static int tun_attach(struct tun_struct *tun, struct file *file) { struct tun_file *tfile = file->private_data; - const struct cred *cred = current_cred(); int err; ASSERT_RTNL(); - /* Check permissions */ - if (((tun->owner != -1 && cred->euid != tun->owner) || - (tun->group != -1 && !in_egroup_p(tun->group))) && - !capable(CAP_NET_ADMIN)) - return -EPERM; - netif_tx_lock_bh(tun->dev); err = -EINVAL; @@ -926,6 +919,8 @@ static int tun_set_iff(struct net *net, struct file *file, struct ifreq *ifr) dev = __dev_get_by_name(net, ifr->ifr_name); if (dev) { + const struct cred *cred = current_cred(); + if (ifr->ifr_flags & IFF_TUN_EXCL) return -EBUSY; if ((ifr->ifr_flags & IFF_TUN) && dev->netdev_ops == &tun_netdev_ops) @@ -935,6 +930,14 @@ static int tun_set_iff(struct net *net, struct file *file, struct ifreq *ifr) else return -EINVAL; + if (((tun->owner != -1 && cred->euid != tun->owner) || + (tun->group != -1 && !in_egroup_p(tun->group))) && + !capable(CAP_NET_ADMIN)) + return -EPERM; + err = security_tun_dev_attach(tun->sk); + if (err < 0) + return err; + err = tun_attach(tun, file); if (err < 0) return err; @@ -947,6 +950,9 @@ static int tun_set_iff(struct net *net, struct file *file, struct ifreq *ifr) if (!capable(CAP_NET_ADMIN)) return -EPERM; + err = security_tun_dev_create(); + if (err < 0) + return err; /* Set dev type */ if (ifr->ifr_flags & IFF_TUN) { @@ -989,6 +995,8 @@ static int tun_set_iff(struct net *net, struct file *file, struct ifreq *ifr) tun->sk = sk; container_of(sk, struct tun_sock, sk)->tun = tun; + security_tun_dev_post_create(sk); + tun_net_init(dev); if (strchr(dev->name, '%')) { diff --git a/include/linux/security.h b/include/linux/security.h index a16d6b7c4ebe..40ba39ea68ce 100644 --- a/include/linux/security.h +++ b/include/linux/security.h @@ -998,6 +998,17 @@ static inline void security_free_mnt_opts(struct security_mnt_opts *opts) * Sets the connection's peersid to the secmark on skb. * @req_classify_flow: * Sets the flow's sid to the openreq sid. + * @tun_dev_create: + * Check permissions prior to creating a new TUN device. + * @tun_dev_post_create: + * This hook allows a module to update or allocate a per-socket security + * structure. + * @sk contains the newly created sock structure. + * @tun_dev_attach: + * Check permissions prior to attaching to a persistent TUN device. This + * hook can also be used by the module to update any security state + * associated with the TUN device's sock structure. + * @sk contains the existing sock structure. * * Security hooks for XFRM operations. * @@ -1597,6 +1608,9 @@ struct security_operations { void (*inet_csk_clone) (struct sock *newsk, const struct request_sock *req); void (*inet_conn_established) (struct sock *sk, struct sk_buff *skb); void (*req_classify_flow) (const struct request_sock *req, struct flowi *fl); + int (*tun_dev_create)(void); + void (*tun_dev_post_create)(struct sock *sk); + int (*tun_dev_attach)(struct sock *sk); #endif /* CONFIG_SECURITY_NETWORK */ #ifdef CONFIG_SECURITY_NETWORK_XFRM @@ -2586,6 +2600,9 @@ void security_inet_csk_clone(struct sock *newsk, const struct request_sock *req); void security_inet_conn_established(struct sock *sk, struct sk_buff *skb); +int security_tun_dev_create(void); +void security_tun_dev_post_create(struct sock *sk); +int security_tun_dev_attach(struct sock *sk); #else /* CONFIG_SECURITY_NETWORK */ static inline int security_unix_stream_connect(struct socket *sock, @@ -2736,6 +2753,20 @@ static inline void security_inet_conn_established(struct sock *sk, struct sk_buff *skb) { } + +static inline int security_tun_dev_create(void) +{ + return 0; +} + +static inline void security_tun_dev_post_create(struct sock *sk) +{ +} + +static inline int security_tun_dev_attach(struct sock *sk) +{ + return 0; +} #endif /* CONFIG_SECURITY_NETWORK */ #ifdef CONFIG_SECURITY_NETWORK_XFRM diff --git a/security/capability.c b/security/capability.c index 1b943f54b2ea..06400cf07757 100644 --- a/security/capability.c +++ b/security/capability.c @@ -706,10 +706,26 @@ static void cap_inet_conn_established(struct sock *sk, struct sk_buff *skb) { } + + static void cap_req_classify_flow(const struct request_sock *req, struct flowi *fl) { } + +static int cap_tun_dev_create(void) +{ + return 0; +} + +static void cap_tun_dev_post_create(struct sock *sk) +{ +} + +static int cap_tun_dev_attach(struct sock *sk) +{ + return 0; +} #endif /* CONFIG_SECURITY_NETWORK */ #ifdef CONFIG_SECURITY_NETWORK_XFRM @@ -1026,6 +1042,9 @@ void security_fixup_ops(struct security_operations *ops) set_to_cap_if_null(ops, inet_csk_clone); set_to_cap_if_null(ops, inet_conn_established); set_to_cap_if_null(ops, req_classify_flow); + set_to_cap_if_null(ops, tun_dev_create); + set_to_cap_if_null(ops, tun_dev_post_create); + set_to_cap_if_null(ops, tun_dev_attach); #endif /* CONFIG_SECURITY_NETWORK */ #ifdef CONFIG_SECURITY_NETWORK_XFRM set_to_cap_if_null(ops, xfrm_policy_alloc_security); diff --git a/security/security.c b/security/security.c index 0e993f42ce3d..f88eaf6b14cc 100644 --- a/security/security.c +++ b/security/security.c @@ -1117,6 +1117,24 @@ void security_inet_conn_established(struct sock *sk, security_ops->inet_conn_established(sk, skb); } +int security_tun_dev_create(void) +{ + return security_ops->tun_dev_create(); +} +EXPORT_SYMBOL(security_tun_dev_create); + +void security_tun_dev_post_create(struct sock *sk) +{ + return security_ops->tun_dev_post_create(sk); +} +EXPORT_SYMBOL(security_tun_dev_post_create); + +int security_tun_dev_attach(struct sock *sk) +{ + return security_ops->tun_dev_attach(sk); +} +EXPORT_SYMBOL(security_tun_dev_attach); + #endif /* CONFIG_SECURITY_NETWORK */ #ifdef CONFIG_SECURITY_NETWORK_XFRM -- cgit v1.2.3 From b3a3ca8ca0c3c29abc5b2bfe94bb14f3f4590df9 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Mon, 31 Aug 2009 23:13:11 -0400 Subject: ext4: Add new tracepoint: trace_ext4_da_write_pages() Add a new tracepoint which shows the pages that will be written using write_cache_pages() by ext4_da_writepages(). Signed-off-by: "Theodore Ts'o" --- fs/ext4/ext4.h | 15 +++++++++++++++ fs/ext4/inode.c | 13 +------------ include/trace/events/ext4.h | 34 ++++++++++++++++++++++++++++++++++ 3 files changed, 50 insertions(+), 12 deletions(-) (limited to 'include') diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 41a76e163b99..81014f4ed22d 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -113,6 +113,21 @@ struct ext4_allocation_request { unsigned int flags; }; +/* + * For delayed allocation tracking + */ +struct mpage_da_data { + struct inode *inode; + sector_t b_blocknr; /* start block number of extent */ + size_t b_size; /* size of extent */ + unsigned long b_state; /* state of the extent */ + unsigned long first_page, next_page; /* extent of pages */ + struct writeback_control *wbc; + int io_done; + int pages_written; + int retval; +}; + /* * Special inodes numbers */ diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index ff659e757578..17802a96af9f 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -1875,18 +1875,6 @@ static void ext4_da_page_release_reservation(struct page *page, * Delayed allocation stuff */ -struct mpage_da_data { - struct inode *inode; - sector_t b_blocknr; /* start block number of extent */ - size_t b_size; /* size of extent */ - unsigned long b_state; /* state of the extent */ - unsigned long first_page, next_page; /* extent of pages */ - struct writeback_control *wbc; - int io_done; - int pages_written; - int retval; -}; - /* * mpage_da_submit_io - walks through extent of pages and try to write * them with writepage() call back @@ -2863,6 +2851,7 @@ retry: mpd.io_done = 1; ret = MPAGE_DA_EXTENT_TAIL; } + trace_ext4_da_write_pages(inode, &mpd); wbc->nr_to_write -= mpd.pages_written; ext4_journal_stop(handle); diff --git a/include/trace/events/ext4.h b/include/trace/events/ext4.h index 15051d2d1219..dd43399288ea 100644 --- a/include/trace/events/ext4.h +++ b/include/trace/events/ext4.h @@ -251,6 +251,40 @@ TRACE_EVENT(ext4_da_writepages, __entry->range_cyclic) ); +TRACE_EVENT(ext4_da_write_pages, + TP_PROTO(struct inode *inode, struct mpage_da_data *mpd), + + TP_ARGS(inode, mpd), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + __field( __u64, b_blocknr ) + __field( __u32, b_size ) + __field( __u32, b_state ) + __field( unsigned long, first_page ) + __field( int, io_done ) + __field( int, pages_written ) + ), + + TP_fast_assign( + __entry->dev = inode->i_sb->s_dev; + __entry->ino = inode->i_ino; + __entry->b_blocknr = mpd->b_blocknr; + __entry->b_size = mpd->b_size; + __entry->b_state = mpd->b_state; + __entry->first_page = mpd->first_page; + __entry->io_done = mpd->io_done; + __entry->pages_written = mpd->pages_written; + ), + + TP_printk("dev %s ino %lu b_blocknr %llu b_size %u b_state 0x%04x first_page %lu io_done %d pages_written %d", + jbd2_dev_to_name(__entry->dev), __entry->ino, + __entry->b_blocknr, __entry->b_size, + __entry->b_state, __entry->first_page, + __entry->io_done, __entry->pages_written) +); + TRACE_EVENT(ext4_da_writepages_result, TP_PROTO(struct inode *inode, struct writeback_control *wbc, int ret, int pages_written), -- cgit v1.2.3 From 1a37f184fa7824982a5f434c06981ec46a66cef7 Mon Sep 17 00:00:00 2001 From: Benjamin Herrenschmidt Date: Mon, 31 Aug 2009 13:48:16 +1000 Subject: lmb: Also remove __init from lmb_end_of_RAM() declaration in lmb.h My previous patch (commit 4f8ee2c9cc: "lmb: Remove __init from lmb_end_of_DRAM()") removed __init in lmb.c but missed the fact that it was also marked as such in the .h Signed-off-by: Benjamin Herrenschmidt Signed-off-by: Linus Torvalds --- include/linux/lmb.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/lmb.h b/include/linux/lmb.h index c46c89505dac..2442e3f3d033 100644 --- a/include/linux/lmb.h +++ b/include/linux/lmb.h @@ -51,7 +51,7 @@ extern u64 __init lmb_alloc_base(u64 size, extern u64 __init __lmb_alloc_base(u64 size, u64 align, u64 max_addr); extern u64 __init lmb_phys_mem_size(void); -extern u64 __init lmb_end_of_DRAM(void); +extern u64 lmb_end_of_DRAM(void); extern void __init lmb_enforce_memory_limit(u64 memory_limit); extern int __init lmb_is_reserved(u64 addr); extern int lmb_find(struct lmb_property *res); -- cgit v1.2.3 From dc1f8bf68b311b1537cb65893430b6796118498a Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Mon, 31 Aug 2009 19:50:40 +0000 Subject: netdev: change transmit to limited range type The transmit function should only return one of three possible values, some drivers got confused and returned errno's or other values. This changes the definition so that this can be caught at compile time. Signed-off-by: Stephen Hemminger Signed-off-by: David S. Miller --- include/linux/netdevice.h | 22 +++++++++++++--------- 1 file changed, 13 insertions(+), 9 deletions(-) (limited to 'include') diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index 60d3aac49ed4..376a2e1ac00d 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -79,17 +79,19 @@ struct wireless_dev; #define net_xmit_eval(e) ((e) == NET_XMIT_CN? 0 : (e)) #define net_xmit_errno(e) ((e) != NET_XMIT_CN ? -ENOBUFS : 0) +/* Driver transmit return codes */ +enum netdev_tx { + NETDEV_TX_OK = 0, /* driver took care of packet */ + NETDEV_TX_BUSY, /* driver tx path was busy*/ + NETDEV_TX_LOCKED = -1, /* driver tx lock was already taken */ +}; +typedef enum netdev_tx netdev_tx_t; + #endif #define MAX_ADDR_LEN 32 /* Largest hardware address length */ -/* Driver transmit return codes */ -#define NETDEV_TX_OK 0 /* driver took care of packet */ -#define NETDEV_TX_BUSY 1 /* driver tx path was busy*/ -#define NETDEV_TX_LOCKED -1 /* driver tx lock was already taken */ - #ifdef __KERNEL__ - /* * Compute the worst case header length according to the protocols * used. @@ -507,9 +509,11 @@ struct netdev_queue { * This function is called when network device transistions to the down * state. * - * int (*ndo_start_xmit)(struct sk_buff *skb, struct net_device *dev); + * netdev_tx_t (*ndo_start_xmit)(struct sk_buff *skb, + * struct net_device *dev); * Called when a packet needs to be transmitted. - * Must return NETDEV_TX_OK , NETDEV_TX_BUSY, or NETDEV_TX_LOCKED, + * Must return NETDEV_TX_OK , NETDEV_TX_BUSY. + * (can also return NETDEV_TX_LOCKED iff NETIF_F_LLTX) * Required can not be NULL. * * u16 (*ndo_select_queue)(struct net_device *dev, struct sk_buff *skb); @@ -580,7 +584,7 @@ struct net_device_ops { void (*ndo_uninit)(struct net_device *dev); int (*ndo_open)(struct net_device *dev); int (*ndo_stop)(struct net_device *dev); - int (*ndo_start_xmit) (struct sk_buff *skb, + netdev_tx_t (*ndo_start_xmit) (struct sk_buff *skb, struct net_device *dev); u16 (*ndo_select_queue)(struct net_device *dev, struct sk_buff *skb); -- cgit v1.2.3 From 25a79c41ce0ce88a4288adf278e9b0e00f228383 Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Mon, 31 Aug 2009 19:50:45 +0000 Subject: usbnet: convert to netdev_tx_t Signed-off-by: Stephen Hemminger Signed-off-by: David S. Miller --- drivers/net/usb/catc.c | 3 ++- drivers/net/usb/cdc-phonet.c | 6 +++--- drivers/net/usb/hso.c | 3 ++- drivers/net/usb/kaweth.c | 3 ++- drivers/net/usb/pegasus.c | 3 ++- drivers/net/usb/rtl8150.c | 3 ++- drivers/net/usb/usbnet.c | 8 ++++---- drivers/usb/gadget/u_ether.c | 3 ++- include/linux/usb/usbnet.h | 3 ++- 9 files changed, 21 insertions(+), 14 deletions(-) (limited to 'include') diff --git a/drivers/net/usb/catc.c b/drivers/net/usb/catc.c index 7abdc4abbe07..0ffc0c6d03be 100644 --- a/drivers/net/usb/catc.c +++ b/drivers/net/usb/catc.c @@ -411,7 +411,8 @@ static void catc_tx_done(struct urb *urb) spin_unlock_irqrestore(&catc->tx_lock, flags); } -static int catc_start_xmit(struct sk_buff *skb, struct net_device *netdev) +static netdev_tx_t catc_start_xmit(struct sk_buff *skb, + struct net_device *netdev) { struct catc *catc = netdev_priv(netdev); unsigned long flags; diff --git a/drivers/net/usb/cdc-phonet.c b/drivers/net/usb/cdc-phonet.c index 792af72da8ac..0ca5916ca8df 100644 --- a/drivers/net/usb/cdc-phonet.c +++ b/drivers/net/usb/cdc-phonet.c @@ -55,7 +55,7 @@ static void rx_complete(struct urb *req); /* * Network device callbacks */ -static int usbpn_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t usbpn_xmit(struct sk_buff *skb, struct net_device *dev) { struct usbpn_dev *pnd = netdev_priv(dev); struct urb *req = NULL; @@ -82,12 +82,12 @@ static int usbpn_xmit(struct sk_buff *skb, struct net_device *dev) if (pnd->tx_queue >= dev->tx_queue_len) netif_stop_queue(dev); spin_unlock_irqrestore(&pnd->tx_lock, flags); - return 0; + return NETDEV_TX_OK; drop: dev_kfree_skb(skb); dev->stats.tx_dropped++; - return 0; + return NETDEV_TX_OK; } static void tx_complete(struct urb *req) diff --git a/drivers/net/usb/hso.c b/drivers/net/usb/hso.c index ffe410635735..123f9b84dd29 100644 --- a/drivers/net/usb/hso.c +++ b/drivers/net/usb/hso.c @@ -771,7 +771,8 @@ static void write_bulk_callback(struct urb *urb) } /* called by kernel when we need to transmit a packet */ -static int hso_net_start_xmit(struct sk_buff *skb, struct net_device *net) +static netdev_tx_t hso_net_start_xmit(struct sk_buff *skb, + struct net_device *net) { struct hso_net *odev = netdev_priv(net); int result; diff --git a/drivers/net/usb/kaweth.c b/drivers/net/usb/kaweth.c index 200fe3d525ca..7f397365b437 100644 --- a/drivers/net/usb/kaweth.c +++ b/drivers/net/usb/kaweth.c @@ -803,7 +803,8 @@ static void kaweth_usb_transmit_complete(struct urb *urb) /**************************************************************** * kaweth_start_xmit ****************************************************************/ -static int kaweth_start_xmit(struct sk_buff *skb, struct net_device *net) +static netdev_tx_t kaweth_start_xmit(struct sk_buff *skb, + struct net_device *net) { struct kaweth_device *kaweth = netdev_priv(net); __le16 *private_header; diff --git a/drivers/net/usb/pegasus.c b/drivers/net/usb/pegasus.c index 69d2df95ac86..7b935b846424 100644 --- a/drivers/net/usb/pegasus.c +++ b/drivers/net/usb/pegasus.c @@ -876,7 +876,8 @@ static void pegasus_tx_timeout(struct net_device *net) pegasus->stats.tx_errors++; } -static int pegasus_start_xmit(struct sk_buff *skb, struct net_device *net) +static netdev_tx_t pegasus_start_xmit(struct sk_buff *skb, + struct net_device *net) { pegasus_t *pegasus = netdev_priv(net); int count = ((skb->len + 2) & 0x3f) ? skb->len + 2 : skb->len + 3; diff --git a/drivers/net/usb/rtl8150.c b/drivers/net/usb/rtl8150.c index bac8b77fb25e..d9f84f22fbc7 100644 --- a/drivers/net/usb/rtl8150.c +++ b/drivers/net/usb/rtl8150.c @@ -727,7 +727,8 @@ static void rtl8150_set_multicast(struct net_device *netdev) netif_wake_queue(netdev); } -static int rtl8150_start_xmit(struct sk_buff *skb, struct net_device *netdev) +static netdev_tx_t rtl8150_start_xmit(struct sk_buff *skb, + struct net_device *netdev) { rtl8150_t *dev = netdev_priv(netdev); int count, res; diff --git a/drivers/net/usb/usbnet.c b/drivers/net/usb/usbnet.c index 7d471fca2743..d166e3385c64 100644 --- a/drivers/net/usb/usbnet.c +++ b/drivers/net/usb/usbnet.c @@ -1007,15 +1007,16 @@ EXPORT_SYMBOL_GPL(usbnet_tx_timeout); /*-------------------------------------------------------------------------*/ -int usbnet_start_xmit (struct sk_buff *skb, struct net_device *net) +netdev_tx_t usbnet_start_xmit (struct sk_buff *skb, + struct net_device *net) { struct usbnet *dev = netdev_priv(net); int length; - int retval = NET_XMIT_SUCCESS; struct urb *urb = NULL; struct skb_data *entry; struct driver_info *info = dev->driver_info; unsigned long flags; + int retval; // some devices want funky USB-level framing, for // win32 driver (usually) and/or hardware quirks @@ -1079,7 +1080,6 @@ int usbnet_start_xmit (struct sk_buff *skb, struct net_device *net) if (netif_msg_tx_err (dev)) devdbg (dev, "drop, code %d", retval); drop: - retval = NET_XMIT_SUCCESS; dev->net->stats.tx_dropped++; if (skb) dev_kfree_skb_any (skb); @@ -1088,7 +1088,7 @@ drop: devdbg (dev, "> tx, len %d, type 0x%x", length, skb->protocol); } - return retval; + return NETDEV_TX_OK; } EXPORT_SYMBOL_GPL(usbnet_start_xmit); diff --git a/drivers/usb/gadget/u_ether.c b/drivers/usb/gadget/u_ether.c index aac69b591aeb..dc3ebd1e68ca 100644 --- a/drivers/usb/gadget/u_ether.c +++ b/drivers/usb/gadget/u_ether.c @@ -465,7 +465,8 @@ static inline int is_promisc(u16 cdc_filter) return cdc_filter & USB_CDC_PACKET_TYPE_PROMISCUOUS; } -static int eth_start_xmit(struct sk_buff *skb, struct net_device *net) +static netdev_tx_t eth_start_xmit(struct sk_buff *skb, + struct net_device *net) { struct eth_dev *dev = netdev_priv(net); int length = skb->len; diff --git a/include/linux/usb/usbnet.h b/include/linux/usb/usbnet.h index 09514252d84e..bb69e256cd16 100644 --- a/include/linux/usb/usbnet.h +++ b/include/linux/usb/usbnet.h @@ -182,7 +182,8 @@ struct skb_data { /* skb->cb is one of these */ extern int usbnet_open (struct net_device *net); extern int usbnet_stop (struct net_device *net); -extern int usbnet_start_xmit (struct sk_buff *skb, struct net_device *net); +extern netdev_tx_t usbnet_start_xmit (struct sk_buff *skb, + struct net_device *net); extern void usbnet_tx_timeout (struct net_device *net); extern int usbnet_change_mtu (struct net_device *net, int new_mtu); -- cgit v1.2.3 From 4c5d502d8b2db8947c44dc44bdc67dbe55cce2b9 Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Mon, 31 Aug 2009 19:50:48 +0000 Subject: hdlc: convert to netdev_tx_t Signed-off-by: Stephen Hemminger Signed-off-by: David S. Miller --- drivers/char/pcmcia/synclink_cs.c | 7 +++---- drivers/char/synclink.c | 7 +++---- drivers/char/synclink_gt.c | 7 +++---- drivers/char/synclinkmp.c | 7 +++---- include/linux/hdlc.h | 8 ++++---- 5 files changed, 16 insertions(+), 20 deletions(-) (limited to 'include') diff --git a/drivers/char/pcmcia/synclink_cs.c b/drivers/char/pcmcia/synclink_cs.c index 77b364889224..caf6e4d19469 100644 --- a/drivers/char/pcmcia/synclink_cs.c +++ b/drivers/char/pcmcia/synclink_cs.c @@ -4005,10 +4005,9 @@ static int hdlcdev_attach(struct net_device *dev, unsigned short encoding, * * skb socket buffer containing HDLC frame * dev pointer to network device structure - * - * returns 0 if success, otherwise error code */ -static int hdlcdev_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t hdlcdev_xmit(struct sk_buff *skb, + struct net_device *dev) { MGSLPC_INFO *info = dev_to_port(dev); unsigned long flags; @@ -4043,7 +4042,7 @@ static int hdlcdev_xmit(struct sk_buff *skb, struct net_device *dev) } spin_unlock_irqrestore(&info->lock,flags); - return 0; + return NETDEV_TX_OK; } /** diff --git a/drivers/char/synclink.c b/drivers/char/synclink.c index 813552f14884..4846b73ef28d 100644 --- a/drivers/char/synclink.c +++ b/drivers/char/synclink.c @@ -7697,10 +7697,9 @@ static int hdlcdev_attach(struct net_device *dev, unsigned short encoding, * * skb socket buffer containing HDLC frame * dev pointer to network device structure - * - * returns 0 if success, otherwise error code */ -static int hdlcdev_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t hdlcdev_xmit(struct sk_buff *skb, + struct net_device *dev) { struct mgsl_struct *info = dev_to_port(dev); unsigned long flags; @@ -7731,7 +7730,7 @@ static int hdlcdev_xmit(struct sk_buff *skb, struct net_device *dev) usc_start_transmitter(info); spin_unlock_irqrestore(&info->irq_spinlock,flags); - return 0; + return NETDEV_TX_OK; } /** diff --git a/drivers/char/synclink_gt.c b/drivers/char/synclink_gt.c index 91f20a92fddf..8678f0c8699d 100644 --- a/drivers/char/synclink_gt.c +++ b/drivers/char/synclink_gt.c @@ -1497,10 +1497,9 @@ static int hdlcdev_attach(struct net_device *dev, unsigned short encoding, * * skb socket buffer containing HDLC frame * dev pointer to network device structure - * - * returns 0 if success, otherwise error code */ -static int hdlcdev_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t hdlcdev_xmit(struct sk_buff *skb, + struct net_device *dev) { struct slgt_info *info = dev_to_port(dev); unsigned long flags; @@ -1529,7 +1528,7 @@ static int hdlcdev_xmit(struct sk_buff *skb, struct net_device *dev) update_tx_timer(info); spin_unlock_irqrestore(&info->lock,flags); - return 0; + return NETDEV_TX_OK; } /** diff --git a/drivers/char/synclinkmp.c b/drivers/char/synclinkmp.c index 8d4a2a8a0a70..2b18adc4ee19 100644 --- a/drivers/char/synclinkmp.c +++ b/drivers/char/synclinkmp.c @@ -1608,10 +1608,9 @@ static int hdlcdev_attach(struct net_device *dev, unsigned short encoding, * * skb socket buffer containing HDLC frame * dev pointer to network device structure - * - * returns 0 if success, otherwise error code */ -static int hdlcdev_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t hdlcdev_xmit(struct sk_buff *skb, + struct net_device *dev) { SLMP_INFO *info = dev_to_port(dev); unsigned long flags; @@ -1642,7 +1641,7 @@ static int hdlcdev_xmit(struct sk_buff *skb, struct net_device *dev) tx_start(info); spin_unlock_irqrestore(&info->lock,flags); - return 0; + return NETDEV_TX_OK; } /** diff --git a/include/linux/hdlc.h b/include/linux/hdlc.h index 6a6e701f1631..ee275c8b3df1 100644 --- a/include/linux/hdlc.h +++ b/include/linux/hdlc.h @@ -38,7 +38,7 @@ struct hdlc_proto { int (*ioctl)(struct net_device *dev, struct ifreq *ifr); __be16 (*type_trans)(struct sk_buff *skb, struct net_device *dev); int (*netif_rx)(struct sk_buff *skb); - int (*xmit)(struct sk_buff *skb, struct net_device *dev); + netdev_tx_t (*xmit)(struct sk_buff *skb, struct net_device *dev); struct module *module; struct hdlc_proto *next; /* next protocol in the list */ }; @@ -51,7 +51,7 @@ typedef struct hdlc_device { unsigned short encoding, unsigned short parity); /* hardware driver must handle this instead of dev->hard_start_xmit */ - int (*xmit)(struct sk_buff *skb, struct net_device *dev); + netdev_tx_t (*xmit)(struct sk_buff *skb, struct net_device *dev); /* Things below are for HDLC layer internal use only */ const struct hdlc_proto *proto; @@ -60,7 +60,7 @@ typedef struct hdlc_device { spinlock_t state_lock; void *state; void *priv; -}hdlc_device; +} hdlc_device; @@ -106,7 +106,7 @@ void hdlc_close(struct net_device *dev); /* May be used by hardware driver */ int hdlc_change_mtu(struct net_device *dev, int new_mtu); /* Must be pointed to by hw driver's dev->netdev_ops->ndo_start_xmit */ -int hdlc_start_xmit(struct sk_buff *skb, struct net_device *dev); +netdev_tx_t hdlc_start_xmit(struct sk_buff *skb, struct net_device *dev); int attach_hdlc_protocol(struct net_device *dev, struct hdlc_proto *proto, size_t size); -- cgit v1.2.3 From 61357325f377889a1daffa14962d705dc814dd0e Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Mon, 31 Aug 2009 19:50:58 +0000 Subject: netdev: convert bulk of drivers to netdev_tx_t In a couple of cases collapse some extra code like: int retval = NETDEV_TX_OK; ... return retval; into return NETDEV_TX_OK; Signed-off-by: Stephen Hemminger Signed-off-by: David S. Miller --- drivers/ieee802154/fakehard.c | 3 ++- drivers/net/8139cp.c | 3 ++- drivers/net/8139too.c | 7 ++++--- drivers/net/82596.c | 5 ++--- drivers/net/8390.c | 2 +- drivers/net/8390.h | 4 ++-- drivers/net/8390p.c | 2 +- drivers/net/a2065.c | 3 ++- drivers/net/acenic.c | 3 ++- drivers/net/acenic.h | 3 ++- drivers/net/amd8111e.c | 3 ++- drivers/net/arcnet/arcnet.c | 3 ++- drivers/net/ariadne.c | 6 ++++-- drivers/net/at1700.c | 6 ++++-- drivers/net/atl1c/atl1c_main.c | 3 ++- drivers/net/atl1e/atl1e_main.c | 3 ++- drivers/net/atlx/atl1.c | 3 ++- drivers/net/atlx/atl2.c | 3 ++- drivers/net/atp.c | 6 ++++-- drivers/net/au1000_eth.c | 2 +- drivers/net/b44.c | 2 +- drivers/net/benet/be_main.c | 3 ++- drivers/net/bnx2.c | 2 +- drivers/net/bnx2x_main.c | 2 +- drivers/net/can/sja1000/sja1000.c | 3 ++- drivers/net/cassini.c | 2 +- drivers/net/chelsio/sge.c | 2 +- drivers/net/chelsio/sge.h | 2 +- drivers/net/cs89x0.c | 4 ++-- drivers/net/cxgb3/adapter.h | 2 +- drivers/net/cxgb3/sge.c | 2 +- drivers/net/defxx.c | 10 ++++------ drivers/net/depca.c | 6 ++++-- drivers/net/dl2k.c | 4 ++-- drivers/net/dnet.c | 2 +- drivers/net/eepro.c | 6 ++++-- drivers/net/eexpress.c | 5 +++-- drivers/net/enc28j60.c | 3 ++- drivers/net/enic/enic_main.c | 3 ++- drivers/net/epic100.c | 5 +++-- drivers/net/eth16i.c | 4 ++-- drivers/net/ethoc.c | 2 +- drivers/net/ewrk3.c | 4 ++-- drivers/net/fealnx.c | 4 ++-- drivers/net/forcedeth.c | 5 +++-- drivers/net/hamachi.c | 6 ++++-- drivers/net/hp100.c | 13 ++++++++----- drivers/net/ibmlana.c | 2 +- drivers/net/ibmveth.c | 3 ++- drivers/net/ipg.c | 3 ++- drivers/net/jme.c | 2 +- drivers/net/ks8842.c | 3 ++- drivers/net/ks8851.c | 5 +++-- drivers/net/lance.c | 6 ++++-- drivers/net/lib8390.c | 3 ++- drivers/net/loopback.c | 3 ++- drivers/net/lp486e.c | 4 ++-- drivers/net/mlx4/en_tx.c | 2 +- drivers/net/mlx4/mlx4_en.h | 2 +- drivers/net/myri10ge/myri10ge.c | 11 +++++++---- drivers/net/natsemi.c | 4 ++-- drivers/net/netxen/netxen_nic_main.c | 5 +++-- drivers/net/ni52.c | 5 +++-- drivers/net/ni65.c | 6 ++++-- drivers/net/niu.c | 3 ++- drivers/net/ns83820.c | 3 ++- drivers/net/pcnet32.c | 6 ++++-- drivers/net/qla3xxx.c | 3 ++- drivers/net/qlge/qlge_main.c | 2 +- drivers/net/r6040.c | 10 +++++----- drivers/net/r8169.c | 12 ++++++------ drivers/net/rrunner.c | 3 ++- drivers/net/rrunner.h | 3 ++- drivers/net/s2io.c | 2 +- drivers/net/sb1000.c | 5 +++-- drivers/net/sc92031.c | 3 ++- drivers/net/seeq8005.c | 6 ++++-- drivers/net/sfc/efx.h | 5 +++-- drivers/net/sfc/selftest.c | 3 ++- drivers/net/sfc/tx.c | 18 ++++++++---------- drivers/net/sfc/tx.h | 3 ++- drivers/net/sis190.c | 3 ++- drivers/net/sis900.c | 5 +++-- drivers/net/skfp/skfddi.c | 6 ++++-- drivers/net/skge.c | 3 ++- drivers/net/sky2.c | 3 ++- drivers/net/smc9194.c | 6 ++++-- drivers/net/smsc9420.c | 3 ++- drivers/net/starfire.c | 4 ++-- drivers/net/sundance.c | 4 ++-- drivers/net/sungem.c | 3 ++- drivers/net/sunhme.c | 3 ++- drivers/net/tehuti.c | 3 ++- drivers/net/tg3.c | 9 ++++++--- drivers/net/tlan.c | 4 ++-- drivers/net/typhoon.c | 2 +- drivers/net/via-rhine.c | 6 ++++-- drivers/net/via-velocity.c | 3 ++- drivers/net/vxge/vxge-main.c | 2 +- drivers/net/yellowfin.c | 6 ++++-- drivers/net/znet.c | 5 +++-- include/linux/arcdevice.h | 3 ++- 102 files changed, 253 insertions(+), 173 deletions(-) (limited to 'include') diff --git a/drivers/ieee802154/fakehard.c b/drivers/ieee802154/fakehard.c index c1c9697f9fde..96a2959ce877 100644 --- a/drivers/ieee802154/fakehard.c +++ b/drivers/ieee802154/fakehard.c @@ -257,7 +257,8 @@ static int ieee802154_fake_close(struct net_device *dev) return 0; } -static int ieee802154_fake_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t ieee802154_fake_xmit(struct sk_buff *skb, + struct net_device *dev) { skb->iif = dev->ifindex; skb->dev = dev; diff --git a/drivers/net/8139cp.c b/drivers/net/8139cp.c index 4a8995aaeca3..462d9f59c53a 100644 --- a/drivers/net/8139cp.c +++ b/drivers/net/8139cp.c @@ -736,7 +736,8 @@ static void cp_tx (struct cp_private *cp) netif_wake_queue(cp->dev); } -static int cp_start_xmit (struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t cp_start_xmit (struct sk_buff *skb, + struct net_device *dev) { struct cp_private *cp = netdev_priv(dev); unsigned entry; diff --git a/drivers/net/8139too.c b/drivers/net/8139too.c index b39ec98345ea..4a3628755026 100644 --- a/drivers/net/8139too.c +++ b/drivers/net/8139too.c @@ -628,8 +628,8 @@ static void mdio_write (struct net_device *dev, int phy_id, int location, static void rtl8139_start_thread(struct rtl8139_private *tp); static void rtl8139_tx_timeout (struct net_device *dev); static void rtl8139_init_ring (struct net_device *dev); -static int rtl8139_start_xmit (struct sk_buff *skb, - struct net_device *dev); +static netdev_tx_t rtl8139_start_xmit (struct sk_buff *skb, + struct net_device *dev); #ifdef CONFIG_NET_POLL_CONTROLLER static void rtl8139_poll_controller(struct net_device *dev); #endif @@ -1687,7 +1687,8 @@ static void rtl8139_tx_timeout (struct net_device *dev) } } -static int rtl8139_start_xmit (struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t rtl8139_start_xmit (struct sk_buff *skb, + struct net_device *dev) { struct rtl8139_private *tp = netdev_priv(dev); void __iomem *ioaddr = tp->mmio_addr; diff --git a/drivers/net/82596.c b/drivers/net/82596.c index 996cc9102215..ea6b139b812c 100644 --- a/drivers/net/82596.c +++ b/drivers/net/82596.c @@ -356,7 +356,7 @@ static char init_setup[] = 0x7f /* *multi IA */ }; static int i596_open(struct net_device *dev); -static int i596_start_xmit(struct sk_buff *skb, struct net_device *dev); +static netdev_tx_t i596_start_xmit(struct sk_buff *skb, struct net_device *dev); static irqreturn_t i596_interrupt(int irq, void *dev_id); static int i596_close(struct net_device *dev); static void i596_add_cmd(struct net_device *dev, struct i596_cmd *cmd); @@ -1054,8 +1054,7 @@ static void i596_tx_timeout (struct net_device *dev) netif_wake_queue (dev); } - -static int i596_start_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t i596_start_xmit(struct sk_buff *skb, struct net_device *dev) { struct i596_private *lp = dev->ml_priv; struct tx_cmd *tx_cmd; diff --git a/drivers/net/8390.c b/drivers/net/8390.c index 21153dea8ebe..7c7518be1756 100644 --- a/drivers/net/8390.c +++ b/drivers/net/8390.c @@ -17,7 +17,7 @@ int ei_close(struct net_device *dev) } EXPORT_SYMBOL(ei_close); -int ei_start_xmit(struct sk_buff *skb, struct net_device *dev) +netdev_tx_t ei_start_xmit(struct sk_buff *skb, struct net_device *dev) { return __ei_start_xmit(skb, dev); } diff --git a/drivers/net/8390.h b/drivers/net/8390.h index 3c61d6d2748a..3d9e8fb4fbee 100644 --- a/drivers/net/8390.h +++ b/drivers/net/8390.h @@ -40,7 +40,7 @@ extern int ei_open(struct net_device *dev); extern int ei_close(struct net_device *dev); extern irqreturn_t ei_interrupt(int irq, void *dev_id); extern void ei_tx_timeout(struct net_device *dev); -extern int ei_start_xmit(struct sk_buff *skb, struct net_device *dev); +extern netdev_tx_t ei_start_xmit(struct sk_buff *skb, struct net_device *dev); extern void ei_set_multicast_list(struct net_device *dev); extern struct net_device_stats *ei_get_stats(struct net_device *dev); @@ -58,7 +58,7 @@ extern int eip_open(struct net_device *dev); extern int eip_close(struct net_device *dev); extern irqreturn_t eip_interrupt(int irq, void *dev_id); extern void eip_tx_timeout(struct net_device *dev); -extern int eip_start_xmit(struct sk_buff *skb, struct net_device *dev); +extern netdev_tx_t eip_start_xmit(struct sk_buff *skb, struct net_device *dev); extern void eip_set_multicast_list(struct net_device *dev); extern struct net_device_stats *eip_get_stats(struct net_device *dev); diff --git a/drivers/net/8390p.c b/drivers/net/8390p.c index d225c291fd93..a2a64ea0b691 100644 --- a/drivers/net/8390p.c +++ b/drivers/net/8390p.c @@ -22,7 +22,7 @@ int eip_close(struct net_device *dev) } EXPORT_SYMBOL(eip_close); -int eip_start_xmit(struct sk_buff *skb, struct net_device *dev) +netdev_tx_t eip_start_xmit(struct sk_buff *skb, struct net_device *dev) { return __ei_start_xmit(skb, dev); } diff --git a/drivers/net/a2065.c b/drivers/net/a2065.c index 174a81187a95..b7ec0368d7e8 100644 --- a/drivers/net/a2065.c +++ b/drivers/net/a2065.c @@ -547,7 +547,8 @@ static void lance_tx_timeout(struct net_device *dev) netif_wake_queue(dev); } -static int lance_start_xmit (struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t lance_start_xmit (struct sk_buff *skb, + struct net_device *dev) { struct lance_private *lp = netdev_priv(dev); volatile struct lance_regs *ll = lp->ll; diff --git a/drivers/net/acenic.c b/drivers/net/acenic.c index 08419ee10290..5f0b05c2d71f 100644 --- a/drivers/net/acenic.c +++ b/drivers/net/acenic.c @@ -2464,7 +2464,8 @@ ace_load_tx_bd(struct ace_private *ap, struct tx_desc *desc, u64 addr, } -static int ace_start_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t ace_start_xmit(struct sk_buff *skb, + struct net_device *dev) { struct ace_private *ap = netdev_priv(dev); struct ace_regs __iomem *regs = ap->regs; diff --git a/drivers/net/acenic.h b/drivers/net/acenic.h index c987c9b5a137..17079b927ffa 100644 --- a/drivers/net/acenic.h +++ b/drivers/net/acenic.h @@ -775,7 +775,8 @@ static void ace_load_jumbo_rx_ring(struct ace_private *ap, int nr_bufs); static irqreturn_t ace_interrupt(int irq, void *dev_id); static int ace_load_firmware(struct net_device *dev); static int ace_open(struct net_device *dev); -static int ace_start_xmit(struct sk_buff *skb, struct net_device *dev); +static netdev_tx_t ace_start_xmit(struct sk_buff *skb, + struct net_device *dev); static int ace_close(struct net_device *dev); static void ace_tasklet(unsigned long dev); static void ace_dump_trace(struct ace_private *ap); diff --git a/drivers/net/amd8111e.c b/drivers/net/amd8111e.c index 61ac671f97bf..98b5f462c092 100644 --- a/drivers/net/amd8111e.c +++ b/drivers/net/amd8111e.c @@ -1300,7 +1300,8 @@ static int amd8111e_tx_queue_avail(struct amd8111e_priv* lp ) This function will queue the transmit packets to the descriptors and will trigger the send operation. It also initializes the transmit descriptors with buffer physical address, byte count, ownership to hardware etc. */ -static int amd8111e_start_xmit(struct sk_buff *skb, struct net_device * dev) +static netdev_tx_t amd8111e_start_xmit(struct sk_buff *skb, + struct net_device * dev) { struct amd8111e_priv *lp = netdev_priv(dev); int tx_index; diff --git a/drivers/net/arcnet/arcnet.c b/drivers/net/arcnet/arcnet.c index 7d227cdab9f8..75a572560909 100644 --- a/drivers/net/arcnet/arcnet.c +++ b/drivers/net/arcnet/arcnet.c @@ -591,7 +591,8 @@ static int arcnet_rebuild_header(struct sk_buff *skb) /* Called by the kernel in order to transmit a packet. */ -int arcnet_send_packet(struct sk_buff *skb, struct net_device *dev) +netdev_tx_t arcnet_send_packet(struct sk_buff *skb, + struct net_device *dev) { struct arcnet_local *lp = netdev_priv(dev); struct archdr *pkt; diff --git a/drivers/net/ariadne.c b/drivers/net/ariadne.c index 47d976cc3d7d..c35af3e106b1 100644 --- a/drivers/net/ariadne.c +++ b/drivers/net/ariadne.c @@ -115,7 +115,8 @@ struct lancedata { static int ariadne_open(struct net_device *dev); static void ariadne_init_ring(struct net_device *dev); -static int ariadne_start_xmit(struct sk_buff *skb, struct net_device *dev); +static netdev_tx_t ariadne_start_xmit(struct sk_buff *skb, + struct net_device *dev); static void ariadne_tx_timeout(struct net_device *dev); static int ariadne_rx(struct net_device *dev); static void ariadne_reset(struct net_device *dev); @@ -589,7 +590,8 @@ static void ariadne_tx_timeout(struct net_device *dev) } -static int ariadne_start_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t ariadne_start_xmit(struct sk_buff *skb, + struct net_device *dev) { struct ariadne_private *priv = netdev_priv(dev); volatile struct Am79C960 *lance = (struct Am79C960*)dev->base_addr; diff --git a/drivers/net/at1700.c b/drivers/net/at1700.c index 5349c58d1fae..544d5af6950e 100644 --- a/drivers/net/at1700.c +++ b/drivers/net/at1700.c @@ -159,7 +159,8 @@ struct net_local { static int at1700_probe1(struct net_device *dev, int ioaddr); static int read_eeprom(long ioaddr, int location); static int net_open(struct net_device *dev); -static int net_send_packet(struct sk_buff *skb, struct net_device *dev); +static netdev_tx_t net_send_packet(struct sk_buff *skb, + struct net_device *dev); static irqreturn_t net_interrupt(int irq, void *dev_id); static void net_rx(struct net_device *dev); static int net_close(struct net_device *dev); @@ -595,7 +596,8 @@ static void net_tx_timeout (struct net_device *dev) } -static int net_send_packet (struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t net_send_packet (struct sk_buff *skb, + struct net_device *dev) { struct net_local *lp = netdev_priv(dev); int ioaddr = dev->base_addr; diff --git a/drivers/net/atl1c/atl1c_main.c b/drivers/net/atl1c/atl1c_main.c index 1d601ce7d5b2..bf7cc83e9836 100644 --- a/drivers/net/atl1c/atl1c_main.c +++ b/drivers/net/atl1c/atl1c_main.c @@ -2055,7 +2055,8 @@ static void atl1c_tx_queue(struct atl1c_adapter *adapter, struct sk_buff *skb, AT_WRITE_REG(&adapter->hw, REG_MB_PRIO_PROD_IDX, prod_data); } -static int atl1c_xmit_frame(struct sk_buff *skb, struct net_device *netdev) +static netdev_tx_t atl1c_xmit_frame(struct sk_buff *skb, + struct net_device *netdev) { struct atl1c_adapter *adapter = netdev_priv(netdev); unsigned long flags; diff --git a/drivers/net/atl1e/atl1e_main.c b/drivers/net/atl1e/atl1e_main.c index 4570749e3d3b..bca127e65f95 100644 --- a/drivers/net/atl1e/atl1e_main.c +++ b/drivers/net/atl1e/atl1e_main.c @@ -1839,7 +1839,8 @@ static void atl1e_tx_queue(struct atl1e_adapter *adapter, u16 count, AT_WRITE_REG(&adapter->hw, REG_MB_TPD_PROD_IDX, tx_ring->next_to_use); } -static int atl1e_xmit_frame(struct sk_buff *skb, struct net_device *netdev) +static netdev_tx_t atl1e_xmit_frame(struct sk_buff *skb, + struct net_device *netdev) { struct atl1e_adapter *adapter = netdev_priv(netdev); unsigned long flags; diff --git a/drivers/net/atlx/atl1.c b/drivers/net/atlx/atl1.c index 8bca12f71390..00569dc1313c 100644 --- a/drivers/net/atlx/atl1.c +++ b/drivers/net/atlx/atl1.c @@ -2349,7 +2349,8 @@ static void atl1_tx_queue(struct atl1_adapter *adapter, u16 count, atomic_set(&tpd_ring->next_to_use, next_to_use); } -static int atl1_xmit_frame(struct sk_buff *skb, struct net_device *netdev) +static netdev_tx_t atl1_xmit_frame(struct sk_buff *skb, + struct net_device *netdev) { struct atl1_adapter *adapter = netdev_priv(netdev); struct atl1_tpd_ring *tpd_ring = &adapter->tpd_ring; diff --git a/drivers/net/atlx/atl2.c b/drivers/net/atlx/atl2.c index 204db961029e..d0bcb572d51e 100644 --- a/drivers/net/atlx/atl2.c +++ b/drivers/net/atlx/atl2.c @@ -821,7 +821,8 @@ static inline int TxdFreeBytes(struct atl2_adapter *adapter) (int) (txd_read_ptr - adapter->txd_write_ptr - 1); } -static int atl2_xmit_frame(struct sk_buff *skb, struct net_device *netdev) +static netdev_tx_t atl2_xmit_frame(struct sk_buff *skb, + struct net_device *netdev) { struct atl2_adapter *adapter = netdev_priv(netdev); struct tx_pkt_header *txph; diff --git a/drivers/net/atp.c b/drivers/net/atp.c index 4beacc9c909f..9043294fe617 100644 --- a/drivers/net/atp.c +++ b/drivers/net/atp.c @@ -199,7 +199,8 @@ static int net_open(struct net_device *dev); static void hardware_init(struct net_device *dev); static void write_packet(long ioaddr, int length, unsigned char *packet, int pad, int mode); static void trigger_send(long ioaddr, int length); -static int atp_send_packet(struct sk_buff *skb, struct net_device *dev); +static netdev_tx_t atp_send_packet(struct sk_buff *skb, + struct net_device *dev); static irqreturn_t atp_interrupt(int irq, void *dev_id); static void net_rx(struct net_device *dev); static void read_block(long ioaddr, int length, unsigned char *buffer, int data_mode); @@ -552,7 +553,8 @@ static void tx_timeout(struct net_device *dev) dev->stats.tx_errors++; } -static int atp_send_packet(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t atp_send_packet(struct sk_buff *skb, + struct net_device *dev) { struct net_local *lp = netdev_priv(dev); long ioaddr = dev->base_addr; diff --git a/drivers/net/au1000_eth.c b/drivers/net/au1000_eth.c index 2aab1ebc6cd1..407fd45f56a1 100644 --- a/drivers/net/au1000_eth.c +++ b/drivers/net/au1000_eth.c @@ -937,7 +937,7 @@ static int au1000_close(struct net_device *dev) /* * Au1000 transmit routine. */ -static int au1000_tx(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t au1000_tx(struct sk_buff *skb, struct net_device *dev) { struct au1000_private *aup = netdev_priv(dev); struct net_device_stats *ps = &dev->stats; diff --git a/drivers/net/b44.c b/drivers/net/b44.c index 07c92f34d8d8..bee510177a3f 100644 --- a/drivers/net/b44.c +++ b/drivers/net/b44.c @@ -946,7 +946,7 @@ static void b44_tx_timeout(struct net_device *dev) netif_wake_queue(dev); } -static int b44_start_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t b44_start_xmit(struct sk_buff *skb, struct net_device *dev) { struct b44 *bp = netdev_priv(dev); int rc = NETDEV_TX_OK; diff --git a/drivers/net/benet/be_main.c b/drivers/net/benet/be_main.c index 7b9efee890f8..e19fe1dcd144 100644 --- a/drivers/net/benet/be_main.c +++ b/drivers/net/benet/be_main.c @@ -427,7 +427,8 @@ static int make_tx_wrbs(struct be_adapter *adapter, return copied; } -static int be_xmit(struct sk_buff *skb, struct net_device *netdev) +static netdev_tx_t be_xmit(struct sk_buff *skb, + struct net_device *netdev) { struct be_adapter *adapter = netdev_priv(netdev); struct be_tx_obj *tx_obj = &adapter->tx_obj; diff --git a/drivers/net/bnx2.c b/drivers/net/bnx2.c index c4e85f694272..fcaf3bc8277e 100644 --- a/drivers/net/bnx2.c +++ b/drivers/net/bnx2.c @@ -6283,7 +6283,7 @@ bnx2_vlan_rx_register(struct net_device *dev, struct vlan_group *vlgrp) * bnx2_tx_int() runs without netif_tx_lock unless it needs to call * netif_wake_queue(). */ -static int +static netdev_tx_t bnx2_start_xmit(struct sk_buff *skb, struct net_device *dev) { struct bnx2 *bp = netdev_priv(dev); diff --git a/drivers/net/bnx2x_main.c b/drivers/net/bnx2x_main.c index b318d16a4d91..e2e50267cc64 100644 --- a/drivers/net/bnx2x_main.c +++ b/drivers/net/bnx2x_main.c @@ -10936,7 +10936,7 @@ exit_lbl: * bnx2x_tx_int() runs without netif_tx_lock unless it needs to call * netif_wake_queue() */ -static int bnx2x_start_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t bnx2x_start_xmit(struct sk_buff *skb, struct net_device *dev) { struct bnx2x *bp = netdev_priv(dev); struct bnx2x_fastpath *fp, *fp_stat; diff --git a/drivers/net/can/sja1000/sja1000.c b/drivers/net/can/sja1000/sja1000.c index b3004de1e5e4..9ce3ddac4e9b 100644 --- a/drivers/net/can/sja1000/sja1000.c +++ b/drivers/net/can/sja1000/sja1000.c @@ -238,7 +238,8 @@ static void chipset_init(struct net_device *dev) * xx xx xx xx ff ll 00 11 22 33 44 55 66 77 * [ can-id ] [flags] [len] [can data (up to 8 bytes] */ -static int sja1000_start_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t sja1000_start_xmit(struct sk_buff *skb, + struct net_device *dev) { struct sja1000_priv *priv = netdev_priv(dev); struct net_device_stats *stats = &dev->stats; diff --git a/drivers/net/cassini.c b/drivers/net/cassini.c index 299a33b914f8..7517dc1da650 100644 --- a/drivers/net/cassini.c +++ b/drivers/net/cassini.c @@ -2918,7 +2918,7 @@ static inline int cas_xmit_tx_ringN(struct cas *cp, int ring, return 0; } -static int cas_start_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t cas_start_xmit(struct sk_buff *skb, struct net_device *dev) { struct cas *cp = netdev_priv(dev); diff --git a/drivers/net/chelsio/sge.c b/drivers/net/chelsio/sge.c index 3711d64e45ef..8c658cf6f62f 100644 --- a/drivers/net/chelsio/sge.c +++ b/drivers/net/chelsio/sge.c @@ -1776,7 +1776,7 @@ static inline int eth_hdr_len(const void *data) /* * Adds the CPL header to the sk_buff and passes it to t1_sge_tx. */ -int t1_start_xmit(struct sk_buff *skb, struct net_device *dev) +netdev_tx_t t1_start_xmit(struct sk_buff *skb, struct net_device *dev) { struct adapter *adapter = dev->ml_priv; struct sge *sge = adapter->sge; diff --git a/drivers/net/chelsio/sge.h b/drivers/net/chelsio/sge.h index 8c9405123992..00cc37fc1f6f 100644 --- a/drivers/net/chelsio/sge.h +++ b/drivers/net/chelsio/sge.h @@ -78,7 +78,7 @@ void t1_sge_destroy(struct sge *); irqreturn_t t1_interrupt(int irq, void *cookie); int t1_poll(struct napi_struct *, int); -int t1_start_xmit(struct sk_buff *skb, struct net_device *dev); +netdev_tx_t t1_start_xmit(struct sk_buff *skb, struct net_device *dev); void t1_set_vlan_accel(struct adapter *adapter, int on_off); void t1_sge_start(struct sge *); void t1_sge_stop(struct sge *); diff --git a/drivers/net/cs89x0.c b/drivers/net/cs89x0.c index 839cac5fd8a5..0c54219960e2 100644 --- a/drivers/net/cs89x0.c +++ b/drivers/net/cs89x0.c @@ -246,7 +246,7 @@ struct net_local { static int cs89x0_probe1(struct net_device *dev, int ioaddr, int modular); static int net_open(struct net_device *dev); -static int net_send_packet(struct sk_buff *skb, struct net_device *dev); +static netdev_tx_t net_send_packet(struct sk_buff *skb, struct net_device *dev); static irqreturn_t net_interrupt(int irq, void *dev_id); static void set_multicast_list(struct net_device *dev); static void net_timeout(struct net_device *dev); @@ -1518,7 +1518,7 @@ static void net_timeout(struct net_device *dev) netif_wake_queue(dev); } -static int net_send_packet(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t net_send_packet(struct sk_buff *skb,struct net_device *dev) { struct net_local *lp = netdev_priv(dev); unsigned long flags; diff --git a/drivers/net/cxgb3/adapter.h b/drivers/net/cxgb3/adapter.h index 74723f2e7431..2b1aea6aa558 100644 --- a/drivers/net/cxgb3/adapter.h +++ b/drivers/net/cxgb3/adapter.h @@ -309,7 +309,7 @@ void t3_stop_sge_timers(struct adapter *adap); void t3_free_sge_resources(struct adapter *adap); void t3_sge_err_intr_handler(struct adapter *adapter); irq_handler_t t3_intr_handler(struct adapter *adap, int polling); -int t3_eth_xmit(struct sk_buff *skb, struct net_device *dev); +netdev_tx_t t3_eth_xmit(struct sk_buff *skb, struct net_device *dev); int t3_mgmt_tx(struct adapter *adap, struct sk_buff *skb); void t3_update_qset_coalesce(struct sge_qset *qs, const struct qset_params *p); int t3_sge_alloc_qset(struct adapter *adapter, unsigned int id, int nports, diff --git a/drivers/net/cxgb3/sge.c b/drivers/net/cxgb3/sge.c index 29c79eb43beb..f86612857a73 100644 --- a/drivers/net/cxgb3/sge.c +++ b/drivers/net/cxgb3/sge.c @@ -1216,7 +1216,7 @@ static inline void t3_stop_tx_queue(struct netdev_queue *txq, * * Add a packet to an SGE Tx queue. Runs with softirqs disabled. */ -int t3_eth_xmit(struct sk_buff *skb, struct net_device *dev) +netdev_tx_t t3_eth_xmit(struct sk_buff *skb, struct net_device *dev) { int qidx; unsigned int ndesc, pidx, credits, gen, compl; diff --git a/drivers/net/defxx.c b/drivers/net/defxx.c index b2e0a8fc21d7..6a6ea038d7a3 100644 --- a/drivers/net/defxx.c +++ b/drivers/net/defxx.c @@ -300,7 +300,8 @@ static int dfx_rcv_init(DFX_board_t *bp, int get_buffers); static void dfx_rcv_queue_process(DFX_board_t *bp); static void dfx_rcv_flush(DFX_board_t *bp); -static int dfx_xmt_queue_pkt(struct sk_buff *skb, struct net_device *dev); +static netdev_tx_t dfx_xmt_queue_pkt(struct sk_buff *skb, + struct net_device *dev); static int dfx_xmt_done(DFX_board_t *bp); static void dfx_xmt_flush(DFX_board_t *bp); @@ -3188,11 +3189,8 @@ static void dfx_rcv_queue_process( * None */ -static int dfx_xmt_queue_pkt( - struct sk_buff *skb, - struct net_device *dev - ) - +static netdev_tx_t dfx_xmt_queue_pkt(struct sk_buff *skb, + struct net_device *dev) { DFX_board_t *bp = netdev_priv(dev); u8 prod; /* local transmit producer index */ diff --git a/drivers/net/depca.c b/drivers/net/depca.c index adb997c5bb5d..9686c1fa28f1 100644 --- a/drivers/net/depca.c +++ b/drivers/net/depca.c @@ -516,7 +516,8 @@ struct depca_private { ** Public Functions */ static int depca_open(struct net_device *dev); -static int depca_start_xmit(struct sk_buff *skb, struct net_device *dev); +static netdev_tx_t depca_start_xmit(struct sk_buff *skb, + struct net_device *dev); static irqreturn_t depca_interrupt(int irq, void *dev_id); static int depca_close(struct net_device *dev); static int depca_ioctl(struct net_device *dev, struct ifreq *rq, int cmd); @@ -928,7 +929,8 @@ static void depca_tx_timeout(struct net_device *dev) /* ** Writes a socket buffer to TX descriptor ring and starts transmission */ -static int depca_start_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t depca_start_xmit(struct sk_buff *skb, + struct net_device *dev) { struct depca_private *lp = netdev_priv(dev); u_long ioaddr = dev->base_addr; diff --git a/drivers/net/dl2k.c b/drivers/net/dl2k.c index 4b6a219fecea..7fa7a907f134 100644 --- a/drivers/net/dl2k.c +++ b/drivers/net/dl2k.c @@ -59,7 +59,7 @@ static int rio_open (struct net_device *dev); static void rio_timer (unsigned long data); static void rio_tx_timeout (struct net_device *dev); static void alloc_list (struct net_device *dev); -static int start_xmit (struct sk_buff *skb, struct net_device *dev); +static netdev_tx_t start_xmit (struct sk_buff *skb, struct net_device *dev); static irqreturn_t rio_interrupt (int irq, void *dev_instance); static void rio_free_tx (struct net_device *dev, int irq); static void tx_error (struct net_device *dev, int tx_status); @@ -600,7 +600,7 @@ alloc_list (struct net_device *dev) return; } -static int +static netdev_tx_t start_xmit (struct sk_buff *skb, struct net_device *dev) { struct netdev_private *np = netdev_priv(dev); diff --git a/drivers/net/dnet.c b/drivers/net/dnet.c index 2818d5de3940..234685213f1a 100644 --- a/drivers/net/dnet.c +++ b/drivers/net/dnet.c @@ -541,7 +541,7 @@ static inline void dnet_print_skb(struct sk_buff *skb) #define dnet_print_skb(skb) do {} while (0) #endif -static int dnet_start_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t dnet_start_xmit(struct sk_buff *skb, struct net_device *dev) { struct dnet *bp = netdev_priv(dev); diff --git a/drivers/net/eepro.c b/drivers/net/eepro.c index 53317a83857a..1e934160062c 100644 --- a/drivers/net/eepro.c +++ b/drivers/net/eepro.c @@ -309,7 +309,8 @@ struct eepro_local { static int eepro_probe1(struct net_device *dev, int autoprobe); static int eepro_open(struct net_device *dev); -static int eepro_send_packet(struct sk_buff *skb, struct net_device *dev); +static netdev_tx_t eepro_send_packet(struct sk_buff *skb, + struct net_device *dev); static irqreturn_t eepro_interrupt(int irq, void *dev_id); static void eepro_rx(struct net_device *dev); static void eepro_transmit_interrupt(struct net_device *dev); @@ -1133,7 +1134,8 @@ static void eepro_tx_timeout (struct net_device *dev) } -static int eepro_send_packet(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t eepro_send_packet(struct sk_buff *skb, + struct net_device *dev) { struct eepro_local *lp = netdev_priv(dev); unsigned long flags; diff --git a/drivers/net/eexpress.c b/drivers/net/eexpress.c index d1b6368faacd..592de8f1668a 100644 --- a/drivers/net/eexpress.c +++ b/drivers/net/eexpress.c @@ -246,7 +246,8 @@ static char mca_irqmap[] = { 12, 9, 3, 4, 5, 10, 11, 15 }; static int eexp_open(struct net_device *dev); static int eexp_close(struct net_device *dev); static void eexp_timeout(struct net_device *dev); -static int eexp_xmit(struct sk_buff *buf, struct net_device *dev); +static netdev_tx_t eexp_xmit(struct sk_buff *buf, + struct net_device *dev); static irqreturn_t eexp_irq(int irq, void *dev_addr); static void eexp_set_multicast(struct net_device *dev); @@ -650,7 +651,7 @@ static void eexp_timeout(struct net_device *dev) * Called to transmit a packet, or to allow us to right ourselves * if the kernel thinks we've died. */ -static int eexp_xmit(struct sk_buff *buf, struct net_device *dev) +static netdev_tx_t eexp_xmit(struct sk_buff *buf, struct net_device *dev) { short length = buf->len; #ifdef CONFIG_SMP diff --git a/drivers/net/enc28j60.c b/drivers/net/enc28j60.c index 372d6c6a4e7f..117fc6c12e34 100644 --- a/drivers/net/enc28j60.c +++ b/drivers/net/enc28j60.c @@ -1276,7 +1276,8 @@ static void enc28j60_hw_tx(struct enc28j60_net *priv) locked_reg_bfset(priv, ECON1, ECON1_TXRTS); } -static int enc28j60_send_packet(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t enc28j60_send_packet(struct sk_buff *skb, + struct net_device *dev) { struct enc28j60_net *priv = netdev_priv(dev); diff --git a/drivers/net/enic/enic_main.c b/drivers/net/enic/enic_main.c index 8005b602f776..49912eb2a338 100644 --- a/drivers/net/enic/enic_main.c +++ b/drivers/net/enic/enic_main.c @@ -622,7 +622,8 @@ static inline void enic_queue_wq_skb(struct enic *enic, } /* netif_tx_lock held, process context with BHs disabled, or BH */ -static int enic_hard_start_xmit(struct sk_buff *skb, struct net_device *netdev) +static netdev_tx_t enic_hard_start_xmit(struct sk_buff *skb, + struct net_device *netdev) { struct enic *enic = netdev_priv(netdev); struct vnic_wq *wq = &enic->wq[0]; diff --git a/drivers/net/epic100.c b/drivers/net/epic100.c index d668ff2af6e3..641a10d2e843 100644 --- a/drivers/net/epic100.c +++ b/drivers/net/epic100.c @@ -298,7 +298,8 @@ static void epic_restart(struct net_device *dev); static void epic_timer(unsigned long data); static void epic_tx_timeout(struct net_device *dev); static void epic_init_ring(struct net_device *dev); -static int epic_start_xmit(struct sk_buff *skb, struct net_device *dev); +static netdev_tx_t epic_start_xmit(struct sk_buff *skb, + struct net_device *dev); static int epic_rx(struct net_device *dev, int budget); static int epic_poll(struct napi_struct *napi, int budget); static irqreturn_t epic_interrupt(int irq, void *dev_instance); @@ -961,7 +962,7 @@ static void epic_init_ring(struct net_device *dev) return; } -static int epic_start_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t epic_start_xmit(struct sk_buff *skb, struct net_device *dev) { struct epic_private *ep = netdev_priv(dev); int entry, free_count; diff --git a/drivers/net/eth16i.c b/drivers/net/eth16i.c index 97d5205edc8f..71bfeec33a0b 100644 --- a/drivers/net/eth16i.c +++ b/drivers/net/eth16i.c @@ -405,7 +405,7 @@ static int eth16i_read_eeprom_word(int ioaddr); static void eth16i_eeprom_cmd(int ioaddr, unsigned char command); static int eth16i_open(struct net_device *dev); static int eth16i_close(struct net_device *dev); -static int eth16i_tx(struct sk_buff *skb, struct net_device *dev); +static netdev_tx_t eth16i_tx(struct sk_buff *skb, struct net_device *dev); static void eth16i_rx(struct net_device *dev); static void eth16i_timeout(struct net_device *dev); static irqreturn_t eth16i_interrupt(int irq, void *dev_id); @@ -1053,7 +1053,7 @@ static void eth16i_timeout(struct net_device *dev) netif_wake_queue(dev); } -static int eth16i_tx(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t eth16i_tx(struct sk_buff *skb, struct net_device *dev) { struct eth16i_local *lp = netdev_priv(dev); int ioaddr = dev->base_addr; diff --git a/drivers/net/ethoc.c b/drivers/net/ethoc.c index 4dbe5f173273..b871aefed9c6 100644 --- a/drivers/net/ethoc.c +++ b/drivers/net/ethoc.c @@ -802,7 +802,7 @@ static struct net_device_stats *ethoc_stats(struct net_device *dev) return &priv->stats; } -static int ethoc_start_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t ethoc_start_xmit(struct sk_buff *skb, struct net_device *dev) { struct ethoc *priv = netdev_priv(dev); struct ethoc_bd bd; diff --git a/drivers/net/ewrk3.c b/drivers/net/ewrk3.c index 9c51bc813ad3..b2a5ec8f3721 100644 --- a/drivers/net/ewrk3.c +++ b/drivers/net/ewrk3.c @@ -298,7 +298,7 @@ struct ewrk3_private { ** Public Functions */ static int ewrk3_open(struct net_device *dev); -static int ewrk3_queue_pkt(struct sk_buff *skb, struct net_device *dev); +static netdev_tx_t ewrk3_queue_pkt(struct sk_buff *skb, struct net_device *dev); static irqreturn_t ewrk3_interrupt(int irq, void *dev_id); static int ewrk3_close(struct net_device *dev); static void set_multicast_list(struct net_device *dev); @@ -764,7 +764,7 @@ static void ewrk3_timeout(struct net_device *dev) /* ** Writes a socket buffer to the free page queue */ -static int ewrk3_queue_pkt (struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t ewrk3_queue_pkt(struct sk_buff *skb, struct net_device *dev) { struct ewrk3_private *lp = netdev_priv(dev); u_long iobase = dev->base_addr; diff --git a/drivers/net/fealnx.c b/drivers/net/fealnx.c index f66da84a9398..18d5fbb9673e 100644 --- a/drivers/net/fealnx.c +++ b/drivers/net/fealnx.c @@ -433,7 +433,7 @@ static void netdev_timer(unsigned long data); static void reset_timer(unsigned long data); static void fealnx_tx_timeout(struct net_device *dev); static void init_ring(struct net_device *dev); -static int start_tx(struct sk_buff *skb, struct net_device *dev); +static netdev_tx_t start_tx(struct sk_buff *skb, struct net_device *dev); static irqreturn_t intr_handler(int irq, void *dev_instance); static int netdev_rx(struct net_device *dev); static void set_rx_mode(struct net_device *dev); @@ -1305,7 +1305,7 @@ static void init_ring(struct net_device *dev) } -static int start_tx(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t start_tx(struct sk_buff *skb, struct net_device *dev) { struct netdev_private *np = netdev_priv(dev); unsigned long flags; diff --git a/drivers/net/forcedeth.c b/drivers/net/forcedeth.c index 3b4e0766c7b2..0a1c2bb27d4d 100644 --- a/drivers/net/forcedeth.c +++ b/drivers/net/forcedeth.c @@ -2137,7 +2137,7 @@ static void nv_gear_backoff_reseed(struct net_device *dev) * nv_start_xmit: dev->hard_start_xmit function * Called with netif_tx_lock held. */ -static int nv_start_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t nv_start_xmit(struct sk_buff *skb, struct net_device *dev) { struct fe_priv *np = netdev_priv(dev); u32 tx_flags = 0; @@ -2257,7 +2257,8 @@ static int nv_start_xmit(struct sk_buff *skb, struct net_device *dev) return NETDEV_TX_OK; } -static int nv_start_xmit_optimized(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t nv_start_xmit_optimized(struct sk_buff *skb, + struct net_device *dev) { struct fe_priv *np = netdev_priv(dev); u32 tx_flags = 0; diff --git a/drivers/net/hamachi.c b/drivers/net/hamachi.c index 635341d6a028..1d5064a09aca 100644 --- a/drivers/net/hamachi.c +++ b/drivers/net/hamachi.c @@ -557,7 +557,8 @@ static int netdev_ioctl(struct net_device *dev, struct ifreq *rq, int cmd); static void hamachi_timer(unsigned long data); static void hamachi_tx_timeout(struct net_device *dev); static void hamachi_init_ring(struct net_device *dev); -static int hamachi_start_xmit(struct sk_buff *skb, struct net_device *dev); +static netdev_tx_t hamachi_start_xmit(struct sk_buff *skb, + struct net_device *dev); static irqreturn_t hamachi_interrupt(int irq, void *dev_instance); static int hamachi_rx(struct net_device *dev); static inline int hamachi_tx(struct net_device *dev); @@ -1263,7 +1264,8 @@ do { \ } while (0) #endif -static int hamachi_start_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t hamachi_start_xmit(struct sk_buff *skb, + struct net_device *dev) { struct hamachi_private *hmp = netdev_priv(dev); unsigned entry; diff --git a/drivers/net/hp100.c b/drivers/net/hp100.c index d1b63387e9bc..a9a1a99f02dd 100644 --- a/drivers/net/hp100.c +++ b/drivers/net/hp100.c @@ -240,9 +240,10 @@ static int hp100_probe1(struct net_device *dev, int ioaddr, u_char bus, static int hp100_open(struct net_device *dev); static int hp100_close(struct net_device *dev); -static int hp100_start_xmit(struct sk_buff *skb, struct net_device *dev); -static int hp100_start_xmit_bm(struct sk_buff *skb, - struct net_device *dev); +static netdev_tx_t hp100_start_xmit(struct sk_buff *skb, + struct net_device *dev); +static netdev_tx_t hp100_start_xmit_bm(struct sk_buff *skb, + struct net_device *dev); static void hp100_rx(struct net_device *dev); static struct net_device_stats *hp100_get_stats(struct net_device *dev); static void hp100_misc_interrupt(struct net_device *dev); @@ -1483,7 +1484,8 @@ static int hp100_check_lan(struct net_device *dev) */ /* tx function for busmaster mode */ -static int hp100_start_xmit_bm(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t hp100_start_xmit_bm(struct sk_buff *skb, + struct net_device *dev) { unsigned long flags; int i, ok_flag; @@ -1635,7 +1637,8 @@ static void hp100_clean_txring(struct net_device *dev) } /* tx function for slave modes */ -static int hp100_start_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t hp100_start_xmit(struct sk_buff *skb, + struct net_device *dev) { unsigned long flags; int i, ok_flag; diff --git a/drivers/net/ibmlana.c b/drivers/net/ibmlana.c index 448098d3b39b..090a6d3af112 100644 --- a/drivers/net/ibmlana.c +++ b/drivers/net/ibmlana.c @@ -812,7 +812,7 @@ static int ibmlana_close(struct net_device *dev) /* transmit a block. */ -static int ibmlana_tx(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t ibmlana_tx(struct sk_buff *skb, struct net_device *dev) { ibmlana_priv *priv = netdev_priv(dev); int tmplen, addr; diff --git a/drivers/net/ibmveth.c b/drivers/net/ibmveth.c index 76b295a18185..5862282ab2fe 100644 --- a/drivers/net/ibmveth.c +++ b/drivers/net/ibmveth.c @@ -887,7 +887,8 @@ static int ibmveth_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd) #define page_offset(v) ((unsigned long)(v) & ((1 << 12) - 1)) -static int ibmveth_start_xmit(struct sk_buff *skb, struct net_device *netdev) +static netdev_tx_t ibmveth_start_xmit(struct sk_buff *skb, + struct net_device *netdev) { struct ibmveth_adapter *adapter = netdev_priv(netdev); union ibmveth_buf_desc desc; diff --git a/drivers/net/ipg.c b/drivers/net/ipg.c index 43019461b776..382c5532e6c5 100644 --- a/drivers/net/ipg.c +++ b/drivers/net/ipg.c @@ -1858,7 +1858,8 @@ static int ipg_nic_stop(struct net_device *dev) return 0; } -static int ipg_nic_hard_start_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t ipg_nic_hard_start_xmit(struct sk_buff *skb, + struct net_device *dev) { struct ipg_nic_private *sp = netdev_priv(dev); void __iomem *ioaddr = sp->ioaddr; diff --git a/drivers/net/jme.c b/drivers/net/jme.c index e7068c7cd627..1d2a32544ed2 100644 --- a/drivers/net/jme.c +++ b/drivers/net/jme.c @@ -1931,7 +1931,7 @@ jme_stop_queue_if_full(struct jme_adapter *jme) * This function is already protected by netif_tx_lock() */ -static int +static netdev_tx_t jme_start_xmit(struct sk_buff *skb, struct net_device *netdev) { struct jme_adapter *jme = netdev_priv(netdev); diff --git a/drivers/net/ks8842.c b/drivers/net/ks8842.c index 39b0aea2aab3..6e74aa9eea44 100644 --- a/drivers/net/ks8842.c +++ b/drivers/net/ks8842.c @@ -551,7 +551,8 @@ static int ks8842_close(struct net_device *netdev) return 0; } -static int ks8842_xmit_frame(struct sk_buff *skb, struct net_device *netdev) +static netdev_tx_t ks8842_xmit_frame(struct sk_buff *skb, + struct net_device *netdev) { int ret; struct ks8842_adapter *adapter = netdev_priv(netdev); diff --git a/drivers/net/ks8851.c b/drivers/net/ks8851.c index 9a1dea60c1c4..547ac7c7479c 100644 --- a/drivers/net/ks8851.c +++ b/drivers/net/ks8851.c @@ -868,11 +868,12 @@ static int ks8851_net_stop(struct net_device *dev) * and secondly so we can round up more than one packet to transmit which * means we can try and avoid generating too many transmit done interrupts. */ -static int ks8851_start_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t ks8851_start_xmit(struct sk_buff *skb, + struct net_device *dev) { struct ks8851_net *ks = netdev_priv(dev); unsigned needed = calc_txlen(skb->len); - int ret = NETDEV_TX_OK; + netdev_tx_t ret = NETDEV_TX_OK; if (netif_msg_tx_queued(ks)) ks_dbg(ks, "%s: skb %p, %d@%p\n", __func__, diff --git a/drivers/net/lance.c b/drivers/net/lance.c index 30fd4f5f1d5a..dcda30338b65 100644 --- a/drivers/net/lance.c +++ b/drivers/net/lance.c @@ -300,7 +300,8 @@ static unsigned char lance_need_isa_bounce_buffers = 1; static int lance_open(struct net_device *dev); static void lance_init_ring(struct net_device *dev, gfp_t mode); -static int lance_start_xmit(struct sk_buff *skb, struct net_device *dev); +static netdev_tx_t lance_start_xmit(struct sk_buff *skb, + struct net_device *dev); static int lance_rx(struct net_device *dev); static irqreturn_t lance_interrupt(int irq, void *dev_id); static int lance_close(struct net_device *dev); @@ -949,7 +950,8 @@ static void lance_tx_timeout (struct net_device *dev) } -static int lance_start_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t lance_start_xmit(struct sk_buff *skb, + struct net_device *dev) { struct lance_private *lp = dev->ml_priv; int ioaddr = dev->base_addr; diff --git a/drivers/net/lib8390.c b/drivers/net/lib8390.c index d6be36000c5c..256119882b1e 100644 --- a/drivers/net/lib8390.c +++ b/drivers/net/lib8390.c @@ -299,7 +299,8 @@ static void __ei_tx_timeout(struct net_device *dev) * Sends a packet to an 8390 network device. */ -static int __ei_start_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t __ei_start_xmit(struct sk_buff *skb, + struct net_device *dev) { unsigned long e8390_base = dev->base_addr; struct ei_device *ei_local = (struct ei_device *) netdev_priv(dev); diff --git a/drivers/net/loopback.c b/drivers/net/loopback.c index 51bbce72bede..1bc654a73c47 100644 --- a/drivers/net/loopback.c +++ b/drivers/net/loopback.c @@ -69,7 +69,8 @@ struct pcpu_lstats { * The higher levels take care of making this non-reentrant (it's * called with bh's disabled). */ -static int loopback_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t loopback_xmit(struct sk_buff *skb, + struct net_device *dev) { struct pcpu_lstats *pcpu_lstats, *lb_stats; int len; diff --git a/drivers/net/lp486e.c b/drivers/net/lp486e.c index c292bad411ee..cc3ed9cf28be 100644 --- a/drivers/net/lp486e.c +++ b/drivers/net/lp486e.c @@ -377,7 +377,7 @@ static char init_setup[14] = { }; static int i596_open(struct net_device *dev); -static int i596_start_xmit(struct sk_buff *skb, struct net_device *dev); +static netdev_tx_t i596_start_xmit(struct sk_buff *skb, struct net_device *dev); static irqreturn_t i596_interrupt(int irq, void *dev_id); static int i596_close(struct net_device *dev); static void i596_add_cmd(struct net_device *dev, struct i596_cmd *cmd); @@ -863,7 +863,7 @@ static int i596_open(struct net_device *dev) return 0; /* Always succeed */ } -static int i596_start_xmit (struct sk_buff *skb, struct net_device *dev) { +static netdev_tx_t i596_start_xmit (struct sk_buff *skb, struct net_device *dev) { struct tx_cmd *tx_cmd; short length; diff --git a/drivers/net/mlx4/en_tx.c b/drivers/net/mlx4/en_tx.c index 0ecc1e1b013e..d3d6e991065b 100644 --- a/drivers/net/mlx4/en_tx.c +++ b/drivers/net/mlx4/en_tx.c @@ -588,7 +588,7 @@ u16 mlx4_en_select_queue(struct net_device *dev, struct sk_buff *skb) return skb_tx_hash(dev, skb); } -int mlx4_en_xmit(struct sk_buff *skb, struct net_device *dev) +netdev_tx_t mlx4_en_xmit(struct sk_buff *skb, struct net_device *dev) { struct mlx4_en_priv *priv = netdev_priv(dev); struct mlx4_en_dev *mdev = priv->mdev; diff --git a/drivers/net/mlx4/mlx4_en.h b/drivers/net/mlx4/mlx4_en.h index 4513fb4960dc..4376147b0ea0 100644 --- a/drivers/net/mlx4/mlx4_en.h +++ b/drivers/net/mlx4/mlx4_en.h @@ -518,7 +518,7 @@ int mlx4_en_arm_cq(struct mlx4_en_priv *priv, struct mlx4_en_cq *cq); void mlx4_en_poll_tx_cq(unsigned long data); void mlx4_en_tx_irq(struct mlx4_cq *mcq); u16 mlx4_en_select_queue(struct net_device *dev, struct sk_buff *skb); -int mlx4_en_xmit(struct sk_buff *skb, struct net_device *dev); +netdev_tx_t mlx4_en_xmit(struct sk_buff *skb, struct net_device *dev); int mlx4_en_create_tx_ring(struct mlx4_en_priv *priv, struct mlx4_en_tx_ring *ring, u32 size, u16 stride); diff --git a/drivers/net/myri10ge/myri10ge.c b/drivers/net/myri10ge/myri10ge.c index 75deef35b1e0..6930c87f362e 100644 --- a/drivers/net/myri10ge/myri10ge.c +++ b/drivers/net/myri10ge/myri10ge.c @@ -360,7 +360,8 @@ MODULE_PARM_DESC(myri10ge_dca, "Enable DCA if possible"); #define myri10ge_pio_copy(to,from,size) __iowrite64_copy(to,from,size/8) static void myri10ge_set_multicast_list(struct net_device *dev); -static int myri10ge_sw_tso(struct sk_buff *skb, struct net_device *dev); +static netdev_tx_t myri10ge_sw_tso(struct sk_buff *skb, + struct net_device *dev); static inline void put_be32(__be32 val, __be32 __iomem * p) { @@ -2656,7 +2657,8 @@ myri10ge_submit_req(struct myri10ge_tx_buf *tx, struct mcp_kreq_ether_send *src, * it and try again. */ -static int myri10ge_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t myri10ge_xmit(struct sk_buff *skb, + struct net_device *dev) { struct myri10ge_priv *mgp = netdev_priv(dev); struct myri10ge_slice_state *ss; @@ -2947,12 +2949,13 @@ drop: } -static int myri10ge_sw_tso(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t myri10ge_sw_tso(struct sk_buff *skb, + struct net_device *dev) { struct sk_buff *segs, *curr; struct myri10ge_priv *mgp = netdev_priv(dev); struct myri10ge_slice_state *ss; - int status; + netdev_tx_t status; segs = skb_gso_segment(skb, dev->features & ~NETIF_F_TSO6); if (IS_ERR(segs)) diff --git a/drivers/net/natsemi.c b/drivers/net/natsemi.c index 32db12a27342..bd41351e4e26 100644 --- a/drivers/net/natsemi.c +++ b/drivers/net/natsemi.c @@ -621,7 +621,7 @@ static void drain_ring(struct net_device *dev); static void free_ring(struct net_device *dev); static void reinit_ring(struct net_device *dev); static void init_registers(struct net_device *dev); -static int start_tx(struct sk_buff *skb, struct net_device *dev); +static netdev_tx_t start_tx(struct sk_buff *skb, struct net_device *dev); static irqreturn_t intr_handler(int irq, void *dev_instance); static void netdev_error(struct net_device *dev, int intr_status); static int natsemi_poll(struct napi_struct *napi, int budget); @@ -2079,7 +2079,7 @@ static void reinit_ring(struct net_device *dev) reinit_rx(dev); } -static int start_tx(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t start_tx(struct sk_buff *skb, struct net_device *dev) { struct netdev_private *np = netdev_priv(dev); void __iomem * ioaddr = ns_ioaddr(dev); diff --git a/drivers/net/netxen/netxen_nic_main.c b/drivers/net/netxen/netxen_nic_main.c index fab51d16f5fb..f824a392bf56 100644 --- a/drivers/net/netxen/netxen_nic_main.c +++ b/drivers/net/netxen/netxen_nic_main.c @@ -63,7 +63,8 @@ static int __devinit netxen_nic_probe(struct pci_dev *pdev, static void __devexit netxen_nic_remove(struct pci_dev *pdev); static int netxen_nic_open(struct net_device *netdev); static int netxen_nic_close(struct net_device *netdev); -static int netxen_nic_xmit_frame(struct sk_buff *, struct net_device *); +static netdev_tx_t netxen_nic_xmit_frame(struct sk_buff *, + struct net_device *); static void netxen_tx_timeout(struct net_device *netdev); static void netxen_reset_task(struct work_struct *work); static void netxen_watchdog(unsigned long); @@ -1599,7 +1600,7 @@ netxen_clear_cmddesc(u64 *desc) desc[2] = 0ULL; } -static int +static netdev_tx_t netxen_nic_xmit_frame(struct sk_buff *skb, struct net_device *netdev) { struct netxen_adapter *adapter = netdev_priv(netdev); diff --git a/drivers/net/ni52.c b/drivers/net/ni52.c index a0ac5d4f27d3..bd0ac690d12c 100644 --- a/drivers/net/ni52.c +++ b/drivers/net/ni52.c @@ -170,7 +170,7 @@ static int ni52_probe1(struct net_device *dev, int ioaddr); static irqreturn_t ni52_interrupt(int irq, void *dev_id); static int ni52_open(struct net_device *dev); static int ni52_close(struct net_device *dev); -static int ni52_send_packet(struct sk_buff *, struct net_device *); +static netdev_tx_t ni52_send_packet(struct sk_buff *, struct net_device *); static struct net_device_stats *ni52_get_stats(struct net_device *dev); static void set_multicast_list(struct net_device *dev); static void ni52_timeout(struct net_device *dev); @@ -1173,7 +1173,8 @@ static void ni52_timeout(struct net_device *dev) * send frame */ -static int ni52_send_packet(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t ni52_send_packet(struct sk_buff *skb, + struct net_device *dev) { int len, i; #ifndef NO_NOPCOMMANDS diff --git a/drivers/net/ni65.c b/drivers/net/ni65.c index 81a061785898..752c2e4d9cf4 100644 --- a/drivers/net/ni65.c +++ b/drivers/net/ni65.c @@ -252,7 +252,8 @@ static void ni65_xmit_intr(struct net_device *dev,int); static int ni65_open(struct net_device *dev); static int ni65_lance_reinit(struct net_device *dev); static void ni65_init_lance(struct priv *p,unsigned char*,int,int); -static int ni65_send_packet(struct sk_buff *skb, struct net_device *dev); +static netdev_tx_t ni65_send_packet(struct sk_buff *skb, + struct net_device *dev); static void ni65_timeout(struct net_device *dev); static int ni65_close(struct net_device *dev); static int ni65_alloc_buffer(struct net_device *dev); @@ -1157,7 +1158,8 @@ static void ni65_timeout(struct net_device *dev) * Send a packet */ -static int ni65_send_packet(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t ni65_send_packet(struct sk_buff *skb, + struct net_device *dev) { struct priv *p = dev->ml_priv; diff --git a/drivers/net/niu.c b/drivers/net/niu.c index 3ada7ea2abca..119fd4e04141 100644 --- a/drivers/net/niu.c +++ b/drivers/net/niu.c @@ -6657,7 +6657,8 @@ static u64 niu_compute_tx_flags(struct sk_buff *skb, struct ethhdr *ehdr, return ret; } -static int niu_start_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t niu_start_xmit(struct sk_buff *skb, + struct net_device *dev) { struct niu *np = netdev_priv(dev); unsigned long align, headroom; diff --git a/drivers/net/ns83820.c b/drivers/net/ns83820.c index 1576ac07216e..c594e1946476 100644 --- a/drivers/net/ns83820.c +++ b/drivers/net/ns83820.c @@ -1077,7 +1077,8 @@ static void ns83820_cleanup_tx(struct ns83820 *dev) * while trying to track down a bug in either the zero copy code or * the tx fifo (hence the MAX_FRAG_LEN). */ -static int ns83820_hard_start_xmit(struct sk_buff *skb, struct net_device *ndev) +static netdev_tx_t ns83820_hard_start_xmit(struct sk_buff *skb, + struct net_device *ndev) { struct ns83820 *dev = PRIV(ndev); u32 free_idx, cmdsts, extsts; diff --git a/drivers/net/pcnet32.c b/drivers/net/pcnet32.c index 6859442f1862..6d28b18e7e28 100644 --- a/drivers/net/pcnet32.c +++ b/drivers/net/pcnet32.c @@ -303,7 +303,8 @@ static int pcnet32_probe_pci(struct pci_dev *, const struct pci_device_id *); static int pcnet32_probe1(unsigned long, int, struct pci_dev *); static int pcnet32_open(struct net_device *); static int pcnet32_init_ring(struct net_device *); -static int pcnet32_start_xmit(struct sk_buff *, struct net_device *); +static netdev_tx_t pcnet32_start_xmit(struct sk_buff *, + struct net_device *); static void pcnet32_tx_timeout(struct net_device *dev); static irqreturn_t pcnet32_interrupt(int, void *); static int pcnet32_close(struct net_device *); @@ -2481,7 +2482,8 @@ static void pcnet32_tx_timeout(struct net_device *dev) spin_unlock_irqrestore(&lp->lock, flags); } -static int pcnet32_start_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t pcnet32_start_xmit(struct sk_buff *skb, + struct net_device *dev) { struct pcnet32_private *lp = netdev_priv(dev); unsigned long ioaddr = dev->base_addr; diff --git a/drivers/net/qla3xxx.c b/drivers/net/qla3xxx.c index 3e4b67aaa6ea..4c610511eb40 100644 --- a/drivers/net/qla3xxx.c +++ b/drivers/net/qla3xxx.c @@ -2572,7 +2572,8 @@ map_error: * The IOCB is always the top of the chain followed by one or more * OALs (when necessary). */ -static int ql3xxx_send(struct sk_buff *skb, struct net_device *ndev) +static netdev_tx_t ql3xxx_send(struct sk_buff *skb, + struct net_device *ndev) { struct ql3_adapter *qdev = (struct ql3_adapter *)netdev_priv(ndev); struct ql3xxx_port_registers __iomem *port_regs = qdev->mem_map_registers; diff --git a/drivers/net/qlge/qlge_main.c b/drivers/net/qlge/qlge_main.c index 8dd266befdc7..220529257828 100644 --- a/drivers/net/qlge/qlge_main.c +++ b/drivers/net/qlge/qlge_main.c @@ -2103,7 +2103,7 @@ static void ql_hw_csum_setup(struct sk_buff *skb, iph->daddr, len, iph->protocol, 0); } -static int qlge_send(struct sk_buff *skb, struct net_device *ndev) +static netdev_tx_t qlge_send(struct sk_buff *skb, struct net_device *ndev) { struct tx_ring_desc *tx_ring_desc; struct ob_mac_iocb_req *mac_iocb_ptr; diff --git a/drivers/net/r6040.c b/drivers/net/r6040.c index 8068a07eb2b3..7dfcb58b0eb4 100644 --- a/drivers/net/r6040.c +++ b/drivers/net/r6040.c @@ -883,13 +883,13 @@ static int r6040_open(struct net_device *dev) return 0; } -static int r6040_start_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t r6040_start_xmit(struct sk_buff *skb, + struct net_device *dev) { struct r6040_private *lp = netdev_priv(dev); struct r6040_descriptor *descptr; void __iomem *ioaddr = lp->base; unsigned long flags; - int ret = NETDEV_TX_OK; /* Critical Section */ spin_lock_irqsave(&lp->lock, flags); @@ -899,8 +899,7 @@ static int r6040_start_xmit(struct sk_buff *skb, struct net_device *dev) spin_unlock_irqrestore(&lp->lock, flags); netif_stop_queue(dev); printk(KERN_ERR DRV_NAME ": no tx descriptor\n"); - ret = NETDEV_TX_BUSY; - return ret; + return NETDEV_TX_BUSY; } /* Statistic Counter */ @@ -928,7 +927,8 @@ static int r6040_start_xmit(struct sk_buff *skb, struct net_device *dev) dev->trans_start = jiffies; spin_unlock_irqrestore(&lp->lock, flags); - return ret; + + return NETDEV_TX_OK; } static void r6040_multicast_list(struct net_device *dev) diff --git a/drivers/net/r8169.c b/drivers/net/r8169.c index 8cd85309c675..ec0092affd5d 100644 --- a/drivers/net/r8169.c +++ b/drivers/net/r8169.c @@ -507,7 +507,8 @@ MODULE_LICENSE("GPL"); MODULE_VERSION(RTL8169_VERSION); static int rtl8169_open(struct net_device *dev); -static int rtl8169_start_xmit(struct sk_buff *skb, struct net_device *dev); +static netdev_tx_t rtl8169_start_xmit(struct sk_buff *skb, + struct net_device *dev); static irqreturn_t rtl8169_interrupt(int irq, void *dev_instance); static int rtl8169_init_ring(struct net_device *dev); static void rtl_hw_start(struct net_device *dev); @@ -3357,7 +3358,8 @@ static inline u32 rtl8169_tso_csum(struct sk_buff *skb, struct net_device *dev) return 0; } -static int rtl8169_start_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t rtl8169_start_xmit(struct sk_buff *skb, + struct net_device *dev) { struct rtl8169_private *tp = netdev_priv(dev); unsigned int frags, entry = tp->cur_tx % NUM_TX_DESC; @@ -3366,7 +3368,6 @@ static int rtl8169_start_xmit(struct sk_buff *skb, struct net_device *dev) dma_addr_t mapping; u32 status, len; u32 opts1; - int ret = NETDEV_TX_OK; if (unlikely(TX_BUFFS_AVAIL(tp) < skb_shinfo(skb)->nr_frags)) { if (netif_msg_drv(tp)) { @@ -3418,13 +3419,12 @@ static int rtl8169_start_xmit(struct sk_buff *skb, struct net_device *dev) } out: - return ret; + return NETDEV_TX_OK; err_stop: netif_stop_queue(dev); - ret = NETDEV_TX_BUSY; dev->stats.tx_dropped++; - goto out; + return NETDEV_TX_BUSY; } static void rtl8169_pcierr_interrupt(struct net_device *dev) diff --git a/drivers/net/rrunner.c b/drivers/net/rrunner.c index d95534655911..20a71749154a 100644 --- a/drivers/net/rrunner.c +++ b/drivers/net/rrunner.c @@ -1401,7 +1401,8 @@ static int rr_close(struct net_device *dev) } -static int rr_start_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t rr_start_xmit(struct sk_buff *skb, + struct net_device *dev) { struct rr_private *rrpriv = netdev_priv(dev); struct rr_regs __iomem *regs = rrpriv->regs; diff --git a/drivers/net/rrunner.h b/drivers/net/rrunner.h index 6173f11218df..28169043ae49 100644 --- a/drivers/net/rrunner.h +++ b/drivers/net/rrunner.h @@ -831,7 +831,8 @@ static int rr_init1(struct net_device *dev); static irqreturn_t rr_interrupt(int irq, void *dev_id); static int rr_open(struct net_device *dev); -static int rr_start_xmit(struct sk_buff *skb, struct net_device *dev); +static netdev_tx_t rr_start_xmit(struct sk_buff *skb, + struct net_device *dev); static int rr_close(struct net_device *dev); static int rr_ioctl(struct net_device *dev, struct ifreq *rq, int cmd); static unsigned int rr_read_eeprom(struct rr_private *rrpriv, diff --git a/drivers/net/s2io.c b/drivers/net/s2io.c index 3138df5773ee..ddccf5fa56b6 100644 --- a/drivers/net/s2io.c +++ b/drivers/net/s2io.c @@ -4072,7 +4072,7 @@ static int s2io_close(struct net_device *dev) * 0 on success & 1 on failure. */ -static int s2io_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t s2io_xmit(struct sk_buff *skb, struct net_device *dev) { struct s2io_nic *sp = netdev_priv(dev); u16 frg_cnt, frg_len, i, queue, queue_len, put_off, get_off; diff --git a/drivers/net/sb1000.c b/drivers/net/sb1000.c index 6a81aec645d9..ee366c5a8fa3 100644 --- a/drivers/net/sb1000.c +++ b/drivers/net/sb1000.c @@ -82,7 +82,8 @@ struct sb1000_private { extern int sb1000_probe(struct net_device *dev); static int sb1000_open(struct net_device *dev); static int sb1000_dev_ioctl (struct net_device *dev, struct ifreq *ifr, int cmd); -static int sb1000_start_xmit(struct sk_buff *skb, struct net_device *dev); +static netdev_tx_t sb1000_start_xmit(struct sk_buff *skb, + struct net_device *dev); static irqreturn_t sb1000_interrupt(int irq, void *dev_id); static int sb1000_close(struct net_device *dev); @@ -1080,7 +1081,7 @@ static int sb1000_dev_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd) } /* transmit function: do nothing since SB1000 can't send anything out */ -static int +static netdev_tx_t sb1000_start_xmit(struct sk_buff *skb, struct net_device *dev) { printk(KERN_WARNING "%s: trying to transmit!!!\n", dev->name); diff --git a/drivers/net/sc92031.c b/drivers/net/sc92031.c index e3156c97bb58..8d6030022d14 100644 --- a/drivers/net/sc92031.c +++ b/drivers/net/sc92031.c @@ -941,7 +941,8 @@ static struct net_device_stats *sc92031_get_stats(struct net_device *dev) return &dev->stats; } -static int sc92031_start_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t sc92031_start_xmit(struct sk_buff *skb, + struct net_device *dev) { struct sc92031_priv *priv = netdev_priv(dev); void __iomem *port_base = priv->port_base; diff --git a/drivers/net/seeq8005.c b/drivers/net/seeq8005.c index 7cc8bb814137..39246d457ac2 100644 --- a/drivers/net/seeq8005.c +++ b/drivers/net/seeq8005.c @@ -81,7 +81,8 @@ struct net_local { static int seeq8005_probe1(struct net_device *dev, int ioaddr); static int seeq8005_open(struct net_device *dev); static void seeq8005_timeout(struct net_device *dev); -static int seeq8005_send_packet(struct sk_buff *skb, struct net_device *dev); +static netdev_tx_t seeq8005_send_packet(struct sk_buff *skb, + struct net_device *dev); static irqreturn_t seeq8005_interrupt(int irq, void *dev_id); static void seeq8005_rx(struct net_device *dev); static int seeq8005_close(struct net_device *dev); @@ -394,7 +395,8 @@ static void seeq8005_timeout(struct net_device *dev) netif_wake_queue(dev); } -static int seeq8005_send_packet(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t seeq8005_send_packet(struct sk_buff *skb, + struct net_device *dev) { short length = skb->len; unsigned char *buf; diff --git a/drivers/net/sfc/efx.h b/drivers/net/sfc/efx.h index da157aa74b83..aecaf62f4929 100644 --- a/drivers/net/sfc/efx.h +++ b/drivers/net/sfc/efx.h @@ -20,8 +20,9 @@ #define FALCON_B_P_DEVID 0x0710 /* TX */ -extern int efx_xmit(struct efx_nic *efx, - struct efx_tx_queue *tx_queue, struct sk_buff *skb); +extern netdev_tx_t efx_xmit(struct efx_nic *efx, + struct efx_tx_queue *tx_queue, + struct sk_buff *skb); extern void efx_stop_queue(struct efx_nic *efx); extern void efx_wake_queue(struct efx_nic *efx); diff --git a/drivers/net/sfc/selftest.c b/drivers/net/sfc/selftest.c index b67ccca3fc1a..817c7efc11e0 100644 --- a/drivers/net/sfc/selftest.c +++ b/drivers/net/sfc/selftest.c @@ -400,7 +400,8 @@ static int efx_begin_loopback(struct efx_tx_queue *tx_queue) struct efx_loopback_state *state = efx->loopback_selftest; struct efx_loopback_payload *payload; struct sk_buff *skb; - int i, rc; + int i; + netdev_tx_t rc; /* Transmit N copies of buffer */ for (i = 0; i < state->packet_count; i++) { diff --git a/drivers/net/sfc/tx.c b/drivers/net/sfc/tx.c index 14a14788566c..489c4de31447 100644 --- a/drivers/net/sfc/tx.c +++ b/drivers/net/sfc/tx.c @@ -138,8 +138,8 @@ static void efx_tsoh_free(struct efx_tx_queue *tx_queue, * Returns NETDEV_TX_OK or NETDEV_TX_BUSY * You must hold netif_tx_lock() to call this function. */ -static int efx_enqueue_skb(struct efx_tx_queue *tx_queue, - struct sk_buff *skb) +static netdev_tx_t efx_enqueue_skb(struct efx_tx_queue *tx_queue, + struct sk_buff *skb) { struct efx_nic *efx = tx_queue->efx; struct pci_dev *pci_dev = efx->pci_dev; @@ -152,7 +152,7 @@ static int efx_enqueue_skb(struct efx_tx_queue *tx_queue, unsigned int dma_len; bool unmap_single; int q_space, i = 0; - int rc = NETDEV_TX_OK; + netdev_tx_t rc = NETDEV_TX_OK; EFX_BUG_ON_PARANOID(tx_queue->write_count != tx_queue->insert_count); @@ -353,14 +353,11 @@ static void efx_dequeue_buffers(struct efx_tx_queue *tx_queue, * * Context: netif_tx_lock held */ -inline int efx_xmit(struct efx_nic *efx, - struct efx_tx_queue *tx_queue, struct sk_buff *skb) +inline netdev_tx_t efx_xmit(struct efx_nic *efx, + struct efx_tx_queue *tx_queue, struct sk_buff *skb) { - int rc; - /* Map fragments for DMA and add to TX queue */ - rc = efx_enqueue_skb(tx_queue, skb); - return rc; + return efx_enqueue_skb(tx_queue, skb); } /* Initiate a packet transmission. We use one channel per CPU @@ -372,7 +369,8 @@ inline int efx_xmit(struct efx_nic *efx, * Note that returning anything other than NETDEV_TX_OK will cause the * OS to free the skb. */ -int efx_hard_start_xmit(struct sk_buff *skb, struct net_device *net_dev) +netdev_tx_t efx_hard_start_xmit(struct sk_buff *skb, + struct net_device *net_dev) { struct efx_nic *efx = netdev_priv(net_dev); struct efx_tx_queue *tx_queue; diff --git a/drivers/net/sfc/tx.h b/drivers/net/sfc/tx.h index 5e1cc234e42f..e3678962a5b4 100644 --- a/drivers/net/sfc/tx.h +++ b/drivers/net/sfc/tx.h @@ -18,7 +18,8 @@ void efx_remove_tx_queue(struct efx_tx_queue *tx_queue); void efx_init_tx_queue(struct efx_tx_queue *tx_queue); void efx_fini_tx_queue(struct efx_tx_queue *tx_queue); -int efx_hard_start_xmit(struct sk_buff *skb, struct net_device *net_dev); +netdev_tx_t efx_hard_start_xmit(struct sk_buff *skb, + struct net_device *net_dev); void efx_release_tx_buffers(struct efx_tx_queue *tx_queue); #endif /* EFX_TX_H */ diff --git a/drivers/net/sis190.c b/drivers/net/sis190.c index 1f040e8a000b..7cc9898f4e00 100644 --- a/drivers/net/sis190.c +++ b/drivers/net/sis190.c @@ -1168,7 +1168,8 @@ static int sis190_close(struct net_device *dev) return 0; } -static int sis190_start_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t sis190_start_xmit(struct sk_buff *skb, + struct net_device *dev) { struct sis190_private *tp = netdev_priv(dev); void __iomem *ioaddr = tp->mmio_addr; diff --git a/drivers/net/sis900.c b/drivers/net/sis900.c index 61ceeaaf104d..d8827323507a 100644 --- a/drivers/net/sis900.c +++ b/drivers/net/sis900.c @@ -214,7 +214,8 @@ static void sis900_check_mode (struct net_device *net_dev, struct mii_phy *mii_p static void sis900_tx_timeout(struct net_device *net_dev); static void sis900_init_tx_ring(struct net_device *net_dev); static void sis900_init_rx_ring(struct net_device *net_dev); -static int sis900_start_xmit(struct sk_buff *skb, struct net_device *net_dev); +static netdev_tx_t sis900_start_xmit(struct sk_buff *skb, + struct net_device *net_dev); static int sis900_rx(struct net_device *net_dev); static void sis900_finish_xmit (struct net_device *net_dev); static irqreturn_t sis900_interrupt(int irq, void *dev_instance); @@ -1571,7 +1572,7 @@ static void sis900_tx_timeout(struct net_device *net_dev) * tell upper layer if the buffer is full */ -static int +static netdev_tx_t sis900_start_xmit(struct sk_buff *skb, struct net_device *net_dev) { struct sis900_private *sis_priv = netdev_priv(net_dev); diff --git a/drivers/net/skfp/skfddi.c b/drivers/net/skfp/skfddi.c index 888a14a045ef..38a508b4aad9 100644 --- a/drivers/net/skfp/skfddi.c +++ b/drivers/net/skfp/skfddi.c @@ -107,7 +107,8 @@ static void skfp_ctl_set_multicast_list(struct net_device *dev); static void skfp_ctl_set_multicast_list_wo_lock(struct net_device *dev); static int skfp_ctl_set_mac_address(struct net_device *dev, void *addr); static int skfp_ioctl(struct net_device *dev, struct ifreq *rq, int cmd); -static int skfp_send_pkt(struct sk_buff *skb, struct net_device *dev); +static netdev_tx_t skfp_send_pkt(struct sk_buff *skb, + struct net_device *dev); static void send_queued_packets(struct s_smc *smc); static void CheckSourceAddress(unsigned char *frame, unsigned char *hw_addr); static void ResetAdapter(struct s_smc *smc); @@ -1056,7 +1057,8 @@ static int skfp_ioctl(struct net_device *dev, struct ifreq *rq, int cmd) * Side Effects: * None */ -static int skfp_send_pkt(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t skfp_send_pkt(struct sk_buff *skb, + struct net_device *dev) { struct s_smc *smc = netdev_priv(dev); skfddi_priv *bp = &smc->os; diff --git a/drivers/net/skge.c b/drivers/net/skge.c index 543af2044f40..1a1e68549f5c 100644 --- a/drivers/net/skge.c +++ b/drivers/net/skge.c @@ -2746,7 +2746,8 @@ static inline int skge_avail(const struct skge_ring *ring) + (ring->to_clean - ring->to_use) - 1; } -static int skge_xmit_frame(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t skge_xmit_frame(struct sk_buff *skb, + struct net_device *dev) { struct skge_port *skge = netdev_priv(dev); struct skge_hw *hw = skge->hw; diff --git a/drivers/net/sky2.c b/drivers/net/sky2.c index dd630cf18b4d..c7c0a5b7b53a 100644 --- a/drivers/net/sky2.c +++ b/drivers/net/sky2.c @@ -1574,7 +1574,8 @@ static void sky2_tx_unmap(struct pci_dev *pdev, * the number of ring elements will probably be less than the number * of list elements used. */ -static int sky2_xmit_frame(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t sky2_xmit_frame(struct sk_buff *skb, + struct net_device *dev) { struct sky2_port *sky2 = netdev_priv(dev); struct sky2_hw *hw = sky2->hw; diff --git a/drivers/net/smc9194.c b/drivers/net/smc9194.c index 0a1b6f401087..934a12012829 100644 --- a/drivers/net/smc9194.c +++ b/drivers/net/smc9194.c @@ -299,7 +299,8 @@ static void smc_hardware_send_packet( struct net_device * dev ); . to store the packet, I call this routine, which either sends it . now, or generates an interrupt when the card is ready for the . packet */ -static int smc_wait_to_send_packet( struct sk_buff * skb, struct net_device *dev ); +static netdev_tx_t smc_wait_to_send_packet( struct sk_buff * skb, + struct net_device *dev ); /* this does a soft reset on the device */ static void smc_reset( int ioaddr ); @@ -487,7 +488,8 @@ static void smc_setmulticast( int ioaddr, int count, struct dev_mc_list * addrs . o (NO): Enable interrupts and let the interrupt handler deal with it. . o (YES):Send it now. */ -static int smc_wait_to_send_packet( struct sk_buff * skb, struct net_device * dev ) +static netdev_tx_t smc_wait_to_send_packet(struct sk_buff *skb, + struct net_device *dev) { struct smc_local *lp = netdev_priv(dev); unsigned int ioaddr = dev->base_addr; diff --git a/drivers/net/smsc9420.c b/drivers/net/smsc9420.c index 60abdb1081ad..514311d67b36 100644 --- a/drivers/net/smsc9420.c +++ b/drivers/net/smsc9420.c @@ -968,7 +968,8 @@ static void smsc9420_complete_tx(struct net_device *dev) } } -static int smsc9420_hard_start_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t smsc9420_hard_start_xmit(struct sk_buff *skb, + struct net_device *dev) { struct smsc9420_pdata *pd = netdev_priv(dev); dma_addr_t mapping; diff --git a/drivers/net/starfire.c b/drivers/net/starfire.c index 5e7645ee8ab0..a36e2b51e88c 100644 --- a/drivers/net/starfire.c +++ b/drivers/net/starfire.c @@ -595,7 +595,7 @@ static int netdev_open(struct net_device *dev); static void check_duplex(struct net_device *dev); static void tx_timeout(struct net_device *dev); static void init_ring(struct net_device *dev); -static int start_tx(struct sk_buff *skb, struct net_device *dev); +static netdev_tx_t start_tx(struct sk_buff *skb, struct net_device *dev); static irqreturn_t intr_handler(int irq, void *dev_instance); static void netdev_error(struct net_device *dev, int intr_status); static int __netdev_rx(struct net_device *dev, int *quota); @@ -1223,7 +1223,7 @@ static void init_ring(struct net_device *dev) } -static int start_tx(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t start_tx(struct sk_buff *skb, struct net_device *dev) { struct netdev_private *np = netdev_priv(dev); unsigned int entry; diff --git a/drivers/net/sundance.c b/drivers/net/sundance.c index d09be481bcc4..e13685a570f4 100644 --- a/drivers/net/sundance.c +++ b/drivers/net/sundance.c @@ -415,7 +415,7 @@ static void check_duplex(struct net_device *dev); static void netdev_timer(unsigned long data); static void tx_timeout(struct net_device *dev); static void init_ring(struct net_device *dev); -static int start_tx(struct sk_buff *skb, struct net_device *dev); +static netdev_tx_t start_tx(struct sk_buff *skb, struct net_device *dev); static int reset_tx (struct net_device *dev); static irqreturn_t intr_handler(int irq, void *dev_instance); static void rx_poll(unsigned long data); @@ -1053,7 +1053,7 @@ static void tx_poll (unsigned long data) return; } -static int +static netdev_tx_t start_tx (struct sk_buff *skb, struct net_device *dev) { struct netdev_private *np = netdev_priv(dev); diff --git a/drivers/net/sungem.c b/drivers/net/sungem.c index d2dfe0ab5106..e0dfdd246470 100644 --- a/drivers/net/sungem.c +++ b/drivers/net/sungem.c @@ -1015,7 +1015,8 @@ static __inline__ int gem_intme(int entry) return 0; } -static int gem_start_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t gem_start_xmit(struct sk_buff *skb, + struct net_device *dev) { struct gem *gp = netdev_priv(dev); int entry; diff --git a/drivers/net/sunhme.c b/drivers/net/sunhme.c index 008bd59fc64b..37d721bbdb35 100644 --- a/drivers/net/sunhme.c +++ b/drivers/net/sunhme.c @@ -2252,7 +2252,8 @@ static void happy_meal_tx_timeout(struct net_device *dev) netif_wake_queue(dev); } -static int happy_meal_start_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t happy_meal_start_xmit(struct sk_buff *skb, + struct net_device *dev) { struct happy_meal *hp = netdev_priv(dev); int entry; diff --git a/drivers/net/tehuti.c b/drivers/net/tehuti.c index 3c2679cd196b..918d4c9e49b3 100644 --- a/drivers/net/tehuti.c +++ b/drivers/net/tehuti.c @@ -1622,7 +1622,8 @@ static inline int bdx_tx_space(struct bdx_priv *priv) * the driver. Note: the driver must NOT put the skb in its DMA ring. * o NETDEV_TX_LOCKED Locking failed, please retry quickly. */ -static int bdx_tx_transmit(struct sk_buff *skb, struct net_device *ndev) +static netdev_tx_t bdx_tx_transmit(struct sk_buff *skb, + struct net_device *ndev) { struct bdx_priv *priv = netdev_priv(ndev); struct txd_fifo *f = &priv->txd_fifo0; diff --git a/drivers/net/tg3.c b/drivers/net/tg3.c index a7d14aae258a..9d5c1786c664 100644 --- a/drivers/net/tg3.c +++ b/drivers/net/tg3.c @@ -5135,7 +5135,8 @@ static void tg3_set_txd(struct tg3_napi *tnapi, int entry, /* hard_start_xmit for devices that don't have any bugs and * support TG3_FLG2_HW_TSO_2 only. */ -static int tg3_start_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t tg3_start_xmit(struct sk_buff *skb, + struct net_device *dev) { struct tg3 *tp = netdev_priv(dev); u32 len, entry, base_flags, mss; @@ -5251,7 +5252,8 @@ out_unlock: return NETDEV_TX_OK; } -static int tg3_start_xmit_dma_bug(struct sk_buff *, struct net_device *); +static netdev_tx_t tg3_start_xmit_dma_bug(struct sk_buff *, + struct net_device *); /* Use GSO to workaround a rare TSO bug that may be triggered when the * TSO header is greater than 80 bytes. @@ -5290,7 +5292,8 @@ tg3_tso_bug_end: /* hard_start_xmit for devices that have the 4G bug and/or 40-bit bug and * support TG3_FLG2_HW_TSO_1 or firmware TSO only. */ -static int tg3_start_xmit_dma_bug(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t tg3_start_xmit_dma_bug(struct sk_buff *skb, + struct net_device *dev) { struct tg3 *tp = netdev_priv(dev); u32 len, entry, base_flags, mss; diff --git a/drivers/net/tlan.c b/drivers/net/tlan.c index 70c9ec45d8fb..49e273bceed6 100644 --- a/drivers/net/tlan.c +++ b/drivers/net/tlan.c @@ -289,7 +289,7 @@ static void TLan_EisaProbe( void ); static void TLan_Eisa_Cleanup( void ); static int TLan_Init( struct net_device * ); static int TLan_Open( struct net_device *dev ); -static int TLan_StartTx( struct sk_buff *, struct net_device *); +static netdev_tx_t TLan_StartTx( struct sk_buff *, struct net_device *); static irqreturn_t TLan_HandleInterrupt( int, void *); static int TLan_Close( struct net_device *); static struct net_device_stats *TLan_GetStats( struct net_device *); @@ -1083,7 +1083,7 @@ static void TLan_tx_timeout_work(struct work_struct *work) * **************************************************************/ -static int TLan_StartTx( struct sk_buff *skb, struct net_device *dev ) +static netdev_tx_t TLan_StartTx( struct sk_buff *skb, struct net_device *dev ) { TLanPrivateInfo *priv = netdev_priv(dev); dma_addr_t tail_list_phys; diff --git a/drivers/net/typhoon.c b/drivers/net/typhoon.c index 2c26b4577e8a..d6d345229fe9 100644 --- a/drivers/net/typhoon.c +++ b/drivers/net/typhoon.c @@ -762,7 +762,7 @@ typhoon_tso_fill(struct sk_buff *skb, struct transmit_ring *txRing, tcpd->status = 0; } -static int +static netdev_tx_t typhoon_start_tx(struct sk_buff *skb, struct net_device *dev) { struct typhoon *tp = netdev_priv(dev); diff --git a/drivers/net/via-rhine.c b/drivers/net/via-rhine.c index 46eb618bbc90..081402cb05fd 100644 --- a/drivers/net/via-rhine.c +++ b/drivers/net/via-rhine.c @@ -408,7 +408,8 @@ static int mdio_read(struct net_device *dev, int phy_id, int location); static void mdio_write(struct net_device *dev, int phy_id, int location, int value); static int rhine_open(struct net_device *dev); static void rhine_tx_timeout(struct net_device *dev); -static int rhine_start_tx(struct sk_buff *skb, struct net_device *dev); +static netdev_tx_t rhine_start_tx(struct sk_buff *skb, + struct net_device *dev); static irqreturn_t rhine_interrupt(int irq, void *dev_instance); static void rhine_tx(struct net_device *dev); static int rhine_rx(struct net_device *dev, int limit); @@ -1213,7 +1214,8 @@ static void rhine_tx_timeout(struct net_device *dev) netif_wake_queue(dev); } -static int rhine_start_tx(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t rhine_start_tx(struct sk_buff *skb, + struct net_device *dev) { struct rhine_private *rp = netdev_priv(dev); void __iomem *ioaddr = rp->base; diff --git a/drivers/net/via-velocity.c b/drivers/net/via-velocity.c index 47be41a39d35..e56cf6b548d6 100644 --- a/drivers/net/via-velocity.c +++ b/drivers/net/via-velocity.c @@ -2465,7 +2465,8 @@ static int velocity_close(struct net_device *dev) * Called by the networ layer to request a packet is queued to * the velocity. Returns zero on success. */ -static int velocity_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t velocity_xmit(struct sk_buff *skb, + struct net_device *dev) { struct velocity_info *vptr = netdev_priv(dev); int qnum = 0; diff --git a/drivers/net/vxge/vxge-main.c b/drivers/net/vxge/vxge-main.c index 094d15548a2b..41dccba50c46 100644 --- a/drivers/net/vxge/vxge-main.c +++ b/drivers/net/vxge/vxge-main.c @@ -812,7 +812,7 @@ static int vxge_learn_mac(struct vxgedev *vdev, u8 *mac_header) * NOTE: when device cant queue the pkt, just the trans_start variable will * not be upadted. */ -static int +static netdev_tx_t vxge_xmit(struct sk_buff *skb, struct net_device *dev) { struct vxge_fifo *fifo = NULL; diff --git a/drivers/net/yellowfin.c b/drivers/net/yellowfin.c index 76764237cde6..9509477f61f4 100644 --- a/drivers/net/yellowfin.c +++ b/drivers/net/yellowfin.c @@ -347,7 +347,8 @@ static int yellowfin_open(struct net_device *dev); static void yellowfin_timer(unsigned long data); static void yellowfin_tx_timeout(struct net_device *dev); static void yellowfin_init_ring(struct net_device *dev); -static int yellowfin_start_xmit(struct sk_buff *skb, struct net_device *dev); +static netdev_tx_t yellowfin_start_xmit(struct sk_buff *skb, + struct net_device *dev); static irqreturn_t yellowfin_interrupt(int irq, void *dev_instance); static int yellowfin_rx(struct net_device *dev); static void yellowfin_error(struct net_device *dev, int intr_status); @@ -808,7 +809,8 @@ static void yellowfin_init_ring(struct net_device *dev) return; } -static int yellowfin_start_xmit(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t yellowfin_start_xmit(struct sk_buff *skb, + struct net_device *dev) { struct yellowfin_private *yp = netdev_priv(dev); unsigned entry; diff --git a/drivers/net/znet.c b/drivers/net/znet.c index 7f9e14131a5c..a0384b6f09b6 100644 --- a/drivers/net/znet.c +++ b/drivers/net/znet.c @@ -156,7 +156,8 @@ struct netidblk { }; static int znet_open(struct net_device *dev); -static int znet_send_packet(struct sk_buff *skb, struct net_device *dev); +static netdev_tx_t znet_send_packet(struct sk_buff *skb, + struct net_device *dev); static irqreturn_t znet_interrupt(int irq, void *dev_id); static void znet_rx(struct net_device *dev); static int znet_close(struct net_device *dev); @@ -534,7 +535,7 @@ static void znet_tx_timeout (struct net_device *dev) netif_wake_queue (dev); } -static int znet_send_packet(struct sk_buff *skb, struct net_device *dev) +static netdev_tx_t znet_send_packet(struct sk_buff *skb, struct net_device *dev) { int ioaddr = dev->base_addr; struct znet_private *znet = netdev_priv(dev); diff --git a/include/linux/arcdevice.h b/include/linux/arcdevice.h index cd4bcb6989ce..7d650a0e3d8f 100644 --- a/include/linux/arcdevice.h +++ b/include/linux/arcdevice.h @@ -337,7 +337,8 @@ struct net_device *alloc_arcdev(const char *name); int arcnet_open(struct net_device *dev); int arcnet_close(struct net_device *dev); -int arcnet_send_packet(struct sk_buff *skb, struct net_device *dev); +netdev_tx_t arcnet_send_packet(struct sk_buff *skb, + struct net_device *dev); void arcnet_timeout(struct net_device *dev); #endif /* __KERNEL__ */ -- cgit v1.2.3 From cb45439977d3602b91dd0aca2d94fa3b32aebba6 Mon Sep 17 00:00:00 2001 From: Yi Zou Date: Mon, 31 Aug 2009 12:31:36 +0000 Subject: net: Add ndo_fcoe_enable/ndo_fcoe_disable to net_device_ops Add ndo_fcoe_enable/_disable to net_device_ops so the corresponding HW can initialize itself for FCoE traffic or clean up after FCoE traffic is done. This is expected to be called by the kernel FCoE stack upon receiving a request for creating an FCoE instance on the corresponding netdev interface. When implemented by the actual HW, the HW driver check the op code to perform corresponding initialization or clean up for FCoE. The initialization normally includes allocating extra queues for FCoE, setting corresponding HW registers for FCoE, indicating FCoE offload features via netdev, etc. The clean-up would include releasing the resources allocated for FCoE. Signed-off-by: Yi Zou Signed-off-by: Jeff Kirsher Signed-off-by: David S. Miller --- include/linux/netdevice.h | 2 ++ 1 file changed, 2 insertions(+) (limited to 'include') diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index 376a2e1ac00d..121cbad0aae5 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -627,6 +627,8 @@ struct net_device_ops { void (*ndo_poll_controller)(struct net_device *dev); #endif #if defined(CONFIG_FCOE) || defined(CONFIG_FCOE_MODULE) + int (*ndo_fcoe_enable)(struct net_device *dev); + int (*ndo_fcoe_disable)(struct net_device *dev); int (*ndo_fcoe_ddp_setup)(struct net_device *dev, u16 xid, struct scatterlist *sgl, -- cgit v1.2.3 From 0f6f290259896afdca30e1ff4a28aff8edd79a14 Mon Sep 17 00:00:00 2001 From: Yi Zou Date: Mon, 31 Aug 2009 12:32:34 +0000 Subject: dcbnl: Add support for setapp/getapp commands to dcbnl This patch adds dcbnl command definitions to support setapp/getapp functionality from the IEEE 802.1Qaz Data Center Bridging Capability Exchange protocol (DCBX) specification. Section 3.3 defines the application protocol and its 802.1p user priority in DCBX, which is implemented here as a pair of setapp/getapp commands in the kernel dcbnl for setting and retrieving the user priority for an given application protocol. The protocol is identified by the combination of an id and an idtype. Currently, when idtype is 0, the corresponding id gives the ether type of this protocol, e.g., for FCoE, it will be 0x8906; when idtype is 1, then the corresponding id gives the TCP or UDP port number. For more information regarding DCBX spec., please refer to the following: http://www.ieee802.org/1/files/public/docs2008/ az-wadekar-dcbx-capability-exchange-discovery-protocol-1108-v1.01.pdf Signed-off-by: Yi Zou Acked-by: Peter P Waskiewicz Jr Signed-off-by: Jeff Kirsher Signed-off-by: David S. Miller --- include/linux/dcbnl.h | 18 ++++++++++++++++++ 1 file changed, 18 insertions(+) (limited to 'include') diff --git a/include/linux/dcbnl.h b/include/linux/dcbnl.h index 7d2e10006188..b7cdbb4373df 100644 --- a/include/linux/dcbnl.h +++ b/include/linux/dcbnl.h @@ -50,6 +50,8 @@ struct dcbmsg { * @DCB_CMD_SNUMTCS: set the number of traffic classes * @DCB_CMD_GBCN: set backward congestion notification configuration * @DCB_CMD_SBCN: get backward congestion notification configration. + * @DCB_CMD_GAPP: get application protocol configuration + * @DCB_CMD_SAPP: set application protocol configuration */ enum dcbnl_commands { DCB_CMD_UNDEFINED, @@ -80,6 +82,9 @@ enum dcbnl_commands { DCB_CMD_BCN_GCFG, DCB_CMD_BCN_SCFG, + DCB_CMD_GAPP, + DCB_CMD_SAPP, + __DCB_CMD_ENUM_MAX, DCB_CMD_MAX = __DCB_CMD_ENUM_MAX - 1, }; @@ -114,6 +119,7 @@ enum dcbnl_attrs { DCB_ATTR_CAP, DCB_ATTR_NUMTCS, DCB_ATTR_BCN, + DCB_ATTR_APP, __DCB_ATTR_ENUM_MAX, DCB_ATTR_MAX = __DCB_ATTR_ENUM_MAX - 1, @@ -338,5 +344,17 @@ enum dcb_general_attr_values { DCB_ATTR_VALUE_UNDEFINED = 0xff }; +#define DCB_APP_IDTYPE_ETHTYPE 0x00 +#define DCB_APP_IDTYPE_PORTNUM 0x01 +enum dcbnl_app_attrs { + DCB_APP_ATTR_UNDEFINED, + + DCB_APP_ATTR_IDTYPE, + DCB_APP_ATTR_ID, + DCB_APP_ATTR_PRIORITY, + + __DCB_APP_ATTR_ENUM_MAX, + DCB_APP_ATTR_MAX = __DCB_APP_ATTR_ENUM_MAX - 1, +}; #endif /* __LINUX_DCBNL_H__ */ -- cgit v1.2.3 From 7114323b1761bdf787ed79323a4a13f787878295 Mon Sep 17 00:00:00 2001 From: Yi Zou Date: Mon, 31 Aug 2009 12:32:55 +0000 Subject: dcbnl: Add support for setapp/getapp to netdev dcbnl_rtnl_ops Adds support of dcbnl setapp/getapp to dcbnl_rtnl_ops in netdev to allow LLDs to implement their corresponding dcbnl setapp/getapp ops to support the IEEE 802.1Q DCBX setapp/getapp commands. Signed-off-by: Yi Zou Acked-by: Peter P Waskiewicz Jr Signed-off-by: Jeff Kirsher Signed-off-by: David S. Miller --- include/net/dcbnl.h | 2 ++ 1 file changed, 2 insertions(+) (limited to 'include') diff --git a/include/net/dcbnl.h b/include/net/dcbnl.h index 775cfc8055be..b36ac7e0914d 100644 --- a/include/net/dcbnl.h +++ b/include/net/dcbnl.h @@ -48,6 +48,8 @@ struct dcbnl_rtnl_ops { void (*setbcncfg)(struct net_device *, int, u32); void (*getbcnrp)(struct net_device *, int, u8 *); void (*setbcnrp)(struct net_device *, int, u8); + u8 (*setapp)(struct net_device *, u8, u16, u8); + u8 (*getapp)(struct net_device *, u8, u16); }; #endif /* __NET_DCBNL_H__ */ -- cgit v1.2.3 From f1ecd5d9e7366609d640ff4040304ea197fbc618 Mon Sep 17 00:00:00 2001 From: Damian Lukowski Date: Wed, 26 Aug 2009 00:16:31 +0000 Subject: Revert Backoff [v3]: Revert RTO on ICMP destination unreachable MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Here, an ICMP host/network unreachable message, whose payload fits to TCP's SND.UNA, is taken as an indication that the RTO retransmission has not been lost due to congestion, but because of a route failure somewhere along the path. With true congestion, a router won't trigger such a message and the patched TCP will operate as standard TCP. This patch reverts one RTO backoff, if an ICMP host/network unreachable message, whose payload fits to TCP's SND.UNA, arrives. Based on the new RTO, the retransmission timer is reset to reflect the remaining time, or - if the revert clocked out the timer - a retransmission is sent out immediately. Backoffs are only reverted, if TCP is in RTO loss recovery, i.e. if there have been retransmissions and reversible backoffs, already. Changes from v2: 1) Renaming of skb in tcp_v4_err() moved to another patch. 2) Reintroduced tcp_bound_rto() and __tcp_set_rto(). 3) Fixed code comments. Signed-off-by: Damian Lukowski Acked-by: Ilpo Järvinen Signed-off-by: David S. Miller --- include/net/tcp.h | 12 ++++++++++++ net/ipv4/tcp_input.c | 5 ++--- net/ipv4/tcp_ipv4.c | 37 +++++++++++++++++++++++++++++++++++++ net/ipv4/tcp_timer.c | 2 +- 4 files changed, 52 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/include/net/tcp.h b/include/net/tcp.h index cbb2a4889fc9..54f212ce8aaf 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -469,6 +469,7 @@ extern void __tcp_push_pending_frames(struct sock *sk, unsigned int cur_mss, int nonagle); extern int tcp_may_send_now(struct sock *sk); extern int tcp_retransmit_skb(struct sock *, struct sk_buff *); +extern void tcp_retransmit_timer(struct sock *sk); extern void tcp_xmit_retransmit_queue(struct sock *); extern void tcp_simple_retransmit(struct sock *); extern int tcp_trim_head(struct sock *, struct sk_buff *, u32); @@ -521,6 +522,17 @@ extern int tcp_mtu_to_mss(struct sock *sk, int pmtu); extern int tcp_mss_to_mtu(struct sock *sk, int mss); extern void tcp_mtup_init(struct sock *sk); +static inline void tcp_bound_rto(const struct sock *sk) +{ + if (inet_csk(sk)->icsk_rto > TCP_RTO_MAX) + inet_csk(sk)->icsk_rto = TCP_RTO_MAX; +} + +static inline u32 __tcp_set_rto(const struct tcp_sock *tp) +{ + return (tp->srtt >> 3) + tp->rttvar; +} + static inline void __tcp_fast_path_on(struct tcp_sock *tp, u32 snd_wnd) { tp->pred_flags = htonl((tp->tcp_header_len << 26) | diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index 2bdb0da237e6..af6d6fa00db1 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -685,7 +685,7 @@ static inline void tcp_set_rto(struct sock *sk) * is invisible. Actually, Linux-2.4 also generates erratic * ACKs in some circumstances. */ - inet_csk(sk)->icsk_rto = (tp->srtt >> 3) + tp->rttvar; + inet_csk(sk)->icsk_rto = __tcp_set_rto(tp); /* 2. Fixups made earlier cannot be right. * If we do not estimate RTO correctly without them, @@ -696,8 +696,7 @@ static inline void tcp_set_rto(struct sock *sk) /* NOTE: clamping at TCP_RTO_MIN is not required, current algo * guarantees that rto is higher. */ - if (inet_csk(sk)->icsk_rto > TCP_RTO_MAX) - inet_csk(sk)->icsk_rto = TCP_RTO_MAX; + tcp_bound_rto(sk); } /* Save metrics learned by this TCP session. diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index 6ca1bc8c3025..6755e29a6dd3 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -332,12 +332,15 @@ void tcp_v4_err(struct sk_buff *icmp_skb, u32 info) { struct iphdr *iph = (struct iphdr *)icmp_skb->data; struct tcphdr *th = (struct tcphdr *)(icmp_skb->data + (iph->ihl << 2)); + struct inet_connection_sock *icsk; struct tcp_sock *tp; struct inet_sock *inet; const int type = icmp_hdr(icmp_skb)->type; const int code = icmp_hdr(icmp_skb)->code; struct sock *sk; + struct sk_buff *skb; __u32 seq; + __u32 remaining; int err; struct net *net = dev_net(icmp_skb->dev); @@ -367,6 +370,7 @@ void tcp_v4_err(struct sk_buff *icmp_skb, u32 info) if (sk->sk_state == TCP_CLOSE) goto out; + icsk = inet_csk(sk); tp = tcp_sk(sk); seq = ntohl(th->seq); if (sk->sk_state != TCP_LISTEN && @@ -393,6 +397,39 @@ void tcp_v4_err(struct sk_buff *icmp_skb, u32 info) } err = icmp_err_convert[code].errno; + /* check if icmp_skb allows revert of backoff + * (see draft-zimmermann-tcp-lcd) */ + if (code != ICMP_NET_UNREACH && code != ICMP_HOST_UNREACH) + break; + if (seq != tp->snd_una || !icsk->icsk_retransmits || + !icsk->icsk_backoff) + break; + + icsk->icsk_backoff--; + inet_csk(sk)->icsk_rto = __tcp_set_rto(tp) << + icsk->icsk_backoff; + tcp_bound_rto(sk); + + skb = tcp_write_queue_head(sk); + BUG_ON(!skb); + + remaining = icsk->icsk_rto - min(icsk->icsk_rto, + tcp_time_stamp - TCP_SKB_CB(skb)->when); + + if (remaining) { + inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS, + remaining, TCP_RTO_MAX); + } else if (sock_owned_by_user(sk)) { + /* RTO revert clocked out retransmission, + * but socket is locked. Will defer. */ + inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS, + HZ/20, TCP_RTO_MAX); + } else { + /* RTO revert clocked out retransmission. + * Will retransmit now */ + tcp_retransmit_timer(sk); + } + break; case ICMP_TIME_EXCEEDED: err = EHOSTUNREACH; diff --git a/net/ipv4/tcp_timer.c b/net/ipv4/tcp_timer.c index c520fb6e06d9..408fa4b7b9ba 100644 --- a/net/ipv4/tcp_timer.c +++ b/net/ipv4/tcp_timer.c @@ -279,7 +279,7 @@ static void tcp_probe_timer(struct sock *sk) * The TCP retransmit timer. */ -static void tcp_retransmit_timer(struct sock *sk) +void tcp_retransmit_timer(struct sock *sk) { struct tcp_sock *tp = tcp_sk(sk); struct inet_connection_sock *icsk = inet_csk(sk); -- cgit v1.2.3 From 6fa12c85031485dff38ce550c24f10da23b0adaa Mon Sep 17 00:00:00 2001 From: Damian Lukowski Date: Wed, 26 Aug 2009 00:16:34 +0000 Subject: Revert Backoff [v3]: Calculate TCP's connection close threshold as a time value. MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit RFC 1122 specifies two threshold values R1 and R2 for connection timeouts, which may represent a number of allowed retransmissions or a timeout value. Currently linux uses sysctl_tcp_retries{1,2} to specify the thresholds in number of allowed retransmissions. For any desired threshold R2 (by means of time) one can specify tcp_retries2 (by means of number of retransmissions) such that TCP will not time out earlier than R2. This is the case, because the RTO schedule follows a fixed pattern, namely exponential backoff. However, the RTO behaviour is not predictable any more if RTO backoffs can be reverted, as it is the case in the draft "Make TCP more Robust to Long Connectivity Disruptions" (http://tools.ietf.org/html/draft-zimmermann-tcp-lcd). In the worst case TCP would time out a connection after 3.2 seconds, if the initial RTO equaled MIN_RTO and each backoff has been reverted. This patch introduces a function retransmits_timed_out(N), which calculates the timeout of a TCP connection, assuming an initial RTO of MIN_RTO and N unsuccessful, exponentially backed-off retransmissions. Whenever timeout decisions are made by comparing the retransmission counter to some value N, this function can be used, instead. The meaning of tcp_retries2 will be changed, as many more RTO retransmissions can occur than the value indicates. However, it yields a timeout which is similar to the one of an unpatched, exponentially backing off TCP in the same scenario. As no application could rely on an RTO greater than MIN_RTO, there should be no risk of a regression. Signed-off-by: Damian Lukowski Acked-by: Ilpo Järvinen Signed-off-by: David S. Miller --- include/net/tcp.h | 18 ++++++++++++++++++ net/ipv4/tcp_timer.c | 11 +++++++---- 2 files changed, 25 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/include/net/tcp.h b/include/net/tcp.h index 54f212ce8aaf..e5319495f15e 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -1252,6 +1252,24 @@ static inline struct sk_buff *tcp_write_queue_prev(struct sock *sk, struct sk_bu #define tcp_for_write_queue_from_safe(skb, tmp, sk) \ skb_queue_walk_from_safe(&(sk)->sk_write_queue, skb, tmp) +static inline bool retransmits_timed_out(const struct sock *sk, + unsigned int boundary) +{ + int limit, K; + if (!inet_csk(sk)->icsk_retransmits) + return false; + + K = ilog2(TCP_RTO_MAX/TCP_RTO_MIN); + + if (boundary <= K) + limit = ((2 << boundary) - 1) * TCP_RTO_MIN; + else + limit = ((2 << K) - 1) * TCP_RTO_MIN + + (boundary - K) * TCP_RTO_MAX; + + return (tcp_time_stamp - tcp_sk(sk)->retrans_stamp) >= limit; +} + static inline struct sk_buff *tcp_send_head(struct sock *sk) { return sk->sk_send_head; diff --git a/net/ipv4/tcp_timer.c b/net/ipv4/tcp_timer.c index 408fa4b7b9ba..cdb2ca7684d4 100644 --- a/net/ipv4/tcp_timer.c +++ b/net/ipv4/tcp_timer.c @@ -137,13 +137,14 @@ static int tcp_write_timeout(struct sock *sk) { struct inet_connection_sock *icsk = inet_csk(sk); int retry_until; + bool do_reset; if ((1 << sk->sk_state) & (TCPF_SYN_SENT | TCPF_SYN_RECV)) { if (icsk->icsk_retransmits) dst_negative_advice(&sk->sk_dst_cache); retry_until = icsk->icsk_syn_retries ? : sysctl_tcp_syn_retries; } else { - if (icsk->icsk_retransmits >= sysctl_tcp_retries1) { + if (retransmits_timed_out(sk, sysctl_tcp_retries1)) { /* Black hole detection */ tcp_mtu_probing(icsk, sk); @@ -155,13 +156,15 @@ static int tcp_write_timeout(struct sock *sk) const int alive = (icsk->icsk_rto < TCP_RTO_MAX); retry_until = tcp_orphan_retries(sk, alive); + do_reset = alive || + !retransmits_timed_out(sk, retry_until); - if (tcp_out_of_resources(sk, alive || icsk->icsk_retransmits < retry_until)) + if (tcp_out_of_resources(sk, do_reset)) return 1; } } - if (icsk->icsk_retransmits >= retry_until) { + if (retransmits_timed_out(sk, retry_until)) { /* Has it gone just too far? */ tcp_write_err(sk); return 1; @@ -385,7 +388,7 @@ void tcp_retransmit_timer(struct sock *sk) out_reset_timer: icsk->icsk_rto = min(icsk->icsk_rto << 1, TCP_RTO_MAX); inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS, icsk->icsk_rto, TCP_RTO_MAX); - if (icsk->icsk_retransmits > sysctl_tcp_retries1) + if (retransmits_timed_out(sk, sysctl_tcp_retries1 + 1)) __sk_dst_reset(sk); out:; -- cgit v1.2.3 From 8aa84ad8d6c740a04386f599694609ee4998e82e Mon Sep 17 00:00:00 2001 From: Thomas Renninger Date: Fri, 24 Jul 2009 15:25:05 +0200 Subject: [CPUFREQ] Introduce global, not per core: /sys/devices/system/cpu/cpufreq Currently everything in the cpufreq layer is per core based. This does not reflect reality, for example ondemand on conservative governors have global sysfs variables. Introduce a global cpufreq directory and add the kobject to the governor struct, so that governors can easily access it. The directory is initialized in the cpufreq_core_init initcall and thus will always be created if cpufreq is compiled in, even if no cpufreq driver is active later. Signed-off-by: Thomas Renninger Signed-off-by: Dave Jones --- drivers/cpufreq/cpufreq.c | 9 ++++++++- include/linux/cpufreq.h | 10 ++++++++++ 2 files changed, 18 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c index bbd5c2164ab6..4da28444b235 100644 --- a/drivers/cpufreq/cpufreq.c +++ b/drivers/cpufreq/cpufreq.c @@ -686,6 +686,9 @@ static struct attribute *default_attrs[] = { NULL }; +struct kobject *cpufreq_global_kobject; +EXPORT_SYMBOL(cpufreq_global_kobject); + #define to_policy(k) container_of(k, struct cpufreq_policy, kobj) #define to_attr(a) container_of(a, struct freq_attr, attr) @@ -1935,7 +1938,11 @@ static int __init cpufreq_core_init(void) per_cpu(policy_cpu, cpu) = -1; init_rwsem(&per_cpu(cpu_policy_rwsem, cpu)); } + + cpufreq_global_kobject = kobject_create_and_add("cpufreq", + &cpu_sysdev_class.kset.kobj); + BUG_ON(!cpufreq_global_kobject); + return 0; } - core_initcall(cpufreq_core_init); diff --git a/include/linux/cpufreq.h b/include/linux/cpufreq.h index 161042746afc..44717eb47639 100644 --- a/include/linux/cpufreq.h +++ b/include/linux/cpufreq.h @@ -65,6 +65,9 @@ static inline int cpufreq_unregister_notifier(struct notifier_block *nb, struct cpufreq_governor; +/* /sys/devices/system/cpu/cpufreq: entry point for global variables */ +extern struct kobject *cpufreq_global_kobject; + #define CPUFREQ_ETERNAL (-1) struct cpufreq_cpuinfo { unsigned int max_freq; @@ -274,6 +277,13 @@ struct freq_attr { ssize_t (*store)(struct cpufreq_policy *, const char *, size_t count); }; +struct global_attr { + struct attribute attr; + ssize_t (*show)(struct kobject *kobj, + struct attribute *attr, char *buf); + ssize_t (*store)(struct kobject *a, struct attribute *b, + const char *c, size_t count); +}; /********************************************************************* * CPUFREQ 2.6. INTERFACE * -- cgit v1.2.3 From 8bc11b491b6cad75e737f1d38bb4b261bd95b291 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Wed, 26 Aug 2009 18:13:17 +0200 Subject: rfkill: relicense header file MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This header file is copied into userspace tools that need not be GPLv2 licensed, make that easier. Signed-off-by: Johannes Berg Acked-by: Alan Jenkins Acked-by: Henrique de Moraes Holschuh Acked-by: Iñaky Pérez-González Acked-by: Ivo van Doorn Acked-by: Jaswinder Singh Rajput Acked-by: Michael Buesch Acked-by: Tomas Winkler Signed-off-by: John W. Linville --- include/linux/rfkill.h | 23 ++++++++++------------- 1 file changed, 10 insertions(+), 13 deletions(-) (limited to 'include') diff --git a/include/linux/rfkill.h b/include/linux/rfkill.h index 21ca51bf4dd2..3392c59d2706 100644 --- a/include/linux/rfkill.h +++ b/include/linux/rfkill.h @@ -6,20 +6,17 @@ * Copyright (C) 2007 Dmitry Torokhov * Copyright 2009 Johannes Berg * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. + * Permission to use, copy, modify, and/or distribute this software for any + * purpose with or without fee is hereby granted, provided that the above + * copyright notice and this permission notice appear in all copies. * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the - * Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. + * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES + * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF + * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR + * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES + * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN + * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF + * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. */ #include -- cgit v1.2.3 From fa8a123855e20068204982596b8fafceb1a67f0b Mon Sep 17 00:00:00 2001 From: Dave Airlie Date: Wed, 26 Aug 2009 13:13:37 +1000 Subject: drm/mm: add ability to dump mm lists via debugfs This adds code to the drm_mm to talk to debugfs, and adds support to radeon to add the VRAM and GTT mm lists to debugfs. I tested with spinlock debugging and it doesn't give out. Signed-off-by: Dave Airlie --- drivers/gpu/drm/drm_mm.c | 21 ++++++++++++++ drivers/gpu/drm/radeon/radeon_ttm.c | 56 +++++++++++++++++++++++++++++++++++++ include/drm/drm_mm.h | 4 +++ 3 files changed, 81 insertions(+) (limited to 'include') diff --git a/drivers/gpu/drm/drm_mm.c b/drivers/gpu/drm/drm_mm.c index 3e47869d6dae..c861d80fd779 100644 --- a/drivers/gpu/drm/drm_mm.c +++ b/drivers/gpu/drm/drm_mm.c @@ -44,6 +44,7 @@ #include "drmP.h" #include "drm_mm.h" #include +#include #define MM_UNUSED_TARGET 4 @@ -370,3 +371,23 @@ void drm_mm_takedown(struct drm_mm * mm) BUG_ON(mm->num_unused != 0); } EXPORT_SYMBOL(drm_mm_takedown); + +#if defined(CONFIG_DEBUG_FS) +int drm_mm_dump_table(struct seq_file *m, struct drm_mm *mm) +{ + struct drm_mm_node *entry; + int total_used = 0, total_free = 0, total = 0; + + list_for_each_entry(entry, &mm->ml_entry, ml_entry) { + seq_printf(m, "0x%08lx-0x%08lx: 0x%08lx: %s\n", entry->start, entry->start + entry->size, entry->size, entry->free ? "free" : "used"); + total += entry->size; + if (entry->free) + total_free += entry->size; + else + total_used += entry->size; + } + seq_printf(m, "total: %d, used %d free %d\n", total, total_free, total_used); + return 0; +} +EXPORT_SYMBOL(drm_mm_dump_table); +#endif diff --git a/drivers/gpu/drm/radeon/radeon_ttm.c b/drivers/gpu/drm/radeon/radeon_ttm.c index 0a85e7b5d592..dc7a44274ea8 100644 --- a/drivers/gpu/drm/radeon/radeon_ttm.c +++ b/drivers/gpu/drm/radeon/radeon_ttm.c @@ -35,11 +35,14 @@ #include #include #include +#include #include "radeon_reg.h" #include "radeon.h" #define DRM_FILE_PAGE_OFFSET (0x100000000ULL >> PAGE_SHIFT) +static int radeon_ttm_debugfs_init(struct radeon_device *rdev); + static struct radeon_device *radeon_get_rdev(struct ttm_bo_device *bdev) { struct radeon_mman *mman; @@ -504,6 +507,12 @@ int radeon_ttm_init(struct radeon_device *rdev) if (unlikely(rdev->mman.bdev.dev_mapping == NULL)) { rdev->mman.bdev.dev_mapping = rdev->ddev->dev_mapping; } + + r = radeon_ttm_debugfs_init(rdev); + if (r) { + DRM_ERROR("Failed to init debugfs\n"); + return r; + } return 0; } @@ -678,3 +687,50 @@ struct ttm_backend *radeon_ttm_backend_create(struct radeon_device *rdev) gtt->bound = false; return >t->backend; } + +#define RADEON_DEBUGFS_MEM_TYPES 2 + +static struct drm_info_list radeon_mem_types_list[RADEON_DEBUGFS_MEM_TYPES]; +static char radeon_mem_types_names[RADEON_DEBUGFS_MEM_TYPES][32]; + +#if defined(CONFIG_DEBUG_FS) +static int radeon_mm_dump_table(struct seq_file *m, void *data) +{ + struct drm_info_node *node = (struct drm_info_node *)m->private; + struct drm_mm *mm = (struct drm_mm *)node->info_ent->data; + struct drm_device *dev = node->minor->dev; + struct radeon_device *rdev = dev->dev_private; + int ret; + struct ttm_bo_global *glob = rdev->mman.bdev.glob; + + spin_lock(&glob->lru_lock); + ret = drm_mm_dump_table(m, mm); + spin_unlock(&glob->lru_lock); + return ret; +} +#endif + +static int radeon_ttm_debugfs_init(struct radeon_device *rdev) +{ + unsigned i; + +#if defined(CONFIG_DEBUG_FS) + for (i = 0; i < RADEON_DEBUGFS_MEM_TYPES; i++) { + if (i == 0) + sprintf(radeon_mem_types_names[i], "radeon_vram_mm"); + else + sprintf(radeon_mem_types_names[i], "radeon_gtt_mm"); + radeon_mem_types_list[i].name = radeon_mem_types_names[i]; + radeon_mem_types_list[i].show = &radeon_mm_dump_table; + radeon_mem_types_list[i].driver_features = 0; + if (i == 0) + radeon_mem_types_list[i].data = &rdev->mman.bdev.man[TTM_PL_VRAM].manager; + else + radeon_mem_types_list[i].data = &rdev->mman.bdev.man[TTM_PL_TT].manager; + + } + return radeon_debugfs_add_files(rdev, radeon_mem_types_list, RADEON_DEBUGFS_MEM_TYPES); + +#endif + return 0; +} diff --git a/include/drm/drm_mm.h b/include/drm/drm_mm.h index f8332073d277..bc5a87e8aeea 100644 --- a/include/drm/drm_mm.h +++ b/include/drm/drm_mm.h @@ -96,4 +96,8 @@ static inline struct drm_mm *drm_get_mm(struct drm_mm_node *block) return block->mm; } +#ifdef CONFIG_DEBUG_FS +int drm_mm_dump_table(struct seq_file *m, struct drm_mm *mm); +#endif + #endif -- cgit v1.2.3 From a3a0544b2c84e1d7a2022b558ecf66d8c6a8dd93 Mon Sep 17 00:00:00 2001 From: Dave Airlie Date: Mon, 31 Aug 2009 15:16:30 +1000 Subject: drm/kms: add explicit encoder disable function and detach harder. For shared tv-out and VGA encoders, we really need to know if the encoder is just being switched off temporarily in blanking or if we are really disabling it hard. Also we need to try harder to disconnect encoders from unused connectors so we can share more efficently. (shared encoders stuff is coming in radeon tv-out support) Signed-off-by: Dave Airlie --- drivers/gpu/drm/drm_crtc_helper.c | 24 ++++++++++++++++++++---- include/drm/drm_crtc_helper.h | 2 ++ 2 files changed, 22 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/drivers/gpu/drm/drm_crtc_helper.c b/drivers/gpu/drm/drm_crtc_helper.c index 205349ea1075..eea5e6c4099c 100644 --- a/drivers/gpu/drm/drm_crtc_helper.c +++ b/drivers/gpu/drm/drm_crtc_helper.c @@ -260,13 +260,27 @@ EXPORT_SYMBOL(drm_helper_crtc_in_use); void drm_helper_disable_unused_functions(struct drm_device *dev) { struct drm_encoder *encoder; + struct drm_connector *connector; struct drm_encoder_helper_funcs *encoder_funcs; struct drm_crtc *crtc; + list_for_each_entry(connector, &dev->mode_config.connector_list, head) { + if (!connector->encoder) + continue; + if (connector->status == connector_status_disconnected) + connector->encoder = NULL; + } + list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) { encoder_funcs = encoder->helper_private; - if (!drm_helper_encoder_in_use(encoder)) - (*encoder_funcs->dpms)(encoder, DRM_MODE_DPMS_OFF); + if (!drm_helper_encoder_in_use(encoder)) { + if (encoder_funcs->disable) + (*encoder_funcs->disable)(encoder); + else + (*encoder_funcs->dpms)(encoder, DRM_MODE_DPMS_OFF); + } + /* disconnector encoder from any connector */ + encoder->crtc = NULL; } list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) { @@ -411,7 +425,7 @@ static int drm_pick_crtcs(struct drm_device *dev, c = 0; list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) { - if ((connector->encoder->possible_crtcs & (1 << c)) == 0) { + if ((encoder->possible_crtcs & (1 << c)) == 0) { c++; continue; } @@ -496,8 +510,10 @@ static void drm_setup_crtcs(struct drm_device *dev) mode->name, crtc->base.id); crtc->desired_mode = mode; connector->encoder->crtc = crtc; - } else + } else { connector->encoder->crtc = NULL; + connector->encoder = NULL; + } i++; } diff --git a/include/drm/drm_crtc_helper.h b/include/drm/drm_crtc_helper.h index e44a4f87303c..4c8dacaf4f58 100644 --- a/include/drm/drm_crtc_helper.h +++ b/include/drm/drm_crtc_helper.h @@ -79,6 +79,8 @@ struct drm_encoder_helper_funcs { /* detect for DAC style encoders */ enum drm_connector_status (*detect)(struct drm_encoder *encoder, struct drm_connector *connector); + /* disable encoder when not in use - more explicit than dpms off */ + void (*disable)(struct drm_encoder *encoder); }; struct drm_connector_helper_funcs { -- cgit v1.2.3 From 388539f3ff0cf1de926b03f94e1eec112358f74d Mon Sep 17 00:00:00 2001 From: Shaohua Li Date: Mon, 27 Jul 2009 09:24:35 +0800 Subject: [libata] add DMA setup FIS auto-activate feature Hopefully results in fewer on-the-wire FIS's and no breakage. We'll see! Signed-off-by: Shaohua Li Signed-off-by: Jeff Garzik --- drivers/ata/ahci.c | 2 +- drivers/ata/libata-core.c | 36 +++++++++++++++++++++++++++++------- include/linux/ata.h | 4 ++++ include/linux/libata.h | 2 ++ 4 files changed, 36 insertions(+), 8 deletions(-) (limited to 'include') diff --git a/drivers/ata/ahci.c b/drivers/ata/ahci.c index fe3eba5d6b3e..7a3124a9abc6 100644 --- a/drivers/ata/ahci.c +++ b/drivers/ata/ahci.c @@ -2869,7 +2869,7 @@ static int ahci_init_one(struct pci_dev *pdev, const struct pci_device_id *ent) /* prepare host */ if (hpriv->cap & HOST_CAP_NCQ) - pi.flags |= ATA_FLAG_NCQ; + pi.flags |= ATA_FLAG_NCQ | ATA_FLAG_FPDMA_AA; if (hpriv->cap & HOST_CAP_PMP) pi.flags |= ATA_FLAG_PMP; diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c index 072ba5ea138f..98af50f16e0c 100644 --- a/drivers/ata/libata-core.c +++ b/drivers/ata/libata-core.c @@ -2299,29 +2299,49 @@ static inline u8 ata_dev_knobble(struct ata_device *dev) return ((ap->cbl == ATA_CBL_SATA) && (!ata_id_is_sata(dev->id))); } -static void ata_dev_config_ncq(struct ata_device *dev, +static int ata_dev_config_ncq(struct ata_device *dev, char *desc, size_t desc_sz) { struct ata_port *ap = dev->link->ap; int hdepth = 0, ddepth = ata_id_queue_depth(dev->id); + unsigned int err_mask; + char *aa_desc = ""; if (!ata_id_has_ncq(dev->id)) { desc[0] = '\0'; - return; + return 0; } if (dev->horkage & ATA_HORKAGE_NONCQ) { snprintf(desc, desc_sz, "NCQ (not used)"); - return; + return 0; } if (ap->flags & ATA_FLAG_NCQ) { hdepth = min(ap->scsi_host->can_queue, ATA_MAX_QUEUE - 1); dev->flags |= ATA_DFLAG_NCQ; } + if (!(dev->horkage & ATA_HORKAGE_BROKEN_FPDMA_AA) && + (ap->flags & ATA_FLAG_FPDMA_AA) && + ata_id_has_fpdma_aa(dev->id)) { + err_mask = ata_dev_set_feature(dev, SETFEATURES_SATA_ENABLE, + SATA_FPDMA_AA); + if (err_mask) { + ata_dev_printk(dev, KERN_ERR, "failed to enable AA" + "(error_mask=0x%x)\n", err_mask); + if (err_mask != AC_ERR_DEV) { + dev->horkage |= ATA_HORKAGE_BROKEN_FPDMA_AA; + return -EIO; + } + } else + aa_desc = ", AA"; + } + if (hdepth >= ddepth) - snprintf(desc, desc_sz, "NCQ (depth %d)", ddepth); + snprintf(desc, desc_sz, "NCQ (depth %d)%s", ddepth, aa_desc); else - snprintf(desc, desc_sz, "NCQ (depth %d/%d)", hdepth, ddepth); + snprintf(desc, desc_sz, "NCQ (depth %d/%d)%s", hdepth, + ddepth, aa_desc); + return 0; } /** @@ -2461,7 +2481,7 @@ int ata_dev_configure(struct ata_device *dev) if (ata_id_has_lba(id)) { const char *lba_desc; - char ncq_desc[20]; + char ncq_desc[24]; lba_desc = "LBA"; dev->flags |= ATA_DFLAG_LBA; @@ -2475,7 +2495,9 @@ int ata_dev_configure(struct ata_device *dev) } /* config NCQ */ - ata_dev_config_ncq(dev, ncq_desc, sizeof(ncq_desc)); + rc = ata_dev_config_ncq(dev, ncq_desc, sizeof(ncq_desc)); + if (rc) + return rc; /* print device info to dmesg */ if (ata_msg_drv(ap) && print_info) { diff --git a/include/linux/ata.h b/include/linux/ata.h index 9c75921f0c16..f5494050df83 100644 --- a/include/linux/ata.h +++ b/include/linux/ata.h @@ -306,6 +306,7 @@ enum { /* SETFEATURE Sector counts for SATA features */ SATA_AN = 0x05, /* Asynchronous Notification */ SATA_DIPM = 0x03, /* Device Initiated Power Management */ + SATA_FPDMA_AA = 0x02, /* DMA Setup FIS Auto-Activate */ /* feature values for SET_MAX */ ATA_SET_MAX_ADDR = 0x00, @@ -525,6 +526,9 @@ static inline int ata_is_data(u8 prot) #define ata_id_has_atapi_AN(id) \ ( (((id)[76] != 0x0000) && ((id)[76] != 0xffff)) && \ ((id)[78] & (1 << 5)) ) +#define ata_id_has_fpdma_aa(id) \ + ( (((id)[76] != 0x0000) && ((id)[76] != 0xffff)) && \ + ((id)[78] & (1 << 2)) ) #define ata_id_iordy_disable(id) ((id)[ATA_ID_CAPABILITY] & (1 << 10)) #define ata_id_has_iordy(id) ((id)[ATA_ID_CAPABILITY] & (1 << 11)) #define ata_id_u32(id,n) \ diff --git a/include/linux/libata.h b/include/linux/libata.h index e5b6e33c6571..d3f7cab4873e 100644 --- a/include/linux/libata.h +++ b/include/linux/libata.h @@ -190,6 +190,7 @@ enum { ATA_FLAG_NO_POWEROFF_SPINDOWN = (1 << 11), /* don't spindown before poweroff */ ATA_FLAG_NO_HIBERNATE_SPINDOWN = (1 << 12), /* don't spindown before hibernation */ ATA_FLAG_DEBUGMSG = (1 << 13), + ATA_FLAG_FPDMA_AA = (1 << 14), /* driver supports Auto-Activate */ ATA_FLAG_IGN_SIMPLEX = (1 << 15), /* ignore SIMPLEX */ ATA_FLAG_NO_IORDY = (1 << 16), /* controller lacks iordy */ ATA_FLAG_ACPI_SATA = (1 << 17), /* need native SATA ACPI layout */ @@ -386,6 +387,7 @@ enum { ATA_HORKAGE_FIRMWARE_WARN = (1 << 12), /* firmware update warning */ ATA_HORKAGE_1_5_GBPS = (1 << 13), /* force 1.5 Gbps */ ATA_HORKAGE_NOSETXFER = (1 << 14), /* skip SETXFER, SATA only */ + ATA_HORKAGE_BROKEN_FPDMA_AA = (1 << 15), /* skip AA */ /* DMA mask for user DMA control: User visible values; DO NOT renumber */ -- cgit v1.2.3 From 6521148c6449724c3b707820b9c535c7e8b8afcd Mon Sep 17 00:00:00 2001 From: Robert Hancock Date: Tue, 14 Jul 2009 20:43:39 -0600 Subject: libata: add command name parsing for error output This patch improve libata's output for error/notification messages to allow easier comprehension and debugging: When ATAPI commands issued through the SCSI layer fail, use SCSI functions to print the CDB in human-readable form instead of just dumping out the CDB in hex. Print out the name of the failed command (as defined by the ATA specification) in error handling output along with the raw register contents. When reporting status of ACPI taskfile commands executed on resume, also output the names of the commands being executed (or not) in readable form. Since the extra data for printing command names increases kernel size slightly, a config option has been added to allow disabling command name output (as well as some of the error register parsing) for those highly sensitive to kernel text size. Signed-off-by: Robert Hancock Signed-off-by: Jeff Garzik --- drivers/ata/Kconfig | 11 ++++ drivers/ata/libata-acpi.c | 7 ++- drivers/ata/libata-eh.c | 128 +++++++++++++++++++++++++++++++++++++++++++++- drivers/ata/libata.h | 1 + include/linux/ata.h | 32 +++++++++++- 5 files changed, 174 insertions(+), 5 deletions(-) (limited to 'include') diff --git a/drivers/ata/Kconfig b/drivers/ata/Kconfig index b17c57f85032..8e64d3c81d53 100644 --- a/drivers/ata/Kconfig +++ b/drivers/ata/Kconfig @@ -26,6 +26,17 @@ config ATA_NONSTANDARD bool default n +config ATA_VERBOSE_ERROR + bool "Verbose ATA error reporting" + default y + help + This option adds parsing of ATA command descriptions and error bits + in libata kernel output, making it easier to interpret. + This option will enlarge the kernel by approx. 6KB. Disable it only + if kernel size is more important than ease of debugging. + + If unsure, say Y. + config ATA_ACPI bool "ATA ACPI Support" depends on ACPI && PCI diff --git a/drivers/ata/libata-acpi.c b/drivers/ata/libata-acpi.c index ac176da1f94e..01964b6e6f6b 100644 --- a/drivers/ata/libata-acpi.c +++ b/drivers/ata/libata-acpi.c @@ -689,6 +689,7 @@ static int ata_acpi_run_tf(struct ata_device *dev, struct ata_taskfile tf, ptf, rtf; unsigned int err_mask; const char *level; + const char *descr; char msg[60]; int rc; @@ -736,11 +737,13 @@ static int ata_acpi_run_tf(struct ata_device *dev, snprintf(msg, sizeof(msg), "filtered out"); rc = 0; } + descr = ata_get_cmd_descript(tf.command); ata_dev_printk(dev, level, - "ACPI cmd %02x/%02x:%02x:%02x:%02x:%02x:%02x %s\n", + "ACPI cmd %02x/%02x:%02x:%02x:%02x:%02x:%02x (%s) %s\n", tf.command, tf.feature, tf.nsect, tf.lbal, - tf.lbam, tf.lbah, tf.device, msg); + tf.lbam, tf.lbah, tf.device, + (descr ? descr : "unknown"), msg); return rc; } diff --git a/drivers/ata/libata-eh.c b/drivers/ata/libata-eh.c index 2c34de841e11..a04488f0de88 100644 --- a/drivers/ata/libata-eh.c +++ b/drivers/ata/libata-eh.c @@ -40,6 +40,7 @@ #include #include #include +#include #include "../scsi/scsi_transport_api.h" #include @@ -2111,6 +2112,116 @@ void ata_eh_autopsy(struct ata_port *ap) ata_eh_link_autopsy(&ap->link); } +/** + * ata_get_cmd_descript - get description for ATA command + * @command: ATA command code to get description for + * + * Return a textual description of the given command, or NULL if the + * command is not known. + * + * LOCKING: + * None + */ +const char *ata_get_cmd_descript(u8 command) +{ +#ifdef CONFIG_ATA_VERBOSE_ERROR + static const struct + { + u8 command; + const char *text; + } cmd_descr[] = { + { ATA_CMD_DEV_RESET, "DEVICE RESET" }, + { ATA_CMD_CHK_POWER, "CHECK POWER MODE" }, + { ATA_CMD_STANDBY, "STANDBY" }, + { ATA_CMD_IDLE, "IDLE" }, + { ATA_CMD_EDD, "EXECUTE DEVICE DIAGNOSTIC" }, + { ATA_CMD_DOWNLOAD_MICRO, "DOWNLOAD MICROCODE" }, + { ATA_CMD_NOP, "NOP" }, + { ATA_CMD_FLUSH, "FLUSH CACHE" }, + { ATA_CMD_FLUSH_EXT, "FLUSH CACHE EXT" }, + { ATA_CMD_ID_ATA, "IDENTIFY DEVICE" }, + { ATA_CMD_ID_ATAPI, "IDENTIFY PACKET DEVICE" }, + { ATA_CMD_SERVICE, "SERVICE" }, + { ATA_CMD_READ, "READ DMA" }, + { ATA_CMD_READ_EXT, "READ DMA EXT" }, + { ATA_CMD_READ_QUEUED, "READ DMA QUEUED" }, + { ATA_CMD_READ_STREAM_EXT, "READ STREAM EXT" }, + { ATA_CMD_READ_STREAM_DMA_EXT, "READ STREAM DMA EXT" }, + { ATA_CMD_WRITE, "WRITE DMA" }, + { ATA_CMD_WRITE_EXT, "WRITE DMA EXT" }, + { ATA_CMD_WRITE_QUEUED, "WRITE DMA QUEUED EXT" }, + { ATA_CMD_WRITE_STREAM_EXT, "WRITE STREAM EXT" }, + { ATA_CMD_WRITE_STREAM_DMA_EXT, "WRITE STREAM DMA EXT" }, + { ATA_CMD_WRITE_FUA_EXT, "WRITE DMA FUA EXT" }, + { ATA_CMD_WRITE_QUEUED_FUA_EXT, "WRITE DMA QUEUED FUA EXT" }, + { ATA_CMD_FPDMA_READ, "READ FPDMA QUEUED" }, + { ATA_CMD_FPDMA_WRITE, "WRITE FPDMA QUEUED" }, + { ATA_CMD_PIO_READ, "READ SECTOR(S)" }, + { ATA_CMD_PIO_READ_EXT, "READ SECTOR(S) EXT" }, + { ATA_CMD_PIO_WRITE, "WRITE SECTOR(S)" }, + { ATA_CMD_PIO_WRITE_EXT, "WRITE SECTOR(S) EXT" }, + { ATA_CMD_READ_MULTI, "READ MULTIPLE" }, + { ATA_CMD_READ_MULTI_EXT, "READ MULTIPLE EXT" }, + { ATA_CMD_WRITE_MULTI, "WRITE MULTIPLE" }, + { ATA_CMD_WRITE_MULTI_EXT, "WRITE MULTIPLE EXT" }, + { ATA_CMD_WRITE_MULTI_FUA_EXT, "WRITE MULTIPLE FUA EXT" }, + { ATA_CMD_SET_FEATURES, "SET FEATURES" }, + { ATA_CMD_SET_MULTI, "SET MULTIPLE MODE" }, + { ATA_CMD_VERIFY, "READ VERIFY SECTOR(S)" }, + { ATA_CMD_VERIFY_EXT, "READ VERIFY SECTOR(S) EXT" }, + { ATA_CMD_WRITE_UNCORR_EXT, "WRITE UNCORRECTABLE EXT" }, + { ATA_CMD_STANDBYNOW1, "STANDBY IMMEDIATE" }, + { ATA_CMD_IDLEIMMEDIATE, "IDLE IMMEDIATE" }, + { ATA_CMD_SLEEP, "SLEEP" }, + { ATA_CMD_INIT_DEV_PARAMS, "INITIALIZE DEVICE PARAMETERS" }, + { ATA_CMD_READ_NATIVE_MAX, "READ NATIVE MAX ADDRESS" }, + { ATA_CMD_READ_NATIVE_MAX_EXT, "READ NATIVE MAX ADDRESS EXT" }, + { ATA_CMD_SET_MAX, "SET MAX ADDRESS" }, + { ATA_CMD_SET_MAX_EXT, "SET MAX ADDRESS EXT" }, + { ATA_CMD_READ_LOG_EXT, "READ LOG EXT" }, + { ATA_CMD_WRITE_LOG_EXT, "WRITE LOG EXT" }, + { ATA_CMD_READ_LOG_DMA_EXT, "READ LOG DMA EXT" }, + { ATA_CMD_WRITE_LOG_DMA_EXT, "WRITE LOG DMA EXT" }, + { ATA_CMD_TRUSTED_RCV, "TRUSTED RECEIVE" }, + { ATA_CMD_TRUSTED_RCV_DMA, "TRUSTED RECEIVE DMA" }, + { ATA_CMD_TRUSTED_SND, "TRUSTED SEND" }, + { ATA_CMD_TRUSTED_SND_DMA, "TRUSTED SEND DMA" }, + { ATA_CMD_PMP_READ, "READ BUFFER" }, + { ATA_CMD_PMP_WRITE, "WRITE BUFFER" }, + { ATA_CMD_CONF_OVERLAY, "DEVICE CONFIGURATION OVERLAY" }, + { ATA_CMD_SEC_SET_PASS, "SECURITY SET PASSWORD" }, + { ATA_CMD_SEC_UNLOCK, "SECURITY UNLOCK" }, + { ATA_CMD_SEC_ERASE_PREP, "SECURITY ERASE PREPARE" }, + { ATA_CMD_SEC_ERASE_UNIT, "SECURITY ERASE UNIT" }, + { ATA_CMD_SEC_FREEZE_LOCK, "SECURITY FREEZE LOCK" }, + { ATA_CMD_SEC_DISABLE_PASS, "SECURITY DISABLE PASSWORD" }, + { ATA_CMD_CONFIG_STREAM, "CONFIGURE STREAM" }, + { ATA_CMD_SMART, "SMART" }, + { ATA_CMD_MEDIA_LOCK, "DOOR LOCK" }, + { ATA_CMD_MEDIA_UNLOCK, "DOOR UNLOCK" }, + { ATA_CMD_CHK_MED_CRD_TYP, "CHECK MEDIA CARD TYPE" }, + { ATA_CMD_CFA_REQ_EXT_ERR, "CFA REQUEST EXTENDED ERROR" }, + { ATA_CMD_CFA_WRITE_NE, "CFA WRITE SECTORS WITHOUT ERASE" }, + { ATA_CMD_CFA_TRANS_SECT, "CFA TRANSLATE SECTOR" }, + { ATA_CMD_CFA_ERASE, "CFA ERASE SECTORS" }, + { ATA_CMD_CFA_WRITE_MULT_NE, "CFA WRITE MULTIPLE WITHOUT ERASE" }, + { ATA_CMD_READ_LONG, "READ LONG (with retries)" }, + { ATA_CMD_READ_LONG_ONCE, "READ LONG (without retries)" }, + { ATA_CMD_WRITE_LONG, "WRITE LONG (with retries)" }, + { ATA_CMD_WRITE_LONG_ONCE, "WRITE LONG (without retries)" }, + { ATA_CMD_RESTORE, "RECALIBRATE" }, + { 0, NULL } /* terminate list */ + }; + + unsigned int i; + for (i = 0; cmd_descr[i].text; i++) + if (cmd_descr[i].command == command) + return cmd_descr[i].text; +#endif + + return NULL; +} + /** * ata_eh_link_report - report error handling to user * @link: ATA link EH is going on @@ -2177,6 +2288,7 @@ static void ata_eh_link_report(struct ata_link *link) ata_link_printk(link, KERN_ERR, "%s\n", desc); } +#ifdef CONFIG_ATA_VERBOSE_ERROR if (ehc->i.serror) ata_link_printk(link, KERN_ERR, "SError: { %s%s%s%s%s%s%s%s%s%s%s%s%s%s%s%s%s}\n", @@ -2197,6 +2309,7 @@ static void ata_eh_link_report(struct ata_link *link) ehc->i.serror & SERR_TRANS_ST_ERROR ? "TrStaTrns " : "", ehc->i.serror & SERR_UNRECOG_FIS ? "UnrecFIS " : "", ehc->i.serror & SERR_DEV_XCHG ? "DevExch " : ""); +#endif for (tag = 0; tag < ATA_MAX_QUEUE; tag++) { struct ata_queued_cmd *qc = __ata_qc_from_tag(ap, tag); @@ -2228,14 +2341,23 @@ static void ata_eh_link_report(struct ata_link *link) dma_str[qc->dma_dir]); } - if (ata_is_atapi(qc->tf.protocol)) - snprintf(cdb_buf, sizeof(cdb_buf), + if (ata_is_atapi(qc->tf.protocol)) { + if (qc->scsicmd) + scsi_print_command(qc->scsicmd); + else + snprintf(cdb_buf, sizeof(cdb_buf), "cdb %02x %02x %02x %02x %02x %02x %02x %02x " "%02x %02x %02x %02x %02x %02x %02x %02x\n ", cdb[0], cdb[1], cdb[2], cdb[3], cdb[4], cdb[5], cdb[6], cdb[7], cdb[8], cdb[9], cdb[10], cdb[11], cdb[12], cdb[13], cdb[14], cdb[15]); + } else { + const char *descr = ata_get_cmd_descript(cmd->command); + if (descr) + ata_dev_printk(qc->dev, KERN_ERR, + "failed command: %s\n", descr); + } ata_dev_printk(qc->dev, KERN_ERR, "cmd %02x/%02x:%02x:%02x:%02x:%02x/%02x:%02x:%02x:%02x:%02x/%02x " @@ -2254,6 +2376,7 @@ static void ata_eh_link_report(struct ata_link *link) res->device, qc->err_mask, ata_err_string(qc->err_mask), qc->err_mask & AC_ERR_NCQ ? " " : ""); +#ifdef CONFIG_ATA_VERBOSE_ERROR if (res->command & (ATA_BUSY | ATA_DRDY | ATA_DF | ATA_DRQ | ATA_ERR)) { if (res->command & ATA_BUSY) @@ -2277,6 +2400,7 @@ static void ata_eh_link_report(struct ata_link *link) res->feature & ATA_UNC ? "UNC " : "", res->feature & ATA_IDNF ? "IDNF " : "", res->feature & ATA_ABORTED ? "ABRT " : ""); +#endif } } diff --git a/drivers/ata/libata.h b/drivers/ata/libata.h index 89a1e0018e71..be8e2628f82c 100644 --- a/drivers/ata/libata.h +++ b/drivers/ata/libata.h @@ -164,6 +164,7 @@ extern void ata_eh_about_to_do(struct ata_link *link, struct ata_device *dev, extern void ata_eh_done(struct ata_link *link, struct ata_device *dev, unsigned int action); extern void ata_eh_autopsy(struct ata_port *ap); +const char *ata_get_cmd_descript(u8 command); extern void ata_eh_report(struct ata_port *ap); extern int ata_eh_reset(struct ata_link *link, int classify, ata_prereset_fn_t prereset, ata_reset_fn_t softreset, diff --git a/include/linux/ata.h b/include/linux/ata.h index f5494050df83..6299a259ed19 100644 --- a/include/linux/ata.h +++ b/include/linux/ata.h @@ -210,15 +210,25 @@ enum { ATA_CMD_STANDBY = 0xE2, /* place in standby power mode */ ATA_CMD_IDLE = 0xE3, /* place in idle power mode */ ATA_CMD_EDD = 0x90, /* execute device diagnostic */ + ATA_CMD_DOWNLOAD_MICRO = 0x92, + ATA_CMD_NOP = 0x00, ATA_CMD_FLUSH = 0xE7, ATA_CMD_FLUSH_EXT = 0xEA, ATA_CMD_ID_ATA = 0xEC, ATA_CMD_ID_ATAPI = 0xA1, + ATA_CMD_SERVICE = 0xA2, ATA_CMD_READ = 0xC8, ATA_CMD_READ_EXT = 0x25, + ATA_CMD_READ_QUEUED = 0x26, + ATA_CMD_READ_STREAM_EXT = 0x2B, + ATA_CMD_READ_STREAM_DMA_EXT = 0x2A, ATA_CMD_WRITE = 0xCA, ATA_CMD_WRITE_EXT = 0x35, + ATA_CMD_WRITE_QUEUED = 0x36, + ATA_CMD_WRITE_STREAM_EXT = 0x3B, + ATA_CMD_WRITE_STREAM_DMA_EXT = 0x3A, ATA_CMD_WRITE_FUA_EXT = 0x3D, + ATA_CMD_WRITE_QUEUED_FUA_EXT = 0x3E, ATA_CMD_FPDMA_READ = 0x60, ATA_CMD_FPDMA_WRITE = 0x61, ATA_CMD_PIO_READ = 0x20, @@ -235,6 +245,7 @@ enum { ATA_CMD_PACKET = 0xA0, ATA_CMD_VERIFY = 0x40, ATA_CMD_VERIFY_EXT = 0x42, + ATA_CMD_WRITE_UNCORR_EXT = 0x45, ATA_CMD_STANDBYNOW1 = 0xE0, ATA_CMD_IDLEIMMEDIATE = 0xE1, ATA_CMD_SLEEP = 0xE6, @@ -243,15 +254,34 @@ enum { ATA_CMD_READ_NATIVE_MAX_EXT = 0x27, ATA_CMD_SET_MAX = 0xF9, ATA_CMD_SET_MAX_EXT = 0x37, - ATA_CMD_READ_LOG_EXT = 0x2f, + ATA_CMD_READ_LOG_EXT = 0x2F, + ATA_CMD_WRITE_LOG_EXT = 0x3F, + ATA_CMD_READ_LOG_DMA_EXT = 0x47, + ATA_CMD_WRITE_LOG_DMA_EXT = 0x57, + ATA_CMD_TRUSTED_RCV = 0x5C, + ATA_CMD_TRUSTED_RCV_DMA = 0x5D, + ATA_CMD_TRUSTED_SND = 0x5E, + ATA_CMD_TRUSTED_SND_DMA = 0x5F, ATA_CMD_PMP_READ = 0xE4, ATA_CMD_PMP_WRITE = 0xE8, ATA_CMD_CONF_OVERLAY = 0xB1, + ATA_CMD_SEC_SET_PASS = 0xF1, + ATA_CMD_SEC_UNLOCK = 0xF2, + ATA_CMD_SEC_ERASE_PREP = 0xF3, + ATA_CMD_SEC_ERASE_UNIT = 0xF4, ATA_CMD_SEC_FREEZE_LOCK = 0xF5, + ATA_CMD_SEC_DISABLE_PASS = 0xF6, + ATA_CMD_CONFIG_STREAM = 0x51, ATA_CMD_SMART = 0xB0, ATA_CMD_MEDIA_LOCK = 0xDE, ATA_CMD_MEDIA_UNLOCK = 0xDF, ATA_CMD_DSM = 0x06, + ATA_CMD_CHK_MED_CRD_TYP = 0xD1, + ATA_CMD_CFA_REQ_EXT_ERR = 0x03, + ATA_CMD_CFA_WRITE_NE = 0x38, + ATA_CMD_CFA_TRANS_SECT = 0x87, + ATA_CMD_CFA_ERASE = 0xC0, + ATA_CMD_CFA_WRITE_MULT_NE = 0xCD, /* marked obsolete in the ATA/ATAPI-7 spec */ ATA_CMD_RESTORE = 0x10, -- cgit v1.2.3 From 051d9fbdd1d1ec85ea18ba20581234cf23f1c217 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Fri, 3 Jul 2009 11:46:12 +0900 Subject: libata: remove spindown skipping and warning This was a hack to give userland shutdown tools time to drop manual spindown. All popular distros updated quite some time ago and the due is well passed. Drop it. Signed-off-by: Tejun Heo Cc: Jaswinder Singh Rajput Signed-off-by: Jeff Garzik --- Documentation/feature-removal-schedule.txt | 18 ----------- drivers/ata/libata-scsi.c | 51 ------------------------------ include/linux/libata.h | 1 - 3 files changed, 70 deletions(-) (limited to 'include') diff --git a/Documentation/feature-removal-schedule.txt b/Documentation/feature-removal-schedule.txt index 09e031c55887..b3b62903f08c 100644 --- a/Documentation/feature-removal-schedule.txt +++ b/Documentation/feature-removal-schedule.txt @@ -206,24 +206,6 @@ Who: Len Brown --------------------------- -What: libata spindown skipping and warning -When: Dec 2008 -Why: Some halt(8) implementations synchronize caches for and spin - down libata disks because libata didn't use to spin down disk on - system halt (only synchronized caches). - Spin down on system halt is now implemented. sysfs node - /sys/class/scsi_disk/h:c:i:l/manage_start_stop is present if - spin down support is available. - Because issuing spin down command to an already spun down disk - makes some disks spin up just to spin down again, libata tracks - device spindown status to skip the extra spindown command and - warn about it. - This is to give userspace tools the time to get updated and will - be removed after userspace is reasonably updated. -Who: Tejun Heo - ---------------------------- - What: i386/x86_64 bzImage symlinks When: April 2010 diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c index d0dfeef55db5..de3a0050760a 100644 --- a/drivers/ata/libata-scsi.c +++ b/drivers/ata/libata-scsi.c @@ -1257,23 +1257,6 @@ int ata_scsi_change_queue_depth(struct scsi_device *sdev, int queue_depth) return queue_depth; } -/* XXX: for spindown warning */ -static void ata_delayed_done_timerfn(unsigned long arg) -{ - struct scsi_cmnd *scmd = (void *)arg; - - scmd->scsi_done(scmd); -} - -/* XXX: for spindown warning */ -static void ata_delayed_done(struct scsi_cmnd *scmd) -{ - static struct timer_list timer; - - setup_timer(&timer, ata_delayed_done_timerfn, (unsigned long)scmd); - mod_timer(&timer, jiffies + 5 * HZ); -} - /** * ata_scsi_start_stop_xlat - Translate SCSI START STOP UNIT command * @qc: Storage for translated ATA taskfile @@ -1338,32 +1321,6 @@ static unsigned int ata_scsi_start_stop_xlat(struct ata_queued_cmd *qc) system_entering_hibernation()) goto skip; - /* XXX: This is for backward compatibility, will be - * removed. Read Documentation/feature-removal-schedule.txt - * for more info. - */ - if ((qc->dev->flags & ATA_DFLAG_SPUNDOWN) && - (system_state == SYSTEM_HALT || - system_state == SYSTEM_POWER_OFF)) { - static unsigned long warned; - - if (!test_and_set_bit(0, &warned)) { - ata_dev_printk(qc->dev, KERN_WARNING, - "DISK MIGHT NOT BE SPUN DOWN PROPERLY. " - "UPDATE SHUTDOWN UTILITY\n"); - ata_dev_printk(qc->dev, KERN_WARNING, - "For more info, visit " - "http://linux-ata.org/shutdown.html\n"); - - /* ->scsi_done is not used, use it for - * delayed completion. - */ - scmd->scsi_done = qc->scsidone; - qc->scsidone = ata_delayed_done; - } - goto skip; - } - /* Issue ATA STANDBY IMMEDIATE command */ tf->command = ATA_CMD_STANDBYNOW1; } @@ -1764,14 +1721,6 @@ static void ata_scsi_qc_complete(struct ata_queued_cmd *qc) } } - /* XXX: track spindown state for spindown skipping and warning */ - if (unlikely(qc->tf.command == ATA_CMD_STANDBY || - qc->tf.command == ATA_CMD_STANDBYNOW1)) - qc->dev->flags |= ATA_DFLAG_SPUNDOWN; - else if (likely(system_state != SYSTEM_HALT && - system_state != SYSTEM_POWER_OFF)) - qc->dev->flags &= ~ATA_DFLAG_SPUNDOWN; - if (need_sense && !ap->ops->error_handler) ata_dump_status(ap->print_id, &qc->result_tf); diff --git a/include/linux/libata.h b/include/linux/libata.h index d3f7cab4873e..76319bf03e37 100644 --- a/include/linux/libata.h +++ b/include/linux/libata.h @@ -143,7 +143,6 @@ enum { ATA_DFLAG_PIO = (1 << 12), /* device limited to PIO mode */ ATA_DFLAG_NCQ_OFF = (1 << 13), /* device limited to non-NCQ mode */ - ATA_DFLAG_SPUNDOWN = (1 << 14), /* XXX: for spindown_compat */ ATA_DFLAG_SLEEPING = (1 << 15), /* device is sleeping */ ATA_DFLAG_DUBIOUS_XFER = (1 << 16), /* data transfer not verified */ ATA_DFLAG_NO_UNLOAD = (1 << 17), /* device doesn't support unload */ -- cgit v1.2.3 From 86393e52c3f1e2f6be18383f6ecdbcdc5727d545 Mon Sep 17 00:00:00 2001 From: Alexey Dobriyan Date: Sat, 29 Aug 2009 01:34:49 +0000 Subject: netns: embed ip6_dst_ops directly struct net::ipv6.ip6_dst_ops is separatedly dynamically allocated, but there is no fundamental reason for it. Embed it directly into struct netns_ipv6. For that: * move struct dst_ops into separate header to fix circular dependencies I honestly tried not to, it's pretty impossible to do other way * drop dynamical allocation, allocate together with netns For a change, remove struct dst_ops::dst_net, it's deducible by using container_of() given dst_ops pointer. Signed-off-by: Alexey Dobriyan Signed-off-by: David S. Miller --- include/net/dst.h | 23 +---------------------- include/net/dst_ops.h | 28 ++++++++++++++++++++++++++++ include/net/netns/ipv6.h | 3 ++- net/ipv6/route.c | 34 +++++++++++++--------------------- 4 files changed, 44 insertions(+), 44 deletions(-) create mode 100644 include/net/dst_ops.h (limited to 'include') diff --git a/include/net/dst.h b/include/net/dst.h index 7fc409c19b37..5a900ddcf10d 100644 --- a/include/net/dst.h +++ b/include/net/dst.h @@ -8,6 +8,7 @@ #ifndef _NET_DST_H #define _NET_DST_H +#include #include #include #include @@ -102,28 +103,6 @@ struct dst_entry }; }; - -struct dst_ops -{ - unsigned short family; - __be16 protocol; - unsigned gc_thresh; - - int (*gc)(struct dst_ops *ops); - struct dst_entry * (*check)(struct dst_entry *, __u32 cookie); - void (*destroy)(struct dst_entry *); - void (*ifdown)(struct dst_entry *, - struct net_device *dev, int how); - struct dst_entry * (*negative_advice)(struct dst_entry *); - void (*link_failure)(struct sk_buff *); - void (*update_pmtu)(struct dst_entry *dst, u32 mtu); - int (*local_out)(struct sk_buff *skb); - - atomic_t entries; - struct kmem_cache *kmem_cachep; - struct net *dst_net; -}; - #ifdef __KERNEL__ static inline u32 diff --git a/include/net/dst_ops.h b/include/net/dst_ops.h new file mode 100644 index 000000000000..d1ff9b7e99b8 --- /dev/null +++ b/include/net/dst_ops.h @@ -0,0 +1,28 @@ +#ifndef _NET_DST_OPS_H +#define _NET_DST_OPS_H +#include + +struct dst_entry; +struct kmem_cachep; +struct net_device; +struct sk_buff; + +struct dst_ops { + unsigned short family; + __be16 protocol; + unsigned gc_thresh; + + int (*gc)(struct dst_ops *ops); + struct dst_entry * (*check)(struct dst_entry *, __u32 cookie); + void (*destroy)(struct dst_entry *); + void (*ifdown)(struct dst_entry *, + struct net_device *dev, int how); + struct dst_entry * (*negative_advice)(struct dst_entry *); + void (*link_failure)(struct sk_buff *); + void (*update_pmtu)(struct dst_entry *dst, u32 mtu); + int (*local_out)(struct sk_buff *skb); + + atomic_t entries; + struct kmem_cache *kmem_cachep; +}; +#endif diff --git a/include/net/netns/ipv6.h b/include/net/netns/ipv6.h index afab4e4cbac7..dfeb2d7c425b 100644 --- a/include/net/netns/ipv6.h +++ b/include/net/netns/ipv6.h @@ -6,6 +6,7 @@ #ifndef __NETNS_IPV6_H__ #define __NETNS_IPV6_H__ +#include struct ctl_table_header; @@ -42,7 +43,7 @@ struct netns_ipv6 { struct timer_list ip6_fib_timer; struct hlist_head *fib_table_hash; struct fib6_table *fib6_main_tbl; - struct dst_ops *ip6_dst_ops; + struct dst_ops ip6_dst_ops; unsigned int ip6_rt_gc_expire; unsigned long ip6_rt_last_gc; #ifdef CONFIG_IPV6_MULTIPLE_TABLES diff --git a/net/ipv6/route.c b/net/ipv6/route.c index 1473ee0a1f51..9ccfef345560 100644 --- a/net/ipv6/route.c +++ b/net/ipv6/route.c @@ -665,7 +665,7 @@ static struct rt6_info *rt6_alloc_cow(struct rt6_info *ort, struct in6_addr *dad net->ipv6.sysctl.ip6_rt_gc_elasticity = 1; net->ipv6.sysctl.ip6_rt_gc_min_interval = 0; - ip6_dst_gc(net->ipv6.ip6_dst_ops); + ip6_dst_gc(&net->ipv6.ip6_dst_ops); net->ipv6.sysctl.ip6_rt_gc_elasticity = saved_rt_elasticity; @@ -970,7 +970,7 @@ struct dst_entry *icmp6_dst_alloc(struct net_device *dev, if (unlikely(idev == NULL)) return NULL; - rt = ip6_dst_alloc(net->ipv6.ip6_dst_ops); + rt = ip6_dst_alloc(&net->ipv6.ip6_dst_ops); if (unlikely(rt == NULL)) { in6_dev_put(idev); goto out; @@ -1060,7 +1060,7 @@ static void icmp6_clean_all(int (*func)(struct rt6_info *rt, void *arg), static int ip6_dst_gc(struct dst_ops *ops) { unsigned long now = jiffies; - struct net *net = ops->dst_net; + struct net *net = container_of(ops, struct net, ipv6.ip6_dst_ops); int rt_min_interval = net->ipv6.sysctl.ip6_rt_gc_min_interval; int rt_max_size = net->ipv6.sysctl.ip6_rt_max_size; int rt_elasticity = net->ipv6.sysctl.ip6_rt_gc_elasticity; @@ -1154,7 +1154,7 @@ int ip6_route_add(struct fib6_config *cfg) goto out; } - rt = ip6_dst_alloc(net->ipv6.ip6_dst_ops); + rt = ip6_dst_alloc(&net->ipv6.ip6_dst_ops); if (rt == NULL) { err = -ENOMEM; @@ -1643,7 +1643,7 @@ out: static struct rt6_info * ip6_rt_copy(struct rt6_info *ort) { struct net *net = dev_net(ort->rt6i_dev); - struct rt6_info *rt = ip6_dst_alloc(net->ipv6.ip6_dst_ops); + struct rt6_info *rt = ip6_dst_alloc(&net->ipv6.ip6_dst_ops); if (rt) { rt->u.dst.input = ort->u.dst.input; @@ -1923,7 +1923,7 @@ struct rt6_info *addrconf_dst_alloc(struct inet6_dev *idev, int anycast) { struct net *net = dev_net(idev->dev); - struct rt6_info *rt = ip6_dst_alloc(net->ipv6.ip6_dst_ops); + struct rt6_info *rt = ip6_dst_alloc(&net->ipv6.ip6_dst_ops); struct neighbour *neigh; if (rt == NULL) @@ -2501,7 +2501,7 @@ static int rt6_stats_seq_show(struct seq_file *seq, void *v) net->ipv6.rt6_stats->fib_rt_alloc, net->ipv6.rt6_stats->fib_rt_entries, net->ipv6.rt6_stats->fib_rt_cache, - atomic_read(&net->ipv6.ip6_dst_ops->entries), + atomic_read(&net->ipv6.ip6_dst_ops.entries), net->ipv6.rt6_stats->fib_discarded_routes); return 0; @@ -2637,7 +2637,7 @@ struct ctl_table *ipv6_route_sysctl_init(struct net *net) if (table) { table[0].data = &net->ipv6.sysctl.flush_delay; - table[1].data = &net->ipv6.ip6_dst_ops->gc_thresh; + table[1].data = &net->ipv6.ip6_dst_ops.gc_thresh; table[2].data = &net->ipv6.sysctl.ip6_rt_max_size; table[3].data = &net->ipv6.sysctl.ip6_rt_gc_min_interval; table[4].data = &net->ipv6.sysctl.ip6_rt_gc_timeout; @@ -2655,12 +2655,8 @@ static int ip6_route_net_init(struct net *net) { int ret = -ENOMEM; - net->ipv6.ip6_dst_ops = kmemdup(&ip6_dst_ops_template, - sizeof(*net->ipv6.ip6_dst_ops), - GFP_KERNEL); - if (!net->ipv6.ip6_dst_ops) - goto out; - net->ipv6.ip6_dst_ops->dst_net = hold_net(net); + memcpy(&net->ipv6.ip6_dst_ops, &ip6_dst_ops_template, + sizeof(net->ipv6.ip6_dst_ops)); net->ipv6.ip6_null_entry = kmemdup(&ip6_null_entry_template, sizeof(*net->ipv6.ip6_null_entry), @@ -2669,7 +2665,7 @@ static int ip6_route_net_init(struct net *net) goto out_ip6_dst_ops; net->ipv6.ip6_null_entry->u.dst.path = (struct dst_entry *)net->ipv6.ip6_null_entry; - net->ipv6.ip6_null_entry->u.dst.ops = net->ipv6.ip6_dst_ops; + net->ipv6.ip6_null_entry->u.dst.ops = &net->ipv6.ip6_dst_ops; #ifdef CONFIG_IPV6_MULTIPLE_TABLES net->ipv6.ip6_prohibit_entry = kmemdup(&ip6_prohibit_entry_template, @@ -2679,7 +2675,7 @@ static int ip6_route_net_init(struct net *net) goto out_ip6_null_entry; net->ipv6.ip6_prohibit_entry->u.dst.path = (struct dst_entry *)net->ipv6.ip6_prohibit_entry; - net->ipv6.ip6_prohibit_entry->u.dst.ops = net->ipv6.ip6_dst_ops; + net->ipv6.ip6_prohibit_entry->u.dst.ops = &net->ipv6.ip6_dst_ops; net->ipv6.ip6_blk_hole_entry = kmemdup(&ip6_blk_hole_entry_template, sizeof(*net->ipv6.ip6_blk_hole_entry), @@ -2688,7 +2684,7 @@ static int ip6_route_net_init(struct net *net) goto out_ip6_prohibit_entry; net->ipv6.ip6_blk_hole_entry->u.dst.path = (struct dst_entry *)net->ipv6.ip6_blk_hole_entry; - net->ipv6.ip6_blk_hole_entry->u.dst.ops = net->ipv6.ip6_dst_ops; + net->ipv6.ip6_blk_hole_entry->u.dst.ops = &net->ipv6.ip6_dst_ops; #endif net->ipv6.sysctl.flush_delay = 0; @@ -2717,8 +2713,6 @@ out_ip6_null_entry: kfree(net->ipv6.ip6_null_entry); #endif out_ip6_dst_ops: - release_net(net->ipv6.ip6_dst_ops->dst_net); - kfree(net->ipv6.ip6_dst_ops); goto out; } @@ -2733,8 +2727,6 @@ static void ip6_route_net_exit(struct net *net) kfree(net->ipv6.ip6_prohibit_entry); kfree(net->ipv6.ip6_blk_hole_entry); #endif - release_net(net->ipv6.ip6_dst_ops->dst_net); - kfree(net->ipv6.ip6_dst_ops); } static struct pernet_operations ip6_route_net_ops = { -- cgit v1.2.3 From 5152fc7de3ae31b46692022ea63ce0501280f5b1 Mon Sep 17 00:00:00 2001 From: Damian Lukowski Date: Tue, 1 Sep 2009 10:24:00 +0000 Subject: RTO connection timeout: coding style fixes and comments This patch affects the retransmits_timed_out() function. Changes: 1) Variables have more meaningful names 2) retransmits_timed_out() has an introductionary comment. 3) Small coding style changes. Signed-off-by: Damian Lukowski Signed-off-by: David S. Miller --- include/net/tcp.h | 19 ++++++++++++------- 1 file changed, 12 insertions(+), 7 deletions(-) (limited to 'include') diff --git a/include/net/tcp.h b/include/net/tcp.h index e5319495f15e..df50bc40b5fd 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -1252,22 +1252,27 @@ static inline struct sk_buff *tcp_write_queue_prev(struct sock *sk, struct sk_bu #define tcp_for_write_queue_from_safe(skb, tmp, sk) \ skb_queue_walk_from_safe(&(sk)->sk_write_queue, skb, tmp) +/* This function calculates a "timeout" which is equivalent to the timeout of a + * TCP connection after "boundary" unsucessful, exponentially backed-off + * retransmissions with an initial RTO of TCP_RTO_MIN. + */ static inline bool retransmits_timed_out(const struct sock *sk, unsigned int boundary) { - int limit, K; + unsigned int timeout, linear_backoff_thresh; + if (!inet_csk(sk)->icsk_retransmits) return false; - K = ilog2(TCP_RTO_MAX/TCP_RTO_MIN); + linear_backoff_thresh = ilog2(TCP_RTO_MAX/TCP_RTO_MIN); - if (boundary <= K) - limit = ((2 << boundary) - 1) * TCP_RTO_MIN; + if (boundary <= linear_backoff_thresh) + timeout = ((2 << boundary) - 1) * TCP_RTO_MIN; else - limit = ((2 << K) - 1) * TCP_RTO_MIN + - (boundary - K) * TCP_RTO_MAX; + timeout = ((2 << linear_backoff_thresh) - 1) * TCP_RTO_MIN + + (boundary - linear_backoff_thresh) * TCP_RTO_MAX; - return (tcp_time_stamp - tcp_sk(sk)->retrans_stamp) >= limit; + return (tcp_time_stamp - tcp_sk(sk)->retrans_stamp) >= timeout; } static inline struct sk_buff *tcp_send_head(struct sock *sk) -- cgit v1.2.3 From 89d69d2b75a8f7e258f4b634cd985374cfd3202e Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Tue, 1 Sep 2009 11:13:19 +0000 Subject: net: make neigh_ops constant These tables are never modified at runtime. Move to read-only section. Signed-off-by: Stephen Hemminger Signed-off-by: David S. Miller --- include/net/arp.h | 2 +- include/net/neighbour.h | 2 +- net/atm/clip.c | 2 +- net/decnet/dn_neigh.c | 6 +++--- net/ipv4/arp.c | 8 ++++---- net/ipv6/ndisc.c | 6 +++--- 6 files changed, 13 insertions(+), 13 deletions(-) (limited to 'include') diff --git a/include/net/arp.h b/include/net/arp.h index c236270ec95e..716f43c5c98e 100644 --- a/include/net/arp.h +++ b/include/net/arp.h @@ -26,6 +26,6 @@ extern struct sk_buff *arp_create(int type, int ptype, __be32 dest_ip, const unsigned char *target_hw); extern void arp_xmit(struct sk_buff *skb); -extern struct neigh_ops arp_broken_ops; +extern const struct neigh_ops arp_broken_ops; #endif /* _ARP_H */ diff --git a/include/net/neighbour.h b/include/net/neighbour.h index 18b69b6cecaf..3817fda82a80 100644 --- a/include/net/neighbour.h +++ b/include/net/neighbour.h @@ -118,7 +118,7 @@ struct neighbour int (*output)(struct sk_buff *skb); struct sk_buff_head arp_queue; struct timer_list timer; - struct neigh_ops *ops; + const struct neigh_ops *ops; u8 primary_key[0]; }; diff --git a/net/atm/clip.c b/net/atm/clip.c index 27f6852ce190..64629c354343 100644 --- a/net/atm/clip.c +++ b/net/atm/clip.c @@ -267,7 +267,7 @@ static void clip_neigh_error(struct neighbour *neigh, struct sk_buff *skb) kfree_skb(skb); } -static struct neigh_ops clip_neigh_ops = { +static const struct neigh_ops clip_neigh_ops = { .family = AF_INET, .solicit = clip_neigh_solicit, .error_report = clip_neigh_error, diff --git a/net/decnet/dn_neigh.c b/net/decnet/dn_neigh.c index 923786bd6d01..794b5bf95af1 100644 --- a/net/decnet/dn_neigh.c +++ b/net/decnet/dn_neigh.c @@ -59,7 +59,7 @@ static int dn_phase3_output(struct sk_buff *); /* * For talking to broadcast devices: Ethernet & PPP */ -static struct neigh_ops dn_long_ops = { +static const struct neigh_ops dn_long_ops = { .family = AF_DECnet, .error_report = dn_long_error_report, .output = dn_long_output, @@ -71,7 +71,7 @@ static struct neigh_ops dn_long_ops = { /* * For talking to pointopoint and multidrop devices: DDCMP and X.25 */ -static struct neigh_ops dn_short_ops = { +static const struct neigh_ops dn_short_ops = { .family = AF_DECnet, .error_report = dn_short_error_report, .output = dn_short_output, @@ -83,7 +83,7 @@ static struct neigh_ops dn_short_ops = { /* * For talking to DECnet phase III nodes */ -static struct neigh_ops dn_phase3_ops = { +static const struct neigh_ops dn_phase3_ops = { .family = AF_DECnet, .error_report = dn_short_error_report, /* Can use short version here */ .output = dn_phase3_output, diff --git a/net/ipv4/arp.c b/net/ipv4/arp.c index 090e9991ac2a..4e80f336c0cf 100644 --- a/net/ipv4/arp.c +++ b/net/ipv4/arp.c @@ -130,7 +130,7 @@ static void arp_solicit(struct neighbour *neigh, struct sk_buff *skb); static void arp_error_report(struct neighbour *neigh, struct sk_buff *skb); static void parp_redo(struct sk_buff *skb); -static struct neigh_ops arp_generic_ops = { +static const struct neigh_ops arp_generic_ops = { .family = AF_INET, .solicit = arp_solicit, .error_report = arp_error_report, @@ -140,7 +140,7 @@ static struct neigh_ops arp_generic_ops = { .queue_xmit = dev_queue_xmit, }; -static struct neigh_ops arp_hh_ops = { +static const struct neigh_ops arp_hh_ops = { .family = AF_INET, .solicit = arp_solicit, .error_report = arp_error_report, @@ -150,7 +150,7 @@ static struct neigh_ops arp_hh_ops = { .queue_xmit = dev_queue_xmit, }; -static struct neigh_ops arp_direct_ops = { +static const struct neigh_ops arp_direct_ops = { .family = AF_INET, .output = dev_queue_xmit, .connected_output = dev_queue_xmit, @@ -158,7 +158,7 @@ static struct neigh_ops arp_direct_ops = { .queue_xmit = dev_queue_xmit, }; -struct neigh_ops arp_broken_ops = { +const struct neigh_ops arp_broken_ops = { .family = AF_INET, .solicit = arp_solicit, .error_report = arp_error_report, diff --git a/net/ipv6/ndisc.c b/net/ipv6/ndisc.c index 44b4c87e5cce..7015478797f6 100644 --- a/net/ipv6/ndisc.c +++ b/net/ipv6/ndisc.c @@ -98,7 +98,7 @@ static int pndisc_constructor(struct pneigh_entry *n); static void pndisc_destructor(struct pneigh_entry *n); static void pndisc_redo(struct sk_buff *skb); -static struct neigh_ops ndisc_generic_ops = { +static const struct neigh_ops ndisc_generic_ops = { .family = AF_INET6, .solicit = ndisc_solicit, .error_report = ndisc_error_report, @@ -108,7 +108,7 @@ static struct neigh_ops ndisc_generic_ops = { .queue_xmit = dev_queue_xmit, }; -static struct neigh_ops ndisc_hh_ops = { +static const struct neigh_ops ndisc_hh_ops = { .family = AF_INET6, .solicit = ndisc_solicit, .error_report = ndisc_error_report, @@ -119,7 +119,7 @@ static struct neigh_ops ndisc_hh_ops = { }; -static struct neigh_ops ndisc_direct_ops = { +static const struct neigh_ops ndisc_direct_ops = { .family = AF_INET6, .output = dev_queue_xmit, .connected_output = dev_queue_xmit, -- cgit v1.2.3 From 6d703a81ad5fdd102334751ddacb053ecc6ff046 Mon Sep 17 00:00:00 2001 From: Alexey Dobriyan Date: Tue, 1 Sep 2009 17:52:57 -0700 Subject: ide: convert to ->proc_fops ->read_proc, ->write_proc are going away, ->proc_fops should be used instead. The only tricky place is IDENTIFY handling: if for some reason taskfile_lib_get_identify() fails, buffer _is_ changed and at least first byte is overwritten. Emulate old behaviour with returning that first byte to userspace and reporting length=1 despite overall -E. Signed-off-by: Alexey Dobriyan Signed-off-by: David S. Miller --- drivers/ide/ide-cd.c | 28 +++- drivers/ide/ide-disk_proc.c | 129 +++++++++++------ drivers/ide/ide-floppy_proc.c | 30 ++-- drivers/ide/ide-proc.c | 330 +++++++++++++++++++++++++++--------------- drivers/ide/ide-tape.c | 31 ++-- include/linux/ide.h | 24 +-- 6 files changed, 365 insertions(+), 207 deletions(-) (limited to 'include') diff --git a/drivers/ide/ide-cd.c b/drivers/ide/ide-cd.c index ad0ab0c0a493..b79ca419d8d9 100644 --- a/drivers/ide/ide-cd.c +++ b/drivers/ide/ide-cd.c @@ -30,6 +30,7 @@ #include #include #include +#include #include #include #include @@ -1389,19 +1390,30 @@ static sector_t ide_cdrom_capacity(ide_drive_t *drive) return capacity * sectors_per_frame; } -static int proc_idecd_read_capacity(char *page, char **start, off_t off, - int count, int *eof, void *data) +static int idecd_capacity_proc_show(struct seq_file *m, void *v) { - ide_drive_t *drive = data; - int len; + ide_drive_t *drive = m->private; - len = sprintf(page, "%llu\n", (long long)ide_cdrom_capacity(drive)); - PROC_IDE_READ_RETURN(page, start, off, count, eof, len); + seq_printf(m, "%llu\n", (long long)ide_cdrom_capacity(drive)); + return 0; +} + +static int idecd_capacity_proc_open(struct inode *inode, struct file *file) +{ + return single_open(file, idecd_capacity_proc_show, PDE(inode)->data); } +static const struct file_operations idecd_capacity_proc_fops = { + .owner = THIS_MODULE, + .open = idecd_capacity_proc_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, +}; + static ide_proc_entry_t idecd_proc[] = { - { "capacity", S_IFREG|S_IRUGO, proc_idecd_read_capacity, NULL }, - { NULL, 0, NULL, NULL } + { "capacity", S_IFREG|S_IRUGO, &idecd_capacity_proc_fops }, + {} }; static ide_proc_entry_t *ide_cd_proc_entries(ide_drive_t *drive) diff --git a/drivers/ide/ide-disk_proc.c b/drivers/ide/ide-disk_proc.c index 19f263bf0a9e..60b0590ccc9c 100644 --- a/drivers/ide/ide-disk_proc.c +++ b/drivers/ide/ide-disk_proc.c @@ -1,5 +1,6 @@ #include #include +#include #include "ide-disk.h" @@ -37,77 +38,117 @@ static int get_smart_data(ide_drive_t *drive, u8 *buf, u8 sub_cmd) return ide_raw_taskfile(drive, &cmd, buf, 1); } -static int proc_idedisk_read_cache - (char *page, char **start, off_t off, int count, int *eof, void *data) +static int idedisk_cache_proc_show(struct seq_file *m, void *v) { - ide_drive_t *drive = (ide_drive_t *) data; - char *out = page; - int len; + ide_drive_t *drive = (ide_drive_t *) m->private; if (drive->dev_flags & IDE_DFLAG_ID_READ) - len = sprintf(out, "%i\n", drive->id[ATA_ID_BUF_SIZE] / 2); + seq_printf(m, "%i\n", drive->id[ATA_ID_BUF_SIZE] / 2); else - len = sprintf(out, "(none)\n"); + seq_printf(m, "(none)\n"); + return 0; +} - PROC_IDE_READ_RETURN(page, start, off, count, eof, len); +static int idedisk_cache_proc_open(struct inode *inode, struct file *file) +{ + return single_open(file, idedisk_cache_proc_show, PDE(inode)->data); } -static int proc_idedisk_read_capacity - (char *page, char **start, off_t off, int count, int *eof, void *data) +static const struct file_operations idedisk_cache_proc_fops = { + .owner = THIS_MODULE, + .open = idedisk_cache_proc_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, +}; + +static int idedisk_capacity_proc_show(struct seq_file *m, void *v) { - ide_drive_t*drive = (ide_drive_t *)data; - int len; + ide_drive_t*drive = (ide_drive_t *)m->private; - len = sprintf(page, "%llu\n", (long long)ide_gd_capacity(drive)); + seq_printf(m, "%llu\n", (long long)ide_gd_capacity(drive)); + return 0; +} - PROC_IDE_READ_RETURN(page, start, off, count, eof, len); +static int idedisk_capacity_proc_open(struct inode *inode, struct file *file) +{ + return single_open(file, idedisk_capacity_proc_show, PDE(inode)->data); } -static int proc_idedisk_read_smart(char *page, char **start, off_t off, - int count, int *eof, void *data, u8 sub_cmd) +static const struct file_operations idedisk_capacity_proc_fops = { + .owner = THIS_MODULE, + .open = idedisk_capacity_proc_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, +}; + +static int __idedisk_proc_show(struct seq_file *m, ide_drive_t *drive, u8 sub_cmd) { - ide_drive_t *drive = (ide_drive_t *)data; - int len = 0, i = 0; + u8 *buf; + + buf = kmalloc(SECTOR_SIZE, GFP_KERNEL); + if (!buf) + return -ENOMEM; (void)smart_enable(drive); - if (get_smart_data(drive, page, sub_cmd) == 0) { - unsigned short *val = (unsigned short *) page; - char *out = (char *)val + SECTOR_SIZE; - - page = out; - do { - out += sprintf(out, "%04x%c", le16_to_cpu(*val), - (++i & 7) ? ' ' : '\n'); - val += 1; - } while (i < SECTOR_SIZE / 2); - len = out - page; + if (get_smart_data(drive, buf, sub_cmd) == 0) { + __le16 *val = (__le16 *)buf; + int i; + + for (i = 0; i < SECTOR_SIZE / 2; i++) { + seq_printf(m, "%04x%c", le16_to_cpu(val[i]), + (i % 8) == 7 ? '\n' : ' '); + } } + kfree(buf); + return 0; +} - PROC_IDE_READ_RETURN(page, start, off, count, eof, len); +static int idedisk_sv_proc_show(struct seq_file *m, void *v) +{ + return __idedisk_proc_show(m, m->private, ATA_SMART_READ_VALUES); } -static int proc_idedisk_read_sv - (char *page, char **start, off_t off, int count, int *eof, void *data) +static int idedisk_sv_proc_open(struct inode *inode, struct file *file) { - return proc_idedisk_read_smart(page, start, off, count, eof, data, - ATA_SMART_READ_VALUES); + return single_open(file, idedisk_sv_proc_show, PDE(inode)->data); } -static int proc_idedisk_read_st - (char *page, char **start, off_t off, int count, int *eof, void *data) +static const struct file_operations idedisk_sv_proc_fops = { + .owner = THIS_MODULE, + .open = idedisk_sv_proc_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, +}; + +static int idedisk_st_proc_show(struct seq_file *m, void *v) { - return proc_idedisk_read_smart(page, start, off, count, eof, data, - ATA_SMART_READ_THRESHOLDS); + return __idedisk_proc_show(m, m->private, ATA_SMART_READ_THRESHOLDS); } +static int idedisk_st_proc_open(struct inode *inode, struct file *file) +{ + return single_open(file, idedisk_st_proc_show, PDE(inode)->data); +} + +static const struct file_operations idedisk_st_proc_fops = { + .owner = THIS_MODULE, + .open = idedisk_st_proc_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, +}; + ide_proc_entry_t ide_disk_proc[] = { - { "cache", S_IFREG|S_IRUGO, proc_idedisk_read_cache, NULL }, - { "capacity", S_IFREG|S_IRUGO, proc_idedisk_read_capacity, NULL }, - { "geometry", S_IFREG|S_IRUGO, proc_ide_read_geometry, NULL }, - { "smart_values", S_IFREG|S_IRUSR, proc_idedisk_read_sv, NULL }, - { "smart_thresholds", S_IFREG|S_IRUSR, proc_idedisk_read_st, NULL }, - { NULL, 0, NULL, NULL } + { "cache", S_IFREG|S_IRUGO, &idedisk_cache_proc_fops }, + { "capacity", S_IFREG|S_IRUGO, &idedisk_capacity_proc_fops }, + { "geometry", S_IFREG|S_IRUGO, &ide_geometry_proc_fops }, + { "smart_values", S_IFREG|S_IRUSR, &idedisk_sv_proc_fops }, + { "smart_thresholds", S_IFREG|S_IRUSR, &idedisk_st_proc_fops }, + {} }; ide_devset_rw_field(bios_cyl, bios_cyl); diff --git a/drivers/ide/ide-floppy_proc.c b/drivers/ide/ide-floppy_proc.c index fcd4d8153df5..d711d9b883de 100644 --- a/drivers/ide/ide-floppy_proc.c +++ b/drivers/ide/ide-floppy_proc.c @@ -1,22 +1,34 @@ #include #include +#include #include "ide-floppy.h" -static int proc_idefloppy_read_capacity(char *page, char **start, off_t off, - int count, int *eof, void *data) +static int idefloppy_capacity_proc_show(struct seq_file *m, void *v) { - ide_drive_t*drive = (ide_drive_t *)data; - int len; + ide_drive_t*drive = (ide_drive_t *)m->private; - len = sprintf(page, "%llu\n", (long long)ide_gd_capacity(drive)); - PROC_IDE_READ_RETURN(page, start, off, count, eof, len); + seq_printf(m, "%llu\n", (long long)ide_gd_capacity(drive)); + return 0; } +static int idefloppy_capacity_proc_open(struct inode *inode, struct file *file) +{ + return single_open(file, idefloppy_capacity_proc_show, PDE(inode)->data); +} + +static const struct file_operations idefloppy_capacity_proc_fops = { + .owner = THIS_MODULE, + .open = idefloppy_capacity_proc_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, +}; + ide_proc_entry_t ide_floppy_proc[] = { - { "capacity", S_IFREG|S_IRUGO, proc_idefloppy_read_capacity, NULL }, - { "geometry", S_IFREG|S_IRUGO, proc_ide_read_geometry, NULL }, - { NULL, 0, NULL, NULL } + { "capacity", S_IFREG|S_IRUGO, &idefloppy_capacity_proc_fops }, + { "geometry", S_IFREG|S_IRUGO, &ide_geometry_proc_fops }, + {} }; ide_devset_rw_field(bios_cyl, bios_cyl); diff --git a/drivers/ide/ide-proc.c b/drivers/ide/ide-proc.c index 021de41655e6..28d09a5d8450 100644 --- a/drivers/ide/ide-proc.c +++ b/drivers/ide/ide-proc.c @@ -30,11 +30,9 @@ static struct proc_dir_entry *proc_ide_root; -static int proc_ide_read_imodel - (char *page, char **start, off_t off, int count, int *eof, void *data) +static int ide_imodel_proc_show(struct seq_file *m, void *v) { - ide_hwif_t *hwif = (ide_hwif_t *) data; - int len; + ide_hwif_t *hwif = (ide_hwif_t *) m->private; const char *name; switch (hwif->chipset) { @@ -53,63 +51,108 @@ static int proc_ide_read_imodel case ide_acorn: name = "acorn"; break; default: name = "(unknown)"; break; } - len = sprintf(page, "%s\n", name); - PROC_IDE_READ_RETURN(page, start, off, count, eof, len); + seq_printf(m, "%s\n", name); + return 0; } -static int proc_ide_read_mate - (char *page, char **start, off_t off, int count, int *eof, void *data) +static int ide_imodel_proc_open(struct inode *inode, struct file *file) { - ide_hwif_t *hwif = (ide_hwif_t *) data; - int len; + return single_open(file, ide_imodel_proc_show, PDE(inode)->data); +} + +static const struct file_operations ide_imodel_proc_fops = { + .owner = THIS_MODULE, + .open = ide_imodel_proc_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, +}; + +static int ide_mate_proc_show(struct seq_file *m, void *v) +{ + ide_hwif_t *hwif = (ide_hwif_t *) m->private; if (hwif && hwif->mate) - len = sprintf(page, "%s\n", hwif->mate->name); + seq_printf(m, "%s\n", hwif->mate->name); else - len = sprintf(page, "(none)\n"); - PROC_IDE_READ_RETURN(page, start, off, count, eof, len); + seq_printf(m, "(none)\n"); + return 0; +} + +static int ide_mate_proc_open(struct inode *inode, struct file *file) +{ + return single_open(file, ide_mate_proc_show, PDE(inode)->data); } -static int proc_ide_read_channel - (char *page, char **start, off_t off, int count, int *eof, void *data) +static const struct file_operations ide_mate_proc_fops = { + .owner = THIS_MODULE, + .open = ide_mate_proc_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, +}; + +static int ide_channel_proc_show(struct seq_file *m, void *v) { - ide_hwif_t *hwif = (ide_hwif_t *) data; - int len; + ide_hwif_t *hwif = (ide_hwif_t *) m->private; - page[0] = hwif->channel ? '1' : '0'; - page[1] = '\n'; - len = 2; - PROC_IDE_READ_RETURN(page, start, off, count, eof, len); + seq_printf(m, "%c\n", hwif->channel ? '1' : '0'); + return 0; } -static int proc_ide_read_identify - (char *page, char **start, off_t off, int count, int *eof, void *data) +static int ide_channel_proc_open(struct inode *inode, struct file *file) { - ide_drive_t *drive = (ide_drive_t *)data; - int len = 0, i = 0; - int err = 0; + return single_open(file, ide_channel_proc_show, PDE(inode)->data); +} - len = sprintf(page, "\n"); +static const struct file_operations ide_channel_proc_fops = { + .owner = THIS_MODULE, + .open = ide_channel_proc_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, +}; - if (drive) { - __le16 *val = (__le16 *)page; +static int ide_identify_proc_show(struct seq_file *m, void *v) +{ + ide_drive_t *drive = (ide_drive_t *)m->private; + u8 *buf; - err = taskfile_lib_get_identify(drive, page); - if (!err) { - char *out = (char *)page + SECTOR_SIZE; + if (!drive) { + seq_putc(m, '\n'); + return 0; + } - page = out; - do { - out += sprintf(out, "%04x%c", - le16_to_cpup(val), (++i & 7) ? ' ' : '\n'); - val += 1; - } while (i < SECTOR_SIZE / 2); - len = out - page; + buf = kmalloc(SECTOR_SIZE, GFP_KERNEL); + if (!buf) + return -ENOMEM; + if (taskfile_lib_get_identify(drive, buf) == 0) { + __le16 *val = (__le16 *)buf; + int i; + + for (i = 0; i < SECTOR_SIZE / 2; i++) { + seq_printf(m, "%04x%c", le16_to_cpu(val[i]), + (i % 8) == 7 ? '\n' : ' '); } - } - PROC_IDE_READ_RETURN(page, start, off, count, eof, len); + } else + seq_putc(m, buf[0]); + kfree(buf); + return 0; +} + +static int ide_identify_proc_open(struct inode *inode, struct file *file) +{ + return single_open(file, ide_identify_proc_show, PDE(inode)->data); } +static const struct file_operations ide_identify_proc_fops = { + .owner = THIS_MODULE, + .open = ide_identify_proc_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, +}; + /** * ide_find_setting - find a specific setting * @st: setting table pointer @@ -240,22 +283,20 @@ static void proc_ide_settings_warn(void) warned = 1; } -static int proc_ide_read_settings - (char *page, char **start, off_t off, int count, int *eof, void *data) +static int ide_settings_proc_show(struct seq_file *m, void *v) { const struct ide_proc_devset *setting, *g, *d; const struct ide_devset *ds; - ide_drive_t *drive = (ide_drive_t *) data; - char *out = page; - int len, rc, mul_factor, div_factor; + ide_drive_t *drive = (ide_drive_t *) m->private; + int rc, mul_factor, div_factor; proc_ide_settings_warn(); mutex_lock(&ide_setting_mtx); g = ide_generic_settings; d = drive->settings; - out += sprintf(out, "name\t\t\tvalue\t\tmin\t\tmax\t\tmode\n"); - out += sprintf(out, "----\t\t\t-----\t\t---\t\t---\t\t----\n"); + seq_printf(m, "name\t\t\tvalue\t\tmin\t\tmax\t\tmode\n"); + seq_printf(m, "----\t\t\t-----\t\t---\t\t---\t\t----\n"); while (g->name || (d && d->name)) { /* read settings in the alphabetical order */ if (g->name && d && d->name) { @@ -269,31 +310,35 @@ static int proc_ide_read_settings setting = g++; mul_factor = setting->mulf ? setting->mulf(drive) : 1; div_factor = setting->divf ? setting->divf(drive) : 1; - out += sprintf(out, "%-24s", setting->name); + seq_printf(m, "%-24s", setting->name); rc = ide_read_setting(drive, setting); if (rc >= 0) - out += sprintf(out, "%-16d", rc * mul_factor / div_factor); + seq_printf(m, "%-16d", rc * mul_factor / div_factor); else - out += sprintf(out, "%-16s", "write-only"); - out += sprintf(out, "%-16d%-16d", (setting->min * mul_factor + div_factor - 1) / div_factor, setting->max * mul_factor / div_factor); + seq_printf(m, "%-16s", "write-only"); + seq_printf(m, "%-16d%-16d", (setting->min * mul_factor + div_factor - 1) / div_factor, setting->max * mul_factor / div_factor); ds = setting->setting; if (ds->get) - out += sprintf(out, "r"); + seq_printf(m, "r"); if (ds->set) - out += sprintf(out, "w"); - out += sprintf(out, "\n"); + seq_printf(m, "w"); + seq_printf(m, "\n"); } - len = out - page; mutex_unlock(&ide_setting_mtx); - PROC_IDE_READ_RETURN(page, start, off, count, eof, len); + return 0; +} + +static int ide_settings_proc_open(struct inode *inode, struct file *file) +{ + return single_open(file, ide_settings_proc_show, PDE(inode)->data); } #define MAX_LEN 30 -static int proc_ide_write_settings(struct file *file, const char __user *buffer, - unsigned long count, void *data) +static ssize_t ide_settings_proc_write(struct file *file, const char __user *buffer, + size_t count, loff_t *pos) { - ide_drive_t *drive = (ide_drive_t *) data; + ide_drive_t *drive = (ide_drive_t *) PDE(file->f_path.dentry->d_inode)->data; char name[MAX_LEN + 1]; int for_real = 0, mul_factor, div_factor; unsigned long n; @@ -388,63 +433,104 @@ static int proc_ide_write_settings(struct file *file, const char __user *buffer, return count; parse_error: free_page((unsigned long)buf); - printk("proc_ide_write_settings(): parse error\n"); + printk("%s(): parse error\n", __func__); return -EINVAL; } -int proc_ide_read_capacity - (char *page, char **start, off_t off, int count, int *eof, void *data) +static const struct file_operations ide_settings_proc_fops = { + .owner = THIS_MODULE, + .open = ide_settings_proc_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, + .write = ide_settings_proc_write, +}; + +static int ide_capacity_proc_show(struct seq_file *m, void *v) { - int len = sprintf(page, "%llu\n", (long long)0x7fffffff); - PROC_IDE_READ_RETURN(page, start, off, count, eof, len); + seq_printf(m, "%llu\n", (long long)0x7fffffff); + return 0; } -EXPORT_SYMBOL_GPL(proc_ide_read_capacity); +static int ide_capacity_proc_open(struct inode *inode, struct file *file) +{ + return single_open(file, ide_capacity_proc_show, NULL); +} -int proc_ide_read_geometry - (char *page, char **start, off_t off, int count, int *eof, void *data) +const struct file_operations ide_capacity_proc_fops = { + .owner = THIS_MODULE, + .open = ide_capacity_proc_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, +}; +EXPORT_SYMBOL_GPL(ide_capacity_proc_fops); + +static int ide_geometry_proc_show(struct seq_file *m, void *v) { - ide_drive_t *drive = (ide_drive_t *) data; - char *out = page; - int len; + ide_drive_t *drive = (ide_drive_t *) m->private; - out += sprintf(out, "physical %d/%d/%d\n", + seq_printf(m, "physical %d/%d/%d\n", drive->cyl, drive->head, drive->sect); - out += sprintf(out, "logical %d/%d/%d\n", + seq_printf(m, "logical %d/%d/%d\n", drive->bios_cyl, drive->bios_head, drive->bios_sect); + return 0; +} - len = out - page; - PROC_IDE_READ_RETURN(page, start, off, count, eof, len); +static int ide_geometry_proc_open(struct inode *inode, struct file *file) +{ + return single_open(file, ide_geometry_proc_show, PDE(inode)->data); } -EXPORT_SYMBOL(proc_ide_read_geometry); +const struct file_operations ide_geometry_proc_fops = { + .owner = THIS_MODULE, + .open = ide_geometry_proc_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, +}; +EXPORT_SYMBOL(ide_geometry_proc_fops); -static int proc_ide_read_dmodel - (char *page, char **start, off_t off, int count, int *eof, void *data) +static int ide_dmodel_proc_show(struct seq_file *seq, void *v) { - ide_drive_t *drive = (ide_drive_t *) data; + ide_drive_t *drive = (ide_drive_t *) seq->private; char *m = (char *)&drive->id[ATA_ID_PROD]; - int len; - len = sprintf(page, "%.40s\n", m[0] ? m : "(none)"); - PROC_IDE_READ_RETURN(page, start, off, count, eof, len); + seq_printf(seq, "%.40s\n", m[0] ? m : "(none)"); + return 0; +} + +static int ide_dmodel_proc_open(struct inode *inode, struct file *file) +{ + return single_open(file, ide_dmodel_proc_show, PDE(inode)->data); } -static int proc_ide_read_driver - (char *page, char **start, off_t off, int count, int *eof, void *data) +static const struct file_operations ide_dmodel_proc_fops = { + .owner = THIS_MODULE, + .open = ide_dmodel_proc_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, +}; + +static int ide_driver_proc_show(struct seq_file *m, void *v) { - ide_drive_t *drive = (ide_drive_t *)data; + ide_drive_t *drive = (ide_drive_t *)m->private; struct device *dev = &drive->gendev; struct ide_driver *ide_drv; - int len; if (dev->driver) { ide_drv = to_ide_driver(dev->driver); - len = sprintf(page, "%s version %s\n", + seq_printf(m, "%s version %s\n", dev->driver->name, ide_drv->version); } else - len = sprintf(page, "ide-default version 0.9.newide\n"); - PROC_IDE_READ_RETURN(page, start, off, count, eof, len); + seq_printf(m, "ide-default version 0.9.newide\n"); + return 0; +} + +static int ide_driver_proc_open(struct inode *inode, struct file *file) +{ + return single_open(file, ide_driver_proc_show, PDE(inode)->data); } static int ide_replace_subdriver(ide_drive_t *drive, const char *driver) @@ -474,10 +560,10 @@ static int ide_replace_subdriver(ide_drive_t *drive, const char *driver) return ret; } -static int proc_ide_write_driver - (struct file *file, const char __user *buffer, unsigned long count, void *data) +static ssize_t ide_driver_proc_write(struct file *file, const char __user *buffer, + size_t count, loff_t *pos) { - ide_drive_t *drive = (ide_drive_t *) data; + ide_drive_t *drive = (ide_drive_t *) PDE(file->f_path.dentry->d_inode)->data; char name[32]; if (!capable(CAP_SYS_ADMIN)) @@ -492,12 +578,19 @@ static int proc_ide_write_driver return count; } -static int proc_ide_read_media - (char *page, char **start, off_t off, int count, int *eof, void *data) +static const struct file_operations ide_driver_proc_fops = { + .owner = THIS_MODULE, + .open = ide_driver_proc_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, + .write = ide_driver_proc_write, +}; + +static int ide_media_proc_show(struct seq_file *m, void *v) { - ide_drive_t *drive = (ide_drive_t *) data; + ide_drive_t *drive = (ide_drive_t *) m->private; const char *media; - int len; switch (drive->media) { case ide_disk: media = "disk\n"; break; @@ -507,20 +600,30 @@ static int proc_ide_read_media case ide_optical: media = "optical\n"; break; default: media = "UNKNOWN\n"; break; } - strcpy(page, media); - len = strlen(media); - PROC_IDE_READ_RETURN(page, start, off, count, eof, len); + seq_puts(m, media); + return 0; +} + +static int ide_media_proc_open(struct inode *inode, struct file *file) +{ + return single_open(file, ide_media_proc_show, PDE(inode)->data); } +static const struct file_operations ide_media_proc_fops = { + .owner = THIS_MODULE, + .open = ide_media_proc_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, +}; + static ide_proc_entry_t generic_drive_entries[] = { - { "driver", S_IFREG|S_IRUGO, proc_ide_read_driver, - proc_ide_write_driver }, - { "identify", S_IFREG|S_IRUSR, proc_ide_read_identify, NULL }, - { "media", S_IFREG|S_IRUGO, proc_ide_read_media, NULL }, - { "model", S_IFREG|S_IRUGO, proc_ide_read_dmodel, NULL }, - { "settings", S_IFREG|S_IRUSR|S_IWUSR, proc_ide_read_settings, - proc_ide_write_settings }, - { NULL, 0, NULL, NULL } + { "driver", S_IFREG|S_IRUGO, &ide_driver_proc_fops }, + { "identify", S_IFREG|S_IRUSR, &ide_identify_proc_fops}, + { "media", S_IFREG|S_IRUGO, &ide_media_proc_fops }, + { "model", S_IFREG|S_IRUGO, &ide_dmodel_proc_fops }, + { "settings", S_IFREG|S_IRUSR|S_IWUSR, &ide_settings_proc_fops}, + {} }; static void ide_add_proc_entries(struct proc_dir_entry *dir, ide_proc_entry_t *p, void *data) @@ -530,11 +633,8 @@ static void ide_add_proc_entries(struct proc_dir_entry *dir, ide_proc_entry_t *p if (!dir || !p) return; while (p->name != NULL) { - ent = create_proc_entry(p->name, p->mode, dir); + ent = proc_create_data(p->name, p->mode, dir, p->proc_fops, data); if (!ent) return; - ent->data = data; - ent->read_proc = p->read_proc; - ent->write_proc = p->write_proc; p++; } } @@ -617,10 +717,10 @@ void ide_proc_unregister_device(ide_drive_t *drive) } static ide_proc_entry_t hwif_entries[] = { - { "channel", S_IFREG|S_IRUGO, proc_ide_read_channel, NULL }, - { "mate", S_IFREG|S_IRUGO, proc_ide_read_mate, NULL }, - { "model", S_IFREG|S_IRUGO, proc_ide_read_imodel, NULL }, - { NULL, 0, NULL, NULL } + { "channel", S_IFREG|S_IRUGO, &ide_channel_proc_fops }, + { "mate", S_IFREG|S_IRUGO, &ide_mate_proc_fops }, + { "model", S_IFREG|S_IRUGO, &ide_imodel_proc_fops }, + {} }; void ide_proc_register_port(ide_hwif_t *hwif) diff --git a/drivers/ide/ide-tape.c b/drivers/ide/ide-tape.c index 7b2032bc357b..9d6f62baac27 100644 --- a/drivers/ide/ide-tape.c +++ b/drivers/ide/ide-tape.c @@ -31,6 +31,7 @@ #include #include #include +#include #include #include #include @@ -1816,22 +1817,32 @@ static void ide_tape_release(struct device *dev) } #ifdef CONFIG_IDE_PROC_FS -static int proc_idetape_read_name - (char *page, char **start, off_t off, int count, int *eof, void *data) +static int idetape_name_proc_show(struct seq_file *m, void *v) { - ide_drive_t *drive = (ide_drive_t *) data; + ide_drive_t *drive = (ide_drive_t *) m->private; idetape_tape_t *tape = drive->driver_data; - char *out = page; - int len; - len = sprintf(out, "%s\n", tape->name); - PROC_IDE_READ_RETURN(page, start, off, count, eof, len); + seq_printf(m, "%s\n", tape->name); + return 0; +} + +static int idetape_name_proc_open(struct inode *inode, struct file *file) +{ + return single_open(file, idetape_name_proc_show, PDE(inode)->data); } +static const struct file_operations idetape_name_proc_fops = { + .owner = THIS_MODULE, + .open = idetape_name_proc_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, +}; + static ide_proc_entry_t idetape_proc[] = { - { "capacity", S_IFREG|S_IRUGO, proc_ide_read_capacity, NULL }, - { "name", S_IFREG|S_IRUGO, proc_idetape_read_name, NULL }, - { NULL, 0, NULL, NULL } + { "capacity", S_IFREG|S_IRUGO, &ide_capacity_proc_fops }, + { "name", S_IFREG|S_IRUGO, &idetape_name_proc_fops }, + {} }; static ide_proc_entry_t *ide_tape_proc_entries(ide_drive_t *drive) diff --git a/include/linux/ide.h b/include/linux/ide.h index 803c1ae31237..e4135d6e0556 100644 --- a/include/linux/ide.h +++ b/include/linux/ide.h @@ -919,8 +919,7 @@ __IDE_PROC_DEVSET(_name, _min, _max, NULL, NULL) typedef struct { const char *name; mode_t mode; - read_proc_t *read_proc; - write_proc_t *write_proc; + const struct file_operations *proc_fops; } ide_proc_entry_t; void proc_ide_create(void); @@ -932,24 +931,8 @@ void ide_proc_unregister_port(ide_hwif_t *); void ide_proc_register_driver(ide_drive_t *, struct ide_driver *); void ide_proc_unregister_driver(ide_drive_t *, struct ide_driver *); -read_proc_t proc_ide_read_capacity; -read_proc_t proc_ide_read_geometry; - -/* - * Standard exit stuff: - */ -#define PROC_IDE_READ_RETURN(page,start,off,count,eof,len) \ -{ \ - len -= off; \ - if (len < count) { \ - *eof = 1; \ - if (len <= 0) \ - return 0; \ - } else \ - len = count; \ - *start = page + off; \ - return len; \ -} +extern const struct file_operations ide_capacity_proc_fops; +extern const struct file_operations ide_geometry_proc_fops; #else static inline void proc_ide_create(void) { ; } static inline void proc_ide_destroy(void) { ; } @@ -961,7 +944,6 @@ static inline void ide_proc_register_driver(ide_drive_t *drive, struct ide_driver *driver) { ; } static inline void ide_proc_unregister_driver(ide_drive_t *drive, struct ide_driver *driver) { ; } -#define PROC_IDE_READ_RETURN(page,start,off,count,eof,len) return 0; #endif enum { -- cgit v1.2.3 From 2fbd3da3877ad8d923b055e5996f80b4d4a6daf4 Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Tue, 1 Sep 2009 17:59:25 -0700 Subject: pkt_sched: Revert tasklet_hrtimer changes. These are full of unresolved problems, mainly that conversions don't work 1-1 from hrtimers to tasklet_hrtimers because unlike hrtimers tasklets can't be killed from softirq context. And when a qdisc gets reset, that's exactly what we need to do here. We'll work this out in the net-next-2.6 tree and if warranted we'll backport that work to -stable. This reverts the following 3 changesets: a2cb6a4dd470d7a64255a10b843b0d188416b78f ("pkt_sched: Fix bogon in tasklet_hrtimer changes.") 38acce2d7983632100a9ff3fd20295f6e34074a8 ("pkt_sched: Convert CBQ to tasklet_hrtimer.") ee5f9757ea17759e1ce5503bdae2b07e48e32af9 ("pkt_sched: Convert qdisc_watchdog to tasklet_hrtimer") Signed-off-by: David S. Miller --- include/net/pkt_sched.h | 4 ++-- net/sched/sch_api.c | 10 +++++----- net/sched/sch_cbq.c | 25 +++++++++++-------------- 3 files changed, 18 insertions(+), 21 deletions(-) (limited to 'include') diff --git a/include/net/pkt_sched.h b/include/net/pkt_sched.h index 7eafb8d54470..82a3191375f5 100644 --- a/include/net/pkt_sched.h +++ b/include/net/pkt_sched.h @@ -61,8 +61,8 @@ psched_tdiff_bounded(psched_time_t tv1, psched_time_t tv2, psched_time_t bound) } struct qdisc_watchdog { - struct tasklet_hrtimer timer; - struct Qdisc *qdisc; + struct hrtimer timer; + struct Qdisc *qdisc; }; extern void qdisc_watchdog_init(struct qdisc_watchdog *wd, struct Qdisc *qdisc); diff --git a/net/sched/sch_api.c b/net/sched/sch_api.c index 92e6f3a52c13..24d17ce9c294 100644 --- a/net/sched/sch_api.c +++ b/net/sched/sch_api.c @@ -458,7 +458,7 @@ EXPORT_SYMBOL(qdisc_warn_nonwc); static enum hrtimer_restart qdisc_watchdog(struct hrtimer *timer) { struct qdisc_watchdog *wd = container_of(timer, struct qdisc_watchdog, - timer.timer); + timer); wd->qdisc->flags &= ~TCQ_F_THROTTLED; __netif_schedule(qdisc_root(wd->qdisc)); @@ -468,8 +468,8 @@ static enum hrtimer_restart qdisc_watchdog(struct hrtimer *timer) void qdisc_watchdog_init(struct qdisc_watchdog *wd, struct Qdisc *qdisc) { - tasklet_hrtimer_init(&wd->timer, qdisc_watchdog, - CLOCK_MONOTONIC, HRTIMER_MODE_ABS); + hrtimer_init(&wd->timer, CLOCK_MONOTONIC, HRTIMER_MODE_ABS); + wd->timer.function = qdisc_watchdog; wd->qdisc = qdisc; } EXPORT_SYMBOL(qdisc_watchdog_init); @@ -485,13 +485,13 @@ void qdisc_watchdog_schedule(struct qdisc_watchdog *wd, psched_time_t expires) wd->qdisc->flags |= TCQ_F_THROTTLED; time = ktime_set(0, 0); time = ktime_add_ns(time, PSCHED_TICKS2NS(expires)); - tasklet_hrtimer_start(&wd->timer, time, HRTIMER_MODE_ABS); + hrtimer_start(&wd->timer, time, HRTIMER_MODE_ABS); } EXPORT_SYMBOL(qdisc_watchdog_schedule); void qdisc_watchdog_cancel(struct qdisc_watchdog *wd) { - tasklet_hrtimer_cancel(&wd->timer); + hrtimer_cancel(&wd->timer); wd->qdisc->flags &= ~TCQ_F_THROTTLED; } EXPORT_SYMBOL(qdisc_watchdog_cancel); diff --git a/net/sched/sch_cbq.c b/net/sched/sch_cbq.c index 149b0405c5ec..d5798e17a832 100644 --- a/net/sched/sch_cbq.c +++ b/net/sched/sch_cbq.c @@ -163,7 +163,7 @@ struct cbq_sched_data psched_time_t now_rt; /* Cached real time */ unsigned pmask; - struct tasklet_hrtimer delay_timer; + struct hrtimer delay_timer; struct qdisc_watchdog watchdog; /* Watchdog timer, started when CBQ has backlog, but cannot @@ -503,8 +503,6 @@ static void cbq_ovl_delay(struct cbq_class *cl) cl->undertime = q->now + delay; if (delay > 0) { - struct hrtimer *ht; - sched += delay + cl->penalty; cl->penalized = sched; cl->cpriority = TC_CBQ_MAXPRIO; @@ -512,12 +510,12 @@ static void cbq_ovl_delay(struct cbq_class *cl) expires = ktime_set(0, 0); expires = ktime_add_ns(expires, PSCHED_TICKS2NS(sched)); - ht = &q->delay_timer.timer; - if (hrtimer_try_to_cancel(ht) && - ktime_to_ns(ktime_sub(hrtimer_get_expires(ht), - expires)) > 0) - hrtimer_set_expires(ht, expires); - hrtimer_restart(ht); + if (hrtimer_try_to_cancel(&q->delay_timer) && + ktime_to_ns(ktime_sub( + hrtimer_get_expires(&q->delay_timer), + expires)) > 0) + hrtimer_set_expires(&q->delay_timer, expires); + hrtimer_restart(&q->delay_timer); cl->delayed = 1; cl->xstats.overactions++; return; @@ -593,7 +591,7 @@ static psched_tdiff_t cbq_undelay_prio(struct cbq_sched_data *q, int prio, static enum hrtimer_restart cbq_undelay(struct hrtimer *timer) { struct cbq_sched_data *q = container_of(timer, struct cbq_sched_data, - delay_timer.timer); + delay_timer); struct Qdisc *sch = q->watchdog.qdisc; psched_time_t now; psched_tdiff_t delay = 0; @@ -623,7 +621,7 @@ static enum hrtimer_restart cbq_undelay(struct hrtimer *timer) time = ktime_set(0, 0); time = ktime_add_ns(time, PSCHED_TICKS2NS(now + delay)); - tasklet_hrtimer_start(&q->delay_timer, time, HRTIMER_MODE_ABS); + hrtimer_start(&q->delay_timer, time, HRTIMER_MODE_ABS); } sch->flags &= ~TCQ_F_THROTTLED; @@ -1216,7 +1214,7 @@ cbq_reset(struct Qdisc* sch) q->tx_class = NULL; q->tx_borrowed = NULL; qdisc_watchdog_cancel(&q->watchdog); - tasklet_hrtimer_cancel(&q->delay_timer); + hrtimer_cancel(&q->delay_timer); q->toplevel = TC_CBQ_MAXLEVEL; q->now = psched_get_time(); q->now_rt = q->now; @@ -1399,8 +1397,7 @@ static int cbq_init(struct Qdisc *sch, struct nlattr *opt) q->link.minidle = -0x7FFFFFFF; qdisc_watchdog_init(&q->watchdog, sch); - tasklet_hrtimer_init(&q->delay_timer, cbq_undelay, - CLOCK_MONOTONIC, HRTIMER_MODE_ABS); + hrtimer_init(&q->delay_timer, CLOCK_MONOTONIC, HRTIMER_MODE_ABS); q->delay_timer.function = cbq_undelay; q->toplevel = TC_CBQ_MAXLEVEL; q->now = psched_get_time(); -- cgit v1.2.3 From 69575d388603365f2afbf4166df93152df59b165 Mon Sep 17 00:00:00 2001 From: Shane Wang Date: Tue, 1 Sep 2009 18:25:07 -0700 Subject: x86, intel_txt: clean up the impact on generic code, unbreak non-x86 Move tboot.h from asm to linux to fix the build errors of intel_txt patch on non-X86 platforms. Remove the tboot code from generic code init/main.c and kernel/cpu.c. Signed-off-by: Shane Wang Signed-off-by: H. Peter Anvin --- arch/x86/Kconfig | 4 + arch/x86/include/asm/tboot.h | 197 ------------------------------------------ arch/x86/kernel/reboot.c | 3 +- arch/x86/kernel/setup.c | 3 +- arch/x86/kernel/smpboot.c | 2 +- arch/x86/kernel/tboot.c | 58 ++++++++++--- drivers/acpi/acpica/hwsleep.c | 2 +- drivers/pci/dmar.c | 2 +- drivers/pci/intel-iommu.c | 2 +- include/linux/tboot.h | 162 ++++++++++++++++++++++++++++++++++ init/main.c | 3 - kernel/cpu.c | 6 +- security/Kconfig | 2 +- 13 files changed, 221 insertions(+), 225 deletions(-) delete mode 100644 arch/x86/include/asm/tboot.h create mode 100644 include/linux/tboot.h (limited to 'include') diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index 738bdc6b0f8b..b66f2102c35d 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -178,6 +178,10 @@ config ARCH_SUPPORTS_OPTIMIZED_INLINING config ARCH_SUPPORTS_DEBUG_PAGEALLOC def_bool y +config HAVE_INTEL_TXT + def_bool y + depends on EXPERIMENTAL && DMAR && ACPI + # Use the generic interrupt handling code in kernel/irq/: config GENERIC_HARDIRQS bool diff --git a/arch/x86/include/asm/tboot.h b/arch/x86/include/asm/tboot.h deleted file mode 100644 index b13929d4e5f4..000000000000 --- a/arch/x86/include/asm/tboot.h +++ /dev/null @@ -1,197 +0,0 @@ -/* - * tboot.h: shared data structure with tboot and kernel and functions - * used by kernel for runtime support of Intel(R) Trusted - * Execution Technology - * - * Copyright (c) 2006-2009, Intel Corporation - * - * This program is free software; you can redistribute it and/or modify it - * under the terms and conditions of the GNU General Public License, - * version 2, as published by the Free Software Foundation. - * - * This program is distributed in the hope it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for - * more details. - * - * You should have received a copy of the GNU General Public License along with - * this program; if not, write to the Free Software Foundation, Inc., - * 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA. - * - */ - -#ifndef _ASM_TBOOT_H -#define _ASM_TBOOT_H - -#include - -/* these must have the values from 0-5 in this order */ -enum { - TB_SHUTDOWN_REBOOT = 0, - TB_SHUTDOWN_S5, - TB_SHUTDOWN_S4, - TB_SHUTDOWN_S3, - TB_SHUTDOWN_HALT, - TB_SHUTDOWN_WFS -}; - -#ifdef CONFIG_INTEL_TXT - -/* used to communicate between tboot and the launched kernel */ - -#define TB_KEY_SIZE 64 /* 512 bits */ - -#define MAX_TB_MAC_REGIONS 32 - -struct tboot_mac_region { - u64 start; /* must be 64 byte -aligned */ - u32 size; /* must be 64 byte -granular */ -} __packed; - -/* GAS - Generic Address Structure (ACPI 2.0+) */ -struct tboot_acpi_generic_address { - u8 space_id; - u8 bit_width; - u8 bit_offset; - u8 access_width; - u64 address; -} __packed; - -/* - * combines Sx info from FADT and FACS tables per ACPI 2.0+ spec - * (http://www.acpi.info/) - */ -struct tboot_acpi_sleep_info { - struct tboot_acpi_generic_address pm1a_cnt_blk; - struct tboot_acpi_generic_address pm1b_cnt_blk; - struct tboot_acpi_generic_address pm1a_evt_blk; - struct tboot_acpi_generic_address pm1b_evt_blk; - u16 pm1a_cnt_val; - u16 pm1b_cnt_val; - u64 wakeup_vector; - u32 vector_width; - u64 kernel_s3_resume_vector; -} __packed; - -/* - * shared memory page used for communication between tboot and kernel - */ -struct tboot { - /* - * version 3+ fields: - */ - - /* TBOOT_UUID */ - u8 uuid[16]; - - /* version number: 5 is current */ - u32 version; - - /* physical addr of tb_log_t log */ - u32 log_addr; - - /* - * physical addr of entry point for tboot shutdown and - * type of shutdown (TB_SHUTDOWN_*) being requested - */ - u32 shutdown_entry; - u32 shutdown_type; - - /* kernel-specified ACPI info for Sx shutdown */ - struct tboot_acpi_sleep_info acpi_sinfo; - - /* tboot location in memory (physical) */ - u32 tboot_base; - u32 tboot_size; - - /* memory regions (phys addrs) for tboot to MAC on S3 */ - u8 num_mac_regions; - struct tboot_mac_region mac_regions[MAX_TB_MAC_REGIONS]; - - - /* - * version 4+ fields: - */ - - /* symmetric key for use by kernel; will be encrypted on S3 */ - u8 s3_key[TB_KEY_SIZE]; - - - /* - * version 5+ fields: - */ - - /* used to 4byte-align num_in_wfs */ - u8 reserved_align[3]; - - /* number of processors in wait-for-SIPI */ - u32 num_in_wfs; -} __packed; - -/* - * UUID for tboot data struct to facilitate matching - * defined as {663C8DFF-E8B3-4b82-AABF-19EA4D057A08} by tboot, which is - * represented as {} in the char array used here - */ -#define TBOOT_UUID {0xff, 0x8d, 0x3c, 0x66, 0xb3, 0xe8, 0x82, 0x4b, 0xbf,\ - 0xaa, 0x19, 0xea, 0x4d, 0x5, 0x7a, 0x8} - -extern struct tboot *tboot; - -static inline int tboot_enabled(void) -{ - return tboot != NULL; -} - -extern void tboot_probe(void); -extern void tboot_create_trampoline(void); -extern void tboot_shutdown(u32 shutdown_type); -extern void tboot_sleep(u8 sleep_state, u32 pm1a_control, u32 pm1b_control); -extern int tboot_wait_for_aps(int num_aps); -extern struct acpi_table_header *tboot_get_dmar_table( - struct acpi_table_header *dmar_tbl); -extern int tboot_force_iommu(void); - -#else /* CONFIG_INTEL_TXT */ - -static inline int tboot_enabled(void) -{ - return 0; -} - -static inline void tboot_probe(void) -{ -} - -static inline void tboot_create_trampoline(void) -{ -} - -static inline void tboot_shutdown(u32 shutdown_type) -{ -} - -static inline void tboot_sleep(u8 sleep_state, u32 pm1a_control, - u32 pm1b_control) -{ -} - -static inline int tboot_wait_for_aps(int num_aps) -{ - return 0; -} - -static inline struct acpi_table_header *tboot_get_dmar_table( - struct acpi_table_header *dmar_tbl) -{ - return dmar_tbl; -} - -static inline int tboot_force_iommu(void) -{ - return 0; -} - -#endif /* !CONFIG_INTEL_TXT */ - -#endif /* _ASM_TBOOT_H */ diff --git a/arch/x86/kernel/reboot.c b/arch/x86/kernel/reboot.c index 9de01c5d9794..18ce5c04242a 100644 --- a/arch/x86/kernel/reboot.c +++ b/arch/x86/kernel/reboot.c @@ -3,6 +3,7 @@ #include #include #include +#include #include #include #include @@ -24,8 +25,6 @@ # include #endif -#include - /* * Power off function, if any */ diff --git a/arch/x86/kernel/setup.c b/arch/x86/kernel/setup.c index 80d6e9e32483..6ce0d6f38f7f 100644 --- a/arch/x86/kernel/setup.c +++ b/arch/x86/kernel/setup.c @@ -66,6 +66,7 @@ #include #include +#include #include